UBC Theses and Dissertations

UBC Theses Logo

UBC Theses and Dissertations

Unbreak my heart : blame and forgiveness in personal relationships Chan, Yu Shing Irwin 2019

Your browser doesn't seem to have a PDF viewer, please download the PDF to view this item.

Notice for Google Chrome users:
If you are having trouble viewing or searching the PDF with Google Chrome, please download it here instead.

Item Metadata

Download

Media
24-ubc_2020_may_chan_yushingirwin.pdf [ 661.52kB ]
Metadata
JSON: 24-1.0387292.json
JSON-LD: 24-1.0387292-ld.json
RDF/XML (Pretty): 24-1.0387292-rdf.xml
RDF/JSON: 24-1.0387292-rdf.json
Turtle: 24-1.0387292-turtle.txt
N-Triples: 24-1.0387292-rdf-ntriples.txt
Original Record: 24-1.0387292-source.json
Full Text
24-1.0387292-fulltext.txt
Citation
24-1.0387292.ris

Full Text

UNBREAK MY HEART: BLAME AND FORGIVENESS IN PERSONAL RELATIONSHIPS by  Irwin Yu Shing Chan  B.A., The University of Hong Kong, 2009 M.A., University College London, 2010  A THESIS SUBMITTED IN PARTIAL FULFILLMENT OF THE REQUIREMENTS FOR THE DEGREE OF  Doctor of Philosophy in THE FACULTY OF GRADUATE AND POSTDOCTORAL STUDIES (Philosophy)  THE UNIVERSITY OF BRITISH COLUMBIA (Vancouver)  December  2019  © Irwin Yu Shing Chan, 2019   	 ii	The following individuals certify that they have read, and recommend to the Faculty of Graduate and Postdoctoral Studies for acceptance, the dissertation entitled: Unbreak my Heart: Blame and Forgiveness in Personal Relationships  submitted by Irwin Yu Shing Chan in partial fulfillment of the requirements for the degree of Doctor of Philosophy in Philosophy  Examining Committee: David Silver, W. Maurice Young Centre for Applied Ethics Supervisor  Matt Bedke, Philosophy  Supervisory Committee Member  Alison Wylie, Philosophy  University Examiner Kiley Hamlin, Psychology University Examiner     	 iii	Abstract Personal relationships are rich, complex, and valuable to our lives; yet, they have been largely overlooked by philosophers. This dissertation therefore hopes to highlight the importance of this neglected territory through exploring blame and forgiveness in personal relationships.   Chapter 1 argues that blame in personal relationships (i) is often and should be affective and (ii) often involves and should involve a rich set of emotions—not just anger, but also disappointment, sadness, and hurt feelings. In non-personal relationships, blame (iii) need not involve any affect and (iv) it involves a narrow set of emotions if it is affective. Recognizing the relationship dependency of blame, this chapter proposes a minimal general account of blame: blame is a response to relational harm.   Chapter 2 argues that understanding the interpersonal aspect of forgiveness is key to any adequate understanding of forgiveness, particularly forgiveness in personal relationships. An adequate account of forgiveness should be able to explain three interpersonal features of forgiveness: that it is communicative, reconciliatory, and performative. Chapter 3 then proposes a performative account of forgiveness—forgiveness as a declarative act that unbreaks relationships. This account focusses on and can adequately explain the three features and hence the interpersonal aspect of forgiveness.   Discussion of blame and forgiveness may make it seem as though these were the primary appropriate responses to wrongs, but relationships—particularly personal ones—may give us reason to neither blame nor forgive the wrongdoer. They may instead give us reason to prioritize maintaining the relationship or supporting the wrongdoer. Chapter 4 discusses such reasons from relationships, which enable us to appreciate the richness, depth, and complexity of personal relationships.   Withholding blame for the above reasons from relationships does not necessarily express disrespect for oneself (the victim) or the wrongdoer, contrary to the arguments given by some blame (and forgiveness) theorists. Chapter 5 proposes a broader understanding of respect, according to which withholding blame because of such reasons is a positive way to express respect.    	 iv	Lay Summary This dissertation studies two typical responses to wrongs: blame and forgiveness. Philosophers have for long overlooked the importance of personal relationships and have been more able to explain blame and forgiveness in non-personal relationships than in personal relationships. This dissertation aims to shed light on and deepen our understanding of blame and forgiveness as they occur in personal relationships. It aims to explore the role of blame and forgiveness in (inter)personal relationships in terms of how they function in and are shaped by/shape our relationships with others.   Blame and forgiveness are not the only appropriate responses to wrongs. Relationships, particularly personal ones, sometimes tell us to set aside blame and forgiveness altogether in order to focus on rebuilding the relationship or supporting the wrongdoer. Though we would be withholding blame against the person who has wronged us, such non-blame responses may yet be legitimate in personal relationships.    	 v	Preface This dissertation is original, unpublished, independent work by the author, Irwin Yu Shing Chan.     	 vi	Table of Contents ABSTRACT	..........................................................................................................................................................	iii	LAY	SUMMARY	...................................................................................................................................................	iv	PREFACE	...............................................................................................................................................................	v	TABLE	OF	CONTENTS	.......................................................................................................................................	vi	ACKNOWLEDGEMENTS	...................................................................................................................................	xi	INTRODUCTION	..................................................................................................................................................	1	CHAPTER	1	BLAME	............................................................................................................................................	8	Section	1	Let’s	Talk	Blame	Realistically	.............................................................................................................................	10	(1.1)	Blame:	Three	General	Features	....................................................................................................................................	10	(1.2)	The	Affectivist/Strawsonian	..........................................................................................................................................	12	(1.3)	The	Non-affectivist	.......................................................................................................................................................	14	(1.4)	Where	Both	Camps	Fail:	Relationship-neutral	Approach	and	Blame	in	Personal	Relationships	..................................	16	(1.4.1)	The	Affectiveness	of	Blame	in	Personal	Relationships	..........................................................................................	16	(1.4.2)	The	Richness	of	Blame	Emotions	in	Personal	Relationships	..................................................................................	19	(1.5)	The	Nature	of	Blame	vs	Blame	in	Practice	....................................................................................................................	21	(1.6)	Summary	.......................................................................................................................................................................	22	Section	2	Blame	It	With	Your	Heart:	The	Affectiveness	of	Blame	in	Personal	Relationships	..............................................	23	(2.1)	Personal	Relationships:	Intimacy	as	Emotional	Vulnerability	.......................................................................................	23	(2.1.1)	Intimacy	as	Emotional	Vulnerability	......................................................................................................................	24	(2.1.2)	Intimacy	and	the	Affectiveness	of	Intimate	Blame	................................................................................................	24	(2.2)	Personal	Relationships:	Particularity	............................................................................................................................	26	(2.2.1)	Seeing	Someone	as	a	Particular	Person	.................................................................................................................	27	(2.2.2)	Particularity	and	Affective	Blame	in	Personal	Relationships	.................................................................................	28	(2.3)	Three	Clarifications	.......................................................................................................................................................	30	(2.3.1)	Two	Difficult	Cases?	...................................................................................................................................................	30	(2.3.2)	Affectiveness	and	Seriousness	...............................................................................................................................	31	(2.3.3)	Personal	Relationships:	Degree	of	Closeness	........................................................................................................	32	(2.4)	Objections:	Affectless	Personal	Relationships?	............................................................................................................	33	(2.4.1)	Conceptual	Difficulty	with	Affectless	Personal	Relationships	................................................................................	33	(2.4.2)	Completely	Affectless	Personal	Relationships	Are	Very	Rare	................................................................................	34	(2.4.3)	Distinguishing	Acts	and	Background	Relationships	...............................................................................................	35	(2.5)	Summary	.......................................................................................................................................................................	35	Section	3	Blame	It	Colourfully:	The	Richness	of	Blame	Emotions	in	Personal	Relationships	..............................................	36	(3.1)	Two	Distinctive	Expectations	in	Personal	Relationships	...............................................................................................	37	(3.1.1)	Care	........................................................................................................................................................................	37	(3.1.2)	Loyalty	....................................................................................................................................................................	38		 vii	(3.2)	Damage	to	Intimacy	......................................................................................................................................................	39	(3.3)	Concern	and	Moral-Aretaic	Blame	...............................................................................................................................	39	(3.4)	Two	Objections	.............................................................................................................................................................	41	(3.5)	Summary	.......................................................................................................................................................................	43	Section	4	Two	Underlying	Flaws	.......................................................................................................................................	44	(4.1)	Insensitivity	to	Relationship	Contexts	...........................................................................................................................	45	(4.2)	Insufficient	Attention	to	Personal	Relationships	..........................................................................................................	46	(4.3)	Summary	.......................................................................................................................................................................	49	Section	5	A	Minimal	Account	of	Blame	.............................................................................................................................	50	(5.1)	Relationship-neutral	yet	-dependent:	Blame	as	a	Response	to	Relational	Harm	.........................................................	50	(5.2)	Two	Features	of	Blame	in	Personal	Relationships	........................................................................................................	52	(5.2.1)	Affectiveness	.........................................................................................................................................................	52	(5.2.2)	Rich	Emotions	........................................................................................................................................................	52	(5.3)	The	“Force”	of	Blame	....................................................................................................................................................	53	(5.3.1)	Personal	relationships	...........................................................................................................................................	54	(5.3.2)	Non-personal	Relationships	...................................................................................................................................	55	(5.4)	Summary	.......................................................................................................................................................................	55	Chapter	Conclusion	..........................................................................................................................................................	56	CHAPTER	2	FORGIVENESS:	PRIVATE	VS	INTERPERSONAL	...............................................................	57	Section	1	Two	Mainstream	Accounts	and	Personal	Relationships	.....................................................................................	58	(1.1)	The	Standard	Account	..................................................................................................................................................	58	(1.1.1)	Overcoming	Resentment	on	Moral	Grounds	.........................................................................................................	58	(1.1.2)	Attempts	to	Expand	the	Standard	Account	...........................................................................................................	59	(1.2)	Change-in-View	Account	..............................................................................................................................................	61	(1.2.1)	Jean	Hampton	and	Margaret	Holmgren	................................................................................................................	61	(1.2.2)	Problem:	Depersonalizing	Forgiveness	in	Personal	Relationships	.........................................................................	63	(1.2.3)	A	More	Particularized	View?	.....................................................................................................................................	67	(1.3)	Non-Mainstream,	Personal	Relationship-friendly	Accounts?	...................................................................................	68	Section	2	Feature	One:	Forgiveness	is	Communicative	.....................................................................................................	71	(2.1)	Three	Interpersonal	Features	of	Forgiveness	...............................................................................................................	72	(2.2)	Two	Mainstream	Accounts:	Forgiveness	as	Non-communicative	................................................................................	74	(2.3)	Performative	Account	as	a	Replacement	.....................................................................................................................	76	(2.3.1)	Revised	or	Alternative?	..........................................................................................................................................	77	(2.3.2)	Complement	or	Replacement?	..............................................................................................................................	78	Section	3	Feature	Two:	Reconciliation	..............................................................................................................................	80	(3.1)	Reconciliation,	Really?	..................................................................................................................................................	81	(3.2)	Reconciliation:	Mostly	Political,	Partly	Personal	...........................................................................................................	82	(3.3)	Restoration—the	Previous	State	..................................................................................................................................	83	(3.4)	Reconciliation	as	Re-establishment	of	Trustworthiness	...............................................................................................	84	(3.4.1)	Linda	Radzik	...........................................................................................................................................................	84	(3.4.2)	Margaret	Walker	....................................................................................................................................................	86	(3.5)	Two	Mainstream	Accounts:	No	Reconciling	with	Reconciliation	..................................................................................	87	(3.6)	Performative	Account:	A	New	Way?	............................................................................................................................	88	Section	4	Feature	Three:	Forgiveness	and	Performative	...................................................................................................	90	(4.1)	Forgiveness	as	Performative	.........................................................................................................................................	90		 viii	(4.2)	Reply	1:	Emotions	and	Perceptions	as	Relationships	...................................................................................................	91	(4.3)	Reply	2:	Expressions	of	Emotions	and	Perceptions	......................................................................................................	92	(4.4)	Reply	3:	Insincere	Forgiveness?	....................................................................................................................................	94	(4.5)	Summary	.......................................................................................................................................................................	95	Chapter	Conclusion	..........................................................................................................................................................	96	CHAPTER	3	PERFORMATIVE	ACCOUNT	OF	FORGIVENESS	................................................................	97	Section	1	Applying	Speech	Acts	to	Forgiveness	.................................................................................................................	98	(1.1)	Speech	Act	Theory	and	the	Performative	Account	of	Forgiveness	...............................................................................	98	(1.2)	Is	Emotion	Necessary?	................................................................................................................................................	100	(1.3)	Emotion	and	Relationships	.........................................................................................................................................	102	(1.3.1)	Forgiveness	as	Overcoming	Blame	......................................................................................................................	102	(1.3.2)	Forgiveness	Without	Internal	Conditions?	..........................................................................................................	103	(1.3.3)	Overcoming:	Complete	or	Partial?	......................................................................................................................	104	(1.3.4)	Internal	Conditions	as	Success	Conditions?	.........................................................................................................	105	(1.4)	Summary	.....................................................................................................................................................................	107	Section	2	Main	Strength	of	the	Performative	Account:	Personal	and	Interpersonal	Forgiveness	.....................................	108	(2.1)	Communicativeness	....................................................................................................................................................	108	(2.1.1)	Objection	1:	Private	Forgiveness	and	“Imputed	Forgiveness”	.............................................................................	108	(2.1.2)	Objection	2:	Completely	Private	Forgiveness	......................................................................................................	110	(2.1.3)	Objection	3:	Significance	of	Completely	Private	Forgiveness	..............................................................................	110	(2.2)	Reconciliation	.............................................................................................................................................................	111	(2.3)	Performativeness	........................................................................................................................................................	113	(2.4)	Accurate	Conceptualization	of	Personal	Relationships	..............................................................................................	114	(2.5)	Summary	.....................................................................................................................................................................	114	Section	3	Performative	Account	and	the	Recognition-Proleptic	Distinction	....................................................................	115	(3.1)	Recognition	and	Proleptic	Forgiveness	.......................................................................................................................	115	(3.2)	Proleptic	Forgiveness:	Distinctive	Features	and	Distinctive	Treatment	.....................................................................	117	(3.3)	Performative	Account	and	Proleptic	Forgiveness	.......................................................................................................	119	(3.4)	Summary	.....................................................................................................................................................................	120	Section	4	Performative	Account	and	The	Normative	Value	of	Forgiveness	.....................................................................	121	(4.1)	The	Paradox	of	Forgiveness	........................................................................................................................................	121	(4.2)	Nussbaum’s	Critique	of	Forgiveness	...........................................................................................................................	122	Chapter	Conclusion	........................................................................................................................................................	124	CHAPTER	4	REASONS	FOR	NON-BLAME	RESPONSES	........................................................................	125	Section	1	How	to	Respond	to	Wrongs:	Reasons	From	Relationships	...............................................................................	127	(1.1)	Relationship	Maintenance	..........................................................................................................................................	127	(1.2)	Concern	.......................................................................................................................................................................	129	(1.3)	Trust	............................................................................................................................................................................	132	(1.3.1)	“It’s	Not	What	You	Think”	....................................................................................................................................	132	(1.3.2)	Not	Blameworthy	.................................................................................................................................................	134	(1.4)	Understanding	............................................................................................................................................................	135	(1.5)	Weighing	the	Strength	of	Reasons	from	Relationships	..............................................................................................	135	(1.6)	The	Depth	and	Richness	of	Our	Personal	Lives	...........................................................................................................	136		 ix	(1.7)	Objection	and	Reply:	Burdening	Victims	....................................................................................................................	137	(1.7.1)	Objection	.............................................................................................................................................................	137	(1.7.2)	Reply	....................................................................................................................................................................	139	(1.8)	Summary	.....................................................................................................................................................................	141	Section	2	Scepticism	of	Blame	........................................................................................................................................	142	(2.1)	Nussbaum’s	Two	Errors	of	Anger	...............................................................................................................................	142	(2.2)	Replies	........................................................................................................................................................................	143	(2.2.1)	Defensible	Retributivism?	....................................................................................................................................	144	(2.2.2)	Specifying	the	Wish	for	Suffering	........................................................................................................................	145	(2.3)	Summary	.....................................................................................................................................................................	147	Section	3	The	Nature	of	Reasons	from	Relationships	......................................................................................................	148	(3.1)	Reasons	for	Blame:	Narrow	and	Wide	........................................................................................................................	148	(3.2)	Reasons	for	Blame:	Narrow	or	Wide?	........................................................................................................................	149	(3.3)	Wide	Not?	...................................................................................................................................................................	150	Section	4	The	Role	of	Reason	in	Personal	Relationships	.................................................................................................	153	(4.1)	Reasons:	Justifying	and	Motivating;	Abstract	and	Particularized	...............................................................................	153	(4.2)	Role	of	Reasons	in	Personal	Relationships	.................................................................................................................	156	(4.3)	Summary	.....................................................................................................................................................................	158	Chapter	Conclusion	........................................................................................................................................................	159	CHAPTER	5	RESPECT	...................................................................................................................................	160	Section	1	Respect	as	Dignity	and	Agency	........................................................................................................................	161	(1.1)	Respect:	Dignity	and	Agency	......................................................................................................................................	161	(1.2)	Respect	in	Personal	Relationship	................................................................................................................................	163	Section	2	Have	Some	Respect	and	Blame	.......................................................................................................................	165	(2.1)	Reasons	from	Relationships	and	Withholding	Blame	.................................................................................................	165	(2.2)	Psychological	Argument	.............................................................................................................................................	166	(2.3)	Normative	Arguments	................................................................................................................................................	167	(2.3.1)	Self-respect	..........................................................................................................................................................	167	(2.3.2)	Other-respect	......................................................................................................................................................	169	(2.4)	Summary	.....................................................................................................................................................................	170	Section	3	A	Broader	Approach	to	Respect	......................................................................................................................	171	(3.1)	Two	Cases	...................................................................................................................................................................	171	(3.1.1)	Prince	Myshkin	....................................................................................................................................................	171	(3.1.2)	Maggie	.................................................................................................................................................................	173	(3.2)	Broader	Approach	to	Respect	and	Reasons	from	Relationships	................................................................................	174	(3.2.1)	Relationship	Maintenance	...................................................................................................................................	175	(3.2.2)	Concern	................................................................................................................................................................	175	(3.2.3)	Trust	.....................................................................................................................................................................	176	(3.4)	Summary	.....................................................................................................................................................................	176	Chapter	Conclusion	........................................................................................................................................................	177	CONCLUSION	...................................................................................................................................................	178		 x	BIBLIOGRAPHY	..............................................................................................................................................	180	   	 xi	Acknowledgements I would like to first and foremost thank my mother and my sister. Without their unwavering support, continuous encouragement, and selfless sacrifices, I would not have been able to begin this PhD in the first place. This PhD has been a long and at times challenging journey. Their kindness, understanding, and belief in me have enabled me to march on and thrive. I owe them everything that I have achieved in these past seven years.   I would also like to thank my supervisor David Silver, who encouraged me to develop my initially messy ideas and helped me identify their value. The beginning stage of this dissertation was bumpy; his faith in my ability, guidance, and ongoing encouragement helped me push through and finish. I am also grateful to the members of my committee, Paul Russell and Matt Bedke. Their fair and insightful comments have helped me clarify my arguments and improve this dissertation.   If there was one place where this dissertation began, it was certainly in my weekly reading group meetings (via video calls) with Tim King Wai Li. Many of my ideas about personal relationships were initially developed during our discussions, and his fair-mindedness and patience have helped me identify the strengths and weaknesses of my arguments.   I would also like to give my thanks to Chris Stephens, my mentor from the start of my PhD, for his generosity with his time and advice. Whether it is about my research, teaching, or personal matters, Chris was always happy to chat. His guidance has helped me navigate through my program and solve many problems.   Last but certainly not least, I would like to thank my partner Erika for her love, companionship, and understanding. I was most fortunate to always have her on my side in my most difficult times. I will never forget our long walk on Spanish Banks during the lowest point of my PhD career (and being surrounded by the water as the tide was quickly coming in).    	 xii	Dedication        Dedicated to my mother 	 1	Introduction Relations between equals and nonintimates will be the moral norm for adult males whose dealings with others are mainly business or restrained social dealings with similarly placed males. But for lovers, husbands, fathers, the ill, the very young, and the elderly, other relationships with their moral potential and perils will loom larger. Annette Baier. “Trust and Antitrust”. (1986, 248)  We should think of the many different kinds of relationship which we can have with other people—as sharers of a common interest; as members of the same family; as colleagues; as friends; as lovers; as chance parties to an enormous range of transactions and encounters. Then we should think, in each of these connections in turn, and in others, of the kind of importance we attach to the attitudes and intentions towards us of those who stand in these relationships to us, and of the kinds of reactive attitudes and feelings to which we ourselves are prone. … The range and intensity of our reactive attitudes towards goodwill, its absence or its opposite vary no less widely.  Peter Strawson. “Freedom and Resentment”. (2008, 6-7)  Personal relationships form an important and valuable part of our lives. They are a place where we spend a significant amount of our time with one another, provide mutual support of all kinds, form deep emotional bonds, and pursue short-term, long-term, if not life-long mutual projects; many of these things may not be possible outside personal relationships. It is therefore not surprising that human beings need and will seek personal relationships, and our flourishing often hinges on such relationships.   But as Annette Baier (1986) insightfully points out in the first quote, there has been a long-standing disregard of personal relationships in moral philosophy. In the blame and forgiveness literature, for example, we see a theoretical slant towards non-personal relationships. Fixated on the non-personal realm, theorists model their accounts on non-personal relationships and present their accounts as “relationship-neutral” such that they can purportedly explain blame and forgiveness in all relationship contexts, giving us the impression that non-personal relationships are all that there is or matters.     But not only do personal relationships matter and deserve more of our attention, they are also in many ways very different from non-personal relationships. What blame is, what forgiveness is, and what responses to wrongs are appropriate are all very different in personal compared with non-personal relationships. This distinctiveness of blame and forgiveness in personal relationships means 	 2	that blame and forgiveness are highly dependent on the relationship context: they have different features in different relationships—personal, non-personal, or anything in-between.  The relationship dependency of blame and forgiveness exposes another problem with accounts in the current literature. If we fail to appreciate the relationship dependency of blame and forgiveness and use these accounts (which are modelled on non-personal relationships) to explain blame and forgiveness in personal relationships, they will at best fail to adequately explain and at worst distort blame and forgiveness in personal relationships.  In light of the above problems, this dissertation specifically emphasizes personal relationships and the relational aspects of blame and forgiveness. First, it recognizes the value of personal relationships. The literature on blame and forgiveness has been skewed towards non-personal relationships, but any attempt to understand blame and forgiveness would be incomplete if it does not pay adequate attention to these phenomena in personal relationships. This dissertation therefore focusses primarily on blame and forgiveness in personal relationships.   Second, this dissertation recognizes the relationship dependency of blame and forgiveness. By delving into the personal realm, one will soon realize that blame and forgiveness are very different in personal relationships compared with non-personal relationships. (Hence, they are very different from what theorists tell us because their accounts are modelled on non-personal relationships.) Because of this relationship dependency, any attempt to explain blame and forgiveness should be attentive to the relationship context in which the blame and forgiveness occur. In other words, if we want to understand blame and forgiveness, we need to first and foremost ask: blame and forgiveness in what relationship? Personal? Non-personal? Or something in-between?   Both focussing on personal relationships and being attentive to the relationship dependency of blame and forgiveness naturally bring our attention to the actual workings of blame and forgiveness in relationships and how they are connected to how we relate to other people. This brings us to the third point: this dissertation aims to explain blame and forgiveness as normative practices situated in interpersonal relationships, particularly in personal relationships. In other words, this dissertation aims to explain the roles of blame and forgiveness in (inter)personal relationships in terms of how 	 3	they function in, (in the case of blame) are shaped by, and (in the case of forgiveness) shape (inter)personal relationships.   This third point has an important implication for the explanatory work that this dissertation aims to do. It does not aim to explain the necessary and sufficient conditions of blame and forgiveness, but to explain features of blame and forgiveness that are significant in relationships. Theorists have been primarily interested in seeking the “nature” of blame and forgiveness (i.e. their necessary and sufficient conditions) and therefore aim to provide accounts that can explain and identify blame and forgiveness in all situations and all relationships.1 This dissertation aims instead to explain features that are crucial to blame and forgiveness as normative practices situated in (inter)personal relationships, even if such features may not be found in all cases of blame and forgiveness.   After explaining the overarching approach of this dissertation, I will give an overview of the chapters of this dissertation below.   Blame The first chapter deals with blame in personal relationships. It first demonstrates that the two major accounts of blame in the current literature—affectivist/Strawsonian and non-affectivist—both fail to adequately explain blame in personal relationships. According to the affectivist, blame is an affect; it is a moral form of anger that responds to violations of moral expectations. On the other hand, the non-affectivist holds that blame does not necessarily involve affect. Blame should instead be identified with some conative elements, i.e. intentions, dispositions, expectations, etc. Both accounts fail to recognize two distinctive features of blame in personal relationships: it (i) is and should be affective and (ii) involves a rich set of emotions—not only anger but also disappointment, hurt feelings, sadness, etc. These two features make up the first role of blame in (inter)personal relationships: how it functions in (inter)personal relationships.   I will then give an in-depth discussion of these two features and explain why blame in personal relationships has such features by referring to several distinctive features of personal relationships. I also attribute the two accounts’ (affectivist/Strawsonian and non-affectivist) failures to explain blame 																																																																		1 There are some exceptions; for instance, Miranda Fricker (2014) aims to explain the “paradigm” of blame, and McGeer (2012) proposes a functionalist account that acknowledges the existence of “exceptional cases”. 	 4	in personal relationships to theorists’ lack of awareness of the relationship dependency of blame and under-appreciation of the importance of personal relationships. I end the chapter with a proposal of a minimal account of blame—blame as a response to relational harm—that explains the general, relationship-neutral feature of blame and yet recognizes its relationship dependency. By using the distinctive features of personal relationships and by conceiving blame in terms of relational harm, I then explain the second role of blame in personal relationships, that is, how blame is shaped by (inter)personal relationships.  Forgiveness The subsequent two chapters are more ambitious in that they aim to give a general account of forgiveness that explains forgiveness in both personal and non-personal relationships, though my focus is on personal relationships. Similar to chapter 1, I first demonstrate in chapter 2 that the two mainstream accounts of forgiveness—the standard account and Change-in-View Account—cannot adequately explain forgiveness in personal relationships. The standard account argues that forgiveness is the overcoming of resentment, whereas the Change-in-View Account argues that forgiveness consists in changing one’s perception of the wrongdoer from a wrongdoer to something more benign, such as a person who is “still decent” or a moral agent. Both accounts fail to explain forgiveness in personal relationships. The standard account, by limiting itself to resentment, fails to explain the rich blame emotions in personal relationships. The Change-in-View Account misconceptualizes how we perceive those who are close to us. We do not see them as “still decent” or as “moral agents”; instead, we see them as particular persons with particular personalities: my honest friend, my unreliable sibling, my humorous wife, etc.   Then I give my main criticism of the two mainstream accounts: they fail to explain the interpersonal aspect of forgiveness—an aspect of forgiveness that is crucial to forgiveness as a practice situated in interpersonal relationships, particularly in personal relationships. Forgiveness, as it functions in our interpersonal relationships, is interpersonal in the sense that it is communicative, reconciliatory, and performative. It is communicative because we, as the forgiver, often express our forgiveness to the person we are forgiving and want the person to know that they have been forgiven; likewise, as the person to be forgiven, we want to know if we have been forgiven. Forgiveness is reconciliatory because forgiveness can in some sense reconcile the relationship damaged by the wrong; and it reconciles the relationship in virtue of its being granted—that is, it is performative. These three 	 5	features of the interpersonal aspect of forgiveness are particularly important in personal relationships because they are all related to and play decisive roles in determining how we are to relate to one another after wrongs have occurred, and such decisions are important in personal relationships.   The two mainstream accounts take what I call the private approach to forgiveness. Forgiveness, on these two accounts, consists in something inside our heads/hearts: either the overcoming of some blame emotions (the standard account) or seeing the wrongdoer in some benign way (Change-in-View Account). This private approach renders the two accounts unable to adequately explain the three interpersonal features of forgiveness for the simple reason that these features are public, “outside-the-head/heart” features. The two accounts could at most give these features a secondary role and would thus fail to recognize the importance of these features.   I then propose in chapter 3 a performative account of forgiveness: forgiveness is a declarative act that unbreaks the relationship. This performative account has several strengths. The most important strength is that it can adequately explain forgiveness in personal relationships. It can explain the interpersonal aspect of forgiveness (that forgiveness is communicative, reconciliatory, and performative), acknowledges the richness of blame emotions in personal relationships, and does not misconceptualize how we perceive those who are close to us.   The role of forgiveness in (inter)personal relationships as revealed in these two chapters is fairly obvious. The interpersonal aspect of forgiveness (i.e. the three features) demonstrates how forgiveness functions in (inter)personal relationships, and my performative account shows clearly how forgiveness shapes (inter)personal relationships through unbreaking them.   Reasons from Personal Relationships and Appropriate Responses to Wrongs Discussions in the blame and forgiveness literature may at times give us the impression that blame and forgiveness are the only appropriate responses to wrongs. The reality is of course not so simple, particularly in relationships as deep and complex as personal ones. The main argument of chapter 4 is that there is a multiplicity of reasonable responses to wrongs in personal relationships. Personal relationships may call for blame and forgiveness when wrongs occur, but they may also call for responses that have nothing to do with blame and forgiveness. For instance, they may give us reason to set blame and forgiveness aside and focus on finding the underlying problems of the relationship 	 6	that gave rise to the wrong in the first place, or they may give us reason to refrain from blaming and instead to show support and trust towards the personal who has wronged us. We should therefore generally construe the question in terms of “appropriate responses to wrongs in personal relationships”, where blame and forgiveness are two among many possible responses, rather than narrowly construe it in terms of “reasons for and against blame and forgiveness”.   Chapter 4 also deals with a few other issues related to blame and reasons from personal relationships. Some theorists might take a more sceptical if not radical stance on blame such that blame (or specifically angry blame) is unjustifiable. I do not hold this strong position. I argue that angry blame is justifiable even in personal relationships. I also clarify in the second half of the chapter that the reasons discussed in this chapter deal not only with whether the person has actually done anything wrong but also broader range of factors, and discuss the role of justifying and motivating reasons in personal relationships.  Respect Most of the reasons from relationships discussed in chapter 4 tell us to adopt a non-blame response wrongs, that is to focus on improving the damaged relationship. Some might argue that, to respect oneself and the wrongdoer, we should respond to wrongs with blame. Does it mean that personal relationships are asking us to sacrifice our respect for ourselves and for the wrongdoer? If so, then these reasons from relationships might not be very good reasons after all. Despite all the complexity of personal relationships, maybe blame is the best response to wrongs.   Chapter 5 addresses this worry about respect. Two aspects of respect are mainly in question: (i) dignity, which deals with whether a person is appropriately treated as a moral equal; and (ii) agency, which deals with whether a person is appropriately treated as someone who is capable of acting in light of reason. I will discuss and reject a few arguments for the claim that dignity and agency require one to respond to wrongs with blame. I will argue for the adoption of a broader approach to respect such that respect for dignity and agency may be expressed in ways other than blame, and I will demonstrate how the reasons from relationships discussed in chapter 4 express respect for dignity and agency in this broader sense.   “Unbreak My Heart”? 	 7	A final word about the title of this dissertation. This metaphor, borrowed from a rather old pop song, is to highlight the importance of the “heart” in personal relationships and how wrongs, blame, and forgiveness in personal relationships all in various ways touch our “hearts”. Being wronged by those close to us break our hearts and we blame them with a broken heart. Though forgiveness cannot always unbreak our hearts—sometimes a broken heart just can’t be unbroken, it can however contribute to it unbreaking through unbreaking the relationship. We may also try to unbreak our hearts by setting blame and forgiveness aside and focussing on maintaining the relationship, supporting the wrongdoer, or trusting them.  	 8	Chapter 1 Blame This chapter has two goals. First, it aims to explore in great detail what blame is like in personal relationships as compared with that in non-personal relationships. In particular, it argues that affectiveness is a psychologically typical and normatively appropriate feature of blame in personal relationships, and its affectiveness is expressed by a set of emotions richer than that in non-personal relationships—not only moral anger, but also disappointment, sadness, and hurt feelings. This exploration of the distinctive features of blame in personal relationships will show that accounts in the blame literature, due to their “relationship-neutral approach”, have failed to adequately explain blame in personal relationships. The second goal of this chapter is therefore to suggest the adoption of a “relationship-dependent approach” to blame—a “Who-are-you-to-me approach”, as I will call it—and propose a relationship-neutral account of blame under this approach.    This chapter is divided into five sections. Section 1 gives a brief overview of accounts of blame in the current literature with an aim to introduce two distinctive features of blame in personal relationships that these accounts fail to explain: the affectiveness of blame and the richness of blame emotions in personal relationships. It also explains two methodological points crucial to this chapter. First, while blame theorists have taken a relationship-neutral approach to blame, which aims to explain features of blame that can be found in all relationships, I propose to adopt a relationship-dependent approach to blame, which recognizes the relationship dependency of blame and aims to explain features of blame in different relationships. Second, this chapter does not aim to explain the nature of blame (i.e. the necessary and sufficient conditions of blame); rather, I aim to explain features that are important to blame as a practice situated in interpersonal relationships, particularly in personal relationships. In other words, I aim to explain the role of blame in (inter)personal relationships in terms of how it functions in and is shaped by (inter)personal relationships.   Sections 2 and 3 substantiate my claim made in section 1 that blame in personal relationships is affective and involves a rich set of emotions. I will use two constitutive features of personal relationships to explain the first feature in section 2, and I will use the distinctive expectations in personal relationships, damage to intimacy, and moral-aretaic blame to explain the second feature in section 3.   	 9	Section 4 moves on to the second goal of this chapter. I will argue that my observations of blame in personal relationships and the failure of accounts of blame in the current literature to adequately explain it point us to the direction of a relationship-dependent approach to blame, and I will sketch a relationship-neutral account of blame under this approach in section 5. 	 10	Section 1 Let’s Talk Blame Realistically  Blame is something that we do and experience in our moral lives, so our attempt to understand it should start there. My approach to blame in this chapter and other topics in the coming chapters will always start from our practice in our moral lives.  The question that this first section addresses is: how do we blame in one important part of our social lives—personal relationships? This section starts with a brief discussion of three general features of blame (1.1), then moves on to an overview of two major accounts in the blame literature: the affectivist (1.2) and the non-affectivist (1.3). Then, in (1.4), it argues that both the affectivist and non-affectivist accounts fail to adequately explain two distinctive features of blame in personal relationships: that blame in personal relationships (i) is and should be affective, and (ii) it does and should involve a rich set of emotions.   Two methodological points crucial to this chapter are discussed. The first one deals with the relationship-neutral approach. This is the approach that theorists in the blame literature take, which aims to explain features of blame found in all relationships. This will be discussed in (1.4). (1.5) explains the second methodological point. While blame theorists have mostly been concerned with explaining the nature of blame—that is, the necessary and sufficient conditions of blame found in all (or almost all) cases of blame, the explanatory work that I aim to do in this chapter is to explain features that are crucial to blame as a normative practice situated in interpersonal relationships, particularly personal ones.  (1.1) Blame: Three General Features When we blame our friends for not helping us when we are in need, the shopkeeper for giving us wrong change, or politicians for their lies and false promises, what exactly are we doing?2 In most 																																																																		2 The discussion in this dissertation focusses mainly on significant but not very serious wrongs (e.g. breaking promises and betraying a friend, but not murder, sexual assault, or genocide) and non-significant but yet non-trivial wrongs (e.g. forgetting a birthday and not showing up to a meeting, but not littering, being late for 5 minutes, or putting a towel in a wrong place), though such cases (very serious wrongs and trivial wrongs) will be discussed every now and then.    There are three reasons for this choice of focus. First, trivial wrongs take place very often in both personal and non-personal relationships, but, given their trivial nature, they may not be the most significant aspect of the  blame and forgiveness phenomena.   	 11	general terms, blame (of the moral kind) is a response to violations of moral standards. Theorists have proposed various accounts to further specify the nature of blame, but they still find a complete account of blame elusive. Before discussing the details of their accounts, I will start by pinpointing the blame phenomenon that theorists are interested to account for. I will focus on three general features of blame.   First, the phenomenon that theorists are interested in is that of moral blame rather than causal blame. We sometimes say that we “blame” the computer for losing our important data, or “blame” the weather for the cancellation of a picnic. Whether or not the use of “blame” in such instances is merely a figure of speech, this is not the kind of blame that interests theorists in the blame literature. What they are interested is moral blame, the kind of blame that involves the belief that some moral agent has violated some moral standards. For instance, when children lie to their parents, when our lovers let us down, and when politicians break their promises.   Second, blame is “forceful”. Some blame theorists, the “cognitivists”, argued that blame is simply a judgement about the fact that the blamee has violated some moral standards. They take blame as a judgement that there is a “discredit” in the wrongdoer’s “moral record”, or that the wrongdoer’s “moral standing” has been diminished (Glover 1970; Haji 1998; Zimmerman 1988). But most blame theorists nowadays agree that blame carries a certain kind of “force” that a merely cognitive activity cannot explain. When we blame a murderer for his crime, when a parent blames their child for misbehaving, and when a husband blames his partner for being uncaring, the blamer is not merely making a cold judgement about the wrongdoer, but is doing something that carries some kind of “force” (Coates and Tognazzini 2012, 9), which can be felt from both the blamer’s and the blamee’s 																																																																																																																																																																																																																																						Second, very serious wrongs do take place and they take place too often—even one case is too many. And when they take place, they give rise to serious consequences and may wreck the lives of the victim and his/her family. But what is more common and what we have to deal with more frequently may be wrongs that are significant but yet not as serious, e.g. lying, betrayal, promise-breaking, verbal abuse, sexual harassment, etc. Collectively, these wrongs may have more profound and significant effects on our lives than the most serious wrongs.   Third, and most importantly, a discussion that focusses primarily on very serious wrongs and trivial wrongs may skew our understandings of the blame and forgiveness phenomena. In cases of very serious wrong, blame and forgiveness may have the most extreme form due to the serious nature of the wrong. In cases of trivial wrongs, blame and forgiveness may have the most casual form, or there may not even be blame and forgiveness at all. Focussing on such cases may make us lose sight of the more common, ordinary cases of wrongs, blame, and forgiveness. Such cases should be our main concern. When we do discuss cases of very serious wrongs and trivial wrongs, we should bear in mind the nature of these wrongs and recognize that blame and forgiveness may well take a form different from those in cases of non-serious and non-trivial wrongs. 	 12	perspective. Blame theorists have yet to specify the nature of this “force” nor to agree on what it is, but they agree that it is an essential feature of blame that any account of blame needs to be able to explain.   Third, blame can be private or public. Identifying blame as negative overt behaviour may seem a nice way to capture the “force” of blame because blame is often associated with activities such as writing to a newspaper to criticize dishonest politicians, condemning multinational corporations’ aggressive tax avoiding strategies, and punishing criminals by putting them behind bars. Nonetheless, such accounts of blame may be able to nicely explain the expressions of blame, but they cannot fully explain blame itself. It is a very common phenomenon we blame the wrongdoer without expressing it. We may be staring at the back of the queue-jumper with distain while not wanting to call him out because we don’t want to “make a scene”, or we may be grumbling inside about the stupid mistakes that our colleague made, while not wanting to openly criticize her for the sake of our work relationship. In general, blame can be private or overt, and a good account of blame needs to account for both forms of blame. Identifying blame as negative overt behaviour can account only for the latter but not the former.   (1.2) The Affectivist/Strawsonian Blame theorists have largely come to a consensus of these three general features of blame, from which they start their projects to build accounts that explain the nature of blame—that is, the “essence”, or the necessary and sufficient conditions of blame. Among accounts that have been proposed, two can be seen as the major ones in the literature: the affectivist/Strawsonian account and the non-affectivist account. I will discuss these two accounts in this sub-section and the next.   The affectivist/Strawsonian account is without doubt the most dominant account of blame in the literature. The affectivists argue that blame is necessarily affective: it involves a distinctive kind of moral emotion3 triggered by violations of moral demands. This account is originated from P. F. 																																																																		3 Blame theorists often adopt a cognitivist account of affects. (For instance, Jay Wallace (1994) and Martha Nussbaum (2016) explicitly endorse a cognitivist understanding of anger.) Roughly speaking, a cognitivist account takes affects to consist in a distinctive form of cognitive/evaluative judgements with certain phenomenological and motivational components. To illustrate it with an example pertinent to our discussion, moral anger consists in a cognitive judgement that some moral expectation has been violated, where this judgement involves certain feelings (e.g. the feeling that you were as if going to explode) and motivation (e.g. to strike back). I will likewise use this cognitivist understanding in my discussion. 	 13	Strawson’s “Freedom and Resentment” (2008). One main contribution of “Freedom and Resentment” is Strawson’s bringing to the fore the importance of the reactive attitudes in our understanding of moral responsibility. He observes that it matters to us, as participants of social transaction, whether other people hold good will or ill will towards us, and that we respond to the quality of their will with affective responses that he calls the “reactive attitudes”. The reactive attitudes are the attitudes that one would hold when one participates in interpersonal relationships in “non-detached” manners, such as “gratitude, resentment, forgiveness, love, and hurt feelings”(5). Strawson’s discussion focusses specifically on three kinds of reactive attitudes. First, resentment. If someone treads on our hand, and if this action reflects ill will, we naturally feel resentment towards him/her (5-6). Strawson calls resentment a “personal reactive attitude”, as it is a response to ill will towards us. Second, indignation. If the ill will is “not towards ourselves, but towards others”, say someone treads on another person’s hand, not your hand, your response will be indignation, a “vicarious/impersonal/moral reactive attitude”. It is “vicarious” because it is a “resentment on behalf of another, where one’s own interest and dignity are not involved” (15). Third, guilt. If one realizes that one holds ill will towards other people, one may turn against oneself and hold “self-reactive attitudes”, like guilt, remorse, and shame, towards oneself (16).  Although Strawson does not use the term “blame” in his paper, it did not stop theorists from taking inspiration from his “Freedom and Resentment” to build their accounts of blame with Strawsonian reactive attitudes. For instance, Jay Wallace (1994), a prominent Strawsonian4 theorist, argues that blaming someone is to be subject to some affects (namely, the reactive attitudes) because the person has violated some moral requirements.5 Note, however, that Wallace’s account of blame is in various ways different from Strawson’s. Instead of following Strawson’s wide interpretation of the reactive attitudes, which take the reactive attitudes to include “gratitude, resentment, forgiveness, love, and hurt feelings” (2008, 5), Wallace narrows them down to resentment, indignation, and guilt (1994, 29; 2011, 122). Moreover, Wallace argues that blame is a response to violations of (Kantian) moral obligations, while Strawson takes blame as a response to ill will. Wallace also takes blame not as a 																																																																		4 I will use the terms “Strawsonian” and “affectivist” interchangeably. 5 Wallace makes a subtle distinction between “blame” and “holding someone morally responsible/blameworthy”. Blaming someone requires that one be actually subject to affects (i.e. the reactive attitudes). By contrast, one may hold someone morally responsible/blameworthy by merely believing that it would be appropriate to be subject to those affects without actually being subject to them (1994, 23; 76-77). 	 14	purely affective matter but contains cognitive elements, namely the belief that moral obligations have been violated (1994, 77).  Wallace defends his affectivist account in a more recent paper (2011). He observes that blame has an intuitive “opprobrium” quality, which he attempts to explain in terms of the “expressive significance” of emotion (357). He argues that blame is a way to care about morality; or, more specifically, the value of the relation of mutual recognition and regard that morality embodies and that wrongs disregard. Emotion is uniquely suitable for expressing that one (the blamer) cares enough about this value such that, if one cares about morality, one must be “emotionally exercised” when wrongs have been committed. Therefore, to count as blaming a person, one has to be emotionally exercised by what the wrongdoer has done (358).  The affectivist account seems to have provided a plausible explanation of the three features of blame. First, the affective response is a response to violations of moral standards and is hence moral. Second, the reactive attitudes, namely resentment and indignation, are attitudes of hostility (or “opprobrium”, as Wallace puts it). Their unpleasant quality can explain the distinctive “force” that blame contains. Third, blame as affect can be expressed or unexpressed. One may take out one’s anger and berates one’s disloyal friend in the strongest terms, or swallow the anger and act as if nothing had happened. Either way, one is blaming the disloyal friend according to the affectivist account, but only in different manners.   (1.3) The Non-affectivist Despite the explanatory power of the affectivist account, a number of theorists, the “non-affectivists”, dispute the affectivist claim that affect is a necessary condition of blame. They argue that what is essential to blame and can ultimately explain the three general features of blame are not affective elements, but are some non-affective, conative elements such as desires, intentions, dispositions, etc.  One prominent non-affectivist, George Sher (2006), observes that he may not blame his daughter or the criminals he reads about on the news with any anger or hostility. He therefore rejects the affectivist’s claim that blame is necessarily affective and instead proposes that what is essential to blame are (i) the belief that the wrongdoer has acted badly or has a bad character and (ii) the desire “that the person in question not have performed his past bad act or not have his current bad 	 15	character” (2006, 95; 112). This account aims to kill two birds with one stone. First, because this combination of belief and desire does not necessarily involve affects, it can explain instances of affectless blame. Second, this account can also explain why blame often (though not necessarily) involves affects. Sher agrees that blame is “closely associated” with “a set of affective and behaviour dispositions” such as anger and hostile behaviour, but he maintains that the essential element of blame that can ultimately explain them are not feelings of anger or hostility, but this combination of belief and (unsatisfied) desires. They may directly motivate anger or hostile behaviour, or they may motivate “second-best desires” which are sufficiently similar to the unsatisfied desire, e.g. reproach or apology.   Another prominent non-affectivist, T. M. Scanlon (2008), proposes that blame consists in modifications of attitudes and dispositions, where affect is just one among many of such attitudes and dispositions. Blame is therefore not necessarily affective. He imagines a scenario where Joe exposed a secret about his friend Tim. Tim may judge that Joe’s behaviour impairs the friendship between them and in turn entails a change in the attitudes and dispositions that would be appropriate for them to hold towards one another. If Tim decides to actually change his attitudes and dispositions towards Joe in light of his judgement, then, on Scanlon’s account, he is blaming Joe. Tim may become less disposed to help Joe when Joe is in need, to see certain reasons for actions as weaker (e.g. loyalty), or hold lower expectations of Joe (e.g. Tim may not expect Joe’s loyalty) (129). Note that Tim need not blame Joe with any affect. He may blame Joe by coldly deciding to see Joe less.     I will briefly explain how the non-affectivist accounts can explain the three features of blame. First, blame on both Sher’s and Scanlon’s accounts is moral, not causal. Second, blame in both accounts has force distinct from mere cognitive judgements. In Sher’s account, the belief-desire pair essential to blame is “closely associated” with the depositions to anger, hostile behaviour, reproach, and apology which in turn can explain the distinctive force of blame (Sher 2006, 112; Coates and Tognazzini 2012, 10). On Scanlon’s account, blame involves taking the relationship with the wrongdoer as impaired by the wrongdoing and modifying one’s attitudes and dispositions towards the wrongdoer accordingly. Such modifications of attitudes and dispositions seem a serious matter in the context of interpersonal relationships, and so stand as a plausible source of the force of blame (Scanlon 2008, 157). Third, blame can be expressed or unexpressed in the non-affectivist accounts. The desire that the wrongdoer not have done the wrong (Sher’s account) and the modifications of 	 16	attitudes and disposition (Scanlon’s account) can be expressed publicly or kept private. For instance, one may simply trust or expect less of the wrongdoer without letting such changes be known.   (1.4) Where Both Camps Fail: Relationship-neutral Approach and Blame in Personal Relationships  Both affectivist and non-affectivist accounts can explain the three general features of blame and correctly describe some aspects of our experience of blame, but they have less success in explaining certain crucial aspects of blame in personal relationships. Both accounts take what I call a relationship-neutral approach to blame; that is, they are accounts that aim to explain blame in all relationship contexts—parent-child relationships, friendship, colleagues, strangers, etc. Affectivist accounts are saying that blame is necessarily affective in all relationships: personal, semi-personal, and non-personal. Similarly, non-affectivist accounts are claiming that blame consists in conative elements and so affect is not a necessary component of blame—again, in all relationships: personal, semi-personal, and non-personal.   But by taking this approach, affectivists and non-affectivists end up paying little attention to blame in personal relationships and failing to adequately explain its two distinctive features: that (1) it is and should be affective and (2) it involves a richer set of emotions. In this sub-section, I briefly discuss these two features and demonstrate how the two accounts fail to adequately explain them.   In fact, these two features of blame in personal relationships show that blame is relationship-dependent; that is, blame has different features in different relationships. To recognize this relationship dependency of blame and more accurately explain features that blame has in different relationships, I argue that we should adopt a relationship-dependent approach and pay closer attention to how blame varies in different relationships (particularly in personal relationships.) And this is the first methodological point of this chapter. A more in-depth discussion of relationship-dependent and relationship-neutral approaches will be given in section 4.   (1.4.1) The Affectiveness of Blame in Personal Relationships By arguing that blame consists in conative elements (e.g. “belief-desire pair” on Sher’s account and modifications of attitudes and dispositions on Scanlon’s account), and affect is merely an accidental feature of blame, non-affectivists fail to recognize the special place that affect has in personal 	 17	relationships. Affect is a very common descriptive feature of blame in personal relationships. When a husband blames his spouse for infidelity, he would not just coldly observe, “this is wrong,” and decide to trust his spouse less or file a divorce. Instead, his remark, “this is wrong!”, will be filled with emotions—anger, disappointment, hurt feelings—and possibly coupled with exclamations such as “how could you do this do me?”   One explanation for this feature is that personal relationships—be it lovers, parent-child relationships, friendship, or siblings—involve a certain degree of emotional engagement, which is reflected in blame. We care about the attitudes those with whom we are personally involved hold towards us and how they treat us, and we do so in an emotionally engaged way such that we will feel resentful and hurt if they do not regard or treat us properly. The closer the relationship, the more emotionally engaged we are, and the more hurt and resentful we become if they wrong us.   Affect is not only descriptively common but also normatively appropriate in personal relationships. The husband should be emotionally engaged in his relationship with his spouse such that he should blame his spouse with emotion. We have reason to think that something has probably gone wrong in the relationship if the infidelity fails to trigger any emotion on the husband’s part. Maybe he does not care enough about his spouse or the relationship, or maybe the relationship is otherwise in bad shape. Either way, it reflects a coldness that is out of place in a relationship that ought to be grounded in mutual love and concern.  One reason why affective blame is normatively appropriate in personal relationships is that emotional vulnerability is a constitutive feature of personal relationships. Part of being in a personal relationship with another person is that what this person does can trigger our emotions, and we do things to this person with emotion. In other words, emotion is something that distinguish acts of personal relationships from acts of non-personal relationships. When we visit our sick friend in the hospital, we would not go to the hospital as if it were merely another item on our schedule. We would visit the friend with heartfelt concern. This heartfelt concern is what makes the visit an act of friendship. And because we have reason to treat those close to us as someone close to us, we have reason to do such things to them with emotion. This constitutive feature also applies to blame. The affect of our blame distinguishes it as blame of personal relationships rather than blame of non-personal 	 18	relationships. When those close to us wrong us, we have reason to blame them as someone close to us, and hence blame them with emotion.   Non-personal relationships, by contrast, do not have emotional vulnerability as a constitutive feature and thus emotion is not a normative criterion of blame in non-personal relationships. This is because emotion does not have as much significance in non-personal relationships as in personal ones. Though we may be emotionally vulnerable to others in non-personal relationships, they do not require that we are emotionally vulnerable to them. There does not seem to be anything inappropriate with responding to their wrongs without any emotion but with only conative elements (e.g. intentions and dispositions). Even if we do respond with emotion, their intensity, depth, and significance are incomparable than that in personal relationships.   Affectiveness may not be a necessary feature of blame in personal relationships—there may well be cases, rare as they are, where we blame those close to us without any affect. But affect is a very common and appropriate feature of blame in personal relationships such that any adequate account of blame should recognize it as a crucial feature of blame in personal relationships. By claiming that blame consists only in conative elements, non-affectivists fail to do so.   Between the two non-affectivist accounts of blame, Scanlon’s account has better potential for explaining blame in personal relationships. His account takes blame to consist in modifications of attitudes and dispositions in response to relationship impairment (caused by the wrong). This could enable him to argue that one of such modifications of attitudes in personal relationships is affect, and hence blame in personal relationships is affective. In fact, he explicitly states that what attitudes and dispositions one modifies when blaming depends on the “ground relationship” (138).   But, strangely enough, he does not think that resentment and other affects are required for blame, even in friendships (136), nor does he think that affect is an important or appropriate feature of blame in friendships. He recognizes the importance of feelings in personal relationships, observing that being a friend involves the dispositions to “take pleasure in the friend’s company, to hope that things go well for the friend and to take pleasure in their going well when they do” (2008, 132), and yet he does not think that the importance of feelings in personal relationships is extended to blame. 	 19	(Stranger yet, he uses an example of blame in friendship to argue that blame need not involve any affect when, as I observe, affect is such an crucial feature of blame in personal relationships.)  While affectivists recognize the importance of affect, they provide the wrong kind of explanation for the affectiveness of blame in personal relationships. Wallace (2011) argues blame is necessarily affective because blame expresses that one cares about the relation of mutual recognition and regard, and emotion is a uniquely suitable way to express such caring. This argument explains the affectiveness of blame as a general, relationship-neutral feature of blame, with heavy reliance on a distant, minimal non-personal relationship (the “relation of mutual recognition and regard”). However, what is crucial to the affectiveness of blame in personal relationships seems to have a great deal to do with the nature of personal relationships. In other words, blame in personal relationships is and should be affective because of something specific to personal relationships—that they involve emotional vulnerability and particularity (to be explained further in the next section.) If this is correct, then the affectiveness of blame in personal relationships cannot be explained in terms of some general, relationship-neutral features blame, not least in terms of non-personal relationships.6   (1.4.2) The Richness of Blame Emotions in Personal Relationships Affectivists also fail to account for a second feature of blame in personal relationships: the richness of blame emotions in personal relationships. As I explained above, Wallace argues that blame emotions are emotions connected to holding the wrongdoer responsible for violating moral obligations, and that these emotions consist in moral anger, i.e. resentment and indignation. (Guilt is also a blame emotion, but it targets oneself.)7 This understanding of blame emotions oversimplifies personal relationships. A relationship as emotionally rich as personal relationships involves a set of blame emotions much richer than just moral anger. We normally do not blame our children, siblings, and 																																																																		6 Wallace’s argument, with some revision, might have more success as an argument for the affectiveness of blame in personal relationships rather than for the affectiveness of blame as a general feature. Personal relationships involve an emotional bond, so the kind of emotional vulnerability that Wallace’s argument alludes to is indeed a feature of blame in personal relationships. (I will make use of emotional vulnerability in the next section (2.2) to explain the affectiveness of blame in personal relationships.) However, as I argued above, non-personal relationships need not involve such emotional bond. Even if non-personal relationships involve emotional bonds, the intensity, depth, and significance of such emotional bonds are incomparable to those in personal relationships. It is therefore not clear that emotion is the only way to show that one cares about (non-personal) morality such that blame must be affective. Why are conative elements (e.g. intentions, dispositions, and expectation, as per Scanlon’s account of blame) not significant enough to show that one cares enough about (non-personal) morality? 7 It is of course questionable whether violations of moral obligations are the only things that we blame people for. We may blame people for their bad moral character, or blame people for not being “nice enough”, even though no obligations are at stake. This seems especially true in personal relationships. 	 20	lovers with only anger, but also with disappointment, sadness, and hurt feelings.8 (In fact, Strawson may have recognized that different kinds of relationships involve different kinds of reactive attitudes (as quoted at the opening of Introduction) and perhaps for this reason would admit of a richer set of blame emotions, such as “gratitude, resentment, forgiveness, love, and hurt feelings” (2008, 5).)   Consider a parent-child example. When a parent blames his/her child for lying, he/she would not only be angry with the child, but would also feel disappointed and hurt. Moreover, it may be normatively appropriate for the parent to be not only angry but also disappointed and hurt. As a good parent, he/she should not be only angry with the child for lying, but should also be disappointed in his/her dishonest child and be hurt by the child’s dishonesty. Blaming with only anger may reveal a lack of concern for his/her child or an inappropriate emphasis on being “respected” by the child.  Or consider love relationships. In one episode of the BBC television series Sherlock, Sherlock Holmes notices that Molly Hooper has prepared a specially-wrapped Christmas present, one that is very different from all of her other presents. Sherlock then makes the deduction, in front of everyone at the party, that this special present must be for someone she loves. He proceeds to open the card attached to the present, only to find out that the present is actually for him. Molly, who is deeply in love with Sherlock, is very embarrassed by this revelation. Does she resent Sherlock’s impertinence and lash out at him? No. Struggling to hold back her tears, she mumbles, “You always say such horrible things. Every time… Always… Always…” (McGuigan 2012). As I see it, Molly is more hurt than angry at Sherlock’s insensitivity and lack of care, and it is in her hurt feelings that her love of Sherlock is revealed.  In short, blame in personal relationships seems to have two features: (1) it is and should be affective and (2) it involves a richer set of emotions than blame in non-personal relationships. Both affectivist and non-affectivist accounts are relationship-neutral accounts and yet fail to account for these two features9. 																																																																		8 While Wallace implicitly acknowledges in a recent paper (2014) that disappointment, hurt feelings, and “sense of betrayal” are emotions that people are distinctively subject to when they blame one another in “thick social relationships” (137), he has not explained nor further discussed his departure from his earlier work or the relationship-dependency of such blame emotions. His acknowledgement is made also not in the context of an account of blame but an explanation of the bipolar nature of moral expectations and reactive attitudes. 9 A few blame theorists have proposed a different theory of blame—the communicative theory of blame (e.g. Fricker 2014, McKenna 2012, MacNamara 2015). This emerging theory takes blame as a form of moral communication and, to 	 21	 (1.5) The Nature of Blame vs Blame in Practice These two features of blame may not be present in some cases of blame in personal relationships, but they remain important features that are worth exploring. This brings up the second methodological point of this chapter. The explanatory work that I aim to do is not to explain the nature of blame (i.e. the necessary and sufficient conditions of blame); rather, I aim to explain features that are important to blame as a normative practice situated in interpersonal relationships, particularly in personal relationships.  Affectivists argue that the nature of blame lies in affect, whereas non-affectivists argue that it lies in some conative elements and so affect is not necessary for blame. The affectivist then challenges the non-affectivist that “affectless blame” is a conceptually oxymoron, and the non-affectivist counters that there are intelligible cases of affectless blame: one might have no emotional energy when blaming, or just happened not to feel anything10. The affectivist then reply with the subtle distinction between “blaming” and “holding accountable”, where the former requires affect but the latter does not. Or they may slightly revise their accounts such that blame requires either affect or, if the blamer does not feel any affect, the judgement that such affect would be appropriate. (See, for instance, Wallace 1994)   Whether or not such manoeuvres are viable, we need to ask a more fundamental question: how would knowing the nature of blame—features of blame that can be found in all (or almost all) cases of blame—help us understand blame as an interpersonal practice? As Miranda Fricker (2014) rightly observes, blame is a significantly disunified practice, so the nature of blame—the “highest common denominator”—may turn out to be very low, and the quest for it may leave out some distinctive or central features of blame that may not be present in all cases of blame and end up not telling us anything philosophically illuminating about blame (2). She then argues for a paradigm approach, which aims to find features of blame that can explain its diverse practices though they may not be present in all cases.   																																																																																																																																																																																																																																						the extent that communication is an important aspect in personal relationships, can explain some features crucial to blame in personal relationships. While this account, as developed by theorists, contains several residual elements from the Strawsonian account, it can potentially become an account that can properly explain blame in personal relationships if it recognizes other distinctive features of blame in personal relationships (namely its affectiveness and rich emotions). 10 Non-affectivists acknowledge that blame often involves affect (e.g. Sher 2006, 94). But their minds are fixed on seeking the necessary and sufficient conditions of them, hence affect loses its place in their accounts. 	 22	 I agree with Fricker’s diagnosis of blame and adopt an approach similar to hers. Given the diversity of blame, the nature of blame, if there is one, may not tell anything illuminating about what blame is like in the context of interpersonal relationships. It will be more illuminating if we focus on the role that blame plays in our relationships, namely, how it functions in and is shaped by our relationships. This explanatory work is an attempt to answer the question “what is blame?”, but it is guided by the broader question about the role of blame in interpersonal relationships, rather than guided by the necessary-and-sufficient-condition question. (The explanatory work that I aim to do in the coming two chapters on forgiveness is similar.) The two features of blame that I am going to explore in this chapter (the affectiveness and the rich emotions of blame in personal relationships) are neither necessary nor sufficient conditions of blame—there will be cases where we blame those close to us without affect and/or without rich emotions. But these features are nonetheless important to blame as an interpersonal practice, and they are important because they are linked to some crucial features that make personal relationships personal. This will become clear in the coming two sections.     (1.6) Summary I have been relying on our moral experience to back up these two features of blame in personal relationships, but I have left untouched the question of why blame in personal relationships has these two features. In the following two sections, I will attempt to answer this question. In section 2, I will argue that blame in personal relationships is and should be affective because of two constitutive features of blame: particularity and intimacy. In section 3, I will argue that blame involves a rich set of emotions because of three features of the wrongs distinctive in personal relationships.  	 23	Section 2 Blame It With Your Heart: The Affectiveness of Blame in Personal Relationships “Personal relationships” encompass a wide spectrum—parent-child relationships, siblings, friendships, lovers, and in some cases even neighbours, and colleagues. The discussion in this chapter does not intend to cover all personal relationships, but focusses only on some paradigmatic personal relationships, such as parent-child relationships, sibling, friendships, etc., arguing that blame in such paradigmatic personal relationships (i) is and should be affective and (ii) involves a rich set of emotions.  This section aims to discuss two features that partly constitute personal relationships—intimacy (2.1) and particularity (2.2)—and use them to explain the affectiveness of blame in (paradigmatic) personal relationships. It is therefore not giving a complete account of personal relationships, nor would it be necessary to achieve my aim. I am not arguing that any relationships with these two features are personal. We may sometimes be emotionally vulnerable to we barely know and see them as a particular person, and it would be ridiculous if that would make the relationship personal. However, intimacy and particularity are distinguishing features of personal relationships; they are partly what makes personal relationships personal and non-personal relationships normally do not have these features.   To strengthen and clarify my explanation of personal relationships, I will then briefly discuss two difficult cases, the seriousness of wrong and the gradability of relationships in (2.3), and reject the suggestion that personal relationships may be affectless in (2.4).   (2.1) Personal Relationships: Intimacy as Emotional Vulnerability In this and the next sub-sections, I use two constitutive features of personal relationships (intimacy and particularity) to explain (i) the descriptive features of blame in personal relationships—that it is often affective; and (ii) the normative features of personal relationships, and in turn the normative features of blame—that it should be affective.     	 24	This sub-section focusses on the first constitutive feature of personal relationships—intimacy. People standing in a personal relationship are intimate (or emotionally vulnerable) to one another and they respond to wrongs with emotion. This explains why blame in personal relationships is and should be affective.    (2.1.1) Intimacy as Emotional Vulnerability It sounds very intuitive if not trivial that intimacy is one feature that makes a relationship “personal”, but what does “intimacy” involve? Theorists have proposed various accounts. Hugh LaFollette (1996) and Laurence Thomas (1987, 1989, 1993) argue that intimacy involves “self-disclosure”—disclosing personal details about oneself. Marilyn Friedman (1993) argues that it involves a commitment to one another as unique persons. Sherman (1987) argues that it involves a “singleness of mind”. These accounts all describe some important features of intimacy in personal relationships, but for the present purposes, which are to explain the affectiveness of blame in personal relationships, my discussion will focus on one feature: intimacy involves exposing one’s personal self to another person.    When we are standing in a personal relationship with another person, we are pooling our lives with the person in various ways such that what the person does, says, thinks, and feels can affect us in profound ways. For instance, a parent’s retirement plans can be affected by which job her daughter chooses to accept. A sister may change her holiday plans because of what her brother thinks of her plans. A husband’s life decisions can be affected by whether his wife still loves him. Exposing ourselves to others and allowing various aspects of our lives to be influenced by them is one important way we relate to others closely and build intimate relationships with them.  Self-exposure is not limited to our practical lives, e.g. where to go for holiday or when to retire, but is also expressed in terms of emotional vulnerability. We care about what they do, say, think, and feel, and we care about these things in such a way that they can influence our emotion in profound ways. If a daughter accepts a job that promises a good career future, her parent would not merely coldly make life-plan adjustments, e.g. preparing for an early retirement, but would feel certain emotions, e.g. relief, excitement, pride, etc.   (2.1.2) Intimacy and the Affectiveness of Intimate Blame 	 25	This emotional vulnerability can explain why affectiveness is a psychologically typical feature of blame in personal relationships. We care about what our intimates do to us and say to us, and how they think of and feel about us, and we care about these things in such a way that, when our intimates do not properly treat us, feel about us, etc., we will respond with emotion.11 For instance, when a husband finds out that his wife no longer loves him, he would not coldly change his life plans, e.g. move out of the house which he has been sharing with his wife, but would feel certain emotions, e.g. feeling hurt, betrayed, despaired, etc. So long as we continue to see and treat the person as someone who stands in a personal relationships, rather than step outside the relationship (at least momentarily) and see and treat her as a non-intimate, we will not coldly throw down words such as “that’s wrong” or simply walk away from the relationship without any feelings. Instead, we will blame her with emotion.   Intimacy also explains why it is normatively appropriate that we blame those close to us affectively. Since intimacy/emotional vulnerability is a constitutive feature of personal relationships, blaming someone close to us affectively expresses that we are blaming her as an intimate and continue to stand in a close relationship with her, and the relationship gives us reason to do so to the extent that we are in it. Blaming without affect, on the contrary, amounts to distancing oneself from one’s intimate and (at least momentarily) stepping outside the close relationship. (This does not mean that we should never do so. Sometimes we may simply not have the emotional energy or resources to deal with the pain intimacy can bring us, and in such cases we may have good reason to close ourselves up (at least momentarily) in order to maintain our emotional well-being. Or the person may have seriously wronged us and the relationship is beyond repair, and the only option left for us is to pack and leave—both physically and emotionally.)   On the other hand, intimacy can explain why blame in non-personal relationships need not be affective. Emotional vulnerability is not a constitutive feature of non-personal relationships. We may be involved in normal social interactions with our colleagues, business partners, and fellow citizens—working and cooperating with them, or even seeing them quite often—without any emotional engagement. Even when a non-personal relationship involves emotional vulnerability, its intensity and depth will normally not be comparable to that in personal relationships. Nor will emotional 																																																																		11 See Shabo 2012a, who makes a similar point. 	 26	vulnerability have the same significance as that in personal relationships. Emotional vulnerability is not required for standing in a non-personal relationship nor for showing that we care enough about such relationships.12 Therefore, when we have been wronged non-personal relationships, we need not, though we may, be “emotionally exercised” and blame affectively.   Note that relating to others non-personally and blaming them without affect do not necessarily amount to the Strawsonian “objective attitude”. The objective attitude is an attitude that we adopt when we see a person not as a normal participant in interpersonal relationships but as an object of “social policy” or “treatment”, or as something to be “managed” or “handled” rather than to be related to. The objective attitude involves the withholding of moral demands that are normally held towards morally responsible agents. This is contrary to the participatory attitude, which we hold towards people we interact with and see as morally responsible agents in normal interpersonal relationships. (Strawson 2008, 9-10)  But standing in a non-personal relationship without emotional vulnerability and blaming without affect do not amount to the adoption of the objective attitude. As I argued in a footnote in (1.4.1), emotions do not have the same moral significance as they have in personal relationships such that blame in non-personal relationships does not require emotions. Indeed, when one judges that someone has done something blameworthy and then blames this person affectlessly (e.g. by becoming less ready to help him or deciding to rely on him less), one continues to see and treat him as a normal human being and a responsible moral agent, but “only as someone who has offended against moral demands”.  (2.2) Personal Relationships: Particularity While intimacy gives us a positive argument for the affectiveness of blame in personal relationships, particularity, a second constitutive feature of personal relationships, gives us a negative argument for the affectiveness of blame in personal relationships: it rules out one typical way to keep our emotion in check and blame affectlessly. When we want to keep our blame emotion in check, we typically try to avoid “taking things personally”, i.e. by seeing oneself, the wrongdoer, and the violation in a generalized way. However, in personal relationships, we see and should see our intimates as 																																																																		12 Hence, in section 1.4.1, I rejected Wallace’s argument that blame needs to be affective in order to show that one cares about morality. 	 27	particular persons. This typical way to blame affectlessly is therefore unavailable in personal relationships.   I will first explain particularity as a constitutive feature of personal relationships in (2.3.1), and then explain in (2.3.2) how this feature rules out “taking things non-personally” and give us one explanation why blame in personal relationships cannot and should not be affectless.   (2.2.1) Seeing Someone as a Particular Person There is a consensus among the few theorists who have proposed accounts of personal relationships that particularity is one defining feature of personal relationships. (See Marilyn Friedman 1993, Hugh LaFollette 1996, Paul Gilbert 1991, John Hardwig 1989) It seems quite intuitive that, in personal relationships, we see others as particular persons rather than as general, substitutable moral agents. This particularity, moreover, is reciprocal. We do not merely see those close to us as particular persons; we also see ourselves as particular persons standing in personal relationships with them.    Seeing someone as a particular person involves at least two things. First, if I see my friend Sarah as a particular person, it means that I would not think that another person, Laura, can substitute Sarah, even though Laura is in various ways very similar to Sarah, say having beautiful eyes, being athletic, intelligent, etc. In non-personal relationships, by contrast, we often see others as general moral agents substitutable by other people who can fulfil the same role, satisfy the same need, etc. We would not care much if the shopkeeper is Brian or Bruce, so long as the eggs he sells are fresh, and we would not care if the technician at the IT help desk is Melanie or Maria, so long as she helps me figure out how to use Canvas.  Second, and relatedly, seeing Sarah as a particular person also means that I do not see her as a mere instantiation of the generalized moral agent, but would focus on the particular person Sarah. For instance, I would see Sarah’s needs and well-being as Sarah’s particular needs and well-being, rather than as mere instances of the needs and well-being of a general moral agent.13    																																																																		13 Friedman makes this observation about focussing on someone as a particular person in her account of friendship (1993, 136;190). However, I believe that her observation applies not only to friendship but to personal relationships in general. 	 28	Personal relationships are therefore relationships where people see and relate to one another as particular individuals rather than as general, substitutable moral agents. And when one is standing in such a relationship with another person, one has reason to see and relate to the person as a particular individual rather than as a general, substitutable moral agent. Doing otherwise would amount to stepping out of the personal relationship and treating the person as a non-intimate, and the relationship gives one reason to refrain from doing so.   In non-personal relationships, by contrast, we normally do not see and relate to one another as particular individuals but as general, substitutable moral agents, nor do non-personal relationships require us to see and relate to one another as particular individuals. We may do so on some occasions, but non-personal relationships do not by their nature give us any specific reasons to do so. When impartiality is a concern, we may even have reason not to see and relate to other people as particular individuals because impartiality requires us to treat other people as morally equivalent.   (2.2.2) Particularity and Affective Blame in Personal Relationships The particularization in personal relationships gives us a negative argument for the affectiveness of blame in personal relationships: it rules out one typical way we employ to blame affectlessly. One typical way to keep our blame emotions under check is to create an emotional distance between us and the wrong by taking things “non-personally”, i.e. by seeing oneself, the wrongdoer, and the norm violated as generalized. Suppose that someone vandalized your front door. You are furious because you just painted the door a few weeks ago. However, you are on your way to an important interview, so you try to make yourself less angry by taking things non-personally. You may remind yourself that the vandal didn’t target you particularly. He may be looking for an easy target for vandalization, and your front door happened to be a golden opportunity, e.g. it was located at a quiet and dark corner of the street, no one was around, etc. He would have vandalized another door had he come across a better opportunity. You may also try to remind yourself that this vandal means nothing to you—he is just another bad guy with nothing better to do.14   																																																																		14 Taking things non-personally does not always work. We may be so worked up that we cannot create the emotional distance that we need. Think about road rage. Sometimes mere bad driving manners can make us very angry. Or the wrong may be so seriously that even taking things non-personally would not help. Imagine a reckless driver running over a few innocent pedestrians and fleeing the scene. He does not target any particular persons, but his entire lack of regard for other people would probably infuriate any normal person. 	 29	Or, for a literary example, think about Michael Corleone in The Godfather the movie. When discussing with his family his plan to kill Virgil Sollozzo, he declares that there is “nothing personal” with his plan but is “just business”. What he means is that he plans to kill Sollozzo not because he is angry that Sollozzo was behind the attempted assassination of his father Vito Corleone (the Godfather), but because he must kill Sollozzo in order to stop Sollozzo from trying to assassinate the Godfather again (Coppola 1972)).  But we cannot take things non-personally when we have been wronged in personal relationships, for three reasons. First, taking things non-personally involves seeing the wrongdoer as a general moral agent, but in personal relationships we see others as particular persons. When Sarah lies to her friend Hugh, Hugh would not say to himself “she is just another liar” or “she would have done it to anyone”. Whether or not Sarah intends to target Hugh particularly, Hugh would see her as targeting him as a particular person—because they are friends. (If Sarah really did not target Hugh particularly, Hugh would still see Sarah to have wronged him as a particular person, but the wrong that she had done to him would be slightly different. The wrong would be that it had not even crossed her mind that Hugh was her friend when she lied to him. This of course is a particularized wrong.)    Second, and relatedly, personal relationships give us reason not to take things non-personally because doing so amounts to (at least temporarily) stepping outside the relationship, and it gives us reason to refraining from doing so. We might sometimes have reason to step outside the relationship and adopt an outsider’s perspective, but this would require other justifying reasons that might not be available whenever we want to keep our blame emotion in check.   Third, non-personal blame may not be able to fully reflect the wrongfulness of offences in personal relationships. Lying to a friend, stealing from one’s parent, and refusing to support a sibling are fundamentally different from lying to, stealing from, and refusing to support a general moral agent. The nature and seriousness of the former three wrongs must be understood within personal relationships. (I will discuss wrongs in personal relationships in greater detail in the next section.)   We therefore cannot and should not take things non-personally in personal relationships. We have to see both the wrongdoer—a person close to us—and ourselves as particular persons. If we blame them accordingly, our blame will also be affective.  	 30	(2.3) Three Clarifications (2.3.1) Two Difficult Cases?  There are some cases where a non-personal relationship has these two features. Consider a long-term colleague. After years of working with this person, one might become emotionally vulnerable to them and see them as a particular person. Colleagues are not a kind of relationship that can be properly called personal, but they may have the two constitutive features, so it is not unclear if they are personal or non-personal according to my understanding of personal relationships.   As I stated in this opening of this section, I do not intend to give a complete account of personal relationships. I am not arguing that any relationship with these two features are personal. However, these two features can explain why long-term colleagues may be a marginally personal relationships. Colleagues are often borderline cases between personal and non-personal relationships; i.e. they are semi-personal relationships, where (some degree of) emotional vulnerability and particularity are possible. If long-term, colleagues may move towards the personal realm if not stepping into it. After years of working together, one may not see a long-term colleague as simply someone who merely fills a (professional) role or satisfying one’s (professional) needs (e.g. managing money, dealing with external relations, etc.), but may well see them as someone with some traits (e.g. patient, responsible, etc.) and know about other non-professional aspects of them (e.g. hobbies). One may also be emotionally vulnerable to how they treat us and whether they care about out interests and feelings, not merely for professional reasons but also for “personal” reasons. In such cases, the relationship seems to have stepped into the personal realm, if not have become partly a friendship. (It is indeed not uncommon to hear people say about their colleagues “we work well together and we are also good friends.”)  Another relationship may have the two constitutive features but does not seem to be a personal relationship is the relationship between a victim and the wrongdoer. A victim may become emotionally vulnerable to and see the wrongdoer as a particular person exactly because of the wrongdoing.   As I said above, I am not claiming that the two constitutive features entirely determine whether a relationship is personal. Moreover, my discussion focusses on paradigmatic relationships, such as parent-child relationships, friendship, colleagues, etc. Victim-wrongdoer relationships, while they 	 31	exist, are not a paradigmatic relationships in our social interactions; instead, they exist alongside paradigmatic relationships—the wrongdoer may be one’s friend, or neighbour, or a stranger.   But I believe that the two constitutive features can partly explain why victim-wrongdoer relationships are different from paradigmatic personal relationships. First, whether wrongs create emotional vulnerability may largely depend on the severity of the wrong. One can imagine that a victim will become emotionally vulnerable to the wrongdoer if the wrong was a traumatic experience. But if the wrong is relatively insignificant, say one’s car being dented by a careless driver, or one has been misled by an irresponsible business partner, the victim may not become emotionally vulnerable to the wrongdoer. Moreover, if the background (paradigmatic) relationship is non-personal (e.g. one had never met the drive or the business partner is one-off), the relationship does not give one reason to be emotionally vulnerable to the wrongdoer.   Second, even if the victim comes to see the wrongdoer as a particular person because of the wrong, the particularization may be limited to aspects that are related only to the wrong. But in paradigmatic personal relationships, by contrast, particularization involves multiple aspects, e.g. our friend Brian is patient to his children, generous to his friends, enjoy playing foosball after work, etc.   (2.3.2) Affectiveness and Seriousness My argument in this section—that blame in personal relationships is and should be affective—does not deny that there may be other factors determining the affectiveness of blame. One such factor is the seriousness of the wrong. When serious wrongs have been committed, say betrayal, domestic abuse, racism, and violations of basic human rights, we would blame affectively and would have reason to do so whether the relationship is personal or non-personal. Similarly, when the wrong that we experience is trivial, say telling a small and inconsequential lie, being a bit late for a meeting, littering, and talking loudly on the phone on the bus, we would likely and have reason to blame without any affect, whether the relationship is personal or non-personal. My argument is merely that the nature of the background relationship is one determining factor of the affectiveness of blame, without denying that there may be other determining factors.   However, I believe that this factor—the nature of the background relationship—is more important than the seriousness factor in the context of interpersonal relationships, particularly personal ones. 	 32	Serious wrongs do take place, unfortunately, and they take place in both personal and non-personal relationships. While we should acknowledge the existence of the serious wrongs in our lives and not trivialize their graveness, we should not overdramatize our lives and focus so much of our attention on serious wrongs. The majority of the wrongs that we encounter and that concern us in our interpersonal relationships, particularly personal ones, are neither serious nor trivial, such as breaking a promise, saying something hurtful, or being negligent. In such cases where the wrong is neither serious nor trivial, the nature of the background relationship plays a larger role than the seriousness of the wrong in determining the affectiveness of our blame. The nature of background relationship should therefore have a larger explanatory role in understanding blame in our moral lives, even though there will be cases where seriousness will determine the affectiveness of blame.   (2.3.3) Personal Relationships: Degree of Closeness My discussion of personal relationships intends to explain the affectiveness of blame in paradigm personal and non-personal relationships, so it might have made it sound as though relationships were binary—that they are either personal or non-personal. However, whether a relationship is personal or non-personal is a matter of degree. Though I do not intend to give a complete account of personal relationships, I will briefly explain how my understanding of personal relationships is consistent with the gradability of relationships.   Relationships can be more or less personal/non-personal, determined by the degrees of intimacy and particularity. The more the parties involved are exposed to one another (e.g. how frequently, how intensely, and the extent to which their lives are in general mutually exposed), and the more extensively the parties see one another as particular, non-substitutable individuals, the more personal their relationship is.   This gradability of relationships can explain two things. First, because relationships can be more or less personal, blame can also be more or less affective (i.e. how intensely affective and how frequently it is affective), depending on how intimate and particularized the relationship is.   Second, there are liminal relationships that lie between personal and non-personal; that is, “semi-personal relationships”. Examples of “semi-personal relationships” include distant relatives, acquaintances, neighbours, long-term business partners, long-term employer-employee relationships. 	 33	On the one hand, these are relationships where there are certain degrees of intimacy and particularity, but their levels are not high enough to qualify the relationships as fully personal; on the other hand, those standing in these relationships to an extent relate to one another as someone who fills a role or satisfies some need. The level of affectiveness of blame in semi-personal relationships therefore lies somewhere between personal and non-personal relationships.    Semi-personal relationships can help reject one possible objection to my argument. I argue that we blame affectively in personal relationships but affectlessly in non-personal relationships. One might object that we sometimes blame affectlessly in relationships that seem personal, such as distant relatives or neighbours, and we sometimes blame affectively in relationships that seem non-personal, such as long-term business partners or colleagues. But these relationships may in fact be neither personal nor non-personal; they are semi-personal relationships. The affectiveness of blame in such relationships lies somewhere between that in personal and in non-personal relationships, and blame in such relationships cannot be sharply categorized as affective or affectless.   (2.4) Objections: Affectless Personal Relationships?  I argued above that personal relationships are necessarily affective and hence blame in personal relationships are typically and appropriately affective. Some, however, might maintain that personal relationships need not be affective. They might argue that some people simply do not have any feelings towards those close to them, and there is no reason to rule out their relationships as not genuinely personal. Though this position isn’t common, I will briefly discuss three points that counter it. Doing so will strengthen and clarify my claim that genuine personal relationships are affective.  (2.4.1) Conceptual Difficulty with Affectless Personal Relationships I want to start with presenting the conceptual difficulty with affectless personal relationships. “Affectless personal relationships” are supposedly personal relationships where emotions and susceptibility to emotion are entirely lacking. Take friendship as an example. The “friend” would not feel any emotion, conscious or unconscious, towards his “friend”. The relationship would be entirely motivated by duties and expressed only in terms of behaviour, dispositions, thoughts, etc. For instance, when his friend was sick in the hospital, his visit would not be motivated by emotion but entirely by a sense of duty. Note that it is not merely that there would not be individual situations where this friend did not experience emotion. Instead, the friendship is one where ongoing emotion 	 34	and susceptibility to emotion are lacking. We might then ask: is this relationship, normatively speaking, a “genuine” friendship?   Such a completely affectless friendship would strike most people as strange and hard to comprehend because it is so different from our experience of personal relationships. It is uncontroversial that most of our personal relationships are to various degrees affective. We have various ongoing feelings towards our close ones, we are susceptible to a range of feelings when they do or say certain things to us and when we do and say things to them. Bar special circumstances, say depression or exhaustion, a personal relationship without any emotion is so far removed from our normal moral experience that it would be difficult for us to conceptualize what such a relationship would be like, not least to normatively adjudicate whether it is a genuine friendship.   (2.4.2) Completely Affectless Personal Relationships Are Very Rare  Even if we could conceptualize affectless personal relationships and did judge that they were, normatively speaking, “genuine” personal relationships, they would at best be empirically rare and therefore lack theoretical and practical significance. Note that by “affectless personal relationships”, I do not mean a relationship where there is a lower-than-normal level of emotion. Someone who is prone to a lower-than-normal level of emotion would still experience emotion; she merely does not get excited easily and does not experience intense episodic emotions. So when her friend is very sick in the hospital, she would experience a low level of sadness, anxiety, etc., but she would be able to eat, sleep, and work.   Such are relationships with “low affect” or “reduced affect”, and they may be what we refer to when we speak of “affectless personal relationships”.15 But such relationships aren’t literally affectless. An affectless relationship is one where the persons involved would not feel any emotions, conscious or unconscious, in the relationship. This complete lack of emotion is very rare and may be found only in 																																																																		15 We may be able to learn something about reduced-affect personal relationships from psychologists. Psychologists recognize two psychological conditions related to reduced emotions. One is “reduced affect display”. Patients with reduced affect display do experience emotions, but are unable to express their emotions verbally or non-verbally. The other psychological condition is “Alexisthymia”. Again, patients with Alexisthymia do not completely lack emotions. They are merely unable to express and identify the emotions that they are experiencing, and sometimes they are unconscious of their emotions. Some patients, however, will experience emotion-related bodily sensations, e.g. muscle tensions. This is an indication that they are not completely insusceptible to emotion. This does not mean that “cold fish” all have these or other similar psychological conditions, but these psychological conditions may explain some reduced-affect personal relationships. 	 35	very special circumstances. This gives us reason to doubt the theoretical and practical significance of such relationships in our accounts of personal relationships.   (2.4.3) Distinguishing Acts and Background Relationships One might give the following argument in support of affectless personal relationships. Genuine acts of friendship can be motivated solely by a sense of duty (e.g. visiting a sick friend in the hospital). This is because duties are an essential part of friendship, and there are times where we just don’t “feel” the right emotions—we may be too tired or too busy. But it does not seem right to say that such acts motivated by a sense of duty would not be genuine acts of friendship at all. They may be an inferior form of friendly act, but they are friendly acts after all. If affectless acts of friendship are genuine acts of friendship, then why aren’t affectless friendships likewise genuine friendships? They may be inferior, but they are a genuine form of friendship after all.   But it is important to clearly distinguish between acts of personal relationships and background relationships. While I agree that the example given in the above argument—visiting your sick friend solely from a sense of duty—may be a friendly act, it may constitute a genuine friendly act only if it takes place within an affective friendship. If, instead, the affectless visit takes place in an affectless relationship where there is no general susceptibility to affect, I suspect that few of us would still think that the affectless visit is a genuine act of friendship. In short, the above argument for affectless personal relationships fails to recognize the background relationship, and thus cannot  on its own establish the claim that personal relationships may be affectless.   (2.5) Summary  The main argument of this section is that the two constitutive features of personal relationships—intimacy and particularity—can explain the first feature of blame in personal relationship: that it is often and should be affective. In the next section, I will move on to the second feature of blame in personal relationships: that it involves a rich set of emotions.  	 36	Section 3 Blame It Colourfully: The Richness of Blame Emotions in Personal Relationships  This section discusses the second feature of blame in personal relationships, that blame in personal relationships involves a richer set of emotions than blame in non-personal relationships.   Strawson proposes that studies of moral responsibility and blame should begin with our interpersonal relationships and see how we interact with one another in our moral lives. He then observes that we respond to wrongs with what he calls the “reactive attitudes” and argues that moral responsibility should be understood in terms of them. Strawsonian/affectivist theorists then propose that blame consists in the reactive attitudes, but they limit them reactive attitudes to moral anger, arguing only resentment and indignation are relevant to (other-)blame (whereas guilt is relevant to self-blame).   But Strawson in fact recognizes that there is a “whole continuum of reactive attitudes” and rightly observes that reactive attitudes vary widely with the relationship context. If we do adopt Strawson’s approach, taking our moral lives as the starting point of our investigation of blame, we should recognize that the richness of blame emotions in personal relationships and expand the range of reactive attitudes relevant to blame.   This section aims to dig a little deeper and gives three reasons to explain why blame in personal relationships involves a rich set of emotions—not only moral anger, but also disappointment, sadness, and hurt feelings. First, there are expectations that are distinctive to personal relationships and whose violations trigger non-anger affective responses. I will discuss two: care and loyalty (3.1). Second, wrongs often damage personal relationships and in turn intimacy, triggering hurt feelings (3.2). Third, people standing in personal relationships often have concern for one another and will blame one another for their moral character; disappointment is a characteristic emotion that blame of moral character triggers (3.3). I will end this section by addressing two possible objections against the admission of non-anger emotions, such as disappointment, as blame emotions (3.4).     	 37	(3.1) Two Distinctive Expectations in Personal Relationships Personal and non-personal relationships involve different expectations. Some expectations that are specific to personal relationships can explain why blame in personal relationships involves a richer set of emotions than in non-personal relationships. I will discuss two in this sub-section: care and loyalty.   (3.1.1) Care Care is one expectation that is central to personal relationships. When our friends, lovers, or siblings show a lack of care, our affective response would not be just anger but also hurt feelings if not disappointment as well.   To see how care is distinctive to personal relationships, consider the following pair of examples. Suppose you see stranger falling ill on the street. Even though you don’t know her, you may have good reason to provide necessary assistance: ensuring her safety, calling her family or an ambulance if necessary, etc. Helping her in such ways is a way to respect her as a person because respect requires minimal beneficence.   Now suppose that the ill person is in fact your friend. As her friend, you should not merely call an ambulance and then leave; instead, you may have good reason to go to the hospital with her and visit her the next day, or driving her home if she doesn’t want to go to the hospital, and so on. Not only should you do more in this case than in the previous one, the reason you have to help her is also different—you should help her not because you should show her minimal “respect” but because you should show her special care.   Or consider another pair of examples. We have good reason to use a borrowed fishing rod with caution and avoid damaging it, but the relevant reason is dependent on the relationship. If the fishing rod is borrowed from the local fishing club, then we would be “negligent” if we have no thoughts about the value of the fishing rod and end up damaging it. But if the fishing rod belongs to our brother, we would not be only negligent but also “uncaring”, uncaring of the importance and the value of the fishing rod to him.  One might argue that care is relevant to non-personal relationships: it makes sense to say that not helping a stranger in distress shows not only lack of respect but also lack of care; the only difference 	 38	between personal and non-personal relationships in this regard is that non-personal relationships require a lesser degree of care (e.g. calling an ambulance rather than going to the hospital with the stranger). But even though this is right, care remains an expectation that is more significant in personal than in non-personal relationships. We are more concerned about whether our friend care about us and our brother cares about the value of our fishing rod, and hence we blame them for lack of care, whereas we are more concerned about whether a stranger shows us minimal respect and our neighbour respects our property right over our fishing rod, and hence we blame them on such grounds. Because our affective response to lack of care is not just anger but also hurt feelings and other emotions, and because care is specific to (or at least more significant in) personal relationships, blame in personal relationships involves a richer set of emotions than in non-personal relationships.   (3.1.2) Loyalty Another expectation distinctive to personal relationships is loyalty, and its violations—betrayal—will trigger not just anger but also disappointment and hurt feelings. We should be loyal to those close to us by showing firm and constant support for them even in adversity. For instance, a good parent should stand by her children when they are being accused of wrongdoing by others and provide moral, financial, and other support for the accused children, even when the accusation is grounded. A stranger or a colleague, by contrast, may not have any reason to do so in such situations.   We have reason to support and not to take advantage others in both personal and non-personal relationships, but they amount to loyalty only in personal relationships. If a colleague tells your boss some compromising information about your past in order to get you fired, she is “badmouthing”. But if it is a friend who does this to you, she is “back-stabbing” and betraying you. Similarly, lying to your neighbour in order to sell him an insurance policy is “dishonest” and “deceptive”; lying to your parent in order to sell him an insurance policy is not only dishonest but also an act of betrayal16.      																																																																		16 Some may argue that “betrayal” is an emotion. We do sometimes say that we “feel betrayed”, and betrayal seems to involve a distinctive feeling that may not be fully described in terms of other emotions. I do not intend, nor have the space, to give a complete account of betrayal, so I won’t discuss whether betrayal is an emotion. It is sufficient for the present purposes to observe that betrayal involves a complex set of emotions, a set that involves more than anger. 	 39	(3.2) Damage to Intimacy The second reason draws on a point that I made in section 2—intimacy—but with a different focus. The rich blame emotions in personal relationships can be explained by the very fact that the background relationship is an intimate one.   Scanlon (2008) argues that wrong acts damage the relationship between the wrongdoer and the victim. Whether or not this applies to all relationships, his claim highlights an important aspect of wrongs in personal relationships. When our close one wrongs us, say a friend was very late to a meeting and did not apologize, our complaint to her may be “do you still see me as your friend?” This complaint shows us that when our friend wrongs us, her wrongful act may strain or damage the relationship, or it may indicate that she no longer sees me as her friend (or as good a friend).   If wrongs in personal relationships strain, damage, or indicate that the relationship is damaged, we can see why we blame our friends with disappointment and hurt feelings. The relationship that such wrongs damage is not a thin and nominal one where two people stand to one another as moral agents. The relationship is much closer to us and more important to our lives. It is a relationship where we share our well-being, identity, and more generally our lives with another person. Damages to such relationships where such things are at stake will not trigger merely anger, but often also hurt feelings.   (3.3) Concern and Moral-Aretaic Blame The second reason that blame in personal relationships involves a richer set of emotions is that we, out of concern for our intimates, blame them for their moral character.   In personal relationships, we do not blame our intimates only for wronging us; we also blame them for the bad moral character that their wronging us reveals. If someone embezzles money from her family business, her parent would not only blame her for wronging him and the family, but would also blame her for being a dishonest and disloyal daughter. If a husband batters his wife, she would not only blame him for injuring her, but would also blame him for being a violent husband. If someone is too lazy to work and always asks his friends for financial assistance, his friends will not only blame him for asking too much from them, but will also blame him for being lazy and irresponsible.   	 40	Why do we blame our close ones on “moral-aretaic” grounds? One reason is concern. This involves the concern for their happiness, well-being, and interests. It also involves the concern for their moral well-being: we are concerned about whether they act morally and are morally upright. We have reason to show moral concern to one another in personal relationships, particularly in those that involve moral education and development, e.g. parent-child relationships, teacher-student relationships, and perhaps Aristotle’s virtue friendship.    Some might find moral-aretaic blame strangely detached and unrealistic. They might argue that moral-aretaic blame requires taking up a perspective of an outsider or a “moral referee”, but not a perspective of a non-detached participant in the moral community. This objection overlooks the relationship context of blame. Moral-aretaic blame may be out of place in non-personal relationships, but it is typical and appropriate in personal relationships. When we have been wronged in non-personal relationships, we normally blame the wrongdoer for wronging us, rather than blame her for the bad moral character that the wrong reveals. This may be because in non-personal relationships people are typically more concerned about their own interests, their claims to others, and whether others have violated their claims (or vice versa). But moral-aretaic blame is by no means out of place in personal relationships. We are and have reason to be concerned about the moral well-being of those close to us, and therefore we do and have reason to blame them on moral-aretaic grounds.    Moral-aretaic blame cannot on its own explain the rich blame emotions in personal relationships, but can do so in combination with moral concern in personal relationships. We may blame a hypocritical, dishonest, and racist politician with nothing but anger (it is very easy to find examples these days), and we may even adopt a detached attitude and blame the recklessness of Napoleon with no emotions at all. But moral-aretaic blame in personal relationships is different: it is motivated by concern for intimates. For instance, it is out of concern for his daughter’s moral character that the parent blames his dishonest daughter, and out of concern for her husband moral character that the wife blames her husband for his violence. Their concern explains the rich emotions of their blame: they blame not only with anger, but also with disappointment and hurt feelings.17 																																																																		17 Theorists in the blame literature focus overwhelmingly on the deontic. Some have even rejected aretaic blame outright (e.g. Wallace 1994). This may have led us to believe that blame deals only with the deontic. This is not true. As our moral experience attests, and some theorists (e.g. Fischer and Tognazzini 2011, Paul Russell 2013) observe, blame has as much to do with the deontic as with the aretaic judgements, and emotions triggered by the aretaic fully qualify as blame emotions. 	 41	(3.4) Two Objections  I will now address two possible objections to my expansion of blame emotions in personal relationships.   First, some might argue that non-angry emotions, such disappointment, sadness, and hurt feelings, are not proper blame emotions in that they express the Strawsonian objective attitude. The objective attitude is an attitude that we adopt when we see a person not as a normal participant in interpersonal relationships but as an object of “social policy” or “treatment”, or as something to be “managed” or “handled” rather than to be related to. (Strawson 2008, 9-10) In other words, we are not treating the person as a morally responsible agent when we adopt the objective attitude and we withhold from him/her the moral demand that he/she treat us properly. “Blaming” someone with non-angry emotions such as disappointment and sadness, the objection goes, amounts to adopting the objective attitude and therefore does not count as blame—because blame presupposes that the person is morally responsible agent. Such emotions might even be condescending—such as the sadness we feel towards a sadistic gangster who became a gangster because he came from a broken family and grew up in a rough neighbourhood.  Disappointment, sadness, and hurt feelings may but need not involve the objective attitude. As David Goldman (2014) observes, we may maintain our moral demands to the wrongdoer when we are disappointed in but not angry with him (10). A more appropriate approach is therefore to limit the expansion of blame emotions only to those forms of disappointment, sadness, and hurt feelings that regard the wrongdoer as a morally responsible agent and hold him/her to moral demands.18 Such restricted forms of non-angry emotions would not express the objective attitude.   The second objection against my admission of disappointment, sadness, and hurt feelings as blame emotions is that these non-angry emotions do not satisfy the constitutive-connection criterion. Some theorists might argue that, for any emotion X, X is a blame emotion if and only if X is an emotion that is triggered by and only by violations of moral expectations. Non-angry emotions such as disappointment, sadness, and hurt feelings are therefore not blame emotions because these emotions may be triggered by things that have nothing to do with moral expectations. For instance, one may be 																																																																		18 Goldman calls such restricted forms of emotion “sharpened form”, a term that he borrows from Justin D’Arms and Daniel Jacobson (2003). 	 42	disappointed when one does not win a lottery or sad that one’s child loses a football game. (See also Wallace 1994, 31)  There are theorists who think that non-angry emotions may satisfy this constitutive-connection criterion. Wallace, for instance, has implicitly acknowledged in a recent article (2014) that disappointment and hurt feelings are constitutively connected to violation of moral expectations, and thereby no longer holds his earlier stance that proper blame emotions are limited to moral anger (1994, 2011). He maintains, however, that sadness does not satisfy this criterion and is hence not a proper blame emotion (2014, 132 n9).  It is unclear why Wallace thinks that disappointment and hurt feelings are but sadness is not constitutively connected to violations of moral expectations. But it may be more fruitful to critically assess the constitutive-connection criterion. I argue that this criterion is too strong. The criterion holds that whether an emotion is a blame emotion is determined on the type level: this emotion must be, as a type, triggered only by violations of moral expectations. In other words, all tokens of this emotion-type must be triggered by violations of moral expectations. If some tokens of an emotion-type may be triggered by anything other than violations of moral expectations, then the entire type of emotion will fail be a blame emotion—even the tokens which are triggered by violations will not be blame. Sadness, for instance, is emotion-type where some of its tokens may be triggered by non-moral matters (e.g. losing a football game); therefore, sadness, as a type, is entirely not a blame emotion.   Why is it that, just because some tokens of an emotion-type may be triggered by non-moral matters, all other tokens of the same emotion-type are also rendered not blame emotions, even if these other tokens are indeed triggered by moral violations? Isn’t it more reasonable to say that an emotion-token is a blame emotion if it is triggered by moral violations, and not a blame emotion if it is triggered by non-moral violations? In other words, why shouldn’t the criterion be set on the token level rather than on the type level? Setting the criterion on the token level allows us to say that non-angry emotions are blame emotions when such non-angry emotions (tokens) are triggered by moral violations. Note that affectivists have already adopted a similar approach: their accounts do not say that “anger” is a blame emotion; they say that “moral anger” (or more specifically “resentment” and “indignation”, which are anger triggered by moral violations) is a blame emotion.  	 43	(3.5) Summary I have given three explanations for the richness of blame emotions in personal relationships. The first one deals with the distinctive quality of expectations in personal relationships, the second one deals with the damage to intimacy that wrongs may cause, and the third one deals with the concern we have for those close to us. I have also addressed two objections against the admission of non-angry emotions as blame emotions.  	 44	Section 4 Two Underlying Flaws  Let me reiterate the two main goals of this chapter. First, it aims to explore what blame is like in personal relationships as compared with that in non-personal relationships, and to ultimately show that accounts in the blame literature have failed to adequately explain blame in personal relationships. Sections 2 and 3 have achieved this first goal.   This section and the next move on to the second goal. I propose that we should give up the relationship-neutral approach to blame and adopt a relationship-dependent approach to blame—a “Who-are-you-to-me approach”. I argue in this section that there are two underlying flaws that give rise to the failure of theaccounts in the literature to adequately explain blame in personal relationships: insensitivity to relationship contexts and insufficient attention to personal relationships.   Taking a relationship-neutral approach to blame, blame theorists aim to explain features of blame that can be found in all relationships. But in their quest for a relationship-neutral account of blame, they pay little attention to relationship contexts, assuming instead that the features of blame that they find in their moral experience (e.g. affectiveness) are relationship-independent. Their theorizing often starts with an observation that blame in a specific relationship context has a certain feature, and then quickly generalizes the observation to all relationship contexts, concluding that blame in all relationships has that feature. But the feature may in fact be specific only to the relationship context in which the original observation was made. Moreover, because theorists tend to privilege non-personal relationships, the features that theorists generalize are often features of blame in non-personal relationships. The resulting accounts, which are presented to be relationship-neutral accounts that are to explain blame in all relationship contexts, are in truth modelled on non-personal relationships.    The combination of these two underlying flaws of the major accounts of blame can explain why blame theorists fail to adequately account for blame in personal relationships. In (4.1) and (4.2), I will explain in detail these two underlying flaws respectively with textual evidence from the accounts of blame that I have discussed. I conclude in (4.3) that we should adopt a relationship-dependent approach as an alternative approach to blame.   	 45	(4.1) Insensitivity to Relationship Contexts As I explained in section 1, blame theorists have taken a relationship-neutral approach to blame, aiming to propose accounts of blame that explain features of blame that can be found in all relationship contexts, overlooking features that may be found only in some specific relationships. This relationship-neutral approach is associated with two flaws of their accounts, which can explain their failure to adequately explain blame in personal relationships. The first of these flaws is their insensitivity to relationship contexts.   Theorists’ accounts of blame are insensitive to relationship contexts in various ways. Wallace argues that blame is necessarily affective and involves only moral anger. But he fails to recognize that both the affectiveness of blame and the variety of blame emotions are dependent on the relationship context.19 Affect is a feature of blame more prominent in personal relationships than in non-personal ones, and blame in personal relationships may involve not only anger but also disappointment, hurt feelings, and sadness.   Theorists may also be insensitive to relationship contexts if they take their moral experience of blame in specific relationships as representing the nature of blame in all relationships, not realizing that their experience may reflect only what blame is like in those specific relationships. We see this insensitivity in Sher’s account. One of his reasons for rejecting the affectivist account is that there are cases where blame may not involve affect. He observes that “[w]e simply do not have the emotional resources to muster even a twinge of hostility toward each of the innumerable miscreants, scoundrels, and thugs…” (2006, 89). While Sher’s observation is accurate, it is an observation about blame in non-personal relationships (e.g. criminals). Not realizing the importance of the relationship contexts of his examples, he generalizes his observation to all relationships and draws the hasty conclusion that blame need not be affective in any relationship contexts.20  																																																																		19 As I wrote in a footnote in section 1.4.2, Wallace now seems to recognize disappointment, hurt feelings, and “sense of betrayal” as blame emotions in personal relationships in a recent paper (2014, 137). However, he has not explained these expanded range of blame emotions or the relationship-dependency of such blame emotions. The context of his discussion in the paper—the bipolar nature of moral expectations and reactive attitudes—may also make one doubt how committed he is to the expansion of blame emotions to non-anger. 20 Some might argue that there aren’t any “relationships” between strangers, even minimal ones. I think it is not unreasonable to say that we stand in non-personal, minimal moral relationships with strangers, including those whom we have never and will never met, and those who flout basic moral requirements. However, my claims that blame in personal relationships has some distinctive features and that blame is dependent on the relationship context would still hold even if such minimal relationships did not exist. For the former claim, even if no such minimal relationships existed between strangers, there would be other non-personal relationships (e.g. business associates) to which I could still compare blame 	 46	Scanlon’s non-affectivist account fares better to an extent than other accounts in regard to its sensitivity to relationship contexts. On his account, blame is the modification of attitudes and dispositions towards the wrongdoer, and what modifications of attitudes and dispositions are involved in blame is dependent on the relationship. Blame therefore varies according to the background relationship (2008, 153).   While I agree with Scanlon that blame varies according to the background relationship, his account still fails to recognize the variability of blame in different relationships. By saying that blame, generally, relationship-neutrally speaking, need not involve affect, his account gives a misleading picture that blame, whether in personal or non-personal relationships, need not involve affect. He also maintains that blame in personal relationships (e.g. friendship) may involve no affect (136), failing to recognize the relationship dependency of blame in regard to affect.   (4.2) Insufficient Attention to Personal Relationships The other underlying flaw with the dominant accounts of blame in the literature is their insufficient attention to personal relationships. Or, as the other side of the same coin, their privileging non-personal relationships. This inattention leads theorists to propose relationship-neutral accounts of blame that skew towards non-personal relationships. Theorists present their accounts of blame as ones that meant to explain blame in all relationship contexts, when in truth their accounts are modelled on blame in non-personal relationships and can at best give a small and partial picture of blame in personal relationships. Worse yet, their accounts of blame may crowd out personal relationships and treat blame in such relationships as an inferior form of blame, while privileging non-personal relationships as the norm of moral relations and blame in such relationships as the paradigmatic form of blame.   There is plenty of evidence that Wallace’s account (1994) privileges non-personal relationships. First, his account adopts a narrow interpretation of blame, limiting blame to be a response only to violations of moral obligations. It therefore deals with moral requirements such as helping people in distress (37), avoiding causing unnecessary suffering, justice (63), respecting rights, keeping 																																																																																																																																																																																																																																						in personal relationships and show its distinctive features. For the latter claim, blame would still vary according to the relationship context, only that, in cases of strangers, the relationship context would be that there is no relationship. This relationship context—no relationship—would still make blame different from other relationship contexts, e.g. in personal relationships. 	 47	promises (64), etc., but it does not deal with moral expectations as kindness, consideration, supererogatory acts and virtues (37). He justifies this narrow scope based on his observation that our judgements of moral responsibility take place “primarily” in cases where moral obligations have been violated (1994, 63-64). This narrow scope of blame in Wallace’s account has been subject to a number of criticisms. (See, for example, Russell (2013)), who argues that Wallace’s narrow account oversimplifies our practices of blame and denies us of the apparatus to challenge them (191).) For our purposes, this narrow scope displays an inattention to personal relationships. It rules out plenty of moral expectations which are central to personal relationships; for instance, kindness, care, thoughtfulness, and sense of responsibility. It also rules out blame on aretaic grounds, which are often grounds for blame in personal relationships. We may blame our child for being lazy, or our spouse for being irresponsible, or our friend for being reckless.   Second, Wallace’s account also ignores special obligations in personal relationships. The moral obligations that Wallace discusses are primarily general moral obligations which belong to non-personal relationships: rights, keeping promises, helping people in distress, the principle of non-maleficence, and justice. Even when Wallace discusses cases where the background relationship is personal, the obligations at stake are not specific to the relationship but are general obligations. For instance, negligently rolling over a child and smothering one’s child while asleep (140), keeping a promise made to one’s sister (22) or one’s friend (141, 144).21  Third, the cases that Wallace uses to illustrate his account are predominantly non-personal: treading on the hands of a thief (137) and an unknown baby (138), agreeing to take care of an unknown baby (139), a bank teller being forced to hand over money at work (143), a plane crash victims stealing food (143), a soldier being ordered to kill innocent civilians (160).   In a recent paper (2014), Wallace gives a more detailed discussion of special obligations and blame in personal relationships. He recognizes that special obligations may be our “primary model of relational requirements” (137) and that personal relationships (and semi-personal relationships) are the 																																																																		21 Some theorists maintain that the moral expectations feature in personal relationships should not be construed in terms of obligations at all (e.g. Hardwig (1989), LaFollette (1996)), so they might argue that Wallace’s account is fundamentally unsuitable for blame in personal relationships.  I do not subscribe to this strong position, nor is it necessary for establishing my claim that Wallace’s account privileges the non-personal. As I see it, moral obligations can at least partly account for the interpersonal morality of personal relationships in term of “special obligations”. 	 48	“paradigmatic contexts” for the exposure to blame emotions (138). This shift in attention from non-personal to personal relationships is a good start, but it does not signal a sufficient attention to the relationship dependency of blame (that is, a relationship approach to blame, which I propose in this chapter). He does not acknowledge the connection between the affectiveness of blame and relationship contexts, but maintains that blame is and should be essentially affective regardless of the relationship context. Note also that Wallace’s discussion is not situated in the context of an account of blame, but in the context of an explanation of the bipolar nature22 of moral requirements and reactive attitudes, and personal relationships are “paradigmatic” only in the sense that it is in such relationships that the bipolarity of moral obligations is most clearly revealed. His discussion is only “by extension” related to the practice of holding people morally accountable (i.e. blame) (2014, 120).  Sher, another major figure in the blame literature, has also given insufficient attention to personal relationships and privileged non-personal relationships in his discussion of blame. First, while he recognizes that blame in personal relationships is more often affective than blame in non-personal relationships (through the example of his daughter) (2006, 88-89), his overall discussion of personal relationships is very limited. Second, he fails to recognize the richness of blame emotions in personal relationships. From the observation that he does not always detect the “flicker of annoyance” when he blames his daughter, he argues that affect is not an essential feature of blame. What he fails to consider is that his blame may involve other emotions, such as disappointment and hurt feelings, which are common and appropriate blame emotions in personal relationships. Third, his account of blame is situated against the backdrop of a morality that is “universal” (2006, 125), which may not square well with the moral expectations in personal relationships.   While I do not want to read too much into Sher’s account of blame and claim that it is an account of blame modelled largely on non-personal relationships, it seems fair to say that his account does not give sufficient attention to personal relationships.   Scanlon’s account is once again a special case. While he states that blame varies according to the background relationship, and he has indeed discussed some cases of blame in personal relationships 																																																																		22 According to the theory of bipolar obligation, moral obligations should be accounted for in terms of one person owing them to another person. For instance, if Sally is obligated to pick up her father at the airport because she has promised him to do so, then she owes it to her father to pick him up at the airport; and she will wrong her father if she fails to do so, rather than simply do a wrong. 	 49	(e.g. friendship and parent-child relationships), the content of his discussion may give one the impression that he is more concerned about non-personal than personal relationships. Recall his example of Joe, who revealed some secret about his friend Tim. Scanlon argues that Tim may blame Joe by coldly walking off the friendship, and uses this as support for his argument that blame may not involve any affect. But as I argue, it is rare that blame in friendship would not be affective because what friends do to us “touches our hearts”. As Scanlon proceeds to conclude that blame, in any relationship, may not involve any affect, one may start to wonder whether Scanlon was in fact thinking of non-personal relationships as the paradigm of his account.   (4.3) Summary I have explained and provided evidence of the two underlying flaws— insensitivity to relationship contexts and insufficient attention to personal relationships—of the major accounts in the blame literature. To avoid these two flaws, an account of blame needs to take seriously the relationship-dependency of blame and recognize the importance of personal relationships in our moral lives. I therefore propose that we abandon the relationship-neutral approach and adopt instead a relationship-dependent approach to blame. I will explain this approach to blame and discuss a relationship-neutral account of blame under this approach in the next section.   	 50	Section 5 A Minimal Account of Blame I concluded in the last section that we should adopt a relationship-dependent approach to blame that recognizes the relationship-dependency of blame and the importance of personal relationships. In this section, I will propose a relationship-neutral account of blame under this relationship-dependent approach. This account can explain the general, relationship-neutral feature(s) of blame while recognizing the relationship dependency of blame (5.1). I will also show how this account can explain the features of blame in personal relationships that I discussed in sections 2 and 3 (5.2) and the force of blame (5.3).   (5.1) Relationship-neutral yet -dependent: Blame as a Response to Relational Harm I have discussed the aspects of blame that theorists fail to explain and examined two underlying flaws that have contributed to their failure. What is the way forward? I propose that we should give up the relationship-neutral approach to blame. If we are too fixated on seeking the general, relationship-neutral features of blame that are present in all relationships, we will likely be unaware of the relationship dependency of blame and more prone to lose sight of features that are specific to particular relationships; namely, the affectiveness and the richness of emotions of blame in personal relationships. Moreover, because blame varies so much in different relationships, if we tries to propose an account that covers blame in all relationships (and particularly if one also adopts the necessary-and-sufficient-condition approach and tries to cover all cases of blame), we will run the risk of achieving generality at the expense of specificity—features of blame that are specific to particular relationships.23  But this does not mean that we should not try to find a general, relationship-neutral account of blame; we only need tread our path to it with great caution. We should adopt a relationship-dependent approach to blame, which recognizes that blame features and functions in different ways in different relationship contexts. The question “what is blame?” has different answers in different relationship contexts, and the answer must start with getting clear on what the background relationship is (or “who is this person to me?”). Moreover, this approach uses the relationship 																																																																		23 This relationship-neutral approach might have contributed to the two underlying flaws discussed in the last section, or it might be the other way round, or they might have reinforced one another. I believe that the last possibility is the most likely one, but I will not take a stance on this issue. What I am arguing is only that maintaining this approach will not be helpful, so we should adopt an alternative approach: a relationship-dependent approach. 	 51	context to explain features of blame. For instance, it uses the features of personal relationships to explain why blame in such relationships is and should be affective and involves a rich set of emotions. This approach therefore avoids the first flaw, insensitivity to the relationship context.24 It also avoids the second flaw. Because this approach recognizes that features of blame cannot be explained without reference to the relationship context, a full account of blame on this approach requires due attention to both personal and non-personal relationships.   This relationship-dependent approach will ultimately take us to a minimal relationship-neutral account. It is minimal in the sense that the features of blame that it specifies as relationship-neutral are very thin. This enables the account to recognize the relationship dependency of blame. This account will also specify the relationship-neutral features of blame in such a way that allows the specific features of blame to be spelt out in individual relationships.    I therefore propose the following account of blame: blame is a response to relational harm. Relational harm is a kind of harm that its nature, the wrongdoer, and the victim are to be understood in terms of the relationship context, and blame, as a response to relational harm, is sensitive to the relationship context. In addition, without attempting to more substantially specify the relationship-neutral features of blame (as the affectivist and non-affectivist accounts do), this account allows for the variance of blame in different relationship contexts—affective in personal relationships and non-affective in non-personal relationships. This account can therefore specify the relationship-neutral feature of blame while recognizing the relationship dependency of blame.    Following the second methodological point explained in section 1.5, this account does not aim to explain the necessary and sufficient conditions of blame. There are cases where blame is a response to general, non-relational harm. We may blame someone simply for deceiving or assaulting us, over and above whether they are a friend or a stranger. So “response to relational harm” is not a necessary feature of blame. Nor is it a sufficient condition. “Responses to relational harm” may include responses such as mere sadness, despair, punishment, change in life goals, and so on, whereas blame is a distinctive response to wrongs. This account of blame does not aim to specify what kind of response blame is as distinguished from these non-blame responses, but aims to highlight some 																																																																		24 Naturally, this approach also does not aim to explain the necessary and sufficient conditions of blame. It aims instead to explain the important features of blame in interpersonal relationships. See section 2.1. 	 52	features important to blame as an interpersonal practice, and it does so by focussing on the relational aspect of blame and hence understands blame as a response to blame.   In the following, I will explain how this account can explain the relationship dependency of blame in relation to the two features of blame in personal relationships: affectiveness and rich emotions.   (5.2) Two Features of Blame in Personal Relationships (5.2.1) Affectiveness I explained in section 2 two reasons for the affectiveness of blame in personal relationships. First, we blame those close to us with emotion because we are emotionally vulnerable to them, that is, we care affectively about whether they treat us appropriately. Blame as a response to relational harm would be sensitive to facts about emotional vulnerability in the relationship. Relational harm is harm that has to be made sense of within the relationship context. It understands harm (partly) in terms of who the harm-doer is in relation to the harmed and what kind of harm it is within the relationship. In the case of personal relationships, the person is someone who is an intimate of the victim and is someone whom the victim cares about how this person treats the victim. Blame understood in such terms is therefore affective.   The second reason that we blame those close to us with emotion is that we see them as particular persons rather then general moral agents. Relational harm understands the nature of the wrong, the wrongdoer, and the victim are all in terms of the relationship context. The wrong would not be simply “breaking a promise” but “breaking a promise made to a (particular) friend (e.g. Sarah)”; the wrongdoer and the victim would not be “a person” or “a moral agent” but “my friend”, or “my sibling”, or “my spouse”. This can explain blame in personal relationships sees the wrongdoer (and oneself) as a particularized person rather than a general moral agent.   (5.2.2) Rich Emotions I gave three reasons in section 3 to explain the rich emotions involved in blame in personal relationships. First, expectations distinctive to personal relationships trigger such rich emotions. I discussed two particular expectations distinctive to personal relationships: care and loyalty. These are expectations that make sense only within personal relationships. Acts of badmouthing and deception may amount to lack of care and disloyalty only in personal relationships. Blame as a response to 	 53	relational harm can cover these cases. It can make the nature of the harm/wrong be sensitive to the relationship context such that, for instance, one would be blaming an act as betrayal in a personal relationship but as merely deception in a non-personal relationships.   Second, blame in personal relationships involves a rich set of emotions because wrongs in such relationships are often linked to relationship impairment. Wrongs in personal relationships are not mere mistreatment of the person but are also harm to the relationship, be it friendship, parent-child relationship, or partnership. Relationship impairment is a kind of relational harm and can be made sense of only in the context of the relationship, and blame as response to relational harm can explain that.   Third, blame in personal relationships may be based on moral-aretaic grounds, which trigger rich emotions. Though moral aretaic blame is not a main focus of this dissertation, we do blame those close to us on moral-aretaic grounds, e.g. that they are irresponsible, lazy, or dishonest. There are some cases of moral-aretaic blame which can be made sense of only within personal relationships. Moral-aretaic blame may be appropriate only in personal relationships. One may appropriately blame one’s friend for being irresponsible or one’s sibling for being lazy, but are in no position to blame a stranger on these grounds. Moreover, one may blame on moral-aretaic grounds only if one stands in a personal relationships with the person. For instance, if a stranger drives recklessly and injures you, you would blame the stranger for driving recklessly and wronging you; but if the driver is your child, you would not only blame her for wronging you—driving recklessly—but would also blame her for her bad moral character—being a reckless driver/person who does not care about your and other people’s safety.   Moral-aretaic blame can be explained in terms of response to relational harm: we hold others to moral-aretaic expectations and would blame them on such grounds because we stand in personal relationships with them.   (5.3) The “Force” of Blame I said in section 1 that there are three general features of blame that accounts of blame should be able to explain. In particular, theorists are concerned to explain the “force” of blame, which is a feature that distinguishes blame from merely making cognitive judgements of moral responsibility. Can this 	 54	minimal account explain the force of blame? I think the force of blame can be more properly appreciated within relationship contexts, and understanding blame as a response to relational harm can help explain the force of blame in different relationships.   (5.3.1) Personal relationships In personal relationships, blame is and should be affective, featuring a rich set of emotions as anger, disappointment, sadness, and hurt feelings. As the affectivist rightly observes, the force of angry blame can be found in the quality of hostility (or “opprobrium”) of anger. It is unpleasant to be angry with someone or be an object of anger, so the force of angry blame is uncontroversial. But some may find disappointment, sadness, and hurt feelings too mild to capture the force of blame. Surely, these emotions are negative, but their force seems to fall far short of that of anger. Being merely sad that someone has stood me up seems hardly an attitude with force.   If we see disappointment, sadness, and hurt feelings outside the relationship context, the force of these emotions may indeed seem far less forceful than anger. But we will be able to appreciate the force of such emotions. if we attend to the relationship context: we are speaking of our parents’ being disappointed in our dishonesty, our friend’s being sad that we have let them down, and our lover’s being hurt by our uncaring attitudes. Such emotions have significant meanings in personal relationships such that they can often provide strong motivations to behaviour and trigger guilt.   In practice, disappointment, sadness, and hurt feelings may at times be even more powerful and meaningful than anger. Because of the unpleasant and hostile quality of anger, our instinctive response to angry blame may be to defend ourselves, to find excuses, or to seek revenge. Some may even dismiss the legitimacy of angry blame out of hand if the angry blamer is a woman and the blamed person is a man: some people have the tendency to dismiss women’s anger as “irrational” if not “hysterical”. These are all distractions from the important issue: that the blamer has been wronged and the wrong needs to be addressed. Disappointment, sadness, and hurt feelings, by contrast, do not as often give rise to these distractions and may be able to more effectively show the wrongdoer that he/she has wronged the victim and to motivate him/her to address the wrong with the victim.    	 55	(5.3.2) Non-personal Relationships Blame in non-personal relationships may be affective, and in such cases its force can then be easily explained by the affect. But blame in non-personal relationships may also be non-affective, involving only cold modifications of intentions, expectations, etc. How can we explain the force of blame without relying on affect? I think we can do that if we have proper regard to the relationship context. On Scanlon’s account, blame involves taking the relationship between the agent and the wrongdoer to be impaired, and revise one’s intentions, expectations, etc. accordingly. Scanlon further observes, in anticipation of the objection that such non-affective blame is not “too weak or mild”, that realizing that what one’s behaviour (i.e. one’s wrongdoing) gives another person good reason to revise their intentions, expectations, etc. towards one is a “serious matter” (2008, 157).   While I agree that revising one’s intentions, expectations, etc. due to (non-personal) relationship impairment is in no way “weak or mild”, I think my minimal account of blame can give a better explanation of the force of blame and strengthen Scanlon’s argument. If we take the harm and its response (blame) in question as taking place in a specific relationship context (non-personal relationships), we can more clearly show non-affective revisions of intentions and expectations as neither “weak” nor “mild”. Non-personal relationships are by nature not intimate, and revisions of intentions etc. have their significance because of the non-intimate nature of such relationships. Relying less on an irresponsible colleague, lowering expectations of a dishonest shopkeeper, and being less disposed to help or even stop greeting a rude neighbour are indeed a serious matter in non-personal relationships, but such non-affective responses would be insufficiently forceful as blame in personal relationships. In other words, the force of such cases of non-affective blame can be explained only with reference to the non-personal nature of the relationships.   (5.4) Summary I have sketched a relationship-neutral account of blame under the relationship-dependent approach. This minimal account of blame—blame as a response to relational harm—specifies the relationship-neutral feature of blame and yet recognizes the relationship-dependency of blame. It explains how and why blame varies in different relationship contexts in terms of features of the relationship.  	 56	Chapter Conclusion  I said in the first section that I aim to explain blame as a normative practice situated in interpersonal relationships, particularly personal ones, in terms of how it functions in and is shaped by (inter)personal relationships. This chapter explains that blame, as it functions in personal relationships, has two distinctive features. First, it is and should be affective; second, it involves and should involve a rich set of emotions. In contrast, blame in non-personal relationships need not be affective and may involve a narrow set of emotions.   This chapter also explains how blame is shaped by (inter)personal relationships. First, it explains that blame in personal relationships has the two distinctive features because of the distinctive features of personal relationships. Second, as a response to relational harm, which is harm that is to be made sense of within the relationship, blame is clearly shaped by (inter)personal relationships.   	 57	Chapter 2 Forgiveness: Private vs Interpersonal This chapter and the next aim to explain forgiveness as a practice situated in interpersonal relationships, particularly in personal ones. This chapter lays the foundation by clarifying the methodology and arguing for the need for an alternative account of forgiveness. The next chapter proposes an alternative account—the performative account of forgiveness.   Similar to chapter 1, my discussion of forgiveness does not aim to look for the necessary and sufficient conditions of forgiveness. Instead, it focusses on explaining forgiveness as an interpersonal practice; that is, the role it plays in (inter)personal relationships in terms of how it functions in and shapes (inter)personal relationships.  I will argue in this chapter that the two mainstream accounts of forgiveness—the standard account and Change-in-View account—fail to adequately explain forgiveness in interpersonal relationships, particularly personal ones. Section 1 gives an overview of the two accounts and argues that they both misconceptualize forgiveness in personal relationships. Sections 2 to 4 launch my main criticism of the two mainstream accounts: they fail to adequately explain the interpersonal aspect of forgiveness—forgiveness is communicative, reconciliatory, and performative. The interpersonal aspect of forgiveness is an important aspect to forgiveness as an interpersonal practice, so an adequate account of forgiveness should be able to explain it. The two mainstream accounts fail to do so and thus fall short of adequate accounts.   This chapter concludes with the proposal that, to explain the interpersonal aspect of forgiveness, we should adopt an interpersonal approach to forgiveness, which conceives forgiveness in terms of its interpersonal aspect. The next chapter presents an account that adopts this approach—the performative account of forgiveness.  	 58	Section 1 Two Mainstream Accounts and Personal Relationships Most theorists of forgiveness endorse the view that forgiveness consists in the overcoming of some negative feelings towards the wrongdoer. This account has been widely accepted as the standard account. Another major account, what I shall call Change-in-View account, takes forgiveness to be the adoption of a benign view of the wrongdoer—i.e. seeing the wrongdoer not as a wrongdoer.25  I will in the following give an overview of these two mainstream accounts. Similar to chapter 1, I will argue that both accounts fail to accurately conceptualize forgiveness in personal relationships and hence fail to adequately explain forgiveness. The upshot of this section is, if we want an account that can better explain forgiveness in personal relationships, we will either need to revise one of the two mainstream accounts, or we will need to look elsewhere for alternatives.   (1.1) The Standard Account (1.1.1) Overcoming Resentment on Moral Grounds Broadly speaking, forgiveness is something offered to a wrongdoer despite that the fact that he/she has genuinely—that is, without any legitimate excuse or justification—wronged the forgiver. The account that has been revered as the standard account of forgiveness argues that forgiveness consists essentially in the forswearing/overcoming of resentment towards the wrongdoer. The origin of the standard account is often attributed to Joseph Butler’s two Sermons on forgiveness, where he argues that forgiveness is the forswearing of the “excess and abuse” of resentment (1896, 151). Butler understands resentment as a kind of anger that (i) is aroused by what one perceives as unwarranted injury and (ii) seeks to defend oneself and punish the injurer. It is justified if the injury is truly unwarranted and if it is proportionate to the injury, but it is “excessive and abusive” if it is based on false beliefs or is disproportionate. Forgiveness is the forswearing of such excessive and abusive resentment, moderating it to a warranted level. In other words, forgiveness is not the complete forswearing of resentment but only its moderation on Butler’s account.   Jeffrie Murphy draws inspirations from Butler’s account of forgiveness and develops it into the standard account in its modern form. In the book Forgiveness and Mercy (1988), which he co-authors 																																																																		25 My categorization of these two accounts may be a bit rough. There are accounts which seem to be a hybrid of the two, e.g. Charles Griswold (2007) and Joanna North (1987, 1998). 	 59	with Jean Hampton, he starts the first chapter with a statement that shows the extent to which he has been influenced by Butler: “Forgiveness, Bishop Butler teaches, is the forswearing of resentment” (15). Murphy argues that wrongs carry the demeaning message that one has a lower moral status than the wrongdoer (e.g. “I count but you do not” (25)), and resentment is a defiant response to such demeaning messages and seeks to defend one’s self-respect (16). Forgiveness, on the other hand, is the forswearing of this self-defensive emotion.   But Murphy emphasizes that forgiveness is not merely the forswearing of resentment simpliciter; it is the forswearing of resentment on moral grounds. One would not have “forgiven” the wrongdoer if one had a head injury and forgot the past wrong, or if one forswears resentment against a wrongdoer in order to improve one’s mental health (22-23). Forswearing resentment amounts to forgiveness only if the resentment is forsworn based on moral grounds; such moral grounds, Murphy adds, must be compatible with self-respect, respect for others as moral agents, and respect for morality (24). One moral ground that Murphy pays most attention to is the repentance of the wrongdoer. By repenting, the wrongdoer no longer endorses his own past wrong and its degrading messages, and has thus separated himself from his past wrong. The separation of the wrongful act from the agent makes forgiving him consistent with self-respect and respect for morality, giving the victim a legitimate moral reason to forgive him (24-26).26   (1.1.2) Attempts to Expand the Standard Account While theorists generally find the standard account appealing, some theorists take note of its narrowness, arguing that forgiveness may involve the overcoming of not only resentment, but also other emotions such as sadness and disappointment. I will add a different but related point here: just as blame in personal relationships involves a set of emotions richer than that in non-personal relationships, so too forgiveness in personal relationships involves the overcoming of a set of emotions richer than that in non-personal relationships—not only resentment, but also sadness, disappointment, hurt feelings, etc.   Theorists have therefore made attempts to expand the standard account to include more emotions other than resentment, but they have yet to agree on what other emotions should be included. Those 																																																																		26 Murphy discusses two other grounds for forgiveness, but he takes them to be either reducible to self-respect or relatively minor. 	 60	who endorse the expansive approach include Norvin Richards (1988), who argues that forgiveness involves the forswearing of all negative feelings towards the wrongdoer, including sadness, disappointment, hurt feelings, contempt, and scorn (77-79); Walker (2006) gives us an even longer list: vengefulness, heartbreak, disgust, humiliation, shame, etc. (155). Even Murphy, who took the narrow approach in an earlier work discussed above, has now conceded that “anger, hatred, loathing, contempt, indifference, disappointment, or even sadness” are all relevant to forgiveness (2003, 59).  There also theorists who expand their accounts not by including negative feelings but other attitudes. For instance, Glenn Pettigrove (2012) and Eve Garrard and David McNaughton (2010, 2014) argue that forgiveness involves a certain degree of goodwill for the wrongdoer. David Novitz (1998) argues that forgiveness involves sympathy/compassion. Proponents of Change-in-view account certainly concur, arguing that forgiveness involves compassion, kindness, or a pro-attitude towards the wrongdoer. (Details in section 1.2)  However, there are theorists who defend the narrow approach; for instance, Paul Hughes (1993, 1995) and Charles Griswold (2007) argue that non-angry feelings, such as disappointment, may not involve any belief that one has been wronged and are therefore not emotions the overcoming of which forgiveness requires.   What these theorists are arguing is that, for any emotion X, X is a forgiveness-relevant emotion (that is, forgiveness involves the overcoming of X) if and only if X is an emotion that is triggered by and only by wrongs. (In other words, X is never triggered by any non-wrongs.) Non-angry emotions such as disappointment and sadness are therefore not forgiveness-relevant because (say) one may be disappointed when one does not win a lottery or when one’s child loses a football game.   I have dealt with a similar objection in Chapter 1 when defending my claim that blame in personal relationships involves a rich set of emotions. Here I will reiterate my argument as applicable to the case of forgiveness. This criterion for forgiveness-relevance is too strong. This point may be better explained in terms of the type-token distinction. For such theorists against the expansive approach (e.g. Hughes and Griswold), whether an emotion is forgiveness-relevant is determined on the type level. For an emotion to be forgiveness-relevant, this emotion must be, as a type, triggered by and only by wrongs. That is, all tokens of this emotion-type must be triggered by and only by wrongs. If 	 61	some tokens of an emotion-type may be triggered by non-wrongs, then the entire type of emotion would fail be forgiveness-relevant—even the tokens which are triggered by wrongs are rendered irrelevant. These theorists are therefore arguing that, because some tokens of non-angry emotions (e.g. disappointment) may be triggered by non-wrongs (e.g. being disappointed in losing a football game), the entire types of these emotions fail to be forgiveness-relevant—even the tokens which are triggered by wrongs are rendered irrelevant.   But it is not obvious that the criterion for forgiveness-relevance should be set on the type level. Why is it that, just because some tokens of an emotion-type may be triggered by non-wrongs, other tokens of the same emotion-type are irrelevant to forgiveness, even if the tokens are indeed triggered by wrongs? Why can’t we say that an emotion-token may be forgiveness-relevant when it is triggered by wrongs, and not forgiveness-relevant when it is triggered by non-wrongs? (Pettigrove 2012, 6 n.12). It seems more reasonable to set the criterion of forgiveness-relevance on the token level. That way, non-angry emotions such as disappointment and sadness may be forgiveness-relevant. They are relevant when, as tokens, they are triggered by wrongs, and not relevant when, as tokens, not triggered by wrongs.    The general direction of the expansive approach therefore seems right, especially in regard to personal relationships. While the question remains what other emotions/attitudes should be included, it seems reasonable to say that, if the standard account is to explain forgiveness in personal relationships, it must include at least emotions such as disappointment, sadness, and hurt feelings.   (1.2) Change-in-View Account Other theorists propose that forgiveness is not merely the forswearing of some negative emotions towards the wrongdoer. What is crucial to forgiveness is instead that we change our way of seeing the wrongdoer when we forgive—from one of blaming to one of forgiving the wrongdoer.27   (1.2.1) Jean Hampton and Margaret Holmgren One prominent proponent of Change-in-View account is Jean Hampton (Hampton and Murphy 1988), who argues that forgiveness consists in changing from seeing the wrongdoer as “not decent” to 																																																																		27 Theorists who have also proposed change-in-view accounts include Lucy Allais (2008), Cheshire Calhoun (1992), Robert Roberts (1995), and Tara Smith (1997). 	 62	seeing her as “still decent” despite what she has done. Hampton rejects the standard account on the grounds that overcoming resentment towards the wrongdoer for moral reasons is not sufficient for forgiveness. A woman who overcomes her resentment and drops her legitimate judgement that her father-in-law has treated her unfairly on grounds of “family peace” is not forgiving but merely condoning him. In addition, Hampton observes that resentment betrays fear and self-doubt in one’s moral value. While a resentful victim may have a desire to defend and reaffirm her moral value, her resentment involves the fear that the wrong is evidence revealing her low value such that she has never been as valuable as she thought she was and such that the treatment she was subject to was in fact appropriate. Or she may think that she was valuable, but she thinks that the wrong has lowered her value such that she has now become not as valuable as she was before (49-51; 59-60). But, contrary to the victim’s beliefs, she is a morally valuable person whose intrinsic value cannot be lowered by anyone. Obviously, if resentment involves such fear and self-doubt, the victim should forswear resentment and regain her self-confidence, and such forswearing of resentment is not forgiveness (83).   Hampton therefore argues that the real obstacle to forgiveness is not resentment but “moral hatred”, or more specifically, the perception of the wrongdoer that moral hatred involves. Moral hatred can be seen as a non-personal analogue of resentment. While resentment deals with and is motivated by one’s value, moral hatred deals with and is motivated by morality, mostly notably justice (59-61). It is an aversion to someone part of or the whole of whom one takes to be “thoroughly identified” with the immoral cause that his/her action promotes, and involves the perception that the person is morally “rotted”, “gone bad”, but it does not involve any fear in one’s own value (80). Forgiveness ultimately consists in overcoming this perception of the wrongdoer. When one forgives, one perceives and judges that the wrongdoer is “something other than or more than the character traits of which she does not approve” and is “still decent despite his action” (83, 85).   But what exactly does it mean to “see the wrongdoer as ‘rotted’ and ‘not decent’” and “take the wrongdoer to be thoroughly identified with the immoral cause that his/her action promotes”? Drawing upon Hampton’s two other works, Linda Radzik (2011) proposes that moral hatred is the perception that wrongdoer is a “committed enemy” of morality. It sees the wrongdoer as a person who freely commits himself to defying the authority of morality, rather than doing it sheepishly or regretfully (4). If moral hatred is seeing the wrongdoer as a “committed enemy” of morality, 	 63	forgiveness then consists in ceasing to see the wrongful act as representing the whole of the wrongdoer but coming to view him as something more benign (4); in other words, he has done something wrongful, but he is short of fully committed to defying morality, and so he is still decent and not rotten.28   Another Change-in-View account that I will discuss is Margaret Holmgren’s (2012). She argues that forgiveness involves regarding certain features of the wrongdoer as salient (4). As she sees it, both resentment and forgiveness are “attitudes” that involves three components: cognitive, affective, and motivational. When one resents a wrongdoer for having wronged one, one regards as salient the fact that the wrongdoer is the person who has chosen to wrongfully harm us (cognitive). Because of this perception, one is subject to the feelings of moral anger (affective), and has a desire for the moral order to be restored (motivational) (30-31). When one forgives a wrongdoer, by contrast, one replaces the attitude of resentment with the attitude of forgiveness. One regards as salient the fact that the wrongdoer is an autonomous moral agent with the capacity to make moral choices, and the fact that the wrongdoer is a sentient being capable of experiencing happiness, pain, etc. (cognitive). One will also have the feelings of compassion and kindness towards the wrongdoer and respect for his personhood (affective) and a desire that things go well for him and that he flourish as a person (motivational) (33-34).  (1.2.2) Problem: Depersonalizing Forgiveness in Personal Relationships Hampton’s and Holmgren’s accounts have both described a way of seeing others that may be accurate in non-personal but not in personal relationships. Whether or not we forgive them, we do not very often see those close to us as general moral agents with some general commitment to morality, moral capacity, or capacity to feel pain and happiness, nor do we abstract them from their past actions, traits, and/or attitudes. Rather, we see them as particular persons with particular past actions, character traits and attitudes (e.g. as “my wife who gave me a bad birthday present last year”, “my brother who is honest but irresponsible”, and “my friend who respects our opinions but are a bit unsympathetic at times”), who stand in particular relationships with us (marriage, siblings, 																																																																		28 Radzik’s reading can help us understand what “rotted”, “not decent”, and “thoroughly identified with immoral cause” are, but it may not tell us what “seeing” and “taking” are on Hampton’s account. Indeed, one difficulty in understanding Hampton’s account, or any Change-in-View account, is how to understand the perception involved in the account. There should be something distinctive about perceiving a person in a certain way, and the perception should not be reduced to judgement, emotion, or motivation lest it would collapse into a cognitive, affective, or conative account. 	 64	friendship, etc.), and who have done some specific wrongful act to us (they lied to us, they didn’t help us when we are in need, they broke a promise, etc.) And we continue to see them as particular persons even when they have wronged us and we will forgive them as such (unless they have done us grave wrongs that severely damage our relationship). To see how such abstract perceptions are at odds with personal relationships, just imagine how unnerving it would be if your lover said he/she forgave you (e.g. for forgetting his/her birthday) because you were not a “committed enemy to morality” or because you were a “sentient being”.   The perceptions of wrongdoer that Hampton’s and Holmgren’s accounts describe are not only inaccurate but also inappropriate in personal relationships. In non-personal relationships, we may often have reason to regard others as abstract moral agents and see them merely in terms of general moral commitment, moral capacity, etc., but it is inappropriate to regard those who are close to us in such ways. As I argued in chapter 1, seeing others as particular people partly constitutes personal relationships. Unless the wrong is major and has seriously damaged the relationship (e.g. backstabbing), we may have reason to continue to see them as particular people whether or not we choose to forgive them. But such are serious cases. In most ordinary situations, where the wrong does not do serious damage to the relationship, we have reason to continue to see them as particular people. The kind of forgiveness that Hampton and Holmgren propose may be appropriate in personal relationships only in cases of serious wrongs. Serious wrongs do occur sometimes, unfortunately, but they account for only a small portion of wrongs that take place in personal relationships. The majority of cases are non-serious yet non-trivial wrongs, and in such cases we should continue to see our intimates as particular people. The perceptions that Hampton’s and Holmgren’s accounts describe are out of place in these cases.   The problem with Hampton’s and Holmgren’s accounts is not that we exclusively see those close to us as particular persons such that we never adopt and never should adopt an abstract perception of those close to us. Nor is the problem that seeing others as particular persons is inconsistent with seeing them as general moral persons. Rather, the problem is that forgiveness on Holmgren’s account (and I believe Hampton’s too) consists in seeing the wrongdoer primarily as an abstract, general moral agent. This perception is at odds with how we see those close to us. When we see the wrongdoer as primarily an abstract, general moral agent, we cannot simultaneously see her primarily as a particularized person. But seeing others as primarily particularized persons is how we see others 	 65	in personal relationships. If forgiveness were how Holmgren and Hampton conceives it, then we would only be able to forgive others non-personally but never personally.   Holmgren might defend the appropriateness of the abstract perception by using the concept of “recognition respect”. She notes in passing that the attitude of forgiveness—seeing the wrongdoer as an abstract moral agent and sentient being—is a form of recognition respect (34). One recognition-respects a person (as a person) if one respects her because she possesses features essential to a person. If we adopt a Kantian concept of persons (as Holmgren does apparently), such features include the capacity to make moral decisions and act autonomously. Recognition respect is then based on such features, rather than on what the person has done, what character traits she possesses, or whether she is our friend or foe. (Respect that is based on the person’s behaviour, character traits, etc. is “appraisal respect”; for instance, we may appraisal-respect a person as honest or a skiier as skilful.) Seeing that regarding as salient the fact that the wrongdoer is a moral agent and a sentient being is a form of recognition respect, and seeing we owe recognition respect to everyone no matter who the person is, what she has done, and how unpleasant her character traits are,29 we should perceive others as abstract moral agents, in both personal and non-personal relationships. (Radzik (2009) gives a more in-depth discussion of this aspect of Holmgren’s account; see also Darwall 1977.)  Holmgren then argues that the attitude of resentment, which involves seeing as salient the person’s acts, character traits, and/or attitudes, disrespects the person because it “objectifies” the wrongdoer by identifying the wrongdoer with her actions, character traits, and/or attitudes (86-87). The wrongdoer is not identical to her actions, character traits, and/or attitudes; she is the agent who chooses her actions, develops her character traits, and holds her attitudes. Such a perception therefore not only makes the conceptual mistake of conflating her with her actions etc. but also fails to recognition-respect her as the moral agent who chooses her actions etc.   One main problem with Holmgren’s argument is that she mistakenly attributes an overriding importance to recognition respect. She argues that we should always and exclusively perceive others as abstract moral agents and disregard their behaviour, traits, etc. But recognition respect does not require that we disregard other people’s behaviour and focus only on their moral agency. It may 																																																																		29 Holmgren adds that one must have fully addressed the wrong by recovering our self-esteem, recognizing the wrongfulness of the wrong, etc. before forgiving the wrongdoer, otherwise one would show disrespect for oneself. 	 66	require merely that we observe certain moral expectations, e.g. do not lie to or manipulate them. Moreover, behaviour, character traits, and attitudes matter in interpersonal relationships, particularly in personal relationships. We cannot and should not disregard such things in how we relate to others.   I can think of two cases where we should exclusively hold the attitude of recognition respect, but neither of them lends weight to Holmgren’s claim about the overriding importance of recognition respect. First, a Buddhist monk may have a distinctive sense of self and others such that he may see others as general moral agents and sentient beings and may be more concerned about their agency and well-being. There may be nothing objectionable about this Buddhist perception of people and way of life—Holmgren does seem to have been influenced by Buddhism—and the Buddhist way of life may carry more nuances than what I can do justice to here. But my point is that (most of) our interpersonal relationships are not Buddhist. In both personal and non-personal relationships, we are concerned not only about other people’s agency and sentience, but also what they do and what character traits they have. So Holmgren’s characterization of forgiveness and its appropriateness, to the extent that they are based on the distinction between recognition and appraisal respect, may be inapplicable to (most of) our interpersonal relationships.    Second, some people argue that parents should love their children “unconditionally”, i.e. no matter what their children do and who they become. When their children did something bad, parents should continue to love them, focussing on their children’s happiness and suffering, and therefore forgive them unconditionally. It is debatable whether this is what parents should do, but even if this is what they should do, this case of unconditional forgiveness would not be very helpful to Holmgren. First, Holmgren’s claim is that recognition respect should always be held and forgiveness be unconditionally given in all relationships, but the case under discussion is one of parent-child relationships. Second, the reason that parents have to unconditionally forgive their children is not that parents should recognize their children as moral agents and sentient beings—such a cold and detached reason has little place in personal relationships, not least in one as intimate as parent-child relationships. Rather, the reason is that the parent stands in a parent-child relationship with their child. This reason comes from their relationship and deals with the particularity of the children, rather than some general features of the children as moral agents and sentient beings.   	 67	Hampton’s and Holmgren’s accounts of forgiveness are therefore unable to fully explain forgiveness in personal relationships—it describes a perception of the wrongdoer that is inaccurate and inappropriate in personal relationships. Wrongs occur in personal relationships every day and plenty of forgiveness is called for, so the failure to explain forgiveness in personal relationships is a major drawback of their accounts.  (1.2.3) A More Particularized View? Some theorists of Change-in-View account have proposed accounts that afford more particularized, “realistic” perceptions of the wrongdoer and might be able to explain forgiveness in personal relationships more accurately. For instance, Lucy Allais (2008) argues that blame involves “lowering the way you affectively esteem or regard [the wrongdoer] as a result of her action” (56), where “affective esteem or regard” is seeing the wrongdoer as trustworthy (or not), as “to-be-censured”, as contemptible, or as having certain character (e.g. inconsiderate, unkind, or nasty) in terms of reactive attitudes (54-55). When we forgive the wrongdoer, on the other hand, we disregard the wrongful act in question in our ways of affectively esteeming or regarding the wrongdoer and we cease to see her as untrustworthy, “to-be-censured”, inconsiderate, etc. (56-57). Allais emphasizes that her account is different from other Change-in-View accounts (such as Hampton’s) in that forgiveness in her (Allais’s) account does not involve coming to a “general” view about the wrongdoer’s “overall” character and moral worth (e.g. that the wrongdoer is still morally decent overall), but involves more specific evaluations of the wrongdoer—that the wrongdoer is trustworthy, considerate, etc.   While I generally agree that forgiveness theorists should adopt a more particularized view of the wrongdoer, theorists of Change-in-View account face a dilemma. On the one hand, if they adopt a more particularized view, forgiveness on their account would require “forgetting” more particularized facts about the wrongdoer. For instance, forgiving a friend who has broken a promise would involve seeing her as once again trustworthy and reliable. While such an account would be able to more accurately describe how we see others in personal relationships, it would fail to accurately describe how forgiven wrongs shape our perception of the forgiven wrongdoer. Wrongs will change the way we see the wrongdoer even after we have forgiven her—in other words, we forgive but we won’t forget. Even though we have forgiven our friend for breaking a promise, her past act, though forgiven, matters to us and to the relationship such that it will inform us of who she is and how we see her, so we may continue to see her as less trustworthy and reliable as before. And we will 	 68	continue to see her in such ways unless we have reason to “disregard” her past wrong, say she has since then become more responsible and sensitive.  We do in some sense “disregard” the past wrong when we forgive, but only in regard to a general view of the wrongdoer; for instance, a forgiver may disregard the past wrong and cease to see the wrongdoer as a committed enemy to morality. Proponents of Change-in-View account may therefore use a more general view to build their accounts in order to avoid the above “forgive-but-not-forget” problem. But this move would only take the account back to the direction of Hampton’s and Holmgren’s accounts and hence to the other horn of the dilemma—that is, misconceptualizing those close to us as general moral agents and thus failing to accurately explain forgiveness in personal relationships. Theorists might be able to craft an account that can strike a balance between the two horns of this dilemma, but this task has yet to be done.   Whether or not Change-in-view account theorists can ultimately solve this dilemma, the most important point from this discussion is that, if an account is to explain forgiveness in personal relationships, it must not limit itself to a general perception of the wrongdoer but must at least be consistent with particularized views.   (1.3) Non-Mainstream, Personal Relationship-friendly Accounts?  There are some accounts in the forgiveness literature that are more personal relationship-friendly than the two mainstream accounts discussed above. For instance, a few theorists have proposed performative accounts of forgiveness (e.g. Joram Haber, Glen Pettigrove, and Brandon Warmke). The performative account argues that saying “I forgive you” (verbally or non-verbally) is performing a speech act that brings about some effects in virtue of the utterance. Haber (1991) argues that forgiveness is an act of revealing that one has overcome or is willing to try to overcome one’s resentment for the wrongdoer’s wronging one; Pettigrove (2012) proposes that forgiveness, in its “highest manifestation”, is an act of revealing that one has overcome hostile reactive attitudes towards the wrongdoer and an act of committing to overcoming such attitudes (18); Warmke (2016a) argues that forgiveness is an act that alters the norms between the wrongdoer and the victim. These accounts do not commit the depersonalization problems and is consistent with the rich emotions found in personal relationships. (Haber, however, limits his account only to resentment.) In particular, Pettigrove, whose account takes good will as a crucial component of forgiveness, further points out 	 69	that background relationships determine the nature and degree of good will that is relevant to forgiveness (2012, 8; 15).   Alice MacLachlan, another non-mainstream forgiveness theorist, recognizes the distinctive features of the practice of forgiveness in personal relationships. She observes that “there are a variety of ways in which people may forgive and a wide range of expressions forgiveness may take” (2009, 191), and one important determining factor is the relationship between the parties(188). She then considers a few examples of personal relationships: in friendship, forgiveness may be communicated non-verbally, but in a “long standing, complex family relationship marked by distrust on both sides, words of forgiveness along might be deeply unsatisfying for all concerned” (188). In a recent paper (2017), she observes the importance of “moral solidarity” in personal relationships such as family and friendship (145); this in turn demonstrates the importance of third-party forgiveness in our moral lives.    Similarly, Margaret Walker’s discussion of forgiveness also shows some sensitivity to personal relationships. Being aware of the disunity of forgiveness, she steers clear of the necessary-and-sufficient-condition approach to forgiveness. (More on this approach in the next section.) She recognizes that forgiveness sometimes (but not always) involves forswearing certain emotions, but what emotions are relevant to forgiveness is dependent on the nature of the wrong, which is in turn dependent on the background relationship (2006, 155-156).    While there are some theorists who recognize the distinctive features of forgiveness in personal relationships and have proposed accounts which are sensitive to them, their voices remain non-mainstream. (I likewise will take a non-mainstream path and propose a performative account in chapter 3.)   (1.4) Summary Both two mainstream accounts misconceptualize and unable to accurately explain forgiveness in personal relationships . The standard account fails to take into account the emotional richness in personal relationships, and should expand the set of emotions relevant to forgiveness if it is to explain forgiveness in personal relationships. Change-in-View accounts misconceptualize persons in personal 	 70	relationships—those who are close to us—as abstract, general moral agents. An account that is at least consistent with a particularized conception of person is called for.    Forgiveness theorists now face a choice. If they want to adequately explain forgiveness in personal relationships, then they will need to either revise the two mainstream accounts to make them more personal-relationship-friendly, or look elsewhere for an alternative, “non-mainstream” account. This is the upshot of this section.   But perhaps theorists do not have a choice after all. In the following three sections, I am going to argue that the two mainstream accounts are unable to explain the interpersonal aspect of forgiveness. Even if theorists managed to revise the two mainstream accounts to solve the misconceptualization problem of forgiveness in personal relationships, they would still need to look for an alternative account that can explain the interpersonal aspect of forgiveness.    	 71	Section 2 Feature One: Forgiveness is Communicative I argued in the previous section that the two mainstream accounts cannot adequately explain forgiveness in personal relationships. In this and the next two sections, I will advance another, indeed more important, criticism of the two mainstream accounts: that they cannot adequately explain the interpersonal aspect of forgiveness. This interpersonal aspect of forgiveness exists in both personal and non-personal relationships, but it is a more significant aspect in personal than in non-personal relationships because of the greater role that forgiveness plays in determining how we relate to those close to us.   I will discuss and highlight the importance of the interpersonal aspect of forgiveness in relation to three features: communicativeness (section 2), reconciliation (section 3), and performativeness (section 4). I will argue that the two mainstream accounts, due to their “private approach”, do not take the interpersonal aspect seriously and therefore cannot explain it adequately. We should take an interpersonal approach in order to explain the interpersonal aspect of forgiveness.   In section 2.1, I will explain the importance of these three interpersonal features of forgiveness, and provide a few reasons for their particular importance in personal relationships. In section 2.2, I zero in on the communicativeness of forgiveness—the topic of this section—and argue that the two mainstream accounts fail to adequately explain the communicativeness of forgiveness due to their “private approach”. Then, in 2.3, I argue that we need an alternative account that can adequately explain the interpersonal aspect of forgiveness, and claim that my performative account (to be discussed in the next chapter) is able to replace the two mainstream accounts because it can explain the interpersonal aspect and (at least partly) the private aspect of forgiveness.   Similar to chapter 1, this and the next chapters on forgiveness are guided by two methodological points. First, my discussion of forgiveness does not aim to explain the nature of forgiveness (i.e. the necessary and sufficient conditions of forgiveness). Instead, it aims to explain forgiveness as a practice situated in interpersonal relationships (particularly in personal ones), and it does so by explaining the interpersonal aspect of forgiveness (i.e. the three features). This first methodological point will be explained in section 2.1.   	 72	The second methodological point is related to the first one. My discussion takes an interpersonal approach to forgiveness, which aims to explain forgiveness as an interpersonal practice through its interpersonal aspect of forgiveness, rather than its private aspect. This second methodological point will be explained in sections 2.1 and 2.2.  (2.1) Three Interpersonal Features of Forgiveness  I argued in chapter 1, section 1.5 that seeking the necessary and sufficient conditions of disunified phenomena such as blame may not be very illuminating because such efforts will likely leave out some features that are crucial to the phenomenon. I therefore argued that we should understand blame as an interpersonal practice and focus on explaining the role it plays in interpersonal relationships—namely, how it functions in, is linked to, and is shaped by our relationships—rather than its nature (i.e. its necessary and sufficient conditions).   As a few forgiveness theorists observe, forgiveness is likewise a disunified phenomenon and effort to find a single account that explains all aspects of forgiveness will not be very fruitful. (Walker, Radzik) I agree with this observation; I therefore do not aim to explain the necessary and sufficient conditions of forgiveness, but focus on explaining the role it plays in interpersonal relationships—namely, how it functions in and shapes our relationships. And this is the first methodological point of this and the next chapters.   One aspect of forgiveness that is crucial to its role in interpersonal relationships is its interpersonal aspect. This interpersonal aspect and the three interpersonal features which it involves may not be present in all and only cases of forgiveness, but they can help illuminate what forgiveness is because, as a practice that is situated in interpersonal relationships, forgiveness often involves a great deal of social interactions and is intertwined with our relationships with others in some profound ways. The explanatory work that I aim to do in this and the next chapters is to explain this interpersonal aspect of forgiveness. And this is the second methodological point of this chapter.   The interpersonal aspect of forgiveness involves three closely related interpersonal features through which the role of forgiveness in interpersonal relationships is expressed: forgiveness is communicative, reconciliatory, and performative. First, forgiveness is communicative because we often do not privately and solitarily consider and process forgiveness; rather, the victim often wants to communicate, verbally or 	 73	non-verbally, his forgiveness to the wrongdoer, and the wrongdoer also wants to know that she has been forgiven. For instance, when we forgive our friend for standing us up,  we want them to know that we have forgiven them so that we can both move on; and they, knowing that we are not pleased by their standing us up, want to know whether they have been forgiven.   Second, forgiveness, in a sense that is hard to pinpoint, reconciles relationships and in turn determines how we are to relate to one another after the wrong. (It is for this reason that forgiveness is often communicated.)   Third, forgiveness is a performative in the sense that it brings about changes to relationships in virtue of being given. The change that forgiveness is reconciliation, which, once again, determines how we relate to one another; and to bring about reconciliation, forgiveness needs to be communicated.    These three interpersonal features—I shall call them collectively the “interpersonal aspect of forgiveness”—exist in both personal and non-personal relationships, but they are particularly important in personal ones, for three reasons. First, in relation to the communicativeness of forgiveness, we do not keep everything in our hearts but will express it to those close to us. We do not just feel grateful to our friends for helping us move houses; we express our gratitude to them (e.g. in the form of treating them to a nice dinner) so that they know we are grateful. We do not just like our sibling; we tell them and do nice things to them (e.g. organizing a big birthday party for them). Likewise, we do not keep our forgiveness in our hearts; we tell our friend that we have forgiven them for their broken promise (e.g. telling them directly, or, more subtly, inviting them to go fishing with us again).   The second and third reasons both deal with all three features. The second reason for the significance of the interpersonal aspect in personal relationships is that questions about the current/future conditions of the relationship is more urgent and less avoidable when we are dealing with people close to us. Non-personal relationships are less resilient to wrongs. They break and end more easily, and we may be more ready to walk away from the relationship if we so choose. But personal relationships are more resilient. Except the most serious ones, wrongs often do not break personal relationships beyond repair—we won’t end a friendship just because our friend has told a small lie or missed an appointment—so the option of making up is most likely to be on the table. In addition, 	 74	some close personal relationships are such that we cannot avoid seeing each other, e.g. if we live in the same house or in the same (extended) family. This means that we have to decide where we stand and how we are to relate to one another after wrongs.   This explains the importance of the interpersonal aspect in personal relationships: reconciliation determines the shape of relationships and how we are to relate to one another, is brought about via the forgiveness performative, and needs to be communicated in order to have such effects and facilitate the decision-making of all those involved.   The third reason is closely related to the second one. Questions about forgiveness are often closely related to questions about our relationships. We often forgive those who are close to us with the intention to improve the relationship damaged by the wrong, and we often seek forgiveness because we want to improve the damaged relationship. To serve this purpose, forgiveness needs to be communicated publicly from the giver to the receiver, and reconciliation and the performativeness of forgiveness are naturally also at stake.   These three features of forgiveness—the interpersonal aspect of forgiveness—are crucial to the role of forgiveness in interpersonal relationships. An adequate explanation of forgiveness as an interpersonal practice should be able to explain these three features. The two mainstream accounts, I will argue in the next sub-section and the coming two sections, cannot explain these three features.   (2.2) Two Mainstream Accounts: Forgiveness as Non-communicative This section focusses on the communicativeness of forgiveness, so I now narrow the discussion down to the communicativeness of forgiveness. I argue that the two mainstream accounts cannot adequately explain the communicativeness of forgiveness, and I attribute their failure to their “private approach” to forgiveness. (I will argue in the next two sections that the two mainstream accounts cannot explain the two other interpersonal features of forgiveness.)   The two mainstream accounts are unable to explain the communicativeness (and the two other interpersonal features) of forgiveness because of their “private approach” to forgiveness. They take some private elements (i.e. emotion or perception) as the most important—the “core”—feature of forgiveness, conceiving forgiveness to be something that can be kept entirely inside one’s head/heart 	 75	and not communicated to anyone. The standard account sees forgiveness as a matter of how we feel about the wrongdoer: forgiveness consists in the overcoming of resentment (Butler and Murphy), sadness, disappointment, hurt feelings (Richards), if not also vengefulness, heartbreak, and disgust (Walker), along with holding some other attitudes such as good will (Garrard and McNaughton). Change-in-View theorists, on the other hand, see forgiveness as a matter of how we perceive the wrongdoer. Hampton says forgiveness consists in seeing the wrongdoer as “still decent”, and Holmgren says forgiveness is seeing as salient the fact that the wrongdoer is a moral agent and a sentient being.30     The two accounts do not deny that forgiveness can be communicated verbally or behaviourally to the extent that emotions and perception can be so expressed and communicated. For instance, Butler and Murphy would say that overcoming resentment is a form of lowering one’s guard against the wrongdoer and can be expressed behaviourally, and Holmgren has made motivations an aspect of forgiveness. However, by saying that forgiveness consists in some emotion and perception, the two accounts treat the private aspect of forgiveness as the most important aspect, and admit of communicativeness as merely a secondary feature of forgiveness that is to be explained by the core feature. They cannot adequately explain the communicativeness of forgiveness given its importance in interpersonal relationships.31   If we are to adequately explain the communicativeness of forgiveness and in turn the role of forgiveness in interpersonal relationships, we should instead treat this feature as crucial to forgiveness and give them a central role in our accounts of forgiveness, rather than relegate them to a secondary role.32 In other words, we should adopt an “interpersonal approach” to forgiveness, which 																																																																		30 Holmgren thinks that forgiveness as an “attitude” that involves cognitive, affective, and motivational aspects, but she seems to take the cognitive aspect—regarding certain features of the wrongdoer as salient—to be the primary aspect of forgiveness. 31 Perhaps what lies under the two accounts’ private approach is their necessary-and-sufficient-condition approach to forgiveness—that is, the main task of their accounts is to find the feature(s) that is (are) found in all (or almost all) cases of forgiveness in all (or almost all) relationships. (One notable exception is Holmgren’s account.) Theorists are then driven to shred their conceptions of forgiveness of all the features that they see as contingent to forgiveness, relegating such features to a secondary place in their accounts—and the communicativeness of forgiveness (and indeed all three interpersonal features) is one of the casualties. For instance, a victim may forgive the wrongdoer without behaving differently to her, or may behave differently to her without forgiving her. Behaviour—a way to communicate forgiveness—is therefore neither necessary nor sufficient for forgiveness, so theorists conclude that it is a mere secondary feature of forgiveness. 32 Michael McKenna has recently proposed a conversation theory of blame that bears some relevance to my analysis of forgiveness. His theory of blame conceives blame as forming part of what he calls “moral responsibility exchange” 	 76	is an approach that conceives forgiveness and explains it as an interpersonal practice in terms of the communicativeness and the other two interpersonal features.   (2.3) Performative Account as a Replacement  In the next two sections, I will argue that the two mainstream accounts fail also to explain the other two interpersonal features of forgiveness—that forgiveness is reconciliatory and performative. I will then conclude at the end of this chapter that, if we are to explain these three interpersonal features of forgiveness, we should adopt the performative account as a replacement of the two mainstream accounts.   However, one might argue that replacing the two accounts with an alternative account is too radical. We should first consider revising and “interpersonalizing” the two mainstream accounts. And even if we may in the end need the performative account as an alternative account, we may need it only as a complement of the mainstream accounts rather than as a replacement. I will argue in the following that we do need a performative account as an alternative and that it can replace the two mainstream accounts.     																																																																																																																																																																																																																																						between morally responsible agents. He observes that actions of morally responsible agents (e.g. right/wrongful acts, blame, and forgiveness) carry meanings and express their regards towards others (2012a, 92-94). To illustrate this with McKenna’s own example, suppose someone tells a racist joke. This joke teller’s act expresses his negative regard towards certain racial groups and initiates a moral responsibility exchange. McKenna calls this initial stage of exchange “moral contribution”. Another person responds to this first person’s contribution by blaming him and showing that she does not approve of what he did. She may even further calls upon him to justify, provide an excuse, or apologize for what he did. McKenna calls this second stage “moral address”. The third stage, “moral account”, is a stage where the first person responds and gives justification, excuses, apology, etc. (Or he may refuse and withdraw from the exchange) (2012a, 89-90; 2012b, 127-128).  McKenna’s discussion primarily deals with blame, so he has merely mentioned in passing that there may be a fourth stage where forgiveness is involved. He states that forgiveness, reconciliation, and apology can all be understood as part of an “unfolding dialogue”, where each of these three “offers the interlocutor reasons to consider in a new light either the moral quality of the agent [the wrongdoer], the relationship [between the wrongdoer and the victim], or, depending on the nature of the account offered, the initiating conduct” (90).   I agree with McKenna’s interpersonal approach to forgiveness, particularly in relation to his observation that forgiveness is communicative partly because both parties would take it as an indicator of the conditions of the relationship and an important factor determining they are to relate to one another. However, while McKenna focusses on the communicativeness of forgiveness, taking forgiveness to be a form of moral exchange, my performative account (to be discussed in the next chapter) emphasizes the performativeness of forgiveness and explains the communicativeness of forgiveness in terms of its performativeness. 	 77	(2.3.1) Revised or Alternative? I argued that the two mainstream accounts, due to their private approach, are unable to explain the interpersonal aspect of forgiveness (the three features). Can the two accounts be revised and “interpersonalized” to explain this important aspect of forgiveness? If so, then we wouldn’t need to look for an alternative account.   I do not think that the two mainstream accounts could be revised and “interpersonalized”. The reason is that the two mainstream accounts are inherently private and therefore incompatible with the interpersonal approach. If the two mainstream accounts were to be revised to explain the interpersonal aspect, they would inevitably depart from their original nature and become at least partially an alternative account. This means that theorists would ultimately be taking the alternative route rather than the revising route.   There may have been attempts to revise and “interpersonalize” the two mainstream accounts. One notable example is Charles Griswold (2007). However, it may be more appropriate to regard Griswold’s account to be an alternative account rather than a revision of the two mainstream accounts. Moreover, even this account faces its own problems in explaining the interpersonal aspect of forgiveness.    Griswold’s account explicitly acknowledges the “intrinsic interpersonal dimension” of forgiveness (58). On his account, forgiveness requires the inputs of both the victim and the wrongdoer, some of these inputs involve communication. If forgiveness is to be warranted, the wrongdoer has to (i) acknowledge that he is responsible for wronging the victim, (ii) repudiate his deeds, (iii) experience and express regret, (iv) commit to moral improvement through words and deeds, (v) understand from the victim’s perspective how he has wronged and harmed the victim, and (vi) provide a narrative of how he came to do now and how he is now worthy of approbation (47-51). And to forgive the wrongdoer, the victim needs to (vii) forswear resentment and (viii) revenge, (ix) commit to letting resentment go all together, (x) change the judgement that the wrongdoer is a “bad person”, (xi) drop any presumption of moral superiority, and (xii) communicate one’s forgiveness to the wrongdoer (54-58).    	 78	Among the above twelve conditions, (iii), (iv), (vi), (xii), and potentially (i), (ii) and (v) involve communication between the victim and the wrongdoer. If Griswold’s account were successful, then it might be able to serve as a revision of one of the two mainstream accounts; even if it is not entirely successful, it might nevertheless show that revising the two accounts is a viable approach to “interpersonalizing” them.   However, Griswold’s account involves both the overcoming of resentment and changing one’s perception of the wrongdoer, along with a number of other elements—there are twelve conditions in total. It is therefore not so much a revision of either of the two accounts as a “hybrid” or even a “mixed” account—that is, an alternative to the two mainstream accounts. The account also faces other problems. First, it is a mix of various elements and lacks unity. Second, while it can explain the communicativeness of forgiveness, it cannot explain the other two features of the interpersonal aspect of forgiveness. It cannot explain how forgiveness constitutes reconciliation—Griswold explicitly denies that forgiveness constitutes reconciliation (111)—and how forgiveness is performative—forgiveness is not a performative on his account.  (2.3.2) Complement or Replacement?  Even if one decides to find an alternative account, wouldn’t the two mainstream accounts still have a part in explaining the role of forgiveness in interpersonal relationships? While communication is important in personal relationships, how we feel about and see the wrongdoer are also important to forgiveness and personal relationships. So it may be more fair to say that both the private aspect and the interpersonal aspect are important to forgiveness as an interpersonal practice, and the two mainstream accounts are able to shed light on the interpersonal role of forgiveness through explaining the private aspect of forgiveness. And didn’t I do something similar in chapter 1, where I explain the role of blame in interpersonal relationships by showing how affect—a private matter—is linked to features crucial to personal relationships?  If this is right, then perhaps the alternative account would only need to complement the two accounts rather than replace them. Such a complementary account would work in tandem with the two mainstream accounts, explaining the interpersonal aspect of forgiveness and leaving the private aspect to the two mainstream accounts. We may not need to take the more ambitious path to find an 	 79	alternative account that will explain both the private and interpersonal aspects and replace the two mainstream accounts altogether.   Finding a replacement may be a tall order, but I think it is viable. The performative account that I propose in the next chapter is a unified account that can adequately explain the interpersonal aspect of forgiveness and also at least partly its private aspect. It therefore has some success in achieving the ambitious goal of replacing the two mainstream accounts. But even if it could not entirely replace the two mainstream accounts, it can at least complement the two mainstream accounts.   But before I propose this performative account, I need to explain in greater detail how the two mainstream accounts fail to explain the other two features of the interpersonal aspect of forgiveness—that forgiveness constitutes reconciliation (section 3) and is performative (section 4).  	 80	Section 3 Feature Two: Reconciliation In this section, I am going to discuss a second feature of the interpersonal aspect of forgiveness—forgiveness constitutes reconciliation—and argue that the two mainstream accounts fail to explain this feature.   I must acknowledge at the very beginning the contentiousness of the idea that forgiveness constitutes (rather than being merely instrumental to) reconciliation. This section does not intend to settle the debate surrounding this idea, nor to specify the sense in which forgiveness constitutes reconciliation. Rather, it intends to keep this idea alive. I argue that, while theorists have given some objections against this idea, they have been too quick to conclude that forgiveness does not in any meaningful sense constitute reconciliation.   I suspect that one underlying reason for theorists’ rejection of this idea is the elusiveness of the concept “reconciliation”. This section therefore also intends to demonstrate the challenge we face in specifying a meaningful sense in which forgiveness does constitute reconciliation. This discussion in turn sets the stage for the next chapter, where I argue that my performative account can take up this challenge and explain this interpersonal feature of forgiveness.   In (3.1), I will discuss the idea that forgiveness constitutes reconciliation and the elusive concept of reconciliation. (3.2) gives a brief overview of the reconciliation literature and then narrows the discussion down to that pertaining to the relationship between reconciliation and forgiveness. (3.3) and (3.4) discuss two accounts of reconciliation in the literature that explain the relationship between reconciliation and forgiveness—reconciliation as restoration and re-establishment of trust-worthiness—and how theorists use them to argue that forgiveness does not constitute reconciliation. Turning to the two mainstream accounts, I argue in (3.5) that they are ill-equipped to explain the idea that forgiveness constitutes a meaningful sense of reconciliation, if there is one. In (3.6), I suggest that we need a new way to explain how forgiveness constitutes reconciliation; my performative account can provide a new way.  Before I begin the discussion, I should reiterate the first methodological point that I made in the previous section. I am not arguing that reconciliation, or any of the three interpersonal features of 	 81	forgiveness, is a necessary or a sufficient condition for forgiveness; instead, I am arguing that all three of them are important features of forgiveness as a practice situated in interpersonal relationships, particularly personal ones.   (3.1) Reconciliation, Really?  The reconciling nature of forgiveness is manifest in our lived experience. When we forgive a friend who stood us up, a sibling who broke a promise, or our spouse for forgetting our birthday, our relationships with them seem to be, in some sense, reconciled. Forgiveness theorists are well aware of this experience in our moral lives, but they are less than receptive to the idea that forgiveness is reconciliatory. They may agree that forgiveness can contribute to or be instrumental to reconciliation (e.g. Griswold 2007; Robert Roberts 1995); as Roberts puts it, reconciliation is the “teleology” of forgiveness. But most theorists33, including proponents of the two mainstream accounts, reject the idea that forgiveness constitutes reconciliation. Their argument is that, using the senses of “reconciliation” that they hold, we may forgive without reconciling with the wrongdoer and that we may reconcile with the wrongdoer without forgiving her. In other words, reconciliation, in the senses that they hold, is neither necessary nor sufficient for forgiveness (e.g. Griswold 2007, Hampton 1988, and Radzik 2009).  But their rejection is premature. Theorists might have successfully argued that forgiveness does not constitute reconciliation in the senses that they hold, but they have yet to establish the stronger claim that forgiveness does not constitute any meaningful sense of reconciliation. Doing the latter requires an exploration of other possible understandings of reconciliation. And before a thorough exploration is undertaken, we should not prematurely reject the intuitive idea that forgiveness constitutes reconciliation, as theorists did.   																																																																		33 One prominent exception is Scanlon (2008), who has in effect given a reconciliation account of forgiveness. As I explained in chapter 1,  Scanlon argues that blame is the modification of relationship-constituting attitudes towards the wrongdoer in light of the wrong act, such attitudes include dispositions, intentions, and expectations. Directly from his account of blame, he derives an account of forgiveness: while blame is the modification of relationship-constituting attitudes in light of the wrong act, forgiveness is the readjustment of relationship-constituting attitudes (160). For instance, you may blame your unreliable friend by avoiding her, and you may forgive her by “un-avoiding” her, say starting to invite her to parties again. Scanlon’s account of forgiveness is thus a reconciliation account of forgiveness: forgiveness is the restoration of relationship. Nevertheless, we may need to treat Scanlon’s account of forgiveness with caution. Scanlon proposes his account of forgiveness in the context of his discussion of his account of blame. Its minimal coverage makes it seem as if Scanlon was merely mentioning his view on forgiveness in passing rather than proposing and endorsing an account of forgiveness. 	 82	But reconciliation is an elusive concept. We use this concept in our lives and have some sense about what it is—a reconciled relationship is a relationship that has been strained, damaged, if not broken, but has gone through some positive changes so it is now good or at least normalized again. But it is difficult to spell out its specifics: what “positive changes” are these? How “good” does a relationship have to be in order to be reconciled? Are there any standards? What is a “normalized” relationship? And how does a relationship become good or normal again and therefore “reconciled”? This elusiveness of the concept reconciliation may partly contribute to theorists’ rejection of the idea that forgiveness constitutes reconciliation.   The challenge, therefore, is to (i) provide a meaningful sense of reconciliation that can (ii) explain the reconciliatory nature of forgiveness found in our lived experience. I argue in this section that the two mainstream accounts cannot do either of these two tasks; by contrast, my performative account, proposed in next chapter, is able to do both. In the following three sub-sections, I will give a brief overview of the reconciliation literature and then narrow the discussion down to that pertaining to the relationship between reconciliation and forgiveness (3.2). I will in particular discuss two accounts of reconciliation that forgiveness theorists have used and their reasons for rejecting the claim that forgiveness constitutes reconciliation (3.3 and 3.4).  (3.2) Reconciliation: Mostly Political, Partly Personal Reconciliation generally speaking consists in some positive changes of the damaged relationship, but the “positive changes” that are essential to reconciliation differ among accounts in the literature. These accounts can be roughly divided five types according to the kind of positive changes that they specify as essential to reconciliation (Linda Radzik and Coleen Murphy 2019): (i) changes in institutional structures (Lu 2017, C. Murphy 2010); (ii) changes in external behaviours (Bhargava 2012, Griswold 2007); (iii) changes in beliefs (Dwyer 1999); (iv) resolving negative emotions and attitudes (Lu 2017, Mihai 2016); (v) adopting or resuming positive emotions and attitudes (Emerick 2017, Griswold 2007, Radzik 2009, Scanlon 2008, Walker 2006) .   Reconciliation admits of different degrees, and each of these five types of accounts aims to explain a specific degree of reconciliation. Generally speaking, type (i) accounts provide the lowest degree of reconciliation and type (v) accounts provide the highest. For instance, the reestablishment of trust (type (v)) is a higher degree of reconciliation than the agreement to cease hostilities (type (ii)) 	 83	(Griswold 2007). However, there are different degrees of reconciliation even within the same type of accounts; for instance, lovers who renew their love reconcile to a higher degree than if they merely reestablish basic respect (type (v)).   Moreover, what reconciliation consists in is dependent on the background relationship. For instance, institutional restructuring is necessary for political reconciliation (C. Murphy 2010) but is not applicable to personal relationships, whereas love and passion are crucial in intimate relationships but not relevant at all in non-personal ones. Therefore, each of these five types of accounts is better at explaining reconciliation in some relationships than others. For instance, types (i) and (ii) are more suitable for political and non-personal relationships, whereas types (iv) and (v) are more suitable for personal relationships.   The reconciliation literature focusses primarily on political reconciliation with extensive references to South Africa (Apartheid), Rwanda (the genocide), Canada (treatment of the first nations), and so on. An emerging subset of the literature discusses reconciliation in non-political, person-to-person relationships, with particular attention to the relationship between reconciliation and forgiveness. In particular, theorists adopt two accounts of reconciliation to explore this relationship. (Both accounts are type (v) and understand reconciliation in strong senses.) I discuss these two accounts in the following two sub-sections.   (3.3) Restoration—the Previous State Most forgiveness theorists understand reconciliation as restoration (e.g. Emerick 2017, Griswold 2007, Scanlon 2008). That is, reconciliation is the restoration of the relationship to its state previous to the wrong. Griswold (2007) argues that reconciliation (in the strong sense) means “affirmation”, meaning “something like friendship and support or a renewal of any previous ties of affection” (111). Scanlon (2008) explains the restoration account in greater detail. He first observes that relationships partly consist in “relationship-constituting attitudes”, e.g. intentions, feelings, expectations, etc. For instance, close friendships involve a readiness to offer help to one another and expectations that the other person would offer help if one needed it. Wrongs change such attitudes. If my friend without good reason refused to take me to the hospital when I was really sick, I might be less ready to help her in the future when she needed my help. My friendship with her is therefore damaged. So to 	 84	reconcile with her, in the sense of restoration, is to restore such attitudes to the state before the wrong, e.g. being as ready to help my friend as before.   As many theorists (e.g. Emerick, Griswold, and Hampton) have pointed out, forgiveness does not constitute reconciliation in this restoration sense. We can forgive a deceitful friend without expecting her honesty as much as before, and we can “go on as before” for the sake of the relationship with an unfaithful spouse without forgiving him. Moreover, if forgiveness required restoration, then forgiveness would be impossible in some cases where we think it should be possible. Wrongs sometimes permanently damage the relationship such that restoration is out of the question. A lie may permanently damage trust, and harsh words may break a loving heart for good. Sometimes we may even have good reason not to trust the wrongdoer like before. A victim of physical and emotional domestic abuse surely have good reason to be more cautious and trust their spouse less in the future (that is, if they choose to stay in the marriage).  Many theorists then conclude at this point that forgiveness does not constitute reconciliation. But the problem may not lie in the idea that forgiveness constitutes reconciliation. It may rather lie in this particular understanding of reconciliation—reconciliation as restoration. “Restoration” is a very high standard such that it not only fails to explain forgiveness but also fails to explain even reconciliation. A lie might make me never trust a friend as much as before, but it would not mean that I could never reconcile with the friend at all. Moreover, some emotions of intimacy cannot be fully restored after wrongs damage them, e.g. love. If reconciliation did require restoration, then reconciliation would be impossible in such intimate relationships. This would be absurd. Reconciliation may still be possible, but just not in the sense of restoration.   Restoration is not the only way to understand reconciliation and forgiveness. It is too soon to conclude at this stage that forgiveness does not constitute reconciliation. What we should do is to explore other possible understandings of reconciliation. Are there any other options?   (3.4) Reconciliation as Re-establishment of Trustworthiness (3.4.1) Linda Radzik Few forgiveness theorists have considered other possible understandings of reconciliation—Linda Radzik is one exception. Her account of reconciliation (2009), which argues that reconciliation 	 85	consists not in the restoration of relationship but the re-establishment of trustworthiness; may be able to provide a weaker sense of reconciliation. Radzik argues that a reconciled relationship is one where the wrongdoer re-establishes herself as a trustworthy person and the victim trusts the wrongdoer again. In a trusting relationship, we regard one another and ourselves as moral equals who (i) deserves to be treated properly and (ii) are capable of and will likely act appropriately. Radzik calls such trusting relationships the “paradigmatically moral relationship”. But when we do wrong, we show that we will act inappropriately and we thereby damages this paradigmatically moral relationship between us and the victim. Reconciliation lies in the re-establishment of the paradigmatically moral relationship, and to do so the wrongdoer needs to re-establish herself as a person who will act appropriately (i.e. re-establishing her trustworthiness) and the victim needs to trusts the wrongdoer again (81-82).   Though Radzik’s account is modelled on the “paradigmatically moral relationship”, which is a relationship where we all stand to one another and are morally required to maintain (81), and is thus modelled on a (very distant) non-personal relationship, she uses an example to briefly describe how her account can incorporate reconciliation in personal relationships. To reconcile with her (ex-)husband, an unfaithful wife need not re-establish her trustworthiness as her (ex-)husband’s wife after the failed marriage because this would require the restoration of the relationship and is too strong a condition of reconciliation. Instead, Radzik argues that this women can reconcile with her (ex-)husband by re-establishing herself as a trustworthy marital partner such that, for instance, her ex-husband would not worry about her faithfulness if she went on and married someone he knows, say a friend (114-115).   Radzik’s account of reconciliation has a few problems. It oversimplifies trustworthiness as invariable among all persons. Conditions of trustworthiness are variable to a very significant degree in different relationships. Her approach of incorporating personal relationships also depersonalizes personal relationships. We do not see those close to us as general persons who has a certain relational role: “a friend”, “a wife”, or “a brother”. Rather, we see them as particular persons who stand in particular personal relationships with us: “my friend”, “my wife”, or “my brother”.   The biggest problem of Radzik’s account in the context of our discussion is that, as Radzik recognizes, it cannot explain forgiveness as constitutive of reconciliation. Forgiveness does not seem 	 86	to require reconciliation in the sense of trustworthiness because forgiveness is compatible with any level of trustworthiness. We can forgive a wrongdoer whom we now trust more, less, just as much as before, or not at all, whether the relationship is personal or non-personal, and whatever the moral expectations in question are. A man can forgive his ex-wife for her infidelity whether or not she has re-established herself as his/a trustworthy romantic partner. Radzik therefore concludes that forgiveness does not constitute reconciliation; that is, to the extent that reconciliation is re-establishment of trustworthiness.   (3.4.2) Margaret Walker Walker (2006) proposes a similar account of reconciliation (or “moral repair”, as she calls it). She argues that wrongs damage our “confidence in shared moral standards, trust in our responsiveness to them and responsibility under them, and hope that our confidence and trust are not misplaced”; reconciliation, on the other hand, is the restoration of such confidence, trust, and hope (191-192).  Walker treads a path that is quite different from most mainstream forgiveness theorists. She steers clear of the traditional necessary-and-sufficient-condition approach, opting instead to explore three common senses of forgiveness (forswearing resentment, restoring relationships, and fixing a wrong in the past). She therefore recognizes that forgiveness, in the second sense, is “essentially about restoring relationships” (158): it “[restores], or [returns] to a functioning state, the conditions of moral relationship” and “[affirms] values and standards (the boundaries) as shared among those with whom we deal, [stabilizes] trust in ourselves and others to be responsive to those standards, and [restores] or instill a hopeful view of our moral values, ourselves, and each other” (162). Her discussion also shows attention to personal relationships and her account of reconciliation applicable to them.  Walker’s discussion has its limitations. Most theorists would agree that forgiveness constitutes reconciliation in some sense and in some cases. The challenge is to explain what forgiveness is in a way that can explain its reconciliatory nature. However, Walker’s main concern is to explore reconciliation/moral repair and how forgiveness may contribute to it, rather than to provide an account of forgiveness and explain how forgiveness constitutes reconciliation. (As she puts it, her approach is to “look at forgiveness through the lens of moral repair” (167).) So while she is right to 	 87	say that there are some cases where forgiveness amounts to reconciliation, her analysis may not to able to help us tackle the challenge we have in hand.   (3.5) Two Mainstream Accounts: No Reconciling with Reconciliation The above discussion shows that reconciliation in both restoration sense and trustworthiness-reestablishing sense cannot explain the reconciliatory nature of forgiveness. Are there any other meaningful senses of reconciliation on the table?   I want to turn the discussion back to the two mainstream accounts for the time being. By getting to the roots of the two accounts, I will demonstrate that the two accounts are fundamentally ill-equipped to explain the idea that forgiveness constitutes (any meaningful sense of) reconciliation. In other words, whatever meaningful sense of reconciliation we come up with (if there is one), the two mainstream accounts will not be able to say that, on their accounts, forgiveness constitutes that sense of reconciliation. They can at best explain a minimal sense of reconciliation, a sense that is too minimal to account for reconciliation in most interpersonal relationships, particularly personal ones. If this is right, it would support my main argument of this section: that the two mainstream accounts fail to fully explain the second important feature of the interpersonal aspect of forgiveness—forgiveness constitutes reconciliation.  The two mainstream accounts each starts with a “bare-bone idea”. The standard account starts with the idea that forgiveness involves ceasing blame (for moral reasons), and Change-in-View Account starts with the idea that forgiveness involves changing from seeing the wrongdoer as a wrongdoer to something more benign. Both accounts accurately point out some crucial components of forgiveness, and proponents of these two accounts try to build their accounts by fleshing out these two bare-bone ideas. For instance, Murphy spells out the first idea in terms of the overcoming of resentment, and Hampton spells out the latter in terms of moral enemy. As I argued in sections 1 and 2, they have had only limited success on this task.   The problem is, in whatever way theorists spell out these two bare-bones ideas, it is hard to imagine that “ceasing blame” and “not seeing the wrongdoer as a wrongdoer” could be substantial enough to explain forgiveness as constituting a meaningful sense of reconciliation. One might say that not blaming the wrongdoer or not seeing her as a wrongdoer is in some sense “reconciling” the 	 88	relationship with the wrongdoer, but doing these two things amount to little more than not hating, relating, and/or seeing the wrongdoer in a negative way. Such a minimal sense of “reconciliation” is far from the meaningful sense of reconciliation in interpersonal relationships (particularly in personal relationships) that we are looking for.   Such a minimal sense of reconciliation might be sufficient for reconciliation of forgiveness in distant non-personal relationships (e.g. the “paradigmatic moral relationship”). Distant non-personal relationships often lie in respecting basic moral expectations (e.g. refraining from harming one another) and holding minimal good will, so overcoming blame emotions may be a significant step towards a normalized non-personal relationships. Similarly, not seeing the wrongdoer as a wrongdoer may also be a significant step, particularly if we adopt Hampton’s account and understand it as “changing from seeing the wrongdoer as a ‘moral enemy’ to ‘still decent’” as Hampton argues. This is because seeing the wrongdoer as a “moral enemy” shows that some serious wrong has been committed and the relationship is in very bad shape. Changing from this perception to seeing the wrongdoer as “still decent” is surely significant in the relationship.   By contrast, personal relationships are much deeper and more complex than non-personal relationships. We would not say we have in any meaningful sense reconciled with our friend if we have merely overcome our resentment towards her and/or we no longer see her as a wrongdoer. Reconciliation in personal relationships, even in a weak but non-minimal sense, would require more than these. For these reasons, I am sceptical that the two mainstream accounts could provide a meaningful sense of reconciliation that can explain the way in which forgiveness constitutes reconciliation.    (3.6) Performative Account: A New Way? I noted at the very beginning of this section that it is a challenge to specify the elusive concept “reconciliation”, not least to do so in a way that can explain that forgiveness constitutes reconciliation. I then explained how forgiveness theorists, using two strong senses of reconciliation, have failed to do so. How should we move forward from here?   As I have been saying, the cause of our failure may not lie in the idea that forgiveness constitutes reconciliation but in us—that we have not come up with a good enough account of reconciliation—so 	 89	we should try hard and find a better account. But it may be helpful to try to see the bigger picture. As I argued in section 2.1, I aim to explain forgiveness as an interpersonal practice, focussing on how it functions in and shapes relationships, rather than looking for its necessary and sufficient conditions. Perhaps theorists’ failure to explain the reconciliatory nature of forgiveness lies in the methodology that they use: looking for the necessary and sufficient conditions of forgiveness and reconciliation, and then explain whether forgiveness constitutes reconciliation. This methodology may be wrong headed. As Strawson reminds us in Freedom and Resentment: philosophical investigations should start from our lives, not from theories. If our moral experience informs us that forgiveness constitutes reconciliation, then our theorizing should be guided by it and our aim should be to find a sense of reconciliation that can explain our experience. More generally, we should refocus on the questions “how do we forgive?” and “how does forgiveness shape our relationships?”. These questions take our discussion to the performative account of forgiveness that I am going to propose in the next chapter. This performative account provides a sense of reconciliation that is weaker than the two senses discuss above, but is yet a meaningful sense that can explain the idea that forgiveness constitutes reconciliation.   But before I give my performative account of forgiveness, I should explain the third interpersonal feature of forgiveness—performativeness.   	 90	Section 4 Feature Three: Forgiveness and Performative The third feature of the interpersonal aspect of forgiveness that I am going to discuss is that forgiveness is performative—that forgiveness “does” something to the relationship. I will first explain how I understand the performative nature of forgiveness and why the two mainstream accounts fail to explain this feature (4.1). Theorists of the two mainstream accounts might dispute my claim and argue that their accounts could explain the performative nature of forgiveness. From (4.2) to (4.4), I will discuss and reject three possible replies that they may give.    (4.1) Forgiveness as Performative  I argued in the last section that forgiveness brings about reconciliation (in a sense yet to be specified). But how exactly does forgiveness bring about reconciliation? It might seem magical that this thing called “forgiveness” could somehow interact with relationships in such a way that it reconciles them. But it isn’t magic; it is the third interpersonal feature of forgiveness: it is performative. If we reflect on our moral experience of forgiveness, we will find that reconciliation does not magically happen when there is forgiveness. Instead, forgiveness directly reconciles the relationship. This is the performativeness of forgiveness.   To see what it means that forgiveness is “performative” and can “do things” to relationships, consider the performative of promising. If I say “I promise to give you $20 tomorrow”, my saying these words performs an act of promise-making: I bind myself to giving you $20 tomorrow. Or consider debt-cancellation. If you say “don’t worry about it”, your saying these words performs an act of debt-cancellation: I no longer owe you $20.   I will give a more detailed discussion of performatives in the next chapter, but the examples of promising and debt-cancellation should be sufficient to show the sense in which forgiveness can “do things” to relationships. Just as you and I can change what obligations I am under in virtue of acts of debt-cancellation and promising, so too we can reconcile relationships in virtue of the offering of forgiveness. Forgiveness is performative in the sense that the offering of forgiveness (e.g. by saying “I forgive you” or by giving an understanding nod) can by itself change the relationship—that it becomes reconciled.   	 91	The two mainstream accounts cannot explain the performative nature of forgiveness for the simple reason that they conceive forgiveness as private. If forgiveness is an entirely private matter (i.e. consisting in having/not having some perception or emotions), then forgiveness would not be able to “do” anything to the relationship because it is hard to imagine how something that is entirely inside our head could “do” anything to relationships in the external world, not least to bring out reconciliation. Forgiveness needs to be expressed “out in the open” if it is to “do” anything to the relationship.   Theorists of the two mainstream accounts might reply that their accounts could explain the performative nature of forgiveness via other channels. I will discuss these possible replies and their problems in the rest of the section.   (4.2) Reply 1: Emotions and Perceptions as Relationships Proponents of the two mainstream accounts might argue that their accounts are able to explain the phenomenon that forgiveness “does” something and reconciles relationships—it does so via the constitutive features of relationships. Forgiveness as emotions or perception can bring about changes to relationships to the extent that emotions and perceptions constitute relationships. Holding blame emotions (e.g. resentment) and seeing the wrongdoer as a wrongdoer constitute bad relationships. But if we forgive someone by overcoming blame emotions or seeing him not as a wrongdoer, then the relationship will be improved. Forgiveness therefore can do things to relationships.   I do not think that this explanation via the constitution of relationship could defend the two accounts. First, this reply seems to misconstrue the sense in which forgiveness “does” things to relationships. It conceives the changes to be via the constitution of relationship: forgiveness consists in something—perceptions or emotions, and this thing is constitutively linked to relationships. But the sense of “doing” in question is more direct. It is that forgiveness, in virtue of being given, directly brings about changes to relationships.   Second, it endorses an account of reconciliation that has already been rejected in section 3.4. Overcoming of blame emotions and seeing the wrongdoer as something more benign can only give us a minimal sense of reconciliation, falling short of the robust sense of reconciliation that we need to adequately explain the reconciliatory nature of forgiveness in personal relationships. Therefore, 	 92	proponents of the two accounts cannot use such emotions or perceptions as constitutive features of relationships to adequately explain the performative nature of forgiveness.   Third, and most importantly, this reply cannot fully explain the performativeness of forgiveness even if we set aside the above two problems. To see this, consider a case where the victim no longer resents the wrongdoer nor does he see her as a wrongdoer. Proponents of the two mainstream accounts would say that the victim has forgiven the wrongdoer and the relationship has been reconciled. Forgiveness has done everything it is supposed to do to the relationship and there is nothing left. But it does not seem right. Something morally significant happens to the relationship when the victim publicly offers his forgiveness to the wrongdoer, even though before that moment he no longer bore any grudges. This morally significant thing is what forgiveness “does” to the relationship, over and above the emotions and perception that the forgiver already has or hold.   (4.3) Reply 2: Expressions of Emotions and Perceptions  Proponents of the two mainstream accounts might have a second reply to my claim that their accounts cannot explain the performative nature of forgiveness. They might argue that it is not forgiveness itself that is performative, but is instead the expressions of forgiveness that are performative. They might then argue that their accounts could say that the expressions of forgiveness, e.g. in behaviour, could do things to relationships.   There are two possible ways that theorists could use expressions of forgiveness to develop this reply. First, expressions of forgiveness reveal the emotions and/or perceptions that the forgiver holds inside her head; such revelations, by themselves, constitutes changes of relationship. However, I find it hard to understand how revealing emotions and/or perceptions can by itself constitute changes of relationship; they may merely be able to give rise to other effects which in turn change the relationship.   This brings us to the second, more plausible way—expressions of forgiveness play an instrumental role in changing the relationship. Revelations of emotions and/or perceptions may communicate to the other party what emotions and/or perceptions one has, changes the dynamics of the relationship, and brings about reconciliation. Suppose my friend Marie has stood me up and I blame her initially. But after a while I start to treat her differently, say I start to invite her to lunch again. Such 	 93	behavioural changes may reveal to her that I no longer resent her. My friendship with Marie may be reconciled when she sees from my behaviour that I no longer blame her; or Marie may be encouraged by such knowledge to work towards reconciling our friendship.   Using the expressions of forgiveness to explain the performativeness of forgiveness has a few problems. First, we seem to think that it is forgiveness itself, rather than its expressions, that is performative. We wouldn’t say “the expressions of my forgiveness has reconciled my friendship with Marie”, but would say “my forgiveness has reconciled it”. Theorists of the two mainstream accounts might counter that our linguistic intuitions are not always reliable. “My forgiveness” may well be a shorthand for “the expressions of my forgiveness”. Theorists might then claim that their accounts would be unproblematic once we clarify this linguistic confusion. They might even argue that their accounts is faithful to the true meaning of our linguistic practices surrounding forgiveness and can contribute to the effort of clarifying our practices.   We might have different intuitions about whether it is forgiveness itself or expressions of forgiveness that are performative. But even if we concede that two mainstream accounts could use expressions of forgiveness to explain the performativeness of forgiveness, the two mainstream accounts would face a more serious and fundamental problem: they would fail to adequately recognize the importance of the performativeness of forgiveness. They would be taking emotions or perceptions as the core of forgiveness and give them a central role in their accounts, while treating the expressions of forgiveness (e.g. how one treats the forgiven person) as merely “extensions” of forgiveness with a secondary role. They would in turn give the performativeness of forgiveness—explained in terms of the expressions of forgiveness—a mere secondary role in their accounts.34 But the performativeness of forgiveness, and indeed the interpersonal aspect of forgiveness in general, plays an important role in our practice of forgiveness. We need an account that recognizes its importance and gives this feature a central role in the account—whether it says it is forgiveness itself or its expressions that have this feature. The two mainstream accounts fail to do so.   The two mainstream accounts might change their approach and regard both the private (emotions and perceptions) and the public (expressions of forgiveness) as central to their accounts, and 																																																																		34 This core-extension division may ultimately be due to the necessary-and-sufficient-condition approach of the two mainstream accounts. 	 94	therefore avoid regarding the performativeness of forgiveness as having mere secondary importance. This move to the direction of the interpersonal approach would indeed be welcomed, but the problem is that this move would fundamentally transform the two accounts, as I argued in section 2.3.1. By recognizing the importance of expressions and communication of forgiveness—an important feature that I discussed in section 2—the two accounts would incorporate public elements and the resulting accounts would likely be hybrid of private and public. Whether or not the two mainstream accounts are flexible enough for such fundamental changes, such a move would strengthen my argument that the private approach needs to be at least complemented by the interpersonal approach to forgiveness.   (4.4) Reply 3: Insincere Forgiveness?  Proponents of the two mainstream accounts might have a third reply. The claim that the mere offering of forgiveness (e.g. by saying “I forgive you”) can change the relationship implies that a person may “truly” forgive someone while still bearing grudges (e.g. being resentful). In other words, “insincere forgiveness”—insincere in the sense that the person is still resentful—could be an “authentic”, “real” form of forgiveness. This is an undesirable implication, proponents of the two mainstream accounts would argue, because it does not seem right that a victim would have truly forgiven the wrongdoer if the victim is still resentful. We can at best call insincere forgiveness non-ideal forgiveness. This performative account, however, cannot make such a claim. It must accept insincere forgiveness as an authentic, ideal case of forgiveness.   I think that we have conflicting intuitions on this point. On the one hand, someone who still bears grudges does not seem to have truly forgiven the wrongdoer. On the other hand, forgiveness that is publicly given is in some sense authentic forgiveness and can change the relationship in question even if the person offering the forgiveness still bear grudges. Consider apology as a comparison. Even if the person does not really feel apologetic deep inside—and she may be appropriately blamed for this reason—her public apology is a public acknowledgement of the wrong that she has done to the victim and can in some sense change the relationship between them.   I believe that the performative account is able to say that some cases of “insincere forgiveness” are not authentic, real forgiveness. The key lies in properly recognizing the relationship-dependency of forgiveness. To put it simply, forgiveness requires “sincerity” in personal relationships but not in non-personal relationships such that insincere forgiveness is not real forgiveness in personal but is real in 	 95	non-personal relationships. We should also recognize that forgiveness should not require the disappearance of the “last ounce of anger” (Haber 1991, 7) but only an aspiration to do so. I will develop these two points in greater detail in the next chapter.   (4.5) Summary I have argued in this and the past two sections that the two mainstream accounts cannot fully explain three important features of forgiveness: communicativeness, reconciliatoriness, and performativeness. I have also considered and rejected some potential objections that theorists of the two accounts might raise. While theorists might be able to come up with other ways to defend and/or revise their accounts in order to better explain the three features, the performative account of forgiveness that I am going to propose in the next chapter can explain these three features of forgiveness in ways better than the two accounts.  	 96	Chapter Conclusion  This chapter has demonstrated the inadequacy of the two mainstream accounts. First, they inaccurately conceptualize forgiveness in personal relationships; second, they fail to explain the interpersonal aspect of forgiveness, namely, that forgiveness is communicative, reconciliatory, and performative. While theorists might be able to “personalize” the two mainstream accounts and solve the first problem, the two accounts are fundamentally incapable of explaining the three interpersonal features, which are crucial to explaining forgiveness as a normative practice situated in (inter)personal relationships (in relation to how it functions in relationships.)    We therefore need to adopt the interpersonal approach that focusses on explaining the interpersonal aspect of forgiveness. In the next chapter, I will propose an account that adopts this approach—the performative account of forgiveness.   	 97	Chapter 3 Performative Account of Forgiveness After arguing in the last chapter that the two mainstream accounts (the standard account and Change-in-View account) fail to adequately explain forgiveness in interpersonal relationships, I am going to propose in this chapter my performative account of forgiveness. The central idea of this account is that forgiveness is an act declaring that the relationship between the victim and the wrongdoer is no longer in a broken state. It focusses on explaining the interpersonal aspect of forgiveness (i.e. forgiveness is performative, communicative, and reconciliatory) and can thus adequately explain forgiveness as a normative practice in interpersonal (particularly personal) relationships. It can at least complement the two mainstream accounts, if not partly replace them.  I present the content of my performative account in section 1. It begins with a brief discussion of John Langshaw Austin’s speech act theory (1962), which forms the backbone of this account. I will then explain three strengths of this account. Section 2 explains its main strength: it can explain the interpersonal aspect of forgiveness and avoid the depersonalization problem. In section 3, I distinguish two kinds of forgiveness: recognition and proleptic forgiveness. I argue that the latter has largely been neglected but should be taken seriously, and my performative account is better equipped than the two mainstream accounts to explain this latter form of forgiveness. Section 4 shows that the performative account can defend the normative value of forgiveness against two sceptical arguments—Aurel Kolnai’s paradox of forgiveness (1973) and Martha Nussbaum’s recent critique of forgiveness (2016).   For the sake of clarity, I want to reiterate one methodological point made in the last chapter. My discussion of forgiveness does not aim to explain the necessary and sufficient conditions of forgiveness; rather, it aims to explain features that are crucial to forgiveness as an interpersonal practice in terms of how it functions in and shapes (inter)personal relationships. Therefore, my performative account is not saying that all and only cases of forgiveness is a declarative act that unbreaks the relationship—for instance, completely private forgiveness may not be a declarative act (discussed in section 2.1); it instead argues instead that this understanding of forgiveness can explain features that are crucial to forgiveness as an interpersonal practice.  	 98	Section 1 Applying Speech Acts to Forgiveness  I am going to present my performative account of forgiveness in this first section of the chapter. In (1.1), I explain how speech acts can be (and have been) used to explain forgiveness. On my performative account, forgiveness is a declarative act—it declares that the relationship is no longer in a broken state. One main contentious point of this performative account is whether the overcoming of emotions (e.g. resentment) should a success condition of the forgiveness performative. The concern is that the overcoming of emotions deals with something “internal”, whereas it seems that whether the forgiveness is successfully performed should be determined only by what is external, e.g. behaviour. I discuss this concern in (1.2). In (1.3), I clarify that the success condition in my account should be framed in terms of “overcoming blame” rather than “overcoming emotions” because, as I argued in chapter 1, blame in personal relationships may involve a rich set of emotions whereas blame in non-personal relationships may not involve any affect. I then address the concern about the internal success condition. In particular, I argue that there are internal success conditions of performatives in personal relationships (e.g. having or not having certain emotions), and hence I defend the internal success condition—successful performance of forgiveness in personal relationships requires overcoming blame emotions—of my performative account.   (1.1) Speech Act Theory and the Performative Account of Forgiveness  The discussion of my performative account of forgiveness should start with a brief discussion of a speech act theory. John Langshaw Austin observes in How To Do Things With Words (1962, 98-102) that we use language to perform a wide variety of linguistic acts. When we utter a sentence, for instance “my glass is empty”, we may not be merely making a statement that is assessable as true or false. We may be simultaneously performing some other linguistic act, for instance, asking a friend to pass me the water jar. Austin therefore proposes that there are three distinctive kinds of speech acts: locutionary, illocutionary, and perlocutionary acts. Roughly speaking, locutionary acts are acts of uttering something, for instance uttering the words “my glass is empty”, or “I promise to pay you back $20 next week”. Illocutionary acts are acts that performed via locutionary acts (that is, by uttering some words). For instance, when I utter the sentence “my glass is empty” to a friend at a dinner party, I may be doing more than just uttering such words—I may be asking my friend to pass me the water jar. The latter act—asking my friend to pass me the water jar—is an illocutionary act (“directive”) performed by my uttering “my glass is empty”.  	 99	There are other kinds of illocutionary acts. If I say “thanks for the lunch Espen!”, I am performing a “behabitive” (or “expressive”) act of expressing my gratitude to my friend Espen for treating me to the lunch by saying these words. If I say “I promise to pay you back $20 next week”, I am performing a “commissive” act of making a promise and now owe it to you to pay you back $20 next week. And when, hearing that I promise to pay you back next week, you replied, “don’t worry about it”, you are performing a “declarative” act that exerts your powers and rights to cancel my debt. Similarly, if I say “I name this boat the Fearless”, I am performing a declarative act of christening: I am declaring that this boat is from now on named the Fearless.   Perlocutionary acts deal with the further consequential effects that are brought about by the performance of an illocutionary act. For instance, by saying “my glass is empty”, I make my friend pass me the water jar; by telling my sister my detailed travel plans, I persuade her I am ready for the trip.   There is nothing new about using speech acts to explain forgiveness. A small number of theorists have used this approach and proposed “performative accounts” of forgiveness. Broadly speaking, the performative account argues that saying “I forgive you” (verbally or non-verbally) is performing an illocutionary act that brings about some effects in virtue of the utterance. Which specific kind of illocutionary act forgiveness performs and what effects (as per “perlocutionary acts”) it brings about depends on the particular account. Richard Swinburne (1989) proposes that forgiveness is a declarative act that “removes the guilt” of the wrongdoer in response to the wrongdoer’s repentance, reparation, apology, and ultimately wrongdoing (85). Joram Haber (1991) argues that forgiveness is an act of revealing that one has overcome or is willing to try to overcome one’s resentment for the wrongdoer’s wronging one (a “behabitive” act) (40). Building on Haber’s account, Glen Pettigrove (2012) proposes that forgiveness, in its “highest manifestation”, is an act of revealing that one has overcome hostile reactive attitudes towards the wrongdoer (“behabitive”) and an act of committing to overcoming such attitudes (“commissive”) (18).   Brandon Warmke (2016a) rejects both Haber’s and Pettigrove’s accounts and sides instead with Swinburne. Warmke argues that forgiveness is a “declarative” act that alters the norms between the wrongdoer and the victim. It alters victim norms such that the victim no longer has the right to blame the wrongdoer, demand apologies and restitution, etc.; it also alters wrongdoer norms by releasing 	 100	the wrongdoer from certain personal obligations owed to the victim, e.g. apology, repentance and remorse, restitution, and penance. Dana Nelkin (2013) has proposed a similar account where forgiveness releases the wrongdoer from a special kind of personal obligation incurred because of the wrongdoing (175).   The performative account of forgiveness that I am proposing takes forgiveness as a declarative act. However, unlike Swinburne and Warmke, I argue that saying “I forgive you” is an illocutionary act of declaring that the relationship is no longer in a broken state35, and the effect that such an act brings about (i.e. the “something” that forgiveness makes happen) is an unbroken relationship. Forgiveness is in this regard similar to debt cancellation and promise making: a debt is cancelled simply in virtue of the creditor’s declaring that the debtor no longer needs to pay him back, and a promise is made simply in virtue of the promiser’s saying “I promise”. Likewise, forgiveness is given and the relationship unbroken simply in virtue of the victim’s saying “I forgive you”, either verbally (e.g. by explicitly saying it) or non-verbally (e.g. as Swinburne observes, with an understanding smile (1989, 85)), The offer of forgiveness enables both the victim and the wrongdoer to “wipe the slate clean”, “get past the wrong”, and perhaps “start over” the relationship. The victim has a right to make such a declaration about the state of the relationship in virtue of her status as the victim. What exactly the state of the unbroken relationship is following the declaration will depend on the circumstances and the negotiation between the victim and the wrongdoer, but it will at least be a neutral, non-negative, and hence improved state.   (1.2) Is Emotion Necessary? One contentious point about the performative account is the role of emotion. Debts can be successfully cancelled and promises be successfully made without any emotion, but it seems that forgiveness is not truly given unless the victim has (for instance) overcome resentment. If performatives do not require emotion but forgiveness does, then performatives will not be able to explain forgiveness.   																																																																		35 As I stated in a footnote in chapter 1 section 1.1, my discussion focusses primarily on significant and non-significant-yet-non-trivial wrongs. Significant wrongs may but need not “break” a relationship—they may merely “damage” it, non-significant wrongs may merely “strain” a relationship. In such cases, the relationship need merely be “undamaged” or “unstrained”.  I will continue to use the terms “break” and “unbreak” as catch-all phrases for the sake of simplicity. My account of forgiveness covers also cases where forgiveness is merely to “undamage” and “unstrain” the relationship. 	 101	Austin’s speech act theory can help us better understand this contentious point and the role of emotion. Austin argues that there are two ways in which a performative utterance can be “infelicitous”, which can be roughly understood as “improperly performed” (13-18). First, a performative utterance is infelicitous if it “misfires” (i.e. it is “unsuccessfully performed”). When a performative utterance misfires, the act will not bring about the effect that it is supposed to bring about. For instance, on the marriage altar, if instead of saying “I take you as my lawful wedded wife” the fiancé says “I take you as my awful wedded wife”,36 then his performative act of taking a marriage vow misfires because he did not make the correct utterance. Or suppose that the fiancé is not entitled to take the marriage vow because he is already married to another person. His performative utterance will also misfire. In both cases, the performative acts are unsuccessful and the fiancé are not yet married to his fiancé(e).  Note that misfire conditions on Austin’s account have nothing to do with anything internal—they do not deal with whether the speaker utters the performative sadly or solemnly. It is the second kind of infelicitous conditions, “abuse” conditions, that deal with the speaker’s emotions, attitudes, and intentions. A performative utterance is “abused” if the person making the utterance fails to have the emotions, attitudes, or intentions that people who make the utterance should have. For instance, if a person says “thanks for the lunch” but does not really feel grateful, then the utterance, which is used to express gratitude, has been “abused”. Or if someone says “I promise to pay you back next week”, but in fact has no intention to do so, then the utterance has been abused. However, even though the performative utterances in these two examples have been abused, the performative acts are successfully performed and their corresponding effects brought about. Consider the second example, for instance. Even though the speaker does not intend to pay back the money, the promise is successfully made and its effect brought about—she is now under the obligation to pay back the money in a week.  If we followed Austin’s understanding of the two conditions of infelicity, then the lack of emotion would be an “abuse” rather than a “misfire” condition of forgiveness. In other words, even if a forgiver said “I forgive you” without having overcome resentment, the performance of forgiveness would be successful37 so long as he made the performative utterance correctly and were entitled to 																																																																		36 I borrow this comical line from Rowan Atkinson in the movie Four Weddings and a Funeral (Newell 1994). 37 I will use “misfire”, “success”, and “necessary condition” interchangeably from this point on. 	 102	forgive. The forgiver would have merely abused the forgiveness utterance because he failed to have the emotion that people who make the utterance should have, namely the overcoming of resentment. Such abused utterances of forgiveness would be “insincere” but nonetheless “authentic” and “real” in the sense that they were successful performative acts and would bring about the intended effect, namely unbreaking the relationship.   Critics would then argue that this is not right. Whatever forgiveness is, the overcoming of resentment (or other blame emotions) should be a success condition for forgiveness such that “insincere forgiveness” is not authentic, real forgiveness at all. A resentful victim cannot be properly said to have forgiven the wrongdoer. If the performative account of forgiveness says emotion is not a success condition, then this account itself is unsuccessful. Haber (1991), a proponent of the performative account of forgiveness, thus concedes that this is the point where his account of forgiveness departs from Austin’s speech act theory. He argues that emotion should be a success condition such that forgiveness is successfully given only if the forgiver has overcome or at least is committed to overcoming resentment (1991, 51).   (1.3) Emotion and Relationships There does seem to be something right about the claim that, if one has not overcome resentment and other blame emotions, one would not have forgiven the wrongdoer no matter how solemnly one says “I forgive you”. Successful performance of forgiveness, on my performative account, should therefore require the overcoming of blame (or at least the aspiration to overcome blame). In the following, I will spell out the details of and defend this success condition.    (1.3.1) Forgiveness as Overcoming Blame The question should not be framed as whether the overcoming of resentment (or even the overcoming of resentment and other blame emotions) is a success condition for forgiveness; rather, it should be framed as whether the overcoming of blame is a success condition for forgiveness. As I argued at length in chapter 1, whether blame is affective is dependent on the relationship context. Because of the importance of emotion in personal relationships, blame is and should be affective in personal relationships; in non-personal relationships, by contrast, blame may not involve any affect but only some conative elements such as desires, intentions, and expectations. If blame may be affectless in non-personal relationships, then emotion is not a success condition for forgiveness in non-personal 	 103	relationships for the simple reason that there may be no blame emotions to begin with. Similarly, blame in personal relationships involves a rich set of emotions (e.g. resentment, disappointment, sadness, and hurt feelings.), whereas in non-personal relationships blame involves only resentment. Forgiveness in personal relationships therefore involves the overcoming of this rich set of emotions, whereas in non-personal relationships it deals only with resentment (when it is affective).   I have also argued in chapter 1 that blame in personal relationships involves a rich set of emotions: not just resentment but also disappointment, sadness, and hurt feelings. In non-personal relationships, however, blame may involve only resentment. Hence, a richer set of emotions—resentment and other blame emotions—may be relevant to forgiveness in personal relationships but not relevant to that in non-personal relationships.   Given these two affective differences between blame in personal and non-personal relationships, the question under discussion should therefore be whether the overcoming of blame is a success condition for forgiveness. The question should then be whether forgiveness, in personal relationships, requires one to have overcome resentment and other blame emotions and, in non-personal relationship, requires one to overcome resentment or conative elements such as desires, intentions, and expectations.  With the question clarified, is overcoming blame a success condition for the forgiveness performative? My answer is yes, with qualifications. It does seem strange to say that one had forgiven a wrongdoer if one continued to resent or desire to seek revenge on the person. This is particularly true in personal relationships because of the importance of emotion in such relationships.   (1.3.2) Forgiveness Without Internal Conditions?  There are mainly three worries about my claim that overcoming blame is a success condition and I will address them in the following. First, while making the overcoming of blame a success condition will make the “internalist” happy, theorists who favour an “externalist” approach—that forgiveness does not require any internal conditions—would likely be not as happy about this move. They might cite cases where merely saying “I forgive” publicly seems sufficient to change the relationship, even if the forgiver says it grudgingly and still has “hard feelings”.   	 104	I think such cases are mostly cases in non-personal relationships, notably minimal relationships such as acquaintances, infrequent or one-off business associates, or political relationships. In such relationships, forgiveness might be given simply by saying “that’s fine”, whether or not the person is still angry or desires revenge. Saying “that’s fine”, sincerely or not, had a moral significance in these minimal relationships such that it would be sufficient to reconcile the relationship. It might be most obvious in political relationships. Two warring communities could reconcile in virtue of a symbolic handshake between their leaders, even if their community members were still resentful of one another.   But are these cases really forgiveness? I do not think that it is clear that they are. The last case is particularly shaky. The relationship between two warring communities is a political relationship, so it may be more appropriate to say that the symbolic handshake can bring about a kind of reconciliation that is different from the kind that is brought about by forgiveness. Even if it did bring about forgiveness, it would be a political kind of forgiveness which is in many ways different from the interpersonal, non-political kind that we have been discussing. Similarly, in the other two non-political cases, merely saying “that’s fine” may be able to bring about some kind of reconciliation, but it may not be the kind that forgiveness brings about. It might be more appropriate to call it “acceptance” rather than forgiveness.   (1.3.3) Overcoming: Complete or Partial?  The second worry deals with whether forgiveness requires a complete or partial overcoming of blame. What if someone has not fully overcome her anger (or desire to revenge etc.) at the wrongdoer but is working on it? Has she forgiven the wrongdoer? Or do we need to wait till she has fully overcome her anger (or desire to revenge etc.) before we can rightly say that she has forgiven the wrongdoer?   We cannot make anger (and many other emotions and desires) disappear by simply clicking our fingers. Overcoming them often takes time and effort; sometimes we may not even know whether it is achievable. If an account of forgiveness did not allow for the stubbornness of our emotions and other attitudes, and required instead the full and complete elimination of blame, it would rule out a considerable amount of cases which we would otherwise call forgiveness. A good account of forgiveness should therefore adopt a broader understanding of “overcoming blame”. Overcoming 	 105	blame should be understood as either the full elimination of blame or an aspiration to eliminate blame;38 that is, one is willing to work on (and continue to work on) eliminating blame and is ready to disown one’s blame if it comes up. This broader understanding of overcoming blame can explain and accept as forgiveness cases where one has forgiven but some residual blame remains, and can put this second worry to rest.   (1.3.4) Internal Conditions as Success Conditions?   The third worry is the most important among the three and requires an in-depth discussion. This success condition of forgiveness—the overcoming of blame—makes the forgiveness performative very different from other performatives in Austin’s account in that none of them has any internal conditions as success conditions (e.g. A false promise is a promise and gives rise to an obligation, despite the speaker’s lack of intention to honour the promise). Would the forgiveness performative be a performative after all, given this difference? More importantly, does it really make sense to say that the forgiveness performative, or indeed any performative, needs to be “sincerely” performed to change the relationship? Isn’t it quite obvious that, for example, a false promise can give rise to a new obligation and change the relationship without being sincere? Why should the forgiveness performative be any different?   By taking an internal condition to be a success condition of the forgiveness performative, my performative account, like Haber’s account, has indeed departed from Austin’s speech act theory. Influential though Austin’s theory has been, departing from it is not as such a problem; what matters is whether this departure is a reasonable one. Hence the main question lies in whether it makes sense to say that performatives require the truth of some internal condition to successfully bring about some effects and change relationships. I will argue in the following that some changes of relationships can be made only if we hold certain thoughts, intentions, or emotions, and the change that the forgiveness performative makes—unbreaking relationships—is one of such changes.  																																																																		38 Haber and Griswold use the term “committed” to describe this phenomenon (Haber 1991, 7; Griswold 2007, 42), but I think “aspire” is a better term. We cannot commit ourselves to doing things that we know we cannot achieve. So we cannot commit ourselves to overcoming anger if we know we will never succeed. If we say forgiveness is commitment to overcoming anger, then there would be no forgiveness if one knows that one can never successfully overcome one’s anger.   This seems to be a wrong result. We have forgiven someone so long as we are willing to keep working on overcoming our anger and will disown it when it comes up, even if we know all too well that we will never achieve full success. That is why “aspiration” is a better term. We can aspire to do the unlikely or impossible, so we can aspire to overcome our anger even if we know it cannot be done. And so long as we aspire to do so, we have forgiven the wrongdoer. 	 106	To make this argument, I need to once again start from the actual workings of our interpersonal relationships. If I said to a friend “I will give you back the money tomorrow” without any intention to do so, how exactly and in what sense would my false promise change my relationship with him? Though I have no intention to pay him back, my false promise will in a sense change the relationship: I am now under the obligation to pay him back tomorrow, I will wrong him if I don’t, and he will have reason to blame me for not paying him back.   But such changes deal only with the formal aspect of the relationship—what obligations formally hold in our friendship. Relationships, particularly personal ones, would be overly legalistic if this formal aspect were all that there was with relationships. The point I am making is not that relationships involve things other than obligations. The point I am making is that, even in regard to obligations, relationships are much deeper and richer such that obligations have a substantive aspect. Suppose my friend, knowing my personality all too well, knows that I won’t give him back the money; and I, knowing my notoriety with loans all too well, know that he knows I won’t pay him back. Does my false promising change our relationship in relation to debt obligations? And in what sense? In a formal sense, it gives rise to an obligation on my part to pay him back. But in a substantive sense, it does not. I won’t honour the obligation, my friend knows that I won’t honour the obligation and does not expect me to pay him back, and I, knowing what my friend knows, know that he won’t expect me to honour the obligation. There is a “mutual lack of expectation” such that the formal obligation about my debt is not operative in the relationship, and my false promising fails to bring about a substantive change of the relationship.   One might agree that the relationship in this case has not substantively changed but disagree with my explanation. One might argue that it isn’t my lack of intention to pay back the loan that gives rise to the failure of the performative; rather, it is how my friend and I understand the relationship, i.e. our “mutual lack of expectations”, that gives rise to the failure. Consider a slightly different scenario: suppose I am a skilful liar and do not have a bad track record of loan repayment, and I deceive my friend into thinking that I sincerely mean to give him back the money tomorrow. My false promising seems to change the substantive obligations in our relationship—my friend expects me to pay him back and I know that he expects me to, even though I have no intention to do so. Hence, the internal condition—my intention to pay back to loan—is not a success condition.   	 107	But why does my friend in this case expect me to pay him back? Some might say that it is because I appear to be sincere and intend to pay him back, but a more reasonable explanation is that he thinks I am actually sincere and intend to pay him back when I make my promise. If he knows later that I have never intended to pay him back, then he will think that, to put it a bit dramatically, the relationship has never been what he thought it was but has been a “scam” in relation to the money and the obligation. His previous understanding of the relationship was mistaken.     If the substantive changes the promising brings about have internal requirements (e.g. intending to repay the loan), and if the change of relationship that forgiveness brings about—unbreaking relationships—are not merely formal changes but substantive changes, then the internal requirement of the forgiveness performative (i.e. the forgiver has overcome or aspires to overcome blame) is not unreasonable.   (1.4) Summary The performative account that I am proposing takes forgiveness as a declarative act that unbreaks relationships. The successful performance of the forgiveness performative requires that the forgiver has overcome blame, where overcoming blame means having eliminated or aspiring to eliminate blame, and where blame may involve a rich set of emotion in personal relationships but may involve only conative components in non-personal relationships. I have also defended the account against the objection that the success condition of the forgiveness formative should not be internal.   	 108	Section 2 Main Strength of the Performative Account: Personal and Interpersonal Forgiveness This performative account has several strengths that make it better than the two mainstream accounts. In this section, I will explain its main strength: it can adequately explain the interpersonal aspect of forgiveness (that is, the three interpersonal features that I identified in the previous chapter) (2.1 - 2.3) and avoid the depersonalization problem (2.4); therefore, it can adequately explain forgiveness as a practice situated in interpersonal relationships, particular personal ones.   (2.1) Communicativeness It is quite clear that performative forgiveness is communicative. Performative forgiveness, if it is to serve its performative function of declaring that the relationship is no longer in a broken state, cannot be kept inside one’s head but needs to be publicly expressed and communicated to the wrongdoer.   (2.1.1) Objection 1: Private Forgiveness and “Imputed Forgiveness”  However, one might object that forgiveness is not always communicative but may be kept entirely private. For instance, the victim might keep his forgiveness private because he has no way to communicate his forgiveness to the wrongdoer because the wrongdoer is dead or because the victim has lost contact with the wrongdoer, or because he does not want to bring up the past wrong so as to avoid reviving hard feelings. One might then argue that, if private forgiveness exists, then my performative account cannot explain it because the account takes forgiveness to be communicative. This objection applies also to the other two interpersonal features because private forgiveness, by being private, is also non-reconciliatory and non-performative. One might then argue that the private approach, which understands forgiveness as primarily a private phenomenon (e.g. in terms of emotion or perception) (discussed in chapter 2 section 2.2) fares better than the performative account in regard to private forgiveness.   My reply is that my performative account is able to explain the above cases of “private forgiveness”. A victim may keep his forgiveness unexpressed because the wrongdoer is dead, but at least in some cases the victim may mean the forgiveness to be expressed and would express and communicate it to the wrongdoer if she could. The forgiver may even imagine how she would offer her forgiveness and 	 109	what the relationship would become were she able to communicate her forgiveness to the wrongdoer. Such are cases of “imputed forgiveness” and can be explained by my performative account.   Let me first draw an analogy to explain. It is not far-fetched to say that we can stand in a relationship with someone with whom we have not interacted and would not interact, say a historical figure. We could not actually communicate or interact with him, but we can imagine how we would communicate and interact with him and how he would respond to us. Such relationships are different from ordinary relationships because they are unilateral and are primarily based on imagined interactions, but it would not be far-fetched to call them relationships. They are just of a different kind.   If such relationships are not far-fetched, then it will also not be far-fetched to say that we can stand in a relationship with a long dead person or a person with whom we have lost contact, and it will also not be far-fetched to say that forgiveness given in such relationships—i.e. imputed forgiveness—is in a sense communicative, reconciliatory, and performative. Even though a victim could not possibly have any contact with his long gone mother, he can still stand in a relationship with her, just like we can stand in a relationship with a historical figure. He may imagine what would happen had there been contact with her. He may imagine what she would be like now compared with before, e.g. whether the long dead mother would be able to understand and acknowledge how much pain her neglect had caused him, how she would behave if they met, or how she would react if we forgave her. Basing on such imagination, he may forgive her mother in an imputed way through which he can unbreak and reconcile the relationship. In addition, though he would not be able to communicate his forgiveness to his long dead mother, he might imagine communicating his forgiveness to his mother and imagine how she would receive it. He might even be willing to communicate his forgiveness to his mother were he able to. His imagined relationship with his long dead mother could be unbroken and reconciled by his imputed forgiveness.  Imputed forgiveness is no doubt different from ordinary cases of forgiveness, but it is forgiveness nonetheless, just like a relationship with a historical figure is different from ordinary relationships but is a relationship nonetheless. There is not an actual, existent person whom the forgiver sees, speaks to, or interacts with in ways that one does in ordinary relationships; the forgiver can only imagine communicating with the person and standing in a relationship. Because the communication is 	 110	imagined, imputed forgiveness is communicative in a non-standard way. Similarly, imputed forgiveness reconciles the relationship through its performative power, but the relationship that it reconciles and performs on is an imagined relationship, so it is reconciliatory and performative in a non-standard way. And because imputed forgiveness is communicative, reconciliatory, and performative in a non-standard way, my performative account categorizes imputed forgiveness as non-standard forgiveness.   (2.1.2) Objection 2: Completely Private Forgiveness Proponents of the two mainstream accounts might then reply that there are cases that are not imputed but are completely private forgiveness. Completely private forgiveness does not involve even imagined communication and my performative account cannot explain it.   My reply is that the significance of completely private forgiveness for forgiveness as an interpersonal practice may be questionable. I am inclined to think the the interpersonal aspect of forgiveness matters more than its private aspect because interpersonal relationships are, well, interpersonal, but I do not wish to argue for this, nor do I think that such a debate would be very fruitful. It would be more constructive and fair to say that both the private and the interpersonal aspects of forgiveness are important, as I said in chapter 2 section 2.3. Completely private forgiveness has only the private aspect. It is kept entirely inside the head of the forgiver such that no one else knows about it, it does not involve any communication between the two people at all, even imagined ones, and it does not reconcile the relationship in any meaningful way. It may lack significance in the context of interpersonal relationships, where both the private and the interpersonal aspects are important.  (2.1.3) Objection 3: Significance of Completely Private Forgiveness Some might insist on the significance of completely private forgiveness. Even if I concede this point, this still would not cause big problems to my account. Of course the private approach can explain completely private forgiveness better—it is a private approach after all. But being better at that may not help the private approach gain many points. First, completely private forgiveness is a non-standard case of forgiveness. To the extent that forgiveness is an interpersonal practice, most cases of forgiveness will likely involve communication (at least imagined one). Completely private forgiveness, on the other hand, may not occur very frequently. So my performative account can say 	 111	that completely private forgiveness deviates from standard cases of forgiveness because of its lack of the interpersonal aspect.   Second, even if the two mainstream accounts could explain completely private forgiveness, they would still be unable to adequately explain the interpersonal aspect of forgiveness, which is an important aspect that requires an explanation—and my performative account can adequately explain it. My account would play a valuable role in complementing the two accounts by highlighting and explaining the importance of the interpersonal aspect of forgiveness, filling the theoretical gap that they have left open. So this concession would at most mean that my performative account would need to take the less ambitious path, i.e. not to replace but to complement the two accounts.   In fact, even if my performative account could not explain completely private forgiveness, it would still have partial success in explaining the private aspect of forgiveness. This is because this account incorporates some private elements as the success condition of forgiveness: forgiveness requires overcoming blame, and overcoming blame involves the elimination or the aspiration to eliminate anger, disappointment, desire to distance herself from the wrongdoer, etc., which are all private elements.   (2.2) Reconciliation Many theorists share the lived experience that forgiveness has something to do with reconciliation. The challenge, as laid out in great detail in chapter 2 section 3, is to explain why the improvement of relationship that forgiveness gives rise to amounts to reconciliation. My performative account provides a sense of reconciliation different from the two strong senses that forgiveness theorists have used. Instead of understanding reconciliation as restoration or reestablishment of trust, this account understands it in a ritualistic sense. It recognizes the importance of rituals in interpersonal relationships and how forgiveness is part of the reconciliation ritual.  To clearly explain the ritualistic role that forgiveness plays, I will start with a brief discussion of Aurel Kolnai’s “paradox of forgiveness”. In his seminal paper “Forgiveness” (1973), Kolnai observes that forgiveness is “paradoxical” in the sense that forgiveness seems either pointless or unjustified. A wrongdoer either repents and goes through a change of heart (e.g. privately acknowledging that what he did was wrong) or he does not. If he does repent and goes through a change of heart, then he is 	 112	now a different person,39 and forgiving him would be merely acknowledging his change and treating him in a way that he should be treated. Forgiving him under such conditions would then be pointless. But if he does not repent, then he has not gone through a change of heart and is still the same person as he was. Forgiving him when he has not changed would then be unjustified because doing so would amount to condoning him.  As part of his solution to the paradox, Kolnai argues that forgiving a repentant wrongdoer may not be pointless after all because forgiving him plays a necessary role in “[eliminating] the offence from the texture of [the] relationship and in that sense to ‘annul’ it” (101). Kolnai’s idea of “eliminating the offence from the texture of the relationship” is rather cryptic and underdeveloped in the paper, and he unfortunately did not have a chance to clarify and develop it after it was published due to his untimely death. It is also quite unfortunate that his idea has hardly been developed since then. Numerous forgiveness theorists after Kolnai found his paradox insightful and have proposed their solutions to it, but very few of them have discussed or addressed Kolnai’s own solution, not least to develop it.40   I believe that my performative account of forgiveness can be seen as a development of Kolnai’s idea of “eliminating the offence from the texture of the relationship”. We should first of all recognize that repentance on its own, especially if it is kept entirely inside one’s head, does not complete the process of unbreaking the relationship and “eliminating” the wrong from the relationship. Even after the wrongdoer has repented, something morally significant happens to the relationship when the victim offers to the wrongdoer his forgiveness.   My performative account can explain what this morally significant thing is—it is that the relationship is now unbroken. Forgiveness is a declaration that makes it the case and publicly communicates to the wrongdoer that the relationship is now unbroken. This public declaration can be seen as a ritualistic act conveying to the wrongdoer (and others) that the victim recognizes the wrongdoer’s repentance (if she repents), that the wrongdoer is no longer treated as a wrongdoer, that the victim is 																																																																		39 Note that one may “change one’s heart” and “become a different person” in various degrees. One might have been a morally corrupt person, then repented and went through some fundamental changes, and ended up becoming a morally decent person. Or the change might be less fundamental. One might simply have made a moral mistake (e.g. telling a non-trivial yet not big lie), then repented and became a “different person” in the sense that one now thought that the lie was wrongful. 40 The only exception, to my knowledge, is Jeremy Watkins (2005). 	 113	willing to “get past the wrong” with the wrongdoer, and that they now stand “with one another” rather than “against one another”. Forgiveness then “eliminates the wrong from the relationship” and in turn reconciles the relationship, and it is in this sense that forgiveness amounts to reconciliation.   There are cases where forgiveness does not constitute reconciliation. After a breakdown in an intimate relationship, one might not want to have any relationship with the person anymore, e.g. one might not want to see the person and would not even trust the person as one would trust a normal stranger. And yet one might forgive the person only to “release” the person from their wrong and guilt feelings. Such cases of forgiveness without reconciliation exist, and my performative account acknowledges that it cannot explain such cases. However, forgiveness is a complex and multifaceted phenomenon, and there are certainly cases that do not involve the three interpersonal features (communicativeness, reconciliatoriness, and performativeness). My account does not aim to explain all cases of forgiveness—in fact, no account of forgiveness can and should aim to do so—but aims only to explain cases of forgiveness that involve these three interpersonal features, which are, as I argue, crucial to forgiveness as an interpersonal practice.41   (2.3) Performativeness Clearly, my performative account can explain the performativeness of forgiveness. It can explain that forgiveness does not bring about reconciliation by magic but by “doing something” to the relationship, and the thing that forgiveness “does” is unbreaking the relationship between the wrongdoer and the victim in virtue of declaration.   One might argue that performativeness is not so much a strength as a weakness of the account. I have acknowledged at the end of the last section that the forgiveness performative is partly affective such that forgiveness necessarily involves the overcoming of blame emotions in personal relationships. This shows that the standard account has pointed out some necessary feature of forgiveness. One might then argue that the standard account is at least in this regard a better account than performative account.   																																																																		41 Other performative accounts, such as Warmke’s (2016a), which takes forgiveness as a performative that changes the norms between the victim and the wrongdoer, may be able to explain such cases. 	 114	My reply will be similar to that I made above (2.1). There may well be some necessary features of forgiveness in the private aspect of forgiveness (e.g. affect) that two mainstream accounts are better equipped to explain. However, the main problem of these two accounts is their over-emphasis on the private aspect of forgiveness. The interpersonal aspect is at least as important as the private aspect, and an ideal account should be able to explain both. The two mainstream accounts can only explain the private aspect, but my performative account can explain the interpersonal aspect and partly the private aspect, and is to that extent superior to the two mainstream accounts.    (2.4) Accurate Conceptualization of Personal Relationships It is also worth noting that this performative account does not commit the misconceptualization problem that I discussed in chapter 2 section 1. It does not limit itself to an abstract perception of the wrongdoer as a moral agent or a “moral enemy”, but is open to a particularized perception of the wrongdoer—as one’s friend, one’s partner, one’s child, etc. Nor does it ignore the rich emotions in personal relationships. It can accept the possibility that (i) the performance of forgiveness often and should come with emotion and (ii) it consists of a rich set of emotions that feature in personal relationships. The performative account of forgiveness is claiming only that emotions cannot alone adequately explain forgiveness in personal relationships.  (2.5) Summary Being able to explain the interpersonal aspect of forgiveness while avoiding the misconceptualization problem is not the only strength of my performative account. I will explain its two other strengths in the following two sections.  	 115	Section 3 Performative Account and the Recognition-Proleptic Distinction I will explain in this section a second strength of my performative account. It deals with an important distinction between two forms of forgiveness: recognition and proleptic. I will first introduce recognition and proleptic forgiveness in (3.1) and further explain the distinctive features of proleptic forgiveness in (3.2); then, in (3.3), I will argue that my performative account is better equipped than the private approach to explain proleptic forgiveness.   (3.1) Recognition and Proleptic Forgiveness  Discussions in the forgiveness literature have largely understood forgiveness as something that is given because of repentance or other moral reasons. One may forgive the wrongdoer because he has repented, i.e. coming to see that what he did is wrong and changing his emotions, attitudes, intentions, etc. towards his wrong accordingly. Theorists often call it “conditional forgiveness”—forgiveness that is given conditional on the repentance of the wrongdoer. (There are theorists who argue that conditional forgiveness is the only legitimate form of forgiveness because forgiveness is consistent with respect for oneself and the wrongdoer only if the wrongdoer repents. The issue of the compatibility of forgiveness and respect will be further discussed in chapter 5.)  One may also forgive the wrongdoer because of other moral reasons, whether or not he has repented. Such reasons may include generosity (Pettigrove 2012), human solidarity (Gerard and McNaughton 2002), and relationship maintenance (that is, forgiving for the sake of maintaining the health of the relationship. This reason for forgiveness will be discussed in detail in the next chapter). Forgiveness given in this way is unconditional on the repentance of the wrongdoer, and hence theorists often call it “unconditional forgiveness”. Lacking a better term, I shall collectively call forgiveness given because of repentance and other moral reasons “recognition forgiveness”.  But reflections on our experience of forgiveness will quickly make us realize that recognition forgiveness does not exhaust the ways we forgive. We quite often forgive not because the wrongdoer has already repented or because of reasons such as generosity or human solidarity, but because we hope that our forgiveness will encourage the wrongdoer to repent. Or we may forgive in order to give 	 116	the wrongdoer a chance to learn from his mistake, “move on”, and do better next time, whether or not he will repent.    To distinguish this form of forgiveness from recognition forgiveness and to better describe it, I shall call it “proleptic forgiveness”.42 I borrow the use of the term “proleptic” from Bernard Williams’s account of proleptic blame (1995). Arguably, one condition of blameworthiness is that the blamed person understands the moral reasons on the basis of which he is blamed. If he does not, then it may be inappropriate to blame him. But when someone proleptically blames a person, she is treating the blamed person as if he had met this condition of blameworthiness (i.e. that he recognized the moral reasons on the basis of which he is blamed) in order to bring it about that the blamed person be able to recognize the reasons as such.  If we strictly followed Williams’s account of proleptic blame, then proleptic forgiveness should be formulated in terms of conditions of forgiveness (or “forgive-worthiness”), and the relevant condition would be repentance. Proleptic forgiveness would then be a form of forgiveness that treats the wrongdoer as if he had repented and is given to the wrongdoer in order that that he repent. But I do not think we should formulate proleptic forgiveness in this narrow way and limit it only to repentance. As I discussed above, we sometimes forgive in order that the wrongdoer learn from his mistake, “move on”, and do better next time, whether or not he will repent. So I will adopt a broader formulation of proleptic forgiveness:   When we proleptically forgive someone, we treat him as if he understood the moral reasons on the basis of which his act in question is wrongful and we hope our forgiveness will bring it about that he understand those reasons.   Formulating proleptic forgiveness in terms of the understanding of moral reasons can explain both cases of proleptic forgiveness. Helping the wrongdoer understand the relevant moral reasons is crucial to encourage the wrongdoer to repent—if he comes to understand those reasons, then he may acknowledge his wrong, apologize, etc. Or we may not be seeking repentance but only the moral 																																																																		42 I would like to thank my supervisor David Silver for bringing my attention to the phenomenon of proleptic forgiveness. 	 117	improvement of the wrongdoer, but understanding relevant moral reasons is still crucial to moral improvement.   (3.2) Proleptic Forgiveness: Distinctive Features and Distinctive Treatment  There are two things distinctive about proleptic forgiveness. First, proleptic forgiveness is given for a distinctive reason: to encourage repentance or moral improvement. The nature of this reason is different from that of those that support recognition forgiveness, e.g. that the wrongdoer has repented, generosity, and human solidarity. Second, proleptic forgiveness cannot be understood in terms of the traditional conditional-unconditional framework in which recognition forgiveness is understood. Because proleptic forgiveness is given to encourage repentance or moral improvement, it is in a sense both conditional and unconditional. On the one hand, proleptic forgiveness is unconditional because it is given without the wrongdoer’s repentance; on the other hand, proleptic forgiveness may have repentance as its aim, so it is in a different sense conditional. Due to these distinctive features of proleptic forgiveness, we should distinguish it from recognition forgiveness.43   Proleptic forgiveness deserves more attention than it has received, not only because of its distinctiveness, but also because of its importance in our moral lives. It is a common phenomenon that plays a valuable role in interpersonal relationships, particularly in personal relationships such as friendships, siblings, and parent-child relationships. When we proleptically forgive someone who has wronged us, we trust that she will come to understand and acknowledge her mistake given time, space, and our encouragement. In other words, proleptic forgiveness expresses our trust in and moral concern for others. In addition, by encouraging the wrongdoer to understand their mistakes, proleptic forgiveness serves an instrumental function of repairing damaged relationships and re-establishing mutual respect. We therefore have good reason to recognize the importance of proleptic forgiveness and appreciate its distinctive features, and the strength of an account of forgiveness should be determined partly based on its capability to explain proleptic forgiveness. 																																																																		43 One might argue that if the distinctiveness of proleptic forgiveness lies in the reason that it is based on, then it is not really so different from recognition forgiveness. Recognition forgiveness is given for some reasons (e.g. generosity) and proleptic forgiveness is given for some other reasons (to encourage repentance). Proleptic forgiveness is therefore merely a kind of recognition forgiveness rather than a different form of forgiveness.  But the distinctiveness of proleptic forgiveness lies not only in its reasons. It is distinctive also because it is both conditional and unconditional. Even if proleptic forgiveness were to be categorized “merely” a kind of recognition forgiveness for terminological reasons, it would be substantively different and should be treated differently from other kinds of recognition forgiveness because its distinctive features and should be treated differently in our discussion of forgiveness. 	 118	However, despite its importance in our moral lives, forgiveness theorists have shown little interest in proleptic forgiveness. Aurel Kolnai (1973) gave a brief discussion of proleptic forgiveness, but his discussion is under-developed and is geared towards solving the paradox of forgiveness. Other theorists, such as Murphy (1988, 30-31) and Radzik (2009, 119), have merely mentioned proleptic forgiveness in passing.44 One likely reason for this neglect is that theorists do not think proleptic forgiveness is a “genuine” form of forgiveness. Critics might argue that proleptic forgiveness is used as an instrument to encourage repentance and is therefore not a “genuine” form of forgiveness. Griswold, for instance, argues that forgiveness requires the wrongdoer’s willingness to “take minimal steps to qualify for forgiveness” (2007, 115; 121-122) and he thus categorically rejects proleptic forgiveness as forgiveness in any form. Theorists might at most regard proleptic forgiveness as a non-standard or non-paradigmatic form of forgiveness.   I suspect that we might not be able to fully spell out what exactly constitutes a “genuine” form of forgiveness and why we need to limit it to recognition forgiveness. Perhaps ultimately this idea of “genuine forgiveness” is based on brute theoretical intuition. But we have reason to believe that proleptic forgiveness is a form of forgiveness that any adequate account of forgiveness should take seriously and be able to explain.  First, Griswold’s argument against proleptic forgiveness—that forgiveness requires the wrongdoer taking “minimal steps to qualify for forgiveness”—is in effect an argument against unconditional forgiveness. It is far from clear and settled that unconditional forgiveness is not a “genuine” form of forgiveness. It is not unusual that we forgive unconditionally, and there are numerous forgiveness theorists who have argued for unconditional forgiveness (Hampton, Pettigrove, Garrard and McNaughton, etc.)  Second, there are theorists who recognize proleptic forgiveness. As I already mentioned, Aurel Kolnai, a prominent forgiveness theorist, explicitly states that what I call proleptic forgiveness is “genuine forgiveness”, even though we might not approve of such form of forgiveness in some circumstances (1973, 106). Another prominent forgiveness theorists who recognizes proleptic forgiveness is Murphy (1988). He takes forgiveness as overcoming resentment due to moral reasons 																																																																		44 Murphy calls it “forgiveness in order to reform” and Radzik calls it “preemptive forgiveness”. 	 119	and seems to accept “reforming” the wrongdoer as a legitimate moral reason for forgiveness, though he is also worried that such form of forgiveness may patronize the wrongdoer (30-31).   Third, and perhaps most importantly, if we aim to explain forgiveness as an interpersonal practice, focussing on how it functions in and shapes relationships, and if we heed Strawson’s insight that analyses of moral concepts should start with our lived experience rather than with brute intuition about theoretical structure, then we should recognize the importance of proleptic forgiveness in our moral lives. Proleptic forgiveness is often found in our moral lives—we often forgive our friends, siblings, partners, children in order that they know they have wronged us. We should therefore treat proleptic forgiveness as such and adequately explain it in our accounts of forgiveness. If we reject proleptic forgiveness and insist that recognition forgiveness is the only “genuine” form of forgiveness, we would be putting the cart before the horse. Our theorizing may at some point need to rely on intuition, but its use should be guided our lived experience. We should also recognize that our moral lives—particularly in personal relationships—may be more complex than we might have initially thought. Rejecting proleptic forgiveness may amount to over-simplifying our moral lives and illegitimately privileging some form of forgiveness while rejecting the rest as sub-standard.    (3.3) Performative Account and Proleptic Forgiveness The phenomenon of proleptic forgiveness enables us to see another strength of the performative account that I am proposing—it can explain proleptic forgiveness. Taking forgiveness to be interpersonal, my performative account says forgiveness is (i) publicly given and communicated to the wrongdoer (and others), (ii) brings about changes to the relationship, and (iii) reconciles the relationship. These three interpersonal features of forgiveness that this account takes to be crucial to forgiveness can explain how forgiveness can be proleptically given and how it can encourage repentance and moral improvement.   The two mainstream accounts, by contrast, will not fare as well. Just as the two accounts have a hard time explaining these three interpersonal features of forgiveness, so too do they have a hard time explaining how overcoming some emotion or adopting some perception inside our heads/hearts can encourage the wrongdoer’s repentance and moral improvement. Only through expressing such emotions or perceptions can one encourage the wrongdoer to repent. That means the two accounts 	 120	could only say that proleptic forgiveness is an expression of forgiveness. But proleptic forgiveness is not an expression of forgiveness; it is a form of forgiveness itself.   This distinction between forgiveness and expressions of forgiveness need not be a significant one, but the two mainstream accounts make it significant because they see expressions of forgiveness as merely secondary in their accounts. If the two accounts use expressions of forgiveness to explain proleptic forgiveness, they would then give proleptic forgiveness a mere secondary role in their accounts. But as I argue, proleptic forgiveness plays an important role in our relationships when wrongs occur, and we need an account that recognizes it and gives it a primary role in the account.   Proponents of the two mainstream accounts might disagree and argue instead that proleptic forgiveness is less important than recognition forgiveness, but denying its importance seems an uphill battle. Alternatively, they might revise their accounts by incorporating expressions of forgiveness into the core of their accounts in order to more adequately explain proleptic forgiveness. But this would in effect be departing from the private approach and going towards the interpersonal direction.   (3.4) Summary I have explained in this section the second strength of my performative account: that it is better at explaining proleptic forgiveness than the two mainstream accounts. The third strength of my account that I am going to discuss in the next section is related to this second strength. Being able to explain proleptic forgiveness enables my account to defend the normative value of forgiveness.  	 121	Section 4 Performative Account and The Normative Value of Forgiveness This section discusses a third strength of my performative account. There are critics who are sceptical of the normative value of forgiveness. They argue that the importance of forgiveness has been over-estimated and its dark side overlooked. I will discuss two of such critiques. First, the paradox of forgiveness, according to which forgiveness is either pointless or unjustified, will be discussed in (4.1). Second, I will discuss in (4.2) a sceptical argument put forward by Martha Nussbaum that forgiveness is normatively problematic in that it is controlling, humiliating, and unproductive. I will argue that my performative account, with its ability to distinguish and explain recognition and proleptic forgiveness, is able to reject both sceptical arguments and defend the normative value of forgiveness.   (4.1) The Paradox of Forgiveness I have discussed in section 2.2 that, according to the paradox of forgiveness, forgiveness is either pointless or unjustified. To reiterate, either the wrongdoer repents and goes through a change of heart or he does not. If he does repent, forgiving him would be merely treating him in a way that he should be treated and thus be pointless. If he does not repent, forgiving him would amount to condoning and thus be unjustified.   I have already demonstrated a solution to the paradox there (section 2.2). I argued that rituals play an important role in interpersonal relationships and performative forgiveness can be seen as a ritualistic act that reconciles relationships. But the recognition-proleptic distinction discussed in the last section can strengthen the solution. On the one hand, both recognition and proleptic forgiveness are justifiable. Recognition forgiveness is justified if the wrongdoer repents; proleptic forgiveness is justified if it is likely to succeed in bringing about the wrongdoer’s understanding of relevant moral reasons. (There are other reasons which may justify recognition and proleptic forgiveness, for instance reasons from relationships. Such reasons are the topic in the next chapter.) On the other hand, recognition and proleptic forgiveness, when they are justified, may have a point. Recognition forgiveness is a form of forgiveness that is publicly given to the wrongdoer in recognition of her repentance and to declare that the relationship is no longer in a broken state. It serves a ritualistic 	 122	function of reconciling the relationship. Proleptic forgiveness is a form of forgiveness that, through unbreaking the relationship, encourages the wrongdoer’s repentance, and her repentance will hopefully lead to further improvement of the relationship. As I argued in section 2.2, forgiveness has a point insofar as rituals play an important role in interpersonal relationships.    (4.2) Nussbaum’s Critique of Forgiveness Nussbaum gives a sustained and insightful critique of forgiveness in her recent work Anger and Forgiveness (2016), where she argues that forgiveness, whether conditional and unconditional, is normatively problematic. Conditional forgiveness is forgiveness given to the wrongdoer only if she repents, whereas the offering of unconditional forgiveness has no such requirement. Nussbaum argues that conditional forgiveness is normatively problematic because the process of its offering is controlling, aggressive and humiliating. Tracing what she takes to be the Judeo-Christian origin of conditional forgiveness, she observes that conditional forgiveness often involves a process of repentance where the wrongdoer is required to confess her wrongdoing, abase herself, and grovel to the victim for forgiveness. Forgiveness is held out as a temporary prize awarded to the wrongdoer if the victim is satisfied with the wrongdoer’s grovelling. Nussbaum therefore calls conditional forgiveness, rather disparagingly, “transactional forgiveness” (72-73).   Forgiveness is normatively problematic even if it is given unconditionally. Unconditional forgiveness does not require the repentance of the wrongdoer, so no grovelling is involved. However, Nussbaum argues that people who forgive unconditionally may assume the moral high ground and hold the secret thought that they are graciously sparing the wrongdoer from begging for forgiveness when the wrongdoer should be doing so. In addition, the forgiver may be trying to use her unconditionally given forgiveness in order to humiliate and shame the wrongdoer; or, as it is put in the New Testament, to “heap coals of fire on his [the wrongdoer’s] head” (2016, 75-78; Romans 12:20 NABRE).   Nussbaum further argues that forgiveness is normatively problematic because it is too backward-looking (or not “transitional”, as she puts it), and we have reason to believe that she takes the backward-looking feature of forgiveness as its primary objectionable feature because “Transitional Anger” is the main theme of her book. Roughly speaking, forgiveness is backward-looking in the sense that it focusses on the past wrong and the anger associated with it. “Life is too short”, as 	 123	Nussbaum proclaims at the beginning of the chapter on forgiveness. What is the more important is to rebuild the damaged relationship between the wrongdoer and the victim, to work out how they are to relate to one another after the wrong, and to construct a productive future; that is, to “transition” to the future. Forgiveness, however, is not geared towards doing any of these things, but can at most removes an impediment—anger—to it (2016, 76), and is therefore not a constructive and valuable way of dealing with wrong.   Although Nussbaum’s critique accurately points out some potentially problematic aspects of forgiveness, it fails to present a full and fair picture of forgiveness and therefore overlooks some valuable aspects of forgiveness. First, forgiveness need not be humiliating nor condescending. Nussbaum is certainly right that forgiveness can be abused. A forgiver may harbour morally condescending thoughts and use the forgiveness process to humiliate the wrongdoer, and a wrongdoer may bow too low in the process of repentance. But it is clearly false that all or even most cases of forgiveness are humiliating or condescending. Forgiveness remains a valuable part of our moral lives when it is neither humiliating nor condescending.   Second, forgiveness on my performative account is constructive and “transitional”. Forgiveness reconciles the damaged relationship and helps bring both parties back together and move forward. Forgiveness in the form of proleptic forgiveness is particularly transitional. It is given in order to encourage repentance and/or moral improvement; if it is successful, memories of the past wrong are less likely to invoke anger and the reconciled relationship will be more stable and solid. Nussbaum’s criticism that forgiveness is not “transitional” may therefore be applicable only to some forms of forgiveness and some accounts of forgiveness, such as the two mainstream accounts. If forgiveness is merely something that is inside our heads—overcoming certain emotion or seeing the wrongdoer in some benign way, then, yes, forgiveness may not be “transitional”. The performative account that I am proposing, by contrast, understands forgiveness in a way that forgiveness is transitional and helps improve the relationship. This, again, is another point where the performative account fares better than the two mainstream accounts.    	 124	Chapter Conclusion My explication of the performative account of forgiveness is now complete. This account takes forgiveness to be a declarative act that unbreaks the relationship between the wrongdoer and the victim. It has several strengths over the two mainstream accounts. Most importantly, it can adequately explain the interpersonal aspect of forgiveness and avoid the misconceptualization problem, and can thus adequately explain forgiveness as a normative practice situated in interpersonal relationships, particularly in personal relationships. It explains how forgiveness functions in relationships (it is communicative, reconciliatory, and performative) and how it shapes relationships (it unbreaks relationships.) Combined with its other strengths, this performative account can at least complement the two mainstream accounts, if not replace them.      	 125	Chapter 4 Reasons for Non-blame Responses The discussion in the previous chapters might give us the impression that, when we have been wronged, we should either blame or forgive the wrongdoer. This is in line with how discussion in the blame and forgiveness literature has been framed. Some blame theorists even argue that blame (as moral anger) is a proper response to wrong because it shows respect for oneself, the wrongdoer, and morality (I shall discuss blame and respect in greater detail in the next chapter). Some forgiveness theorists argue accordingly that forgiveness is appropriate only if the wrongdoer repents, e.g. acknowledging and denouncing her own wrongdoing. Other forgiveness theorists disagree, arguing instead that we have reasons to forgive unconditionally; for instance, reasons of human solidarity and generosity may give us reason to forgive the wrongdoer even if she does not repent.    But whether or not theorists do think blame and forgiveness are the only or primary appropriate responses to wrongs, reflections on our experience of personal relationships will quickly reveal to us how abstract and narrow this approach is. First, abstract talks of “reasons for and against blame and forgiveness” overlook reasons that are specific to personal relationships, that is, reasons that cannot be generalized to all moral agents but are applicable only to those who are standing in the specific personal relationship. Second, and more importantly, treating the issue of how we should respond to wrongs as a matter of only blame and forgiveness will easily make us lose sight of the depth and complexity of personal relationships. Personal relationships do sometimes give us reason to blame and/or forgive our intimates, but other times they render apparent wrongs morally innocuous, and yet other times they give us reason to set aside blame and forgiveness entirely but to do other things. In other words, there is much more to personal relationships than just blame and forgiveness.   This chapter thus argues that there is a multiplicity of reasonable responses to wrongs in personal relationships. Blame (angry or non-angry; affective or affectless) and forgiveness are not the only appropriate responses; at times they may even be inappropriate. We need to expand our discussion and take a broader approach to our enquiry of appropriate responses to wrongs. Instead of asking “should I blame?” and “should I forgive?”, we should ask instead “what responses to wrong are appropriate in personal relationships?”   	 126	The first section of this chapter discusses four reasons in personal relationships that call for various responses to wrongs. Though these reasons are more prominent in personal relationships, which are the focus of discussion in this chapter and indeed in this dissertation, they are applicable to both personal and non-personal relationships, and appreciating the multiplicity of responses to wrongs in one helps us appreciate it in the other. I will therefore at times turn to non-personal relationships. The first section also addresses one major concern about reasons against blame: that telling the victim not to blame may amount to shifting the burden from the wrongdoer to the victim. I argue that, though the idea that victims may have reason not to blame does not necessarily amount to victim-burdening, it can easily slide into victim-burdening and so we should approach this idea with caution.   This chapter also addresses a few other issues that deal with reasons for and against blame in personal relationships. Some theorists have taken a sceptical stance on angry blame, arguing that angry blame is unjustifiable. While I believe that angry blame is sometimes unjustified, I do not believe that it is unjustifiable. I reject such a sceptical argument against angry blame in section 2 and clarify my stance on the justifiability of angry blame. Section 3 clarifies the nature of the reasons discussed in this chapter. It also makes an observation about blame theorists’ lack of interest in reasons from personal relationships. Section 4 discusses the role of justifying and motivating reasons in personal relationships.   	 127	Section 1 How to Respond to Wrongs: Reasons From Relationships This section discusses four reasons that govern how we should respond to wrongs in personal relationships (1.1 to 1.6). These reasons call for various responses to wrongs—they may tell us to blame the wrongdoer (privately and/or publicly), to forgive the wrongdoer, or to do something else. Blame and forgiveness are therefore merely two among many appropriate responses to wrongs. Moreover, because these four reasons all focus on what the victim rather than what the wrongdoer should do in response to wrongs, this section will address the legitimate concern that such reasons may amount to overburdening the victim (1.7).   Before I begin discussing reasons from relationships, I will state two points, to be further elaborated in (1.5), in order to preempt some objections. First, the reasons to be discussed are not conclusive reasons but are to be weighed against other reasons; second, the strength of these reasons are dependent on the value of the relationship in question.   (1.1) Relationship Maintenance A healthy, unbroken personal relationship is a valuable relationship. It is where people can pursue mutual goals, share emotions, build and develop their lives together, among other things. More importantly, personal relationships give people the space and opportunities to do these things in ways and depth that are not available in non-personal relationships, e.g. there are goals and emotions that non-personal relationships cannot easily accommodate. To the extent that healthy, unbroken personal relationships are valuable, we have reason to take good care of them when they are in good shape. For instance, we may have a reason to “catch up with” our friends every once in a while, or spend a sunny afternoon at the beach with our family because these are things that develop, renew, and strengthen relationships. We also have reason to avoid damaging personal relationships and to unbreak them when they are broken.   Discussions in the blame literature sometimes give us the impression that blame is a natural and proper response to wrongs so long as wrongs have indeed been committed. Treating blame as merely a matter of whether the person is in fact blameworthy and as a straightforward answer to wrongs will 	 128	likely make us lose sight of considerations that are pertinent in personal relationships.45 Even when it is clear that someone close to us has in fact wronged us and so they are indeed blameworthy, we will not always immediately jump to the blame mode. This is because we are acutely aware of the damage that blame can do to personal relationships. Blame in personal relationships involves anger, disappointment, and hurt feelings. If expressed, blame in such forms could create tensions and/or distance between people; anger, in particular, is an oppositional attitude that may give rise to rifts and further conflicts. For instance, blaming a family member when the family relationship is already wraught by previous wrongs may not be conducive to the fragile relationship but will only further damage it. Even if it remains unexpressed, blame may still change the behaviour of the people harbouring such hard feelings and damage the relationship between them. These are considerations that we face and should take into account when wrongs occur, and we may sometimes have reason to refrain from blaming a clearly blameworthy wrongdoer. In addition, we may even have reason not only to withhold blame but also to forgive the wrongdoer in order to unbreak and rebuild the relationship; or, as one might put it, to press the “reset” button.   Relationships maintenance may sometimes even give us reason to go beyond blame and forgiveness and perform other more important relationship-maintenance tasks. Sometimes the occurrence of wrong in a personal relationship signals the existence of some underlying problems with the relationship, and preserving the relationship may require identifying and solving these problems. In such cases, we may need to refrain from expressing blame in order to focus on such relationship-maintenance tasks. We may even have reason to set aside blame (private and public) and forgiveness all together if dwelling on such matters would divert our attention away from the underlying problems if not jeopardize our efforts in identifying and solving them.  Nussbaum explains how setting aside blame may be crucial to relationship maintenance with the following case from the psychologist Harriet Lerner (Lerner 1985; Nussbaum 2016, 107-109). The case involves Maggie and her overly-critical, fault-finding mother. She criticizes Maggie’s messy kitchen and her management of money, complains about the fancy dinners that Maggie and her husband prepared, and doubts her ability to take care her child. Maggie, naturally and 																																																																		45 In other words, the reasons discussed in this section do not deal with blameworthiness, but instead deal with “now that this person has wronged me, how should I respond? Should I blame? Forgive? Or respond in some other ways?” In section 3, I will discuss in greater detail the difference between reasons that deal with blameworthiness and that deal with such non-blameworthiness considerations. 	 129	understandably, is angry at her mother, but the problem with their relationship has much more to do with her mother’s criticisms and complaints. The underlying problem lies in how Maggie and her mother relate to one another: her mother fails to respect Maggie’s independence and treats her as if she was still a child. What Maggie needs to do is to solve this underlying problem and establish a healthy, mutually respectful relationship with her mother. Blaming her mother, though she is indeed blameworthy, cannot solve the problem but will only perpetuate it. Following the advice of her counsellor, Maggie sets aside blame (both private and public) and speaks calmly with her mother in order to work out a relationship that respects Maggie’s independence.   The value of relationship maintenance is relevant not only in personal relationships but also in non-personal relationships. Relationships with colleagues, neighbours, and fellow citizens can all be valuable relationships that are worthy of preserving. Imagine an absent-minded colleague who is otherwise responsible, or a noisy neighbour who is otherwise friendly and approachable. We may have good reason to withhold blame, or at least withhold it till the right time, in order to preserve the generally good relationship. Or consider the recent polarized political climates in several western countries. There are certainly people, politicians and ordinary citizens who hold blameworthy beliefs, made bad decisions, and/or acted inappropriately. But continuing to blame those who do not share the same political beliefs as ours might not be the best way to deal with the differences with our fellow citizens nor to improve the strained relationships. There might be reason to start engaging in other relationship-maintenance tasks before our relationships worsen.   To briefly summarize, relationships (personal and non-personal) give us reason to keep them in good shape. When wrongs occur, we may have reason to withhold blame, to forgive, or to set aside blame and forgiveness in order to pursue other relationship-maintenance tasks.   (1.2) Concern Concern is a very important component of personal relationships. Friends and family members have and should have strong mutual concern for one another. It may be difficult to define what “concern” is, but it will suffice for the present purposes that an important component of concern for someone is the disposition to promote the person’s interests and well-being. When those who are close to us have committed wrongs, concern may give us reason to respond to the wrong in ways that have little to do with blame. For example, we may have reason to give them “moral support”. By “moral support”, I 	 130	do not mean “psychological encouragement” but support that promotes the person’s moral well-being; for example, encouraging her moral improvement. When a friend lied to us, or our partner said hurtful things to us, we may have reason to explain to them how their behaviour has harmed us and is inappropriate, over and above whether we are to blame them.   The reason to give moral support is particularly strong in personal relationships where part of our role is to provide moral education, such as parent-child relationships and, arguably, teacher-student relationships. The educative nature of these relationships sets the priority of the relationships and gives the person with the educator role good reason to focus on educating the wrongdoer rather than on blaming her.   Consider the following case of third-person wrong. Suppose your daughter was caught embezzling a large sum of money from the firm that she was working for. She has been fired, publicly and privately condemned by her colleagues, and the case has gone into legal proceedings. Perhaps what she has done is so outrageous that you want to join the condemners, but this may not be the best course of action for a parent. If she does not see that what she did was wrong—say she exploited some legal loopholes to dishonestly obtain money from her firm and she thinks what she did was legal and therefore moral. As her parent, you have reason to promote her moral growth by explaining to her the wrongfulness of her behaviour, e.g. how harmful/dishonest/disloyal it was, how it made her colleagues feel, how she has taken advantage of their trust, etc. If she comes to understand the wrongfulness of her behaviour, you may have reason to help and encourage her to learn from her mistakes. You will also have reason to help and encourage her to learn from her mistakes if she understands the wrongfulness of her act. For instance, you may give suggestions on how to avoid making the same mistakes in the future, what steps she can take to do so, its implications for other kinds of behaviour, etc. In addition, you may provide moral support by advising your daughter on how to deal with the moral fallout of her wrongdoing. You may suggest that she apologize for her dishonesty and taking advantage of her colleagues’ trust, make compensations (returning the money and paying for damages), turn herself to the police if she hasn’t done so already, and cooperate with the investigation by providing them with evidence of her wrongdoing.   All of these responses have nothing to do with blame (at least expressed blame) or forgiveness, but they are reasonable because they show the concern that a parent should have. The parent might still 	 131	have reason to express blame to their daughter, but it should come second to moral support. One may need to wait for a suitable time and occasion to do so, say when the case is all settled. I must add, however, that moral advising and encouragement can easily descending into lecturing or even blaming, so one should pay careful attention to the direction that one’s moral support is going.  Moral support is not applicable only to personal relationships but also non-personal relationships. This may initially sound wrong because it seemed inappropriate to give unsolicited moral advice to our neighbours or a random taxi driver. It is plainly intrusive to teach our neighbours how to organize their garbage and to advise a taxi driver to develop his career to take better care of his family. But the reason to provide moral support exists in non-personal relationships; it is just weaker than in personal relationships and we need give our support in the right way and at the right time. Perhaps in most cases we have stronger reason not to be intrusive. Even when we should support others in non-personal relationships, we should do it subtly. For instance, we may support someone by simply giving him time and space to learn from his mistakes and to take care of himself. (“Support” may often connote with positive acts, but non-intrusion can promote the interest and well-being of other, and it is in a sense a way to give “support”.)  One might argue that (expressed) blame can also serve the function of providing moral education for the wrongdoer. Just as some theorists have proposed the moral education theory of punishment, so too some might propose a moral education theory of blame. Blame may be a robust way to convey to the wrongdoer the message that the act in question is wrong, and through this message the wrongdoer will be able to learn from his mistake and make moral improvement. Therefore, if we want to promote the moral well-being of our intimates, we should not avoid blaming but should blame them when they do wrong.   I do not deny that blame can sometimes serve the function of moral education. I also think that sometimes moral support gives us strong reason to blame those close to us. If someone is a bully and no one calls him out, I think those close to them, e.g. his parent or his partner, have good reason to blame him so that he knows his bullying is unacceptable. But there are also times when blame is not the best way to morally educate. Blame is often filled with emotion and tension. The blamer may easily get carried away, and the blamed person may perceive the blame as an attack and become defensive. The lesson may be over before it starts, and the important moral message is buried deep 	 132	under anger and exasperation. Refraining from blame is sometimes a safer, clearer, and more direct way to convey one’s moral message to the wrongdoer, showing the wrongdoer that we stand with them and are here to support them.   (1.3) Trust Philosophers may not have agreed on a definition of trust, but they recognize the importance of trust in personal relationships, especially since the publication of Annette Baier’s “Trust and Antitrust” (1986).46 We trust that our friends, siblings, etc. have our interests at heart and will not harm us. We also think that we have reason to trust them and not to doubt their good will towards us because we are friends, sibling, etc. For the present purposes, I will roughly define trust as an attitude of dependence on the goodwill of the trustee and lowering one’s guard against the potential harm by them. I want to argue in this sub-section that we have reason to respond to their (apparent) wrongs in ways that show trust.   Trust governs our responses to wrongs in ways different from relationship maintenance and concern. Relationship maintenance and concern both deal with whether one should blame, forgive, or respond in some other ways when wrongs have been committed. Trust, by contrast, primarily deals with whether one should believe that wrongs have been committed in the first place when outsiders who are not in the same relationship may have sufficient reason to believe they have been committed. Trust also deals with whether the person in question has done anything blameworthy.   (1.3.1) “It’s Not What You Think” Trust gives us reason to believe that wrongs might not have been committed in the first place because trust gives us reason to interpret apparent wrongs as not genuine wrongs. Trusting those who are close to us means believing that they bear goodwill towards us and would not do wrong to us. If they do something that appears to be wrongful to us, trusting them means, absent strong evidence, (i) believing that they did not do it, (ii) they have legitimate excuses or justification, or (iii) that the apparent wrong expresses innocuous attitudes.  																																																																		46 Baier, along with others who have written on trust, argues that trust exists in all relationships, personal and non-personal. This seems right, but trust is a stronger reason in personal relationships than in non-personal ones. There are also different kinds of trust in different relationships. We may trust a stranger not to put our lives in danger and not to intentionally show us the wrong direction, but we would trust our friends and family members in a deeper way. We may trust that they would make a sincere effort to find the direction for us, or that they would warn us when we are about to make a big mistake. 	 133	Suppose you heard a rumour that your good friend Sam had been secretly sabotaging your romantic pursuit. The rumour sounds credible, and other people believe it and are blaming Sam. But because Sam is your good friend, it may be wrong for you to immediately believe the rumour and blame him. Instead, you may have reason to trust him. You may have reason to be sceptical of the rumour and to hold off your belief till more evidence becomes available, even though the currently available evidence may be sufficient for other people who are not Sam’s good friends to believe the rumour.47   Trust also gives us reason to believe that the apparent wrong has a legitimate excuse or justification. Suppose your friend Marie calls you to cancel your long-planned dinner date and then quickly hang up without explaining the reason for the cancellation, trust gives you reason to believe, at least initially, that Marie has a legitimate excuse for the cancellation—perhaps she needs to finish a big project at work, or that she may have a family emergency. Unless you have strong evidence to believe otherwise, say Marie has a track record of cancelling dates last minute, jumping immediately to the conclusion that Marie has wronged you may show lack of trust.   Or consider a rather well-worn example. Suppose, when walking on the street with your friend Andrew, he suddenly gives you rather strong push to the side that you lose your balance and fall. Trust gives you reason to assume that he has an excuse for his push—perhaps he tripped—or a justification—perhaps you were going to step on a kitten lying on the ground. You would make such assumptions because you trust that Andrew is your friend and he won’t do you wrong.   Trust also gives us reason to interpret apparently wrongful acts as expressing innocuous attitudes. When an intimate does something that seems to reveal a bad quality of will and seems to have no excuse or justification for it, we may still have reason to believe that they do not “mean” it. For instance, when a friend makes fun of us in an apparently mean way, we would assume the joke is not meant to be offensive or hurtful, but is only a friendly tease meaning to express friendliness.   If trust gives us reason to interpret apparent wrongs as not genuine wrongs, then our responses to (apparent) wrongs in personal relationships should surely not be blame nor forgiveness. But more importantly, trust gives us reason to trustfully assume that the wrongful act in question did not take 																																																																		47 You might also have special epistemic access to evidence that Same did not or would not do it, and such evidence gives you reason not to blame Sam. I will call such a reason “understanding”, which I discuss in the next sub-section (1.4). 	 134	place, has a legitimate excuse/justification, or expresses innocuous attitudes. Even if the apparent wrong turned out to be a genuine wrong, trust gives us reason to at least initially assume that it is a mere apparent wrong. To use a rather legalistic expression, our friends and family members should be “assumed innocent until proven guilty”. We may also have reason to find out facts relevant to whether our friend or family member has actually done anything that genuinely wrongs us (maybe only to show that they are indeed innocent). And if we cannot find out those facts, then we may have reason to continue to assume that they are innocent.   (1.3.2) Not Blameworthy Trust also gives us reason to believe that those who are close to us hold general good will towards us despite the particular wrongs that they did to us. It matters to us very much whether those close to us care about us. We appreciate the good things that they do to us—say giving us presents or buying us lunch—not merely because those things are nice, but more because they reflect how much our friends and siblings care about us.  On the other hand, when a friend wrongs us, say breaking an important promise, what matters to us is more than the harm—financial, social, etc.—that these things cause. What matters to us the most is whether the broken promise is an indication that the friend does not care about us such that she would not have our interests at heart. So what we blame her for may be less of the broken promise itself but her lack of care.   But wrongs do not necessarily reflect a lack of general good will. We are all fallible beings, and in personal relationships where we frequently interact with one another, we will time and again make mistakes no matter how much we like and care about them. Wrong is simply an inevitable and unfortunate fact of personal relationships.   A certain level of trust in the general good will of those close to us is therefore not only reasonable but also crucial in any personal relationship. Trust gives us reason to believe that the wrongs that others do to us are mere one-time mistakes and are not indications that they do not care about us. It therefore gives us reason to respond to wrongs in various ways, some of which do not involve blame. For instance, maybe we should still blame our friend for saying some very hurtful things to us on one occasion, but perhaps we should blame them less severely because we have reason to believe that they still care about us. Or we may have reason to forgive them and give them a “second chance” because they do care about us and will learn from their mistake and behave better in the future. Or 	 135	we may set aside blame and forgiveness entirely but work with them to improve our relationship, say establishing better communication and mutual understanding.   (1.4) Understanding  Similar to trust is what I shall call “understanding”. Through regular interactions and mutual disclosure with our friends, partners, and family members, we develop special epistemic access to their psychological states, behavioural patterns, and traits. This special epistemic access enables us to interpret one another’s behaviour and infer from their behaviour the desires, needs, thoughts, etc. that motivate their behaviour, and in turn their quality of will.   If we have this special epistemic access to details about those close to us, then we have reason to show understanding by going under the surface of their behaviour when they (apparently) wrong us; that is, we should “know better”. When a friend cancels a meeting on short notice, we should know whether they have any legitimate excuses or justification; or when a sibling makes a seemingly cruel joke about us, we should know whether they mean well or ill. We have reason to “know better” because, first, we should blame only genuine wrongs, and, second, personal relationships give us reason to know one another.   If we have reason to show understanding, then we should try to find out the facts about apparent wrongs when we see them. We should try to ascertain whether our intimate has genuinely wronged us before jumping to the conclusion that they have and blaming them. Failing to do so shows lack of understanding and possibly also lack of trust. We may also have reason to improve our communication with them and knowledge of their behavioural patterns such that we will be able to know better when conflicts arise.   (1.5) Weighing the Strength of Reasons from Relationships I stated in the introduction of this section that (i) the reasons discussed above are not conclusive reasons but are to be weighed against other reasons and (ii) the strength of these reasons are dependent on the value of the relationship in question. I will now further clarify these two points.   First, the reasons discussed above—relationship maintenance, trust, etc.—are reasons to be weighed against other reasons. Sometimes not blaming the wrongdoer risks reinforcing the wrongful 	 136	behaviour and inviting future wrongs; blaming the wrongdoer will instead tell her clearly that her behaviour is unacceptable and will deter future wrongs. In such cases, the victim may have good reason to firmly stand up for himself and blame the wrongdoer. Moreover, if it is unsafe for the victim to work with the wrongdoer to solve the underlying problems of the relationship, say the victim is dealing with an abusive spouse who will be provoked by any mentioning of the abuse, then the victim surely has good reason to first of all protect himself and not to risk any further harm.   Second, the strength of reasons from relationships are dependent on the value of the relationship in question. Relationships give the victim good reason to refrain from blaming in order to show concern, trust, etc. only if they are valuable, e.g. they are generally equal, respectful, and caring. Consider, as a point of contrast, an abusive and hierarchical marriage where one spouse regularly abuses the other and the other spouse is regularly expected to show “love” and “caring”. Or a hierarchical parent-child relationships, where the child’s independence is not respected by their parents but the child is simply treated as the extended self of their parents. They (particularly daughters) are supposed to “respect” their parents and never defy their will, and should follow their parents’ “advice” on career path, marriage, and so on. There is no mutual respect nor mutual caring in such relationships, and the victims may have no good reason to rebuild them or show special concern to the wrongdoer. The only “reasons” that such relationships give may be prudential ones—say to avoid further abuse.48   (1.6) The Depth and Richness of Our Personal Lives Perhaps one thread that runs the above points together is this: personal relationships would be too rigid, moralistic, if not prosecutorial if the only reasonable responses to wrongs are blame and forgiveness. Our personal relationships ought to be richer than this. When our daughter was caught embezzling, we have reason to show her our support—emotional, physical, financial, and moral. When our child lied to us, we have reason to find out the causes of the lie—is it because of fear, insecurity, or anxiety? When our spouse has been unfaithful to us, we may have reason to find out whether she is emotionally unsatisfied with the marriage. And when our friend says hurtful things to us, we may have reason to wonder if there has been a breakdown in communication.  																																																																		48 It may not always be obvious how to determine the value of a relationship and in turn the strength of reasons from it. For instance, it is not clear whether one’s emotional investment in a hierarchical marriage would give one strong reason to maintain the relationship despite it is a hierarchical marriage. Such questions about how exactly to determine and weigh the value of and reasons from relationships are beyond the scope of this dissertation. What I intend to put forward is only a framework—that relationships do give us legitimate reasons regarding how to respond to wrongs. 	 137	I am by no means suggesting that determining who is in the right and who is in the wrong, and following that blame and forgiveness are unimportant in personal relationships, nor am I suggesting that it would be easy not to think about the wrongfulness of the treatment that we have been subjected to. All I am suggesting is that there ought to be more to personal relationships.   (1.7) Objection and Reply: Burdening Victims (1.7.1) Objection One might worry that the idea “victims should not blame” may impose extra burdens on the victim. First, asking the victim not to blame may expose the victim to further danger. Blame sometimes serves as a protest and deterrence against wrong. It is a way victims tell the wrongdoer that the act in question is inappropriate and should not be done again in the future. Opting not to blame makes the victim lose this important means of deterring future wrongs and runs the risk of perpetuating the wrongdoer-victim relationship. Hampton therefore warns us against the danger of withholding blame due to “family peace” (Hampton 1988, 39). In addition, attempts to work out the “real problem” with the wrongdoer may be met with resistance, ridicules, if not further abuse, causing more physical and/or emotional damage to an already harmed victim. Such moves are particularly risky in hierarchical and abusive relationships, where one party does not see or treat the other as an equal and there is a pattern of wrongful treatment. Blame certainly carries its risk in that it may upset the wrongdoer and subject the victim to further harm, but it may sometimes be a relatively safe way to deal with problems in such relationships. “Relationship maintenance”, “concern”, and so on do not seem legitimate reasons to ask victims to put themselves in further danger.   The clarification I made in (1.5) is able to provide a reply to this objection. The victim should consider other reasons, such as their own safety. Also, if the relationship is clearly and gravely abusive and/or hierarchical, then the relationship is clearly not reason-giving. The only “reasons” that such relationships give the victim may be prudential one, e.g. to maintain the relationship in order to avoid further conflicts, confrontations, and danger, or to leave the relationship altogether.   But a stronger objection and more worrying problem lie in the idea that it is the victim, rather than the wrongdoer, who has the responsibility and reason to deal with the fallout of the wrongdoing. If there is anyone who should be responsible for rebuilding the relationship and showing concern, it should be the wrongdoer who did the wrong and damaged the relationship, and who may well be the 	 138	cause of the “underlying problem” and the unhealthy relationship in the first place. Asking the victim to maintain the relationship and show concern, or even suggesting that there are things that the victim should do is shifting the focus from the wrongdoer to the victim, illegitimately transforming the problem of the wrongdoer into a problem of the victim. Now the victim faces even more problems. Not only does she need to deal with the physical and/or emotional harm that the wrongdoer has caused her, she also has on her weary if not scarred shoulders the responsibility to improve the relationship, a relationship that is damaged not due to her fault. And if she refuses to bear the extra burdens and goes on to blame the wrongdoer, then she is “selfish” and should be blamed for it. There is a term for this form of blame: victim-blaming.   The problem of imposing responsibility on victims is particularly common in hierarchical and abusive relationships. I do not wish to make any generalizations about any culture; I am only speaking from cases of responsibility-imposing and victim-blaming that I witnessed. In marriage, for instance, wives are often expected to show “love” and “concern” and forbear their husbands’s wrongful behaviour, even when the husbands have been uncaring, unfaithful, or even physically and/or emotionally abusive. Other family members (particularly those in power) will often enforce such norms by blaming the wife if she tries to stand up for herself. Third parties are discouraged from intervening; at most, they should only remind the wife of the importance of “family peace”. Similarly, children (even adult children) are expected to show respect and filial piety to their parents even if the parents have wronged them or have failed to respect their children’s independence (such as in Maggie’s case). Children are expected to accept their parents’ wrongdoing and to know that their parents are doing it “for their own good”; standing up for themselves is “disrespectful” and “filially impious”. And filial impiety is regarded as a very serious character flaw.   There is a story from Kongzi Jiayu (15:10, reprinted in Sturgeon 2011) that can illustrate responsibility-imposing in hierarchical parent-child relationships. There may be other ways to interpret the text, but my purpose is not to do interpretive work nor to criticize Confucianism. My purpose is only to use the story to illustrate the norms of responsibility-imposing that are often found in hierarchical societies—Confucian or otherwise, and there are still people who hold such norms nowadays—in eastern, western, and other societies.   	 139	Zeng-zi is a student of Confucius. One day, while working with his father in the field, Zeng-zi accidentally cuts the roots of a plant. His father is infuriated by Zeng-zi’s carelessness and starts beating him with a long stick. Zeng-zi does not escape and is beaten out of consciousness. When he finally regains consciousness, he “happily” gets up and goes to apologize to his father for making his father exert himself to teach Zeng-zi a lesson. To reassure his father that he has been unharmed, Zeng-zi returns to his room to sing and play music.   Zeng-zi thought what he did was filial, but, on the contrary, it angered his teacher Confucius. Confucius says that a filial son should escape from heavy beatings in order to prevent his father from being unkind to the son. Zeng-zi’s taking his father’s heavy beating led his father to act unkindly and potentially commit manslaughter. Such behaviour is gravely unfilial. Hearing of his teacher’s criticisms, Zeng-zi went to apologize to his teacher for his mistakes.   Some Confucians have used this story to argue that parent-child relationships (or father-son relationships) in Confucianism are not as one-sided as many people think. There are norms governing both the father and the son—the father should act kindly, and the son should act filially. But even if these Confucians were right, parent-child relationships in Confucianism as represented by this story would still be deeply problematic. Beating one’s child out of consciousness is clearly wrongful, but Zeng-zi does not think his father has done anything wrong (he “happily” gets up, according to the story), and instead actively conveys the message to his father that the beating was acceptable (e.g. by singing). Now Confucius does find the beating wrongful, but his response is even more problematic than Zeng-zi’s. He does not defend Zeng-zi, who is his own student and the victim of the beating, nor does he condemn the father for inflicting potentially serious bodily harm to his own son. He does not even stop short of blaming Zeng-zi for mistaking the beating as appropriate. He chooses instead to accuse Zeng-zi of committing the “gravely unfilial act” of  “leading his father into wronging him”. To draw a contemporary comparison, Confucius’s accusation is akin to blaming rape victims for dressing “immodestly” and “inviting” men to rape them (though there may be nothing not ancient about rape and victim-blaming).  (1.7.2) Reply The idea of reasons from relationships against blame does not and should not shift the focus from the wrongdoer to the victim. If the relationship is abusive or hierarchical, then of course it is the 	 140	wrongdoer’s fault that the relationship is damaged and requires rebuilding, and it is surely the wrongdoer’s responsibility to show concern to the victim by not having wronged her in the first place. The wrongdoer should now acknowledge his wrongdoing and make amends, and should be blamed if he fails to do so. It is surely unfair to the victim that she, rather than the wrongdoer, has to take the initiative to rebuild and improve the relationship after she has been wronged. The idea of reasons from relationships against blame fully recognizes that the burdens should be on the wrongdoer rather than on the victim in such situations.   But moral lives are most of the time not so black and white. Most ordinary personal relationships are neither abusive nor hierarchical, and neither particularly good nor particularly bad. Suppose, in such relationships, the wrong and/or the underlying problems are clearly the wrongdoer’s fault. One might want to say that it is the wrongdoer’s responsibility to apologize, make amends, and take the initiative to rebuild the relationship, and that it is unfair to ask the victim to take the initiative. But what if the relationship is a very intimate one, such as a parent-child relationship? Would the victim—whether it is the parent or the (adult) child—really have no reason at all to refrain from blaming the wrongdoer and work on the underlying problems? Is it always unfair to ask the victims to do so?     In addition, it is often not so clear-cut that the wrong and the underlying problems are entirely the one person’s fault. What if the wrong is a symptom of some underlying problems that both parties have contributed to? Or they are neither parties’ fault, but are due to personality incompatibilities? Wouldn’t blaming do nothing but cycle the problems “round and round”? Wouldn’t the victim have at least some reason to refrain from blaming but focus on improving the relationship instead?   Or suppose the relationship is hierarchal. What if there are some positive aspects to the relationship? Say a parent does not respect her children’s independence, but is otherwise very caring and supportive? Or what if a husband holds patriarchal views of marriage, but is otherwise very loving, honest, and responsible? Are such relationships reason-giving? If so, how much?  I am afraid I am not able to give any straight-forward answers. Personal relationships are complex, ambiguous, and full of dynamics—one purpose of this dissertation is to bring this point to the fore. I can only make a vague claim that the strength of reasons from a relationship is dependent on how 	 141	valuable the relationship is, what relationship it is, and how much potential it has for improvement. Other factors may also be relevant, e.g. the safety of the victim.  But the complexity and ambiguity of personal relationships give us all the more reason to discuss and take seriously reasons from personal relationships. We do recognize such reasons in our relationships and will try to take them into consideration when deciding how to respond to wrongs, even though we do not always have a good grip on them. Moreover, abusers often take advantage of the complexity and ambiguity of personal relationship to advance their interests and reinforce their abuses. Rejecting the legitimacy of reasons from relationships and avoiding discussing them would only play into the hands of abusers. When the time inevitably comes that we have been wronged by someone close to us and she refuses to acknowledge her wrong, we will have to consider such reasons and decide how to respond to the wrong: should we blame her? Keep silent? Or do some other things? Having a clear understanding of reasons from relationships—the content of such reasons, the strength of such reasons, and how they stack up against other reasons such as safety—will be very helpful for our decision-making in such situations. It will also help us keep in clear sight who the real culprit is when wrongs occur and be vigilant of victim-imposing—whether we are the victim or the wrongdoer.  (1.8) Summary I have discussed in this section four reasons that are pertinent and specific to personal relationships —relationship maintenance, concern, trust, and understanding. Each of these four reasons call for various responses to wrong, some of which are neither blame or forgiveness. There is indeed a multiplicity of reasonable responses to wrongs in personal relationships. Blame and forgiveness are not always the right ones.   Some might hold a stronger position such that blame is not merely sometimes but always unjustified. In the next section, I am going to discuss and reject the claim that blame is always unjustified.  	 142	Section 2 Scepticism of Blame In her recent book Anger and Forgiveness, Martha Nussbaum (2016) launches a sceptical argument against angry blame—that blame, to the extent that it involves anger, is unjustifiable. While the scope of Nussbaum’s sceptical argument is limited—it deals only with angry blame but not non-angry nor affectless blame, it is a significant argument in that angry blame is not an uncommon form of blame both in personal and non-personal relationships and Nussbaum categorically rejects its justifiability.   I do not subscribe to this strong position. While I argue that (angry) blame is sometimes unjustified, I think that, in both personal and non-personal relationships, (angry) blame can be justified in some cases. In this section, I will clarify my stance on the justifiability of angry blame and demonstrate the lack of success of Nussbaum’s sceptical argument against angry blame.    (2.1) Nussbaum’s Two Errors of Anger Nussbaum’s sceptical argument consists of two claims. First, anger conceptually involves the wish for suffering. Second, the wish for suffering is unjustifiable. Therefore, angry blame is unjustifiable. These two claims will be explained in detail in the following.   Nussbaum first observes that moral anger conceptually involves a wish that the object (i.e. the person with whom one is angry; or simply, the wrongdoer) suffers (15).49 Her characterization of anger as a wish for suffering (or the “payback wish”, as Nussbaum calls it) follows a long tradition of “first-rate thinkers” such as Aristotle, the Stoics, Butler, Adam Smith, Richard Lazarus, and James Averill (22), but she further backs up the characterization by four stories for illustration (23-26).   The main thrust of Nussbaum’s sceptical argument lies in her second claim—that the wish for suffering is unjustifiable. She argues that the wish for suffering may come in two forms, both of which are unjustifiable. In its first form, it is a wish that, by inflicting pain on the wrongdoer, the angry person (who has been wronged by the wrongdoer) can improve the state-of-affairs. The wish 																																																																		49 Nussbaum treats “Transition-anger” as a “borderline species of anger” which is a pure forwarding “anger” and does not involve the wish for suffering. It seems confusing if not inconsistent to categorize “Transition-anger” as a kind of anger because the wish for suffering is a conceptual part of anger in Nussbaum’s account. Moreover, Nussbaum acknowledges that Transition-anger is a “borderline case” that is “rare” in our lives (262). I will therefore treat Transition-anger as a different emotion in order to avoid confusion, and my discussion of Nussbaum’s account of “anger” will not include Transition-Anger. 	 143	for suffering in this form commits the “payback error”, which is a form of metaphysical error: the suffering of the wrongdoer will not make things better. As Nussbaum puts it, “[medical malpractice] litigation does not resurrect the dead, nor does a punitive divorce settlement restore love” (29).   In its second form, the wish for suffering is a wish that, by inflicting pain on the wrongdoer, the angry person/victim can lower the wrongdoer’s relative (social) status and in turn restore the equal (social) status between the two. While this second form of the wish for suffering does not commit any metaphysical errors—inflicting pain on the wrongdoer can in a sense improve the state-of-affairs, it commits the “status error”, which is a form of moral error. The wish for suffering in this second form construes the wrong and the relationship between the wrongdoer and the victim in terms of their relative (social) positions. This construal is morally problematic in that it betrays a narcissistic attention to oneself and a desire to dominate and control other people while losing sight of the intrinsic wrongfulness of the wrong—that it did harm to the victim (28-29).   To sum up, anger conceptually involves the wish for suffering; the wish for suffering commits either of two errors—the payback error or the status error—and is therefore unjustifiable. Since angry blame conceptually involves the wish for suffering, it is unjustifiable.   (2.2) Replies I want to first of all make two clarifications. The first one deals with the sense of “justifiability” of anger that Nussbaum is dealing with. Her sceptical argument against anger does not imply that anger is never apt. Roughly speaking, (moral) anger is an affective response to violations of moral expectations, and it is apt if the object to which it responds is a violation of moral expectations. Nussbaum’s critique—that anger commits either the payback error or the status error—does not touch on the aptness of anger. Her critique does not say that anger is a response to something other than moral violations, nor does it say that anger never accurately targets any genuine moral violations. The point that her critique makes is that anger may be problematic even when it accurately targets a genuine moral violation.50 For instance, would anger make the state-of-affairs worse for both the victim and the wrongdoer? Would it betray narcissism or a desire to manipulate?   																																																																		50 Srinivasan (2018) describes Nussbaum’s critique of anger as dealing with “reasons of prudence” of anger (127). However, some of the reasons that Nussbaum discusses are not prudential reasons; for instance, the “status error”—anger betraying narcissism—is not an error of prudence. 	 144	Second, I want to clarify once again that Nussbaum’s sceptical argument targets only angry blame. It therefore does not extend to blame that involves other emotions (e.g. disappointment) nor to affectless blame, where the former is prominent in personal relationships and the latter may occur in non-personal relationships. (Nussbaum recognizes that angry blame is merely one form of blame among others. See Nussbaum 2016, Appendix B).   With these two clarifications, I will now evaluate both claims of Nussbaum’s sceptical argument, starting with the second.   (2.2.1) Defensible Retributivism?  If we adopted her analysis of anger, taking anger to be necessarily a wish to inflict suffering on the wrongdoer in order to either improve the state-of-affairs or restore one’s relative status, then anger would seem to be problematic in personal relationships. We would face the challenge of explaining or at least making sense of the idea that the mere suffering of a spouse or a friend would be an appropriate response to cases of lack of care, disloyalty, untruthfulness, or lack of punctuality. And wouldn’t even the thought of “relative status” be inappropriate in personal relationships?   But Nussbaum does not think that anger is entirely hopeless. She thinks that anger, if it involves forward-looking elements, is justifiable. One form of forward-looking anger is “Transition-anger”, which she proposes and defends in the book. Another form of forward-looking anger that she finds palatable is provided by “expressive retributivism”. Proponents of expressive retributivism believes that “talk is cheap” and imposing suffering on the offender is the only proper way to express negative evaluations of wrongs and to affirm the value of morality. Anger is therefore a wish to use suffering for this purpose (190). Two expressive theories of retributivism that Nussbaum seems to be sympathetic with are Antony Duff’s communication theory (2001) and Jean Hampton’s moral education theory (1984).   Nussbaum has expressed doubt about the effectiveness of suffering as a means to express moral evaluation, but expressive retributivism can at least make a sensible case of the justifiability of anger. It makes sense to say, at least in some cases, that our anger involves a wish for suffering that partly aims to express our condemnation of the wrong and the values that it represents. And expressive retributivism makes sense in both personal and non-personal relationships. Anger may express our 	 145	condemnation of an abusive spouse, an authoritarian parent who does not respect our independence, a colleague who tries to gossip himself into a promotion, or a burglar who broke into our house.   But questions may remain. Such an understanding of anger as an expression of condemnation may make sense in non-personal relationships and some cases of personal relationships, but it seems to make less sense in cases of non-serious wrongs in personal relationships. It does not sound right to say that we are angry with a disloyal friend or an uncaring spouse because we want to express our condemnation of their wrongs through their suffering. In addition, our anger may at times not involve any forward-looking elements, such as in cases of serious wrongs and when we are very angry. Expressive retributivism would not be able to justify such cases of anger. Using expressive retributivism or other similar theories of retributivism to defend the justifiability of anger would therefore give us only a partial defence of anger: it may not work very well in cases of non-serious wrongs in personal relationships and in cases where anger does not involve any forward-looking elements.   (2.2.2) Specifying the Wish for Suffering  Nussbaum’s critique of anger, as far as the second claim (i.e. the wish for suffering is unjustifiable) is concerned, can be seen as a continuation of the age-old debate between retributivism and non-retributivism. This dissertation does not deal with retributivism; instead, it deals with how we respond to wrongs in personal relationships, so Nussbaum’s first claim (i.e. anger conceptually involves the wish for suffering) may be more pertinent to our discussion.   There is reason to question Nussbaum’s claim that moral anger necessarily (or “conceptually”) involves a wish for suffering. She cites a number of “first-rate thinkers” to support this claim, but we may have reason to ask for a more substantive defence. In fact, there are dissenting voices in the literature, some of whom are no doubt “first-rate thinkers”. Peter Strawson, for instance, observes that resentment and indignation (i.e. moral anger) “tend to inhibit or at least to limit our goodwill towards the object of these attitudes, tend to promote an at least partial and temporary withdrawal of goodwill” and involve “the modification… of the general demand that another should, if possible, be spared suffering” (2008, 23). In other words, anger does not necessarily involve ill will (the wish for suffering) but may involve merely reduced good will or indifference. An angry person may not want the person she is angry with to lose a finger or to be spat at; she just does not want him to win a 	 146	lottery or to have good luck in his job hunting. Or she is simply indifferent to him, having neither good nor ill will towards him.   Nussbaum’s first claim becomes even more questionable when we consider how anger is expressed in personal relationships. When we are angry with our sibling for leaving all the dirty dishes unwashed, or our friend for excluding us from a gathering, does our anger really involve a desire for their suffering as a form of “payback” for their irresponsibility and uncaringness? Or for the hurt that they cause us? This does not seem to be an accurate characterization of our anger we have to those close to us.   Anger in personal relationships may at times involve a wish for suffering, but the suffering in question has a more specific form that cannot simply be described as “payback”. Susan Wolf (2011) argues that the suffering that an angry person desires may be the “painful feelings of guilt and remorse”:   [A]lthough the angry emotions and attitudes do seem to me to be conceptually tied to a disposition to punish, and therefore with a willingness to make the object of blame suffer in a particular way, it would be a serious mistake to identify this with a general withdrawal of good will. Even in the midst of my fury at my daughter’s repeated raids of my closet, there was never a moment when I wanted harm to come to her, or when I was indifferent to her well-being. If I wanted her to suffer, it was in a specific way, with a specific kind of significance: I may have wanted her to experience the painful feelings of guilt and remorse. I never wanted her to break her leg, or even scratch her knee. (338, italic in original text)  Similarly, Amia Srinivasan (2018) argues that the suffering that we want to inflict on our friends when we are angry with them is the pain that we experience because of their wrongdoing:   Suppose my friend betrays me, and I am angry with her. I might want revenge. But might I not want—have we not all wanted—the friend to recognise the pain she has caused me, the wrong she has done me? It might be that this sort of recognition itself involves suffering. If so then, in a sense, I want my friend to suffer. But I don’t want her to suffer willy-nilly; my anger hardly calls out for her to break her leg, or fall ill. Rather I want her to experience that suffering that comes precisely from taking part in my own. (129)   The desires for the suffering of guilt and the pain of the victim are justifiable. To be repentant, one must acknowledge one’s wrong, and part of acknowledging one’s wrong is to feel guilt for one’s 	 147	wrong and to feel the pain that one has caused the victim. So to desire that the wrongdoer feel guilt and the pain of the victim is partly to desire that she repent. To the extent that repentance is valuable, anger, construed in terms of the desire for such suffering, is also valuable and justifiable.     (2.3) Summary Nussbaum’s sceptical argument against angry blame is unsuccessful. Contrary to the first claim of her argument, anger need not involve a wish for suffering; contrary to the second claim, the wish for suffering that angry blame may involve is justifiable. I therefore maintain that angry blame is not categorically unjustifiable.  	 148	Section 3 The Nature of Reasons from Relationships  The previous two sections discuss several reasons from relationships that govern responses to wrongs. This section aims to clarify the nature of these reasons and contrast them with the narrowly construed “reasons for and against blame” in the blame literature. This contrast will reveal the lack of attention to the actual workings of personal relationships in the blame literature51 and highlight the need for a broader scope of discussion.   (3.1) Reasons for Blame: Narrow and Wide  Suppose that you had lent your bike to your friend Dan for the weekend, and he promised to return it to you on Sunday. He forgot and did not return it till Tuesday, so for two days you could not use it and had to take the long walk to work. Do you have reason to blame Dan? Surely. He promised to return the bike to you on Sunday but didn’t. He had no excuse—it was not that he had something urgent come up or he was sick—and he is a full-grown adult, not a child. So he has clearly done something wrong to you and you have reason to blame him.   But, in another sense, you may not have reason to blame him. Yes, he has done something wrong by returning the bike late and making you walk for two days, and yes, walking to work is quite inconvenient, but is it really so big of a deal? You two are good friends. Will blaming him good for the friendship between you two? Is blaming him really worth it? Also, who has not returned things late? You may have forgot to give him back the money that you borrowed from him in the past.   This case helps illustrate the wide range of reasons for and against blame that we face when we have been wronged. One of which—one that most blame theorists spend most of their time if not exclusively on—deals with whether the person has in fact violated the relevant moral norms. In this case, you have a reason to blame your friend Dan to the extent that he has in fact violated norms that deal with returning borrowed property. He was supposed to return your bike on Sunday but did not; he had no legitimate excuse nor justification; and he is a fully developed moral agent, so the moral norms in question do apply to him.52 																																																																		51 For the sake of simplicity, I will discuss only reasons for blame, but the discussion applies also to reasons for/against other responses to wrong. 52 One may call such reasons “reasons of aptness”. 	 149	But there are other reasons governing whether we ultimately have good reason to blame a wrongdoer, even though the wrongdoer has in fact violated some moral norms. Angela Smith (2007) discusses three kinds of such wider range of reasons. First, one may have reasons not to blame the wrongdoer if one lacks standing to do so. For instance, you might not have good reason to blame Dan because it would be hypocritical—maybe you forgot to give him back some money last week. Or a stranger might not have reason to blame Dan because it is not this stranger’s business at all. Second, the moral fault in question may be too insignificant to warrant blame. Getting you to walk to work for two days is unpleasant, but it may not really be a big deal. Third, the wrongdoer’s response to his/her own violation may give us a reason not to blame him/her. If Dan has already reproached himself, it would be uncharitable for you to blame him (478-483). In all of the above cases, although the wrongdoer—Dan—has wronged you, but you should, all-things-considered, not to blame him after all.   The reasons from relationships discussed in the first section of this chapter—relationship maintenance, concern, trust, and understanding—are among this wider range of reasons. They do not deal with whether the person in question has in fact violated any moral norms; in fact, they presuppose that the person has done some wrong. Instead, they deal with factors that determine how one should respond to the wrong. For instance, after our friend has been caught stealing, we may have reason to show her support rather than join others in blaming her, even though she has in fact done something wrong.   (3.2) Reasons for Blame: Narrow or Wide?  Discussion in the blame literature has primarily focussed on reasons that deal with whether the person has in fact violated any moral norms but rarely mentioned this wider range of reasons that govern whether we should blame. If we look at the works of the influential figures in the literature, we will notice that they primarily treat the issue of appropriate blame as a matter of whether the person has truly violated any moral norms. Strawson (2008) observes that blame is an affective response (the reactive attitudes) to the quality of will revealed in the behaviour of the wrongdoer, and blame is appropriate so long as her behaviour truly reveals ill will (i.e. no excuse nor justification) and if she is an appropriate object of our demands for good will (i.e. no exempting conditions). Wallace (1994), taking inspiration from Strawson, likewise argues that blame is appropriate if (i) the person in question has truly violated a moral obligation (i.e. she does not any 	 150	legitimate excuse) and (ii) she possesses the power to grasp and apply moral reasons and the power to control or regulate her behaviour in the light of such reasons (i.e. she is a morally responsible agent) (85). Sher (2006), though taking an approach to blame different from that of Strawson’s and Wallace’s, treats reasons of blame as a narrow matter just as they do. Sher argues that blame involves (i) a belief that the wrongdoer has acted badly or has a bad character and (ii) a backward-looking desire “that the person in question not have performed his past bad act or not have his current bad character” (95; 112). Blame is therefore appropriate if (i) the wrongdoer has truly acted badly or has a bad character, rendering the belief involved in blame true and (ii) the moral principles that have been violated by the bad act or the bad character are justified, rendering the backward-looking desire appropriate (130).  Scanlon’s account (2008) is an exception among the accounts that I have discussed in that his account has to an extent extended into a wider range of reasons (he calls issues dealing with such reasons the “ethics of blame”) (2008, 166-179). Scanlon argues that wrongs impair relationships, and the impaired relationship renders appropriate some adjustment of attitudes and dispositions; such adjustment is blame. One reason that warrant such adjustment is of course whether the wrongdoer has truly wronged the victim and impaired the relationship, but he has discussed other reasons, such as generosity and standing to blame. He has also briefly discussed some reasons that deal with parent-child relationships, such as loyalty, sympathy, and encouragement (171-174). Such reasons are among the wider set of reasons that I have been discussing.   I certainly agree with the direction of Scanlon’s ethics of blame. What I am proposing is that such ethics of blame could and should be further developed, and in particular with a focus on reasons from personal relationships. In addition, discussion of reasons from (personal) relationships should not focus narrowly on reasons for and against blame, but should be widely construed in terms of appropriate responses to wrongs.   (3.3) Wide Not? Why are most blame theorists not interested in the wider range of reasons? I suggest that there are two reasons. First, theorists see their accounts of blame as part of the free will project. Second, theorists take a necessary-and-sufficient-condition approach to blame.   	 151	The free will problem deals with the issue whether we would be morally responsible for what we do (and what we think, who we are, etc.) if determinism were true. A number of theorists suggested that the free will problem can be resolved as a blame problem. The pioneer of this approach is, of course, Peter Strawson. His seminar paper “Freedom and Resentment” (2008), which has inspired the whole literature on blame, was originally meant to be a paper on the free will problem. There he proposes that the right approach to the free will problem is to start by asking how we blame in our moral lives, and then ask, according to our practice of blame, whether the truth of determinism is compatible with moral responsibility. Since our practice is such that we withhold blame only if excusing or exempting conditions hold, and the truth of determinism would not entail the truth of excusing or exempting conditions, moral responsibility is compatible with the truth of determinism.   Due to the influence of Strawson, blame has been regarded by many blame theorists as more a part of the free will project than an independent project. Wallace (1994) and, more recently, McKenna (2012) are notable examples. Blame is slowly becoming an independent project, but even when theorists treat blame as a (relatively) independent project, they feel the need to at least give a nod to the free will problem. Scanlon (2008) and Sher (2006) are good examples. If blame theorists treat blame not as an intrinsically valuable project but as merely a solution to the free will problem, then they will naturally not be interested in the wider range of reasons. Such wider reasons that deal with how we should blame, when we should blame, who has the standing to blame, and relationship contexts are simply irrelevant to their concern.   Second, even when theorists treat blame as an independent project, they take a necessary-and-sufficient-condition approach to it. One main goal of most blame theorists is to explain the nature of blame, that is, the necessary and sufficient conditions of blame. Their task is to explain the response that we called “blame” when and only when wrongs occur; all “extraneous components” are to be sifted out. Those wider reasons that we discussed are among those that are irrelevant to theorists’ purposes and are to be sifted out.   There are of course theorists who recognize the importance of and have given some attention to such wider reasons. But, first, most of their discussions do not deal with reasons specific to personal relationships. For instance, Smith (2007), G. A. Cohen (2006), Wallace (2010), and Gary Watson (2012) discuss general reasons such as moral standing, insignificance of fault, etc. Second, for those 	 152	few theorists who do discuss reasons specific to personal relationships, their discussion is far from adequate for something as complex as personal relationships. For instance, Scanlon (2008) briefly argues that parents may have more reason to encourage and trust their children than to blame them, but his brevity does not do justice to the complexity of parent-child relationships. It might rather indicate that he gave more attention to the free will problem than reasons from relationships.    Personal relationships are an important part of our lives. They govern how we should treat one another and how we should respond to mistreatment. If we take blame seriously and start treating it as an independent project, and if we take personal relationships seriously and start thinking carefully about how blame works in personal relationships, we will soon realize the depth and complexity of blame in personal relationships and the considerations we need to take into account when determining how to respond to wrongs. There is so much more to blame in our personal lives than whether moral norms have truly been violated. This chapter aims not only to explore such wider range of reasons that govern blame in personal relationships but also to call for the attention that such reasons deserve.  	 153	Section 4 The Role of Reason in Personal Relationships I have been arguing in this chapter that personal relationships give us reason to respond to wrongs in various ways. Some might find this approach to personal relationships all wrong headed. They might think that personal relationships should be guided by emotions and desires but not by “reason”.   While the considerations that I discussed in the first section (relationship maintenance, concern, etc.) need not function as reasons but may function as emotions and/or desires, reason does play a valuable and significant role in personal relationships. It justifies what we do to one another and at times motivates us to do what we should. This section discusses and defends the importance of reason in personal relationships.   I will first clarify the issue at stake in (4.1). The kind of “reason” that one might find problematic in personal relationships is not justifying reasons but motivating reasons (that is, being motivated by a sense of duty), and such motivating reasons are problematic only if they are abstract but not if they are particularized. In (4.2), I will discuss more generally the role of (motivating) reasons in personal relationships. I argue that a psychologically realistic theory of personal relationships would recognize the value of motivating reasons. Reasons serve as valuable fall-back mechanisms (Hardwig 1989, 76) when we face weakness of will or moral indecision in our dealing with those who are close to us.   (4.1) Reasons: Justifying and Motivating; Abstract and Particularized There is an extreme romantic view of personal relationships (notably love) according to which there is no reason nor rationality in personal relationships; there are only “feelings, nothing more than feelings”. So everything should come down to emotions and desires. Talks of reason are not only cold but also simply wrong.   This extreme romantic view paints an inaccurate if not misleading picture of personal relationships, and may skew our conception of what is appropriate and inappropriate in our personal relationships and our expectations of those close to us. There are people who hold this view (especially as applied to love), but it can barely stand any critical assessment. There is at least one sense in which reason is appropriate in personal relationships: the justifying sense. Justifying reasons are, simply put, considerations that count for or against something. For instance, the importance of a parcel gives me 	 154	a justifying reason to go to the post office today. Justifying reasons are ubiquitous in personal relationships. When we say things such as “I should help my friend move houses”, “I should go home to take care of my sick child”, or even “I should feel bad about lying to my sister”, we are already talking reason in this sense.   But there seems to be something right about this idea. Especially in close personal relationships, don’t we think that we should help our friends and spend time with our spouse because we want to, rather than because we think we should? Perhaps the idea is that I should not be motivated by reasons—in the sense of thoughts of duty—such as “I should help my friend move houses” but should instead be motivated by feelings and desires. We should “feel like” and “want” to help our friends (Hardwig 1989, 74).   Why is it that we should not be motivated by thoughts of duty in personal relationships? I think two examples given by Bernard Williams (1981) and Michael Stocker (1976) can help illustrate the idea. Though their examples were originally meant for discussion about general moral theory and the moral worth of act, they can give us some insights into personal relationships.  Williams (1981) discusses a case where someone has to choose between saving his wife’s or a stranger’s life. This person would have “one thought too many”, as Williams puts it, if he makes his decision to save his wife based on the thought that doing so would be “morally justified”. Theorists may have different ideas on how exactly to understand Williams’s claim, but the general idea seems to be that the person should not think that he needs to “morally justify” favouring his wife. He should not consider anything more than that one of the two people is “his wife” (17-19).   Stocker (1976) has given another oft-cited case that may help shed more light on the problems with motivating reasons in personal relationships. Suppose that you are sick in the hospital and your friend, Fred, comes and visit you. You are very pleased that Fred comes and visits you, but suppose Fred says that he comes and visits you not because he is concerned about you and wants to see you, but because “it is the right thing to do” (e.g. it is his duty as a fellow Christian or Communist, or it is the best way to maximize overall happiness (462)). Stocker takes it that you will be very disappointed when you hear this and your enthusiasm will immediately vanish.   	 155	Note that the problem is not merely that the duty in question is a non-personal duty (e.g. duty as a fellow Christian). You would be just as disappointed if Fred visited you because “friendship requires it”. The problem may instead be that Fred’s visit is motivated by abstract reasons. Recall one of the constitutive features of personal relationships which I discussed in chapter 1 section 2: personal relationships involve seeing someone as a particular person rather than a general, abstract moral agent. If Fred’s visit were motivated by thoughts such as “it is my duty as a fellow Christian”, or “friendship requires me to do so”, he would be motivated by general, abstract reasons, and in turn seeing you as a general, abstract moral agent rather than as a particular, “concrete” person—his friend. This way of seeing you is “dehumanizing” (Stocker 460); “the concrete person [i.e. you] drops out of the person” (Pettigrove 2011, 200).  If, on the contrary, Fred were motivated by reasons that come from your needs, desires, etc. that are particular to you, you might not be disappointed by him after all. Suppose Fred further explained why he came to visit you: “Coming to visit you is the right thing to do, my friend. It will make you feel better. Your happiness and well-being matter to me, and they give me reason to come here.” Hearing his explanation, I imagine that you would not think that he had “one thought too many” because, as Fred said, it is your happiness and your well-being, not anyone else’s, that motivate him.53   The above discussion show us that it is only abstract, general motivating reasons that are out of place in personal relationships. “Particularized reasons”, reasons that come from the needs, desires, etc. of particular persons, are not inappropriate in personal relationships. However, our discussion so far has shed more light on what is wrong about abstract reasons than on what is right about particularized reasons. There might be alternative interpretations of Williams and Stocker which could reject both abstract and particularized reasons in personal relationships. To further strengthen my case about the appropriateness of particularized (motivating) reasons in personal relationships, I will directly engage the question: what role does reason play in personal relationships? The answer lies in the undeniable fact that we are all imperfect beings. We often do not feel in the right way, have the desire to do the right act, or even know what the right course of action is.54  																																																																		53 In fact, Stocker even goes further and admits in a later paper (1981) that being motivated by “friendship” rather than by the particular friend may still constitute a friendly act so long as there is a background sensitivity to the friend’s well-being, needs, etc. (762). 54 Note that I do not and need not endorse a non-cognitive account of emotions, according to which emotions are beyond rational control. I am only arguing that, even on a cognitive understanding of emotions such that emotions are susceptible 	 156	(4.2) Role of Reasons in Personal Relationships  There are at least three ways in which reasons (both generalized and particularized ones) play an important and appropriate role in personal relationships. The first one deals with cases of weakness of will, where we lack the desire to do what we should. For instance, we may know that we should visit our sick friend in the hospital, but we just don’t want to—we would rather stay home and sleep. But our sense of duty, the thought that “I should go because I am his friend”, may override our desire to stay home and end up motivating us to visit our friend. Our visit, motivated by a pure sense of duty, might not be as good as a visit motivated by concern about our friend and a desire to see him and cheer him up, but it is an appropriate act of friendship after all because it is motivated by a sense of what friends should do to one another.   The second deals with moral indecision. Sometimes we may be torn between two courses of action and cannot determine which one we should take. For instance, we may be unable to decide if we should go to visit our sick friend or stay home to help our spouse with their work. In such situations, one may try to decide by considering “what would a good friend/husband/wife do?” or “what does friendship/loving partnership require me to do?”, to which one may answer “a good friend would go and visit the sick friend in this situation”, and one may then be motivated by this thought to decide the sick friend. Deliberating and being motivated in this way in such situations are appropriate. There seems nothing wrong if your friend Fred said, “I didn’t know if I should come or stay home to help my wife with her work, but I thought that coming to visit you is what a good friend should do, so here I am.”   Third, reason needs to step in when feelings do not serve as a reliable guide of action. Consider personal relationships that are full of frustration and disappointment, namely parent-child relationships and spousal relationships. We often have many and a high degree of expectations of others in such relationships, but, time and again, they will let us down and test our patience. If we let feelings and desires guide all of our decisions, conflicts and arguments would never end and the relationship would not possibly be sustained. A sense of duty is hence necessary in these relationships. It will help us stay calm and be more alert to what we should do and how to maintain the relationship.  																																																																																																																																																																																																																																						to rational control, there will be times when we fail to have the appropriate emotions or emotions fail to inform us of the right course of action. In such cases, a sense of duty can play a crucial role in guiding our behaviour. 	 157	I will use the example I gave in the section 1 to illustrate this point. Your daughter has been caught embezzling a large sum of money in the firm that she was working for. You are furious and very disappointed that her disgraceful behaviour has caused great harm to herself and her company. Understandably, you may have a strong desire to condemn her behaviour or even cut ties with her. But letting your desires decide what to do might not be the best course of action as a parent. As her parent, maybe you should give her your support and help her learn from her mistakes. Having a sense of duty and reflecting on where one’s reasons lie may help you keep your anger (appropriate though it may be) in check and make the right decision.    Even in situations where feelings are appropriate, they may need to work in tandem with reason. I will use parent-child relationships as an example again. Love, care, and desiring the best for one’s child are appropriate in parent-child relationships. But, at the same time, parents should respect the autonomy of their children (particularly adult children). Love and care, if too much, would become interference in the child’s personal life and infringe her autonomy; they therefore need to be guided by a sense of duty.   Weakness of will is not uncommon in personal relationships. When we have been wronged by those close to us, we may have a strong desire to blame them even though it might not be the best response to the wrong. Such desires might be all the stronger exactly because the person who wronged us is someone close to us. Moral indecision is also common in personal relationships. Personal relationships are full of problems and complications, giving us countless hard decisions where we are faced with various conflicting obligations, responsibilities, and interests. Our feelings, even appropriate one, also often need to be guided by a sense of duty lest we overstep.   A realistic account of personal relationships should therefore take this into consideration. It should recognize that we are all imperfect beings and we at times need to use our sense of duty to navigate our moral lives in personal relationships. It should recognize the necessity and the value of reason as a source of motivations in personal relationships.  However, even though reason is not inappropriate in personal relationships, it may be merely a “fall-back mechanism” (Hardwig, 1989, 76) such that it should not be triggered too often. If situations such as weakness of will arises and the fall-back mechanism is triggered too often, it may indicate a 	 158	lack of appropriate feelings and desires and, in turn, a lack of care. If a friend, a lover, or a parent frequently lacks the proper desire to do what they should, it may be an indication that the underlying concern and care crucial to personal relationships are lacking. It is true that personal relationships are complicated, but dramatic scenarios do not occur in real life as often as in academic philosophy. Similarly, moral dilemmas do occur in personal relationships every once in a while, but if we frequently experience moral indecision, chances are there is something wrong with our decision-making, or the underlying care and concern of the relationship, or both.   (4.3) Summary Reason plays an important role in personal relationships. In the justifying sense, reason determines what we should and should not do. In the motivating sense, there may well be nothing inappropriate with being motivated by reason if they are the right kind of reason. Reasons from relationships may also serve as a fall-back mechanism and be appropriate motivations so long as such mechanisms are not used too often. We are imperfect beings and will at times need to rely on fall-back mechanisms in our dealing with those close to us. Any psychologically realistic theory of personal relationships needs to take our imperfection into account.     	 159	Chapter Conclusion This chapter argues that there is a multiplicity of reasonable responses to wrongs in personal relationships. In section 1, I discussed four reasons that feature in personal relationships and show that each of them calls for various responses to wrong, many of such responses are not blame or forgiveness. I also argued that the idea of reasons from relationships against blame does not necessarily amount to victim-burdening, but I recognize the legitimacy of this worry and suggest that we approach this idea with caution.   I then discussed other issues related to reasons for and against blame in the rest of the chapter. In section 2, I discussed and rejected Nussbaum’s sceptical argument that angry blame is unjustifiable. I argued, contra Nussbaum, that anger may be a valuable and justifiable response to wrongs in personal relationships.   Section 3 clarified the nature of the reasons under discussion. The reasons governing responses to wrongs that I have been discussing deal not merely with whether wrong has indeed been done but a wider range of reasons. I argued that such a wider range of reasons play an important role in the workings of personal relationships and therefore deserve more attention than they have been given in the blame literature.   Section 4 discussed the role of reason in personal relationships. I argued that reason plays both a justifying and motivating role in personal relationships. A psychologically realistic theory of personal relationships should recognize the necessary and valuable role that reason plays in motivating behaviour. 	 160	Chapter 5 Respect I argued in the last chapter that personal relationships may give us reasons to respond to wrongs in ways other than blame, e.g. forgiveness, or setting blame and forgiveness entirely aside. A number of theorists, however, have argued that withholding blame against a(n) (unrepentant) wrongdoer is morally problematic because it shows a lack of respect for oneself and for the wrongdoer. This argument, if successful, may cast doubt on the legitimacy of non-blame responses to wrong in personal relationships.   This chapter aims to defend the legitimacy of non-blaming responses in personal relationships. Just as there are a multiplicity of appropriate responses to wrong, so too there are a multiplicity of appropriate ways to respect oneself and the wrongdoer—blame is merely one among many ways to do so.   I will start in the first section with a brief discussion of the concept of respect, with particular attention to its two important aspects—dignity and agency. In section 2, I will explore some typical arguments that theorists have given to establish the claim that withholding blame shows disrespect for oneself and the wrongdoer. Then, in section 3, I will argue that we should take a broader approach to respect and recognize that blame is merely one way among many to show respect in cases of wrongdoing. I will argue that the non-blaming responses to wrong that I discussed in the last chapter are alternative ways to respect the dignity and agency of oneself and the wrongdoer. 	 161	Section 1 Respect as Dignity and Agency Though respect is an important concept, discussion of respect in the blame and forgiveness literature does not always rely on a clear conceptual framework of respect, so theorists may at times be talking past one another. I do not intend in this section to give a complete account of respect, nor do I think it would be possible. What I am going to do is instead to highlight two important aspects of respect—dignity and agency—which I think theorists generally take as essential to respect (1.1). In (1.2), I will briefly describe how these two aspects of respect can be applied to personal relationships.  (1.1) Respect: Dignity and Agency The term “respect” is used very loosely. We respect the law; we respect the prime minister because he/she is the leader of the country; we respect our enemy for his good strategic manoeuvres; we respect a good skier for her skiing skills; we respect a honest person for his honesty; and we respect the rights of a person. We also say we respect ourselves: we respect ourselves as a member of a profession (scholar, lawyer, forester, etc.); we respect ourselves as a good skiier; and we respect ourselves as a person with rights.   Stephen Darwall’s widely discussed article “Two Kinds of Respect” (1977) may help us categorize the wide variety of respect listed above. He observes that respect can be categorized into recognition respect and appraisal respect. Recognition respect is the kind of respect most relevant to our discussion. We recognition-respect something (say a person or a non-animate object) by giving appropriate weight to some relevant features of the object in our deliberations and act accordingly. For instance, when one recognition-respects a referee as the referee of a match, one will take into account the referee’s role as the referee and will act accordingly. One would therefore take his decisions seriously and defer to his decisions. Or one may recognition-respect the property right of a person and will then take this into account when deciding how to treat the person’s property (e.g. whether to return her lost wallet to her.) Other examples of recognition respect include respect for the law, a snow-capped mountain, the prime minister, and a member of a profession. Appraisal respect, by contrast, is a positive evaluation of a person’s traits and/or character. For instance, one may appraisal-respect a referee as fair and skilful in handling disputes, or appraisal-respect an enemy as fierce and crafty. Other examples of appraisal respect include respect for a good skier, a honest person, a handy electronic device, etc.   	 162	Darwall emphasizes that respect involves both “regarding” and “responding”. Respect involves regarding the object as worthy of respect (e.g. by giving appropriate weight to the features of the object in one’s deliberations), but merely “regarding” is not sufficient for respect. To use the referee example again, if one regards the referee as the referee of the match by taking his position into one’s deliberation, but refuses to defer to his decisions or even tries to sabotage his position as a referee, one is not really respecting the referee. Respect therefore requires proper response from the respecting person. One may respond to an object of respect by acting in certain ways or have certain emotions. For instance, deferring to the referee’s decision, or having a feeling of awe towards the fair-mindedness of the referee. For Darwall, one properly responds to an object of respect by acting according to the appropriate weight that one gives the object in one’s deliberation.   Among the various kinds and instances of respect discussed above, the one that is most relevant to our discussion about blame is recognition respect for persons as persons. We hold (or should hold) such respect for all persons (including ourselves) regardless of their social positions, their traits and characters, and whether we are friends or foes. We respect them simply because they are persons, that is, they have features that are shared by all persons in virtue of being persons. But what are the features which all persons have in virtue of being persons? And which of these do we have to properly regard and respond to if we are to respect them? The answer to this question depends on the concept of persons that one adopts, but theorists tend to take a broadly Kantian approach. First, a person is a being with dignity. They are a being with high moral worth deserving to be treated in certain ways, and their moral worth is equal to all other persons; that is, persons are all moral equals. Second, a person is a being with agency. A person is capable of judging what is worthy and what is not, and acting in light of such judgements.   According to this broadly Kantian approach, to respect a person as a person is therefore to properly recognize and respond to the dignity and agency of the person.55 One might argue that this Kantian concept of person is merely one concept of person and gives us merely one account of respect for persons. There may well be other concepts of persons generating alternative accounts of respect for persons (Dillon 1992, 113). However, this broadly Kantian account of respect is widely shared in the 																																																																		55 Dignity and agency are closely related. A Kantian would say that agency is the basis of dignity. It is because persons are agents capable of acting in light of their judgements that they deserve to be treated in some ways but not others. But dignity and agency remain largely conceptually distinct, and I will treat them as different concepts in my discussion. 	 163	blame and forgiveness literature, so this account is sufficient for our purposes. Moreover, theorists have good reason to use this account because dignity and agency do seem to be two crucial aspects of respect. It seems hard to imagine that we would be respecting a person as a person if we did not properly their dignity and agency.  But how exactly do we “properly recognize and respond” to the dignity and agency of others and ourselves? This is where things are tricky and contentious. It seems fairly obvious that respecting other people’s dignity requires us to recognize their moral worth such that they deserve to be treated in some ways but not in others. It certainly also requires us to treat them appropriately, e.g. avoiding harming them, keeping our promise made to them, etc. Similarly, respect our own dignity requires us to recognize ourselves as deserving to be treated in certain ways but not in others; it also requires us not to allow ourselves to be mistreated without good reason. As for agency, respecting other people’s agency requires us to recognize their capacity to act in light of their own judgements. This means that we should avoid intervention without good reason. It may also mean that we should help protect their agency and create an environment where their agency will be improved and exercised properly. Similarly, respecting our own agency requires us to recognize our capacity to act in light of our judgements and to protect, develop, and try to exercise our agency properly.   What is tricky and contentious—and most pertain to our discussion—is whether dignity and agency require us to blame the wrongdoer when she has wronged us. In the next section, I will discuss and evaluate some arguments for the claim that dignity and agency, and in turn respect, require us to blame the wrongdoer. But before I move on to that, I will add a short note on respect in personal relationships.   (1.2) Respect in Personal Relationship “Respecting persons as persons”, “dignity”, and “agency” may all sound concepts that belong only in non-personal relationships, but respect is applicable to personal relationships. First, we should respect those close to us as persons just like we should respect strangers as persons. As Hampton (2007) rightly observes, there is justice in personal relationships. We should treat our friends and partners as persons with intrinsic worth such that we should not assault, enslave, or belittle them. (Hampton even goes further and argues that genuine personal relationships are possible only if the persons in the relationship are treated as persons with intrinsic worth (31).)  	 164	 Second, we should respect our intimates as our intimates: parent, sister, husband, etc. This is an analogue of respect for persons and can be similarly analyzed in terms of dignity and agency. In common parlance, “respect your parent” or “respect your husband” may have a ring of hierarchy to it, but it can be properly understood as treating them appropriately (dignity) and respecting their agency as our parent or husband, over and above giving them the minimal respect that is owed to all persons. For instance, caring about whether our parents fare well and offering help when they need it are certainly ways to treat them appropriately as our parents, and not allowing one’s husband to have a life outside home and work (e.g. spending time with his friends or developing a hobby) is a clear case of not respecting his agency as a husband. Similarly, self-respect has a proper sense in personal relationships. If someone always obeys his parents without any objection, even their most unreasonable whims, thinks that his wife’s interests should always take precedence over his, or will gladly allow his friend to take advantage of him in order to please them, his attitudes seem to be exemplars of not respecting oneself as a child, a husband, and a friend.  Personal respect and non-personal respect may sometimes conflict. Consider the potential conflict between “respecting personal boundaries” and “showing concern”. Suppose a friend won a prize and was considering spending the money on buying a luxurious car, rather than spending it on paying down their (hefty) debts or on their children’s education. Should you avoid being intrusive and do nothing other than congratulating them? Or should you show your concern through your financial advice. There may not be a general answer to the question whether personal respect should take precedence over non-personal respect. We might be able to judge only on a “case by case” basis. It depends on what is at stake—whether it is a “life-and-death” issue or a small matter. It also depends on the nature and how close relationship is.   	 165	Section 2 Have Some Respect and Blame  The two aspects of respect discussed in the last section—dignity and agency—can help us understand the arguments that theorists have used to support the claim that withholding blame disrespects oneself (i.e. the victim) and the wrongdoer. I will explore and examine their arguments in this section. To facilitate the discussion, I will first briefly clarify in (2.1) how the reasons from relationships that I discussed in chapter 4 involve withholding blame. I will then discuss two main kinds of arguments that theorists have advanced—psychological (2.2) and normative (2.3). My discussion of these arguments will pay specific attention to their reference to dignity and respect.   (2.1) Reasons from Relationships and Withholding Blame  Two components of blame are relevant to our discussion. First, blame involves the belief that one has been wronged; second, blame involves an affective (e.g. anger) or conative (e.g. a desire and intention to distant oneself from the wrongdoer) response to the wrong, depending on the relationship context. One may therefore withhold blame by (i) not having the belief that one has been wronged, (ii) not having the affective/conative response to the wrong, or (iii) not having both.   Not holding the belief that one has been wronged when one has in fact been clearly shows disrespect for oneself. (E.g. thinking that there is nothing wrong with a good “friend” spreading rumours about you, or a colleague denying your right to express yourself at an important work meeting, etc.)This may indicate that one does not see where one’s rights and claims lie.   What is less obvious and more contentious is whether withholding blame in the second way shows disrespect for oneself and the wrongdoer. If one believes firmly that one has been wronged and deserves to be treated in a better way, but does not get angry (either feeling it privately or expressing it publicly) nor change one’s intentions, expectations, or behaviour towards the wrongdoer, would one be disrespecting oneself or the wrongdoer? If it did amount to disrespect, then the non-blame responses to wrongs discussed in the last chapter (e.g. not expressing or even feeling one’s anger in order to maintain a relationship) might in the end be inappropriate.     	 166	(2.2) Psychological Argument Several theorists have argued that the disrespect that withholding blame involves (by not having the relevant affective/conative response) lies in the person’s psychology—that she, deep down, does not respect herself. Murphy claims that “proper self-respect is essentially tied to the passion of resentment, and that a person who does not resent moral injuries done to him… is almost necessarily a person lacking in self-respect” (1988, 16). In later writing, he adds that the victim’s “lack of resentment reveals a servile personality—a personality lacking in respect for himself and respect for his rights and status as a free and equal moral being” (2003, 19). Similarly, Haber (1991) claims that resentment is a “natural [expression] of aversion to mistreatment” (89) such that self-respecting person will express her self-respect “in the form of resentment when she is the object of moral injury” (73). In the same vein, David Novitz (1998) points out that “people who forgive too readily… do not manifest the right degree of self-respect; they underestimate their own worth and fail to take their projects and entitlements seriously enough.”  The argument launched by these theorists seem to be the following. (1) Failing to blame (e.g. failing to be resentful) when one has been wronged indicates that one does not see oneself as a moral equal and thus lacks self-respect. (2) Lacking self-respect is morally objectionable. (3) Since withholding blame indicates lack of self-respect, withholding blame is thus also morally objectionable (Pettigrove 2012, 113). (This psychological argument can also be applied to other-respect—that withholding blame indicates that one does not see the wrongdoer as one’s moral equal and as a moral agent, and therefore respect for others requires us to blame them when they wrong us.)    There are at least two problems with this psychological argument. First, in relation to claim (1), it is not obvious that withholding blaming always or even often indicates that one does not value oneself enough. If, after a long day of work, I feel no anger at the driver who ran a red light, does my lack of anger really reveal my “servile personality”? Or if an exceptionally patient person feels no anger nor disappointment at his flaky friend, does it necessary imply that this person is a “doormat”?   Second, in relation to the other two claims, claim (3) does not follow from claim (2). As Pettigrove observes, “the qualities of the origin of an activity do not translate so neatly into the qualities of the activity” (2012, 114-115). Someone may be extremely polite, for instance he may ask for permission before using things that belong to other people (consider table salt or sugar in an office pantry), or he 	 167	will make sure to give other people the option to say no even when he is making a very small proposal (consider after work drinks or making tea for them). His extreme politeness may come from his timidity or over-cautiousness. But his timidity and over-cautiousness do not necessarily imply that he should not be polite, nor that his politeness is any less valuable. Whether he should be polite in those ways should be determined by whether he has good reason to be polite. If he does have good reason, then he should continue to be polite but at the same time work on his timidity and over-cautiousness. Similarly, someone may refrain from blaming the wrongdoer because she does not have enough self-respect, but this does not necessarily mean that she should not refrain from blaming nor that her refraining blame is morally objectionable. Whether she should refrain from blaming should be determined by whether she has good reason to.  (2.3) Normative Arguments (2.3.1) Self-respect I believe that the normative argument are stronger and more interesting than the psychological arguments. Rather than claiming what withholding blame reveals about our psychology, a normative argument deals with the appropriate way to express (privately and/or publicly) our respect for ourselves. Blaming the wrongdoer when we have been wronged is an appropriate way to express our respect for ourselves, whereas withholding blame expresses disrespect for ourselves.    But why is it that withholding blame expresses disrespect for ourselves? One might argue that withholding blame amounts to approving the wrong. For instance, Kolnai (1973, 96) argues that failure to blame a wrong is “acquiescing” in it, and acquiescing is “virtually conniving” in it—in other words, approving it. While approving of the wrong that one has been subjected to is clearly not properly valuing oneself as person who should be treated appropriately, we do not have good reason to believe that withholding blame necessarily amounts to approving of the wrong. If an exceptionally nice and calm person did not feel angry at a gossipy colleague but continued to maintain a good work relationship with him, why would her non-blaming response necessarily amount to approving the bad-mouthing, rather than reflecting her patience and generosity?   Other theorists argue that blame is an appropriate way to respect oneself because blame defends and shows commitment to one’s value. First, observe the insulting if not degrading messages that wrongs may express. When someone drives recklessly and puts everyone else’s safety in danger, takes other 	 168	people’s property, blasting loud music late at night, or lies and manipulates others in order to achieve some career goals, such behaviours are communicating the messages that “I count but you do not,” “I can use you for my purposes,” (Murphy 1988, 25) or, simply put, “I may do such things to you and there is nothing wrong with it.” Such messages are certainly outrageous and should be rejected. Blame is then a response that seeks to reject such messages and to defend the value of oneself (Murphy 1988, 25; Hieronymi 2001, 546-547) and to emotionally express that one does not endorse such insulting/degrading messages (Murphy 2003, 44). In other words, blame is a way to show commitment to the worth of the victim.   There are at least two ways to understand these claims about commitment, both are problematic. First, theorists may be claiming that withholding (private or public) blame necessarily shows lack of commitment to the worth of the victim. This claim seems implausible. If I patiently wait at the bus stop for my tardy friend rather than steaming with anger, or if someone with a growing family refrains from blaming his boss in order to focus on his work and avoid trouble (Thomas 1989b, 161), why would withholding blame in these two cases show a lack of commitment? As Holmgren (2012, 69-70) observes, if withholding blame did necessarily amount to a lack of commitment to one’s value, then respect would require us to blame each and every wrongful act that we encountered, and this would be absurd. We can at most say that regularly withholding blame or withholding blame against serious wrongs may amount to a lack of commitment to the value of the victim (and to approving the wrong), but this is far from the claim that withholding blame is always disrespectful.  The second possible interpretation is that withholding (private and public) blame sometimes (or often) shows a lack of commitment to the worth of the victim. While this interpretation sounds more plausible, it implies that withholding blame would not be normatively problematic when it does not show such lack of commitment. But when would withholding blame not show such a lack of commitment? Would it be when one withholds blame because of one of the reasons from relationships discussed in the last chapter? Moreover, even when withholding blame did show some degree of lack of commitment to the worth of the victim, would those reasons render it all-things-considered appropriate to withhold blame? Either way, there is room for the legitimacy of withholding blame in personal relationships.    	 169	(2.3.2) Other-respect Theorists have also argued that withholding blame amounts to disrespecting the wrongdoer. A number of them have argued that to blame a wrongdoer is to see the person as one’s moral equal and is hence a form of respect. As Strawson puts it, the reactive attitudes (i.e. a form of blame) is exactly “the consequence of continuing to view him [the wrongdoer] as a member of the moral community; only as one who has offended against its demands” (2008, 23). Withholding blame, on the contrary, may amount to adopting the “objective attitude”, which is an attitude that sees the wrongdoer as “an object of social policy” or as “a subject of treatment” (9). Hieronymi has similarly argued that blame is necessary for respecting the wrongdoer as a moral equal. She observes that “so long as you continue in your standing as moral peer, I will protest [the wrong that you have done to me].” I won’t protest it (that is, in her understanding, not blaming you) “only if it concerns an unimportant matter or if you fall from the status of moral peer.” This is because “[r]espect for you as a fellow human being commits me to caring about your evaluation”, whereas “[t]o disregard your evaluation is to disregard you” (2001, 549)   Other theorists build their argument based on moral agency. Part of being a moral agent is to be capable of governing oneself by moral reasons and reflecting the quality of one’s will via one’s acts (Scanlon 2008, 168; see also Mason 2014). Blame presupposes these two aspects of moral agency and is therefore a way of recognizing the wrongdoer’s agency. By contrast, not blaming the wrongdoer when she has mistreated me amounts to treating her as lacking moral agency. Or, as Strawson puts it, it amounts to taking the objective attitude towards her and treating her as merely someone to be “dealt with”. Treating the wrongdoer as lacking moral agency when she is a moral agent is patronizing and clearly disrespectful, so withholding blame disrespects the wrongdoer.   Like the normative arguments that deal with self-respect, these arguments face the same problems. First, they are implausible if they are claiming that withholding (private or public) blame necessarily amounts to treating the wrongdoer as a moral inferior or patronizing her as lacking moral agency. These arguments do not provide us with good reason to believe that blaming a tardy friend or a gossipy colleague is the only way to express respect for their agency. Again, if respect for agency did require us to blame every wrong, we would need to spend so much of our time and energy on blaming every wrong that we encountered.   	 170	Second, if the arguments are claiming that withholding (private or public) blame sometimes (or even often) treats the wrongdoer as a moral inferior or patronizes her as lacking moral agency, then there will in fact be room for the legitimacy of withholding blame. Withholding blame may not be problematic when it does not treat the wrongdoer in such ways, and perhaps it does not treat the wrongdoer in such ways exactly when blame is withheld because of one of those reasons from relationships.   (2.4) Summary  Withholding blame may not amount to disrespecting oneself and the wrongdoer, so there is room for the legitimacy of non-blame responses to wrongs (say forgiving or doing something else) for reasons from relationships. In the next section, I am going to take a step further. I will argue for a broader approach to respect such that, in cases of wrongdoing, respect can be expressed through such reasons, i.e. through improving the damaged relationship, show moral and non-moral support, and so on.            	 171	Section 3 A Broader Approach to Respect What underlies the objection that withholding blame amounts to lack of respect for oneself and the wrongdoer may be a very narrow approach to respect. It conceives respect as merely a matter of (i) “minimum requirements”—one must respond in a certain way when wronged in order to meet such requirements—and (ii) blame, which is the required response to wrongs. Narrowly conceiving respect in this way makes us lose sight of alternative, and indeed more valuable ways to express respect. I therefore suggest that a more effective and illuminating way to engage this objection is to go beyond this narrow conception of respect and consider some more positive ways to express respect. I therefore propose in this section a broader approach to respect; in particular, I argue that this broader approach shows that those non-blame responses that I discussed in chapter 4 are positive ways to express respect for oneself and the wrongdoer.   I will first discuss in (3.1) two concrete cases that both point to a broader approach to respect. I will then in (3.2) demonstrate that, under this broader approach, withholding blame for the reasons from relationships that I discussed in chapter 4 may show respect for one’s (victim’s) dignity and the wrongdoer’s agency.  (3.1) Two Cases (3.1.1) Prince Myshkin I will start the discussion with two cases, both of which aim to show alternative ways to express respect via reasons from relationships. The first case comes from Fyodor Dostoevsky’s masterpiece The Idiot. The protagonist Prince Myshkin is a kind, candid, and intelligent young man, but is often perceived as an “idiot” because of his simplicity and past illness (epilepsy). In one scene, quoted below, Prince Myshkin is to be hosted by Ganya Ardalionovich and they are on their way to Ganya’s house. Ganya is a rather unpleasant fellow. He is jealous of Myshkin’s ability to build an instant rapport with the three Epanchin sisters, one of whom Ganya is in love with. Ganya’s jealousy soon turns into rage and verbal abuse—he verbally abuses Myshkin, Myshkin’s friend, and his doctor in Switzerland. But in spite of Ganya’s repeated abuses, Myshkin does not get angry; rather, he calmly tells Ganya his feelings about what Ganya said:   	 172	“I think I ought to tell you, Gavrila Ardalionovitch," said the prince, suddenly, "that though I once was so ill that I really was little better than an idiot, yet now I am almost recovered, and that, therefore, it is not altogether pleasant to be called an idiot to my face. Of course your anger is excusable, considering the treatment you have just experienced; but I must remind you that you have twice abused me rather rudely. I do not like this sort of thing, and especially so at the first time of meeting a man, and, therefore, as we happen to be at this moment standing at a crossroad, don't you think we had better part, you to the left, homewards, and I to the right, here? I have twenty-five roubles, and I shall easily find a lodging. (Dostoevsky 2001)  One might argue that Myshkin’s telling Ganya that he is displeased by Ganya’s abuses and his proposal to part company are a form of blame, but I do not think it is a right way to understand the scenario. Calmly telling someone that one is displeased is not the same as blaming. On the other hand, one might argue that Myshkin does not take offence to begin with because he says Ganya’s behaviour is “excusable”; I, again, do not agree with this reading. Myshkin is tactfully telling Ganya that his behaviour is inappropriate while avoids calling him out “to his face”. (I suppose that many of us have said “it isn’t your fault” when it clearly is and want them to know that it is.)  But I suppose that some might still disagree with my reading. So let’s suppose that Myshkin did not raise any objections against Ganya’s abuses but merely came up with some excuses to decline Ganya’s offer and calmly leave to find a place to stay. Would Myshkin’s decision not to blame Ganya show any lack of respect?   There does not seem to be any lack of respect. For one thing, Myshkin’s proposal shows his confidence. It is the first day he returns to Moscow from Switzerland and he knows no one, but he is confident that he does not need to rely on Ganya. More importantly, his proposal is a way to avoid further conflicts with Ganya. He knows that Ganya, rightly or wrongly, is angry with and jealous of him, so staying at Ganya’s place under such circumstances will likely lead to further conflicts which will not do any good to either one of them. The best solution, therefore, is not to place blame on him but to simply part company. This attempt to avoid further conflicts shows respect for his own dign