Open Collections

UBC Theses and Dissertations

UBC Theses Logo

UBC Theses and Dissertations

Affixed interjections in English and Polish : a corpus-based study of emotional talk in digital communication… Lockyer, Dorothy (Dorota) 2018

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


24-ubc_2018_september_lockyer_dorothy.pdf [ 1.94MB ]
JSON: 24-1.0368976.json
JSON-LD: 24-1.0368976-ld.json
RDF/XML (Pretty): 24-1.0368976-rdf.xml
RDF/JSON: 24-1.0368976-rdf.json
Turtle: 24-1.0368976-turtle.txt
N-Triples: 24-1.0368976-rdf-ntriples.txt
Original Record: 24-1.0368976-source.json
Full Text

Full Text

AFFIXED INTERJECTIONS IN ENGLISH AND POLISH: A CORPUS-BASED STUDY OF EMOTIONAL TALK IN DIGITAL COMMUNICATION AND LITERARY DIALOGUE by  Dorothy (Dorota) Lockyer  B.A., The University of Victoria, 2010 M.A., The University of Victoria, 2013  A THESIS SUBMITTED IN PARTIAL FULFILLMENT OF THE REQUIREMENTS FOR THE DEGREE OF  DOCTOR OF PHILOSOPHY in THE FACULTY OF GRADUATE AND POSTDOCTORAL STUDIES (English)  THE UNIVERSITY OF BRITISH COLUMBIA (Vancouver)  July 2018  © Dorothy (Dorota) Lockyer, 2018    ii  The following individuals certify that they have read, and recommend to the Faculty of Graduate and Postdoctoral Studies for acceptance, the dissertation entitled: Affixed interjections in English and Polish: a corpus-based study of emotional talk in digital communication and literary dialogue  submitted by Dorothy Lockyer in partial fulfillment of the requirements for the degree of Doctor of Philosophy in English  Examining Committee: Jessica de Villiers, English Supervisor  Laurel Brinton, English Supervisory Committee Member  Katarzyna Dziwirek, University of Washington Supervisory Committee Member Barbara Dancygier, English University Examiner Caroline Rieger, Central, Eastern and Northern European Studies University Examiner  Additional Supervisory Committee Members:  Supervisory Committee Member  Supervisory Committee Member iii  Abstract  Extensive cross-linguistic research documents a wide range of functions and semantic-pragmatic meanings of interjections in English and Polish that typically correspond with a primary function of conveying emotion. With many forms that have changed over time and appear in a variety of written and spoken mediums, interjections have been mainly considered ‘morphologically simple’; that is, they typically do not take on affixes. However, recent research has shown that interjections do, indeed, acquire various slang, diminutive and augmentative suffixes to change the register, to intensify or diminish the base interjections’ meaning, and/or to convey jocularity and non-serious meanings associated with play. This dissertation addresses some major gaps in the descriptive and empirical research of the semantic and pragmatic functions and meanings of affixed interjections in Polish, a synthetic Slavic language, and English, an analytic Germanic language. These two languages are compared and analyzed by examining the core and peripheral meanings of affixed interjections, their typology, and their pragmatic potential as attitudinals. The morphology and pragmatics of these affixed interjections are examined qualitatively and quantitatively by examining definitions from online dictionaries and standard corpora. It is argued that three fundamental semantic constraints underlie the formation of affixed interjections: [+INFORMAL], [+EMOTION], and [+ATTITUDE]. These three features can be subdivided into secondary semantic and pragmatic features that may or may not always apply, including [+PLAYFUL], [+CUTE], [+SILLY], [+GOOD HUMOUR], [+INTIMATE] and [+WARM]. Given the volume and variety of forms considered, affixed interjections would be methodologically challenging to gather in naturally occurring spontaneous speech; therefore, the study combines data from corpora of online sources that provide novel, current and slang words, including the micro-blogging site Twitter, Google Books, fanfiction, blogs, and blog comments. The dissertation also examines the Appraisal resources (Martin and White 2005) used for diminutive interjections. It is argued that English uses diminutive interjections mainly for positive APPRECIATION (positive meanings), negative JUDGEMENT (sarcasm), and negative AFFECT (negative meanings). These interjections are relatively rare compared to Polish. In comparison, Polish diminutives are much more frequent and conventional, and are used to facilitate social bonding, and show warmth and affection.  iv  Lay Summary  This dissertation investigates the linguistic formations, functions and meanings associated with English and Polish interjections that receive endings which convey emotion and/or attitude, e.g. English wowie!, whoopsie! and Polish jejku! These diminutive interjections are examined in fiction, Twitter and other modes of written communication, along with dictionary entries. After examining the range of core contexts in which diminutive interjections occur (e.g. pain, sarcasm, admiration), I posit that these interjections are linked with the overarching concepts of emotion, attitude, and informality, but that their specific meanings and functions depend on context. In some cases, they intensify positive attitude, while in others they may intensify sarcastic and negative attitudes. Overall, diminutive interjections tend to appear more frequently in Polish than in English. v  Preface  This dissertation is an original, unpublished, and independent work by the author, Dorothy (Dorota) Lockyer. vi  Table of Contents  Abstract ................................................................................................................................... iii Lay Summary ............................................................................................................................ iv Preface ....................................................................................................................................... v Table of Contents ..................................................................................................................... vi List of Tables ............................................................................................................................. x List of Figures ........................................................................................................................... xi List of Abbreviations .............................................................................................................. xii Acknowledgements ............................................................................................................... xiii Dedication ............................................................................................................................... xv  Purpose and significance of the study .................................................................... 1 1.1 Definitions and types of interjections ............................................................................... 5 1.1.1 (Primary) emotive interjections ............................................................................... 17 1.1.2 Cognitive interjections ............................................................................................. 20 1.1.3 Conative interjections .............................................................................................. 20 1.1.4 Secondary emotive interjections .............................................................................. 21 1.1.5 Contact interjections ............................................................................................... 24 1.1.6 Descriptive and onomatopoeic interjections .......................................................... 26 1.2 Morphology of interjections ............................................................................................ 27 1.3 Interjections in the Slavic languages ............................................................................... 28 1.3.1 Polish interjections .................................................................................................. 29 1.3.2 Czech and Slovak interjections ................................................................................ 31 1.3.3 Russian interjections ................................................................................................ 32 1.4 Chapter conclusions ........................................................................................................ 34 1.5 Organization of the thesis ............................................................................................... 35  Background to evaluative and slang suffixes ...................................................... 37 2.1 Introduction ..................................................................................................................... 37 2.2 Defining diminutives and augmentatives ....................................................................... 38 2.3 Diminutive affixes in English .......................................................................................... 42 2.3.1 -Y (-IE) ...................................................................................................................... 44 vii  2.4 Slang suffixes in English .................................................................................................. 48 2.4.1 Diminutive-slang suffixes -O, -ER(S), -S ................................................................. 50 2.5 Diminutive affixes in Polish ............................................................................................ 53 2.6 Czech, Slovak, and Russian diminutives ......................................................................... 56 2.7 Pilot studies ...................................................................................................................... 58 2.8 Chapter conclusions ........................................................................................................ 60  Research methods and texts ................................................................................. 62 3.1 Purpose of the present case study ................................................................................... 62 3.2 Research questions .......................................................................................................... 63 3.3 Theoretical background ................................................................................................... 65 3.3.1 Semantic and pragmatic features ............................................................................ 66 3.3.2 Appraisal ................................................................................................................... 67 3.4 Datasets ............................................................................................................................ 74 3.4.1 Twitter ...................................................................................................................... 75 3.4.2 Blogs ......................................................................................................................... 78 3.4.3 Fanfiction ................................................................................................................. 79 3.4.4 Literary dialogue (Google Books) ............................................................................ 79 3.4.5 Reference corpora .................................................................................................... 82 3.4.6 Online dictionaries .................................................................................................. 84 3.5 Methods ............................................................................................................................ 85 3.5.1 Methods for chapter 4 ............................................................................................. 86 3.5.2 Methods for chapters 5 – 7 ...................................................................................... 89 3.6 Methodological issues...................................................................................................... 93 3.7 Chapter conclusions ........................................................................................................ 95  Corpora and definitions ....................................................................................... 97 4.1 Introduction ..................................................................................................................... 97 4.2 Primary emotive interjections ......................................................................................... 99 4.2.1 English ..................................................................................................................... 100 4.2.2 Polish ....................................................................................................................... 108 4.3 Contact interjections ....................................................................................................... 110 4.3.1 Greetings and farewells ............................................................................................ 110 viii  4.3.2 Requests ................................................................................................................... 117 4.3.3 Apologies .................................................................................................................. 118 4.3.4 Gratitude ................................................................................................................. 120 4.4 Secondary interjections .................................................................................................. 124 4.5 Chapter conclusions ....................................................................................................... 133  Admiration, astonishment, and delight ............................................................. 136 5.1 Introduction .................................................................................................................... 136 5.2 Positive surprise/shock ................................................................................................... 136 5.2.1 English ..................................................................................................................... 136 5.2.2 Polish ....................................................................................................................... 143 5.3 Compliments and expressions of appreciation .............................................................. 144 5.3.1 English ..................................................................................................................... 144 5.3.2 Polish ....................................................................................................................... 157 5.4 Chapter conclusions ....................................................................................................... 170  Sarcasm and play .................................................................................................. 173 6.1 Introduction .................................................................................................................... 173 6.2 (Non-serious) accidents and mistakes: Friendships and children ................................ 174 6.2.1 English ..................................................................................................................... 174 6.3 Playful accidents and mistakes ....................................................................................... 176 6.3.1 English ..................................................................................................................... 176 6.4 (Non-serious) accidents and mistakes: Politics ............................................................. 183 6.4.1 English ..................................................................................................................... 183 6.5 Sarcasm and irony ........................................................................................................... 185 6.5.1 English ..................................................................................................................... 185 6.5.2 Polish ....................................................................................................................... 195 6.5.3 (Veiled) threats and intimidation .......................................................................... 197 6.6 Chapter conclusions ....................................................................................................... 201  Pain, sympathy, and disgust ............................................................................... 204 7.1 Introduction ................................................................................................................... 204 7.2 Disgust ............................................................................................................................ 204 7.2.1 English .................................................................................................................... 205 ix  7.2.2 Disgust in Polish .................................................................................................... 208 7.3 Negative surprise (fear and worry) ................................................................................. 210 7.3.1 English ...................................................................................................................... 211 7.4 Pain .................................................................................................................................. 214 7.4.1 English ..................................................................................................................... 214 7.4.2 Polish ....................................................................................................................... 221 7.5 Sympathy and empathy .................................................................................................. 223 7.5.1 English ..................................................................................................................... 223 7.5.2 Polish ...................................................................................................................... 226 7.6 Anger, bitterness, and dislike ........................................................................................ 228 7.6.1 English .................................................................................................................... 228 7.6.2 Polish ...................................................................................................................... 230 7.7 Chapter conclusions ....................................................................................................... 232  Conclusion........................................................................................................... 234 8.1 Summary and contributions .......................................................................................... 234 8.1.1 Formation of affixed interjections ......................................................................... 234 8.1.2 Semantic and pragmatic features ........................................................................... 237 8.1.3 Appraisal ................................................................................................................. 239 8.2 Limitations ..................................................................................................................... 242 8.3 Future directions ............................................................................................................ 243 Bibliography .......................................................................................................................... 246  x  List of Tables  Table 1: English primary emotive interjections ...........................................................................107 Table 2: Polish primary emotive interjections ........................................................................... 109 Table 3: Polish greetings and farewells ........................................................................................ 114 Table 4: Polish 'plis' ...................................................................................................................... 118 Table 5: English 'sorry' ................................................................................................................. 118 Table 6: Polish 'sor(r)y' ................................................................................................................ 119 Table 7: English 'thanks' and 'thank you' .................................................................................... 121 Table 8: Polish ‘dziękuję’ .............................................................................................................. 123 Table 9: English secondary emotive interjections ...................................................................... 130 Table 10: Polish secondary emotive interjections ....................................................................... 132  xi  List of Figures  Figure 1: An overview of appraisal resources (© 2005 Peter R.R. White and James R. Martin, by permission) ................................................................................................................................... 68  xii  List of Abbreviations  AmE = American English AUG = augmentative affix AusE = Australian English BNC = British National Corpus BrE = British English CanE = Canadian English CDS = child-directed speech CMC = computer-mediated communication COCA = Corpus of Contemporary American English CS = child-speech DE = diminutive expletive DI = diminutive interjection DIM = diminutive affix DM = discourse marker GloWbE = Global Web-Based English IrE = Irish English NKJP = National Corpus of Polish NSM = Natural Semantic Metalanguage NZE = New Zealand English OED = Oxford English Dictionary PDE = present-day English PDP = present-day Polish Pol. = Polish POS = part of speech SJP = Słownik języka polskiego [Dictonary of the Polish language] SOAP = Corpus of American Soap Operas Strathy = Strathy Corpus of Canadian English VOC = Vocative case  xiii  Acknowledgements  There are many people I need to thank, and I doubt I will be able to name everyone who has made a positive impact towards the completion of this dissertation. There are two people that I owe the greatest gratitude to: my supervisor, Dr. Jessica de Villiers, and my previous supervisor and current committee member, Dr. Laurel Brinton. Second, I owe huge thanks to my committee member, Dr. Katarzyna Dziwirek for her invaluable advice about the Polish language, and for her suggestions throughout the process of writing this dissertation. Without them I would not have finished this dissertation. I also am grateful to Dr. Stefan Dollinger, a pro-tem committee member, who provided me with invaluable support and grounded me in Canadian English, lexicography, and corpus linguistics. The dissertation is a product of years of work, including coursework and teaching. I took several courses that challenged my knowledge and perspective during these past five years. I would like to thank the amazing faculty at UBC, including Dr. Barbara Dancygier, whose class on cognitive poetics unpacked different possible approaches to translation and emotion. Also to my translation professor in Creative Writing, Rhea Tregebov, who re-inspired my passion for translation. In addition, I had the privilege of working with Dr. Laurie McNeill as a TA at UBC, where I learned a lot about teaching memoirs to first-year students, along with the support from my fellow TAs: Loren Gaudet, Dallas Hunt, and Michael de Santis. When I first arrived at UBC, I had the pleasure of meeting my supportive PhD cohort – Loren Gaudet, Rhonda Shanks, Szu Shen, Jade Standing, and Lora Moon – to begin this amazing adventure together. Then, while working on the Dictionary of Canadianisms on Historical Principles (DCHP-2), I worked with inspirational colleagues at the Canadian English Lab, namely Emily Briggs, Sasha Gaylie, Gabrielle Lim, and Baillie Ford. We had many good times at our ‘jour fixes’ and writing up word entries! Also, a big thank you to fellow Language graduate students, Natasha Rose, Kristan Newell, and Tomoharu Hirota.  For two semesters, I had the amazing opportunity to teach as a sessional instructor at the University of Victoria. The amazing faculty there, namely Dr. Megan Swift, Dr. Julia Rochtchina, Dr. Helga Thorson, Dr. Serhy Yekelchik and Dr. Olga Pressitch – just to name a few – provided me with guidance and a warm environment to work in. I am grateful to Dr. Ewa Czaykowska-Higgins, who encouraged my interest in diminutives in the first place. xiv  Of course, there are many people in the background who I would like to thank for their support. First, to our amazing Graduate Admin, Louise Soga, for making everything run smoothly. Friends outside of academia provided wonderful moral support. Becky Rogers and Irene Peinhopf - thank you for your encouragement and support. To Mihye Park – the Housing Supervisor at St. Andrew’s Hall – and Sumarme Goble, your kind words often lifted me up when I was feeling discouraged. Also, to all my family friends who have uplifted me and kept me in thoughts and prayers – Eiddwen, Mavis, Val, Sophie, and many others – thank you. You constantly inspired and cheered me up. And to my family and friends in Europe who often helped me with Polish and answered my repetitive questions about ‘which of these two words is “smaller” or more endearing’, namely Ewa Sikora, Jana Grygová, Barbara Byrtusová, Agata Byrtus, Gosia Bocková and Tereza Mlčochová, serdecznie dziękuję! My final and dearest thanks go to my mom, who was there for me from all the ups to the downs, and back up again. While my dad could not be with me, I am sure he was looking down on me during these times, and I know that I have made him proud. And last, but most importantly, my enduring thanks go to God, without whom nothing would be possible.  . xv  Dedication  For my father, Francis Ernest Lockyer (1922-2013). The night before you went Home, I told you that I would be fine, and that I would complete a PhD.  I promised, and here it is. Love you always.1   Purpose and significance of the study The following English interjections with a -y or -ie ending have meanings of ‘silly mistake’ in (1), ‘non-serious joy’, ‘play’ and/or ‘teasing’ in (2) and ‘emphatic shock’ in (3).  (1) I try to use first initials if they do not already have a web presence or if they are parents of young children or they are, actually, young children. I sometimes forget about this. Whoopsy. [GloWbE: US B] (2) # Lizzie: Oh goody! It's groom hunting season. # Jonathan: Whoa. Is that thing loaded? # Alan: You don't want to find out, now do ya? [SOAP: 2006] (3) criminal offense in Japan for foreign lawyers to violate certain discovery rules. # "Wowie!" says Turbin, recounting his shock at finding out about the law. [COCA: ACAD: 2003] In the Polish example in (4), the writer establishes an emphatic emotional and social connection by offering extensive congratulations to another person on an online forum with jejku! (< jej! ‘oh dear’).  (4) oo gratulacje !!!!!!!!!!  jejku wszystkiego naj naj naj lepszego na nowej drodze zycia!  ‘oo congratulations !!!!!!!!!!  jejku everything the very very very best for your new road in life!’ [NKJP: 2005] This dissertation centers on the little words shown above in bold – that is, diminutively-affixed interjections – that are ways of expressing emotions in many languages. However, few studies have examined their use in internet language and fictional dialogue, particularly comparing the analytic Germanic language of English and the synthetic Slavic language of Polish. In both languages, these words are interjections with evaluative suffixes. ‘Interjections’ are broadly defined here as “little words or non-words which in terms of their distribution can constitute an utterance by themselves and do not normally enter into construction with other word classes” (Ameka 1992: 124). A suffix, on the other hand, is defined as “[a] bound morpheme that is added to the end of a free morpheme to form a new composite word, of either a similar or different word class” (Hamawand 2011: 121). One such type of suffix is the ‘evaluative suffix’, 2  an umbrella term typically used in evaluative morphology (e.g. Grandi and Körtvélyessy 2015), a subfield of derivational morphological studies that examines evaluative constructions. A construction is considered evaluative if it conveys a meaning of size (‘small’ or ‘big’) or emotion (‘good’ or ‘bad’) through a prefix, suffix, reduplication, etc. I use the term ‘evaluative suffix’ throughout this dissertation to refer to various suffixes that generally add a certain emotional and/or attitudinal colouring to the base interjection. The suffix, thus, need not create a new interjection completely different in meaning from the base, but creates a new form of the interjection that often has a new function or additional (emotive) meaning. This dissertation examines the forms, emotive meanings, functions and uses of suffixed emotive interjections comparatively in English and Polish. On a secondary level, the dissertation applies cultural values and social contexts to its examination of suffixed interjections.  Both English and Polish have evaluative affixes that add additional emotional colouring to nouns and sometimes other parts of speech. They are ‘diminutive’ (sometimes termed ‘hypocoristic’) affixes which typically convey the meaning of [+LITTLE] and/or [+AFFECTION] (see, for example, Schneider 2003, Wierzbicka 2003, and Jurafsky 1996 for in-depth discussions of diminutives), such as e.g., oh goodie! (< oh good!), whoopsy! (< whoops!) and wowee! (< wow!). According to Bakema and Geeraerts (2004: 1045), “the term ‘diminutive’ refers to any formation in a language expressing the referential meaning 'small', and possibly a variety of derived evaluative shades of meaning.” It may be formed analytically or synthetically. The term ‘diminutive’ is opposite to ‘augmentative’, which Bakema and Geeraerts (2004: 1045) define as “the semantic counterpart of the diminutive [which] denotationally expresses the concept ‘big’, and may have derived readings such as evaluative exaggeration and intensification.” To achieve the emotional colouring conveyed through the diminutive suffixes, affixed interjections are frequently formed with the most widely-used and productive -y diminutive suffix in English and -k- diminutive affix in Polish. Alternatively, affixed interjections may be formed by augmentative suffixes such as English -rama and -ola (cf. Bauer, Lieber and Plag 2013: 411) and Polish -isko/-icho that generally convey the meaning of [+LARGE], e.g. crapola! (< crap!) and shitorama! (< shit!), and Pol. narazicho! (< na razie! ‘bye’). So-called ‘slang’ suffixes such as -aroo also appear and are sometimes added to interjections, e.g., whoopsaroo! (< whoops!) for a playful effect. Diminutive and augmentative prefixes also exist in English and Polish, e.g., mini-, hyper- and ultra-, but I have found no evidence of interjections receiving prefixes. Thus, this 3  study discusses only those affixes that are used with interjections, especially English -y, -ee, -ie and Polish -k- to convey additional emotional meaning to the base interjection in its dialogic context. The common view assumes no difference between affixed and unaffixed forms (cf. Nübling 2004: 16, 29, Stange 2016: 61);1 however, it is reasonable to assume that speakers use these forms instead of the base interjection for a purpose, arguably to convey additional attitude or emotion. In English and Polish, we can easily see the extremely frequent use of the affixed forms in baby-talk or child-directed speech (CDS); in fact, we would assume that it would be only natural for mothers to use loving endearments to their young children. Support for this view in child-directed speech and adult speech is provided by Lappe (2007: 13), who argues that “y-hypocoristics do not have the same meaning as their unsuffixed counterparts.” Furthermore, Alber and Arndt-Lappe (2012: 323) “see a meaning difference between the base and the derived form.” Although Alber and Arndt-Lappe are not specifically referring to interjections, we can infer that the rule applies to interjections as well. It can be assumed that there is a difference in meaning between, e.g., wow! and wowie!, and oops! and oopsie! The central questions that will be addressed in this dissertation are as follows: ● With which suffixes and base interjections are affixed interjections typically formed and what are the main causes for these word-formation choices? ● Which fundamental semantic and pragmatic features underlie the formation of affixed interjections? In other words, what emotional connotations do affixed interjections convey? ● What can digital communication and literary dialogue, viewed through the Appraisal framework, reveal about the main evaluations of affixed interjections? In which contexts are they found in most often, and how do they differ between English and Polish? The main aims and goals of each chapter relate back to these three essential questions about form, meaning, and function. I will argue that affixed forms of interjections, which have only been sparsely touched upon in previous research, have more important functions than has been previously thought, especially in expressing playful, cutesy, and jocular types of attitude.                                                  1 Stange (2016: 61) suggests that affixed interjections are spelling variants [which] “merely reflect” pronunciation variation, not variation in meaning. She speculates that ouchie “is probably only a hypocoristic form, like horsie or doggie”. 4  The affixed forms discussed in this dissertation were collected by mining2 written language: internet language, literary dialogue from adult fiction and, least frequently, established corpora. These text types were chosen because they contain the highest number of affixed interjections in written language with sufficient context for linguistic analysis of the items. The scope of this data mining was further restricted to searches of the following text types; recipe blog comments, fanfiction comments, Twitter posts, and literary dialogue from adult novels using a Google Books search. In general, recipe blogs and their comments tend to be positive and convey expressions of strong delight, admiration, and praise, while Twitter is by far the largest source of affixed interjections. Fiction, especially from the romance (e.g. Harlequin) genre, provides additional contexts of male/female dichotomies not found in the short online comments. Established corpora, due to insufficient examples, were primarily used for the frequency charts in Chapter 4. These sources for data mining were important to this study because they provide insight into how the affixed interjections function across several text types. The major goal of this dissertation is to refute the idea that interjections in the English and Polish languages are ‘morphologically simple’ and do not receive suffixes. An equally important second goal is to show the uses of these forms across different contexts and written language for various functions – though I do not claim to identify every possible use for each form. Although previous research has rarely addressed affixed interjections, we have numerous examples of different interjections with different endings.  Little scholarly research has attempted to describe the affixed forms of interjections; in fact, from the previous scholarly treatment of these forms, it could be argued that scholars have typically viewed affixed forms of interjections as non-existent or morphologically impossible as lexical formations because interjections are not expected to take endings. In looking at the morphology and semantics-pragmatics of affixed interjections across two rather unrelated languages, this dissertation contributes to an understanding of the use and functions of these emotionally loaded, ‘little’ words with endings. A central claim is made that affixed interjections do, in fact, exist to convey various emotional connotations – also referred to as attitude, evaluation, affect, and sentiment - and are therefore a tool in written language to convey subtle nuances of meaning.                                                   2 The procedures and methods for data mining are described in Chapter 3. 5  A foundational aspect of this dissertation is to investigate interjections in English and Polish. Thus, the rest of Chapter 1 is divided into three main sections. The first section (1.2) defines the interjection; specifically, it identifies the semantic and pragmatic functions and meanings of the interjection, the various types of words that are considered interjections, and why the interjection needs further study. Section (1.3) discusses the morphology of interjections to establish why they may or may not accept suffixes and why we find many interjections in different parts of speech. The final section of this chapter (1.4) leaves the English-centred focus of the previous sections and instead examines interjections from the Slavic perspective, namely Polish, Czech, Slovak, and Russian. The objective of 1.4 is to establish other points of view towards interjections that may be unfamiliar to English scholars, and to present a counterpoint to English interjections, illustrating the importance of having a wider-range of perspectives when approaching the formations, functions, and meanings of interjections as a universal part of speech (POS). The chapter ends with a description of the structure of the dissertation (1.5). 1.1 Definitions and types of interjections There is a substantive body of literature about the interjection (see, e.g. Ameka 1992) – a term that comes from Latin interjectio and signifies ‘a throwing in between’ because they are “thrown in between parts of discourse” (Fowler 1850: 330, original emphasis), almost as if interjections were ‘non-words’ in discourse.  The part of speech (POS) interjection comprises many items that were treated by Latin and Greek grammarians in antiquity as para-linguistic or nonlinguistic phenomena having minor importance in language (cf. Wharton 2003: 175, Ehlich 1986). This contributed to Müller’s (1836: 366) early claim that “language begins where interjections end.” By the nineteenth century their status had changed very little (Wharton 2003: 175). This notion persisted into the twentieth century. For instance, Jakobson (1960: 354) views interjections as existing purely in the “emotive stratum of language” and Sapir (1970: 7) writes that interjections are “never more, at best, than a decorative edging to the ample, complex fabric [of language]” because of the view that interjections “are linguistically somewhat primitive expressions of feeling” (Leech et al. 1982: 6  53).3 It was not until Ameka’s (1992) special issue of the Journal of Pragmatics that interjections were properly approached as important features of language and as serious subjects of linguistic analysis (e.g. Ameka 1992, Wierzbicka 1992, Wilkins 1992). Fowler (1850) was one of the earliest scholars to correctly note that interjections have not received proper scientific treatment in language. He argues that should language have more than an intellectual basis, then interjections should be viewed as an integral part of language. He forms his conclusion based on three main (philosophical/spiritual) factors that would now be seen as functional factors. First, interjections “express the multiplied emotions of the human mind, and lend their aid where all other language fails in this respect” (382). Secondly, interjections are the only way that humans may communicate verbally with animals, or animals communicate with other animals. Thirdly, “[interjections] are a natural universal language” (382). As aptly mentioned by Stange (2016: 7), “if the focus is on function as opposed to form, it is defensible to view interjections as part of language.” This is the view adopted in this study. Interjections have been affected by language change and linguistic trends, which have caused many interjections to disappear and new ones to arise. Despite how the forms of interjections have changed, their emotional content and short length, among their other characteristics, have remained the same. Because of their typical length and emotive features, interjections are generally considered ‘little’ linguistic expressions of emotion (also referred to as passions or feelings) produced “because there has been an object, event, action, or state of affairs that has caused […] a certain feeling or emotion at a certain moment” (Padilla Cruz 2010: 55).4 These include wow!, whoa!, yay!, oops! and gee!  Norrick (2009: 888) writes that “the class of interjections must always remain open to fresh entries and new combinations” – that is, interjections are a class of words highly susceptible to                                                  3 Polish researchers of interjections had similar views. Milewski (1965: 117) writes that “interjections are elements of old, pre-language codes, which took on language forms” (my trans.), and Kopczyński’s (1817: 38) view reflects the purely emotive view of interjections in his definition of interjections as “the human language of the heart: for they express affect or passions” (my trans.). 4 Words that convey emotion include the following by Clark (1852: 185): oh!, ah!, alas!, alack!, welldone!, ha!, whew!, eh!, what!, poh!, pshaw!, pish!, tut!, humph!, gammon!, fie!, ho!, ahoy, hurrah!, huzza!, whoh!, hush!, hist!, oho!, farewell!, fee-fo-fum!, gee-up!, tallyho!, no!, among others. Many of these went out of use and are now considered archaic, namely fie! (often found in Shakespeare), poh!, gammon!, and hist! To the modern English ear, most of Clark’s interjections sound old-fashioned. 7  replacement and innovation. They capture our attention because of their novelty and ability to add a ‘splash of color’ to language.5 They are ‘recycled’ to create new forms that can convey the emotional impact that the previous interjections lost over time. Thus, LOL! (< laugh out loud), a 21st century internet invention, helps get rid of the mundane with novelty. Although many interjections have been replaced with new ones, the core functions and characteristics of interjections have changed very little, if at all. Several different definitions of interjections have been proposed, which focus on a number of characteristics. Although there is no consensus on how to define interjections (Wharton 2009: 176), scholars from various backgrounds, disciplines and approaches usually agree upon several characteristics of interjections that specify where an interjection typically appears in a sentence (syntax) and what it does in an utterance (i.e. its function). Below I present five standard definitions of the interjection, beginning with an early quote from 1832 and ending with a recent definition of the interjection proper by Stange (2016). ● Interjections are words thrown in between the parts of a sentence, to express the passions, or emotions of the speaker; as, ah! I have alienated my friend: alas! I fear for life: O virtue! How amiable thou art! (Murray and Benedict 1832: 83) ● [An interjection is] a conventional lexical form which (commonly and) conventionally constitutes an utterance on its own, (typically) does not enter into construction with other word classes, is (usually) monomorphemic, and (generally) does not host inflectional or derivational morphemes (Wilkins 1992: 124) ● An interjection can be defined as a linguistic sign expressing the speaker's current mental state (1) which can be used on its own, (2) which expresses a specifiable meaning, (3) which does not include other signs (with a specifiable meaning), (4) which is not homophonous with another lexical item that would be perceived as semantically related to it, and (5) which refers to the speaker's current mental state or mental act (for example `I feel...', 'I want...', 'I think...', 'I know...'). (Wierzbicka 1992: 164)                                                  5 Some recent interjections that attest to the creation of new words include as if!, an interjection that was popularized by the 1995 movie Clueless, bazinga! (from The Big Bang Theory), d’oh! (from The Simpsons). In Polish, interjections such as, e.g. hejo! ‘heyo’ and the suffixed siemka! (a greeting equivalent to ‘hi, how are you’). 8  ● [An interjection is] a term used in the traditional classification of parts of speech, referring to a class of words which are unproductive, do not enter into syntactic relationships with other classes, and whose function is purely emotive. (Crystal 2003: 239) ● [An interjection is] a syntactically independent, meaningful, semi-automatic exclamation providing an insight into the speaker’s current emotional state of mind (Stange 2016: 20) From these definitions, we may extract several prominent characteristics of interjections. Specifically, the standard definitions presented suggest that interjections express the mental/emotional state of the speaker, are meaningful, exclamatory, and syntactically independent. Wilkins’ (1992) definition contains many hedges which show the multi-functionality of interjections; that is, it is not always possible to provide a firm definition of all interjections because they are not all the same. Below I address each of the features in turn except for the morphological issues, which are treated in section 1.3. There is an underlying assumption that interjections convey passions or feelings (cf. Schourup 1985); we generally tend to think of emotive wow! and whoops! when giving examples of interjections. Although interjections are fundamentally linked with emotion, there are many words in the POS of ‘interjection’ that do not function to convey emotion or a mental state. Interjections including psst! or shh! that are used to get attention or indicate that the hearer should be quiet are communication devices. Likewise, interjections including boom! and whack! describe sounds that may be made by inanimate objects and are often depicted in comics, such as a car explosion or the sound of a shotgun. Therefore, Crystal’s (2003) assertion that interjections have only emotive functions, along with Murray and Benedict’s (1832) focus on emotions, do not apply to interjections as a whole. They apply to emotive interjections – namely those interjections which primarily function to convey emotion  (cf. 1.2.1). It is generally agreed that interjections are ‘meaningful’; that is, “they always have meaning, whether this meaning is identifiable through linguistic analysis or not” (Stange 2016: 20). However, the meaning that an interjection may have differs between scholars. Wierzbicka (1992) takes a slightly different approach from many linguists because she argues that interjections have ‘specifiable meaning’ through semantic primitives in her Natural Semantic Metalanguage (NSM) theory. She claims that “the subtlest shades of meaning encoded in interjections [can be 9  captured] by relying exclusively on universal or near-universal concepts such as ‘good’ and ‘bad’, ‘do’ and ‘happen’, ‘want’, ‘know’, ‘say’ or ‘think’” (Wierzbicka 1992: 163). Other scholars have used semantic primitives to show the meaning of an interjection; for example, Ameka (1992: 109) presents the following explication for wow! using Wierzbicka’s theory: Wow! I now know something has happened I wouldn’t have thought it would have happened I think this is very good I feel something good because of that I say this: [wow] because I want people to know this The explication suggests that wow! conveys three core meanings and functions: 1) to convey surprise and suddenness (‘I now know’); 2) positive feeling (‘I think this is very good’); and, 3) a desire to let others know how the speaker is feeling.  Norrick (2011: 248) aptly comments that “such semantic analysis of interjections misses the pragmatic point: who uses wow, in what context, to what effect?” Ameka’s (1992) explication accounts for a frequent use of wow!, such as in a situation where a speaker feels surprised and good from receiving a surprise gift. The explication does not apply in sarcastic, teasing, or negative contexts, such as when a speaker utters wow! upon hearing surprisingly bad news. In sarcastic usage, wow! can indicate that the speaker is not surprised at all. Thus, while it can be agreed upon that interjections are ‘meaningful’, I adopt Stange’s (2016) view that interjections do not have specifiable ‘basic meanings’, but their meaning is always contextually determined (Schourup 2016 [1985]: 18). There is always an audience for interjections because they may be used in self-talk where the speaker is one’s own audience (cf. Goffman’s ‘response cries’) or in everyday thinking (‘inner talk’). Goddard (2014: 73) notes frequent uses of interjections in his ‘inner talk’. This is unsurprising, as it is generally agreed that interjections show a speaker’s mental state – or, put in another way, the non-vocalized language only occurring in the speaker’s mind. When these 10  interjections are vocalized, they arguably come from one’s thought process directly to the speaker’s audience - be it to oneself, one person, or a large group of people.6  Interjections typically function as a type of ‘window’ into the current7 emotional state of mind of the speaker (Stange 2016: 20),8 and are thus ‘evincives’. They convey emotion in the same way a soccer player might pump her fist in the air after scoring a goal or clap her hands in excitement. In other words, an interjection shows the current, sudden, and spontaneous reaction or mental state of the speaker in her immediate situation. Because of this feature, Schourup (2016 [1985]: 18) argues that at the time an interjection is uttered the speaker has either thought or is thinking and/or feeling something, “but does not completely specify its content.” Schourup provides the example of ‘Aha, the century plant bloomed’, where the aha does not simply express “the existence of undisclosed thought but address[es] something about the current contents of the private world” (original emphasis). Likewise, the oh! in Oh! didn’t make the phone call you asked me to “establish[es] the speaker’s accountability [because oh indicates] that a thought expressed in the sentence following oh just entered the speaker’s mind [and] the speaker’s failure to make the call was due to forgetfulness, not malevolent intent” (Schourup 2016 [1985]: 21).  However, interjections may be feigned and thereby become an “interactive display [of emotion]” (Drescher 1997: 241, original emphasis). That is, interjections are not always true expressions of a speaker’s feelings and therefore must be evaluated in contexts where an interjection may be feigned or insincere. As stated in a well-known passage by Jakobson (1987: 66), the emotive function “tends to produce an impression of a certain emotion whether true or feigned.” If a speaker says “Ouch!” she indicates that she has certain feelings; if she does have                                                  6 Goddard (2014) calls for a deeper exploration of words, including interjections, in the everyday thought dialogue of humans; linguists should “open [their] eyes to the role of words in everyday thinking” (2014: 73). While it is difficult to record thoughts, I agree with Goddard that “literature may provide a valuable point of entry into the ‘inner’ uses of language” (2014: 74), despite some positions which claim that literature does not provide ‘real’ or ‘natural’ examples of language use. Obviously, language use in the thought process is best left to neurolinguistics, but the inner dialogue as recorded in literary dialogues cannot be underestimated.  7 Arguably, there is the possibility of a slight lag if the other interlocutor is dominating the conversation and the person wanting to say an interjection must wait a while for their turn to speak. 8 Ameka (1992) supports this characteristic by his semantic interpretation of ‘I now know’ (emphasis mine) in the explication of wow! that was shown above. 11  those feelings, then the interjection is genuine, but if not, then they are feigned (Tsur 2003: 45). This seems to suggest that interjections may be seen as windows into a speaker’s mind, but also as nothing more than displays of emotion not felt by the speaker. That interjections may be used insincerely calls for a more rigorous analysis of interjections beyond the meanings attributed to them in linguistic studies. The view that interjections are ‘exclamatory’ in the sense that a speaker typically puts force into their utterance through greater volume is a persuasive position, given that (emotive) interjections tend to be said more loudly than words from other parts of speech (Stange 2016: 20). The idea appears in Stange’s earlier work (2009: 40), where she comments that a striking feature of interjections is that they correspond to a “seeming loss of control of the voice” that causes them to be “often rather loud in comparison to the rest of the lexical utterances.” She notes that pain interjections including ow! and ouch! can be screamed or even shouted very loudly (2009: 40). It may also be observed that emotive interjections like whoopee! are rarely, if ever, whispered or said in a soft voice. However, a speaker may also say wow! with a regular or decreased volume, especially if the recipient is emotionally unmoved and unimpressed, or may say oh! faintly from disappointment. By hedging with ‘usually’ and ‘often’, Stange (2009) argues that loudness is frequent enough to be a ‘striking’ feature of interjections, though the feature is arguably context-specific and determined by the speaker’s feelings and/or mental state.  In formal terms, interjections are syntactically independent and holophrastic.9 That is, interjections are “self-contained, autonomous utterances that resist integration into syntactic structures” (Stange 2016: 48), and contain the meaning of a whole sentence (Gehweiler 2010: 316) that is separate from the rest of the utterance (cf. Wilkins 1992, Biber et al. 1999, Ameka 1992, Quirk et al. 1985).10 Interjections only integrate into the sentence after conversion to other parts of speech (e.g. she wowed the audience). On their own, interjections remain syntactically outside of the rest of a speaker’s utterance.                                                  9 That is, they express the meaning of an entire utterance in one word. 10 This is more difficult to see in spoken discourse than in written discourse because “[s]peakers do not always use syntactically well-formed messages, but instead they may be found to not finish their sentences, to use only fragments of sentences or clauses, to break off an utterance and start in anew” (Stange 2016: 48) and the like. 12  Based on the viewpoints addressed above, it is not fallacious to reason that emotive interjections, at least in English, are either displays of or windows into a speaker’s mental state (depending on the situation), have contextually-determined meanings, usually affect the volume of the speaker’s voice, and are syntactically independent and holophrastic. These characteristics, however, are not necessarily applied to other types of interjections or in other languages like Polish. The other types of interjections will be discussed in the following sections, and Polish (and Slavic) interjections discussed in 1.4.  Different emphases and approaches have led to the difficulty in deciding which words and particles are interjections, and has subsequently caused some disagreement (Joue and Collier 2006; see also Kowal and O’Connell 2004: 5, Wharton 2009: 72). There is disagreement about the classification of interjections “both in regard to identifying what belongs to [the] class and in terms of finding language to describe or designate [the] class” (Joue and Collier 2006: 144) because not all scholars agree about which items are interjections. This has caused Nübling (2004: 11) to observe that the term interjection has become a “dumping ground for particles which are otherwise difficult to classify,” as we may observe for ‘inserts’ and ‘discourse markers’ below.  The items which belong to the class differ because of unclear and inconsistent terminology. Biber et al. (1999) views interjections as ‘inserts’ and consequently includes oh! and ah! as real interjections but excludes well!, right!, hey!, yo!, um, er, good grief! and shit! (see Jucker 2012: 201). Goffman (1971) calls them ‘response cries’. Jucker (2002) refers to oh! as a discourse marker, while Culpeper and Kytö (2010) refer to it as ‘pragmatic noise.’ Other terms used to refer to some interjections include ‘pragmatic markers’, ‘pragmatic particles’, ‘attention signals’ and ‘response signals’ (cf. Joue and Collier 2006). In my view, ‘discourse markers’ and ‘formulae’/ ‘conversational routines’ are the most difficult to define in regard to interjections.  Discourse markers (DMs) - also termed pragmatic markers - are a class of words identified by function instead of POS that can significantly contribute towards the discursive direction of a conversation for a number of reasons. These are shown in the definitions of DMs in the following: • [Pragmatic markers] implicitly anchor the act of communication to the speaker’s attitudes towards aspects of the on-going interaction (Östman 1981: 5)  13  • [Pragmatic markers are] those expressions used to indicate how the relevance of one discourse segment is dependent on another (Blakemore 1987: 125) • [Discourse markers are] sequentially dependent elements which bracket units of talk (Schiffrin 1987: 31) • [Discourse markers] are short lexical items, used with a pragmatic meaning on a metalingual level of discourse in order to signal for the hearer how the speaker intends the present contribution to be related to preceding and/or following parts of the discourse (Lenk 1998: 52) •  [Discourse markers are] lexical expressions drawn primarily from the syntactic classes of conjunctions, adverbs, and prepositional phrases. With certain exceptions, they signal a relationship between the segment they introduce, S2, and the prior segment, S1 (Fraser 1999: 950) The definitions show that DMs are ‘short’ items, have a metalinguistic and/or pragmatic function in discourse, and communicate an attitude towards ongoing discourse. For many scholars, the discursive functions are the most important characteristics of DMs. While the above definitions may appear straightforward because they tend to focus on the discourse aspects (e.g. relationships between segments), DMs are difficult to categorize because of the lack of unification in terminology and definitions. Their definitions can be confusingly similar to interjections in several ways, including their placement in sentences, syntactic independence, and potential for expressing emotion. In fact, the overlap and confusion between interjections and DMs is daunting for anyone encountering the literature on DMs. Scholars routinely ask and are unable to unite on an answer to ‘what are discourse markers?’ (see, e.g. Fraser 1999) and texts use a number of confusing or seemingly contradictory terms. It would be impossible to create a complete list of DMs because many linguistic items can be used as DMs (cf. Jucker 1993). Some of the most frequently-used discourse markers in present-day English are by the way, besides, however, well (Urgelles-Coll 2012: 23-24), right, well, great, good, there you go, anyway, because (or cos), fine, like, now, oh, okay, so, then, and, but, you know, I mean, as I say, for a start, and mind you (Tagg 2012: 104) but the “set is by no means a fixed or clearly demarcated one” (Tagg 2012: 105). Many of these DMs have been studied in-depth, e.g. 14  D’Arcy’s (2017) monograph on like shows the wide semantic-pragmatic variation between items within this class.11  Discourse markers and interjections share “formal characteristics: both are invariable, typically monomorphemic forms that are syntactically independent from the rest of the utterance” (Brinton and Brinton 2010: 287; see also Wilkins 1992: 124). In terms of their semantics, DMs are not holophrastic. DMs instruct the hearer how to process the information in order to comprehend it - the so-called ‘procedural’ meaning (Schröder 2016: 84; Fraser 1999: 950). While the difference helps several DMs and interjections to be put into their respective groups, the fact that DMs relate to discourse and interjections to emotion does not solve the problem because the same forms have been both discourse markers and interjections in earlier periods of the English language (Brinton and Brinton 2010: 287).  One reason that DMs and interjections overlap is because DMs are able to convey attitude and interpersonal functions instead of merely showing relations in segments of discourse.12 The interpersonal function appears in works of Early Modern English authors who used DMs “to construct character attitudes, relationships, emotions and so on” (Culpeper and Kytö 2010: 396). The items oh and ah overlap the most because they are used extremely frequently and in a plethora of contexts. In the view of some scholars such as Schiffrin, the interpersonal function of oh puts the form into the group of DMs (cf. Fairbanks 2016: 27). Likewise, well is often viewed as a DM and interjection because “it contributes an attitude towards a proposition” (Fairbanks 2016: 36). Thus, DMs and interjections cannot be simply demarcated. Interjections have relationships with other linguistic resources that relate to emotions, especially with respect to the distinction between emotion talk and emotional talk (see Bednarek 2008). Interjections are categorized as forms of emotional talk along with intonation, punctuation, emphatic particles, intensifiers, affective derivation, and vagueness; simply put, they are “linguistic expressions that conventionally signal the speaker’s emotions” (Bednarek                                                  11 Of course, DMs are by no means limited to English; instead, DMs exist in most languages. In present-day Polish, the items grouped into DMs are similar to English. These include, but are not limited to, okej ‘okay’, no ‘well’, ale ‘but’, to znaczy ‘well’ (lit. ‘that is’), (no bo) wiesz ‘(well) you know’, tak? ‘right?’, gdzieś tam ‘in a way/sense’ (lit. ‘somewhere there’) and jakby ‘like’ (cf. Adamczyk 2015). 12 This may be a reason why a few DMs receive diminutive and slang suffixes in English, e.g. indeedy, alrighty, righty, righto, whatevs, howevs, and okaysies (on Twitter). Such formations have yet to receive (sufficient) scholarly interest and would be an interesting avenue for future study. 15  2008: 12). Terms of emotion, including he’s wonderful and she loved him, denote emotions through those dictionary and fixed expressions of affect or emotion. Emotion and emotional talk often co-occur in an utterance, such as Ugh! I hate prunes, where ugh! signals disgust/frustration, and I hate prunes contains the emotion verb hate, which makes an Appraisal or involvement analysis particularly relevant to this study.13  The daunting task of dividing these ‘little’ linguistic items also appears in the two main forms of interjections, namely ‘primary’ and ‘secondary’ interjections. Ameka (1992: 105)14 argues that “it seems useful to distinguish between those words that are primary interjections, that is, they are not used otherwise; and [secondary interjections] which come to be used as interjections by virtue of their notional semantics.” Primary interjections “in terms of their distribution can constitute an utterance by themselves and do not normally enter into construction with other word classes” (Ameka 1992: 105) such as for example ouch! and ow! They tend to be “phonologically and morphologically anomalous” (Ameka 1992: 105)15 and can be divided into three subtypes: ‘conative’, ‘phatic’, and ‘expressive’. ‘Expressive’ interjections are further divided into ‘emotive’ and ‘cognitive’ interjections. Although Ameka (1992) does not explicitly divide types of secondary interjections, arguably this functional division applies to them as well. Secondary interjections – interjections which are found as other parts of speech, such as nouns – have semantic values that can be identified independently of the surrounding context to convey attitudes (Ameka 1992: 111). They often include swear words, such as blimey! (a corruption of blind me! or blame me!, according to the OED), damn!, shoot! in English and cholera! (‘damn’; lit. ‘cholera’) and psiakrew! (‘damn’; lit. ‘dog’s blood’) in Polish. Calls of alarm                                                  13 Alternatively, these can be expressed through ‘expressive’ and ‘descriptive’ linguistic resources (see Jakobson 1960), which may be loosely compared with emotional talk and emotion talk respectively. 14 Primary and secondary interjections were first so distinguished by Wundt (1912), cf. Stange (2009).  15 For example, the curious interjection tsk-tsk is formed by dental clicks, which otherwise do not appear in English (Ameka 1992: 106). The fact that tsk-tsk and other interjections, such as psst! and shh! do not contain any vowels has contributed in the past to a view that they are ‘non-words’ (cf. Ameka 1992: 106). I would argue that this claim shows an Anglo-centric view of interjections because many languages (e.g. Czech and Slovak), have many nouns, verbs and words from other parts of speech that do not contain any vowels. For example, the well-known Czech tongue-twister, strč prst skrz krk ‘stick your finger down your throat’, contains four vowel-less verbs, nouns and a preposition.  16  and attention getters also fall under secondary interjections, including help! (Pol. pomocy!), careful! (Pol. uwaga!) and emotive words such as shame! and drat! (cf. Ameka 1992: 111).  The distinction between primary and secondary interjections is not as clear-cut as it might appear because of the origin and use of interjections in other parts of speech. Norrick (2014: 255) points out that “the primary-secondary distinction provides no hard-and-fast, exceptionless classification for interjections.” He has two main reasons for suggesting that the distinction is not hard-and-fast. First, forms considered primary interjections in present-day English (PDE) sometimes derive historically from full lexical items or phrases. For example, golly!, gee! and jeez! have become primary (euphemistic) interjections in present-day English (PDE) but originally developed from the secondary interjections Jesus! and God! Similarly, the interjection LOL! is considered a primary interjection with a more complex development,16 but we may more suitably refer to it as a secondary interjection because it originated as an acronym for ‘laughing out loud’ (Norrick 2014: 255). Second, a primary interjection can be “pressed into service as another part of speech, so that it would appear to be a secondary interjection from a purely synchronic perspective” (Norrick 2014: 255). For example, the interjection wow! can become an interjection-based delocutive verb17 in expressions including he wowed the audience and wowing the customers. Or, a primary interjection such as ow! may receive a diminutive suffix that creates the diminutive form owie! that is found in other parts of speech (e.g. a noun in ‘I have an owie’). In the following subsections, I provide background for (primary) emotive interjections (1.2.1), cognitive interjections (1.2.2), conative interjections (1.2.3), secondary emotive interjections (1.2.4), contact interjections (1.2.5), and descriptive interjections (1.2.6). However, the interjections discussed in the rest of this dissertation are a) primary emotive interjections, b) contact interjections, and c) secondary emotive interjections because they are affixed most frequently.                                                  16 While Norrick (2014) describes LOL as ‘primary’ in PDE, he notes that since it is used as a verb in expressions such as he LOLed so hard caused the item to be better described as having developed from a ‘secondary’-type acronym into a primary interjection, and then developing back into a ‘secondary’ interjection after being pressed into service in other parts of speech.  17 According to Brinton (2014: 140), interjection-based delocutive verbs are “verbs converted from interjections with the meaning ‘to say or utter [interjection]’”. Their historical development is thoroughly examined in Brinton (2014), who provides such examples as hey’ed and aww’ed in PDE.  17  1.1.1  (Primary) emotive interjections Emotive interjections have recently been thoroughly discussed by Stange (2016) and in many other shorter studies (e.g. Wierzbicka 2003). They are one of the two subcategories of ‘expressive’ interjections which focus on the mental state of the speaker (Stange 2009: 29; cf. Ameka 1992: 113). The main function of emotive interjections is to convey emotion, ranging from disgust (e.g. ew!, yuck!, ugh!) to joy (e.g. yay!) to surprise/shock (e.g. whoa!). Sadowski (2001: 71) comments that primary emotive interjections “have arisen in the course of evolution as communicative responses to typical life situations such as survival, sexual behaviour, search for food, group integration and so on.” At the very least, they express an emotion to someone – if not to another person, then to oneself. Stange and Nübling (2014: 1986) write that verbalized interjections are always accompanied by body movement such as the facial movements of rounding the lips or pinching the nose.18 In fact, it is universal for humans to express their emotions through various sounds formed through the movement of lips and other facial features, and gestures. Darwin (1872: 258) describes how disgust is shown on the face by, e.g. “by blowing out of the protruded lips; or by a sound as of clearing the throat … [which are then] written ach or ugh; and their utterance is sometimes accompanied by a shudder.” Fowler (1850: 332), a contemporary of Darwin, suggests that “[w]onder or astonishment rounds the lips; hence results the interjection Oh! with a downward intonation.” More recently, Stange and Nübling (2014: 1986) point out that “the face displays a number of expressions when speakers feel revolted: they may wrinkle up or pinch their nose, screw up their face, cover mouth and nose with their hand, poke out their tongue” and physically move away from the disgusting object, while likely verbalizing their revulsion by an interjection including yuck!, ew! or ugh! Thus, the expression of emotions causes the facial muscles to contract.19                                                  18 According to Motley (1993), a facial expression, such as pinching one’s nose, functions as a nonverbal interjection. Thus, verbal and nonverbal interjections go hand in hand. 19 In digital communication, such as social media, where there are limitations in expressing body movements beyond the text, these facial movements and gestures are represented through the addition of emoticons (e.g. happy faces) and emojis (e.g. a thumbs-up image). 18  In her corpus-based study of emotive interjections in British English, Stange (2016) argues that ‘interjectionality’ is determined by five factors: the context of use, the variant used (if alternatives are available, e.g. ow! and ouch!), the physical absence/presence of the stimulus, the nature of the stimulus (concrete or abstract), and whether the interjection is spoken for oneself or another. She uses these five factors for the interjections of surprise, disgust, and pain. Stange (2016) suggests that the interjections of surprise - namely wow!, whoops!, and whoopsadaisy! - show the discursive reaction to a particular situation more often than emotion. Wow! functions as an intensifier, is found in the speech act of complimenting, and is used to express unemotional surprise in spoken language, though the largest differences appear to lie between adult and child speech. For example, wow! is used for praise or admiration only by adults (Stange 2016: 199); and, wow! is used to express ‘more or less X than usual’, ‘praise/admiration’ and ‘neutral surprise’.20 Stange’s results show that wow! is not as emotional as we would expect; instead, wow! seems to be linked with politeness formulae in its function as a compliment. In comparison, the interjection whoopsadaisy! is particularly rare and mainly said to children, and whoops is used in the contexts of “situation changed noticed, situation caused by own action and mistake noticed” (199, original emphasis), which, like wow! above, seems to function more as a response to a situation or action than an expression of deep feeling or passion.21 The interjections of disgust (yuck! and ugh!) in adult speech are used in discursive contexts, but each variant is a response to a different sense or physical closeness (cf. Stange 2016; Goddard 2015), particularly between smell and sight. They are typically responses to a physical/sensory exposure to, for example, “tasting something very bad, confronting decaying food in the fridge, or finding vomit in a public toilet” (Goddard 2014: 87). Goddard (2014: 88) posits that ugh! “is frequently used in response to smells,” while yuck! is most frequently caused by ‘oral revulsion’. Ugh! activates and sounds like the gag response of retching that is induced by smells, while yuck! is closely linked to the mouth and body; for example, “[c]hildren come out with Yuck! […] when                                                  20 Children, on the other hand, used wow! when experiencing a ‘pleasant event’ and ‘positive surprise’.  21 As with wow, the children’s use was slightly different; the most frequent contexts were “situation caused by own action and failure of intended situation” (Stange 2016: 198-199). 19  rejecting the prospect of some kind of unwanted food, as if to indicate that they could not stand to think of such food entering their mouths” (Goddard 2014: 88). Likewise, bird poop landing on a person’s arm would also invoke yuck!, but could also invoke ugh! if the bird poop landed close enough to the nose to smell.22 In sum, a person can feel disgust from one or more senses, especially smell, vision, and taste, which tend to be associated with the perceived physical closeness to the object of disgust. These reactions of disgust also have a ‘discursive function’, whereby they can occur in imagined “situations in which the stimulus is not something in the immediate context, either a physical-sensory stimulus or a human action or behaviour, but rather something the speaker is thinking about” (Goddard 2014: 90). For example, Goddard’s example of I have to clean the toilet. Ugh! is an imagined situation that stimulates the visual and olfactory senses, while the imagined scenario in Sex with DFK? Yuck! stimulates the oral senses and has “a mouth-related component” (Goddard 2014: 88). This function is primarily used by adults (Stange 2016: 198). Stange (2016: 135) and Goddard (2014: 87) find that the interjection ugh! in the yuck-ugh interjectory pair is more ‘interjectory’ than yuck! which tends to be used more frequently due to its sound, automaticity, and spontaneity. It is more frequent than yuck! overall, occurring at least twice as often in the Wordbanks corpus than yuck!, likely because ugh! is automatic and “noise-like, while [yuck!] is word-like23 [and] seems intuitively a bit less spontaneous and a bit less visceral than Ugh!” (Goddard 2014: 87-88). Goddard also finds frequent attestations of yuck yuck yuck! in Wordbanks, but not ugh ugh ugh!  The expressions for pain, ow! and ouch!, have a discursive function used only between adults (Stange 2016: 198) to express three types of pain: imagined (not ‘real’ pain24 – or ‘pretend’ pain when playing with toys, etc. – experienced by the speaker or someone else), experienced (real                                                  22 See Goddard (2014: 88-89) for an in-depth explication of ugh! and yuck! using semantic primitives. 23 In addition, “the feature position of explanation was in fact highly significant as regards the distribution” (original emphasis) of the two forms.  That is, despite the same placement of the two interjections in Goddard’s examples above, yuck! was typically preceded by an explanation of why the speaker found something disgusting (e.g. All that hair. Yuck), while ugh! either had no explanation or the explanation was placed after the interjection (e.g. Ugh look, it’s horrible) (Stange 2016: 135). 24 ‘Real’ pain is the presence of “a clear physical basis for their pain” (Loland 2006: 51). ’Pretend’ pain, in contrast, is feigned pain where the speaker only pretends to be hurt.  20  pain experienced by the speaker or someone else) or anticipated (empathetic use for not real pain experienced by the speaker or by someone else). In addition, the interjection ouch! “was found more frequently in empathetic use” (Stange 2016: 198) than ow!, the interjection favoured by children. (Interestingly, using ouch! or ow! as a means of ‘expressing frustration’ and ‘getting attention’ was only employed only by children.) This difference between adult and child speech functions raises the question as to whether adding an affix to ow! or ouch! provides an adult speaker a means for ‘getting attention’ and ‘expressing frustration’.  In sum, the position that ‘primary’ emotive interjections express different meanings depending on the specific interjection used, the context, and whether the speaker is an adult or a child is persuasive. The many possible functions and meanings show the multi-functionality and discursive functions of emotive interjections in general, and these are likely applicable to their affixed forms as well. 1.1.2 Cognitive interjections Another ‘expressive’ interjection is the ‘cognitive’ interjection, which, following Ameka (1992: 113), “pertain[s] to the state of knowledge and thoughts at the time of the utterance”, (e.g. aha!, oho!) and shows the speaker’s mental state. Wierzbicka (2003: 326) defines cognitive interjections as “global (unanalysable) expressions which express the speaker's mental state without reference to feeling or wanting.” In other words, cognitive interjections typically mean ‘now I know it’. While Wierzbicka claims that emotive or attitudinal meaning can be attached to cognitive interjections in a suitable context to invoke irony or satisfaction, other studies assume that these interjections convey the state of knowledge more than feelings or emotion because ‘emotive’ interjections fulfil the need for a subcategory of interjection which conveys the speaker’s immediate emotions at the time of utterance.  1.1.3 Conative interjections ‘Conative’ interjections (called ‘volitive’ interjections by e.g. Wierzbicka 2003) are directed at another person (Ameka 1992: 113) in order to instruct “what the speaker wants the auditor to do (or not to do), thus also containing items that have semantics similar to that of vocatives and imperatives” (Nordgren 2015: 80). They are said to a person or animal from the basic meaning of ‘I (don’t) want you to do something now’ (Wierzbicka 2003: 293) for six main reasons. They are: 21  a) ‘I want silence’ (shh!; Pol. sza!, cii!),25 b) ‘I don’t want you in this place’ (e.g. shoo!; Pol. sio!), c) ‘I want you to jump’ (e.g. Pol. hop!, hopla!), d) ‘I urge you to X’ (e.g. Pol. hej!, nuże), e) ‘I want you to hear me (from a long distance) (e.g. hello!; Pol. halo!, ahoj!), and, f) ‘I give it to you’ (e.g. here!; Pol. na!). Because of this ‘command’ function, conative interjections “are often followed by a vocative identifying its target” (Cuenca 2011: 187), e.g. Psst, Annie! or in Polish, Sza, mamo! (‘Hush, mom!’).  Wierzbicka (2003) divides conative interjections into two classes in Polish, namely those used with animals (e.g. prr! ‘whoa!’ to get a horse to stop, sio! ‘shoo!’, and reduplicative cip-cip-cip to get birds to come) and the class used with human beings (e.g. shh! or psst!). Unlike expressive interjections, conative interjections do not inherently convey a mental state, but they may, as Stange (2016: 13) suggests, “have an emotive component in that their utterance may express annoyance at the noise level (Shh!), or the speaker's desire to share delicate news (Psst!).” Thus, these interjections may have various layers of nuances and function that require further exploration.  Finally, conative interjections are not necessarily always primary because the imperative verb quiet! (Pol. cicho!), for example, is secondary. In fact, conative secondary interjections typically derive from nouns, imperatives, and adverbs (cf. Cuenca 2011: 187). While the secondary forms sometimes have an emotive component (e.g. an angry shout of quiet!), I do not include them in the secondary emotive interjection group discussed in the following subsection. 1.1.4 Secondary emotive interjections Secondary emotive interjections can be single or multi-word expressions that derive from other parts of speech (Gehweiler 2011) to primarily convey emotion.26  Several English and Polish emotive secondary interjections function as euphemisms or expletives (swearwords, profanity, or foul language) (see Hughes 2006: 154). I include complex and phrasal interjections (also termed ‘interjectional phrases’) as types of secondary interjections (following Ehlich 1986 and                                                  25 It should be noted that sza! likely derives from cisza ‘silence’, while cii! likely derives from cicho ‘quiet’ (cf. Wierzbicka 2003). Wierzbicka (2003: 293-295) discusses each form in more detail. 26 Secondary emotive interjections do not include secondary interjections from other types, e.g. the secondary conative interjections quiet! 22  Bloomfield 1933), such as for example, dear me! and by golly! Other emotive secondary interjections are derived from syntagms (e.g. blimey!) and phrases (e.g. LOL!). Therefore, the most commonly cited secondary emotive interjections include the following: • (oh) boy, (oh) no, (oh) hell, damn, crap, shit, fuck, goddamn; LOL; Jesus, God, Christ; blimey • cholera ‘damn’, psiakrew ‘dog’s blood’; Jezu(s) ‘Jesus’, Boże ‘God’, Maria ‘Mary’ Secondary interjections in PDE generally appear to receive less scholarly attention than primary interjections, though Mohr (2013), Hughes (1991), McEnery (2006) and Allan and Burridge (2003) provide exemplary research into swear words throughout the history of English. Smaller-scale studies of secondary interjections typically focus on sociolinguistic distribution, language-dependent properties, meaning in social contexts, and their origins through the processes of ‘grammaticalization’ and ‘subjectification’ (Traugott 1995b).27  As we have seen earlier, there is no fixed and demarcated line agreed on by all scholars about which words are ‘secondary’ interjections and how they precisely fit with the group of words known as ‘swear words’. These have many alternate labels, including ‘bad words’, ‘cuss words’, ‘dirty words’, ‘expletives’, ‘profanity’, ‘blasphemy’ and ‘rude language’ (Fägersten 2012: 3). Therefore, they may potentially “be offensive, inappropriate, objectionable, or unacceptable in any given social context” (Fägersten 2012: 3). Of course, not all swear words are secondary interjections because they are not all exclamatory, holophrastic, or syntactically independent. However, secondary emotive interjections seem to have a tendency to be socially unacceptable in many contexts and/or have a higher level of emotional involvement.28                                                   27 Grammaticalization is “the process whereby lexical items and constructions come in certain linguistic contexts to serve grammatical functions, and, once grammaticalized, continue to develop new grammatical functions” (Hopper and Traugott 1993: xv). (See, e.g., Traugott 1995, Brinton and Traugott 2005). 28 The argument for the idea that secondary forms are more intense than primary is made by Drescher (1997: 238), who suggests, in a discussion of French interjections, “that secondary interjections indicate affectivity in a more intense and differentiated way than the less expressive primary interjections. The most frequent primary interjections ah and oh seem to have lost part of their expressive potential due to their abundant use.” Thus, oh! is less emotive than damn!, and ugh! is less emotive than shit!, particularly in the amount of anger and negative emotion expressed. However, interjections tend to weaken in force over time as they wear out from frequent use (Burridge and Mulder 1998). This lessens the emotive difference between the two types of interjections. Even Drescher suggests that the age and frequent use of primary interjections may account for the fact that secondary interjections may be more emotionally 23  Crystal (2003: 239) writes that there is “an unclear boundary between [interjections] and other types of exclamation, where some referential meaning may be involved, and where there may be more than one word.” The term ‘exclamation’ is defined as “a vocal act compared to a ‘cry,’ which is itself defined as a loud utterance” (Ozzello 1978: 8). In addition, exclamations “are primarily for expressing the speaker’s own feelings” (Quirk et al. 1972: 386) and “express surprise and unexpected or otherwise remarkable information” (Aikhenvald 2015: 238). Sometimes the adjective ‘exclamatory’ is used because secondary interjections and exclamations are typically defined as ‘exclamatory’ items; that is, they suddenly or strongly vocalize an emotion usually of surprise or shock, and often ends with an exclamation mark in written language.29 There are two approaches taken towards exclamations and interjections that have not solved the terminological problem: in one approach, they are viewed the same, with “no line of distinction between them” (Charleston 1960: 49). In the other, perhaps more popular approach, scholars suggest that “[m]any interjections, and among them all ejaculations, are typically uttered with an exclamatory intonation. But this does not imply that every interjection is an exclamation or the other way around” (Poggi 2009: 180; see Fowler 1850: 331). We would not consider the exclamatory response of shocking! or a person’s glad exclamation of tulips! when being given a bouquet of tulips in early spring as a surprise gift to be secondary interjections, even though they derive from various parts of speech, including nouns. However, unlike secondary interjections, almost any word can become an exclamation, and there is no rule which states that secondary interjections can never be exclamations.  In sum, secondary emotive interjections often overlap with exclamations, swear words, discourse markers, and other emotive words with alternative aliases; they either derive from or are pressed into service in another part of speech; they are highly innovative; and, typically have a high level of emotional content. In later chapters, I will show how the addition of evaluative suffixes adds a further emotional component to these secondary interjections, e.g. in dear me! → dearie me! and Lord! → Lordy!                                                   intense. 29 I exclude exclamatory sentences here; that is, exclamations beginning with what (e.g. what a shock!) or how (e.g. how horrid!) because they would not be considered secondary interjections (cf. Leech and Svartvik 1975: 230). Thus, I only consider exclamations that consist of one or two words, e.g. dear me!, hell! or oh boy! 24  Next, I turn to ‘contact interjections’ (i.e. conversational formulae) that are considered by some researchers, e.g. Norrick (2011), to be secondary or phatic interjections. 1.1.5 Contact interjections Contact interjections are generally viewed as linguistic items typically found in conversational routines (e.g. sorry!, hello!, goodbye!) and have received much scholarly attention (e.g. Aijmer 1996). Although they were traditionally viewed as interjections (cf. Ameka 1992), they are currently rarely included in the class of interjections. Rather, formulae are generally defined as “intentional and (socially) expected reactions to situations” (Ameka 1992: 109) that are “among the most conventionalized and perfunctory doings we engage in” (Goffman’s 1971: 90).30  In other words, formulae are viewed as ritualized linguistic items that do not (generally) show the speaker’s current and spontaneous mental state or emotions. However, not all scholars leave formulae completely outside of interjections. Wilkins (1992) and Sauer (2012: 158), for instance, draw on the similarities between conversational formulae and interjections to view formulae as a subdivision or function of interjections. This is a persuasive position because conversational formulae, like interjections, may express or display the speaker’s immediate mental state and are generally holophrastic and syntactically independent. Because of this, I refer to conversational formulae as ‘contact interjections’ (a direct translation of the Czech term citoslovce kontaktová) which includes them as a type of interjection.  There is a social expectation to say certain contact interjections, such as greetings when meeting someone. However, a speaker who utters a greeting is arguably in a state of mind at that moment to express a greeting – whether the greeting is an enthusiastic hi!, neutral hi, cold hello, or a mere grunt of acknowledgment. Conversely, emotive interjections may also be ritualized in discourse. The emotive interjections wow! or yay! may also be said as a type of ritualized social convention; that is, a speaker may say wow! after receiving a gift only out of politeness. In addition, one could arguably fake an excited wow! the same way as one could force an unwanted hello. Thus, wow! and hi! can both be automatic (ritualized) responses (cf. Ferguson 1981: 22) or (genuine) expressions of feelings or mental states. If the meaning of ‘emotion’ or                                                  30 Sometimes they are viewed as secondary interjections; for example, Norrick (2011) includes yeah, okay and hey as ‘secondary interjections’. 25  ‘feeling’ is a prime characteristic of interjections, then arguably contact interjections fall into the broad group of interjections.  Like most interjections, contact interjections are syntactically independent and convey meaning. Even Ameka (1992: 109) admits that contact interjections are indexical (though towards an addressee) and “can [like true interjections] constitute non-elliptical utterances by themselves.” In addition, contact interjections are holophrastic, unlike DMs. Because of these characteristics, I consider conversational/politeness formulae to be interjections – albeit not generally expressive interjections or interjections proper. In this, I am following Wilkins (1992), and consider Ameka’s (1992) categorization of interjections vs. formulae as too narrow. It needs to be mentioned that contact interjections have also been considered separate from interjections because of their historical origins; many derive from full phrases (e.g. I am sorry and I thank you). However, I follow Bloomfield (1933), Leech et al. (1982) and descriptive grammars by putting these current and shortened forms of polite formulae – for example, thanks and please31 – under the umbrella term of ‘interjection’. In purely diachronic terms, these contact interjections underwent a process associated with grammaticalization and/or pragmaticalisation,32 e.g. I thank you → thanks/thank you (Jacobsson 2002), “whereby linguistic items […] gradually lose their original, propositional meaning and take up textual and interpersonal functions” (Furkó 2014: 291, original emphasis).  The original politeness phrases are now equated with interjections because they “have lost their original sentential elaborations (“if you please”) [and now] often occur as pure isolates” (Leech 2014: 66). Not all formulae are interjections, particularly multi-word expressions, such as I thank you, but it would be counterintuitive not to acknowledge the similarities. The contact interjections in English and Polish include the following: Greetings and farewells: In English and Polish, some frequent greetings are the following: • good morning, good afternoon, good day, good evening, hello, hi, hey, yo                                                  31 The origin of please is debated – see Allen (1995). 32 Some scholars prefer the term pragmaticalisation (e.g. Aijmer 1997), while others prefer the term grammaticalization (e.g. Traugott 1995a) for the process that items such as y’know and I mean undergo to become discourse markers. Since this dissertation is not diachronic, I consider either term suitable because both refer to a change in the pragmatics and grammar of the linguistic item. 26  • cześć ‘hi/bye’, dzień dobry ‘good morning / good day’, dobry wieczór ‘good evening’, hej ‘hey’, siema ‘hi how are you’ In comparison, some frequent farewells are the following: • goodbye, bye, (see you) later, goodnight, so long, cheers, cheerio, bye-bye • cześć ‘hi’/ ‘bye’, dobranoc ‘good night’, pa ‘bye’, do widzenia ‘goodbye’, do zobaczenia ‘goodbye’, na razie ‘so long’, do jutra ‘until tomorrow’ Requests: Note that there is only one contact interjection of request in English, but there are two contact interjections of request in Polish: • please • proszę ‘please’ (also can mean ‘you’re welcome’ and ‘here you go’, lit. the verb ‘I ask’), plis (< Engl. please) It should be noted, however, that proszę for ‘please’ is the standard form, while the newer form, plis, would not appear in a standard Polish dictionary. Polish speakers have recently borrowed please from English as plis, though plis is informal and typically used among youth on social media.  Apologies: The English and Polish interjections of apology are the following:  • sorry, excuse me, pardon me • sory ‘sorry’, przepraszam ‘(I’m) sorry’ Gratitude and appreciation: The contact interjections which convey gratitude and appreciation in English and Polish are the following: • thank you, thanks • dziękuję ‘thank you’, dzięki ‘thanks’ 1.1.6 Descriptive and onomatopoeic interjections Last, there is a group of ‘descriptive interjections’ which consists essentially of “ideophones [which] display phonetic symbolism, often iconic, based on onomatopoeia or phonaesthetics” (Award 2001: 728). Unlike the previously-discussed interjections, they do not “index aspects of the speech event” (Award 2001: 728). They include interjections frequently found in comics to represent sounds, such as e.g. bam!, wham!, boom! and thud! Some have received slang suffixes: in the OED, we find whammo! (wham + -o suffix) which first appeared in 1932. Bammo! (bam + -o), in comparison, appears on Twitter but not in the OED. From these results, we can argue that 27  several interjections from this group take suffixes that add an expressive nuance. Since they are not (typically) expressive, but rather tend to be descriptive (Ameka 1992), they do not fit within the scope of this dissertation.   1.2 Morphology of interjections It is commonly thought that interjections “resist virtually all morphological processes, i.e. that they allow neither inflection nor derivation” (Stange 2016: 36); in other words, they are “morphologically simple”. In his definition of the interjection, Wilkins (1992: 124) claims that interjections do not generally “host inflectional or derivational morphemes”. For the Slavic languages, Stankiewicz (1986: 223) writes that primary (‘simple’) interjections, such as Russian ax! ‘ah’, fu! ‘yuck’, tfu! ‘yuck’, “have the lowest degree of lexical specificity, and are morphologically the least characterized elements of the system.”  A closer examination shows that English interjections are subject to morphological processes because they can undergo at least one process (conversion33 from one POS to another) and to some extent are also subject to (zero-)derivation and/or inflection. One may observe instances of conversion frequently; such forms include interjections to nouns (e.g. ow!int → owien), verbs (e.g. ouch!int → ouchedv), and adjectives (e.g. ugh!int → ughadj). These are found in phrases such as the ugh factor and I have an owie.  Derivation is somewhat more complicated. Stange (2016: 38) takes the view that “[i]n a first step the [emotive] interjections allow conversion, and only in a second step derivation and/or inflection,” but it seems that some interjections allow derivation before conversion to other parts of speech. The addition of diminutive suffixes to interjections (e.g. whoops!int → whoopsy!int) show derivation, and interjection-based delocutive verbs such as in she oohed and aahed over the baby demonstrate interjection → verb conversion (see Brinton 2014). When conversion is involved, then it is acceptable for converted interjections to become more morphologically                                                  33 There are several different terminologies and theoretical views on conversion. Bauer, Lieber and Plag (2013) discuss these theoretical views. It is a ‘morphological operation’ and ‘derivational process’ that “links an input form to an output form: [...] it turns one thing into another” (562). In its broadest sense, it takes a form (the input) and converts it into another form (the output). 28  complex. Without conversion, the process is extremely limited, but this does not mean that non-converted interjections cannot receive diminutive/augmentative and slang suffixes.  In sum, what can be agreed upon is that while it is uncommon that English34 interjections receive derivational affixes, they are able to do so. To quote Stange (2016: 32), “speakers subject interjections to morphological processes only rarely, but the matter of the fact is that they do it in the first place.” They can be found in advertising slogans, e.g. the television commercial which claims that the product advertised is “Tough on Yuck, Gentle on Everything Else”. Here, yuck is used as a noun. These processes also depend on the suffix used in the formation of the affixed form, and I will provide an in-depth look at these affixes in Chapter 2.  1.3 Interjections in the Slavic languages The Slavic languages currently consist of at least thirteen languages that comprise three main groups: a) East Slavic (Russian, Belarusian, Ukrainian), b) West Slavic (Polish, Czech, Slovak), and, c) South Slavic (Slovenian, Serbian, Croatian, Bulgarian, Bosnian, Montenegrin).  Originally, the Slavic languages came from one language (called Proto-Slavic or Common Slavic) that changed over time; it was not until after the 12th century that the different groups of Slavs were no longer able to understand each other. Therefore, the languages in each group are often very similar because of shared history, geographical situation, and religion. For example, several of the South Slavic languages are nearly identical, but are different due to their socio-political, historical, and religious backgrounds. Their religious and political divisions often were reflected in language, such as the choice of alphabet (Cyrillic or Roman). Besides natural linguistic changes, there are a variety of ways that one group attempted to establish itself as separate from another.35  Each Slavic language has numerous interjections, most of which came from the same Proto-Slavic roots, that are shared with other Indo-European languages. Many primary interjections                                                  34 I note ‘English’ here because the morphological processes described here are not always applicable to Polish. Unless a process or feature has been shown to be universal by cross-linguistic analysis, it seems reasonable to specifically state which language(s) such descriptions may be applied to. 35 See Stankiewicz (1986), and Sussex and Cubberley (2006) for a more detailed discussion of the history of the Slavic languages. 29  are easily transferable because they are still shared by most, if not all, of the Slavic languages, such as for example, o!, ach!, och!, oho! (Karamysheva 2012: 232), and the secondary interjection Boże! ‘God.VOC’ (< Bóg). Other interjections are more localized to a Slavic language or group and may not easily transfer into other Slavic languages. However, translating an interjection from a Slavic language to a non-Slavic language is more difficult. While there may be several equivalents between, e.g. Polish and Russian, these are unlikely to carry over to a different language group. The differences between a Slavic language such as Polish and a Germanic language such as English are notably greater and are a rich source for the cross-linguistic investigation of interjections. 1.3.1 Polish interjections Interjections have received much attention by Polish scholars (see, e.g. Daković 2006, Wierzbicka 2003, Grochowski 1993).  Milewski’s (1965) grammar of interjections is the most frequently referred to text on interjections by Polish scholars (cf. Kryk-Kastovsky 1992). It shows many similarities between Polish interjections and mainstream thought about interjections, e.g. their division into ‘primary’ and ‘secondary’ forms. Milewski’s categorization of interjections is not unlike contemporary views by English scholars; namely he distinguishes between primary interjections (e.g. ej! ‘hey’, ah! ‘ah’), appeals (e.g. hop-hop! ‘hey-hey’, halo! ‘hello’), onomatopoeic forms (e.g. sza! ‘hush’, hu-hu! ‘boo’) and secondary interjections (e.g. Jezus, Maria! ‘Jesus, Mary’). The categories are the same, but some interjections are categorized differently than, e.g. Ameka (1992). For example, sza! ‘hush’ is viewed as an onomatopoeic form instead of a conative interjection. Only slight modifications have been made to Milewski’s categories, namely by Wierzbicka (1991), who divides interjections further into emotive (‘I feel something’), volitive (‘I want something’), and cognitive (‘I think something’). Overall, it appears that mainstream Polish scholars have not made any momentous deviations from our general understanding of interjections. There are a number of Polish interjections that do not appear in English (and vice versa) because “interjections (with the exception of the simplest forms) do not have a universal scope, but are specific to each individual language” (Weinsberg 1983: 15, my trans.). Cultural and social issues (e.g. the importance of the Virgin Mary in Polish Catholicism) thus distinguish Polish 30  from not only English but other Slavic speakers. For example, Goddard (2011: 184) writes that the English interjections gee! and wow! do not have equivalents in Polish.36 Likewise, the slightly outdated Polish interjection hejże! ‘hey’, which urges the listener to do something, does not have an equivalent in English. English well! and Polish no! ‘well’ are also seen as not precise equivalents. Tabakowska (2000: 321) compares the frequent use of Jezus, Maria, Józefie Święty! ‘Jesus, Mary, St. Joseph!’ in Polish to its rare occurrence in English where it is used with (slight) reverence towards these religious figures. Wierzbicka (2003) compares emotive interjections of disgust such as tfu, fe, fu, Russian fu, and English yuck and notes that there is no direct correspondence between the Polish and English interjections of disgust. Broadly, tfu! expresses “contempt and moral disgust” (312), fe “expresses a mild moral disgust”, while fu! is used “when one’s nose or mouth comes close to something perceived as ‘bad’ and when one feels something bad because of that and wants to avoid the offensive contact (or closeness)” (303), particularly in ‘disgusting’ food, smells and eating habits. In more specific terms, Wierzbicka claims that the Polish interjection fe! places the speaker on a higher moral ground because it roughly “expresses a mild moral disgust [and] is always a reaction to human behaviour” (306), and is used particularly to shame children into changing their behavior. In this way, fe! is not interchangeable with the similar Russian interjection of disgust, fu!, or even its Polish counterpart fu!, both of which have a broader sense of being an almost physical reaction to an object or thought that the speaker finds repulsive or, at the very least, annoying.  Sieradzka and Hrycyna (1996) include the contemporary fuj! in their discussion of interjections of disgust. In addition, they suggest that tfu! is significantly different than the other Polish interjections of disgust; namely that it has the meanings of ‘I despise’ and ‘I am surprised I said that’. It only joins the other interjections of disgust in the meaning of ‘I loathe it, I am disgusted by it’, as shown below:  • ‘I am dissatisfied’: fe, fu (fuj), pfe, pfu (pfuj) • ‘I am surprised I said that, how could I have said that, that’s a stupid mistake’ and ‘I despise’: tfu                                                  36 See Szkapienko (2016) for a study of borrowed English interjections wow! and yes! in present-day Polish and Russian. 31  •  ‘I loathe it, I am disgusted by it’: fe, fu (fuj), pfe, pfu (pfuj), tfu •  ‘be ashamed of yourself, it is wrong, I negate it, I do not like it’:  fe, uj (202-203).  The interjection (o)jej! is, perhaps, an ‘overarching’ interjection that can convey many different meanings, ranging from dismay to delight depending on context. According to the Słownik Języka Polskiego (SJP),37 it is an interjection that conveys “various emotional states, e.g. admiration, helplessness, fear, threat, surprise, astonishment” and has several synonyms, including oj!, ojoj!, ojojoj!, ojeju!, joj!, jej!, and the diminutive forms ojejku!, jejciu! and jejciu!  In sum, while views of Polish interjections and where they fit into the typical classification and functional typology are similar to those of ‘standard’ English perspectives, the Polish language possesses a wide variety of interjections that do not correspond well to their standard dictionary English equivalents. Polish no does not quite have the same colouring as English well, and the three Polish interjections of disgust – tfu!, fu!, fe! - each have a meaning that does not correspond to English yuck! without explanation. The Polish interjections also have slightly different meanings or connotations than their Slavic counterparts, such as Czech and Slovak, or Russian, which I briefly discuss below. 1.3.2 Czech and Slovak interjections There have been numerous studies of interjections in the Czech and Slovak languages (see, e.g., Trávníček 1930, Uhriková 2012). Some of the major findings of these studies have been that Czech and Slovak share many interjections in common with each other and with Polish. Overall, the definition of an interjection and its typology between English and the West Slavic languages of Polish, Czech and Slovak are very similar. For example, Trávníček (1930) focuses on two important aspects that do not appear often in English, namely the ‘verbalisation’ of interjections and interjections that appear in other parts of speech. In addition, Trávníček (1930: 11) uses the (perceived) origin of Czech interjections to place them into two main groups: subjektivní ‘subjective’ (e.g. ‘original’ (primary) interjections such as ach! ‘ah’ and ‘non-original’ (secondary) interjections such as pst! ‘psst’) and objektivní ‘objective’ (e.g. sounds not naturally produced by humans, such as mú ‘moo’). For Slovak, Uhriková (2012: 114) concludes that spoken Slovak                                                  37 The Dictionary of the Polish language. 32  interjections “are more evenly distributed than in English.”38 It is fair to conclude that these results would be similar, if not the same, in Czech. In the following chapter, I discuss the many suffixes available for Czech and Slovak interjections. 1.3.3 Russian interjections Russian interjections have been well studied by linguists across the 20th century (e.g. Wierzbicka 1991, Liston 1971). These linguists have characterized Russian interjections as having no connections with words in a sentence, no denotational meaning, and no derivative or inflectional morphology – much like the typical English perspectives of grammarians discussed earlier. In fact, Rochtchina (2012: 72) notes that interjections, adverbs, modals, and conjunctions differ from the main synthetic parts of Russian and rather “appear as islands of analyticity (or analytism), as they are indeclinable and express their grammatical meaning in context by means of syntax.” Russian linguists also note several features that often are left out by English scholars; for example, they examine conversion of interjections to verbs, reduplicative morphology (e.g. oj-oj-oj), marginal suffixation with affixes (e.g. nate-ka), the particle -ka, and special sounds, such as in t’fu to express disgust (cf. Liston 1971: 479). While some of these features, namely the particle -ka and diminutive suffixation through the -k- affix are not applicable to English interjections, they are a frequent and well-established aspect of Russian interjections. Russian verbal interjections and many primary interjections are easily and frequently converted to (semelfactive) verb infinitives by the addition of the verb ending -kat’ or -nut’. The verbs may then be easily conjugated for first-person, second-person, etc. forms in the past, present, or future tenses. This process of converting interjections to verbs is an example of how the Russian “interjection is productive and constantly enlarges its domain at the expense of the others” (Matthews 1953: 75). The process creates verbs such as, e.g., стукнуть [stuknut’] ‘to knock’ and ахнуть [axnut’] ‘to exclaim’). Pontoppidan-Sjövall (1959: 56) adds several more ‘imitative verbs’ that derive from interjections, namely увы [uvy] > увыкать [uvykat’] ‘to cry alas!’, да [da] > дакать [dakat’] ‘to say yes’, усь [us’] > уськать [uskat’] ‘to cry woof!’. In addition, Liston (1971: 482) adds ajkat’ ‘to pronounce or exclaim aj’ and nukat’ ‘to say nu [well] to someone;                                                  38 Uhriková (2012: 107) also compares spoken Slovak and English interjections, and finds that the ten most frequent Slovak interjections are aha!, jaj!, jáj!, joj!, jéj!, jój!, ó!, ach!, fuj!, and wow! 33  to hurry a person up by saying nu’. In sum, through these verbs we can see some different processes at work in Russian than an English or Polish; notably, the easy conversion of interjection to verb for many, if not most, interjections.  The Slavic slang/informal particle -ka appears frequently in Russian (but not in Polish) and is joined with a hyphen at the end of words (particularly verbs) for additional emphasis or attitude. While it looks like the diminutive suffix -ka, the “Slavic emphasizing particle =ka/=ko […] [is] cliticized to pronouns, adverbs and imperatives; e.g. Russian mne=ka ‘to ME’, Ukrainian tút=ka ‘here’, Bulgarian áz=ka ‘I’, Czech dnés=ka ‘today’; Russian podí=ko! ‘do come!’” (Hewson and Bubenik 2006: 182). Generally, the particle adds a certain attitudinal nuance to the word; for example, the translation of skaži-ka by Koktova (1999: 110) adds the attitudinal just in “just tell me”, whereas skaži straightforwardly means ‘tell me’. The particle can also serve to emphasize the interjection nu ‘well’. Andrews’ (1989: 130) example of Nu-ka družok, sognite-ka ruku v lokte ‘Well friend, bend your arm at the elbow’ shows the interjection nu ‘well’ and the verb sognite ‘bend’ using the particle -ka in a speech situation explicitly between friends. Generally, like diminutives, they “possess a unified general meaning [as both] require a special between the speaker and receiver” (Andrews 1989: 130) such as conveying a feeling of emotion and intimacy. Overall, the -ka particle changes a verb from an imperative command to an appeal and an utterance with an interjection to a more intimate, lower register utterance that would not be found in formal language. Despite its popularity in the East Slavic languages, the particle is not used in the Polish language. Overall, interjections are thought of in similar ways across the Slavic languages and English. The individual properties found in each language contribute to some differences, such as for example the formation of verbs or other parts of speech from interjections, and their likelihood of diminutive affixation. The syntactic property of interjections as interjectio ‘thrown in between’ and function as a ‘cry’ is shown through the word for ‘interjection’ across the Slavic languages. Russian междометия [meždometija] contains the word meždo ‘between’, Czech citoslovce contains the word cit ‘feeling/emotion’ and Polish wykrzyknik contains the word krzyk ‘cry’. This further contributes to the view that interjections have properties and functions we may apply to both languages under study.  34  1.4 Chapter conclusions In this chapter, I have introduced interjections by identifying the semantics, pragmatics and morphology of interjections across English and Polish with some attention to other Slavic languages. I presented the classification of interjections as primary emotive, secondary emotive, contact, cognitive, conative, and descriptive, and described terminological overlaps with other similar words such as DMs, exclamations, and swear words.  Some differences arise in the morphological potential and transfer of interjections between English and Polish. For example, Polish does not have interjection-based delocutive verbs, while English, though not to the same extent as Russian, sometimes converts interjections to verbs, e.g. she oohed and ahhed over the baby (see Brinton 2014). Polish has also borrowed interjections from English to fill a gap in the lexicon, namely the contact interjection sorry! and wow! An important feature arises during translation and finding equivalents, namely adequately capturing meaning when transferring between the two languages. There are some interjections that exist in PDE that do not (yet) exist in Polish (e.g. wow!), while interjections such as Polish no have more meanings and different nuances than the typical English equivalent well, or are simply not used in English (e.g. Jezus Maria! ‘Jesus Mary’) because of cultural and historical reasons.  I have also shown that English and Polish interjections are fundamentally similar; that is, scholars in both English and Polish have similar ideas about interjections – namely their classification, types, and syntax. English and Polish have interjections that can be considered either primary or secondary, which is usually determined by the use of the interjection in other parts of speech. Also, both have a wide range of interjections that fall into the subcategories of (primary) emotive interjections (e.g. oj! ‘oh’), cognitive interjections (e.g. aha! ‘aha’), conative interjections (e.g. sza! ‘shh’), contact interjections (e.g. dobranoc! ‘good night’), secondary emotive interjections (e.g. cholera! ‘damn’) and descriptive interjections (e.g. bum! ‘boom’). Finally, both English and Polish interjections may be suffixed, elongated, reduplicated, converted, or be altered in other ways, especially to make an interjection more expressive – and, in the case of Polish, to convey a greater degree of ‘warmth’ and affection. Rules of word formation are applicable to interjections in English and Polish, even if the individual interjections may differ in acceptability in adult language, certain contexts, and the like. 35  Thus, based on the above characteristics of interjections in English, Polish, Czech, Slovak and Russian languages, my definition of interjection for English and the Slavic languages is the following:  an interjection is a holophrastic, meaningful, and syntactically independent linguistic item which either expresses or displays the speaker’s current mental state, and may become polymorphemic by hosting derivational (i.e. diminutive, augmentative, and slang) morphemes. While outside the scope of this dissertation, it would be beneficial in future to compare this definition with unrelated languages such as Greek, Spanish, and non-Indo-European languages to establish a universal definition of interjection without Anglo bias. In this way, we can avoid a faulty assumption that some characteristics (e.g. hosting affixes) apply to all languages. In the following chapter, I focus on the various affixes that may be added to interjections. Therefore, the following chapter will expand on the central concept of affixation by providing a detailed background about the formation, meanings, and types of suffixes that are added to words (with a focus on interjections) to support my argument that the overall trajectory of each language, religious and historical influence, and contemporary views on expressing emotion play a vital role in determining which interjections across English and Polish are suffixed and when. 1.5 Organization of the thesis In this chapter, I introduced interjections in English and Polish; specifically, their definitions and types, their morphology (or lack thereof), and some comparison with the other Slavic languages of Czech, Slovak and Russian.  Chapter 2 provides background on diminutive, augmentative and slang suffixes in the English and Polish languages historically and in present-day language.  Chapter 3 lays out the methods of data collection used. It provides a description of ‘digital communication’, ‘literary dialogue’, established corpora and dictionaries that are the focus of this study. Thus, I discuss the three main types of language: the microblogging site Twitter, blogs and blog comments, and literary dialogue from fiction and fanfiction. I give an overview of each web and literary genre, and suggest limitations and restrictions of each. Other resources used in this study, including online corpora and dictionaries, are also described. It also introduces Appraisal theory and semantic and pragmatic features that are used in the following chapters. 36  Chapter 4 shows the suffixes that create affixed forms of interjections in English and Polish. It examines the meanings of affixed interjections in online slang dictionary entries along with their raw frequencies in established online corpora. The results of the analysis of dictionary meanings and feature analysis suggest that in both English and Polish, many affixed forms are considered ‘new’ items (although a few first appeared in the 19th century). Also, they are thought to function as ‘extreme’ forms of the base interjection which add additional features such as [+SILLY] and [+CHILD] in English. The following three chapters examine the functions and meanings of diminutive interjections based on Appraisal analysis (Martin and White 2007). I begin with positive themes and end with the most negative themes. Thus, in Chapter 5, “Admiration, astonishment, and delight”, I focus on the positive APPRECIATION and AFFECT of diminutive interjections, as found in praise, expressions of admiration, joy, and emphatic positive nuances. Chapter 6, “Sarcasm and play”, addresses the jocular, playful and non-serious meanings of diminutive interjections in the contexts of (non)-serious threats and intimidation shown through negative JUDGEMENT and raised FORCE and FOCUS, male-female gender roles that (playfully) reinforce the stereotype of the large male and the small female, and other ‘playful’ functions such as sarcasm and irony. Last, Chapter 7, “Pain, sympathy, and disgust” deals with the negative and serious aspects of affixed interjections that occur in contexts of physical and emotional pain, disgust, and revulsion. These are frequently expressed through negative AFFECT. While we would consider the offer of sympathy a positive gesture to make, I view it also as painful because a speaker is drawn into the other person’s pain when reaching out in solidarity to sympathize with another who is suffering.  Chapter 8 concludes the dissertation.   37   Background to evaluative and slang suffixes 2.1 Introduction In this chapter, evaluative (i.e. diminutive) suffixes are discussed in terms of their morphology, semantics-pragmatics, emotional connotations, and sound symbolism. Augmentative and slang suffixes that may attach to interjections are also briefly considered. This chapter begins by providing the definition of evaluative suffixes and their two main types (diminutive for ‘small’ and ‘little’, and augmentative for ‘big’ and ‘large’), followed by the emotional connotations, senses, and their interactional functions. It should be noted that while diminutives are the focus of the present study, they are linked to augmentatives by their semantics and pragmatics. Because of this, I include some background for augmentatives along with diminutives. The chapter continues with a discussion about the continued disagreement as to which English affixes are diminutive based on their productivity, integration and use in English, history, and diminutive senses. There is also some terminological confusion that contributes to the dispute about the inventory of diminutive suffixes used in slang, nursery talk, and so on. Section 2.3. looks at various English diminutive suffixes per Schneider (2003) with their definitions and brief history.  Evaluative suffixes are often conflated with slang suffixes that produce slang words typically found in the colloquial and stylistically-neutral register. Slang suffixes are widespread and more extensively-studied than the evaluative diminutive and augmentative suffixes. Many scholars tend to classify diminutive suffixes as ‘slang’ because they might appear to lack the typical senses of diminutive or augmentative suffixes, such as [+LITTLE] or [+CHILD]. In order to show how diminutive suffixes may be (mis)construed as slang suffixes, I briefly look at the known slang suffixes and then turn to the ‘diminutive-slang’ suffixes -o, -er(s), and -s in section 2.3. As diminutive and slang suffixes are prevalent in the Slavic languages, section 2.5 examines diminutive/augmentative suffixes in Polish, and the scope is extended to Czech, Slovak, and Russian diminutives in 2.6 to better situate Polish diminutives among the Slavic languages. Suffixed interjections have received some previous scholarly examination in the pilot studies (Lockyer 2014, 2015a, and 2017). In section 2.7, I present the main findings from these previous studies.  38  2.2 Defining diminutives and augmentatives There is a substantive body of literature about evaluative affixation (see, e.g. Grandi and Körtvélyessy 2015, Lieber 2015) - a term that refers to “a special subset of derivational morphology [that is, evaluative morphology] in which affixes are attached to bases to form new derivatives that convey a meaning of either size or emotion,39 in the form of diminutives or augmentatives” (Albair 2010: 1). Diminutives have received much scholarly attention across the languages of the world (see, e.g. Lockyer 2013, Schneider 2003, Dressler and Merlini Barbaresi 1994, Bratus 1969), though it is arguably useful to examine augmentatives to properly situate evaluative suffixation within a language or cross-linguistically. The OED provides some useful history and definitions of the terms ‘diminutive’ and ‘augmentative’. The term ‘diminutive’ entered English through French in the 14th century from the Latin past participle dī-, dēminuĕre ‘to lessen’ (OED). ‘Augmentative’40 was partly borrowed from Middle French in the 14th century from classical Latin augmentāre + -īvus ‘causing increase’ (OED). For ‘diminutive’, the definition of “[e]xpressing diminution; denoting something little: usually applied to derivatives or affixes expressing something small of the kind denoted by the primitive word” (OED) continues from its first use in the late 16th century, as does “[c]haracterized by diminution; hence, of less size or degree than the ordinary; small, little. In later use, generally, a more forcible expression for ‘small’: = minute, tiny. (Usually in reference to physical size.)” from the early 17th century (OED). The term ‘augmentative’ still fits the definitions from the 16th and 17th century, namely “[h]aving the property of increasing or adding to something; characterized by augmentation” and “[o]f an affix or derivative: that augments or increases in force the idea conveyed by a root word; esp. that indicates something large of its kind. Also, more generally: (of a word) that expresses augmentation or intensification of an idea” (OED). The traditional meanings of diminutives given by the OED are too restrictive in present-day English and Polish because they place diminutives and augmentatives as binary opposites that                                                  39 Emotion refers to the ‘evaluation’ or ‘attitude’ directed towards a referent or hearer. 40 The term ‘diminutive’ in Polish is zdrobnienie (< drobny ‘small’); the term ‘augmentative’ is zgrubienie (< gruby ‘thick’ or ‘fat’).  39  never converge (e.g. ‘to lessen’ and ‘causing increase’). Recent studies of the diminutive have shown much less predictability; that “[diminutives’] ample polysemy, often paradoxically contradictory, [has made] identifying stable connotations […] difficult” (Merlini Barbaresi 2015b: 35). In a similar way, there are too many pragmatic functions of diminutives, and differences between languages, to propose a ‘universal’ model that can adequately show diminutive functions in all situational and/or social contexts. Diminutives may be simplified to a certain extent. First, it is possible to describe them in terms of three-fold functions: denotational meanings (to specify properties of size and related extensions), emotional connotations (to show the speaker’s emotional shading of the situation), and interactional functions (to manage the effects of politeness and speech acts) (Ponsonnet 2018: 7). The denotational function has received ample study and diminutives have come to be seen as having denotative meanings which differ noticeably from their source words (Kryk-Kastovsky 2000: 167) without downgrading the importance of small size in noun diminutives and large size in noun augmentatives.41 Diminutive forms of interjections have yet to receive any such study. Diminutives typically convey positive emotional connotations such as endearment, tenderness, and affection; for example, He’s the cutest little doggie! But the emotional connotations are not restricted to positive meanings, as it is not far-fetched to use a diminutive in derogatory/ironic and angry ways, such as in Merlini Barbaresi’s (2015a: 1134) examples: He’s got a wife and a couple of wifies (‘girlfriends’) and Do me a teensy weensy little favour! Get out. Thus, although we typically presuppose diminutives to occur in positive (especially affectionate) contexts, we must examine the situational context to correctly identify the meanings and functions expressed. Ponsonnet (2018: 9) presents another inventory of connotations by dividing the emotional connotations of diminutives into two ‘levels’ of emotions: ‘mild’ emotions and ‘serious’ or ‘deep’ emotions. ‘Mild’ (child-related) emotions include fun (jocular), affection, endearment, approval                                                  41 Merlini Barbaresi (2015a: 1135) cautions against the downplay of denotative meanings “in favour of emotional values whose meanings and effects depend on the context, the participants’ attitudes and the type of speech act” in the study of diminutives. In this study, the ‘emotionalization’ of diminutives and augmentatives do not apply (much) to interjections, which primarily convey emotion and cannot be evaluated in terms of physical size. 40  (positive judgement), familiarity and proximity, and disapproval (negative judgement). The ‘deep’ (serious) emotions derive from the ‘mild’ ones: affection → compassion or romantic and/or sexually-oriented love; approval → admiration and respect, familiarity → comfort of familiar routines; and, disapproval42 → contempt, humility, and self-irony. It is not unreasonable, then, to suggest that diminutives in adult-to-adult speech may convey ‘deep’ emotional and psychological emotional connotations that have more layers to them than previously thought. Scholars of evaluative morphology have attempted to classify the ‘universal’ semantic and pragmatic meanings/features of diminutive categories in numerous ways. Regardless of the theory used or representation of these meanings, they generally include the following semantic and pragmatic features: [little], [child], [female] as the most frequent, followed by [intense], [dear], [flirt], [commiserate], [attenuate], [euphemism], [pejorative], and [irony] (Prieto 2005), along with [affection], [pets], [intimacy], [contempt], [hedges] and the semantic [imitation], [exactness], [partitive] and [approximation] (Jurafsky (1996). There has been a movement away from the idea that the central meaning or association of diminutives is the feature [child]; more recently, Prieto (2005) has aptly argued that [child] should be replaced with the feature [little]. This is a persuasive position because diminutives are not primarily used to or by children about physical size, but are often used to refer to ‘little’ things that evoke affection (e.g. [young] animals, [cute] objects and places). Also, the emotional distance between two people (e.g. between mother and child) may demonstrate the feature [little].43                                                   42 The connotation of disapproval is linked to the denotational semantics of ‘small’, for example in it’s only a short-term boom-let, where the diminutive conveys ‘unattractiveness’ (Merlini Barbaresi 2015b: 35) because a ‘short’ boom of increased economic or political interest, activity, or growth is less appealing or attractive than a steady, long-term boom. 43 In comparison, augmentatives convey emotional connotations typically associated with negative meanings because the increase in size intrudes on our personal space and causes feelings of hostility (Inchaurralde 1997: 139). This causes an ‘augmentative effect’ that includes “exaggeration and intensification” (Bakema and Geeraerts 2004: 1045), negative judgement, contempt, revulsion, and fear (Ponsonnet 2018: 24). These connotations tend to stem from the metaphor BIG IS UGLY - that is, ‘big’ lacks positive aesthetic qualities and thereby causes augmentatives to become associated with excess, disgust, revulsion, contempt, and vulgarity (Ponsonnet 2018: 25), to name a few. It is, however, possible for augmentatives to convey positive emotions, including admiration for high social status, endearment, and compassion (Ponsonnet 2018: 25), e.g. a huge success. Inchaurralde (1997: 139) explains that the object’s increase in size may also allow others to enter and thereby creates positive emotional connotations. 41  The meanings of diminutives typically are not based on empirical studies of diminutive interjections. Instead, the majority of studies focus on noun diminutives, followed by adjectives, and then adverbs and verbs. Arguably, nouns most frequently undergo diminutivization, and finding examples of the much rarer diminutive forms such as interjections is a difficult task, but is required to present a complete view of the potential meanings and functions of all diminutive forms without bias towards a POS. Based on the data presented in Chapter 4, these features (with the exception of Jurafsky’s strictly semantic meanings) may be applied to DIs and further elaborated on with associations such as [silly], [cute] and [naive]. Thus, the so-called ‘universal’ meanings require re-evaluation when applied to diminutive interjections and could be labelled with tags to indicate those which are specific to one or more languages.  Sometimes these emotional connotations resemble the affection found in pet names or the meaning of familiarity. Such approaches have developed terms such as ‘hypocoristic’, ‘affectionate’ and ‘familiarizing’ (suffixes) that replace ‘diminutive’. The term ‘hypocoristic’ is strongly associated with pet names and endearment terms used as vocatives. Because of the loosely and broadly-applied term ‘hypocoristic’ in scholarly studies, I use the definition supplied by the OED. This is: “[o]f the nature of a pet-name; pertaining to the habit of using endearing or euphemistic terms,” such as the pet name Patty for Patricia. It may convey affectionate connotations in pet names or more rarely be used sarcastically or as a put-down. In my view, the use of such terms undermines the diminutive category and causes undue confusion; thus, I do not use ‘hypocoristic’ in this dissertation except for truncated pet names. Interactional functions are “crucial to the understanding of diminutive use in conversation” (Schneider 2003: 118). These functions include, but are not limited to, the following: • [Diminutives] more often than not [make] an emotional comment on the relationship of the object and/or person mentioned and the speaker (Castro 2006: 6);  • Diminutives function as praise minimisers, which maximise, so to speak, the success of a compliment (Schneider 2003: 233); • Diminutives are used in requests and orders to mitigate the strictness of the speech act (Dabašinskienė and Voeikova 2015: 218); • [Diminutives] express the speaker’s wish to maintain or establish common ground and solidarity with the addressee (Dabašinskienė and Voeikova 2015: 223) 42  In sum, diminutives are typically ‘other-oriented’ (i.e. they seek to maintain positive relationships).  Diminutives may be formed synthetically (with an evaluative affix) or analytically (with an analytic marker) in different parts of speech, and with divergent emotional strength. On many occasions, speakers may be able to include both the analytic marker (e.g. little, small) and evaluative suffix in the same utterance; e.g. What a darling little doggie! and Jaki malutki domek! ‘What a little.DIM house.DIM!’. The more suffixes or analytical markers a diminutive has, the stronger its emotional meaning (Schneider 2003: 118). Of course, the POS determines the emotional strength as well, for diminutive adverbs, e.g. Polish prędziutko ‘quickly.DIM’ conveys the meaning of ‘very’ or ‘absolutely’ in an affectionate way. Although the meanings and functions of diminutives overlap to some extent in all languages, as I have shown, as a whole they are too complex and tied to society and culture to fit into one ‘universal’ mold, as some scholars (e.g. Jurafsky 1996) have attempted. Some functions may appear more often in one language than another, and some meanings may be more prevalent in one language than another (as shown by Wierzbicka 2003, Sifianou 1992) due to many factors, such as the attitudes towards emotion and cultural dimensions of rationality/emotionality (Wierzbicka 1997), uncertainty avoidance (Hofstede, Hofstede and Minkov 2010), and politeness strategies. However, such a cultural analysis is outside the scope of this study. Therefore, section 2.3. turns to various diminutive affixes in English, especially to the productive -y affix. In 2.5., the study points out key features of Polish diminutives. 2.3 Diminutive affixes in English Schneider (2003: 76-77) writes that “[t]here is no agreement in the literature regarding which affixes of English are diminutive suffixes and how many diminutive suffixes exist in the English language” because scholars adopt “different criteria in deciding which suffixes should be included” (Schneider 2003: 76-77).44                                                   44 Scholars agree that outdated suffixes from Latin and earlier English and French are no longer diminutive: the Latin -ul-, -ll-, -eolus, -idium, and -podicum, and the non-productive -ina, -en, -incel, -ol, -oon, -ot and -rel suffixes from Old French, Old English, Proto-Germanic and Middle English (cf. Schneider 2003). These may be easily eliminated from the inventory of PDE diminutive suffixes. 43  Wierzbicka (1985) supports the popular and restrictive view that, although creating childish diminutives, -ie is the main diminutive suffix in PDE. Dressler and Merlini Barbaresi (1994), in addition to -ie/-y, look at -ette, -let, and -s. Quirk et al. (1985) view -ette, -let and -ling as the only diminutive suffixes. Marchand (1969) identifies ten different diminutive suffixes, plus four variants. Bauer (2013: 392) claims that -o, -a, -er, -s and -kin are diminutive suffixes. Schneider (2003) provides a substantively longer list of eighteen diminutive suffixes plus combinations: -a, -chik, -een, -er, -et, -ette, -ie, -ing, -kin, -le, -let, -ling, -o, -ock, -poo(h), -pops, -s, and -sky. He acknowledges the potential for many second-degree forms that can be formed with -s, such as -kins, -ies, -ers, -sy, and so forth. Charleston (1960) includes 34 diminutive suffixes.45  The lack of agreement about diminutive suffixes in PDE stems from different criteria for inclusion in this category, such as: register (e.g. the claim that -ette used more often in business language or child language, cf. Schneider 2003: 77), polysemy (e.g. -ette for a female) (Schneider 2003: 92), productivity (whether the suffix creates new diminutives) and phonological distortion (i.e. truncation). Some consider the prominence of [+CHILD] conveyed by the suffix as the defining criterion of the diminutive or as grounds for exclusion from the term diminutive (e.g. Jurafsky 1996). Others claim that English diminutive use must match that in Polish or Spanish in terms of frequency. Overall, it appears that many scholars simply follow (older) English grammars and notions of cross-linguistic frequency and do not provide much – if any – definite criteria for their inventory of diminutive suffixes (Schneider 2003: 77).  Based on the differing views on diminutives in English, it appears that only -y, -let, and -ling can be claimed with any certainty to be diminutives (in the strictest sense) in PDE based on the denotational semantics of [+LITTLE] or [+SMALL]. However, if we view diminutives as an onomasiological category, following Schneider (2003), then a longer list of suffixes may be counted as diminutive or diminutive-slang suffixes. This study only considers y-suffixed interjections as ‘diminutive’ because interjections do not receive -let or -ling suffixes, and the additional diminutive suffixes posited by Schneider need to be critically evaluated in terms of                                                  45 There are also diminutive prefixes in English; for example, we can find the prefix mini- in miniskirt and minibus, nano- in nanobug, and micro- in micro-cosmos. However, these formations generally belong to technical language (Schneider 2003: 7) and are never affixed to interjections, e.g. *miniwhoops, *nanowow, or *microwhoops. See Schneider (2003) for a more comprehensive overview of diminutives in English. 44  their strong slang (e.g. in-group and colloquial) and hypocoristic (e.g. pet names in Rosiepops) associations before we add, for example, whoopser! or whoopso! as diminutive interjections.  In contrast to the lack of unanimity over the inventory of diminutive suffixes, scholars agree that English does not have any (prototypical) augmentative suffixes (cf. Schneider 2003: 16) as a rule. Augmentatives are typically formed using the adjectives big, large, and so on. However, it can be argued that there are augmentative-like suffixes in English, namely -rama (e.g. fawnorama), -zilla (e.g. bridezilla) and -ola (e.g. problemola) (Bauer, Lieber and Plag 2013: 411). The -ern suffix may be viewed as a non-productive augmentative suffix, as Bauer (2014: 4) comments that “the strange thing about -ern as a suffix is that it appears to mark an augmentative” such as in cavern, which the OED defines as “vaguer and more rhetorical [than cave], usually with associations of vastness.” Likewise, the -ard suffix (e.g. dullard ‘a very dull person’) was sometimes seen as an augmentative suffix (Duxbury 1884: 178). The -ern and -ard suffixes are attached to nouns only, while -rama, -zilla and -ola may be added to interjections.  This does not mean that augmentatives (and diminutives) cannot be formed in English. Analytic constructions (e.g. little house) are typically viewed as the “most common equivalent for [synthetic] diminutives in other languages” because little “suggests that the speaker harbours feelings of affection, sympathy, or pity for the thing or person in question” (Charleston 1960: 112), creates a term of endearment, or shows feelings of “friendly amusement” or of scorn (Charleston 1960: 113). Like the diminutive suffix, the marker little focuses more on emotion than size while the “non-emotional” marker of size is small (Charleston 1960: 112). Augmentatives are formed with the analytic markers big, large, and huge, among others before nouns. Analytic markers never precede interjections (e.g. *oh, little oops! or *big wow!) because analytic constructions can only be formed with nouns. 2.3.1 -Y (-IE) The -y suffix is “certainly the most productive diminutive category in English” (Bauer 2013: 392). Diminutives formed with -y most often include proper names, such as Bertie (< Albert) and Annie (< Anna), and nouns, such as beddie (< bed) and beastie (< beast), but can also include adjectives and discourse markers, such as comfy (< comfortable) and alrighty (< alright) (cf. Plag 2003: 120). Diminutives formed from kinship terms include auntie (< aunt), mommy (< mom) and hubby (< husband), regional memberships include Aussie (< Australian) and Newfie (< Newfoundlander), 45  and we infrequently find (lexicalized) diminutives from professions, e.g. goalie (< goalkeeper) and bookie (< bookmaker) (Schneider 2003: 89).  There is no widespread consensus as to whether -y should be considered a diminutive suffix. The suffix is sometimes termed a ‘familiarizing suffix’46 by Mattiello (2013b), who supports her choice of terminology with words including goodie (“an expression of delight”). Dressler and Merlini Barbaresi (2017: 507) cannot seem to decide on one term and use three terms in “the evaluative/familiarizing/nursery suffix -y/ie” likely because the -y suffix conveys evaluation towards a referent (e.g. sweet wifie), typically shows that the speaker and hearer know each other (e.g. in lover’s talk; sweetie, duckie), and in many cases, are used towards babies and small children (e.g. birdie, doggie). In comparison, Bauer (2013) and Schneider (2003) argue that -y is a productive diminutive suffix.  The origin of -y in the Middle English period helps little in the quest for a consensus because the suffix remains an “etymological mystery” (Shields 2001: 141). Sunden (2010: 162) hypothesizes that the suffix originated with Scottish personal names - but not as hypocoristics - rather, -y started as "a general onomastic suffix tending to increase its sphere of application according as weak final -e [with hypocoristic value] was dropped." Jespersen (1922: 402) uses sound symbolism as a lens to conclude that "the vowel [i], especially in its narrow or thin variety, is particularly appropriate to express what is small, weak, insignificant, or, on the other hand, refined or dainty.” (Shields 2001: 141) argues that -y originated with “caretaker speech,” a speech register used by adult speakers to help children understand their linguistic environment and learn new words. Thus, the first attestation of the -y suffix occurred with baby (< babe, first recorded in 1377) because the suffix was “simply a caretaker speech (or nursery speech) variant of babe, the older of the two variants” (Shields 2001: 141) that eventually became lexicalized into the standard form found in PDE.  Why, then, should the -y suffix be put in the diminutive category? Schneider (2003: 87) argues that that the -y suffix “cannot be referred to as a hypocoristic suffix and is, therefore, best defined as a diminutive suffix which usually indicates familiarity between speaker and hearer.” Moreover, it has been claimed that -y is the most common suffix used to create diminutive forms                                                  46 A ‘familiarizing’ marker is a suffix that has a highly informal tone and implies close, affectionate affiliation between speakers (cf. Schneider 2003: 80).  46  (Naciscione 2010: 135). But there are other aspects of the suffix which fulfill criteria for admission into the diminutive category: its associations with infants/children in the form of semantic associations such as silly and non-serious, its sound-symbolism, its associations with ‘female’, use in different varieties of English, its history as a diminutive suffix, and its relatively high frequency of use in words with emotional connotations.  Diminutives formed with the -y suffix are often simple, clipped, and reduplicated diminutives that most frequently appear in nursery language, child-directed speech, and language used to infants (baby-talk). In baby-talk, “rhymed lexical pairs [...] function as diminutives in the discourse such as: meany-weeny, tigey-wigey, and wormy-worm” (Varga 2012: 366). The repetition of syllables in the rhymed pairs could form hundreds of words (MacWhinney 2000: 48), including goody! to express delight and beddie(bye) for ‘go to sleep’. However, the rhyming scheme is not necessary for the creation of diminutives in child language as for example in the simplified consonant clusters in tummy (< stomach), milky, potty, shoesies (Fader 2009: 106), and the childish names for animals in birdie and horsie (Wierzbicka 1992: 384).  There is a dichotomy between the unsuffixed/suffixed -y form corresponding to adult/childish, serious/non-serious, foolish/sophisticated, and wise/foolish because children feature predominantly in the semantic associations of diminutives (Jurafsky 1996) rather than serious and sophisticated adults. Moreover, the y-affixed diminutive “conveys a childish effect” (Wierzbicka and Goddard 2008: 220). The y-diminutive is also viewed as ‘cutesy’; it is associated with subjective feelings of emotion (affect) alongside ‘cute’, ‘clever’ and ‘twee’ (OED, s.v. “cutesy”) that make a speaker of ‘cutesy language’ seem immature and unsophisticated. For this reason, using a diminutive in formal contexts would give the term depreciatory or comical meanings, and might weaken a speaker’s argument. In this way, there is an association between -y diminutives and adults speaking immaturely and irrationally, like small children. The high front/i/ vowel produced by the -y/-ie suffix has been claimed to be “symbolic of smallness” (Jespersen 1922) and often indicates that which is “small, slight, insignificant or weak” (Jespersen 1922: 283).47 The connection extends beyond the suffix, however, as Matisoff (1994:                                                  47 Sound symbolism also is important to Polish. Stolarski (2015) suggests a strong association between ‘smallness’ and /i/ in diminutive suffixes; that is, the /i/ diminutives are perceived as ‘smaller’ than /a/, for example. Stolarski concludes that “the theory of sound symbolism […] relevant to the Polish diminutive 47  124) suggests a strong “universal sound-symbolic connection between high-front vowels and notions of smallness.” The connection can be seen in words with ee, including wee, teeny, twee, peep, seed, peek (Crystal 2002), and also the slightly different sound in skimpy, flimsy, slim, slinky, spindly, piddling (Jespersen 1922: 283). Moreover, humans sometimes describe the sounds of small or baby animals by words with ee; for example, baby birds peep, cf. Jespersen (1922).48  So far, I have shown how the -y diminutive suffix has a strong association with children. Empirically, the use of the suffix in TV dramas such as Buffy the Vampire Slayer (see Adams 2003) suggests a departure from playground language. The -y diminutive suffix is also used differently in Australia and New Zealand,49 and is extremely frequent in Scottish English.50 A sociolinguistic survey of diminutives in different varieties and their influence on y-diminutive interjections, however, is outside the scope of this study.                                                  system”, but that “[s]uffixes containing the high front /i/ [are viewed] as “smaller” than suffixes containing the open central /a/” (114).  48 The notions of smallness do not apply to all words that end in -ee; for example, addressee, referee, and refugee do not have associations with smallness because here the suffix does not function as a diminutive. As such, we must not give too much importance to sound symbolism. 49 The -y suffix has become firmly established in Australian English (AusE) and New Zealand English (NZE). Wierzbicka (2003) and Peters (2009) include many curious formations: gummies ‘gumboots’ in NZE and sunnies ‘sunglasses’ which is iconic in the culture of Australian beach life (Peters 2009: 116). Peters (2009: 116) concludes that the momentum that keeps the -(e)y suffix alive and productive is Australian English. In contrast with British English and other Englishes, “in the antipodes its use as a hypocoristic is greatly extended into a general marker of familiarity and social belonging among adults [and] are characteristically colloquial and often ad hoc or situation-bound” (Peters 2009: 115). However, Wierzbicka (2003) is adamant that the -y suffix is not diminutive, but hypocoristic, in AusE because of its lack of the feature [+little]; instead, it conveys [+AFFECTION] that is implied through familiarity and hypocorism. Most words formed with the -y suffix are a form of slang that developed differently in NZE and AusE and cannot be considered diminutive. 50 Bratus (1969: 2) writes that “English dialects are richer in diminutives than standard English”; this is particularly true of Scottish English. In this variety, the -y diminutive suffix is not seen as fundamentally belonging to a child’s environment; instead, it “happens with a freedom and frequency which is startling to speakers of other [varieties] of English” (Dorian 1993: 134). The diminutives formed with -y in Scottish English include: bairnie, hillie, housie, knifie, laddie, lassie, lambie, ninnie and diminutives with double suffixes, such as mitherikie, bittikey, housikie, lasseckie or wifiekie, among many others (cf. Bratus 1969). Most, if not all, nouns may receive the diminutive suffix. Some view this extensive inventory of diminutives in Scottish English a remnant of the “large number of diminutives which [English] formerly possessed” (Mackay 1888: 300), though such a view is more hopeful speculation than proven fact. The suffix can also create other words besides diminutives; for example, riddie ‘a red face; a blush’, kiltie ‘a highlander’, cludgie ‘a toilet’, grippie ‘greedy’, and goskie ‘luxuriant’ (Jones 2002: 38). 48  The y-suffix appears in three different variants orthographically: -y, -ie, and -ey (Lappe 2007: 14, Merlini Barbaresi 1999, Marchand 1969); these are orthographic variants only. They are “one suffix” (Schneider 2003: 71) where the choice of orthographic variant used is merely caused by “preferences in different varieties of English” (Lappe 2007: 15). The -ey variant tends to be preferred in AusE and typically appears in words including matey (but not *matie, < mate), lovey (< love), and dovey (< dove). The -y and -ie orthographic variants do not seem to differ across varieties of English, as found in Charlie (< Charles), kitty, dolly, and doggie. I refer to these three orthographic variants as the -y suffix unless specifically referring to another variant. 2.4 Slang suffixes in English Affixes which create words labelled as ‘slang’ are group-related, (typically) short-lived, innovative, playful, and metaphorical words “that are below the level of stylistically neutral language” (Stenström, Andersen, Hasund 2002: 67, original emphasis). Slang endings change the register to informal and colloquial, and are typically invented and used by youth, ‘rejuvenate’ words that have become old-fashioned or stale (Akmajian et al. 2017: 204), and raise levels of playfulness and jocularity. From these characteristics, the term ‘slang’ here is meant to be viewed without pejorative connotations. A language requires new words to maintain a certain linguistic and emotive balance, which gives rise to slang forms. Intensifying adjectives (e.g. awful, fearful, terrible) “wear thin” (Charleston 1960: 114) and gradually lose their emotive impact. Moreover, “[i]n order to be vehicles of emotion, these words must also continually be changed […] and must be replaced by something stronger, racier, or more novel” (Charleston 1960: 115). Words from various parts of speech are also ‘recycled’ by the addition and replacement of suffixes to keep the emotional impact fresh and vivid.  In English, there are many well-documented suffixes that are socioculturally tied to various regions, particularly America, Britain, and Australia (Belladelli 2013: 215). An American English productive suffix is -eroo, and its variants -aroo, -roo and -oo, which convey “playfulness and jocularity” (Belladelli 2013: 215) in flopperoo (< flop) ‘a failure’ and pipperoo (< pip) ‘a particularly remarkable or pleasing person or thing’. American colloquial language has borrowed the Australian suffix -o, a so-called ‘depreciative’ hypocoristic of Australian and New Zealand in 49  words such as nutso.51 British English prefers the -er suffix (e.g. rugger < rugby) and -ers (e.g. nutters < nut).  Slang suffixes include the -aroo, -aroonee, and -io suffixes, the (borrowed) Slavic -ski suffix briefly outlined below: • -aroo (or: -eroo, -roo, -oo, and -amaroo): This suffix has been most notably studied by Wentworth (1942), who suggests that -eroo is a “neo-pseudo-suffix” with slang origins. By the term “neo” suffix, Wentworth likely refers to a new or modified form of a suffix because of the sudden influx of words suffixed with -eroo after 1939. The suffix is “pseudo” because of its “deceptive resemblance” to real suffixes that include -ness and -er (Wentworth 1942: 11) and because the suffix did not yet have a “firmly settled meaning” (Cassidy 1978: 51).52  These are thus “curious toy-words” that are most often found53 in “certain circles – notably radio, sports, advertising, and motion pictures” (Wentworth 1942: 10) for the purposes of “dramatic heightening” (Cassidy 1978: 51) and to create amusing, slangy variations of whatever is prefixed” (González 1995: 421). Words affixed with -aroo include checkeroo (a night-club check), crusheroo, kisseroo, to jokeroo, the adjective snoozemarooed, and the interjection jiggeroo, a dialectal word used among vagrants to warn each other “to be careful or escape” (Wentworth 1942: 10). • -aroonee:  This suffix may be a combination of -aroo + -(n)ee, and in modern English includes, e.g. switcharonee ‘a switch, swap’) that has been in use since 2006, if not earlier (Coleman 2012: 36). According to Quinon (2008), -aroonie (-aroo + diminutive -ie) has been used in the USA since the 1960s, e.g. smackaroonie.                                                   51 Several of these, particularly the -y, -o, -er, and -s are considered diminutive suffixes (e.g. Schneider 2003, Bauer, Lieber and Plag 2013), though often with a hypocoristic function. I take the view that while the -y suffix is diminutive due to sound symbolism; the other suffixes should be analyzed in context and by the basis of use and the base word. 52 Cassidy (1978: 51), considers the -aroo suffix to have been introduced “due to a twentieth-century perception of the cowboy”, but does not provide a strong case for this claim.  53 Peters (2009: 117) writes that the suffix is informal and Australian in origin, having appeared first in AusE in the late 19th century, “which seems to have had some early takeup in NZE and in spoken US English.” The various forms, likewise, are caused by “its equivocal origins, involving the convergence or coincidence or several morphemes” (Peters 2009: 117). 50  • -ski (or: -sky, -skie): This slang suffix includes brewski ‘beer’ and buttinski ‘one who butts in’, and is thought to be a “linguistic parody” (Akmajian et al. 2017: 287) of the Slavic languages, particularly Polish or Russian. In addition, the suffix “applies to the same classes of words as slang -s, leading to diminutive forms” (McCumber 2010: 127). Like the -s suffix, -ski makes the language more casual and informal (McCumber 2010). Sometimes the suffix combines with -s to create an -skie + -s ending.  • -za (alt. -sa): This suffix tends to be a popular slang suffix, especially in the slang interjection wowza. Used with personal names in AusE (Australian English), e.g. Barry > Bazza, Sharon > Shazza (Rojo 2009: 230). 2.4.1 Diminutive-slang suffixes -O, -ER(S), -S The -o, -er(s) and -s suffixes viewed as diminutive by Schneider (2003) are relatively controversial. They are sometimes viewed as slang suffixes or so-called embellished clippings because they are used between youth and adults or in informal register and/or are attached to clippings. I take the common position that they are likely not ‘true’ diminutives. -O The origin of the -o suffix is somewhat complex and is extensively discussed in the OED (updated 2004). According to the OED, this suffix can be attached to full and truncated words, and has three main and different origins, namely the following: i. as the final syllable of words of chiefly Romance origin;  ii. as the vowel that became final after the shortening of a word by dropping the syllables following a medial o, especially in compounds truncated after a prefix or combining form ending in -o;  iii. < ho int.1, O int., and oh int., occurring as a second element in various exclamatory phrases. (OED, accessed 22 September 2016) In the first origin, words such as camisado (< Spanish camisada) and lingo (< Portuguese lingua) appeared in the 16th century and became assimilated into English by the late 17th century. According to the OED, the second source, -o “first appears in the late 17th cent. and early 18th centuries” namely in words such as memo and hypo, which “probably established an association of the ending -o with causal or light-hearted use which it has retained ever since.” Thus, as with 51  the -eroo endings, these words are generally considered non-serious words and casual. In the third origin, the OED links the suffix to the call signal ‘oh!’ and thus has a strong reference to the interjections ho, oh and o. These attached to words and formed cries during late Middle English, e.g. heave ho, hey-ho, alive ho, and the expressions including righto, cheero from the later 19th century. Later, interjections such as rabbit-o, milko, whizzo, socko and the nouns kiddo and daddy-o appeared. After the 19th century, the suffix became attached to words from all parts of speech, creating ammo, arvo, whacko, and cheapo.54  -ER(S) The -er (and -er +-s form) suffix originated at Rugby School and Oxford in the late 1800s and remains productive as a “British upper-middle-class all-purpose suffix” (Ayto 2002: 364). According to the OED, the suffix originated “at University College, in Michaelmas Term, 1875; used to make jocular formations on ns., by clipping or curtailing them and adding -er to the remaining part, which is sometimes itself distorted.” Some of the earliest formations are associated with university life and sports; they include footer (< football, 1863), rugger (< rugby football, 1893), brekker (< breakfast, 1889) and ekker (< exercise, 1891) from the mid to late 1800s. Soccer was formed in this way from Association Football in 1891 (Mattiello 2013b: 81). Over time, these words lost their jocular meanings and are now “stylistically neutral” (Benson, Benson and Ilson 1986: 27). The -er + -s formation originated together with -er, and creates humorous or familiar nicknames, nouns, or adjectives. For example, words suffixed with -er + -s include preggers (< pregnant), starkers (< stark naked), shampers (< champagne) (Schneider 2003: 111),                                                  54 One of the stages in the development of -o not mentioned by the OED is its link to Australia and Ireland. In fact, there are two conflicting views on the origins of the suffix in Australia. The first view posits that the Irish brought the -o suffix to Australia, where it caught on and developed differently from Irish English (IrE) (Peters 2009: 117). The expanded use of -o did not catch on in New Zealand but was used in Dublin (Fritz 1996), and the Australian use likely was borrowed from IrE (Taylor 2001: 336). The second view stems from the observation that Australia was not influenced to the extent as Britain by the Spanish; thus, the so-called “calling-out” origin is another probable explanation. The “calling-out” origin posits that the “suffix originated from the early nominal uses of the cries of various street vendors. Thus, the milkman used to sing out “milk-oh!” and so became the milko” (Dalzell and Victor 2012: 560) and so on. However, the second view is rather simplistic, as it does not account for expanded uses of the -o suffix in AusE after WWII, namely the following: “as an abbreviatory device for familiar inanimate concepts” (e.g. susso, bizzo, rego); “a way of referring to eccentricity or social difference in others” (e.g. troppo, reffo, derro); and, thirdly, its “essential agentive role” (e.g. garbo, journo, muso) (Peters 2009: 117). As such, it can be safely argued that the suffix has strong historical roots in the Romance languages, and developed independently and/or from Irish in Australia. 52  AusE bathers (< bathing shorts), Honkers (< Hong Kong), Johnners (Brian Johnston), Aggers (Jonathan Agnew) and Athers (Atherton) as popularized on BBC’s Test Match Special of cricket (Rundell 2009). -S Langenfelt (1942: 210) claims that the suffix “originated in slang” and the OED claims that it is a “shortened form of the hypocoristic [diminutive] suffix -sy” seen in Babs, Toots; ducks and moms. In more current usage, -s is “becoming popular for creating informal, slang versions of common words and phrases” (McCumber 2010: 124), such as oh noes! (< oh no!) and slang terms in other parts of speech including totes (< totally), peeps (< people) and maybs (< maybe) (McCumber 2010: 124). However, the fact that the -s suffix serves many functions can cause some confusion as in English, the -s suffix can mark the following: the plural (e.g. cat-s, door-s), the third singular present tense of verbs (e.g. he serve-s), the formation of nouns from adjectives (e.g. acrobatic-s), and the possessive (e.g. George’s) (Lett 2009: 162). Some problems arise when the -s suffix appears in utterances where it is difficult to distinguish between the slang and plural marker, e.g. when used by itself in Wows! Does the speaker mean ‘many wows’ or a slang wow?  The slang -s suffix tends to be added to clipped base forms (e.g. turps < turpentine or Becks < David Beckham). The -s suffixed interjections change the register without any distortion, conversion, or clipping of the base form; hells no/yeah (< hell no/yeah), (oh) noes (< (oh) no), okays (< okay), and (oh) wows (< (oh) wow!) (McCumber 2010). There is also a spelling variant, -z, which appears in, e.g. lolz, and (oh) noez. Likely, these words are not clipped because of the short base (but see Trast 2010 for a comprehensive overview). In sum, there is a lack of consensus as to which suffixes in English are diminutive, and as such there are several suffixes that seem to straddle the diminutive-slang divide. That is, they create ‘toy-words’ that are mainly jocular, informal, and function to increase dramatic connotations. This study does not attempt to solve this long-standing issue, but rather takes the common position that -y, -let, and -ling are diminutives, while the rest in Schneider’s (2003) list are potentially diminutive and/or slang. The most frequent affixes that appear with interjections (see Chapter 4) have been discussed: -y, -aroo(nee), -io, -sky, -za, -o, -er(s), -s, and the pseudo-53  augmentative suffixes -ola, -rama, and -zilla. Next, the discussion turns to Polish diminutive affixes in 2.5. before turning to other Slavic languages in 2.6.  2.5 Diminutive affixes in Polish  Diminutive affixes in Polish (and most, if not all, of the Slavic languages) are particularly productive and have been studied extensively (see, e.g. Biały 2015; Lockyer 2013; Gorzycka 2012; Szymanek 2010; Kreja 1969). There is agreement about which affixes are diminutives and augmentatives in Polish. Diminutive derivation is so frequent and productive in Polish that any new formation is accepted “as a correct, acceptable word in standard adult Polish” (Hamas 2003: 43).  Polish diminutives are put into two groups: diminutives proper (Pol. deminutywa właściwe) and diminutives formal (Pol. deminutywa formalne). Diminutives proper include those from nouns that mainly signify small, e.g. domek < dom ‘house’. Diminutives formal fall into three main subgroups. They are, from Długosz (2009: 274): • Diminutives where there are some commonalities between the basic expression and the derivative, as in shape or function. E.g. dzwonek ‘bell’ < dzwon ‘gong’. • Diminutives which carry the meanings of singularity, attenuation, a small amount of the object, or part of a whole (Pol. syngulatywności and partytywności), which may be simply a single element of the concept encoded in the basic expression, or its part; e.g. ciastko ‘cookie’ < ciasto ‘dough’ or ‘pastry’. • Emphatic diminutives, which are divided into three subtypes: ○ those fixed in terms of size, e.g. słonko < słonce ‘sun’ ○ abstract nouns, e.g. idejka < idea ‘idea’ and ambicyjka < ambicja ‘ambition’ ○ names of young things, e.g. prosiątko < prosię ‘pig’ Across most Slavic languages, the -k- affix forms diminutives mainly from nouns; e.g. the grammatically masculine Polish domek (< dom ‘house’), Czech domek (< dům ‘house’), Ukrainian dubok (< dub ‘oak’), Russian domik (< dom ‘house’), Kashubian grzibk (< grzib ‘mushroom’) and Slovene čajček (< čaj 'tea'). The -k- affix changes for the masculine, feminine and neuter genders 54  as -ek, -ka and -ko respectively.55 Feminine and neuter diminutives include Polish główka (< głowa ‘head’), drzewko (< drzewo ‘tree’) and masełko (< masło ‘butter’). Diminutives can also be formed from adjectival diminutives (e.g. bialutki < biały ‘white’), adverbs (szybciuśko < szybko ‘quickly’), and the rare verb (e.g. płakusiać < płakać ‘to cry’). In non-noun words, however, the -k- diminutive affix tends to be replaced with other related diminutive affixes.  A diminutive formed by the -k- affix may receive more diminutive suffixes to change the ‘degree’ of the diminutive. Degrees of diminutives are formed as follows: “(basic) NOUN → DIM1 ‘small NOUN’ → DIM2 ‘small DIM1’ → DIM3 ‘small DIM2’” (Manova and Winternitz 2011: 116). Each degree adds on to the diminutive from the previous step and therefore adds emotional meanings.56 In the formation of DIM2 and DIM3, Polish has the masculine second-degree diminutive suffix -(ecz)ek, -(icz)ek, -uszek; feminine -(ecz)ka, -(icz)ka, uszka; and, neuter double diminutive -(ecz)ko, -(icz)ko, uszko. Second-degree diminutives include koteczek ‘dear little cat’ (< kotek < kot), buciczek ‘dear little boot’ (< bucik < but), and wierszyczek ‘dear little poem’ (< wierszyk < wiersz) (Kreja 1969: 85-87). Diminutives do not have to end with the third degree, though, a fourth-degree (and higher) diminutive is rare and would not be accepted as adult language. The suffix alone does not necessarily indicate the degree of the diminutive or diminutive intensity of an utterance. Many diminutives have a long history where the base word evolved into different words or became lexicalized. We might identify króliczek ‘dear little rabbit’ as a double diminutive because of the -iczek suffix, but the base is królik ‘rabbit’, not król ‘king’. The base makes króliczek a first-degree diminutive (Kreja 1969: 86). A similar process may be seen in sałata ‘lettuce’ and sałatka ‘salad’’; or, chusta ‘a head-kerchief usually worn by old women’, chustka ‘a scarf fashionably worn by younger women’ and chusteczka ‘facial tissue’. Many first-degree diminutives are now the standard form as in jajko ‘egg’ < jajo, which now functions as an augmentative57 form (a ‘big egg’), and jajeczko is now the first-degree diminutive form.                                                   55 However, the differences in the South Slavic languages of Bosnian, Croatian and Serbian are more obvious, e.g. nosić (< nos ‘nose’) (cf. nosek in Polish). 56 The choice of suffix within each grammatical gender tends to depend on phonological rules of diminutive affixation (see Manova and Winternitz 2011). Of course, the endings also change according to the case of the base word. 57 Augmentative suffixes are well-attested in Polish and are used to produce opposite meanings than diminutives, namely [+LARGE] rather than [+SMALL]. Polish augmentative suffixes include -icho, -ucho, -55  Personal names may have up to “ten different derivatives, all commonly used with respect to the same person, each of them implying a slightly different emotional attitude, and ‘emotional mood’” (Wierzbicka 2003: 51).5859 The diminutive suffixes that tend to occur with names are -ątko, -ulka, -uchna, -ycha, -ychna, -ś, and -usia. Wierzbicka (1992)60 provides extensive analysis of the specific emotional colourings for each suffix; for example, she argues that the -eńka suffix in e.g. Marysieńka implies that the speaker thinks of the person named as ‘someone small’, while the -ulka suffix in e.g. Marysiulka suggests ‘something small’. As such, only the former can be used to both children and adults. Not every personal name may be suffixed to the extent of Anna or Maria. Anna may have ten derivatives, but Agata and Ewa have fewer diminutive versions partially due to phonological rules of affixation, and partially due to whether the full first name                                                  ucha, -(i)sko, and -ensje (Piekot 2008: 104) to create, for example, kocisko (< kot ‘cat’) and konisko (< koń ‘horse’). In addition, Dziwirek and Lewandowska-Tomaszczyk (2010: 30), include 20 different augmentative suffixes, including (but not limited to): -al (brzydal ‘an ugly person’), -ara (nudziara ‘boring woman’), -isko/-ysko (dziadisko ‘unpleasant old man’), -sko (babsko, ‘unpleasant, rude woman’), -uch (leniuch ‘lazy-bones’), -ucha (starucha ‘unpleasant, rude woman) and -ula (brzydula ‘ugly woman’).  Augmentatives in Polish tend to be similar to the English analytic expressions ‘big X’ (size) and/or ‘old X’ (affection), but are difficult to translate from Polish to English. English translations of Polish augmentatives through adjectives such as unpleasant, rude, ugly, boring, and lazy show the metaphor big is bad. Wierzbicka (1985: 125) notices these associations with her meaning of “I can say that I think of you as of someone big and that I feel bad feelings towards you as one does towards something big.” Yet augmentatives may also convey the positive associations of maturity, stature, magnificence, softness, and wisdom typically attributed to ‘old’ referents (cf. Lockyer 2015b) such as szare nosisko ‘grey nose.AUG’.   58 Several diminutive suffixes are used more often with proper names and relationship terms; these include -cia, -sia, -zia, -unia, -la, -ina, and -ątko. The -ątko suffix exists in the realm [+YOUNG] people as in dzieciątko ‘little child’ (< dziecię ‘child’) or [+YOUNG] animals as in kociątko ‘little kitten’ (< kocię) (Kreja 1969: 99-100). 59 Perhaps one of the familial contexts that tends to enforce the idea that diminutives are vastly more extensive in Polish is the possible derivations in Polish for matka (mother) compared to English. In English, only one suffixed form exists – mommy – and it is used by children. In contrast, in Polish, the number amounts to 47: mateczka, matunia, matuńka, matuńcia, matunieczka, mateńka, matuchna, matusia, matuś, matuśka, matusieńka, matusina, matusiczka, matusieczka, matusienieczka, matyńka, matyneczka, matuchniczka, matuniczka, matula, matulka, matuleńka, matulejka, matejka, matulina, matulinka, matuleczka, matuluś, mamcia, mamka, mamusia, mamuś, mamuśka, mamusieńka, mameczka, mamunia, mamuchna, mameńka, mamunieczka, mamusieczka, mamiczka, mamula, mamulka, mamuleńka, mamunieńka, mamusińka, mamuliczka. (See Handke 2008: 105). However, the difference is not always as extensive. 60 See also Grabias (1981) for a comprehensive survey of Polish diminutive suffixes. 56  is marked or unmarked.61 By the time a person becomes an adult, she tends to go by a certain form (e.g. Tosia instead of Tereska < Teresa), and the other diminutive forms are for special occasions where the speaker feels particularly affectionate, tender, or the like. I have already established the overall emotional connotations and pragmatic functions of diminutives and augmentatives in 2.1. The Polish diminutive system tends to amplify those emotional connotations and interactive functions as Polish diminutives primarily function to “personalize the human interaction and make it affectionate” (Hřebíčková et al. 2002: 71) and to add ‘warmth’ and ‘emotional spontaneity’ to linguistic expressions (cf. Wierzbicka 2003). Slang suffixes may do the same for youth and in-group affiliation, but I do not discuss them here.62 Next, I turn to diminutives in the neighbouring Slavic languages of Czech and Slovak (also West Slavic languages like Polish) and the East Slavic language of Russian. 2.6 Czech, Slovak, and Russian diminutives The other two West Slavic languages, Czech and Slovak, and the East Slavic language of Russian and have numerous ways of creating diminutives (see Bratus 1969 for Russian and Gregová 2015 for Slovak). All Slavic languages are rich in evaluative suffixes, particularly diminutive and augmentative suffixes (Panocová 2011: 176) that are affixed to words from various parts of speech. Diminutives in the Slavic languages express ‘endearment’ and ‘smallness’ and are typically formed by the -k- diminutive affix.                                                  61 Full first names may be marked or unmarked. Those with marked names, such as Jan, Jerzy and Zofia, are more likely to be referred to by a first-degree diminutive form, e.g. Janek, Jurek and Zosia in normal adult-adult interactions. Those with unmarked names, such as Adam, Michał and Andrzej are more likely to be typically referred to by their full first names because the diminutive versions, Adamek, Michałek and Andrzejek convey affectionate and childish connotations (Wierzbicka 1997: 270-279). 62 Since the fall of the Soviet Union in 1989, English has contributed many ad-hoc neologisms and calques into non-standard Polish, including dens ‘dance’, totalnie ‘totally’ and most notably the contact interjection sory/ sorki ‘sorry’. Clipping typically forms Polish slang, e.g. impre < impreza ‘party’, spoko < spokojnie ‘no panic!’ (which can also mean ‘you’re welcome’), and komp < komputer ‘computer’ (cf. Fellerer 2014: 171). Also in effect is the depreciating slang suffix -ol, found in głupol ‘a stupid person’ (< głupi ‘stupid’) and psychol ‘a mentally ill or unstable person’ (Kucelman 2015: 32). The slang and hypocoristic affixes -ki and -ś are thought to create forms including koleś < kolega ‘matey, friend’ and sorki ‘sorry/sorries’, though they may be viewed as diminutive suffixes.   57  Czech exhibits derivational processes such as suffixation, in e.g. onomatopoeic bums! (< bum ‘boom’), the contact interjection čaues! (< čau! ‘hi/bye’) and the secondary interjection jémináčku! (< jemine! ‘gosh’) (Krzempek 2015: 122), and “diminutives are most often formed from interjections of contact (greetings), followed by emotional […] and volitive (encouraged by exhortation)” (Káňa 2016: 289, 292, my trans.). The suffixes that can be added to interjections are: -ek: nazdárek! (< nazdár! ‘hi’/’bye’), -ček: páček! (< pá! ‘bye’), -ík: ahojík! (< ahoj! ‘hi’/’bye’), -ka: ahojka! (< ahoj! ‘hi’/’bye’),63 -íčko: (na) zdravíčko! (< (na) zdraví! ‘cheers’), -ičky: pacičky! (< pac! ‘clap’), -ky: ahojky! (< ahoj! ‘hi’/’bye’), čauky! (< čau! ‘hi’/’bye’), fujky! (< fuj! ‘yuck’), jupínky! (jupí! ‘hurray’), mňamky! (< mňam! ‘yum’), -ku: jéminku! (< jémine! ‘golly’), -í/anku: božínku! (< božé! ‘God’), jeminánku! / jéminánku! (< jémine! ‘golly’) and -ity. There is likely little difference between the various diminutive forms of a certain word except for emotional ‘flavoring’ and personal preference.  Slovak, like Czech, frequently forms diminutives that are considered colloquial and/or acceptable in adult speech. Although Bóhmerova (2011: 76) suggests that diminutivization in Slovak “is systematically and communicatively extremely marginal,” Slovak diminutives can be found across many parts of speech (cf. Kačmárová 2010, Štolc 1958). Bóhmerova lists the “colloquial diminutives” from contact interjections ahojček! and ahojko! ‘hi’/’bye’, čauko! ‘hi’/’bye’ and dobré ránko! ‘good morning’ as Slovak suffixed diminutive interjections,64 though the latter ránko ‘morning’ is a substantival diminutive. Slovak has verbal diminutives that also follow a linear structure, e.g. driemkat’ (< driemat’ ‘to doze’) “where driem- is a root morpheme, -a- is the so-called thematic morph, -t’ is a form morpheme and the segment -k- is a diminutive suffix” (Gregová 2015: 296). The -k- affix is so linguistically embedded in the Slovak speaker’s mind that Slovak learners of English may say, Look, snailik here? instead of Look, is there a little snail here? (Bobčáková 2017: 154). The ‘error’ made by the Slovak learner of English suggests that Slovak diminutives are not marginal to the language.                                                  63 The multiple diminutive versions of the base contact interjection ahoj ‘hi’/’bye’ tend to be found in between intimate and jovial contact (Káňa 2016: 292). 64 According to Schneider (2003: 214), diminutive greetings are possible in German, e.g. hallöchen (< hallo ‘hello’), tschüs(s)chen (< tschüs(s) ‘goodbye’) and the Swabian farewell of adele (< ade). They appear in other languages: Italian salutini (< saluti ‘greetings’), Slovene pozdravček (< pozdravi ‘greetings’) and Lithuanian labukas (< labas ‘hello’) and labanaktukas (< labanaktis ‘goodnight’). 58  Russian diminutives have been studied extensively (e.g. Lockyer 2013, Volek 1987, Bratus 1969), and it has been found that diminutive suffixes can be added to several primary emotive interjection bases (with processes including reduplication) to greatly raise levels of emotion. Diminutive interjections that may be found in Russian include the following double diminutives: ойойошеньки! [oyoyoshen’ki!] ‘oy.oy.DIM’, oхохонюшки! [oxoxoniushki!] ‘oh.oh.DIM’, oхохошеньки! [oxoxoshen’ki!] ‘oh.oh.DIM’ (Dufková 2011: 168). Doludenko (2012: 22), in her study of the social media site “VKontakte” (similar to Facebook, but in Russian), found that the site “is notable by the high number of expressive suffixes – a tendency typical for modern colloquial Russian” across all parts of speech. In addition, translations from English tend to add diminutives, especially in works of children’s literature that is comparable to similar translations into Polish. In sum, diminutive interjections differ from (Slavic) language to language: Slovak and Czech have a high frequency of diminutive contact interjections with various suffixes, while Russian boasts of reduplicated oj ‘oh’ interjections with first and second-degree diminutive suffixes. Likewise, Polish speakers are also different as to which interjections become diminutive forms, along with their meanings and functions, as briefly addressed in the pilot studies discussed in section 2.7 below.  2.7 Pilot studies Considering the formal and functional properties described above, the question remains: how do these various diminutive affixes function when applied to interjections? To answer this question, I completed several pilot studies (2014, 2015, 2017) in anticipation of the larger study that follows.  My 2014 study briefly examines the existence and core meanings of DIs in English from an analysis of Twitter posts, while my 2015 and 2017 studies turn to Polish (o)jejku!, (o)jejciu! and affixed forms of the expletive Jezu(s)! ‘Jesus’, e.g. Jezuniu!, Jezusiczku! All the studies find similar evidence: DIs are multifunctional and emotionally-varied items that make the emotional colouring of the base interjection and/or utterance more vivid. They may lessen the emotional impact and soften any ‘blow’, or intensify the emotional meaning of the base. They convey emotions across the spectrum from positive to negative. They tend to convey sympathy and support, exaggeration and joy or excitement, the concept of ‘children’ or ‘childish’, affection 59  (love and appreciation), appreciation and gratitude, feelings for pets and young animals, and embarrassment due to a mistake. The affixed interjection can also be more lighthearted and a reaction to something less serious in some contexts, while ‘deeper’ in emotional loading in others; the DI “comes from a deeper place emotionally” (Lockyer 2015a: 215) than the base.  My studies have found that diminutives, as markers of affection and other aspects of emotivity, are particularly compatible with interjections to form synthetic diminutive interjections. Emotive interjections are ideal candidates for diminutive (and slang) suffixation for the following reasons: • They express emotive meaning(s) ranging from positive (appreciative) to negative (depreciative).  • They are linguistic features that can express the mental state of a speaker, a speaker’s attitude towards an action or referent, or a speaker’s reaction to a situation.  • They are generally used in informal and casual speech and are commonly avoided in formal speech (perhaps with the exception of lexicalized diminutives including droplet and some primary interjections including oh).    (Lockyer 2014: 71) Following Dressler and Merlini Barbaresi’s (1994) ‘complex’ and ‘simplex’ forms, I argue that diminutive interjections - and, indeed, all diminutives - are ‘complex’ phenomena in comparison with their non-diminutive ‘simplex’ counterparts because of a greater psychological processing. That is, base interjections are ‘simplex’ because they “are the shortest phonologically, can be more quickly exclaimed, and have less emotional depth” (Lockyer 2015a: 204). In comparison, affixed interjections are ‘complex’ because “they consist of more elements which add additional speaker attitude” (Lockyer 2015a: 204), are longer phonologically and take longer to exclaim. The ‘size’ of an interjection (i.e. the number of morphemes) is "inversely proportional to the amount of information that is recoverable from context” (Stange 2009: 46). To show this concept, I developed the following formula:  INTJ (e.g. jej!, ojej!) = ‘emotional reaction’  INTJ ‘emotional response’ + -k- > INTJ ‘emotional reaction that conveys [+LITTLE]  (deeper/‘softer’) emotional reaction (e.g. jejku!) (Lockyer 2015a: 204). The rest of this dissertation takes the results found in my small-scale pilot studies and applies them to larger corpora and other forms of written communication. It uses a different theoretical framework – Appraisal theory (Martin and White 2007) – to fill in the gap left by the pilot studies 60  and previous research on diminutives and interjections. Namely, the dissertation shows which interjections may receive suffixes and how frequently in Chapter 4, and qualitatively analyzes positive, playful and negative y-affixed diminutive interjections across written corpora in Chapters 5 through 7. 2.8 Chapter conclusions In this chapter, I introduced diminutives in English and Polish by their formal and functional aspects. For the English language, I presented the extremely productive -y diminutive suffix and the various other suffixes that often overlap with slang, including -o, -er(s), and -s. For Polish, I gave a background of the diminutive suffixation, most notably through the -k- diminutive affix that produces endearing and positive words which help build positive relationships among individuals in both written and oral communication. Augmentatives, which are the “opposites” of diminutives (Schneider 2003) also were briefly mentioned as words meaning [+BIG] instead of [+SMALL/LITTLE].  At the word-formation level, phonological rules and available suffixes largely determine which base words can be suffixed, and with which suffixes. Polish is easy to describe in this regard because the language contains a vast array of affixes that can be added onto each other to create first-, second-, or third-degree diminutives. While there are lexicalized diminutives, and some words do not accommodate certain suffixes, these issues are typically minimal. Polish speakers have many opportunities to choose a certain diminutive suffix to convey emotive colouring, especially in the formation of diminutives from personal names. In some communication, the use of a diminutive is a conventionalized cultural norm. This freedom in diminutive formation is shared among Czech, Slovak, and Russian speakers as well.  The chapter showed that there is disagreement about the inventory of diminutive suffixes in English. The term ‘diminutive’ has associations that differ amongst English speakers who typically form diminutives analytically using little or small. Some scholars use the term ‘diminutive’ suffix loosely and with criteria that allows for many suffixes to be considered diminutive (e.g. Schneider 2003); others have a restricted set of criteria and allow for very few diminutive suffixes in English. The -y suffix, despite being extremely productive in forming new words used to children, by children, and among adults, has received the most criticism as a potential diminutive suffix because of its frequent use to and by children. In this dissertation, I 61  follow Schneider (2003) in my interpretation of -y as a true diminutive suffix that conveys [+SMALL/LITTLE]. The other suffixes that have been suggested to function as diminutives, namely -o, -er(s) and -s do not appear particularly affectionate or endearing; rather, they often merely add splashes of colour to an utterance.  The following chapter presents the methods used for the study and its theoretical background. The datasets that were mined for y-affixed diminutive interjections are also introduced.   62   Research methods and texts 3.1 Purpose of the present case study The present case study will consider the use of affixed interjections as they are used by adults and teenagers within two different written text types: online and literary discourse. It first identifies all the affixed interjections in use in reference corpora, then turns to focus specifically on diminutive interjections in Twitter, blogs, fanfiction, and fiction digitized in Google Books. The major goal of the study is to show how, when, and why diminutive interjections are used in fiction and in written online communication.  The purpose of the present case study is to add to the findings in Lockyer (2014, 2015a, and 2017). These pilot studies began to combine research on diminutives and interjections to suggest meanings and emotional connotations for diminutive interjections. Using Twitter and reference corpora, the pilot studies investigated the English diminutive interjections wowee!, whoopsie!, oopsie!, ouchies!, and Polish (o)jejku!, (o)jejciu!, jezusku! and other diminutive forms of Jezu(s)! While the small Twitter corpus and the national Polish corpus were only able to provide enough data for the main diminutive interjections, the pilot studies made it possible to suggest some basic (emotional) semantic and pragmatic meanings of these linguistic items, including ‘children’, ‘sympathy’, and ‘(augmented) excitement’.  The first part of the present study (Chapter 4) combines qualitative and quantitative analysis for all affixed interjections in English and Polish; that is, those interjections with diminutive, augmentative, and slang affixes. Qualitatively, it sets out to investigate which interjections receive affixes, how they are defined by users in online dictionaries, and which semantic, pragmatic and registerial features can be extrapolated from user definitions. Quantitatively, the chapter presents raw and normalized frequencies of affixed and unaffixed forms from several reference corpora. These results form the basis of the three chapters that follow.  The second part of the present study applies the semantic system of Appraisal to only the y-diminutive interjections in English and -k- affixed diminutive interjections in Polish. It will attempt to extend the semantic and pragmatic analysis beyond the few emotional meanings identified in the small-scale pilot studies. In order to understand and explain emotions, emotional evaluations are polarized and divided into ‘positive’ and ‘negative’ (Ben-Ze'ev 2000). Those emotions that are humorous (e.g. sarcasm) can be considered playful evaluation. 63  Therefore, the chapters are divided into three main groups: positive (Chapter 5), playful (Chapter 6), and negative (Chapter 7). These chapters will focus on situational and contextual factors to review the lexical Appraisal resources used with diminutive interjections. 3.2 Research questions The present study presents the first comprehensive qualitative and corpus-based study of affixed interjections in English and Polish. The study is qualitative because the aim is to provide a complete, detailed description of the linguistic items, and is corpus-based because it “assumes the validity of linguistic structures derived from linguistic theory; the primary research goal is to analyze the systematic patterns of use for those pre-defined linguistic features” (Biber 2009: 276). In this study, the linguistic features are affixed interjections, which refers to those interjections that have any expressive affix, e.g. diminutive, augmentative, and slang. In other words, the term includes all interjections with endings. The term diminutive interjection, then, refers to interjections with diminutive affixes. This is the focus of Chapters 5 to 7.  Written language differs from spoken language in various ways, and digital communication tends to fall in-between the two (see Biber 1988, Crystal 2006). The findings of the present study are limited to written language, notably internet language and adult fiction. Online spoken interactions through, e.g. Skype and YouTube are excluded. Children’s literature is likewise excluded. Thus, I do not claim that the results from this study apply to spoken language, though to a certain extent they may inform which affixed interjections may be used, and some of their corresponding functions.  1a. With which affixes are affixed interjections typically formed? The first part of this study relates to the variety of expressive affixes available in English and Polish that are added to interjections. Before we proceed to the analysis of diminutive interjections, we must consider the broad palette of all potential affixed interjections. To answer this question, prior research of English and Polish diminutive, augmentative and slang affixes informed potential affixed interjections, and the search function in online slang dictionaries, reference corpora, and the micro-blogging site Twitter show the variety of affixed interjections. 1b. Why are these affixes used, and how can online slang dictionaries inform these choices? 64  Since there are many affixes available to an English or Polish speaker, there is a point where the writer of a text decides to choose one affix instead of another, or decides to use an affix instead of the unaffixed base form. Since affixed interjections rarely appear in standard dictionaries, user definitions in online slang dictionaries tentatively answer the research question by decomposing affixed interjections into semantic and pragmatic features (see 3.5).  2. Which fundamental semantic, pragmatic and registerial features underlie the formation of affixed interjections? Interjections and diminutives convey semantic and pragmatic features and emotional attitudes. They are also often linked to register, particularly slang-affixed interjections. This focal research question aims at investigating the meanings and attitudes conveyed by diminutive interjections based on previous scholarly work on the emotional connotations of diminutives (e.g. Ponsonnet 2018, Biały 2015, Jurafsky 1996). It also adds to the pilot studies of diminutive interjections, especially those in Polish, which have suggested the key semantic and pragmatic functions of: – conveying sympathy and giving support (e.g. ‘I feel for you; I  feel grief-stricken, sorrowful, etc.’); – exaggerating and intensifying an emotion or reaction; – conveying the sense of ‘cute’ or ‘infantile;’ – showing affection (e.g. ‘I like/love you’; ‘I appreciate and am grateful for you’); – reacting to a mistake (e.g. ‘I am embarrassed’). (Lockyer 2015: 209) These may be expressed more simply as the features [+COMMISERATE], [+INTENSE], [+CUTE], [+CHILD], [+DEAR], and [+MISTAKE]. However, other semantic and pragmatic features that have been previously identified with either diminutives or interjections should apply to affixed interjections as well and are identified in Chapter 4.  3. What can Appraisal reveal about the evaluations expressed through diminutive interjections in digital communication? The discourse semantic systems of Appraisal (and, to an extent the related system of Involvement) consist of resources that provide speakers with ways to express attitudes and emotions towards each other and the world (Eggins and Slade 1997: 116), especially through lexical choices. Following Eggins and Slade (1997: 116), I suggest that the expression of attitude in contexts with diminutive interjections “is an important device for constructing and signaling 65  degrees of solidarity and intimacy in relationships.” It allows for the discussion of playful devices, including teasing, swearing, and using nicknames/endearments. It is suggested that these devices, which lessen (emotional) distance and serve a diminutive function, are particularly relevant in using diminutive interjections. Indeed, diminutive interjections are likely by themselves important lexical choices that fit under Appraisal (see 3.3.2). The internet is a focal resource in this study because it presents a kaleidoscope of communicative, lexical, syntactic, and pragmatic uses of language and interacts with trending ideas and socio-political and cultural discourse. When linguists began examining computer-mediated communication (CMC) in the 1980s, a strong interest developed to identify the characteristics of its “micro-linguistic structural features” (Bieswanger 2013: 463) that have now broadened to include interjections and diminutives. While literary dialogue and (casual) communication are the most-frequently analyzed modes of communication when studying diminutives (e.g. Schneider and Strubel-Burgdorf 2011 about the diminutive suffix -let) and for enacting a text analysis of Appraisal (Martin and White 2005), this study looks at literary dialogue and its online counterpart, fanfiction, in addition to Twitter conversations and messages, (culinary) blog posts and comments.  It is important to note that this study follows Tagg’s (2015: 5) definition of the term digital communication as “interactions between people that are mediated by digital communications technology (tools which transmit information in digital form),” such as tweets and blog comments. Interactions between characters in novels are termed literary dialogue. As written forms of communication, it is possible to explore what writers do with affixed interjections through and with language.  3.3 Theoretical background This dissertation analyzes the terms under study through a semantic lens, and uses semantic and pragmatic features to explain the meanings and uses of affixed interjections. The theoretical approach that is adopted for the textual/digital investigation of diminutive interjections is Appraisal framework (Martin and White 2007), from Systemic Functional Linguistic theory (Halliday 1985). The descriptive standpoint of this study thereby employs some theoretical considerations of Appraisal resources, emotions and semantics and pragmatics.  66  This section provides background for semantic and pragmatic features in 3.3.1, and then for Appraisal in 3.3.2. 3.3.1 Semantic and pragmatic features Diminutives, interjections, and other linguistic items are often described in terms of their semantic features. Lexical semantics posits that semantic features are “theoretical constructs which can characterize the vocabulary of a language" (Leech 1974: 96). They “are usually presented as a matter of opposition, paired positive and negative features, denoting the presence or absence of the particular feature in the meaning of the word” (Brinton and Brinton 2010: 139). Pragmatic features are also concerned with meaning, but with consideration to context, utterance, and/or speech act (e.g. Dressler and Merlini Barbaresi 1994 for the pragmatic feature +FICTIVE). In this study, these features are recognized as typical associations of affixed interjections. These associations are valuable because they “predict precisely the attested range of pragmatic implicatures associated with the semantic representation” (Olsen 1997: 21).   Semantic and pragmatic features often appear in studies of diminutives, such as the model of universal semantic (and pragmatic) features by Jurafsky (1996). The basic semantic feature of [+LITTLE] or [+SMALL] is at the core of diminutives, and other scholars have proposed other features, as seen with Prieto (2005) in section 2.2. Dressler and Merlini Barbaresi (2001: 44) propose the pragmatic feature [+FICTIVE], which they view as a “more basic” and “common” feature for Italian, German, and other languages including English. Interjections are also defined in terms of their emotional semantic meanings, e.g. [anger], [annoyance], [approval], [contempt], [delight], [disgust], [joy], [pain], and [pity] (Jovanović 2004: 22-23). These semantic and pragmatic features are important for the results in Chapter 4 because, following Fillmore (2003), they answer ‘when’ and ‘why’ one affixed interjection may be used instead of another, or an affixed form instead of an unaffixed form.  As with any model, semantic features have weaknesses. The erosion of emotive force and nuances and the tendency of words to adapt and gain new meanings over time cause a semantic feature to become less valid. As Brinton and Brinton (2010: 139) point out, semantic features are not likely to comprehensively capture meaning. These features are not necessarily universal or translatable; in many contexts, they do not ‘mean the same’ when translated or understood by someone from a different culture. Scholars like Anna Wierzbicka and Cliff Goddard follow the 67  idea that ‘semantic primes’ are needed to explain words in a clear and understandable manner. While I think that in theory, these NSM explications are helpful, I do not view such explication necessary for the present systematic and detailed description of affixed interjections. 3.3.2 Appraisal  Appraisal was developed over the course of over a decade in the 1990s by researchers in Australia.65 The theory began within the idea of ‘interpersonal metafunction’ in Systematic Functional Linguistics (SFL)66 and was seen as “a particular approach to exploring, describing and explaining the way language is used to evaluate, to adopt stances, to construct textual personas and to manage interpersonal positionings and relationships” (White 2015). Martin and White (2007) provide the first comprehensive account of Appraisal.67   Despite its roots in SFL, appraisal theory has diversified: it has been applied widely by researchers who are not systemic linguists because it “can be adopted [as a methodological tool] in a more theory-neutral way to the analysis of language” (Bednarek 2008: 12) alongside complementary tools of pragmatic and corpus analysis. Some view Appraisal as a form of discourse semantics “oriented towards mapping semantic domains as they operate in discourse” (White 2011: 21). It “seeks to describe the linguistic resources speakers or writers use for evaluative purposes, always bearing in mind at least one real or virtual person as a co-active recipient” (Becker 2007: 123). In discourse analysis, it is most closely linked to stance – the idea that “whenever speakers (or writers) say anything, they encode their point of view towards it” (Martin and White 2005: 92).  Some of the questions that Appraisal can investigate include, from White (2015): • the linguistic basis of differences in a writer/speaker's ‘style’ by which they may present themselves as, for example, more or less deferential, dominating, authoritative, inexpert,                                                  65 See Iedema, Feez, and White (1994), Martin (1995), Christie and Martin (1997), and Martin (2000). 66 See Halliday (1985) for more on SFL. 67 Although this study uses the term ‘Appraisal’, there is a more recent tendency to avoid terms such as ‘Appraisal theory’ or ‘Appraisal framework’ because it is more precisely considered as a discourse semantic system. 68  cautious, conciliatory, aloof, engaged, emotional, impersonal, and so on; • how the different uses of evaluative language by speakers/writers act to construct different authorial voices and textual personas; • how different genres and text types may conventionally employ different evaluative and otherwise rhetorical strategies; • the underlying, often covert value systems which shape and are disseminated by a speaker/writer's utterances; • the different assumptions which speakers/writers make about the value and belief systems of their respective intended audiences. The resources/categories that are offered through the appraisal framework are represented in Figure 1; namely the wider categories of Engagement, Attitude, and Graduation.   Figure 1: An overview of appraisal resources (© 2005 Peter R.R. White and James R. Martin, by permission) Attitude Attitude divides positive and negative assessments (i.e. attitudinal meanings) into three categories, namely JUDGEMENTS (moral evaluations), AFFECT (emotional responses) and APPRECIATION (opinions about ideas or things) and combines with Graduation to measure their 69  force and/or focus (Tagg 2015: 170).68 Despite the distinctions between the three categories, they are fundamentally linked; that is, the framework “sees the three categories as fundamentally interconnected in that they are all to do with the expression of ‘feelings’. It is just that the ground of that feeling varies across the three modes” (White 2011: 19). AFFECT is concerned with the emotional response, such as I’m sad, I’m happy, or I hate sweets. It is fundamentally linked to the speaker’s emotion – indicating “positive or negative views via either reports of the speaker/writer’s emotional responses or reports of the emotional responses of third parties” (White 2011: 16). In this category, “the action of emotion is directly indicated – feelings are presented as the contingent, personalized mental reactions of human subjects to some stimulus” (White 2011: 19). AFFECT is perhaps the most ‘natural’ category of linking emotion and evaluation. In the other two categories, JUDGEMENT, and APPRECIATION, feelings “are institutionalized in some way and are recast as qualities which inhere in the evaluated phenomenon itself” (White 2011: 19).  JUDGEMENT uses positive and negative feelings as “proposals about correct behaviour – how we should or shouldn’t behave” (White 2011: 19) based on “the social acceptability of the behaviour of human actors, assessment by reference to some system of social norms or morality” (Martin 2011: 16). The statement, He cruelly set up his dog for an Ice Bucket challenge, redirects the negative feelings into a moral judgement on the concept of ‘right and wrong’ behaviour towards dogs (cf. White 2011: 19). Typically, judgements are made on the moral, legal, or religious basis, and can be made about ‘social esteem’ (having to do with normality, capacity, and tenacity) and ‘social sanction’ (veracity, propriety). The category acknowledges adverbials (e.g. positive justly, honestly, truthfully and negative oddly, stupidly, deceitfully), attributes and epithets (e.g. a corrupt politician), nominals (e.g. a brutal tyrant) and verbs (e.g. to cheat, to triumph) (cf. Martin 2002).  APPRECIATION occurs when “meanings by which assessments are made of semiotic and natural phenomena by reference to their value in a given field, perhaps most typically by reference to their aesthetic qualities” (White 2011: 17). Furthermore, “feelings are reconstrued as propositions about the value of things” (White 2011: 19). The category subdivides into five                                                  68 I follow SFL conventions by using SMALL CAPS for appraisal categories. 70  assessments: reaction (impact: ‘did it grab me?’, e.g. arresting, dull), reaction (quality: ‘did I like it?’, e.g. lovely, plain), composition (balance: ‘did it hang together?’, e.g. balanced, irregular), composition (complexity: ‘was it hard to follow?’, e.g. simple, ornate), and valuation (‘was it worthwhile?’, e.g. profound, shallow). Other words that are employed in this category include a masterpiece, dramatic proportions, a gorgeous arrangement. We can see positive APPRECIATION occur with compliments, admiration and praise, an negative APPRECIATION in e.g. an ugly garment, boring proceedings, a flop. Martin (1997; 2000) developed the following six factors that classify different instances of AFFECT. They are replicated by White (2011: 22) as the following: i. Are the feelings popularly construed by the culture as positive (enjoyable) or negative ones (unenjoyable)? ii. Are the feelings represented as a surge of emotion involving some kind of paralinguistic or extralinguistic manifestation (for example, weeping or trembling), or more internally experienced as an emotive state or ongoing mental process? iii. Are the feelings represented as targeting or responding to some specific emotional stimulus or are they represented as a general ongoing mood? iv. Where do the feelings lie on a scale from low to high intensity? v. Do the feelings involve intention (rather than reaction), with respect to a stimulus that is not yet actualised (irrealis) as opposed to an actual stimulus (realis)? vi. Finally, emotions can be grouped into three major sets having to do with un/happiness, in/security and dis/satisfaction. The un/happiness variable covers emotions concerned with ‘affairs of the heart’ – sadness, anger, happiness, and love; the in/security variable covers emotions concerned with ecosocial well-being – anxiety, fear, confidence and trust; the dis/satisfaction variable covers emotions concerned with telos (the pursuit of goals) – ennui, displeasure, curiosity, respect. In this study, these six factors form the basis of analysis of diminutive interjections. The construction of positive-negative feelings, the addition of paralinguistic manifestations (e.g. emoticons in tweets), the existence of a stimulus, and the intensity of feelings are considered for the diminutive interjections under analysis in Chapters 5 to 7. They are considered as such because of the inherent emotional connotations conveyed by both interjections and diminutives in both English and Polish. 71  Graduation The semantics of scaling are found in the graduation of FORCE (slightly, somewhat, very) and FOCUS (a true friend or pure folly) and in the progression of verbs, e.g. to like → to love → to adore. The category of FORCE, thus, can be viewed as a type of amplification system, whereby evaluations and the strength with which we view something or someone as upgraded or downgraded (cf. Bednarek 2008: 30). These affect interpersonal relationships; for example, Crespo-Fernandez (2015: 180) suggests that one function of upgrading FORCE in an online forum could be to “make the participants in the thread agree with his views” and to establish dominance. The graduation of FOCUS, on the other hand, either ‘sharpens’ (e.g. award-winning, all alone) or ‘softens’ (e.g. sort of, kind of). Both amplify ATTITUDE. It has been shown by Tagg (2015) and Zappavigna (2012) that Twitter users frequently ‘upscale’ their emotions with FORCE (e.g. exclamation marks, capital letters). A tweet showing the concept is given by Zappavigna (2011: 170): HOLY CRAP [FORCE]. OBAMA WON [JUDGEMENT+] HE WON [JUDGEMENT+]!!! [FORCE] IM SO [INCREASED FORCE] HAPPY [AFFECT+]!!! [FORCE]69 The tweet suggests that the secondary interjection holy crap! works as FORCE (signaling commitment to the statement) that is relatively mild (or less explicit) than holy shit!, which in turn is milder than fucking hell! This item is used as an emotional reaction that precedes the descriptive context that Obama won the election, and other such emotive features as exclamation marks, intensifiers (so), positive AFFECT (happy) and positive JUDGEMENT (won). The choice of all capital letters is often described as yelling.70 Expletives – alongside terms of address, slang, idioms, euphemisms and interjections – tend to be viewed through the interpersonal system of Involvement. This system complements and works together with Appraisal because it uses nongradable resources to negotiate tenor                                                  69 Unlike semantic features, the plus (+) sign following the term indicates a positive evaluation, and the minus (-) sign a negative evaluation. 70 In the present study I modify these features of graduation slightly for clarity. Intensification (raised force) is indicated by [R-FORCE], while de-intensification (lowered force) is indicated with [L-FORCE]. Instead of indicating [FOCUS], I use [SOFT] or [SHARP]. 72  relations, e.g. solidarity (cf. Martin and White 2005: 33). Involvement also offers “interactants ways to realize, construct, and vary the level of intimacy of an interaction” (Eggins and Slade 1997: 143-144). Because diminutives (along with augmentatives) express a high degree of intimacy between speakers, they may be assessed best under Involvement. In this study, the focus remains on Appraisal. It does, however, consider how Appraisal resources, slang, naming, and swearing may negotiate what is termed ‘emotional involvement’; that is, the level of emotional solidarity and intimacy between tweeters or characters in literary dialogue. Appraisal seems a useful tool in the linguistic analysis of interjections and diminutives; however, few scholars have applied it to these emotional linguistic items (see, e.g. Vian 2008) apart from passing mentions in the larger context of emotions. When the appraisal-emotion relationship is studied, scholars tend to discuss emotions in general; some look at one language specifically, while others look at the role of culture in feeling emotions (e.g. Mesquita and Ellsworth 2001). When analyzing lexis, Bednarek (2008: 13) writes that “its classification of resources of interpersonal meaning lends itself in particular to discourse analytical purposes [and] is also specifically suited to the analysis of emotion talk.” I would argue that it is also suited to the analysis of emotional talk (e.g. affixed interjections) because of the inherent evaluative component of these linguistic units.  Prior literature on Appraisal treats interjections in three different ways. Hood and Martin (2007) write that interjections and related phenomena are not included in Attitude. In contrast, Martin and White (2005: 69, original emphasis) take the approach that interjections (i.e. prototypical interjections, expletives, and related euphemisms) should be viewed “as outbursts of evaluation which are underspecified as far as type of attitude is concerned.” According to Zappavigna (2012), the secondary interjection holy crap! can be classified as (raised) FORCE. Thus, the position that interjections fall into the Appraisal framework because they convey a speaker’s appraisal of a particular situation (Michaelis 2001: 1039) is persuasive. However, interjections arguably can be classified under Graduation (as emotional FORCE) and/or a type of Attitude because of their capacity to express, for example, positive AFFECT through joy (e.g. yay!) and positive APPRECIATION/JUDGEMENT through surprise about something (e.g. wow!). This study takes the position that situational context determines the Appraisal resource(s) of the interjections, and they are typically a combination of Attitude and Graduation. 73  Appraisal considers the “changing variables, interaction, and context” (Alba-Juez and Thompson 2014: 12) of evaluation in a way unafforded by semantic features. That is, Appraisal can suggest why some lexical items have a critical and evaluative stance when “their semantic features do not include anything that could be labeled as negative or critical” (Alba-Juez and Thompson 2014: 12). Thus, while semantic and pragmatic features can give a broad overview of meaning, it is often one-sided and is not able to consider interaction, context, and mode of communication, which are important aspects of any study of (emotional) language. As Alba-Juez and Thompson (2014: 13) point out, “the resources deployed to realize evaluation may vary according to the medium that is used […] [causing] a relatively new use of the evaluative resources of a language” found in, for example, Twitter and blogs. Thus, Appraisal is suitable for a dynamic study of diminutive interjections across written modes. It is important to consider whether Appraisal, a system inductively developed on the basis of analyzing English texts, can simply be applied to the analysis of Polish texts. The answer is arguably in the affirmative.71 Polish texts employ FORCE through intensifiers (e.g. bardzo ‘very’), and positive/negative AFFECT (e.g. kapitan był smutny/szczęśliwy ‘the captain was sad/happy’), APPRECIATION (e.g. łzawe wykonanie piosenki ‘a weepy rendition of the song’) and JUDGEMENT (e.g. on grał umiejętnie ‘he played skillfully’). The Polish language is not different enough from English that Appraisal could not be applied to it. In future, it would be of value to comprehensively assess the application of an Appraisal analysis to texts beyond the English language. Another Appraisal category is ENGAGEMENT, which positions a speaker relative to her proposition, and is briefly presented next. Engagement Engagement is a newer sub-system of Appraisal that “is concerned with the ways in which resources such as projection, modality, polarity, concession and various comment adverbials position the speaker/writer with respect to the value position being advanced and with respect to potential responses to that value position” (Martin and White 2005: 36). They may be grouped                                                  71 Scholars have developed and extended Appraisal to other languages, e.g. Vian 2008 for Brazilian Portuguese. I am, however, not aware of studies using Appraisal in Polish. 74  together under four headings: disclaim, proclaim, probabilise, and attribute. Disclaim includes ‘denial’ and ‘counter-expectation’: Amazingly/Bizarrely, this damaged the trust between the President and his bodyguards. Proclaim includes ‘expectation’ and ‘pronouncement’: The action will, of course, damage the trust between President and bodyguard. Probablilise includes ‘evidence’, ‘likelihood’ and ‘hearsay’: It seems that this damaged the trust, and I hear that this has damaged the trust. Attribute includes ‘extra-vocalisation’: The head of Clinton's security division says this will damage trust.  Engagement is not used in the present study; rather, the resources of FORCE, FOCUS, AFFECT, APPRECIATION, and JUDGEMENT are most frequently accessed for diminutives and interjections. These Appraisal resources can also show (emotional) evaluation in the literary and digital communication found from the datasets described in the next section (3.4) for the analysis in Chapters 5 to 7. 3.4 Datasets This study uses corpus linguistics methodology, which investigates language by using a large collection of texts (Deignan 2005: 5) in order to conduct “a thorough and exhaustive analysis of the feature as it occurs” (Wynne 2005: 2). Four main sources72 form the basis for the case study in Chapters 5 through 7, and the following sections will briefly describe them. They are: • Twitter • blogs and blog comments • fanfiction and fanfiction comments • Google Books (adult fiction) As points of reference, various standard corpora and online slang dictionaries were employed: they included The Corpus of Contemporary American English (COCA), The National Corpus of Polish (NKJP), the Polish Web Corpus 2012 and Urban Dictionary. These were searched                                                  72 At an early point in the study, other corpora were somewhat compiled but not used. A corpus of online consumer reviews was compiled from Amazon; however, the number of diminutive interjections was very low. A corpus of TV show dialogues in English was partially compiled, using online transcripts from The Big Bang Theory and Once Upon a Time, but there were no online transcripts of Polish TV shows. In future, a comparative study of affixed interjections in contemporary TV dialogue, especially from a multimodal perspective, would be useful. 75  for each affixed interjection: the number of occurrences in corpora were used for quantitative results, and the definitions were decomposed into keywords, and semantic and pragmatic features, for the qualitative results for Chapter 4. The methods for Chapter 4 are discussed in 3.5.1., while methods for Chapters 5 through 7 are discussed in 3.5.2. It is important to stress that the four sources in this study are ‘live’. That is, unlike typical corpora which are built by the researcher and fixed in terms of size, live sources are online websites/platforms that are constantly in flux (e.g. being added to). In live mining, the researcher repeatedly revisits the sources for new instances of the linguistic item over a period of time when needed, adds useful examples to the study as they appear, and may interact with users by asking questions about their choice of lexical item, etc. On Twitter in particular, entire or parts of conversations may be used or referred to for context. This is a variation on using data mining applications that gather published tweets and retrieve tweets as they are published (see Bonzanini 2016 on using Python for data mining). This method was chosen in lieu of typical corpora because of the particularly rare instances of affixed interjections in standard corpora, dictionaries, and so on. 3.4.1 Twitter The microblogging site Twitter “is defined as an information network [...] open to all users” (Demirhan 2014: 288). Since its beginnings in 2006, Twitter has become a useful resource in corpus-based linguistic studies of language use and change because of the sheer number of tweets posted daily by 313 million monthly active users (as of December 2016). This amounts to thousands of tweets per hour by all types of users: individuals, businesses, news networks, celebrities, bots and so on.73 Because of this thriving online environment, “Twitter potentially offers access to massively big data sets, of people’s passing thoughts, ideas, comments, interjections, observations and information sources (e.g. new articles) that they are interested in. There is really no offline equivalent to tweets as a source of data” (Hewson, Vogel and Laurent                                                  73 According to Demirhan (2014: 288), Twitter remains the most popular micro-blogging site on the internet. Individual users can communicate with each other in Twitter ‘conversations’ and ‘re-tweet’ information that they want to share with their followers. Businesses use hashtags and retweets, and buy ‘promoted tweets’ to extend message reach (Chaston 2015: 172). The Twitter bot, an unmanned account, posts the information programmed into it. 76  2016: 67). Because of this, Tagg (2015), Zappavigna (2012) and many other scholars have studied Twitter extensively from various perspectives (e.g. sentiment analysis, digital communication).  Compared with other social media platforms, Twitter is unique as “a short-form genre74 […] better suited than blogs for quick, near-synchronous exchanges” (Lomborg 2013: 96). Twitter imposes a strict word limit of 140 characters or fewer, though as of November 2017 the character limit has been extended to 280 characters for most of its users. However, Twitter, like other forms of digital communication and literary dialogue, are asynchronous, meaning that the interlocutors do not have to be present at the same time. At the same time, they present an illusion of synchronous communication. The asynchronous mode is perhaps more frequent on the internet in direct contrast to synchronous CMC, e.g. in instant messaging, online gaming, and chat rooms.75 From the sources mined in the present study, Twitter is perhaps the most synchronous online platform that may provide context through substantial interaction. Zappavigna (2012: 67) notes that Twitter users tend to ‘upscale’ emotion by “[intensifying] interpersonal systems as a way of increasing solidarity through emphasising both positive and negative appraisal as shared experience. Users deploy playful typography and punctuation to assist with this upscaling.” These “[substitute] for paralinguistic cues (such as volume, proxemics, and facial expression) for expressing emotion” (Baron 2009; qtd. in Zappavigna 2012: 67). Users tend to use these resources to intensify emotion, “perhaps with the exception of the kind of interpersonal softening seen with emoticons”76 (Zappavigna 2012: 67). Thus, punctuation                                                  74 I adopt the common view that the concept of ‘genre’ is an elastic and flexible one, generally referring to “a group of texts that are connected by a network of overlapping similarities” (Lomborg 2013: 16), 75 In comparison, CMC is sometimes viewed through a three-dimensional space, where Skype, FaceTime, webinars, and voice chat for gamers, along with text-based communication, are included under the term (Segerstad 2002). Scholars who apply discourse analysis to CMC often refer to the mode as ‘computer-mediated discourse’ (CMD), where they focus on language and how CMC affects language use (Segerstad 2002). 76 Emoticons differ from these features as they do not necessarily enhance emotion. Rather, they are used “to connect with others in convivial, friendly and generally interpersonally positive ways” (Zappavigna 2012: 71) and “can be used to diffuse tension” (78). 77  and other typographic resources – sometimes together with the immediate context – frequently serve to intensify the emotive content.77  Danet (2010: 148, cf. Sindoni 2014: 39) adds that digital writing tends to be “informal and playful” (original emphasis). She lists several features of ‘digital writing’ that often appears on Twitter that I have modified in the following: • reduplicate punctuation marks for emphasis, playfulness: e.g. What????!!!!??? • reduplicate phonemes for emphasis, playfulness: e.g. Whaaaat? Oooh • non-standard spelling for emphasis, playfulness: e.g. txt me, 4u, bizy • capitalization for expression of anger, resentment, aggressive behavior or, if it concerns a single word, emphasis: e.g. oh my fucking GOD i hate messages like this • acronyms, shortenings, abbreviations for playfulness, e.g. LOL (laughing out loud), bff (best friends forever) • descriptions of actions, events, states of mind for emphasis to a certain word or attitude, e.g. *pantones*; <pantones>; >>pantones>> • emoticons, emojis and other visual items (e.g., sending a photo, a picture, a text file) for visual contents, e.g. 😊 😂 🎄 🔥 🤷 • onomatopoeias for playfulness, to mimic speech, e.g. argh, mwuhahahahahaha! These features occur because writers have difficulty expressing emotions and nuances without the paralinguistic gestures found in spoken language (e.g. facial expressions, intonation and body language). Emoticons, as representations of facial expressions, replace visual cues in CMC; “with the help of emoticons writers can emphasize the tone of what they write and show the                                                  77 Twitter has been a cause for concern to those who fear that technology is ‘ruining’ language. As Tagg (2015: 25) argues, there are three myths and misconceptions which contribute to underlying fears of digital communication; namely that “all young people” use the kind of digitalese that is portrayed by the media (for example, ‘If u cn rd ths, u mst b gr8 at txt spk, u gk u! Jk:-.’ from the Daily Herald, March 2001), that adults cannot enter into this illicit ‘code’, and “that digitalese and ‘proper’ forms of a language are somehow in competition with each other.” While a study from 2013 by the social media monitoring company Bandwatch showed that Twitter users are the worst spellers, the results showed that the spelling ‘mistakes’ were typically frequent shortening of words such as omitting apostrophes and using acronyms such as LOL and YOLO ‘you only live once’ (O’Mahoney 2013). The short character limit likely causes users to condense the number of letters, which in turn presses the user to become creative with alternative spelling (O’Mahoney 2013).  78  reader if something is meant to be funny, ironic, or sad” (Greiffenstern 2010: 43). Emojis, as digital images that express an idea, emotion, etc. also contribute to the overall tone. Emoticons of laughter (e.g. smiley faces) are a type of linguistic ‘punctuation’ that occurs at specific syntactic points in an utterance for a punctuation effect (Provine 1993). In other words, “[l]aughter occurs at places in the speech stream associated with pauses, phrase boundaries, and the beginnings and ends of statements and questions [such as] “You are going where? – ha-ha,” but rarely “You are going – ha-ha – where?”’ (Provine et al. 2007: 300). Through its placement in a written sentence, laughter as punctuation is similar to interjections. It can be argued that laughter and interjections are frequently inseparable and function to enrich the emotive content.  3.4.2 Blogs A blog is a type of web-based genre of written discourse, and is created by “a distinctive Web application which came into prominence in the early 2000s” (Crystal 2006: 15).78 According to Crystal (2006: 15), a blog is “a personalized web page where the owner can post messages at intervals.” Myers (2010: 160-161) writes that “blogs provide a vast source of data already in electronic form, so it is easy to download material, save it as text, and use concordance tools to find keywords and collocations” [...] [and] blogs are hard to sample. There is no list of the whole blogosphere, no ‘representative sample’ from that list.” There are many different types of blogs, and opportunities for different types of authorship. For example, many blogs can be viewed as online personal diaries. Others are interactive, have shared authorship, and yet others are static monologues.  A specialist type of blog typically has some distinctive features from other blogs. Food bloggers use different words and phrases than, e.g. music blogs. They more likely engage with words of smell and taste, which are more affectively loaded than those of other senses; “describing something as ‘pungent’ or ‘delicious’ has an inherent evaluative component” Winter (2016: 1). Authors of these blogs build strategies of involving the reader, e.g. “directly thanking the audience” and adding ‘a light-hearted tone to the text, presenting the author as less of a professional authority than an approachable person sharing food-related everyday experienced with likeminded readers” (Diemer and Frobenius 2013: 72). Thus, a food blogger tries to identify                                                  78 See Myers (2010) for a detailed overview of the history of blogs. 79  with her audience, be likeable and friendly, and come across as approachable to bring more traffic to her blog.  Lexical innovation is a salient feature in blogs whereby bloggers may construct norms and personal/group identities. Diemer and Frobenius (2013: 66) describe the feature of “the coining of new, often non-standard, lexical items [which] usually start occurring as hapaxes and then gradually spread through the blogosphere and beyond.” In food blogs, Diemer and Frobenius (2013: 66) identify curious coinages that are typically suffixed, e.g. superfine, wilty, superhumanly, summerish, scoopable, awesomer and melty. In addition, amplification resources include “spelling variations used for emphasis or punning (weddddddddding, gorrrrrrgeous, thissssssssssssss, verrrry, Chaaahles)” (Diemer and Frobenius 2013: 66).  3.4.3 Fanfiction Fanfiction (also fanfic or fic) is a form of online literature that is rooted in community-based enjoyment of a variety of stories that individuals want to re-tell with their own endings, explore what could have happened, change characters’ sexual orientations, or to meld together two or more different worlds. It “may be defined as prose fiction of any length, style, genre, and narrative technique, produced by fans of a wide range of cultural products including TV shows, movies, video games, Japanese manga, and ‘classic’ literature” (Thomas 2011: 205). Thus, fanfiction is as broad as imagination itself.  The idea of embellishing and re-writing stories has been a part of human culture since antiquity (Thomas 2011: 205). It is generally accepted that Jane Austen and Sherlock Holmes fan societies date back to the 1920s, while the Star Trek media fandom dates back to 1967 (cf. Derecho 2006: 62). Since then, the genre “has substantially gained in widespread popularity as a result of the affordances of digital technology, which enables fans from across the world to come together and disseminate their work” (Tagg 2015: 37) on web platforms such as and personal blogs. However, writing fanfiction is a hobby that is written mainly for oneself. While Fifty Shades of Grey and Fifty Shades Freed by E.L. James were produced as films and received international acclaim, most remain in relative obscurity. 3.4.4 Literary dialogue (Google Books) Google Books claims to be “the world's most comprehensive index of full-text books” (Google Books). In 2007, the Google Books project set out its ambitious goal of scanning all the world’s 80  books and making them available online. While many books were scanned, the project eventually lost its momentum due to criticism, copyright battles, and other issues (Yu 2015). Because of these issues, the project was never finished. By 2015, Google Books contained thirty million volumes, but much of it was - and still is - inaccessible (Yu 2015). Many out-of-print books only provide snippet views or a limited selection of pages, and still others do not provide page numbers. Books without page numbers included in this study are marked with ‘np’ after the year in the citation. Polish literature in Google Books is sparse; most literary texts are translations from English, and from those texts, most only provide snippet views. This often makes it difficult to analyze situational contexts and longer interactions. Overall, it is difficult to find a satisfactory number of contemporary Polish works to build a personalized corpus of Polish fiction, especially if the researcher intends to examine rare linguistic items. Because of this, the study focuses more on English than Polish; most of the examples of literary dialogue in this study are from English texts. It would be of use to digitize a larger number of recent Polish novels of fiction for future corpus-based or corpus-driven studies of diminutive interjections and similar linguistic phenomena in literary dialogue. Fiction dialogue is a text-type that is not a web-based genre, and allows for more developed and described speech acts because of longer length and descriptive context. The author of a text has had time to edit and rewrite the text, and unless the book is self-published, there are numerous people involved in the process before it appears on bookshelves. Editors, proofreaders and possibly the author’s writing group or friends weigh in on the characters, dialogue, and other elements of the story. Thus, fiction dialogue has garnered much interest by scholars from many perspectives and disciplines such as in creative writing, stylistics, book editing, and linguistic analyses of literature. Interjections have been discussed in the context of fiction dialogue as well, and studies that have discussed interjections in literary dialogue include Taavitsainen (1995, 1998).  81  Interjections are used in literary dialogue to produce a certain (emotional) effect and contribute to the aesthetics of the text.79 In his works, Chaucer “uses [interjections] in a variety of functions; in some cases, he modifies and colours their meanings with shades of irony” (Taavitsainen 1997: 601). In Early Modern romances, interjections can “convey heightened emotional excitement” (Taavitsainen 1998: 203), add “to the overall affective tone of the text” (207), “enforce the emotive loading” (206) of a passage, and create “narrative suspense” (213). Taavitsainen (1998) shows that interjections have a “foregrounding function […] but sometimes they are just emotive asides directed to the reader to encourage compassion” (211). Furthermore, she finds that “speech quotations and passages that imply interjections seem to be consciously targeted at manipulating readers’ emotions [...] interjections may gain an almost conative function in guiding audience reactions” (211). In sum, interjections play a prominent role in the emotional ‘landscape’ of a text. Writing guidelines in manuals and books for aspiring authors80 provide advice for using interjections which downplays their significance and meanings in a text as opposed to when they are used in natural, spoken conversation. Linguistic studies of interjections in fiction conclude that well-placed interjections create “the illusion of real conversation” (Leech and Short 2007: 132). Thus, studies of interjections in literary dialogue arguably function as commentaries of the literary conventions of creative writing as much as they are expressions of evaluation, attitude and emotion in various societies, cultures, and psychological thought.                                                  79 It should be mentioned that this is a generalization because some genres are arguably less edited for style and emotive content than others, e.g. Harlequin romances; however, all publishing companies have some expectation towards the quality of their authors’ works. 80 Books and articles on writing suggest that “[i]nexperienced fiction writers seem to begin nearly every sentence of dialogue with interjections such as "well," "umm," "huh," "oh," or "ah.” In their attempt to replicate speech, they are overburdening it with words that, when actually spoken, usually aren’t heard” (Sharpe and Gunther 1994: 155). This follows writing conventions that have a negative attitude towards “overburdening” dialogue with interjections; not because they are not actually used in natural discourse, but because it does not fit into literary conventions. Style manuals instead urge writers to convey hesitation and uncertainty without interjections, with “dialogue that moves slowly, that sounds uncertain or groping: Think of dialogue in fiction as what is left when the extraneous verbiage is stripped away” (Hough 2015: np, original emphasis).  82  3.4.5 Reference corpora The reference corpora used in this study are: the Corpus of Contemporary American English (COCA), Global Web-based English (GloWbE), the Corpus of American Soap Operas (SOAP), the British National Corpus (BNC), the Strathy Corpus of Canadian English (Strathy), the Polish National Corpus (NKJP) and Polish Web 2012. They are large balanced corpora that contain a certain number of words per year, are divided by different genres, and include various registers. In the present study, the data from the main sources needed to be tested against a reference corpus, i.e. an alternative data source. Thus, several corpora were used as background corpora to set a baseline for the types and number of affixed interjections for the quantitative section of Chapter 4. a. English  Most of the corpora consulted were developed by Mark Davies from BYU. They are briefly outlined below (from ✓ Corpus of Contemporary American English (COCA): 520 million words, American, 1990-2015. It is the largest freely-available corpus of English, and the only large and balanced corpus of American English. COCA is probably the most widely-used corpus of English [...] and it is equally divided among spoken, fiction, popular magazines, newspapers, and academic texts;81 ✓ Global Web-based English (GloWbE): 1.9 billion words, 20 countries/web, 2012-13. It is unique in the way that it allows [the carrying out of] comparisons between different varieties of English. GloWbE is related to many other corpora of English that we have created, which offer unparalleled insight into variation in English. [Its size] makes it about 100 times as large as other corpora like the International Corpus of English, and it allows for many types of searches that would not be possible otherwise. ✓ Corpus of American Soap Operas (SOAP): 100 million words, American, 2001-2012. It contains 100 million words of data from 22,000 transcripts from American soap operas from the early 2000s, and it serves as a great resource to look at very informal language. ✓ British National Corpus (BNC): 100 million words, British, 1980s-1993. It was originally                                                  81 COCA was updated in December 2017. 20 million words of data from both 2016 and 2017 were added. 83  created by Oxford University press [...] and it contains 100 million words of text texts from a wide range of genres (e.g. spoken, fiction, magazines, newspapers, and academic). ✓ Strathy Corpus of Canadian English (Strathy): 50 million words, Canadian, 1970s-2000s. It is a product of the Strathy Language Unit at Queen's University [with] words from more than 1,100 spoken, fiction, magazines, newspapers, and academic texts. b. Polish ✓ The Polish National Corpus / Narodowy Korpus Języka Polskiego (NKJP): Over 1 billion words. Its website notes that it “is a shared initiative of four institutions: Institute of Computer Science at the Polish Academy of Sciences (coordinator), Institute of Polish Language at the Polish Academy of Sciences, Polish Scientific Publishers PWN, and the Department of Computational and Corpus Linguistics at the University of Łódź. [...] The list of sources for the corpora contains classic literature, daily newspapers, specialist periodicals and journals, transcripts of conversations, and a variety of short-lived and internet texts” (from  ✓ Polish Web 2012: “a Polish corpus made up of texts collected from the internet. Data was crawled by the SpiderLing web spider in June 2012 and comprise of more than 7 billion words in 22 million documents” (from The corpus was made available in March 2018 and searched 24 March 2018.  Established corpora have the downside of not often containing examples of affixed interjections. Despite this, corpora are good indicators of their frequency and potential meanings and functions across different registers. As put by Pulcini, Furiassi and Rodriguez Gonzalez (2012: 18), “large representative corpora of present-day European languages have become an indispensable resource for modern lexicology and lexicography […] because large bodies of authentic texts can be stored and linguistic information can easily be retrieved [and] because they offer up-to-date source material from which new Anglicisms or new meanings/senses of Anglicisms may be detected.” In the context of this study, established corpora, such as COCA or GloWbE, can suggest new meanings/senses of affixed interjections.  There are four major limitations with the use of established corpora besides their limited number of affixed interjections. First, while the BNC is the only British corpus used in this study, the language in it comes from before the focus dates of this study (1980s-1993). I include this 84  corpus, along with Strathy (CanE), only for the comparison of regional varieties. Second, the Polish corpora are fewer than the English corpora; there is only one corpus available of contemporary fiction.82 There are no other Polish corpora likely to contain affixed interjections. Third, the corpora sometimes do not easily differentiate between parts of speech; for example, the interjections dear! and good! gave the ‘no instances’ response when selecting the POS of ‘interjection’. Last, censorship applies to the media language in the spoken part of these corpora as a language-external factor which prevents use of strong swear/taboo words. 3.4.6 Online dictionaries Rare lexical items are typically not found in standard dictionaries and corpora (cf. Zappavigna 2012: 129). The dictionaries used here contain words, definitions and examples submitted by online language users and are some of the most popular online dictionaries (cf. Cotter and Damasco 2007) in English and Polish respectively. They are: Urban Dictionary and its Polish counterpart, Miejski słownik slangu i mowy potocznej [Urban dictionary of slang and contemporary language] – Miejski for short. I also referred to the Oxford English Dictionary (OED) for the few affixed interjections that are well-established in the language, e.g. lordy! and oh goodie!  Zappavigna (2012: 129) writes that Urban Dictionary is “[o]ne non-scholarly, collaborative resource […] of slang and popular internet-related expressions.” It can be viewed as a newer populist-type of dictionary (Damaso 2005); that is, “an emergent dictionary genre that joins lexicographic principles with Web-only communication technologies to provide a context in which users collaborate, cooperate, and compete for meaning-making” (Cotter and Damaso 2007: 1). As with any dictionary, Urban Dictionary has a unique set of benefits and problems that have to do with, for example, slang, online collaboration, lexicographical practice, and idiosyncratic contribution methods.                                                  82 Most Polish corpora seem to be web-based (e.g. Polish Web 2012) or consist of types of texts that do not tend to use emotional language, such as the European Parliament Proceedings (Europarl parallel corpus), news summaries (Polish Summaries Corpus), Wikipedia (Polish Wikipedia Corpus), stock market reports (gpwEcono Corpus), Those that may include emotional language may be too small and specific (e.g. Polish Corpus of Suicide Notes).  85  Urban Dictionary is helpful to linguists in several ways. It offers “insight into slang practices and their contexts of use, useful for obtaining a starting point for exploring a given instance of slang and also for assessing its stability” (Zappavigna 2012: 130). De Schryver (2003: 157) points out the relevance of Urban Dictionary as it potentially never goes out of date like printed dictionaries “and can as such represent the ultimate dynamic repository of knowledge.” Perhaps most importantly, Urban Dictionary “captures what most traditional English dictionaries fall short of: recording ephemeral quotidian spoken language and representing popular views of meaning” (Cotter and Damaso 2007: 1).  Coleman (2016: 335) points out that there is no documented proof that “these casual slang lexicographers [in, e.g. Urban Dictionary] actually use the terms they are documenting.” Urban Dictionary has a “voting system for editorial filtering whereby an accepted term receives a certain proportion more ‘accept’ than ‘reject’ votes from users” (Zappavigna 2012: 129), and users can rate an entry with a thumbs-up or thumbs-down in Miejski. Despite these community ratings which are supposed to show the perceived accuracy of a definition, at worst, the “contributors’ energies are often diverted into abuse and uninformed arguments about etymology and usage” (Coleman 2016: 335). It is possible for an individual to create an entry for a fake word, or for a word which only has meaning for a small group of people.  Many Urban Dictionary definitions include references to sex, drugs, and alcohol. Notably, they tend to reflect “the interests of the site’s demographic, which Urban Dictionary identifies as 53 percent male and mostly under twenty-five” (Baron 2009: 203). For example, one definition for ow! is “the word michael jackson likes to say a lot in his songs” [8 February 2004, zee] and yay! is “[s]lang for Cocaine, popular in California's Bay Area” [2 February 2005, Bookworm]. Thus, it must be emphasized that Urban Dictionary will be used as “a useful starting point for definitions of unfamiliar slang terms” (Coleman 2016: 335) and only those entries with an overall good rating will be included. 3.5 Methods This study used corpus linguistic techniques, semantic and pragmatic feature analysis and Appraisal analysis to study affixed interjections in English and Polish. Chapter 4 used corpus linguistic techniques as well as feature analysis to identify and study affixed interjections in English and Polish corpora. The basic method for feature analysis involved decomposing lexical 86  items into semantic, and sometimes pragmatic, features using mostly abstract level and basic level semantic systems. Existing inventories of semantic and pragmatic features were compiled and, additionally, keywords from user entries in online dictionaries were consulted to identify some features, including specific level semantic or pragmatic features. In Chapters 5 through 7, an Appraisal analysis is used to look at diminutive interjections in four data sources. The method follows the functional approach for classifying appraisal items into categories (AFFECT, APPRECIATION, JUDGEMENT and Graduation) developed by Martin and White 2007 and described in Eggins and Slade 1997, informed by pilot research of types of affixed interjection (Lockyer 2015, 2016). Details of the methods and procedures for data collection and analysis are presented below in sections 3.5.1 (methods for Chapter 4) and 3.5.2 (methods for Chapters 5-7). Section 3.6, Methodological issues, discusses methodological considerations and limitations. 3.5.1 Methods for chapter 4 Data collection and procedures Twitter In this study, the entire Twitter database was systematically mined for the first time in October 2016 to discover which affixed interjections were used. However, the English DIs wowee!, ouchies!, (wh)oopsie! and owie! were mined several times each month from 2013 to 2014 for the pilot study published in 2014, and the Polish DIs (o)jejku! and (o)jejciu! were subjected to a similar search between 2014 and 2016 for the pilot study published in 2016. In the first step for Chapter 4, a list of interjections was put together based on interjections found in previous research on interjections that are used in present-day English and Polish (e.g. Ameka 1992, Jovanović 2004 for English; Wierzbicka 2003, Sieradzka and Hrycyna 1996 for Polish). The initial list of the study was divided into each type of interjection (cf. 1.2). This included primary emotive interjections such as wow!, (wh)oops!, yuck!, ugh!, yay!, ow!, ouch!, whoa!, ah!, ew!, and oh!, cognitive interjections such as oho! and aha!, conative interjections including shh! and shoo! Contact interjections were also added, including hello!, sorry!, goodbye!, good morning!, good afternoon!, later!, hi!, sorry!, please!, thanks!, and descriptive/onomatopoeic interjections such as bam!, wham!, boom!, and secondary interjections including lord!, god!, 87  jesus!, oh dear!, oh boy!, oh no!, crap!, shit!, lol!, hell!, and fuck! All Polish counterparts were included in their own list.  Next, all the diminutive, augmentative, and slang affixes were gathered from prior research (e.g. Schneider 2003). These included suffixes including -ie, -ey, -let, -ling, -o, -ola, -rama, -aroo, and -s (see 2.3 and 2.4) for English, and the -k- affix for Polish (along with its palatalized form and the slang affixes). Prefixes were also included, namely mega-, mini-, micro-, hyper-, and super- for English and Polish. Each interjection from the list was systematically given each affix and entered into the search function of Twitter to find out how many tweets had the item and over what period of time. Occasionally, when the form appeared to have many occurrences in another language, the form was put into the Advanced Search function to only display results from a selected language – either English or Polish. A form was included in the study if it occurred at least once or twice a month (or the equivalent in one year) by different users in the language under study. From the first full and systematic pass, done in 2015,83 it was determined that interjections did not receive prefixes in either English or Polish; thus, the diminutive and augmentative prefixes were excluded. Some primary emotive interjections, especially oh! (Pol. och!), eh! (Pol. ech!), ah! and ew! did not receive suffixes, and were taken from the list. It was found that primary emotive, contact, and secondary emotive interjections were most likely to receive suffixes, and these finalized affixed interjections are listed in 4.1. Occurrences with affixed forms of conative, cognitive, and descriptive/onomatopoeic forms were very sparse (e.g. whammo!) and it was deemed that they belonged to a future study. Next, the list was taken to the reference corpora for a similar systematic analysis. Reference corpora Because of the sheer size and colloquial language of Twitter, the affixed forms that never appeared on Twitter were arguably unlikely to appear in any reference corpora. Thus, the list compiled from the Twitter results were searched on the BNC, COCA, SOAP, Strathy, GloWbE for English, and the NKJP in October 2016, and Polish Web 2012 was added later in March 2018.                                                  83 It should be noted that I had sporadically looked at affixed forms of interjections and other linguistic items earlier, and looked at these results again several months later, but this full pass from 2015 represents the first complete and systematic search of all possible forms. 88  Each affixed form was placed into the search function and the raw frequency added to the corresponding Table. Because it is typical in corpus linguistics methodology to include both raw and normalized frequencies, the normalized frequency counts (normalized to 100 million) were added in square brackets if the raw frequency was above zero.  After the affixed forms were put in their respective Tables, COCA and GloWbE, as representatives of contemporary American English and web-based English, were selected for a search of the base forms of each already-listed affixed interjection. A POS search for ‘interjection’ was first applied, but sometimes this function did not work. When this occurred, the POS function was removed and the entire corpus was searched and all occurrences (including the form in other parts of speech) were included in the Table and marked with an asterisk. Together, the results for the base and affixed forms in the reference corpora were able to paint a quantitative picture of affixed interjection use that cannot be applied to the ‘messy’ internet data found on platforms such as Twitter (cf. Hoffmann 2007: 69).  Dictionary meanings Chapter 4 focuses on meanings by laypersons who contributed to online slang dictionaries. Arguably, the demographic on these websites are older teens and young adults in their early to mid-twenties, and therefore the definitions proposed by them may be narrow in scope and should be regarded with caution (see 3.4.6). They may be, at best, seen as a rough picture of popular trends in colloquial and informal use by young adult laypersons that contribute to identifying the semantic feature(s) associated with an affixed interjection. Clearly, definitions from Urban Dictionary and Miejski do not reflect the use of affixed interjections by young children and those adults that rarely, if ever, go online and/or use social media. Each affixed interjection, from the list compiled through the systematic search of Twitter, was searched for in either Urban Dictionary (for English) or Miejski (for Polish). If the form returned any occurrences that had a high approval rating, was not clearly a fabricated definition, and was used as an interjection, then it was included in the study. Most affixed interjections did not receive more than one or two usable definitions because of creative neologisms submitted by users. (See Cotter and Damaso 2007 for a discussion of neologisms in online slang dictionaries.) Furthermore, Urban Dictionary vastly outsurpassed Miejski in terms of 89  thoroughness (e.g. quality of definition and example sentence) and general number of affixed interjections. The user entries were then read and analyzed for keywords in the definition (e.g. new) and in the corresponding hashtags (e.g. #accident). These words helped in identifying the semantic and pragmatic features of each affixed interjection and to group together the most frequent uses of affixed interjections.  3.5.2 Methods for chapters 5 – 7  In Chapters 5, 6, and 7, the study turns to diminutive interjections that are found in four sources (see 3.4) in a descriptive, discourse-analytic and interactional way. Following Schneider and Strubel-Burgdorf (2012) for the analysis of the English diminutive suffix -let, this means that the affixed interjections surrounding utterances or tweet messages are taken into consideration, and the situational contexts shows the time, place, and occasion of the dialogue. The approach is synchronic and therefore only examines present-day (21st century) language. The method differs from typical studies of diminutives and interjections by turning to Appraisal analysis (sometimes combined with Involvement) for a new perspective on these lexical items. Appraisal (see 3.3.2) was applied to the literary excerpts and digital communication using the following method (following Eggins and Slade 1997: 138, 140, adapted from Martin 1994): • identifying all the Appraisal items in a conversation/text, • classifying the Appraisal items into the four categories of Attitude as found in [AFFECT+/-], [APPRECIATION +/-], [JUDGEMENT +/-], and Graduation (raised or lowered FORCE and FOCUS). • summarizing Appraisal choices made by the different speakers (including the author in a literary text) • interpreting the patterns and use of Appraisal by the speaker(s) The affixed interjections are divided into three groups: ‘positive’, ‘playful’, and ‘negative’ based on their emotional content and situational context. The three groups are based on the suggestions made in my pilot studies (2014, 2016). These broadly cover all the potential emotional uses of diminutive interjections. The groupings of ‘positive’ and ‘negative’ is “a basic division expressing the centrality of the evaluative component in emotions” (Ben-Ze’ev 2000: 99). This may be seen in the positive and negative potential for Appraisal resources (e.g. positive 90  or negative AFFECT). The group of ‘play’ was added for those uses that did not fit into the conventional mold or could be considered both positive and negative, such as sarcasm, where the emotion shown is contrary to the speaker’s true feelings, and contexts of (humorous) irony, and romance.  Appraisal is applied in these chapters for several reasons. They “[enable] exploration of resources for evaluative meaning” (Schleppegrell 2013: 26), such as Attitude and Graduation. Moreover, Appraisal may show the frequency and type of Appraisal items used by speakers who use diminutive interjections to describe and explore the attitudinal lexis and other language used with diminutive interjections to evaluate, adopt stances, and construct interpersonal relationships in Twitter, blog, fanfiction, and fiction dialogue data.  Data collection  For this study, each dataset was manually searched for clear and appropriate examples that had enough context to apply the Appraisal analysis. They were searched thoroughly and then routinely after that for a period of two years, beginning in late 2015, to look for new examples and to note how the interjections were being used. Over a period of two years each dataset was searched a minimum of 24 times, except for Twitter, which was searched two to three times a month.84 Because these datasets were ‘live’, not part of a fixed corpus, and at times difficult to find, it was not possible to give a clear picture of how representative the sample used is for each dataset. Twitter  For Chapters 5 through 7, the initial tweets found in the full systematic pass for Chapter 4 were identified as positive, playful, or negative and placed into their respective groups. As the study progressed, Twitter was searched three or four times a month between 2016 and 2018 for new tweets and to gain a clearer picture of the frequent contexts that the diminutive interjections were used in. Sometimes the DI with a particular word was entered into the search function to find examples of the DI in a particular context or with a certain emotion, e.g. wowee + sarcasm.                                                  84 The pilot studies employed searches of Twitter from an earlier date (beginning in 2014); however, the present study, with its current datasets, began in late 2015. 91  This helped to sort through the variety of possible contexts for DIs, especially those used more frequently, such as wowee! and whoopsy! It was important that the tweets included in the study were most likely written by adults or teenagers.85 Many of the diminutive interjections searched for brought up usernames, misspellings, retweets, and ambiguous tweets because the Twitter search function does not allow a POS search. The advanced search feature, however, allowed for searches of an exact phrase, within a time period, and in a particular language.86 In addition, it was necessary that the diminutive interjections were used as such because they were frequently used as hashtags and could therefore be important for purposes of alignment to a certain judgement or social movement (e.g. the recent #MeToo hashtag to denounce sexual assault and harassment). In this study, they were viewed simply as part of the message. For example, the #StupidDonald hashtag was understood as a negative judgement of Trump that could be paraphrased as ‘I think Donald Trump is stupid’. The study generally did not include diminutive interjections with hashtags (e.g. #whoopsie) because it was unclear whether the term was being used as an interjection.  Blog and blog comments  Beginning in 2016, and monthly afterwards until March 2018, a Google search was run to find examples of each diminutive interjection. This involved entering the DI and and ‘recipe’ into the Google search engine and finding an occurrence, used correctly, that was from a blog comment. (Frequently the food item was named after the DI, e.g. ‘Maui Wowee Dipping Sauce’, which was not included in the study.) Once a suitable match was found, the blog was read at least once and the ‘about the author’ section was checked to confirm that the blog author was an adult or youth over thirteen years of age. The immediate context was read at least twice to understand the                                                  85 This was checked by looking at the tweeter’s bio, which sometimes gives personal and demographic information.  86 It should be noted that computer programs and algorithms have been developed to extract tweets with certain words or hashtags. These are commonly used in sentiment analysis and/or computational linguistics. In addition, there are websites that pull tweets continuously based on the word or hashtag entered. One of these websites was used for the 2014 pilot study. However, there was a problem with the site: like a regular Twitter search, it brought up misspellings, usernames, and was not always accurate. Navigating through thousands of tweets from a .txt file this way seemed counterproductive, especially since the landscape of Twitter is in constant flux as tweets are added and deleted. A manual search of Twitter seemed a better way to proceed with the present study. 92  emotional connotations of the diminutive interjection. Because a limited number of blogs had diminutive interjections, it was deemed unnecessary to download and save each blog. Instead, the examples were chosen randomly. Those that were used had the immediate context (that is, a couple sentences before and after the DI) copied into the dissertation file with the date and author.  Blog comments were searched for and treated in a similar way: they were read for correct interpretation, and the previous and following comments were also read in case they could provide clues for the use of the DI. The blog was also read for emotional clues that would lead to a blog commenter using a diminutive interjection. Once the initial survey was completed, the blog comment was extracted into the study along with the username and date of posting. As with the blogs, the blog comments were chosen randomly. If important to the use of the diminutive interjection, parts of the original blog or the blog author’s response to the blog comment were also extracted.  Fanfiction  For two years (early 2016 to March 2018), the fanfiction data was mostly found through the search engine, Google, and some were found through a search of Because the search feature on only provides data from the title and summary, there was an extremely limited number of usable occurrences from the search. Next, Google was searched for the DI +, which provided several results from fanfiction (comments) on Next, the DI was searched together with the term ‘fanfiction’, ‘fanfic’, and ‘fic’ to find examples from other fanfiction websites or blogs containing fanfiction.  Examples of diminutive interjections were copied with their situational context to the dissertation file, along with the name of the fanfic writer, the fanfic title, and the link. That is, the situational context means that the characters and situation are specified and the DI speaker’s Appraisal resources included. The rest of the fanfic – or several pages before and after the DI depending on the length of the entire fanfic – was briefly scanned to get an improved picture of the story, characters, and style, and also to check that the character who used the diminutive interjection was an adult. The surrounding context was read at least twice to limit the possibility for misunderstanding. Google Books  93  For examples of literary discourse, each diminutive interjection was entered into the Google Books search function, beginning in late 2015 and searched monthly until March 2018. All the examples were scanned through the brief previews offered: these ranged from a page to ten or more pages from an entire book (though sometimes with missing consecutive pages). Those from dialogue in adult fiction novels were examined further by reading the context in more detail. The DI was searched within the book in order to find out whether any other character used that lexical item; in (nearly) every instance, the DI was only used once or by one character in a particular scene. If the DI was used clearly, with little possibility of misinterpretation, the literary dialogue excerpt was inserted into the dissertation file with enough situational context to provide a clear picture of the speaker(s) and situation. If the available excerpts were too many to include in Chapters 5 through 7, such as for example lordy!, then they were chosen randomly. In many instances, the books only offered snippet previews that barely put the term into context. This occurred most frequently for Polish fiction. The snippets were not included, causing the amount of Polish literary examples to be much smaller in number than the English ones. Because of this limitation, there is a broader discussion of English literary dialogue than there is of Polish.  3.6 Methodological issues Before moving on to the results, I would like to briefly discuss the four issues of misinterpretation, cherry-picking, representativeness of datasets, and anonymity/privacy that arose during the study. (Mis)interpretation To analyze what emotional connotations and evaluations were being conveyed by the users of diminutive interjections, the study followed that of Stange (2015: 82), where “occurrences were interpreted in terms of why the speaker had used them in this particular situation.” Obviously, misinterpretation was possible; e.g. short tweets were likely to be interpreted differently than the tweeter intended. In addition, sarcasm and irony can be easily misinterpreted. To minimize the chances of misinterpretation, examples with sufficient context were primarily used, and when possible, the tweeters were asked why they chose the diminutive form. This proved difficult because interjections and diminutives are highly dependent on context and tend to 94  appear in isolation, and most tweeters did not respond to the question. Many texts were subsequently excluded because the emotional connotations or reasons for using the diminutive interjection were ambiguous.87  Cherry-picking A researcher may analyze more data than she may include in a paper or book, and therefore at a certain point must choose examples from her corpus that present her findings in the clearest way. ‘Cherry-picking’ is a term used when a researcher chooses examples that best illustrate her personal views but overlooks inconvenient data (Groom, Charles and John 2015: 9; see also Baker 2006). This method produces skewed results.  Creating a balanced and well-represented corpus that is put through a concordancing program such as Wordsmith Tools is typically seen as the solution to this methodological issue because it offers collocations, keywords in context, and other features that help analyze the linguistic items under study. Indeed, this method is important for corpus studies of recurrent linguistic items. When the linguistic item is rare enough that several large datasets are required to find enough examples to show the variety of meanings, then a concordancing program is arguably not the best option. Instead, if the researcher already has a solid basis of meanings from other sources such as dictionaries and reference corpora, then choosing examples that further the argument to supplement results should be acceptable. In the present study, ‘cherry-picking’ did not occur: inconvenient data was not ignored, evidence was not suppressed, and arguments were not made by selective observation. The most typical semantic and pragmatic features of diminutives were found in previous scholarly research (see Chapter 2, cf. Schneider 2003, Jurafsky 1996), which were then corroborated with the affixed interjections found in Urban Dictionary and Miejski (see Chapter 4) and their                                                  87 An obvious way to minimize the risk of misinterpretation would be to ask the writers why they used the diminutive interjection. However, as was discussed in Lockyer (2014), it is difficult to do so. Asking Twitter users proved widely impractical; assuming the user decided to respond honestly, they either forgot why they used it or did not want to participate. Blog writers and blog commentators do not generally have a way of being contacted except by commenting, and most examples used in the study had usually been written at least several weeks prior, and the writers would likewise have forgotten what they had been feeling at the time of writing. Asking them to remember could cause misinterpretation as well. Literary discourse therefore proved the most informative because the text describes how the diminutive interjection is said, and why. 95  frequencies with quantitative data from reference corpora. The examples for Appraisal analysis gathered from Twitter, blog/fanfiction comments were presented in their entirety; the fiction and fanfic excerpts provided enough situational context and, when needed, a description of context before and after the excerpt. As a whole, these datasets were subjected to random sampling – that is, the excerpts, tweets, etc. were chosen randomly, provided that there was enough context to be able to adequately analyze the text. Representativeness of datasets It may appear problematic that four main datasets were mined for examples, and each has a different number of tokens. Twitter is significantly larger than fanfiction, and blogs fall somewhere in the middle. Nor does the literary discourse found in Google Books include all adult fiction written from the year 2000 onwards. Many social media sites (e.g. Facebook), online forums, and consumer reviews were deliberately excluded, even though to some degree the language that appears on each type of site differs slightly from others.   Anonymity and privacy Anonymity and privacy presented limitations when using Twitter as a ‘live’ corpus. It is not possible to ensure the privacy of each user and there is no guarantee that each user’s identity will be completely protected because of the search function on Twitter and other factors (see Tagg 2015). To protect privacy, I follow Zappavigna’s (2012) method of replacing usernames with, e.g., ‘user’ and ‘user1’. The only usernames I have kept are those mentioned within tweets to refer to public figures or organizations (for example, @realDonaldTrump or the Trump organization @Trump) that are crucial to understanding the tweet correctly. 3.7 Chapter conclusions This chapter presented the particularities of the datasets/corpora, methods and theoretical considerations that are used in the present study. The semantic and pragmatic feature analysis used in Chapter 4 was discussed, followed by a discussion of the Appraisal method used in Chapters 5 to 7.  The four main sources used for the Appraisal analysis were presented and identified, namely Twitter, blogs, fanfiction, and fiction dialogue. In addition, the sources for the semantic and pragmatic features were briefly discussed: Urban Dictionary, Miejski, and established corpora.  96  The following chapter turns to corpora and definitions to show which interjections receive affixes in English and Polish. The background presented in this chapter and earlier chapters informs the affixed interjections that I find. Following the structure of Chapter 1, the chapter is organized according to type of interjection: primary interjections (4.2), contact interjections (4.3), and secondary interjections (4.4).   97   Corpora and definitions 4.1 Introduction A search of Urban Dictionary, Miejski, and Twitter shows which interjections can receive diminutive, augmentative and/or slang suffixes (see 3.5.1 for Methods). They are divided into groups of primary emotive interjections, contact interjections, and secondary emotive interjections, and presented in the following bulleted lists.  The affixed primary emotive interjections in English are: • EEK > eeks, eeksy, eeksie, eeky, eeker   • OW > owies, owie • OUCH > ouchies, ouchers, oucher • UGH > ugha, ughers, ughies, ughz, ughs • WHOA > whoaies, whoaee, whoazers, whoaz • (WH)OOPS > (wh)oopsie, (wh)oopsy, (wh)oopsies, whoopso, whoopser • WOW > wowie, wowee, wowies, wowsie, wowza, wowzers • YAY > yayee, yayza, yayzor, yayzors, yayzers, yayzies, yayzella • YUCK > yuckie, yucks, yucko   In Polish, they are: • FUJ ‘yuck’ > fujka, fujki • (O)JEJ ‘(oh) dear’ > (o)jejku, (o)jejciu, (o)jejkuś Affixed contact interjections in English are: • HELLO > hellosies • HEY > heysies • (GOOD) MORNING > (good) morningsies • (GOOD) AFTERNOON > (good) afternoonsies • (GOOD) NIGHT > (good)nighties, (good)nightsies, (good)nighters, nighty-night, nighto • BYE > byesies, byesy-bye, byedy-bye • LATER > laters 98  • SORRY > sorries • THANKS / THANKYOU > thanksies, thankyousies, thankies, thankie, thanky, thankee In Polish, they are: • HEJ ‘hey’ > hejcio, hejcia, hejka, heja, hejeczka, hejo • SIEMA ‘hi how are you’: siemka, siemanko, siemaneczko, siemcia • WITAM ‘I greet you’ > witanko, witka • ELO ‘’ello’ > eloszka • DZIEŃ DOBRY ‘good day/morning’ > dzieńdoberek • DOBRY WIECZÓR ‘good evening’: miłego wieczorku, dobry wieczorek • DOBRANOC ‘goodnight’ > dobranocka, miłej nocki • PA ‘bye’ > papatki • NARA (< NA RAZIE) ‘so long’ > narka, narazka, narazicho, nareczka • DO JUTRA ‘until tomorrow’ > do juterka • PLIS ‘please’ > pliska • SORY ‘sorry’ > sorenjca, sor(r)ka, so(r)rki, soreczki, sorewicz • DZIĘKI ‘thanks’ > dziękowa, dziękujki, dziekowka Affixed secondary interjections in English are the following: • DEAR > dearie (me) • LORD > lordy, lawdy • GOOD > (oh) goodie, goody, goody me • OH NO > oh noes, oh noez • LOL > lolies, lolers, lolski, lolskies, lolz • CRAP > crappers, crapola, craporama, craparoo • SHIT > shitola, shitorama, shitaroo • FUCK > fuckola, fuckorama, fuckaroo, fuckaroonie, fuckarooski In Polish, they are: • JEZU(S) ‘Jesus’ > jezusku, jezusicku, jezuniu, jezusieńku • CHOLERA ‘damn’ > cholerka, cholercia 99  Other types of interjections are not suffixed; for example, cognitive interjections including aha! and oho! are not affixed in English or Polish; and, neither are conative interjections, e.g. shh! or any other types of interjection except to create nonce forms.88  In the following sections, the discussion turns to the frequent meanings and associations suggested for affixed interjections by users of Urban Dictionary and Miejski. That is, it shows how online users on sites such as Urban Dictionary or Miejski define affixed interjections in relation to their unaffixed forms, and which semantic and pragmatic features may be glossed from the main keywords and associations of these definitions. Each section also takes into consideration quantitative data analysis by showing the raw and normalized frequencies for affixed and unaffixed interjections in established corpora. The data from these sources help to answer the above questions about the meanings and frequencies of primary emotive (4.2), contact (4.3), and secondary interjections (4.4). By using Urban Dictionary and Miejski as my starting points – along with any definitions found in the OED and the affixed forms found on Twitter – I avoid showing results merely based on my intuition as a native speaker of English and heritage/near-native speaker of Polish. I do not claim to include all possible meanings or affixed forms, only those that appear most frequently. In addition, I do not look at traditional dictionaries because these affixed forms rarely appear in them. 4.2 Primary emotive interjections English seems to have a greater number of affixed primary emotive interjections than Polish, likely because the primary interjection (o)jej! in Polish serves to convey many specific meanings attributed to English eek!, ow!, ouch!, ugh!, oops!, yay!, etc. (SJP; see also Lockyer 2015a). There is also a greater number of suffixes: in English, the affixes that are most frequent are -y/-ie, -s, -er(s), -za, and -ella.89 The diminutive suffixes -let, -ling, or -ette are not attached to any of these interjections (e.g. *eeklet!, *whoopsling! or *ughette!).                                                   88 In Slovak, there is the suffixed interjection tiško ‘hush’ (< ticho). 89 Several affixed forms of a primary interjection can also be used as nouns (e.g. I had an owie) or adjectives (e.g. she was feeling whoopsy); Polish fujki! (< fuj ‘ew’) may also function as a noun that could be roughly translated as ‘disgusting things’ or ‘litter’ in colloquial speech because -i is also a plural suffix. A clear-cut 100  4.2.1 English In Urban Dictionary, there are a number of affixed forms that appear from the following primary emotive interjections: yay!, ugh!, ouch!, wow!, whoa!, oops! and eek! The semantic and pragmatic features that arise in the dictionary definitions most frequently include [+INTENSE], [+EXTREME], followed by [+CHILD], [+CUTE], [+AFFECTIONATE], and [+FEMALE]. However, diminutive forms, in comparison with slang-affixed forms, are less often found in Urban Dictionary and are less likely to be defined as different from the base form. Affixed forms from the base interjection yay! appear in Urban Dictionary. The top (and only) definition for yayza! defines the term as a “[w]ay of expressing joy, similiar [sic] to ‘w00t’. Branches off Yay” [12 September 2006, NekoLynx]. The writer of the entry links the term to yay, yahoo, w00t, wee and joy. In addition, yayzello! is “an exclamation of joy” [28 April 2006, Plunketto] and yayzers! “[m]akes the word yay cooler. An exclamation of happiness” [8 January 2004, Kathy Redden] and “[e]xtreme happiness. Excited for something coming up” [26 January 2014, CharBear]. Yayzies!, defined as “[a]n expression used to describe a state of immense pleasure, such as one would experience on receipt of a two-for-one McDonald's voucher” [2 December 2010, BigMacluvrrr] also is included, along with yayzor(s)!, “[a] reserved, yet mildly excited exclamation of cheer, agreement, anticipation, or celebration” [20 February 2004, KiloWatts].  As expected from the base yay!, defined in Urban Dictionary as “an exclamation of pleasure, approval, elation, or victory” [11 March 2004, IAMSODOT], the keywords for the affixed forms are very positive: joy, happiness, exclamation, cheer, agreement, anticipation, and celebration. The keywords cooler, extreme, and immense suggest that the affix amplifies the positive base emotion and convey the semantic feature of [+INTENSE] and [+EXTREME].                                                  example is whoopsy! from whoops!, an exclamation of surprise caused by an accident or mistake. The addition of the -y suffix creates the diminutive whoopsy!. However, whoopsy is attested in slang and regular dictionaries as (primarily in British) the noun meaning “an act of defecation” (Thorne 2009) or “a piece of excrement” (Ayto and Simpson 2010). In the 1940s it appeared in American English to mean ‘a homosexual’ or as an adjective, meaning ‘homosexual, effeminate’. It is also used as an adjective that is synonymous with ‘queasy’ or ‘nauseous’. Likewise, wowie (< wow) is part of the term maui wowie, which is another name for marijuana, and owie (< ow), which appeared in the 1990s, is also a noun referring to a minor injury (Dalzell 2009). 101  The interjection ugh!, defined as “[u]sed to describe disgust or boredom” [26 April 2003, Bob], is typically intensified emotionally and used to convey negative feelings. The slang-suffixed ugha! is “[a] word people use to express anger or fear in a moment when they can't think of anything logical to say” [19 November 2013, Rainbows_Butterflies_Larry] and carries the semantic feature [-LOGICAL]. The quasi-diminutive ughers! is defined as “a new version of ugh […] used during physical activity…of all sorts” [4 June 2009, fjadkfjsdkfla] with the hashtags #ugh, #damn, #dang, #fuck, and #ouch.90 This definition seems to imply that ughers! is a newer form of ugh! as of 2009.  Slang-affixed forms of ouch!, defined as “[a]n interjection said when hurt” [22 March 2003, Aaron W.], also tend to convey a more intense meaning than the base and/or are considered child-speech. One Urban Dictionary user defines oucher! as “[s]o intense and painful that a normal ‘ouch’ will not suffice” [28 October 2009, Mr. Munroe], with the hashtags #ouch, #pain, #more, #ow, and #damn. Ouchers! is seen as used by small children, but is spreading more widely to teenagers as well [MastaChief93, 11 December 2009]. Therefore, the semantic features associated with slang affixed ouch! and ow! are generally [+INTENSE] and possibly [+CHILD]. The diminutive interjections owie! (< ow!) and ouchies! (< ouch!) are mainly defined in the same way as their unaffixed forms. Ouchies! is defined nearly in the same way as the unaffixed ouch!; the Urban Dictionary user writes that it is “[a] word used when you hurt yourself” [27 May 2004, Mr Miyagi]. The alternate affixed interjection of pain, owie!, is likewise similar to its unaffixed base form: “said when hurt” [11 December 2002, Aaron W.] with the example sentence Owie, I cracked my skull! There are three other somewhat reliable definitions for the affixed interjection owie! in Urban Dictionary, all with similar phrasing and hashtags including #ouch, #hurts, #pain, #feelings, #painful, and #uncomfortable. Only one newer definition (with a two thumbs-up rating) differs from the unaffixed form by suggesting that owie! “is the sound a mentally deficient child makes after promptly hitting him/her self in the head with either there [sic!] fist or various objects” [17 February 2018, Help_me].91 Thus, the interjection owie! might                                                  90 There are many slang combinations of interjections that will not be discussed in this dissertation but merit further investigation since the second interjection seems to function like a suffix in creating a new interjection, e.g. ughew (ugh + ew), ughblah (ugh + blah), ughwoo (ugh + woo) and ughfuck (ugh + fuck). 91 These results support the discoveries I made in an unpublished (2013) study that looked at the results of an online, written questionnaire filled out by respondents in Vancouver, BC about their attitudes towards 102  have the semantic feature [+CHILD], likely because of the strong child/infant connotations from the noun owie. The unaffixed form wow! may have several different meanings. An Urban Dictionary user describes the popular meanings this way:  “the word ‘wow’ can be used when somebody says something stupid, when you realize something you should've known a looong time ago, when you are told something unexpected, when you are insulted very very badly, whenever someone says something pervy and stupid, when you see something amazing, when you are kissed..XD, and just whenever you feel like it. WOW!” [21 February 2009, mz. Zoi].  These eight different uses for wow! has likely caused the interjection to have several affixed forms: wowsie!, wowsa!, wowsas!, wowoza!, wowzers!, and wowee! Several users of Urban Dictionary compare wow! with the base form and suggest that the affixed form has the semantic feature [+INTENSE], e.g. wowsa! and wowsas!, which “shows something to be a further breath taker than wow” [1 December 2007, Pebbles_14] and occurs “[w]hen you think something/someone is beautiful, fantastic, pretty...When something or someone is WOW!” [13 May 2008, Maura Taylor]. In addition, wowoza! is defined as “a form of wow that creates a really surprised and wonderful imagery. a beautiful word. similar to wow” [14 January 2008, queen lucy], and wowzers!, “[which] [m]eans the same thing as wow but with more "oomph" & enthusiasm” [17 July 2014, wavyg]. In sum, the slang affixed interjections of wow! are defined on Urban Dictionary as intensifiers that are generally positive with hashtags including #surprised, #beautiful, #breath taker, #fantastic, #pretty, #wonderful, and #enthusiasm. The diminutive forms of wow!, namely wowsie! and wowee!, are not as differentiated from the unaffixed form as the slang affixed forms. That is, wowsie! is defined as “[s]omething you say to express shock, disbelief, or empathy” [25 May 2009, Empathetic Fool], which may easily be applied to wow! The diminutive form wowee! is also seen as relatively synonomous with its unaffixed form through the definition of “another form of wow. You can use wowee wherever                                                  and uses of whoopsie!, owie! and ouchies!. The respondents described whoopsy! as ‘cute’, ‘whimsical’, ‘condescending’, ‘playful’, ‘humorous’, ‘comedic’, ‘ironic’, and ‘minor’. In comparison, wowie! was considered ‘silly’, ‘not serious’, ‘sarcastic’, ‘playful’, ‘extreme’, ‘small’, ‘childish’, ‘cool’ and conveying ‘pleasure’, ‘excitement’, ‘delight’ and ‘amazement’. 103  you can use wow. It also means "thats cool’” [15 January 2005, simma]. Only recently has a new definition suggested difference in meaning by defining wowee! as “[t]he proper response when something is so shocking you can’t think of any other response” [2 January 2018]. This suggests that wowee! can be used to mean more X than its unaffixed form, and may therefore have the semantic feature of [+EXTREME]. 92 Whoa!, like wow!, also has several different uses, including “1. To express surprise (interj) 2. To express astonishment(interj) 3. To indicate a desire for one to end that of which they are speaking (interj)” [3 October 2004, Z]. This Urban Dictionary user also helpfully notes the alternate spelling woah! and gives the example Whoa! You scared me. The interjection has two affixed forms in Urban Dictionary: whoazers!, defined as “1. to be disappointed 2. to be surprised” [15 December 2007, Lostgirlforever92] and whoaz!, “cooler and more awesome way of saying whoa” [10 June 2008, born2ride xo]. The definition for whoazers! does not compare the affixed form with its base, but does suggest associations with negative surprise and disappointment, while the second user views affixed whoaz! as an improvement to the unaffixed form through words such as cooler and more awesome, which suggest the feature of [+EXTREME].93  The unaffixed oops! is defined as “[s]omething people say when they screw up” [1 April 2004, Hannah Banana]; that is, it is a reaction to a mistake. The top definition for diminutive interjection oopsie! in Urban Dictionary is similar: “1. A Mistake 2. A[n] Accident” [31 March 2010, Johnny Davison], with the hashtags #accicent, #mistake, #regrets, #error, #problems, and #failed. Other definitions of (wh)oopsy! refer to the term as a noun. For example, the top definition for whoopsy is “[a] screaming metrosexual, who uses a female shield (pretend girlfriend) to cover up his secret obsession with men” [7 June 2011, Waz Goody2011]. It appears that the sexual                                                  92 Anderson (1998: 126) writes that wowee! is an “intensified variant” of wow! 93 In established corpora, we find zero instances for both affixed forms in GloWbE and COCA, suggesting extremely rare use. A search on Twitter for whoazers, brings up approximately one hit per week, on average and at most; whoaz, in contrast, is more frequent, with approximately one hit per day. In sum, the two resources fit together to show this affixed form as ‘slanguage’ (broadly defined as ‘slangy lexis’ typically used by youth (Zappavigna 2012)). The affixed form is considerably rare and a neologism, but not (quite) an idiosyncratic nonce form. Thus, for the purposes of this study, such a form would be included as an affixed form in English. Other forms, such as whoaee, do not appear in Urban Dictionary, but are relatively frequent on Twitter; likewise, these also are included. 104  connotations of the noun are focused on by users of Urban Dictionary, and users of the site seem to find that the affixed and unaffixed form of the interjection oops! are similar, if not the same. The interjection eek!, defined as “[a]n expression of unpleasant revelation” [17 January 2003, cruznsquaw], is stereotypically associated with the image of a small girl surprised by seeing a mouse because of its sound symbolism. It appears in Urban Dictionary with several diminutive suffixes that mimic the lengthened and high-pitched /i/ sound in the base interjection. The affixes that can be added here are -s, -y, -s + -ie and -s + -ie + -s. The affixed forms include eeks!, a “[m]ild expletive. Often used by girls when surprised” [2 January 2009, ToniReal], eeky!, “[a]n expression showing excitement, surprise and/or joy. Must always be expressed in a loud, high pitched voice” [12 March 2007, Sunshine & Simon]), and eeksie!, “1. a mild obscene, profane words [sic!] 2. a cuter way of saying eeks” [24 January 2010, primaverabebe]. While not listed in Urban Dictionary, users of Twitter sometimes use eeksies!94 Therefore, emphasis is placed on (positive) surprise, ‘high pitched voice’, the features [+SMALL], [+CUTE], [+AFFECTION], and ‘used by girls’ [+FEMALE].  The interjection yuck!, defined as “something gross” [22 March 2003, Aaron W.], appears mainly with the suffixes -ola and -poo. The definition of affixed yuckola! suggests [+EXTREME] in “[s]omething that is extremely disgusting and or disturbing. Normally used by preps”95 [30 October 2006, That one crazy kid]. However yucky-o-la! suggests the feature [+CUTE] because “'o-la' is meant to soften the word, or make it a little bit cute sounding” [21 January 2006, Aistra Oleander]. The affixed interjection yuckapoo! is defined as “[a]n expression often used to categorize a person, thing, or event as unpleasant or undesirable” [25 February 2008, Stanley P. Tencza, Jr.], while yucky-poo! Is defined as “an exclamation of dislike or distaste when one is informed that they must perform a tedious, time-consuming or difficult task. More commonly used sympathetically when others tell of tasks they dread doing soon” [25 November 2006, Jake & David]. These definition do not expressly differ from that of the base yuck!, but the fact that the user’s example uses yuckapoo! in capital letters does seem to suggest an intensifying effect.                                                  94 Urban Dictionary also has an entry for eeko!, which is supposedly a synonym for ‘yikes’. However, it is substantially rarer than affixed forms of eek! with a diminutive suffix. 95 The definition comments on the social aspect by mentioning that yuckola! is typically used by preps, i.e. a wealthy class of teens that attend prepatory school on the East Coast of the US in preparation for an Ivy League university (cf. ‘prep’, 27 December 2003, anonymous). 105  There are other affixed forms of yuck that do not appear in Urban Dictionary, or are disqualified because of a low rating. The diminutive interjection yucky!, defined as “something gross” [6 December 2002, Aaron W.], is disqualified because of 441 thumbs-down votes to 163 thumbs-up. Likewise, yuckers!, defined as “[a] fun word thats [sic] used instead of ‘ew’ or just plain ole ‘yuck’” [18 June 2009, CintaaStarr] receives 8 down-votes to 7 up-votes. This suggests that Urban Dictionary users likely take issue with either the words or their definitions. The suffixes -o, -sy, and -ie also appear, but only appear for the adjective yuck, e.g. “not feeling good today” [2 March 2005, LuCkiEjOaN].  In sum, entries on Urban Dictionary paint a picture of the popular meanings ascribed to affixed primary emotive interjections in English that range from the frequent semantic feature [+INTENSE] to the less frequent [+CUTE], [+FEMALE], and [+CHILD]. Although several definitions were similar, or the same, as the unaffixed counterpart, others implied a foundational idea that the suffix(es) used have intensifying functions that are ‘new’, ‘fun’, and ‘cool’ in popular usage of the site’s younger demographic. However, this is expected, as newer forms of words generally are formed or used for the purpose of expressing stronger attitude and emotion from a form that seems outdated and without its original emotional strength. In the case of emotive interjections which primarily convey emotion, it is not surprising that the affixed forms intensify the emotion (e.g. disgust, surprise) of the unaffixed forms. The number of and types of affixed interjections in Urban Dictionary is not dissimilar from the frequencies found in standard corpora. Standard corpora do not typically have many primary emotive affixed interjections. The affixed interjections with at least one occurence in at least one corpus are presented in Table 1.96 In the Table (and the following Tables in the present chapter), the raw frequency is the first number, and the normalized frequency is presented in square brackets if the number is above zero. For comparison, the frequency for the base interjection is included as well.97 Here, the GloWbE corpus is most likely to contain an occurrence of an affixed interjection because of its huge size at 1.9 billion words. The results for affixed yummy! are                                                  96 This table, and all following tables, provide raw frequencies and normalized frequencies. See Chapter 3 for information about each corpus. 97 Numbers with an asterisk are all the instances for the term and likely include instances from other parts of speech, while numbers in brackets are the (approximate) number of interjections from a manual count. 106  inconclusive because of the many occurrences of the adjective in the search results; the affix creates a frequently-used secondary interjection with the same form as an adjective. The least frequent affixed forms that appear in at least one corpus are affixed forms of ouch! In addition, yayee! has one instance in GloWbE.  107     ENGLISH Primary interjection  COCA; GloWbE  Suffixed interjection  BNC  COCA  SOAP  Strathy  GloWbE eek 25 [4.46]; 1133* eek-s 0 2 [0.35] 0 2 [0.4] 28 [1.4] oops 1119* [199.8]; 6208* oops-ie oops-y 0 1 [1] 7 [1.25] 6 [1.07] 21 [21] 8 [8] 0 0 28 [1.4] 21 [1.1] whoops 282* [50.3]; 1135* whoops-ie whoops-y 0 0 6 [1.07] 1 [0.17] 13 [13] 3 [3] 0 0 37 [1.94] 4 [0.2] ouch 983 [175.5]; 3468 ouch-y ouch-ie ouch-ee ouch-ie-s 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 11 [0.57] 8 [0.42] 1 [0.05] 9 [0.47] ow 2318 [413.9]; 18 ow-ie 0 2 [0.35] 6 [6] 0 6 [0.31] wow 13859* [2474.8], 57616 wow-ie wow-ee wow-za wow-sa wow-sie 0 6 [6] 0 0 0 2 [0.35] 8 [1.14] 3 [0.5] 0 0 2 [2] 6 [6] 3 [3] 0 0 0 0 0 0 0 15 [0.78] 73 [3.84] 121 [6.36] 15 [0.78] 1 [0.05] yuck 360* [64.2]; 1486* yuck-s yuck-y yuck-o yuck-ola 0 6 [6] 0 0 0 8 [1.42] 0 0 0 2 [2] 0 0 0 0 0 0 28 [0.14] 14 [0.73] 3 [0.15] 2 [0.10] yum 615 [109.8]; 4445 yumm-y 41* (24) 703*  315* 29* (5) 7691* Table 1: English primary emotive interjections 108   4.2.2 Polish The Polish language does not have specific interjections such as, e.g. oops!, whoops!, ouch! or ow! Instead, these have been borrowed into the language as ups! ‘oops’, auć! ‘ouch’ and au! (or ał!) ‘ow’. It is rare for the (relatively recent) borrowings to receive suffixes, though there is some evidence for ups! > upsik! or upski! ‘oops’ and fuj! > fujka! ‘yuck’. Rather, the multi-functional oj(ej)! is used in contexts such as of e.g. pain, accidents, and surprise. In fact, a user of Miejski defines ojej! as “1. Gdy coś sie nam nie uda 2. Gdy kogoś chcemy ośmieszyć” (‘1.When something goes wrong 2. When we want to make fun of someone’) [26 October 2006]. The affixed forms of ojej! or other interjections do not have any entries on Miejski with the exception of fujka!, which is defined as a synonym for fuj ble [3 November 2009, Młodsza], which can be best translated as ‘ugh yuck’.  As in English, the primary interjections o! and och! ‘oh’, oj! and aj! ‘oy’, ech! ‘eh’, uf! ‘ooph’ and ach! ‘ah’ do not receive suffixes to create *ochku! or *achki!; neither are there any instances for affixed interjections in Miejski or the established corpora, such as fe! ‘expression of disgust’, tfu! ‘expression of disgust, like spitting’, ba! ‘expression of disgust’. Unfortunately, none of the affixed forms of primary emotive interjections can be glossed according to their semantic and pragmatic features because of their rare occurrences in Miejski, and the fact that definitions on Miejski tend to be shorter and less informative than those on Urban Dictionary. Instead, we have to turn to other analyses (see Lockyer 2016) and the semantics of Polish diminutive suffixes to posit features of [+WARM], [+AFFECTIONATE] and an intensifying function. Table 2 only presents the two main affixed forms of the interjection (o)jej! It also shows that (o)jejku! is used more frequently than the palatalized form (o)jejciu! The diminutive forms fujka! and upsik! do not have any occurances in the NKJP or Polish Web 2012.109   POLISH Primary interjection  English translation  NKJP  Suffixed interjection  Suffix(es)  NKJP  Polish Web 2012 jej98 wow!, gee -- jejku jejciu jejcia -ku -ciu -cia 1894 [124.2] 166 [10.8] 4 [2.62] 18771 [200] 1680 [23] 76 [1] ojej oh wow!, oh geez, oh dear 3817 [250.3] ojejku ojejciu -ku -ciu 547 [35.8] 29 [1.90] 895 [32] 250 [3] Table 2: Polish primary emotive interjections                                                  98 The unaffixed form jej is also the word for ‘her’ and is extremely common; thus, I do not include the frequency tokens for jej in Table 2. 110   4.3 Contact interjections When English contact interjections receive diminutive suffixes, they often come across as child-speech or silly nonce forms. Slang suffixes are more frequently used in English because they indicate the social and age-related linguistic identities of the speaker (e.g. laters! < later! < see you later!) instead of associations with children and/or ‘cute’ things. In Polish, diminutive contact interjections are used to create positive social interactions across a variety of situations.  4.3.1 Greetings and farewells English The affixed forms of English contact interjections rarely appear in Urban Dictionary or established dictionaries. For this reason, I do not include a table that lists their frequencies from established corpora. The GloWbE corpus produces no instances for the affixed byesies!, morningsies!, nightsies!, nighto!, heysies!, hellosies!, or afternoonsies!; they only appear, infrequently at best, on Twitter and a few forums and blogs from a Google search. Only hiyee! appears in Urban Dictionary.99 Several affixed forms – especially morningsies! and afternoonsies! – appear on Twitter just enough to be included in the list of English affixed interjections, but also could be viewed as too rare to be included. When affixed, English contact interjections tend to receive the -ie and -s diminutive suffixes, though these do not have any instances in the BNC or COCA. Only several contact interjections with the first-degree -y suffix appear in standard corpora. One hit of nighty night! is in the BNC, five in COCA and 11 in SOAP; furthermore, there are three instances of laters! in the BNC and two in COCA.100 Interjections with the -s suffix are typically found in Urban Dictionary. Otherwise, only the affixed byes! and nighty-night! appear in Urban Dictionary.                                                   99 The above seven affixed forms were found through an extensive search of each possible affixed form on Twitter. In the selection process, I only included valid neologisms, not idiosyncratic nonce formations that are nearly non-existent; thus, I only look at forms with the newest tweet posted during the past year and which have at least several instances per month. 100 Surprisingly none in SOAP, perhaps because it seems British. 111  Three affixed contact interjections appear on GloWbE: laters! produces 123 instances, 38 for nighty night!, and 1 for hiyee! They are also the only affixed contact interjections of greeting or farewell that appear in Urban Dictionary. Here, hiyee! suggests heightened emotional force and the semantic feature [+INTENSE] in its definition as “[it] means hiiiiiiii-eeeeeeeeee. Pronounced: Hi-eee. A hyper sense of Hi” [5 April 2011, Zscar]. In comparison, nighty-night! is defined as “Good night! Usually said when it's late into the night” [20 October 2006, Peter Lu] and “[a] cute way of saying good night to somebody or group of people” [17 February 2015, sikknastyy], with the example sentence Nighty night, sleep well now. :). These two entries suggest that nighty night! is said later at night than the unaffixed ‘(good)night’ or ‘night night’ and has the semantic features of [+CUTE] and [+GOOD HUMOUR]. The more frequently-used slang suffixed laters! is defined in Urban Dictionary as “[t]alk or see you later. Talk to you laters” [11 October 2005, KristinaJ] with the example Like when you say goodbye to a friend you would say "laters" dude that suggests social group bonding.  In sum, contact interjections of greeting and farewell in English are sporadic at best, and can be seen to be ‘cute’ expressions, to intensify the meaning to rather humorous levels, and/or to maintain positive solidarity between friends or a group. Polish Polish affixed contact interjections of greeting and farewell are significantly more numerous than their English counterparts.101 They include miłego ranka! ‘good morning’, dobry wieczorek! ‘good evening’, dobranocki/dobranocka! ‘goodnight’, hejcia/hejka! ‘hey’, and papatki! ‘bye’. None of these appear on Miejski or other dictionaries (except for siemka!, siemanko!, siemaneczka!), which suggests that they are not standard diminutives or slangy enough to merit an entry on Miejski. Table 3 provides a list of the Polish contact interjections that are affixed, their rough English translation, the raw and normalized frequencies of the unaffixed interjection in the NKJP and Polish Web 2012 corpus, and the frequency of the suffixed form. The affixed forms are rarer than                                                  101 In general, they are shared across the West Slavic languages, specifically Slovak and Czech, but also are found in other Slavic languages. In Russian, privetik! (< privet! ‘hi’) is generally considered less formal and childish than privet! Also, Czech and Slovak have dobre ranko! ‘good morning’, which does not appear in Polish. 112  their unaffixed forms by a fair amount except for siema! Only three affixed interjections do not have any instances in the NKJP: cześcik!, witka!, and eloszka! Overall, siemka! and hejka! have the highest raw and normalized frequencies of all affixed interjections.113   POLISH Contact interjection  English translation  NKJP  Suffixed interjection  NKJP  Polish Web 2012 Witam  welcome (lit. ‘I welcome you’) 147289 [9660] witan-ko wit-ka wit-ecz-ka 51 [3.34] -- 1 [0.06] 711 [8] --102 -- Dzień dobry good morning/afternoon, hello 13443 [881.6] dzieńdober-ek 50 [3.27] 525 [6] Elo  hello (< engl. hello) 1974 [129.4] elo-sz-ka 0 107 [1] Dobry wieczór  good evening 1681 [110.2] miłego wieczor-ku dobry wieczor-ek 24 [1.57] 3 [0.19] 996 [11] 110 [1] Dobranoc  goodnight 7432 [487.4] dobranoc-ka miłej noc-ki 429 [28.13] 36 [2.36] -- 772 [8] Cześć  hi / bye 68791 [4511] ? cześc-ik 0 8 [0.00] Hej103 hey 38408 [2519] hej-ka hej-a hej-o hej-cia hej-ecz-ka hejc-io 3950 [259.06] 747 [48.