UBC Theses and Dissertations

UBC Theses Logo

UBC Theses and Dissertations

Realism, not relativism : a critique of Gilbert Harman Rawlings, Adam D.H. 2003

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

Item Metadata


831-ubc_2004-0043.pdf [ 3.54MB ]
JSON: 831-1.0091327.json
JSON-LD: 831-1.0091327-ld.json
RDF/XML (Pretty): 831-1.0091327-rdf.xml
RDF/JSON: 831-1.0091327-rdf.json
Turtle: 831-1.0091327-turtle.txt
N-Triples: 831-1.0091327-rdf-ntriples.txt
Original Record: 831-1.0091327-source.json
Full Text

Full Text

REALISM, NOT RELATIVISM: A CRITIQUE OF GILBERT H A R M A N 1 by A D A M D. H. RAWLINGS B.A. , The University of British Columbia, 2001 A THESIS SUBMITTED IN PARTIAL F U L F I L M E N T OF THE REQUIREMENTS FOR THE D E G R E E OF MASTER OF ARTS in THE F A C U L T Y OF G R A D U A T E STUDIES (Department of Philosophy) We accept this thesis as conforming to the required standard THE UNIVERSITY OF BRITISH C O L U M B I A November 2003 © Adam D. H. Rawlings 2003 Library Authorization In presenting this thesis in partial fulfillment of the requirements for an advanced degree at the University of British Columbia, I agree that the Library shall make it freely available for reference and study. I further agree that permission for extensive copying of this thesis for scholarly purposes may be granted by the head of my department or by his or her representatives. It is understood that copying or publication of this thesis for financial gain shall not be allowed without my written permission. /(-CIAW, ( a v w / l ' S . ^ 7^/(7/ Name of Author (please print) Date Title of Thesis: Ic^lfl^, Vw.f (iflAl C- . l W / IL,-i-Degree: A U k - u f hrh Year: ZOCJ<-\ A B S T R A C T This thesis will critique Gilbert Harman's moral relativism. Harman argues for a form of moral relativism he calls a "conventionalist" account of morality. He supports this by defending a view of explanations, a view of simplicity, and a view of the moral "ought." However, the anthropological literature contains strong evidence against his drawing of this relativist conclusion — and in support of a contrary one. According to anthropologists, there is a universal belief in the moral wrongness of incest, the "incest taboo": its existence suggests that Harman may have better supported a form of moral realism than the relativism he endorses. Thus, at the very least, Harman's argument does not prove that relativism is true; more strongly, it may prove that relativism is false, and realism true. i i T A B L E O F C O N T E N T S Abstract u Table of Contents i i i I Realism, Relativism, and Moral Observations 1 1 Introduction 1 2 Definitions 1 3 Outline 3 II Moral Oughts and Best Explanations 5 1 Introduction 5 2 Explanations 5 3 Inner Judgements and "Oughts" 10 4 Moral Conventionalism 15 III Incest Taboos and Moral Reality 24 1 Introduction 24 2 Data 24 3 Harmanian Realism 32 IV Further Issues and Responses • 36 1 Introduction 36 2 Roman Egypt 37 3 Inbreeding Theory 48 4 Universal Slavery 53 5 Conclusions 56 Endnotes 59 Works Cited 65 ii i I: R E A L I S M , R E L A T I V I S M , A N D M O R A L OBSERVATIONS 1: Introduction This thesis will critique Gilbert Harman's moral relativism. Harman argues for a form of moral relativism he calls a "conventionalist" account of morality. He supports this by defending a view of explanations, a view of simplicity, and a view of the moral "ought." However, the anthropological literature contains strong evidence against his drawing of this relativist conclusion — and in support of a contrary one. According to anthropologists, there is a universal belief in the moral wrongness of incest, the "incest taboo": its existence suggests that Harman may have better supported a form of moral realism than the relativism he endorses. Thus, at the very least, Harman's argument does not prove that relativism is true; more strongly, it may prove that relativism is false, and realism true. 2: Definitions Moral realism and relativism are particular species of more general views. So, to begin, the general realism and relativism will be defined in the following ways. First, realism is the view that: a, b, and c and so on exist, and the fact that they exist and have properties such as F-ness, G-ness, and H-ness is (apart from mundane empirical dependencies of the sort sometimes encountered in everyday life) independent of anyone's beliefs, linguistic practices, conceptual schemes, and so on (Miller 2002). Realism can thus be specified in several ways. Take a sort of "common-sense" realism about ordinary objects. Ordinary objects — such as apples (a), buildings (b), and chairs (c) — 1 exist, and the fact that they exist and have properties — such as firmness (F-ness), girth (G-ness), and height (H-ness) — is (apart from mundane empirical dependencies of the sort sometimes encountered in everyday life) independent of anyone's beliefs, linguistic practices, conceptual schemes, etc. This particular realism is probably quite sensible, at least for most objects and most properties. Moral realism, though a more limited sort of realism than the "common-sense" realism, is also a species of realism in general. Moral facts — such as "stealing is wrong" (a), "feeding your children is right" (b), "murder is wrong" (c) — exist, and the fact that they exist and have properties — such as being true (F-ness), and being false (G-ness) — is (apart from mundane empirical dependencies of the sort sometimes encountered in everyday life) independent of anyone's beliefs, linguistic practices, conceptual schemes, etc. To avoid needless technicality, though, for the purposes of this thesis, moral realism will be defined more simply as the view that moral facts exist independently of people. Relativism, by contrast, can be classified with: the relativistic schema Y is relative to X. and so involves three things: 1. Y: The thing that is relative 2. X: what it's relative to 3. Type of connection between X and Y (reflected in the phrase is relative to) A complete characterization of a species of relativism requires specification of all three things, telling how changes in one or more independent variables lead to changes in one or more dependent variables (Swoyer 2 0 0 2 ) . The relativistic schema can be specified in the following way. Take the uncontroversially 2 relativist Protagorean maxim, "Man is the measure of all things." Whatever else this may mean, it at least could mean that all things (Y) are created by (is relative to) people (X). In this view, then, all experiences people believe they have, all beliefs people have, all objects people seem to encounter, are actually creations — conscious and otherwise — of people themselves. Everything is relative to people. This is, of course, a very extreme sort of relativism — and likely a superficial reading of Protagoras — but it serves to show how relativism in general can be specified in the views of a particular relativist. Harman's moral relativism, discussed in much more detail below, is a much more moderate view, but is nonetheless a species of relativism in general. In his view, all moral facts require that there be some sort of understood, tacit contract between people. Thus, moral facts (Y) depend on implicit agreements formed between (is relative to) people (X). 3: Outline Chapter 2 will outline Harman's views in more depth, tracing his conclusions from his account of explanations, pragmatic simplicity, and the moral "ought." The emphasis will be on discussing the connections between these various views and how they lead to his conventionalist view of morality. Chapter 3 will discuss the anthropological data on the incest taboo, and how the existence of a universal belief in the wrongness of incest shows that Harman's arguments may actually support moral realism. Chapter 4 will consider three possible counter-arguments Harman, or a defender of 3 Harman, might make to Chapter 3's claims. First, that the belief in the wrongness of incest is not universal. Second, that the belief, even i f universal, does not damage Harman's relativism. And third, that the reasoning used to support the realist conclusion in Chapter 3 is dangerously flawed. It will be argued that all of these fail to help Harman support his relativist conclusions, and Harman has, therefore, better argued the truth of moral realism than that of moral relativism. 4 II: M O R A L OUGHTS A N D B E S T E X P L A N A T I O N S 1: Introduction Harman holds that the truth or falsity of moral observations depends, essentially, on people. Harman's view stems largely from his account of explanations. This account leads him to conclude that moral observations must be explained by beliefs, not facts. And this is, he argues, due to beliefs being implicitly referenced by moral claims, through the moral "ought." Harman holds that moral conventionalism follows from these claims. 2: Explanations Harman's view of explanations is rather layered and complex, including an account of "best" explanations and an unusual definition of "simplicity." It begins with the seemingly3 simple claim that there is an ambiguity in the word "explanation," as it is ordinarily used (Harman 1968, 530). In the statement, "that water is wet explains why one observes water to be wet," it may be that "that water is wet" alone is an explanation — it explains the observation4. It may also be that the whole statement is an explanation — it expresses an explanatory claim. In both senses, though, there may be several alternative explanations. In the former sense, "that water is wet" may explain why one observes water to be wet, but "that one's senses are functioning properly" may also explain the same observation. In the latter sense, "that water is wet explains why one observes water to be wet" is one possible explanation; "that one's senses are functioning properly explains why one observes water to be wet" is another. 5 According to Harman, one of several alternative explanations may, in a particular situation, be the "best" explanation. The best explanation should always be preferred, in his view, as "inference to the best explanation is ultimately inference to the best total explanatory picture of the world" (Ibid., 533). That is, inferring the best explanation is of overall benefit to understanding and knowing about the world. However, for Harman, there can only be a best explanation in cases where multiple explanations conflict. The "best" explanation is the best of competitors, not merely of alternatives — non-conflicting explanations may be simultaneously true, and thus equally "best" (Ibid., 530). In cases where explanations conflict, Harman holds that the best explanation is the simplest explanation. Sometimes we infer an explanation on the grounds that competing explanations are too complicated, ad hoc, etc.... Sometimes we infer that something we now know about wil l lead to, and hence explain, something else, on the grounds that to assume otherwise leads to too much explanatory complication.... Sometimes we even give up something previously accepted, on the same grounds: reduction of explanatory complication... (Ibid., 533). Harman has, however, a rather idiosyncratic view of simplicity, one he terms a "pragmatic" or "computational" account (Harman 2000c, 83). Under this view, a hypothesis is simpler than a competitor i f it is easier to use in answering the questions that there is interest in answering (Ibid., 84). This usefulness in answering questions must not, Harman insists, be measured in terms of usefulness in extrapolating new observations, or in accurately describing current observations. The relevant sort of usefulness is usefulness in increasing understanding of data (Ibid., 87) — and thus increasing the ease of answering questions at interest. So, for example, i f there is interest in answering questions regarding the nature of water, then explanations referring to a person's sensory apparatus may not lead to as much understanding of water as explanations referring to facts about water. Thus, the explanation, "water is wet," may be the simplest, and 6 thus best, explanation of one's observation that water is wet. If, however, there is interest in answering questions regarding a person's neurological state, then explanations referring to the person's sensory apparatus may be better than explanations referring to wetness. Thus, the explanation, "one's senses are functioning correctly," may be the simplest, and thus best, explanation of one's observation that water is wet. Beyond these claims, Harman leaves the exact nature of "ease of use" unstated. This is quite deliberate, though, for: "[w]hether a hypothesis is easier... to use depends on what it is to be used for" (Ibid., 84). Which questions a hypothesis is supposed to help answer will determine, at least in part, if 'the hypothesis is easier or harder to use, and thus more or less simple, than competing alternatives.5 Harman argues for relativism from this view of explanations. He first claims there is an explanatory distinction between science and ethics.6 In the case of science, according to Harman, observations cannot be explained by appeal only to beliefs of the person. At least some scientific observations must be explained by assuming certain facts. For example, says Harman, suppose a physicist is performing an experiment that produces certain patterns in a cloud chamber. The physicist observes these patterns and makes the observation, "that's a proton." On Harman's view, the physicist's observation is explained by the physicist's beliefs — about protons, cloud chambers, and the behaviour of protons in cloud chambers — and also by there really being a proton in what really is a cloud chamber (Harman 1977, 6). This is because there is something in the (physical)7 world that is an ineliminable cause of the physicist's observation. By contrast, no appeal to any sort of fact is, for Harman, ever needed to explain ethical observations. Suppose that an ordinary person were to round a corner, and suddenly see several 7 boys setting a cat on fire (Ibid., 4). This person makes the observation, "that's wrong." In Harman's view, this moral observation can be explained solely by reference to the causal mechanisms of the psychology of the observing person (Ibid., 6). Since no other causes of the observation need to be hypothesized, Harman argues, it can be concluded that moral observations arise solely out of the beliefs held by people. As Harman puts it, the assumption of moral "facts" is "totally irrelevant" to the explanation of moral observations (Ibid., 7). For Harman, the source of this distinction between the explanations of scientific and moral observations lies in the causal inefficacy of moral facts. In the case of the physicist observing the cloud chamber, some account is available of how the proton assumed to be present in the chamber causally interacts with the chamber and thus with the sensory apparatus of the physicist in order to produce his observation, "that's a proton." There is, Harman claims, similarly no mystery about how psychological causal mechanisms might bring about moral observations (Ibid., 6-7). But, he argues, no sufficient causal account for how either non-reducible or reducible moral facts might do so is available. Thus, he concludes that moral facts cannot be taken to have any effects, including moral observations (Harman 1986, 59-67). Urrfortunately, Harman presents little evidence to support his insistence that a causal account of moral facts be required in order to suppose that they can produce moral observations. To take a simpler example, suppose a person is watching television, and making certain observations about the program. Whether or not the person watching television — or any other person — has some causal account of how the television-watcher makes these observations is really beside the point. The television-watcher is making these observations. It could be argued, though, that in the moral case there is some significant disanalogy — 8 something like: morality and moral observations are so much more complex than television and television observations that a causal account is needed in order to be reasonably sure putative moral observations are really moral observations. Or perhaps there is some good reason to suppose the person watching television may be deluded in some way, and their observations are faulty: perhaps their senses are being deceived by a clever holographic device that projects not only the image of a television, but the images on the television; or perhaps they are simply hallucinating, or dreaming. And so, due to the many ways observation can go awry, a causal account is always needed in order to demonstrate that observations are reliable, in any case. There are clearly many complex questions about causality and observation lying within this issue. For example: How reliable are our senses in ordinary cases (such as watching television)? In scientific cases (such as observing cloud-chambers)? In moral cases (such as seeing boys setting cats on fire)? How full a causal account is needed before observations can be considered reliable in ordinary cases? In scientific cases? In moral cases? Is observation usually presumed to be reliable, until proven otherwise; or unreliable, until proven otherwise? In ordinary cases? In scientific cases? In moral cases? These, and many, many others are questions which Harman does not address here, but, perhaps, should. On the other hand, the particular matter of having or not having a causal account is one Harman can, within the context of his overall argument, ignore, in favour of a broader issue. Harman raises the question, as elaborated above, of whether it is beliefs or facts that explain scientific and moral observations. This could be settled, on Harman's view, by determining which is the best of these two explanations — which are, it should be noted, explanations in Harman's first sense, as they (purport to) explain observations. To do this, no causal considerations need be offered or entertained. The best explanation is just the explanation that is 9 pragmatically simpler. Harman could therefore draw his stated conclusions in both cases — that moral observations are best explained by the beliefs of the observer, while (at least some) scientific observations must be explained with reference to facts — i f he holds that the questions of interest in ethics render beliefs to be a simpler explanation of ethical observations than facts, while the questions of interest in science render facts to be a simpler explanation of scientific observations than beliefs. 3: Inner Judgements and "Oughts" It should be noted immediately that it is, of course, very rare — and not only in ethics — for only one question to be of interest in any given case of observation. One question may be of immediate interest, but only in that answering it leads to further knowledge. For example, a physicist may wish to know if the particle in a cloud-chamber is a proton. But it is likely that he only wishes to answer this question in order to advance his understanding of fundamental particles, so that he may answer many other questions, not only about protons in cloud-chambers, but about many different areas of science. In the same way, a person may wish to know what is the right thing to do now, not just to decide on an immediate course of action, but also to decide on courses of action in similar situations in the future. So, to be fully accurate, in order to determine which explanation is, under Harman's account, simpler, it must first be determined which questions are of interest overall, and then why these questions render beliefs to be the best explanation of moral observations. This detail does not seem, however, to add much to the discussion at hand beyond complicated technical maneuvering; any discussion that assumes only one question is of interest could be extended systematically to encompass the possibility that many questions are of interest. Thus, the discussion below will consider only the questions of immediate interest, and show that, in the moral case at least, Harman has sufficient 10 argument to claim that beliefs are a better explanation of observations than facts, causal considerations left entirely aside.9 In Harman's view, there is an often-unappreciated feature10 of the "ought" used in certain moral judgements, a feature which accounts for all questions that are of interest in morality rendering beliefs to be the best explanation of moral observations. Thus, beliefs are always the best explanation of moral observations. Harman thinks that the "ought" of morality can be distinguished from several other senses of "ought": the moral "ought" expresses that one ought to do what is morally right, and ought not to do what is morally wrong, rather than making any statement about what it is prudent for one to do, or preferable for one to do, or expected that one do (Harman 2000a, 6). Harman also tMnks the moral "ought" can be used in two different ways in two sorts of moral judgements11, what Harman calls "inner judgements," and what shall be called "other 1 0 judgements," as Harman neglects to give these a particular name (Ibid., 8). Inner judgements have two important characteristics. First, they imply that the [person]1 ... has reasons to do something. Second, the speaker in some sense endorses these reasons and supposes that the audience also endorses them.... [For example, if] someone S says that A (morally) ought to do D, S implies that A has reasons to do D and S endorses those reasons.... Other moral judgements about... [a person], on the other hand, do not imply that [the person]... has reasons for acting that are endorsed by the speaker.... [For example, i f someone] S says that B was evil in what B did, 51 does not imply that the reasons S would endorse for not doing what B did were reasons for B not to do that thing. In fact, S implies that such reasons were not reasons for B (Ibid., 8). So, to say that a particular person ought not to kill is always, in Harman's view, to make an inner judgement about them. Consider the judgement, "Jean ought not to kil l Paul." 11 Harman holds that a speaker making this judgement implies, rightly or wrongly, that Jean has some reason not to kill Paul, that the speaker accepts these reasons, and that the speaker supposes whoever he speaks to also shares these reasons.14 But, one might also say that killing a certain person is evil, entirely apart of who ought not to do it. That is, one might say, "it would be evil for Jean to kill Paul." This is not an inner judgement — it is an other judgement — for Jean is not being told not to kil l , nor that he should not kil l . The judgement makes no implications about reasons that Jean does have (Ibid., 9 ) . Rather, according to Harman, it must imply that Jean doesn't have the reasons that the speaker does — else it would be an inner judgement.15 Confusingly, though, Harman seems to waver on whether "other judgements" count as moral judgements or not. For example, he states that the "other judgements" do not "represent a full-fledged moral judgement on the part of the speaker. In such a case, we have an example of what [R. M.] Hare calls an inverted-commas use of'ought'" (Ibid., 10). And yet he also seems to hold that inner judgements can be distinguished from other judgements, as shown in the citation above. To be fair, then, Harman's view should probably be interpreted as holding that "other judgements" are moral judgements in a bare, technical sense, while "inner judgements" are moral judgements in a fuller, more robust sense. But, since Harman also doesn't seem to take "other judgements" very seriously when considering moral judgements, the focus from here on in this discussion will be on inner judgements exclusively. The "ought" used in inner judgements is, according to Harman, a four-place relation of the form "ought (A, D, C, M)" — wherein^ is a person,16 D is an act-type (such as stealing or murder or charity), C is a set of considerations (such as "considering A made a promise", or "all things considered"), and M is A's set of motivating attitudes, at least at the time the judgement is 12 made. A statement such as "A ought to D" is thus equivalent to "given that^ has motivating attitudes M, and given C, D is the course of action for A that is supported by the best reasons" (Harman 2000a, 10). C and Mare, he notes, usually given contextually, and not explicitly. For example, consider the claim, "that boy ought not to have set the cat on fire". Here, A is clearly the boy, and D not setting the cat on fire. 1 7 Thus, for Harman, this claim would be equivalent to "given that that boy has a certain set of motivating attitudes M, and given certain circumstances C, not setting the cat on fire is the course of action for that boy that is supported by the best reasons". Presuming that the child is a fairly normal child in an ordinary setting, M might be "to be a good boy" or "to not make his mother angry," and C might be "all things considered." So, "that boy ought not to have set the cat on fire" is equivalent to "given that the boy wishes to be a good boy, or wishes to not make his mother angry, and all things considered, not setting the cat on fire is the course of action for that boy that is supported by the best reasons." This view of the moral "ought" connects to Harman's views of explanations as follows. Since moral observations employ the moral "ought," they contain implicit reference to reasons for acting (i.e. M). For any particular person, these reasons, according to Harman, stem in some way from principles the person accepts — in other words, from some of their beliefs. "[Principles a person accepts are the principles that give a person reasons to act in various ways...[though] a person [may not] always... [have] sufficient reasons to act in accordance with the principles he or she accepts" (Harman 2000b, 67). Additionally, moral questions, in the same fashion as moral observations, also employ the moral "ought." Harman defines this "ought" as the "ought" relating to whatever one morally should or should not do and, surely, any moral question that might be of interest — "what ought I to do?", "ought I to steal this wallet?", "ought 13 she to have given money to the Salvation Army?" — must employ this "ought" rather than any other. So, the questions of interest implicitly refer to beliefs, through the moral "ought," and moral observations implicitly refer to beliefs, through the moral "ought" and principles accepted by people. Thus it seems clear that the explanation that outlines and elaborates on these implicit beliefs should do most to increase understanding of moral observations; these beliefs are prominent in not only the observations, but the questions underlying the search for a best explanation. To draw from this the conclusion that beliefs are the best explanation of moral observations, Harman must argue in something like the following way: for any x, x will be the best outline or elaboration of itself in all cases where x can outline or elaborate itself. For example, i f x is a thing, x cannot outline anything — x simply is. So, some description or account of x — such as the fact that x — will be the best outline available of x. If, by contrast, x is a fact, x can outline things. So, x is the best outline — in the sense of communicating most understanding about x — of itself. It might be argued that x plus some theory of grammar and semantics might communicate more information. But this information would not increase understanding of the sentence, x. Rather, it would increase understanding of many sentences, or sentences in general. In the same way, beliefs can outline things, granting that beliefs are linguistic; thus, the best outline of a given belief, x, must be x itself, x plus some account of psychology and the mechanisms by which beliefs interact would be enlightening and informative, but would increase understanding of many beliefs or of beliefs in general — not of the given belief, x.19 The other competitor — facts — is not nearly as strong. The moral "ought," according to Harman, contains no reference, explicit or implicit, to facts. A is not a fact — A is a person. D is 14 not a fact — D is an action. M is not a fact — M is a set of motivating attitudes. C is the only possibility and C, though it might seem to be factual, is heavily dependent on^'s psychology. For example, the circumstance, "if A made a promise," makes reference to a decision that the person, A, made, and (presumably) an intention A has to adhere to the promise — else that A made the promise would be irrelevant. Other such examples are easy to construct. Hence, C relies heavily on M; hence, C is not an independent fact that can serve to explain a moral observation. Despite this, C might still explain a moral observation to some extent — the problem is, however, that it seems C cannot explain moral observations as well as the beliefs inherent in moral observations and the questions of interest. After all, there is just no implicit presumption regarding facts in the moral "ought." It is thus easy to see why Harman states that the assumption of any facts is "totally irrelevant" to explaining moral observations (Harman 1977, 7). Even i f facts were able to explain moral observations — and, since they are not implied by the moral "ought", it is hard to see how they might — they are simply not as strong a competitor as beliefs. Hence, because of the "unappreciated" feature of the moral "ought," and its implicit connections to beliefs, beliefs must be the best explanation of moral observations. 4: Moral Conventionalism The structure of the moral "ought" — once again, M i n particular, but also C — and the account of explanations are key for Harman's full development of a conventionalist view of morality.20 Morality is, for Harman, fundamentally a matter of endorsing social conventions. Moral rules, in Harman's view, take the form of either prohibitions or requirements (Ibid., 15 57). These moral prohibitions and requirements are enforced by implicit agreements between people, agreements arrived at by a process of "implicit bargaining and mutual adjustment" (Ibid., 104). This bargaining process, put briefly, is a matter o f compromising in response to others' actions. There are some elements in our currently-held moral beliefs that, Harman argues, suggest they may have arisen through implicit bargaining. For example, he claims that we seem to believe deliberately inflicting harm (as in specifically and directly inflicting harm on others 2 1) is morally worse than withholding aid (as in not giving aid to someone to prevent them from coming to harm ). Harman suggests as an example a doctor who has five critically i l l patients that can be returned to ful l health, i f the doctor is wi l l ing to k i l l one perfectly healthy patient, and redistribute his organs amongst the five i l l (Ibid., 110). Even i f the doctor's actions w i l l have the effect of saving five people by inflicting harm on one, as opposed to letting five die to avoid inflicting harm on one, we would hold, says Harman, that we should let the five die. K i l l i ng the one is the morally worse action. What could account for this, asks Harman, save his theory that this moral belief— harming is worse than not helping — must have arisen from a process of bargaining? A group of unequal people — some strong, some weak, some rich, some poor — w i l l all have an interest in preventing harm being inflicted upon them. However, only the weak and the poor w i l l have an interest in aid being rendered to them whenever needed — the strong can simply take what help they need, and the rich can simply buy it. Further, the strong and the rich, as they have more resources available to them, w i l l be more likely to be called upon to give aid, and less likely to receive it (Ibid., 111). So, as a result of compromise between people, a strong duty to not harm (as little compromise is needed) and a weak duty to help (as compromise is needed between the 16 group consisting of the weak and the poor, and the group consisting of the strong and the rich) will come about. Because of the presence of this process, moral right and wrong must, Harman says, be relative to particular agreements, which he calls "moral framework[s]" (Harman 1998a, 161). A "moral framework" is defined as, a system of moral coordinates — a set of values, standards, and principles conceived on the model of the laws of one or another state. Whether something is right or wrong in relation to a given moral framework is determined by the system together with the facts of the case in something like the way in which whether something is illegal in a given jurisdiction is determined by the laws of that jurisdiction together with the facts of the case (Ibid., 163).23 Many such frameworks are possible, says Harman, but none of these is more "correct" than the others. As Harman puts it, "A's doingXis wrong" is "elliptical" for "in relation to a given moral framework, A's doingXis wrong" (Ibid., 162).24 Since, "A ought to D," is similarly "elliptical" for, "given M and C, A ought to D," it follows that, "A ought to D," is, overall, equivalent to, "given M and C and a given moral framework, A ought to D." This can be simplified by eliminating specific reference to a "given moral framework,", as Mmust contain the motivation to adhere to the agreements that make up the moral framework, and C must contain the condition that A be party to these agreements. The agreements that make up these moral frameworks, Harman says, can be held by many people, or only a few: there is no set number of people that must hold to an agreement in order for it to constitute a moral framework. For instance, one can be subject to certain obligations as part of a family, a workplace, a neighbourhood, a city, or a nation. There is a 17 looming issue here, of how to differentiate groups from individuals in a principled way, but Harman dodges it adroitly.25 He first claims that there is a "limiting case" of a group of only one person. Thus, one may be morally obliged due to personal principles alone (Harman 1977, 113). Harman also allows that there may be a "limiting case" in the other direction, of a group containing all people (Harman 1998b, 209): It is compatible with moral relativism for certain moral claims to hold in relation to all moral frameworks just as certain claims about motion hold in all spatio-temporal frameworks, eg. [sic] that nothing travels faster than the speed of light. Any moral claim that is incompatible with such universal claims would be false in all moral frameworks.... Can a moral relativist agree... that a judgment [sic] whose truth value is the same in all moral frameworks has that truth value absolutely and not merely in relation to one or another such framework? Maybe, although that is in some ways like agreeing that 'The universe is at least as large as' is true absolutely because true in relation to everything (Ibid., 211; emphasis removed). Harman's two "limiting cases" thus imply there is no distinction between groups and individuals save for quantity. A group is nothing over and above an aggregate of individuals. And an individual can be counted, in some sense, as such a group — an aggregate of one, so to speak. A person can be part of many groups, and hence part of many such agreements. It is very likely that these agreements will conflict, even though one might not be aware of such conflicts. But, Harman argues, as one has agreed to be considered part of all these groups, and thus part of all the agreements between members of these groups, one must decide, however one can, which principles are to take precedence when conflicts become apparent. A person's decisions whether or not to agree with a given set of conventions are central to Harman's account. Harman has argued that beliefs, not facts, are the best explanation of moral observations. When coupled with his account of moral frameworks, it must be that these frameworks are, in Harman's view, the best explanation of moral observations. Consider, for 18 example, the observation, "Bob ought not to have stolen the wallet." According to Harman's own views, this must be best explained by whatever will most increase understanding of the observation, through increasing the ease with which we can answer the moral questions that are of interest —for example, "should Bob have stolen the wallet?" Elaborating on and identifying the moral framework — the agreement — that prohibits stealing wallets would do this. Thus, the framework best explains the observation. And, as the framework is a sort of implicit agreement, • 26 particularly on "values, standards, and principles," it at least contains some beliefs"0 held by the parties to the agreement — and, thus, as Harman does conclude, beliefs best explain moral observations. Thus far, then, Harman's views on moral frameworks mesh neatly with his views on explanations and the moral "ought." However, Harman's requirement that we look to frameworks for the relevant moral standards is not yet enough for a full-blown relativism. There must be some way to locate the moral standards within the relevant frameworks, or they cannot be (easily) used to form moral judgements or make moral decisions. After all, the agreements are implicit, which makes it quite possible, without such a method of location, for people to be party to agreements and yet have absolutely no idea what the agreements require they do. For Harman, this boils down to a matter of locating agreements, and thus the conventions they support. As long as this can be done, one can find the standards that must be taken into account when evaluating the truth of moral claims. But this is, he argues, a (fairly) straightforward, sociological/psychological matter — a matter for empirical science. An agreement, in Harman's view, is based on intentions; to agree is to intend to do something. That is, if one agrees to do X then one intends to do X, given that others keep up their ends of this bargain (Harman 2000a, 13). Agreements may be "inconsistent, incoherent, 19 or self-defeating," and thus in need of alteration over time (Ibid., 14). Agreements need not be •yet historical , or conscious, but can exist in implicit convergence of intentions arising through interactions with other individuals (Ibid., 16). These agreements are, however, vague — Harman likens them to the understanding existing between members of a symphony orchestra that allows them to play as a unit (Ibid., 17). In practice, as disputes over what it implies are met and resolved, a given agreement will become less and less vague. The agreements are maintained by vaguely social forces. Society — that is, the individual people in a society — requires the adults to stay within the confines of these agreements (Harman 1977, 61). The agreements are also maintained by children trying to please their parents (Ibid., 60). Their parents have certain expectations, as participants in certain implicit agreements, and children attempt to predict these expectations and live up to them, in order to receive the parents' love and avoid their parents' punishment. Harman postulates that children may even internalize a sort of Freudian "superego," a kind of imaginary parent that has certain expectations, expectations that the child hopes will mirror the actual parents' expectations. The child can either live up to, or violate, these expectations, and the superego will mete out praise and punishment as appropriate, in the form of (broadly) pleasure (such as elation due to success) and pain (such as shame due to failure). However, there are ways these agreements may be altered. The parents may make mistakes with regards to what their children are actually obliged to do. That is, the agreement that the parents believe requires this conduct actually does not (Ibid., 61). Similarly, the children may make mistakes with regards to predicting what their parents expect, and thus make mistakes in trying to shape their conduct to meet parental expectations (Ibid., 61). In both cases, if the mistakes are not corrected, the children will be party to slightly altered agreements — agreements 20 which transmute these mistakes about acceptable behaviour into norms for acceptable behaviour. For example, i f a father mistakenly believes that it is immoral for his son to lie, under any circumstances, the son may take on this mistake as a norm for his future behaviour — that is, the son will come to believe that it is immoral to lie. And the son may be perfectly correct to do so, for the father's mistake has, in effect, created an altered agreement that governs the son. Moreover, people may disagree with the agreement, and press for it to be changed deliberately, by altering their behaviour or asking that expectations of behaviour be altered (Harman 1998a, 164). And, finally, people may disagree about what the agreement implies in any given situation — after all, the agreement is, ex hypoihesi, vague and imprecise (Harman 2000a, 17) — and thus either deliberate in order to decide how the agreement should be interpreted, or "agree to disagree" and create differing agreements. It seems to be in the spirit of Harman's view to hold that the existence of these sorts of agreements gives the appearance of universality to moral claims. Agreements on such matters as honesty may be very widely-held, and so claims that it is wrong to lie may seem universally true. Similarly, agreements on matters such as the immorality of murder may also be very widely-held, and so claims that it is wrong to murder other people may seem universally true. However, under Harman's view, this is only an appearance. If there are some groups of people under agreements, and thus conventions, that do not place a high value on honesty, or on human life, and similarly for other "key" moral claims, then it follows that the claims are not, in fact, universal after all. Anthropologist Ruth Benedict has a few cases that seem to support this, and show that moral claims which seem universally true are, in fact, not. The most significant of these is that of a society living on Dobu, an "island of northwest Melanesia." This society is "built upon traits 21 which we regard as beyond the border of paranoia" (Benedict 1985, 207), but, for its citizenry, are perfectly acceptable, and even laudable. For example, the groups within this society "look upon each other as prime manipulators of black magic," and do not trust gifts of food from another group even during times of starvation and famine (Ibid., 207-8). By contrast, an individual of "sunny, kindly disposition who liked work and liked to be helpful" is considered "silly and simple and definitely crazy" (Ibid., 208). Given the existence of such a society, perhaps it is right to say that the appearance of universality to many of our moral claims is just that — an appearance. There are societies that may diverge from our seemingly universal beliefs, and so there are societies that may be operating with entirely different (even opposite) conventions in place. In summary, Harman holds that moral rules are a sort of socially-maintained convention, implicit agreements between people which Harman terms, "moral frameworks." He supports this claim through an account of explanations, and an analysis of moral observations of the form, "A (a person) ought to D (an act-type)." According to Harman, "A ought to D," is equivalent to, "given M (a set of motivating attitudes) and C (a set of conditions), A ought to D." Harman also holds, though, that understanding of a given moral observation would be most increased by determining the best explanation of the observation. More precisely, he holds that by increasing the ability to answer moral questions that are of interest — such as, "what ought one to do?" — understanding of moral observations will be most increased. A s moral questions will involve the same "ought" as is present in moral observations, A, D, C, and M will also be implicit in moral questions. Thus, elaborating on 4 D, C, or M w i l l best explain, and thus most increase understanding of moral observations. However, plausibly, C contains the fact that A is party to a particular framework, F, and, similarly, M contains A's intention to adhere to F. Therefore, elaborating on F — through identifying the conventional moral rules that make up F — will best 22 e x p l a i n , t h r o u g h m o s t i n c r e a s i n g u n d e r s t a n d i n g , m o r a l o b s e r v a t i o n s . F u r t h e r , as F— as w e l l as C a n d M— i s not a fact but a b e l i e f ( o f a sort) , it i s therefore H a r m a n ' s v i e w , as he has stated, that b e l i e f s best e x p l a i n m o r a l o b s e r v a t i o n s . O v e r a l l , H a r m a n ' s c o n v e n t i o n a l i s t v i e w i s a moral relativism, as the truth o r f a l s i t y o f all m o r a l o b s e r v a t i o n s i s r e l a t i v e to s o m e t h i n g about p e o p l e — n a m e l y t h e i r b e i n g party to c e r t a i n m o r a l f r a m e w o r k s . 23 I l l : INCEST T A B O O S A N D M O R A L R E A L I T Y 1: Introduction It will be argued in this chapter that Harman's argument above is insufficient to support his relativist conclusion. Anthropologists have uncovered a universal belief in the wrongness of incest — the "incest taboo" — which seems to undermine Harman's desired conclusion. That this prohibition exists shows, prima facie, that there is a moral fact that is independent of any framework or agreement between people, a claim that Harman has denied. 2: Data It should be noted that, for simplicity, the focus from here onwards will be on prohibitions against sibling incest, rather than taboos against incest in general. The anthropological literature indicates that conceptions of "sibling" vary slightly between societies. For example, in Tallensi society — discussed in more detail below — one's siblings include not only one's siblings by blood, but also other female relatives of the patrilineal line (Fox 1962, 140). Assuming, reasonably enough, that there is similar variation in conceptions of spouse and child, trying to include all forms of incest would clearly lead to a complicated discussion, as each point made would require several permutations to deal with the shifting conceptions of "sibling," "spouse," and "child." By focusing the discussion instead on just one slightly-shifting concept — "sibling" — it is quite likely that nothing is lost; and, with appropriate adjustments, the conclusions suggested here should apply to incest in general. The following four cases make clear that there exists some sort of universal moral 24 prohibition or restriction on incest. The first is that of an Israeli "kibbutz," Kiryat Yedidim, and particularly a single "kevutza," a "bi-sexual children's peer group" with a common living and sleeping area (Ibid., 136). At young ages, the children in the kevutza lived together without any sort of shame or sexual separation. For example, from ages one to five, the children would sleep in the same rooms, shower together, sit on the toilet together, run around nude in the morning before dressing, and in the evening after undressing, and engage in overt heterosexual behaviour. The most frequent form of this was "a single embrace of one child by another" — not terribly remarkable — but it was followed in frequency by "stroking or caressing, kissing and touching of genitals" (Ibid., 136). As the children age, though, their attitudes change. While on the verge of puberty, the two sexes barely spoke to each other, and fought constantly over insignificant matters. The girls would, contrary to their earlier behaviour, try to conceal their nudity from the boys (Ibid., 137). At age 15, the girls began to take an interest in sex, but did not direct this interest towards the boys in the kevutza, whom they considered "immature, hence, asexual" (Ibid., 137). And after 15, the relations eased between the sexes, possibly due to finding alternative sexual outlets (Ibid., 137). However, even with this easing of tension, "sabras" — age-mates within in the kibbutz — never married each other, nor did they have intercourse with each other. Their cited reason is that they consider their sabras to be siblings. Moreover, only one sabra interviewed by Fox's source expressed an overt sexual interest in other sabras; all the rest denied even this (Ibid., 137). These prohibitions cannot, it seems, have their root in some sort of parental authority, as the parents of the sabras, the founders of the kibbutz, emphasized "sexual freedom" rather than sexual restraint of any kind (Ibid., 137). To go against this freedom, Fox concludes, must be an 25 entirely voluntary action on the part of the sabras — they chose to engage in sexual relations, and to marry, outside the kevutza and the kibbutz (Ibid., 137). Sexual relations between those considered to be siblings are strongly, voluntarily controlled. The second case31 is in quite significant contrast. In Tallensi society, children spend much of their time in play groups of both sexes, including all siblings and half-siblings — as polygamy is apparently prevalent in Tallensi society, a child may have several half- as well as full siblings — at least up to, and probably beyond, the age of 8. In these groups, there is "full freedom of contact, [with] squabbling, fighting, and soothing being common" behaviours (Ibid., 139). Until about age 9 or 10, siblings are thus together for "sleeping, playing, [and] eating" (Ibid., 139). The attitude towards sex in general is relaxed, but in a curious way. It is inappropriate for boys to engage in premarital sexual intercourse with unrelated women, and it is believed that the "only possible relationship a man could want with an unrelated woman is marriage" (Ibid., 140). This reverses in the case of related females. Sexual contact with "sisters," that is, "related females in the patrilineal line" is permitted, but marriage is strictly forbidden (Ibid., 139). However, this sexual contact seems limited to all but intercourse, as evidenced by this passage Fox cites from his source: there is probably a strong and partly overt sexual component in the affective relation of brother and sister. I was struck by this when I once had an opportunity of observing the behaviour of [an 11 year old boy]... and his [real sister, 9 years of age]... who were playing together in a corner of the room where I was chatting to their father. The children, who were quite naked, stood embracing each other, the boy with his legs round his sister's, and they twisted and wriggled about as i f they were engaged in a mixture of an orgiastic dance and a wrestling match. They were both in a state of high excitement, panting and giggling and muttering to each other, with obvious sexual pleasure. They seemed oblivious of their surroundings. This game went on for about twenty minutes, after which they separated and lolled back as i f exhausted. These children were most attached to each other (Ibid., 139-40). 26 Further, Tallensi children are taught that to "play at actual coitus with their 'real' sister... is wrong," and will result in smacks from their parents and ridicule from other children (Ibid., 140). In all other respects, though, complete intimacy of contact is allowed (Ibid., 140). There is no legal or mystical sanction involved in Tallensi beliefs about incest. However, incest is held to be "incompatible with the pattern of co-operation and the structure of disciplinary and affective relations in the family" (Ibid., 140). The Tallensi hold that incest is simply not possible — any man committing incest is "obviously" immature and not in full control of themselves (Ibid., 140). They do not hold that incest is as bad a crime as sex with a "wife of a member of the lineage," but it is still a disgusting act (Ibid., 140). The Tallensi argue incest is "disgraceful, scandalous" and "deny the temptation [to incest] exists" (Ibid., 140). Some internal sanction, something not contained in an overt law or social rule, seems to be limiting incestuous behaviour here. The third case is that of the Pondo. In this society, brother and sister play together; even though the boys go to herd at age six, boys and girls tend to play and gather in groups of their own age and sex (Ibid., 142). Though brothers and sisters may not see each other much, they are still on intimate terms in the home. "Often I have seen a small sister snuggling up to an older brother, keeping warm under the cover of his blanket and getting titbits [sic] from his plates. Even when grown up sisters will sit and chat with their brothers..." (Ibid., 142). There is a high indulgence of premarital sex amongst the Pondo, and "plenty of objects" with which to engage in it; and, although virginity is an ideal at marriage, it is not a deciding factor — there is even a "division of opinion on the examination of girls" to determine this (Ibid., 27 142). Similarly, sanctions against incestuous behaviour are rather mild. There is an old ceremony, no longer performed, which served as punishment for incestuous and adulterous couples alike. The two were forced to eat scorched meat, possibly as some sort of purification, and then "sworn at, exhorted, and told, 'There your filthiness has been exposed'" (Ibid., 142). The belief seems to have changed towards a supernatural sanction, whereby the children of incestuous relations "will not suckle unless they [the parents] confess" (Ibid., 142). So, there seems to be some general belief that incest is not "all right" behaviour, but only at a very mild, almost benign, level. The fourth, and final, case33 is that of the Tikopia. In Tikopia society, there is great freedom of interaction allowed for children, who go unclothed for many years. There is no taboo on bodily contact, and children will even simulate intercourse with each other (Ibid., 144). The brother-sister relation is highly familiar and free, yet, they do not mention sex matters to each other. They may freely take part in all household matters together, cook and eat together, sit together, and sleep side by side under one blanket (Ibid., 145). The intimacy of the brother-sister relation goes beyond that of most other relations amongst the Tikopia, and a sister may see her brother nude, and even dress him, while other woman may not (Ibid., 145). There is a high premium on virginity amongst the Tikopia, though, and brother-sister incest is abhorred. In fact, the occurrence of such incest is outright denied (Ibid., 145). Informants may admit that the temptation towards incest can overwhelm a man, but always phrase it in terms that imply it is a result of a momentary sexual desire, having no relation to some long-held desire for the sister. Intercourse with a sister only occurs i f she is the most immediately available female, not because she is a man's sister (Ibid., 145). 28 Any actual occurrences of incest are generally punished by the people themselves, through suicide (Ibid., 145). And incestuous dreams, which do occur, are taken as a sign of an evil spiritual influence, and the reaction is not shameful, but furious and fearful (Ibid., 145). Fox interprets this anger, plausibly, as the result of a strong temptation to intercourse with a close, opposite-sex adult, which must nonetheless be denied due to some sort of restriction, either in the form of an overt social rule, or something less explicit and more internal, on incestuous behaviour (Ibid., 146). Common to all four is, clearly, some sort of prohibition on committing incest. It might be argued, though, that this prohibition is not a moral prohibition, but of some other kind. Perhaps it is simply irrational to commit incest, or it is not in one's best interests — that is, it is imprudent — to commit incest. However, usually, the prohibition against incestuous behaviour in our own society is taken to be a moral prohibition. Further, i f prohibitions in our own society mirrored the superficial features of the incest prohibitions in the societies examined above, they would surely be taken as moral prohibitions, just as legal prohibitions in other societies are taken to be legal prohibitions if they mirror superficial features of the legal prohibitions in our own society. Consider first, then, the case of Tikopia society, wherein incest is punished by people themselves through suicide. In our society, i f a person were to kil l themselves after committing a certain act — for example, killing their parents — this would normally be taken as evidence of great guilt due to the person's at least believing they had committed a great moral wrong. Similarly, consider the case of Pondo society, wherein incest is believed to be punished supernaturally. In our society, i f a person is supposed to be punished supernaturally — by the Christian God, for example — then this is usually taken as evidence the person has committed some great moral wrong. Homosexuality is the obvious example, with a number of very vocal 29 Christians expressing the belief that homosexuals wil l be punished eternally in Hell for their "sin." Thus, since incest is usually taken to be morally wrong in our own society; and, in addition, prohibitions in our own society which mirror superficial features of the incest prohibitions in the societies described above are usually taken to be moral prohibitions, there is thus good reason to suppose that the incest prohibitions described so far are moral prohibitions. And, therefore, i f these incest prohibitions are but instances of a universal prohibition, then the universal prohibition is itself a moral prohibition. That the four societies described above have many differences between their incest prohibitions is clear enough. It is not "the practice of, or the reaction to, intra-familial intercourse" that is uniform — there is "no uniformity" on this matter (Ibid., 129). Further, there is "no uniformity in the desire for such intercourse" (Ibid., 129). And the "anxiety displayed in the presence of the incest stimulus (thought, dream, suggestion, or occurrence of incest) varies widely" (Ibid., 129). People in Kiryat Yedidim claim to not even consider incest; in Tallensi society, they consider incest disgraceful, and deny the temptation; in Pondo society, they believe in only a supernatural sanction; and in Tikopia society, incest is punished through suicide. And yet, there is something that is shared between all four of these cases, as they all prohibit incest in some way — they seem to share a sort of "moral disapproval" of incest. On the strength of these, and many other, examples, it can thus be concluded that, "[s]exual intercourse between members of the family [such as siblings]... is nowhere condoned as a regular and systematic activity" (Ibid., 129; emphasis added). This prohibition has been called the "incest taboo." The incest taboo cannot, however, be taken as "simply [another] cultural rule... put 'there' [in the person] by the highly intentioned efforts of parents and others..." (Spain 1988, 297), for explicit "rules and laws have often an origin and an autonomy which is 30 divorced from, or only indirectly related to, behaviour which the rules sanction..." (Fox 1962, 131). To take a highly divergent example, current attempts to apply existing copyright and trademark law to the computer software industry demonstrate this point strongly. Advocates of so-called "abandonware" claim that software which is no longer actively produced (i.e. "abandonware") should be freely copied and distributed. Members of the software industry claim this is violating their copyrights and thus falls under "software piracy". However, the laws • in question were brought about in order to protect physical works — books, papers, paintings, and the like — which would endure for some time, long before electronic works, that can be obliterated in a matter of minutes, even existed. Regardless of whether existing copyright and trademark laws should apply to electronic works, it is clear that these laws were not intended to so apply. Regulating the copying and distribution of copies of electronic works is behaviour entirely apart from the laws which are used, or are attempted to be used, to govern this behaviour; so, the behaviour regulated is quite separate from the rule used to regulate it. Thus, instead, the incest taboo is a "tacit rule", a disposition that is "actively created by [an] individual... who appear[s] to be biologically predisposed to create" it (Spain 1988, 297-8; emphasis added). These tacit rules are taken by anthropologist David Spain to be analogous to the rules of grammar, in the following way (Ibid., 298). One implicitly follows rules of grammar of the language one speaks. Clearly there are some rules in place, guiding linguistic behaviour — some expressions are grammatical (e.g. "This is a good sentence"), and some are not (e.g. "sentence good This a is"). But why this is so is extremely difficult to spell out. Similarly, says Spain, there are implicit rules governing other behaviour — rules such as the incest taboo — which cannot easily be stated or formulated, but are nonetheless present. 31 3: Harmanian Realism Given the existence of this tacit rule, it seems quite likely that all people in all societies will generally and regularly, upon exposure to an "incest-stimulus" — e.g. a description or dream of an act of incest — make the observation, "that is (morally) wrong." To return to Harman's account, the incest-stimulus, in his view, must be psychological. Citing a non-psychological cause of an observation could, plausibly, serve to most increase understanding of the observation (in light of the questions that are of interest, of course) — but this would mean the non-psychological cause is, for Harman, the best explanation of the observation. As has been shown, though, beliefs are the best explanation of all moral observations — and beliefs are psychological. Thus, the incest-stimulus must be a description of an act of incest, or a dream, or a belief about incest, or similar. Further, as discussed above, Harman holds that beliefs are the best explanation of all moral observations. Since the above observation, "that is (morally) wrong," is made universally, in the presence of similar stimuli (i.e. incest-stimuli), the same belief must be its explanation in all cases in which the observation is made.34 It is not at all open to Harman to claim that different beliefs may explain this observation as made by different people. All people make this observation in always the same circumstances. It would thus be highly imparsimonious to claim, without argument, that the observation is best explained differently i f made by different people. Now, though, Harman faces a problem. If he is correct, then each and every person throughout history and across widely varying regions of the planet must have, or have had, a belief such that each and every person, when in the presence of an incest-stimulus, made the 32 observation, "that (act of incest) is (morally) wrong." On the face of it, this is highly implausible. Consider the following, non-moral case. Suppose that all people, in all societies, have made the observation, "rain is wet." A Harmanian sort of explanation35 of this observation would require that each and every person throughout history and across widely varying regions of the planet must have, or have had, a belief such that each and every person, would, when in the presence of a psychological "rain-wetness stimulus" — for example, "being told rain is wet" or "having the belief that rain is wet" — make the observation, "rain is wet." Note that this "rain-wetness stimulus" must be psychological, in the same way as the incest-stimulus must be psychological. The "rain-wetness stimulus" is thus not equivalent to some property such as, "feeling the wetness of rain." Contrast this with an "anti-Harmanian" sort of explanation, which would require that each and every person throughout history and across widely varying regions of the planet must have observed (the fact) that rain is wet, and thus each and every person, when in the presence of this non-psychological "rain-wetness stimulus" (i.e. (the fact) that rain actually is wet), would make the observation, "rain is wet." The Harmanian explanation requires a coincidence: people separated by time and space come, through some unknown mechanism, to have the same belief — the "rain-wetness stimulus" — which best explains their observation. This coincidence is left unaccounted for. The anti-Harmanian explanation presumes there is a fact, which people are able to observe, that best explains the observation. These concepts are also unaccounted for. But they are, however, what might be called "deeply mysterious." That the "common-sense" explanation does not expound upon the nature of facts, of "things," of properties, of reality, or of observation is no real mark against it. For instance, as one walks away from a stereo playing music, the music becomes quieter, which is explained by the inverse-square law. 3 6 The inverse-square law, though, seems to presume (at least) that there is some thing (music) emanating from the source (the stereo); that 33 it has a property (intensity) at a given distance from the source (determined by the inverse-square law); and that there is some fact of what the intensity of the music emanating from the stereo is. The inverse-square law does not account for the nature of the "thing" nor of properties nor of facts, but surely this does not mean that the inverse-square law is a bad explanation. Leaving an unaccounted-for coincidence of belief is a different story — it is "superficially mysterious." If many people just happen to have the same belief, which is presumed by few, i f any, good explanations, there must surely be some account available for it. Perhaps the people have all been taught the same things at the same schools, or they have had similar experiences leading to their forming the same beliefs. The Harmanian explanation needs to be augmented with just such an account; else, it is the anti-Harmanian explanation that contributes more to understanding observations of rain being wet. It is thus, by the Harmanian account's own requirements, the best explanation. This conclusion applies equally well to the moral case. Suppose Harman's explanation of the universal observation that incest is morally wrong is accepted. Central to this explanation, similar to the Harmanian explanation of the observation of rain's wetness, is the claim that people separated by time and space came to all form the belief that sibling incest is wrong (i.e. the incest taboo). This convergence of belief must, somehow, be accounted for, and the Harmanian explanation simply does not offer one, nor is it even clear what such an account might look like. 3 8 There is also an anti-Harmanian explanation of this same observation parallel to that in the rain-wetness case: that each and every person throughout history and across widely varying regions of the planet must have observed the fact that incest is wrong, such that each and every person, when in the presence of this non-psychological incest-stimulus (i.e. the fact that incest is wrong), would make the observation, "that (act of incest) is (morally) wrong." This 34 explanation is not a conventionalist one. The fact that incest is wrong does not depend on any framework or agreement between people — it is independent of frameworks and agreements. Thus, as it fits the definition of realism given above, this explanation wil l be referred to as the "realist explanation." The realist explanation of the observation, "that (act of incest) is (morally) wrong," is better than Harman's explanation for similar reasons as those favouring the anti-Harmanian explanation of the observation, "rain is wet," over the Harmanian explanation. It contributes most to understanding moral observations, principally by not hypothesizing some not understood coincidence of belief, and it is thus pragmatically simplest. By Harman's own requirements, then, the realist explanation is best.39 35 I V : F U R T H E R ISSUES A N D RESPONSES' 1: Introduction There are three possible avenues Harman could take to escape the conclusion reached in the previous chapter. First, he could claim that, contrary to the given anthropological evidence, the belief in the wrongness of incest is not universal. He actually does seem to claim something along these lines, holding that there is, "[n]o substantive central core of morality [that] is universally recognized in all societies.... [For any] universally accepted principle must verge on triviality, saying, for example, that one must not kill or harm members of a certain group, namely, the group of people one must not kill or harm" (Harman 1999,162). This may not be pure dogmatism on Harman's part; there is some controversy in the anthropological literature about the truth of the claim that there is a universal incest taboo. Keith Hopkins, for one, argues that incestuous marriage was widely-practiced, and widely-accepted, in Roman Egypt, and, thus, there is no universal prohibition against incest. His argument will be considered in-depth below. Second, even i f Hopkins' case does not hold together, and there is no good reason to assume that the Egyptians — or anyone else — did believe incest is morally permissible, there may be another means for Harman to save his relativist conclusion. It may be possible for Harman to account for the belief in the wrongness of incest without questioning its universality, and yet also without hypothesizing any moral facts. Thus, Harman could happily assume, along with the realist, that there is a universal belief that incest is wrong, and still strongly support his relativism. Melvin Ember, another anthropologist who has studied the incest taboo in depth, argues in favour of this point. He puts forward a view, which he terms "inbreeding theory," that may show that incest is universally wrong, and yet that there are no moral facts. 36 Finally, third, Harman might argue that the consequences of endorsing the conclusion are so unacceptable as to render his own position tenable by default. As argued above, the universal belief in the wrongness of incest, coupled with Harman's views on explanations, simplicity, and the moral "ought," seems to give good evidence to suppose that Harman has successfully defended a species of realism rather than relativism. But there may be a problem with the reasoning used to employ the belief in the wrongness of incest in this way, particularly with the implicit claim that any universal belief in either moral lightness or wrongness must be endorsed as a true belief. If there is such a problem, then Harman may have a good reason to reject the realist conclusion. 2: Roman Egypt First the possibility of undermining the universality of the belief in the wrongness of incest will be addressed. Without yet delving into the data, though, suppose that the case holds together. In Roman Egypt, people did not believe in the wrongness of incest; thus, clearly, Roman Egypt serves to undermine the argument employing a universal prohibition against incest as evidence against relativism. There is a phenomenon — call it "belief convergence" — wherein many people in many societies come to have similar beliefs on a given subject. For example, in many societies, people have developed a belief in a supernatural force — "magic" — which certain, special people can manipulate to bring about bad results. People in Christian societies have held these beliefs, calling the people "witches", but so have people in radically different societies, such as the Azande, an African tribal society (Winch 1964). If all societies believed in magic and witches, there would be a strong case to be made for the actual existence of magic and witches. It would be very unlikely that all societies, across the world and throughout history, 37 came to believe in the existence of the same things, and yet were all mistaken. However, in actuality, there are societies, such as our own, that do not generally believe in witches and magic. And our society does not suffer for this lack — for example, all phenomena that might be accounted for by magic, such as crop failures and disease, can be accounted for in other ways. Since societies that do not believe in magic and witches can be found, that many societies believe in magic and witches is not (by itself) evidence for the existence of some real magic. Rather, it is just a coincidence: these societies have, probably from very different causes, "just happened" to have produced the same belief. The belief in the wrongness of incest may be an unusually extreme case of this sort of convergence. Beliefs that incest is wrong are clearly more wide-spread than beliefs in the existence of witches — it is not hard to find a society that does not believe in witches, though quite difficult to find one that does not believe incest is wrong. But, i f Keith Hopkins' conclusions are correct, Harman would be able to employ the case of Roman Egypt against the belief in the wrongness of incest, just as one might employ the case of our society against the belief in magic and witches. That is, the Egyptians, according to Hopkins, did not believe incest was morally wrong; and, the Egyptians did not, seemingly, suffer much due to this lack of belief. The existence of such a society that does not believe incest is wrong — following the same reasoning as in the case of belief in witches and magic — shows that the fact that nearly all other societies believe incest is wrong is no evidence that incest is really wrong. 4 1 It could thus be claimed, contrary to the position of moral realism, that the belief that incest is wrong is a belief no different in kind from beliefs in witches, only different in degree. Hopkins spells out a tremendous amount of data to support this conclusion, data relating to the practice of incestuous marriage in Egypt while under Roman domination. Every 14 years, 38 at least, from A D 19-20 to 257-8 — though, A D 33-4 is the first certain instance of a census — the Roman governors of Egypt ordered their officials to take a census of the whole population, likely as a basis for taxation (Hopkins 1980, 303). Egyptian males became subject to poll-tax at age 14, hence the standard interval, though censuses were sometimes taken more frequently, in order to account for deaths, exemptions due to advanced age, and registration of newborn children (Ibid., 313). Hopkins also notes that there was an old custom of censuses as a basis for taxation in Egypt, while under the rule of the Greeks. Firm evidence of the practice as extended to the entire population is, however, only available starting with the Roman conquest in 31 BC. It is worth noting that the Romans took censuses of their own population, and of some other conquered provinces, for these same purposes (Ibid., 313). Severe penalties were levied for failure to completely reveal all information required, hence the truthfulness of the surviving census returns — approximately 270, written in Greek, the official language of the Roman administration in Egypt — is on (fairly) solid foundation (Ibid., 303). These returns show a consistent practice of "lawful, publicly celebrated marriage between full brother and sister, replete with wedding invitations, marriage contracts, dowries, children, and divorce" (Ibid., 304). The inclusion of "children" here shows strongly that these are not cases of formal, ritual marriage. As Hopkins puts it, "brother-sister marriages were obviously fertile, not formal, and were declared openly, not only in family matters but also in business [such as crop sales, lawsuits, and official petitions]" (Ibid., 323). By the time Rome fully assimilated its conquered territories, by conferring citizenship on all males, and thus forced Roman prohibitions on incest onto the Egyptian people, evidence of the practice vanishes (Ibid., 353-4). Hopkins reports that, according to the censuses, brother-sister marriages made up 39 approximately 15-21% of all ongoing marriages (Ibid., 304). He estimates that, due to high mortality of children (in particular), only about 40% of all families had a son and a daughter, or both sons and daughters, surviving to marriageable age. By his calculations, these figures imply that one-third — or more — of all brothers with marriageable sisters chose to marry their sisters rather than take an unrelated wife. If these figures can be seen as representative of the Egyptian population as a whole — rather than a statistical aberration — then Roman Egypt presents a serious challenge to claims attributing universality to the incest taboo (Ibid., 304). There is, obviously, no way to prove that the surviving returns are random, in the statistical sense of "random" (Ibid., 314). The best that can be achieved, then, is a demonstration that no significant bias exists in the sample that have, seemingly by fate, survived. However, Hopkins, rather cautiously, notes that a sample that is not obviously biased can only be taken as a rough guide to the overall population; the "lottery of survival", in his view, draws censuses "oddly", no matter what. The historical and geographical changes over 300 years are, he thinks, significant enough that a sample that is not fully random cannot be fully relied upon (Ibid., 314). But it seems that Hopkins might be rather over-cautious here. While the changes he notes are surely not to be ignored, it is very hard to see why they would make any tremendous difference, as he seems to believe they will . Of the 270 surviving returns, only 172, covering 880 people, are informative and legible enough to be usable (Ibid., 315). Of these, 79% are from the 2 n d century A D ; however, all censuses taken between A D 19 and 257 (with the sole exception of 229) are represented to some extent in the sample (Ibid., 315). The historical representation is, thus, Hopkins says, fair. With regards to the geographical representation, no returns from Alexandria, the largest 40 city, have survived — papyri, according to Hopkins, did not tend to survive the climate. However, the geographical distribution is otherwise quite fair — 8 towns (54% of the returns) and 20 villages (41%) are represented (Ibid., 315). Hopkins notes that the Fayum region is somewhat over-represented — 63% of surviving returns — but this is common, he claims, amongst ancient Egyptian papyri. The region was a natural depression some 100 kilometres south of Cairo, irrigated artificially for a time in order to be fertile; but, during the 3 r d and 4 t h century A D , the irrigation system was neglected, leading to whole towns and villages being deserted. During these desertions, papyri were either carefully buried or even discarded on rubbish piles, to be covered and preserved by the sands until modern times (Ibid., 315). As this is a general problem facing all Egyptian records, it is probably fair to grant Hopkins that this is no troubling bias. Further, as he notes, village parish records are frequently relied on during historical demographical research in European areas, which represent an even more homogenous sample than these (Ibid., 315). If this sort of geographical homogeneity is fair for research into European history, then it is fair for research into the Egyptian. There is a further possibility of bias, namely a bias of social class. There may be a bias towards the well-to-do in the sample — after all, the wealthy would tend to keep more records on papyrus, and these records would likely be better cared for, and thus better preserved (Ibid., 315-6). While no clear answer is available here, Hopkins says, three indicators of social class can be explored: literacy, occupation, and ownership of slaves ~ the wealthy being more likely to be literate, more likely to own slaves, and have quite different occupations from the poor (Ibid., 316). Hopkins calculates the literacy rate — judging by, first, the number of returns written by professional scribes; second, the number only endorsed, rather than written by heads of household; and, third, those outright written by heads of household — as about 30%. This, Hopkins judges, is a little bit high, even granting that the spread of bureaucracy across the 41 Roman provinces might have boosted the literacy rate (Ibid., 316). So, this possibly indicates a bias. The occupations listed do little to clear this up — they are unfortunately rather inconclusive. Only 24% of the returns even list an occupation. Of these, 1/2 are either "respectable" or too general to be useful — such occupations as landowner, farmer, or priest. However, the other 1/2 are craftsmen and unskilled workers, indicating that at least some (about 12%) of these returns came from poor, or poorish, households (Ibid., 316). Finally, 81% of the households covered by the returns had no slaves, which would be a key indicator of wealth (Ibid., 316). From this, Hopkins concludes, that, although there is a slight bias towards the literate and well-to-do, the returns are generally not biased towards any particular social class, and thus cover a broad spectrum of Egyptian society. The final possible bias Hopkins considers is that of gender or age. A bias in either of these areas would skew any general conclusions, regarding people of both genders and many ages, that might be drawn from the data. The ratio of males to females reported is, however, 415 to 387 — about 100 to 95 (Ibid., 316). So the bias here is likely negligible. With regards to age, Hopkins notes that there are a disproportionate number of young females registered, but the ratio balances out at greater ages, leading him to conclude that this is a problem of under-registration (Ibid., 317). He offers no explanation for this anomaly of young females, but does for another case of potential under-registration, that of males aged 10-19. In this age group, young males were either just before or just after eligibility for the poll-tax at age 14. Thus, their fathers or mothers might have deliberately neglected to register them, as an attempt to deceive the tax officials (Ibid., 318). Perhaps a similar explanation might be available for the under-registration of young females. Hopkins discusses this problem at length, but only concludes that there very probably is a 42 bias regarding the ages — the distribution of the ages of both males and females reported in the returns fit no usual demographic pattern he is familiar with — a bias which he has no idea how to correct (Ibid., 318-20). However, he makes some attempt to discharge the possibility of this bias, pointing out that the patterns are close to those of actual human populations; and, moreover, the pattern is closer than any found from records of the Romans themselves (Ibid., 320). Thus, "[i]n sum, these Egyptian census data are far from perfect, but they are the best data we have; and I think they constitute a sample representative of a much wider Egyptian population" (Ibid., 320). What, then, do these data say about incestuous marriage? Of 113 on-going marriages noted in the returns, 23 (17 for certain; 6 "perhaps to probably") are between some sort of brother and sister — so, 15% for certain, and 21% i f less than certainty will suffice (Ibid., 320). Of these, 11 or 12 are between full brother and sister (9 or 10 of the 17 certain), and 8 are between half-brother and -sister (6 of the same father, 2 of the same mother); the remainder are unclear (Ibid., 320). Further, there are 13 marriages between brother and sister of some kind traceable in the genealogies mentioned in the censuses (Ibid., 320). Hopkins points out that there really is no ambiguity in most of these cases; the formula used, consistently, in the returns is: "my (or 'his') wife and sister of the same father and of the same mother" (Ibid., 320). Further, these marriages were considered perfectly normal. Other surviving documents, Hopkins claims, show this clearly. "In birth certificates, in claims for privileged status which were accompanied by explicit genealogies, in contracts of marriage and of divorce, in wedding invitations and in private letters, brother-sister marriages were recorded, mentioned without fuss, and apparently taken as a matter of course" (Ibid., 321). Oddly enough, though, wedding invitations and letters are not considered good evidence of the acceptance of the practice. Calling one's wife "sister" was apparently a widespread practice in Egypt; only in cases of royal 43 letters and wedding invitations, as royal genealogies are clear and available, can actual brother-sister connections be identified from such evidence (Ibid., 324). Hopkins also considers the issue of the, clearly, harmful effects of inbreeding that must have resulted from the Egyptian practice of incestuous marriage. He admits that brother-sister matings produce offspring five times more likely to suffer harmful, genetic effects than those of randomly mated parents, but argues that this tendency is most evident in societies with otherwise low mortalities (Ibid., 325). If mortality is generally high, such as in Roman Egypt, the deleterious effects of inbreeding would likely not be visible (Ibid.,326). Further, as it is extremely unlikely, given this high mortality, that two consecutive generations would have a brother and sister pair able to mate, the effects of inbreeding would be mild, and obliterated by a long series of non-familial matings (Ibid., 327). Hopkins' "tentative" conclusion, on the basis of this data, is that brothers and sisters must have married each other because they wanted to (Ibid., 353). If the incest taboo is an explicit rule, such as a law, against the activity, then Hopkins' data will clearly refute the universality of the belief, and any argument that depends upon it. The Egyptians did not, as shown by their regular endorsement of the practice, have laws or similar rules against incest between full brothers and sisters. If explicit rules are at stake, Roman Egypt is a devastating counter-example. But, as outlined above, the incest taboo can — and should — be understood as a tacit rule and not as an explicit rule or law. For example, the Pondo have no explicit rules against incest, only a somewhat vague sense that the act wil l be punished by supernatural forces. Hopkins could not, without begging the question, claim that the Pondo are just an exception to some explicit incest taboo; thus the Pondo can be taken as evidence that the 44 incest taboo is not an explicit rule. On the other hand, if the incest taboo is a tacit rule, Hopkins' data suggests three possibilities. First, that the Egyptians believed that incest was morally wrong, but, nonetheless, were acting contrary to a tacit rule against incest. In cases of extreme hunger, for example, a person can, though wracked with guilt, overcome a belief that stealing food is wrong. Perhaps social forces, such as a desire to keep family property within the family, 4 2 or a desire to only marry within the relatively small community of Egyptians — and not, for example, marry a Roman — had a similar sort of effect on Egyptian beliefs in the wrongness of incest. In this case, Roman Egypt is not a counter-example at all; rather, it serves to elucidate a subtlety of the nature of moral beliefs, namely that moral beliefs, though they may powerfully influence behaviour, and even be perfectly true, can be overridden by other considerations. Second, that the Egyptians may not have believed that incest was wrong, and instead developed and followed some other rule, one that permitted incest. In this sort of case, Roman Egypt seems to defeat the argument for moral realism — if the Egyptians did not believe that incest is wrong, then there is no universal moral belief in the wrongness of incest, and thus the reason to believe it is a fact that incest is wrong is undermined. Of course, it might still be possible to revise the argument and save the realist conclusion. If the argument were appended with some systematic way of separating reliable moral beliefs from unreliable ones, and, as it turned out, the beliefs of the Egyptians were not reliable (due to being malformed in some way, or corrupted by the influence of persons in authority, or some such), Roman Egypt would no longer cause difficulties for the realist. And, third, that the Egyptians had both the incest taboo and some contradictory tacit rule 45 regulating their behaviour, and, though most chose to follow the former, some Egyptians chose to follow the latter. Though one should not be guided by contradictory rules, this does not, of course, imply that people are never within the reach of contradictory rules. In other words, the Egyptians may have believed contrary things — both that incest was wrong, and that incest was not wrong. Hopkins' data is non-committal between these three cases, as he has no data which would clearly indicate which case amongst the three is the correct one, such as data on the beliefs of the Egyptians concerning the wrongness of incest, or on the tacit rules that regulated Egyptian incestuous behaviour. And, there is, apparently, no further evidence of incestuous marriage in Roman Egypt available, which could be added to that Hopkins covers; all the evidence that has survived is in the census returns examined by Hopkins. Anthropologist Brent D. Shaw makes this clear: "[o]f all these studies [of brother-sister marriage in Roman Egypt].. .by far the most sophisticated and analytically satisfying (and this by a considerable margin over the others) is the detailed description and analysis that has been provided by Keith Hopkins.... [H]is study must form the basis of any further serious discussion of the subject" (Shaw 1992, 269). So, as Hopkins simply has no data on the tacit rules that guided the Egyptians' behaviour, there is no way to decide between the three possibilities. Roman Egypt may be a counter-example to the universality of beliefs in the wrongness of incest. But, it equally may not be. The most reasonable position to take on this case is an agnostic one: to suspend judgement until further evidence of the beliefs of the Egyptians comes to light. This is an unfortunate result for Harman. His first avenue of escape — the claim that there is no significant case of a non-trivial, non-framework relative moral belief — needs some 46 clear, strong counter-example to the universality of the belief that incest is wrong; else, he must rest it on the unsupported assumption that the Egyptians believed incest was morally acceptable. And, even more unfortunately for Harman, the anthropological literature presents no other cases that might show the belief in the wrongness of incest to be non-universal — the focus is entirely on Roman Egypt, which seems to suggest it is the only interestingly controversial case. Even anthropologists who disagree that the belief is universal have little evidence to support the claim. For example, Russell Middleton holds that although Roman Egypt, "is probably the most significant example, the Egyptian case does not stand alone as an exception to the universality of the brother-sister incest taboo" (Middleton 1962, 61 l i) . Middleton points to a small community on a Caribbean island that seems to endorse brother-sister incest, but immediately undermines this by stating that the situation, "apparently is an aberration... which developed because of special circumstances, and the original... standards are now beginning to be reasserted" (Ibid., 611). He then suggests there is other evidence of "societies which have sanctioned unions between brothers and sisters", but provides no details of these societies ~ not mentioning even one by name (Ibid., 61 Iii). His claims thus seem hopeful, at best. It thus seems highly unlikely that there are any other possible exceptions to the universality of the belief in the wrongness of incest. At the very least, given the failure of Roman Egypt, the most exhaustively-examined case, it is up to Harman to show that there is some other possibility, some other civilization or social group which clearly did not believe in the wrongness of incest, that might save his argument. The first line of counter-argument Harman might adopt against the realist conclusion can therefore be judged a failure. 47 3: Inbreeding Theory Harman may, of course, attempt to pursue the second avenue of counter-argument, by arguing that the acceptance of the universality of the belief in the wrongness of incest is no real challenge to his moral conventionalism. The clearest path to this conclusion lies in other anthropological theories, theories which purport to show that beliefs in the wrongness of incest are universal, but do not require the existence of any moral facts. Anthropologist Melvin Ember has such a view, which he terms "inbreeding theory." According to Ember, inbreeding theory is the view that the emergence of incest taboos can be explained by considering the generally harmful effects of inbreeding. "Genes that regularly arise by mutation are generally harmful and recessive; the offspring of close relatives may inherit a double, and possibly lethal, dose of such a recessive gene" (Ember 1975, 256). Hence, the percentage of early deaths will be consistently greater in the offspring of related parents than those descending from unrelated spouses. The risk is high enough in the offspring of tertiary relatives (first cousins), but even moreso in the offspring of secondary relatives (uncle/aunt and niece/nephew). The offspring of second cousins and more distant relatives are at much less risk. Thus, not only do "the recessive traits which come to light in the offspring of relatives result... in lessened viability," but "the closer the inbreeding, the greater the frequency of early deaths" (Ibid., 256). Ember argues that a population of mixed incestuous/non-incestuous unions would eventually be entirely dominated by non-incestuous unions, as offspring of these matings would be most likely to survive, and hence their numbers would, in time, be greater. Thus societies which had only non-incestuous matings would grow at an even greater rate; so natural selection 48 would select for a prohibition against incest. Societies with these prohibitions would be most likely to survive and dominate their competitors. Further than this, though, Ember claims that it's very likely the consequences of inbreeding would be so great as to be recognized by members of a given society. And, upon recognizing that incest would lead to death, these societies would likely enact, consciously, explicit incest taboos. He constructs a genetic/statistical model to establish this, which will be assumed to be accurate (Ibid., 271-6). According to Ember, the model "suggestfs] that our forebears may have recognized the deleterious consequences of inbreeding and therefore may have consciously adopted the incest taboo and extended it to other close relatives" (Ibid., 271). More specifically, on the strength of the model Ember concludes that (a) as the risk of familial matings resulting in infant deaths due to genetic causes is increased, so, too, the advantage of non-familial matings43 increases; (b) as more births, generally, survive to reproductive age, and the society's population is thus expanding, the advantage of non-familial matings also increases; and, (c) as there are fewer familial matings that result in surviving offspring, as compared to the offspring surviving from non-familial matings, the advantage of non-familial matings again increases (Ibid., 276). In other words, as the features of the modelled populations come closer to meeting the features of actual early populations, populations that have not yet enacted taboos — namely, that familial matings are very likely to result in deaths from genetic causes; that early populations were likely increasing very rapidly, rather than remaining stationary; and that familial matings are highly risky when compared to non-familial matings — non-familial matings have a more and more noticeable advantage over familial matings. Ember's own figures give a "best case" advantage of 300% to non-familial matings, and a "worst case" of 16% (Ibid., 273). A conscious 49 adoption of an incest taboo — whether in implicit or explicit form — is therefore quite likely. If the advantage of an incest taboo is very great, then one would have to be willfully blind to the deformities and deaths of offspring of familial couples, and the (relatively) healthy offspring of unrelated spouses, in order to not see the benefits of forbidding incest, at least for oneself. Ember also argues that inbreeding theory can account for a feature of incest taboos not often discussed, namely their extension to first cousins. This feature of incest taboos is variant: there are societies that prohibit sexual intercourse between first cousins, and others that allow it freely (Ibid., 250). Ember holds that the existence of a prohibition on first-cousin incest is a "default," and there are three main conditions which, i f any one is in place, will lead to a relaxation of the incest taboo such that it no longer extends to first-cousins. First, the taboo is more likely to be relaxed in societies with relatively large "genetic isolates" — approximately equivalent to the popular notion of "gene pools." Societies with relatively small genetic isolates are more likely to prohibit first-cousin incest (Ibid., 269). Second, mid-size societies that have suffered sudden and severe depopulation are likely to relax the taboo (Ibid., 269). And, third, very small societies may relax the taboo (Ibid., 270). If Ember is taken to be addressing consciously created, explicit incest taboos — as he seems to be, in places — he cannot be taken as supporting Harman's position. The objection is the same as one raised against Hopkins: the prohibition on incest that is under investigation here is a tacit not explicit rule, hence any arguments relating to explicit rules are simply irrelevant. Unlike Hopkins, though, Ember could simply eliminate all traces of this view — such as the claim that the incest taboo is a conscious creation — without doing much damage to his overall theory. But, once this change has been made, it becomes clear that his argument does not actually conflict with moral realism. Instead, Ember can be seen as offering little more than an 50 evolutionary account of how the tacit rule against incest came to be. Tying moral realism to evolutionary theory may even serve to help support realism, rather than undermine it, depending on one's views on evolutionary theory. Some people, says Ember, were able to recognize that incest resulted in fewer children surviving to produce their own offspring, and in more genetic disease, and thus developed tacit rules prohibiting incest. Other people did not recognize this, and did not develop such rules. Those that did develop tacit rules were able to quickly dominate those that did not — prohibiting incest results, as Ember argues, in a competitive advantage due to the deleterious consequences of incestuous mating. But this is only a causal account of the emergence of a belief in the wrongness of incest, much as a causal account of the belief in the importance of protecting one's children might arise, through likely similar genetic and biological circumstances. Giving such an account does not address the fact that protecting one's children is important, only that it is believed to be important, and that this belief has many useful consequences. Similarly, Ember's account does not address the fact that incest is wrong, only that it is believed to be wrong, and that this belief is useful. He does not touch the issue underlying the universality of the belief: the fact that incest is morally wrong. Thus, this does nothing to help support Harman's relativist claims. And, given the failure of as sophisticated a treatment as Ember's, it is hard to see what sort of anthropological theory could support Harman's views. So, without a supporting theory, Harman would only have one means of maintaining that a universal belief in the wrongness of incest is no threat to his relativism: the simple, stubborn claim that the apparently universal belief in incest is, contrary to all other argument, actually only held due to many hundreds — even thousands — of moral agreements made between different people, in different times and places. Putting aside, for the moment, questions of the 51 plausibility of this tactic (as they have already been addressed above, p. 33-4), there is a further difficulty if this is the strategy used to defend Harman's conclusion. It is, in effect, a concession of the argument to the realist. There is no morally significant difference between the claim that a belief is made universal due to its being conventional everywhere, and the claim that a belief is universal due to its being about some recognizable fact. Take the belief that Mount Everest is big, and, for the moment, assume that this belief is universal. It might be claimed, sensibly enough, that Mount Everest is universally believed to be big because it is a fact that Mount Everest is big. It might also be claimed that Mount Everest is universally believed to be big because everyone is under some "Mount Everest agreement" or another, and, as all these agreements imply that Mount Everest is big, accepting any of these agreements requires believing that Mount Everest is big. Investigations of Mount Everest can proceed smoothly on the basis of either of these claims; whichever is adopted makes no difference. Similarly, i f a realist wishes to assert that some moral belief is true due to some fact, and a relativist wishes (rather dogmatically) to assert that the same moral belief is true relative to everything, then there is really no point in continuing an argument on the issue. Though there may be some very minor ontological wrangling left to do, there is no interesting moral point at stake. Realists could roll their eyes and talk about moral beliefs that are "relative to everything" in order to keep relativists happy, and relativists could similarly hold their noses and talk about moral beliefs that are "about facts," and both could move on to other ethical discussions. It is, however, quite unlikely that Harman would wish this outcome ~ he does, after all, wish to defend relativism, not show that it and realism aren't different in any morally significant way. But it is worth requoting the following passage, also quoted above on p. 19. 52 It is compatible with moral relativism for certain moral claims to hold in relation to all moral frameworks just as certain claims about motion hold in all spatio-temporal frameworks, eg. [sic] that nothing travels faster than the speed of light. Any moral claim that is incompatible with such universal claims would be false in all moral frameworks.... Can a moral relativist agree... that a judgment [sic] whose truth value is the same in all moral frameworks has that truth value absolutely and not merely in relation to one or another such framework? Maybe, although that is in some ways like agreeing that 'The universe is at least as large as' is true absolutely because true in relation to everything (Harman 1998b, 211; emphasis removed). In this passage, Harman seems oddly unaware of the tremendous impact the admission that "absolutely true" or "universally true" amounts to the same as "true relative to everything" wil l have on his overall argument. Once he concedes this point, it seems he really has nothing to say that is in opposition to a realist view. It seems better, then, to treat this passage as only admitting the possibility of some universal moral belief as a theoretical "limiting case." Once faced with the actuality, perhaps Harman would fight harder to preserve his view, and abandon the sort of thinking shown here. It is sensible, therefore, to conclude that he would not take this second avenue of counter-argument, denying that a universal incest taboo is problematic, against the realist. 4: Universal Slavery Harman's only remaining avenue of counter-argument, then, is the third, which questions the reasoning underlying Chapter 3's argument in support of moral realism. The key section of the reasoning is as follows: given that x is a universally-held moral belief, it follows that "the fact that x" is, in Harman's sense — i.e. it is pragmatically simpler, in that it leads to greatest understanding of the questions of interest, etc. — the best explanation of x. If x is "incest is wrong," this formula then becomes: given that "incest is wrong" is a universally-held belief, it follows that "the fact that incest is wrong" is, in Harman's sense, the best explanation of (the belief that) "incest is wrong." 53 Suppose, however, that there is another universal belief, established by as detailed and far-reaching a set of data as the belief in the wrongness of incest: the belief that enslaving other people is always morally permissible. And suppose that, as with the belief in the wrongness of incest, this belief exhibits all the same superficial characteristics as other, uncontroversially moral beliefs. It can thus be concluded that this is a genuinely universal moral belief. And it could be used as further evidence in favour of moral realism, and against Harman's moral conventionalism; there is, ex hypothesi, no significant difference between this belief and the belief in the wrongness of incest. With x as "enslaving other people is morally permissible," the formula becomes: given that "enslaving other people is morally permissible," it follows that "the fact that enslaving other people is morally permissible" is, in Harman's sense, the best explanation of (the belief that) "enslaving other people is morally permissible." It follows that, under this reasoning, it is a fact that enslaving other people is morally permissible. Harman might at this point level the charge that, since this line of reasoning clearly leaves open the possibility that slavery is morally permissible, this is reasoning that must contain some deep flaw — logical or otherwise. On the face of it, this is a possibly fruitful attack for Harman. Regardless of what one might think of slavery, it is surely possible to construct other counter-examples which fit the given formula, and seem to imply that egregiously wrong-headed beliefs — such as in the moral permissibility of slavery, or in the moral wrongness of charity, or in the moral permissibility of casual, random murder — can, i f found to be universal and to exhibit similar superficial characteristics as other moral beliefs, be shown to be true, and thus statements of moral fact. If this is a consequence of this kind of reasoning, it may be necessary to remove its conclusions from consideration, much as a line of argument that could be used to 54 show 1=2 is judged to be highly suspect, whatever else it might prove, on the basis of that unacceptable conclusion. But, although there is a problem with the reasoning used to reach the realist conclusion, it would be far too quick to immediately conclude the problem lay in the reasoning simpliciter, rather than in some part of the reasoning. The formula, once again, is: given that x is a universally-held moral belief, it follows that "the fact that x" is, in Harman's sense, the best explanation of x. While Harman might try to lay the blame for unacceptable conclusions on the formula as a whole, it is just as plausible that the blame should be laid on the views referenced by the clause, "in Harman's sense." It is only as a best explanation in Harman's sense that "the fact that x" can be so closely linked to the universally held moral belief that x. For purposes of comparison, consider Hempel's Deductive-Nomological (D-N) model of explanation. The model, stated very briefly, holds that an, explanatory account may be regarded as an argument to the effect that the event to be explained (let me call it the explanandum-event) was to be expected by reason of certain explanatory facts. These may be divided into two groups: (i) particular facts and (ii) uniformities expressed by general laws.... If we imagine these various presuppositions explicitly spelled out, the idea suggests itself of construing the explanation as a deductive argument of this form: C], C2,Ck Li, L2,Lr E ...The kind of explanation thus characterized I will call deductive-nomological explanation..." (Hempel 1968, 55-6). If the D-N- model is applied to this case, the universally-held moral belief that x will be treated as explanandum, and "the fact that x" as part of the explanans. A n explanation of the 55 explanandum will only be an explanation in Hempel's sense at all — never mind the best explanation — if there is some law that, as part of the explanans, completes a deductive argument from the explanans to the explanandum. The least complex possibility is, it seems,"all universally-held moral beliefs are true (or factual)." If this were an actual law, it would have to be an exceptionally bizarre one — whether a law of science, ethics, nature, or what have you. Though a beliefs being held universally may be good evidence for its truth, it is quite a large leap from this to the claim that universal beliefs are always true, by some sort of "law." Further, more complex possibilities, whatever they might be, would surely be even harder to take seriously. Clearly, then, the argument in Chapter 3 depends crucially on Harman's own views of explanations and simplicity; and, once these are replaced with other possibilities, such as Hempel's, it becomes much harder — though perhaps not impossible — to endorse the realist conclusion solely on the basis of the anthropological data. So, for Harman, or a defender of Harman, to try to discredit the reasoning behind the realist conclusion will only serve to do more damage to Harman's position. The flaw in the reasoning lies in its employment of Harman's highly idiosyncratic views of explanations and simplicity. While Harman is perfectly free to call these into question, doing so undermines his conclusion of moral conventionalism just as quickly and definitively as it undermines the conclusion of moral realism. Thus, this final possible response to the realist conclusion fails, as did both other alternatives. 5: Conclusions Harman argues that his views on explanations, simplicity, and the moral "ought" imply, strongly, that only beliefs can explain moral observations. However, when confronted with the 56 evidence of the incest taboo, and its associated observations, it begins to look as i f a realist explanation — that there is a fact that incest is wrong — is, by Harman's own standards, the best explanation. Harman might try to argue that there is no universal incest taboo — that it is simply a very widely-held belief, and thus not harmful to his view. The only possible study of the only possible counter-example is Keith Hopkins' work on Roman Egypt. And Hopkins' data are, unfortunately, insufficient to serve as the sort of strong case Harman would need to demonstrate beliefs in the wrongness of incest are not universal. Without any clear, strong case, Harman cannot show that the incest taboo is not a universal belief. Harman might then try to argue that a universal belief in the wrongness of incest is not damaging to his view, and look to a theory such as Melvin Ember's, which seems to account for the universality of the belief in the wrongness of incest without hypothesizing any moral facts, for support. But, Ember's view does not conflict, seemingly, with moral realism. If Harman were to continue to pursue this line, leaving Ember's views behind, he would be in danger of conceding any interesting moral ground to the realist. Harman's only remaining alternative is to question the reasoning underlying the realist conclusion, and argue that it leads to the endorsement of various unacceptable conclusions, such as the possible endorsement of the moral permissibility of slavery. The flaw here, though, is that Harman would be too quick to dispose of the entire chain of reasoning, rather than focus on its weakest link — namely, his own views on explanations and simplicity. While these could be eliminated and replaced, with, for example, Hempel's views on explanations, this leaves Harman's moral conventionalism just as undermined as the realist conclusion of Chapter 3. 57 Thus, Harman cannot bolster his arguments against the argument made in Chapter 3 — there is nothing he can add or argue about to support his conclusions. His only open options are: (i) to assume, without any evidence, that there is some society that does not believe that incest is morally wrong; or (ii) to dogmatically insist that there is some moral point at stake, even i f universal moral beliefs can be glossed as "relative to everything;" or (iii) try to locate a flaw in the reasoning other than his own accounts. Given the paucity of all these options, it follows that Harman's own arguments, when coupled with the evidence relating to the universal incest taboo, do not show that moral relativism is true. Instead, they show the truth of moral realism. 58 E N D N O T E S 1 Catherine Wilson and Scott Anderson have kindly made extensive comments and suggestions on earlier versions of this thesis. The work has also benefited from comments and suggestions by participants in Catherine Wilson's "Ethics Teas" at UBC, Fall-Spring 2002-03. 2 This is not to say that moral facts are not affected by people. For instance, it may be true that soldiers are morally permitted to kill soldiers of opposing armies during times of war, or it may be true that employers are morally obligated to provide living wages for their employees. "Soldier", "army", "war", "employer", "wages", and "employees" are all socially-created categories — in other words, categories created by people. So, at least some, and probably many, moral facts will vary due to features of people and their social world — what counts as a soldier, for example, can and will change, as will whether there are any actual people that fit this category. The realist claim is, though, that, even if there are no soldiers or armies or wars, and even if the definitions of a soldier and an army and a war can be altered, what a soldier in an army in a time of war is morally permitted to, obligated to, and forbidden from doing is true or false independent of people. This does seem to have some far-reaching consequences. For example, suppose a soldier were, somehow, redefined as "a person who is morally permitted to shoot anyone ". This could be seen as a creation of a new social category, using an old name, rather than a redefinition of the old category — as, if it is true that a soldier is only morally permitted to shoot certain people at certain times, it cannot also be true that a soldier is morally permitted to shoot anyone. Thus, it must be that a soldier is only morally permitted to shoot certain people at certain times, while a soldier* is morally permitted to shoot anyone. Alternately, this could be seen as proof that truth is dependent on time — at one time, it was true that a soldier was only morally permitted to shoot certain people at certain times, and, at a different time, it is true that a soldier is morally permitted to shoot anyone. There may be other possibilities as well. A detailed discussion of this issue is, however, beyond the scope of this thesis; it is only possible to note that the issue is there and move on. ' "Seemingly" as, in making this claim, Harman is riding roughshod over Carl Hempel's more standard distinction between an explanation (Harman's second sense of "explanation") and an explanans (Harman's first sense of "explanation") (Hempel 1968, 55). 4 As do many other things, such as "that one's senses are functioning correctly," "that one's brain is in the appropriate state(s)," etc. 5 Harman himself uses a far more confusing, mathematical example to illustrate; cf. Harman 1994, 79. 6 It is important to take note that this is one of the few distinctions Harman allows between science and ethics. Unlike other relativists — Stevenson, for example (Stevenson 1997) — Harman endorses the claim that morality is as cognitive as science, and is thus not an affective or emotive matter. 7 "Real" is, in most cases, a better term than "physical," in that it encompasses a greater number of metaphysical leanings. Harman, though, seems to be a fairly staunch physicalist, so "physical" is used here. 8 Or, more strictly speaking, in the causal inefficacy of whatever features in reality moral facts are supposed to be about. 9 The scientific case is more difficult to analyze than the moral, as Harman does not present much argument regarding the questions of interest in science, and how they affect the facts or beliefs issue. Presumably, though, scientific observations are explained in order to answer questions about the way things are: observations of protons in cloud-chambers are explained in order to answer questions about how protons are, observations of weather patterns are explained in order to answer questions about how weather is, and so on. Since these questions are directed towards discovering facts about the world, it is, prima facie, plausible to conclude, as Harman does, that facts will generally be more enlightening as an explanation of scientific observations than beliefs. However, there may be considerations that will render this conclusion quite a bit more dubious, particularly as attention is shifted away from physical sciences and into areas such as psychology. Are psychological observations, particularly of people suffering from delusions, best explained by facts? Or by beliefs? Shouldn't facts about beliefs play a role here? How do facts about beliefs differ from facts and from beliefs, anyway? 59 Harman does not, apparently, discuss this issue in depth in any of his work; and, moreover, while it might suit Harman to be able to maintain scientific realism and moral antirealism, it may simply prove to be an impossible distinction to endorse. If so, Harman would either be forced to adopt his own relativism with respect to science as well as ethics, or accept the conclusion of this thesis and endorse moral realism. Thus, given Harman's lack of clear elaboration of his views in the philosophy of science, and the possibility he cannot support the distinction he makes between science and ethics, it would be less than fruitful to spend too much time in discussing the scientific case, rather than the more fully-developed metaethical positions. 1 0 Cf. Harman 1977, 118-24. 1 1 Cf. Harman 1977, 84-5. 1 2 Harman, infamously, illustrates this distinction between inner judgement and other judgements with the example of two people, whom he calls Hitler and Stalin and clearly bases on the historical figures (Harman 2000a, 7). It would be a mistake, he claims, to hold that Hitler "morally ought not to have ordered the extermination of the Jews" or "that it was morally wrong of him to have done so" (Ibid., 7). It seems, he says, '"too weak' a thing to say" (Ibid., 7). It would be more correct, Harman argues, to hold that Hitler was evil, or even that Hitler's ordering of the extermination of the Jews (in other words, the action Hitler took) ought never to have happened. It is the inner judgement alone that seems too weak (Ibid., 7). This, according to Harman, is due to Hitler's being "beyond the pale" — Hitler was "beyond the motivational reach of the relevant moral considerations" (Ibid., 7). By contrast, Stalin, another murderer on a mass scale, could have been "reached". Harman argues that Stalin may have "realised [sic] that the course he was going to pursue would mean the murder of millions of people and he dreaded such a prospect; however, the alternative seemed to offer an even greater disaster" (Ibid., 7). As Stalin seems to have been responsive to the relevant reasons — regarding the value of human life, for example — inner judgements about him are entirely appropriate. Hitler was not so responsive, in Harman's view, and hence inner judgements about his mass-murder are not. I j Here, and elsewhere, Harman actually uses the term "agent" rather than "person". However, as Scott Anderson has pointed out (private communication), employing "agent" — a more technical, even logical, concept — would seem to require some explicit account of agency, while employing "person" — a more familiar concept — does not seem to similarly require an explicit account of personhood. In a sense, "person" has some account of personhood "built in" to the concept, while an account of agency must be spelled-out if "agent" is to be used in any important sense. Whether or not this point is accepted, it seems that nothing of significance is lost through this switch in terminology, and much clarity may be gained by use of the more familiar and less technical term. 1 4 If one or several of these implications is false, then the inner judgement is somehow inappropriate. In Harman's example of Hitler, discussed in note 12, Harman does not claim that inner judgements about Hitler are outright false, even if their implications are. Rather, he claims that it is "too weak" a thing to say about Hitler, and that Hitler was "beyond the reach" of the relevant moral considerations. 1 5 It might be argued, as Scott Anderson has suggested (private communication), that is it grossly unfair of Harman to build motivations into the moral "ought" as he clearly does here. There are many realists who would take exception to this claim. Peter Railton, for example, holds, as part of his "stark, raving moral realism" that "a rational agent may fail to have a reason for obeying moral imperatives, although they may nonetheless be applicable to him" (Railton 1997, 138ii-139i). It is even possible that some antirealists would object to Harman's assumption. However, it would be quite difficult to remove or alter this part of Harman's view without drastically altering his overall conclusion. He needs a psychological underpinning within the moral "ought" in order to claim, as he does, that beliefs — and only beliefs — best explain moral observations. Without this underpinning, the assumption that "oughts" motivate in and of themselves, it is very hard to see how Harman could support his claim that beliefs always best explain moral observations without radically changing or significantly adding to his other views. Hence, for the sake of argument, it seems fairer to grant Harman this rather biasing assumption, and show that, nevertheless, he provides neither sufficient argument nor sufficient evidence to support his moral relativism over moral realism. 1 6 Again, Harman's actual nomenclature refers to agents, not people, hence the use of A rather than P here. 1 7 To be technically accurate, the example should be "the boy ought to not set the cat on fire." Phrased in this non-idiomatic way, the example is more clearly of the form "A ought to D," where A and D are as stated. 60 Interestingly enough, the discussion here would apply equally well to moral judgements as to moral observations. Moral judgements are, for Harman, a kind of moral observation: "[i]f you round a corner and see a group of young hoodlums pour gasoline on a cat and ignite it, you do not need to conclude that what they are doing is wrong; you do not need to figure anything out; you can see that it is wrong" (Harman 1977,4). Judging that a person ought not to have done as they did — to continue Harman's example, that the "young hoodlums" ought not to have poured gasoline on the cat and set it on fire — is not a conclusion, it is a sort of observation. Therefore, the "ought" in moral judgements is the same as the "ought" in moral observations. 1 9 If the point on p. 10, that many, and more general, questions are usually of interest rather than one immediate question, this line of argument must surely change dramatically. With the given reasoning, if many questions are under consideration instead, anything that increases understanding of beliefs in general is surely more apt than understanding of one particular belief. It is likely, though, that Harman can still conclude beliefs best explain moral observations, even if many questions are taken to be of interest, although the point must be argued with more subtlety and care. 2 0 Cf. Harman 1977,91-3. 2 1 Harman is very clear that he does not intend a broader definition; for example, that one inflicts harm if one stands by and allows some other cause — another person, perhaps — to directly inflict harm (Harman 1980, 369). 2 2 Again, Harman is clear that he does not intend a broader definition; for example, that one gives aid if one gives help to people to achieve their ends or life's goals (Harman 1980, 369). 2 3 This, on the face of it, undermines Harman's claim that beliefs, and not facts, best explain moral observations. As Scott Anderson has noted (private communication), it is surely a matter of fact whether an agreement, of any sort, exists between people. So couldn't this fact best explain moral observations? However, as pointed out on p. 19, C — according to Harman, implicit in the moral "ought" — seems it must contain the condition that A be party to particular agreements. In other words, the fact that A is party to some agreements is part of the conditions, C, implied by making an inner judgement about A. And, as discussed on p. 15-6, C cannot be the best explanation for moral observations. 2 4 This, on its own, has interesting consequences for the issue of conflicting moral judgements. Take the two claims, "A's doing X is right," and, "A's doing X is wrong." If these expressions are elliptical for "in relation to moral framework M, A's doing X is wrong" and "in relation to moral framework M, A's doing X is right," of course they conflict over truth, and the conflict must be resolved. But, if they are elliptical for "in relation to moral framework M, A's doing X is wrong" and "in relation to moral framework N, A's doing X is right," AfeTV, then the conflict is only superficial. So, some, but not all, conflicts between moral judgements are, therefore, unresolveable rationally on the grounds that the conflict is not cognitive. 2 5 Harman has thus apparently managed to take on the strengths of a group-oriented relativism, by emphasizing the sense in which morality is shared between individuals, a mutual experience rather than something peculiar and particular; and avoided the problems associated with trying to split off groups from "mere" individuals in some systematic way. This is a clear benefit for Harman's position, as such a "splitting off' criterion can easily fail. Will Kymlicka, for example, tries to argue for the following — it must be noted that he does so entirely apart from any metaethical issue — as a means of picking out, to use his term, a "societal culture" from mere masses of individuals. A societal culture must be an (1) "intergenerational community," that is (2) "more or less institutionally complete," that (3) occupies "a given territory or homeland," and that (4) shares "a distinct language and history" (Kymlicka 1995, 18). These are mostly straightforward; the only slightly obscure point is (2). As Kymlicka himself says in elaboration, though, (2) is nothing more than the requirement that a societal culture must be made up of individuals who share common institutions and practices (Ibid., 76). But, suppose there is a lifelong business partnership between two men, who pass this business on to their sons when they retire. Is this, using Kymlicka's criterion, a group? There may be common (business) practices. There may be a particular territory, if the place of business has not relocated in many years. There may be a distinct language and history, such as is shared between any two people with a deep-rooted relationship. In-jokes and unique phrasings and terminology abound, which creates a somewhat divergent "language" (or dialect, to be technically accurate), 61 and, obviously, their personal history is distinctly theirs. The issue of common institutions is not easy to resolve without more detail in the example; but, for the moment, let it be assumed that institutions are shared. The inheritance of the business fulfills the intergenerational requirement. This is, therefore, a legitimate societal culture. And, hence, it must count as a group. From this, it would follow, if a more "group-oriented" moral relativist were to adopt Kymlicka's criterion, (again, note that Kymlicka himself does not do this) as presented, that the people in this business could have their own distinct and true moral beliefs, quite apart from those held by any other societal culture (such as, for example, the city in which the business operates). Something must have gone wrong if this is the result of coupling group-oriented relativism with Kymlicka's criterion of a moral group. A simple family business surely shouldn't be counted as morally distinct. This case serves to highlight the difficulty in delineating legitimate groups from mere masses of people in such a way that neither delineates what clearly should be illegitimate groups, nor fails to delineate what clearly should be legitimate groups. This is no more and no less than a grain problem: group-oriented relativism needs a criterion of moral groups that is neither too coarse, in allowing too many groups to count as legitimate moral groups, nor too fine, in not allowing enough groups to count as legitimate moral groups. Kymlicka's criterion is just one example of a criterion that fails to meet this requirement. Harman avoids this, however, by explicitly denying group-oriented relativism — the focus is on individual people — but implicitly borrowing many of group-oriented relativism's strengths, as morality does not end up as a peculiarly individual matter. The possible danger with such a moderate position, though, is that it may, on further examination, prove to be actually one or the other extreme, and so lose its perceived advantage. 2 6 What is it, for example, to accept a principle if not to believe in the principle? 2 7 It may seem rather strange to take agreements as being this sort of convergence of intention. It might even be possible, as Scott Anderson suggests (private communication), for a person to create or destroy a binding moral agreement simply by altering, consciously or not, their intentions — whatever these intentions might be. So, for example, a professor may intend that a student refrain from cheating on an exam; a student may similarly intend to refrain from cheating. This creates a moral obligation on the student to not cheat, under Harman's view. But, if the student were to alter his intentions — perhaps by no longer expecting the professor to hold up his end of the agreement, or by simply intending to cheat after all — then it seems that the original agreement no longer exists. Thus, it would not be immoral for the student to cheat on the exam (though it might be terribly imprudent). Also, note that the convergence of intentions may have been entirely accidental in this case — yet, for Harman, the convergence would still count as an agreement. However, Harman stands quite firmly behind this loose conception of agreements; as he says, "[a moral] understanding [i.e. agreement] will not in general be constituted by absolute rules but will take a vaguer form, specifying goals and areas of responsibility..." (Harman 2000a, 17). So, if Anderson's suggestion is a consequence of Harman's view, it is a consequence Harman must accept. However, Harman does go on to say that the "vague nature of moral understandings is to some extent alleviated in practice.... [as, for example, expectations] are adjusted to other expectations" (Ibid., 17). So, perhaps, the development of agreements over time will lead to something firmer than mere convergence of intentions. 2 8 Note that Harman appears to have changed his mind on this point. Elsewhere, Harman argues that agreements must be real, not hypothetical (Harman 1977, 111). The earlier view is being adopted here as it seems clear that holding the agreements must be real and historical invites a host of easy objections, questioning the whens, wheres, whos, and hows of these "real, historical" agreements. 2 9 Fox's source is Melford E. Spiro, and Audrey G. Spiro, Children of the Kibbutz (Cambridge, MA: Harvard University Press, 1958). j 0 Fox does not define this term. ' l Fox's source is M. Fortes, The Web of Kinship among the Tallensi: The Second Part of an Analysis of the Social Structure of a Trans-Volta Tribe (New York, NY: Oxford University Press, 1949). 62 j 2 Fox's source is M. H. Wilson, Reaction to Conquest: Effects of Contact with Europeans on the Pondo of South Africa (New York, NY: Oxford University Press, 1961). 3 3 Fox's source is R. Firth, We, The Tikopia (London, UK: Allen & Unwin, 1957). j 4 And, perhaps this one belief is even all there is to the incest taboo. There would, though, need to be a small adjustment made to the usual concept of "belief in order for Harman to make the claim that the incest taboo is a belief. The incest taboo, as a tacit rule, is not conscious — as with many of the rules of grammar, people are not aware of the incest taboo. Beliefs, by contrast, are usually thought to be conscious. However, Harman could simply argue that not all beliefs are conscious. For example, psychotherapy of various types seeks to, ostensibly, "bring out" beliefs that people may not be aware they possess, beliefs which are affecting their behaviour in adverse ways. The incest taboo may be another such belief that could be "brought out," perhaps through psychotherapy or similar mechanism. j 5 Of course, Harman would not put forward a similar sort of account of this non-moral observation as he does moral observations. Hence, "a Harmanian sort of explanation," rather than "Harman's explanation." j 6 This is the inverse-square law, as it applies to sound. Given a point source of power, P, and a sphere with P as its centre and with radius, r, the intensity of the sound at the surface of the sphere, /, is determined by the formula, / = PMnr2. Thus as r increases (i.e. the distance from the source increases), I decreases — the energy of the sound is spread over a larger area the further it is from the source, so the intensity of the sound decreases. j 7 Not "the Harmanian." The discussion is now back to Harman's actual views. 3 8 But see p. 48-53. , 9 Whether or not any explanatory picture incorporating Harman's background views is true is not at issue. The argument in this thesis is intended only to show that Harman's relativist conclusions do not follow from this background. 4 0 This section in particular benefitted greatly from suggestions by Catherine Wilson. 4 1 That is, an argument that proceeds in the following way — there is a universal moral practice (prohibition of incest); this is caused by a universal moral belief (belief that incest is wrong); this is caused by a real moral fact (that incest is really wrong); therefore, moral realism — would be defeated. There may be other arguments for realism available, even if a case such as Hopkins' does turn out to be correct. 4 2 Both Hopkins (Hopkins 1980, 322) and Brent D. Shaw (Shaw 1992, 276) deny the claim that economic factors could override the incest taboo. Shaw, in particular, argues that the economic circumstances usually cited to account for the allowance of incestuous marriage in Roman Egypt are not significantly different from those in other societies which do not allow incestuous marriage. Calculated as the difference between the reproductive rate for non-familial matings and the reproductive rate for familial matings, divided by the reproductive rate of familial matings (Ember 1975,273). 4 4 There is an old mathematical trick which, in a series of operations, proceeds from a statement of equality of some combination of variables to the claim that 1=2. There are many variations on this trick, such as the following. 1. heta=b. 2. Then a2=ab, 3. a2+a2=a2+ab, 4. 2a2=a2+ab, 5. 2a2-2ab=a2+ab-2ab, 6. and 2a2-2ab-a2-ab. 7. This can be written as 2(a2-ab)=l(a2-ab), 8. and cancelling the (a2-ab) from both sides gives 1=2 (Spencer 1998). The trick lies in the last line. The equation is divided by a fairly innocent looking collection of variables and coefficients, (a2-ab). But, by (1), (a2-ab)=(b2-b2)=0. So, the division was actually division by zero, which is 63 technically undefined (as (q x 0)=0 for any q, thus 0/0 is indeterminate; and (q x 1)=0 for no q, thus 1/0 is undefined). Therefore, 1 does not actually equal 2; rather, the reasoning used to support the conclusion was sophistic. The reasoning is therefore considered unreliable and suspect, due to that mistaken step. 64 W O R K S CITED Benedict, Ruth. 1985. A Defense of Moral Relativism. In Vice and Virtue in Everyday Life: Introductory Readings in Ethics, ed. C. H . Sommers, 204-12. San Diego, C A : Harcourt-Brace. Originally published in The Journal of General Psychology, 10(1934): 59-82 Ember, Melvin. 1975. On the Origin and Extension of the Incest Taboo. Behavior Science Research. 10(4): 249-81. Fox, J. R. 1962. Sibling Incest. The British Journal of Sociology. 13 (2): 128-50. Harman, Gilbert. 1968. Enumerative Induction as Inference to the Best Explanation. The Journal of Philosophy. 65 (18): 529-33. 1977. The Nature of Morality: An Introduction to Ethics. New York, N Y : Oxford University Press. 1980. Moral Relativism as a Foundation for Natural Rights. Journal of Libertarian Studies, 4: 367-72. 1986. Moral Explanations of Natural Facts--Can Moral Claims be Tested Against Moral Reality? The Southern Journal of Philosophy, 24 (Supp.): 57-68. 1998a. Precis of Moral Relativism and Moral Objectivity: Precis of Part One. Philosophy and Phenomenological Research, 58(1): 161-9. 1998b. Responses to Critics. Philosophy and Phenomenological Research, 58 (1): 207-13. 1999. Wide Functionalism. In Reasoning, Meaning, and Mind, ed. Gilbert Harman, 235-43. New York, N Y : Oxford University Press. Originally published in Cognition and Representation, ed. Stephen Schiffer and Susan Steele (Boulder, CO: Westview Press, 1988): 11-20. 2000a. Moral Relativism Defended. In Explaining Value, ed. Gilbert Harman, 3-19. 65 Oxford, U K : Clarendon Press. Originally published in Philosophical Review, 84 (1975): 3-22. 2000b. Justice and Moral Bargaining. In Explaining Value, ed. Gilbert Harman, 58-76. Oxford, U K : Clarendon Press. Originally published in Social Philosophy and Policy, 1 (1983): 114-131. 2000c. Simplicity as a Pragmatic Criterion for Deciding what Hypotheses to Take Seriously. In Reasoning, Meaning, and Mind, ed. Gilbert Harman, 75-92. New York, N Y : Oxford University Press. Originally published in Gruel: The New Riddle of Induction, ed. Douglas Stalker (Peru, IL: Open Court, 1994): 153-171. Hempel, Carl G. 1968. Explanation in Science and in History. In The Philosophy of Science, ed. P. H. Nidditch, 54-79. New York, N Y : Oxford University Press. Originally published in Frontiers of Philosophy, ed. R. G. Colodny (Pittsburgh, PA: University of Pittsburgh Press, 1962): 7-33. Hopkins, Keith. 1980. Brother-Sister Marriage in Roman Egypt. Comparative Studies in Society and History, 22: 303-54. Kymlicka, Wil l . 1995. Multicultural Citizenship. Oxford, U K : Clarendon Press. Middleton, Russell. 1962. Brother-Sister and Father-Daughter Marriage in Ancient Egypt. American Sociological Review, 27 (5): 603-11. Miller, Alexander. 2002. Realism. The Stanford Encyclopedia of Philosophy, (Winter 2002). Edited by Edward N . Zalta. http://plato.stanford.edu/arcMves/win2002/entries/realism/. Railton, Peter. 1997. Moral Realism. In Moral Discourse and Practice: Some Philosophical Approaches, eds. Stephen Darwall, Allan Gibbard, and Peter Railton, 137-63. New York, N Y : Oxford University Press. Originally published in Philosophical-Review, 95 (1986): 163-207 Shaw, Brent D. 1992. Explaining Incest: Brother-Sister Marriage in Graeco-Roman Egypt. 66 Man, 27: 267-99. Spain, David H . 1988. Taboo or Not Taboo: Is That the Question? Ethos, 16(3): 285-301. Spencer, Philip. 1998. Classic Fallacies. http://wvvw.math.toronto^ Stevenson, Charles. 1997. The Emotive Meaning of Ethical Terms. In Moral Discourse and Practice: Some Philosophical Approaches, eds. Stephen Darwall, Allan Gibbard, and Peter Railton, 71-82. Edited by Stephen Darwall, et al. New York, N Y : Oxford University Press. Swoyer, Chris. 2003. Relativism. The Stanford Encyclopedia of Philosophy, (Spring 2003). Edited by Edward N . Zalta. http://plato.stanford.edu/archives/spr2003/entries/relativisrn/. Winch, Peter. 1964. Understanding a Primitive Society. American Philosophical Quarterly, 1 (4): 307-24. 67 


Citation Scheme:


Citations by CSL (citeproc-js)

Usage Statistics



Customize your widget with the following options, then copy and paste the code below into the HTML of your page to embed this item in your website.
                            <div id="ubcOpenCollectionsWidgetDisplay">
                            <script id="ubcOpenCollectionsWidget"
                            async >
IIIF logo Our image viewer uses the IIIF 2.0 standard. To load this item in other compatible viewers, use this url:


Related Items