Neo-Meilandian Truth-Relativism of a Weak Sort

Peter Davson-Galle

University of Tasmania



[1] Relativism occupies a confusing place in our cultural and intellectual life. In some ways, Jack Meiland's 1977 remark that "our age is an age of relativism" (Meiland 1977: 568) applies still. Relativism certainly seems prominent among literary theorists and social-scientists (including a good number of meta-scientists) yet it is rather swiftly dismissed by most philosophical commentators. One such writer, Jamie Whyte, recently went so far as to observe that
relativism is even sillier than it at first appears. Indeed, if relativism were not so popular, it wouldn't be worth discussing at all. And, even given its popularity, it isn't worth discussing for long. The only excuse can be that you intend to provide the last words on the matter. And that is my excuse: I hope to provide the last words on relativism. (Whyte 1993: 112)
Perhaps regrettably, Whyte is too sanguine and, although the crude varieties of relativism which he criticises are unworthy of belief, that is not to say that no version of relativism fares better.{1} Most positions in philosophical controversies have something to be said for them, in some version or other, and I find the generally swift dismissal of relativism by philosophers uncomfortably swift. As Alisdair MacIntyre has remarked in this context, "relativism... is one of those doctrines that have by now been refuted a number of times too often" (MacIntyre 1989: 182). So, the burden of what follows is to see whether some form of relativism can be crafted which is not open to the usual objections. I am confident that this paper will by no means constitute the last words on the subject; rather, to borrow Whitehead's phrase, it will be but another "footnote to Plato," or, more aptly, to Protagorus.

What Sort of Relativism?

[2] Relativism is, of course, not so much a thesis as a shifting collection of theses of dubious clarity and unclear inter-relationships.{2} I shall be concerned with but one of these: relativism about truth. The thesis{3} of relative truth has been called : "the Achilles' heel of relativism"{4} and, in some versions at least, has been criticised from Plato on. Can a viable form of it be crafted? Further, if it is relativism about truth under discussion, then, presumably, whatever conception of it that one comes up with will sit within one of the three broad analytic theories of truth: the correspondence, coherence and pragmatic theories. On the face of it, the last two theories look to be the more congenial homes{5} but, in what follows, I shall attempt the unpromising looking task of developing a notion of correspondence truth relativism. The suggestion offered here has its genesis in a proposal by Jack Meiland (1977) that relative truth be explicated as a three place relation. The key passage of his paper is as follows:
(1) The concept of absolute truth seems to be a concept of a two-term relation between statements (or perhaps propositions) on the one hand and facts (or states of affairs) on the other. But the concept of relative truth, as used by some relativists, seems to be a concept of a three-term relationship between statements, the world and a third term which is either persons, world views, or historical and cultural situations.
(2) The relation denoted by the expression 'absolute truth' is often said to be that of correspondence. The relativist can make use of this type of notion and say that "P is true relative to W" means something like "P corresponds to the facts from the point of view of W" (where W is a person, a set of leading principles, a world view, or a situation). (Meiland 1977: 571){6}

From Meiland to Neo-Meiland

[3] The first adjustment I wish to make to Meiland's suggestion is to restrict what the third relatum, the W, might be. Closest to what I want on Meiland's list is: "a set of leading principles" or: "a world view". Indeed, a largish, but loose, family of such notions are used by relativists and, in a later work, Meiland speaks of the relativiser (or third argument place in the relation of relative truth) as: "conceptual schemes, conceptual frameworks, linguistic frameworks, forms of life, modes of discourse or thought, Weltanschaungen, disciplinary matrices, paradigms, constellations of absolute presuppositions, points of view, perspectives, or worlds" (Meiland and Krausz 1982: 84). And even that list is not complete; one could add: "world versions, categorial frameworks, networks, episteme etc." Which of this bewildering array do I offer as the W?

[4] Israel Scheffler comes close to stating what I wish to say when he draws a distinction. Of it, he says:

This distinction may be drawn in different ways, but the variations are irrelevant to the main issues at hand. We may express it, for example, as a distinction between concepts on the one hand and propositions on the other, between general terms or predicates on the one hand and statements on the other, between a vocabulary on the one hand and a body of assertions on the other, between categories or classes on the one hand and expectations or hypotheses as to category membership on the other. (1967: 36)
My proposal is that the W of Meiland's formula be the former part of Scheffler's distinction and I shall adopt the turn of phrase "categorial web" ("W" for short) for this. "Categorial web" because it is plausible that various concepts are linked to others such that the applicability (or otherwise) of some has implications for the applicability of others. How far this might go is something I see no point in pursuing as, for present purposes, debates between semantic holists and semantic "discretists" are an un-necessary complication. (I avoid the more common "conceptual scheme" to avoid the unwanted suggestion of holism.)

[5] I seek recourse to this distinction as I am persuaded that having a body of substantive assertions (a theory, say) as the relativiser has intractable difficulties.{7} Having drawn the distinction, a key feature for what follows is that the one conceptual web can be used to make conflicting statements. For instance, whatever the theoretical motivations two interlocutors might have for their statements, the conflicting assertions: "phlogiston exists" and: "phlogiston does not exist" have the same associated conceptual web. We shall return to this point.

[6] The "W" given some initial clarification, let's return to Meiland's three place relation suggestion for explicating the concept of relative truth.

[7] Although admitting that his explication is incomplete and problematic, Meiland takes it to be adequate "to begin to show that Husserl and other absolutists are making a great mistake by assuming that relative truth must be either nothing at all or else a variety of absolute truth" (Meiland 1977: 572). Crucial to the viability of the explication then is that the notion of relative truth does not include the notion of absolute truth (574). In further clarification of his idea, Meiland says that we should not view the form of a relative truth claim as appropriately expressed by: "P is true for W", for that raises the question: "what does 'truth' mean in this claim?" and encourages us to answer: "absolute truth" and thus have our notion of relative truth include that of absolute truth. Rather, we should hyphenate, to get: "P is true-for-W". Here "true" has no independent semantic role and is to be considered a mere part of a term "true-for-W" much as "cat" in "cattle" is but a semantic fragment and not to be taken to mean the same as "feline" (574).

[8] In comment upon Meiland's paper, Harvey Siegel complains about this that the cases are disanalogous (Siegel 1987: 13-14). Though "cat" in "cattle" is not a meaningful part, surely "true" in "true-for-W" is. "For it is, after all, a conception of relative truth..."(14) being offered. Siegel continues to say that, this being so, it has not been shown by Meiland that the concept of truth involved in the "true" of "true-for" is not that of absolute truth. I have some sympathy with Siegel's complaint here. In what sense, if any, are the concepts of absolute and relative truth both concepts of truth? What seems required of Meiland is a more general notion of truth which has (at least) two sub-species. If something can be done along these lines, then he could either withdraw his hyphenation and have it that "true" in "true for W" meant that broader concept, or retain his hyphenation and have "true-for-W" meaning a sub-species of the more general notion of truth (with absolute truth as another sub-species). On this latter option "truth-for-W" would be meaningful only as a whole and the whole would be a label for the sub-species, relative truth, of the broader notion, truth. Use of the hyphenated label would be just a visual reminder that this is, after all, a (relativistic) variety of the (more general) notion of truth. But, were it to prove muddling to absolutists (who might tend to write in another sub-variety, absolute truth, as what's meant by "true" in "true-for-W") then perhaps some other, less misleading, technical label like "relth" (or "Jack", or "Harvey"!) ought be substituted.

[9] All of this seems to me to be fine and dandy except that it leaves obscure just what this more general notion of truth might be such that it allows of these two sub-varieties. Meiland does, to my mind, go some way towards meeting this concern with his discussion of correspondence. Speculating as to why Husserl might see any notion of truth as having to include the notion of absolute truth he says that Husserl "perhaps thinks of absolute truth as correspondence with reality and also thinks that any form or variety of truth has to involve correspondence with reality" (Meiland 1977: 573). Meiland responds by "distinguishing between two-term correspondence and three-term correspondence. In other words we can bring both absolute truth and relative truth under the more general concept of correspondence with reality, although these two types of correspondence may differ considerably from one another" (573-574). This is fine as far as it goes, but, as Meiland recognises (580), it's still obscure what might be meant by the putatively three-term relation of correspondence, "P corresponds to reality for W", or, for that matter, what the more general, or generic, notion of correspondence might be. Meiland's response is to claim that, obscure though the three-term relativist notion of correspondence (and presumably the more general notion) is, "the relativist is in no worse a position than the absolutist at this point" (580).

[10] Siegel is unimpressed by this and claims that the relativist is worse off but, as he doesn't here consider conceptual web variants of W, his discussion doesn't quite mellifluously fit in to our present one and I won't directly address his remarks.{8} Rather, I will attempt to develop Meiland's notions further.

[11] The traditional correspondence theory of truth has been widely criticised concerning the obscurity of each relatum and of the relation of correspondence between them but, at an intuitive level at least, it seems clear enough. There's a world "out there", just one world and there's an objective way that it is (it is, if you like, one only of a range of "possible worlds" or ways that things might have been). Moreover, people have conceptions of it{9}, and make statements about how it is, and these statements might correspond to how the world actually is or they might not.

[12] Whatever difficulties might emerge upon closer analysis there is nonetheless an obvious commonsense clarity to this picture. Does the relativist have any such intuitively graspable three-term notion of correspondence, one that as easily gets to an intuitive "first base" of understanding? It's not clear to me that she has. The hard thing to grasp, it seems to me, is just how the conceptual scheme, the W, fits in. One obvious way is that any P will employ the categories and concepts of a W but I don't see how this helps distinguish the relativist's views for even the absolutist will say that of course any P employs some W but, that done, and P's sense established, the truth or falsity of that P is a two-place matter of correspondence of P and reality. Either reality is, or it is not, such that P corresponds to it.{10} If it is not, the problem might well be that the W employed by P is the culprit, that the world just does not contain the kinds of thing postulated by W (and presupposed by P) and thus P, presupposing W, won't correspond to it.{11} But to admit this is merely to note one source of non-correspondence of P and the world{12} and constitutes no stimulus for conceiving of truth/falsity as some sort of peculiar three-term relation.

[13] But if this involvement of W won't suffice, what is the contribution of W to a three-term correspondence? It's not obvious. If I'm right in the above, then it is a legitimate challenge to Meiland to say that there is more of a problem for the relativist than the absolutist in providing a way of understanding his suggestion about truth, even admitting it to be embryonic.

[14] Also, if we have both a two and a three-term relation of correspondence, what warrants us in deeming both to be sub-varieties of a more general notion? What, if you like, are the similarity and difference relationships of the three notions?

[15] I think that something can be done in answering these challenges that is plausibly construable as relativistic and also seems a development of Meiland's ideas, and thus is, as I shall call it, "neo-Meilandian".

[16] As I've said, I favour the third relatum being something like a conceptual web but even without exhaustive treatment of the hydra-like multitude of issues and controversies that emerge when one explores the suggestion, something more has to be said to clarify matters before it can be judged that the idea is at least a "starter". Even so, and donning a Meilandian cloak of humility, I acknowledge that what follows is incomplete and problematic; it is, however, promising, and may be the most that an aspiring truth relativist can have, at least within the context of some sort of correspondence conception of truth.

[17] Before I proceed any further with the development of my neo-Meilandian proposal, it will help avoid mis-understanding if I make some background assumptions explicit. I have earlier signalled two things about the to-be-proposed three place relation of relative truth. The first was that the third relatum would be the conceptual web associated with the candidate statement. The second was the conception would be that of a sort of correspondence truth. I take the latter point to mean that, in some sense, the world plays a role as truth-maker for statements. That is, the truth of a statement will, in some sense be a matter of it being answerable to how reality is. If we go back to Meiland's formulation, we find the phrase: "corresponds to the facts from the point of view of W". Now, it might seem that, were there to be multiple realities tied to, or even constructed by, differing conceptual webs, then the correspondence relativist's task would be made simpler. Perhaps so, although I have my doubts.{13} In any event, it strikes me as too much like thievery instead of good honest philosophical toil (to adapt Russell's quip) and I shall attempt to craft a viable notion of relative truth that is applicable even in less ontologically lush circumstances. (As we shall see, this will help us avoid Whyte-style worries about flat earth conceivers living on a flat earth, while round earth conceivers live on a round earth and so on).

[18] However, though I shall assume there to be perhaps "just one lot of stuff out there", and not rival, or alternative, worlds, I shall make no stipulation about that stuff's nature (beyond something I shall return to soon). That is, the world's stuff will not be assumed to be restricted to the "gritty" stuff of common sense or even the oddities of theoretical physics. It might include gods, hobgoblins, non-corporeal minds, abstract entities (like numbers and abstract propositions) and so on. Accordingly, I shall henceforth just speak of the world and its stuff.

[19] So, given all of this, can a viable concept of relative truth be offered? I think so, though it might be more boring than relativists might hope.

[20] In explication of Meiland's idea of a general notion of correspondence with two-place (absolute truth) and three-place (relative truth) sub-varieties, consider the following.

[21] First, the general notion, what I shall call: "The generic concept of correspondence truth" (GCCT).{14}

GCCT: "P is true (simpliciter)" means "the world is such that it is as P portrays it to be". (And the corresponding concept of falsehood {15} is simply that the world is not as P portrays).
This is just a version of the usual vague correspondence theory's intuition of true statments corresponding to the facts or to the way things are. It is vague, yes, but all participants in this particular dispute are correspondence theorists and thus presumably consider the usual difficulties in spelling out a tighter formulation to be just that, difficulties, not catastrophic objections to the core intuition. Also, it is more the task of the following, more specific, varieties to cash out this general conception. Let us move on, then, to the explication of the notions of absolute truth and relative truth; consider the following.

[22] I take it that part of the positive idea of truth that absolutists have is that there just is a way that the world is, that it just has structures and features and its bits stand in relations and so forth and it is this natural patterning of, and relationships among, the world's stuff that, if we're lucky, we might, in part, capture with a true statement. So far, this is pretty close to a simple (though still vague) expansion of the generic notion; so let me select a little. A distinction can be drawn between two sorts of categorisation of the stuff of the universe. In the first, the categorisation might capture a natural kind, in which nature is, as it is sometimes put, "carved at the joints". Whatever the difficulties for the notion of a natural kind{16}, I take it that the core intuition is clear: the stuff of the universe falls naturally into kinds and these categories are thus, in some sense, reified and not merely imposed upon the universe as some sort of useful artefactual classification of ours. Natural kind terms (should there be any, as opposed to putative natural kind terms) and conventional classification terms differ, then, in the ontological status of the categories of stuff they correspond to. And, of course, putative natural kind terms might fail to be natural kind terms yet succeed in referring to stuff which forms a conventional class.{17}

[23] A statement (such as many would conceive of some statements of science to be) which, for example, purports to pick out natural kinds and contend that they stand in some relationship seems able to be true/false in a clear-cut world matching way. Either the world is such that it contains those kinds and they stand in that relationship or it is not. This seems a two place correspondence notion. And I shall employ it in the crafting of a concept of absolute truth (CAT).

CAT: "P is absolutely-true" means "the world is such that its stuff forms kinds as P portrays and is otherwise as P states".
And the corresponding notion of absolute falsehood is that the world is such that its kinds are not as P portrays or is not otherwise as P states it to be.

[24] In contrast, consider statements which do not purport to pick out natural kinds but just conventional classes{18} and, again, say, contend that those classes stand in some relationship. Such a statement does not seem to be capable of truth/falsity in quite as clear-cut a world matching way because, although the world-stuff so classified is there to be matched, the classifications are not, or not in the same ontological sense. The status of the third element, the conceptual web, is different; it is, so to speak, more autonomous of the nature of the world than a natural kind categorisation. It is not, however, totally autonomous in that the world might resist being classified in some ways even if they are understood as mere conventional classifications. That is, not just any old attempt at conventional classification of the world's stuff will succeed; try as one might, "phlogiston" and "hex" will, perhaps, be empty classifications.

[25] My suggestion is that this conceptual distinction{19} between "kinds" and "classes" be drawn upon to craft a notion of relative truth.

[26] Accordingly, I offer, as a three-place relation between a statement, the world and the conventional conceptual web associated with the statement, the following concept of relative truth (CRT):

CRT: "P is true-relative-to its associated conventional conceptual web, W" means "the world is such that its stuff is able to be conventionally classified by W as P portrays and is otherwise as P portrays".{20}
The corresponding notion of relative falsehood is that the world is such that its conventional classifiability is not as portrayed or is otherwise not as P portrays.

[27] Having outlined these three concepts (GCCT, CAT and CRT) some discussion and clarification is in order before I employ them to frame a substantive relativist thesis about truth.

[28] First, note that (almost) any participant in the dispute would accept that some of the ways in which we categorise the universe are mere artefactual classifications. Moreover, on pain of fussing verbally to no great point, almost any participant might, however grudgingly, go along with the above concept of relative truth and thus accept that some statements are but true/false relative to their associated conventional conceptual web.

[29] Second, note that for a conceptual web to be associated with a statement is not for that statement to be thereby committed to the applicability of any particular one of those concepts. Just how big a web is associated with any given statement is, as I have indicated, not a matter that I see any present point in pursuing, however, I ought re-emphasise an earlier point that I take it as axiomatic that the presence of any concept in the conceptual web automatically entails the presence of the complement or negation of that concept. This is not of great importance when considering statements of the form: "A's are B's" but it becomes more significant when we think of disputes between statements of that sort and "A's are not B's" or "no A's exist". For such statements to be such that, say, it can be coherently claimed that "A's are B's" and "no A's exist" are contraries, their truth/falsity have to be relativised to the same conceptual web, though clearly the former statement holds that the classification "A" is applicable and the latter does not. So, in short, a distinction is to be drawn between the ontological commitments of a statement (which might include that some conceptual classification concept has application, that the stuff of the world is such that it can be so classified) and the conceptual web associated with the statement. Thus, "A's exist" and "no A's exist" share an associated conceptual web containing at least the pair A/not-A, though they differ in ontological commitment.

[30] Thus, if one goes back to the statement of CRT, the turn of phrase: "...classified by W as P portrays..." is hopefully clear. It has been chosen to allow for a P like: "no A's exist" to be true-relative-to P's W (which would include the conventional classification "A") even though the world's stuff is not able to be classifiable, even conventionally, by "A". That it is not so classifiable is indeed the key statement of P. (In this case the denial of class term applicability is quite explicit but in other cases it might be less direct, by entailment).{21} Similarly, one way in which such a P might be false would be if, contrary to its claim, more of its W were to be applicable than portrayed by P. (If, in the above case, "A" were to constitute a way of conventionally classifying the world).

[31] Note further that, as I make no particular assumption about the mutual exclusiveness of absolute and relative truth, it might be that a statement is both absolutely and relatively true. W might pick out categories that form not just artefactual classes but also natural kinds. Moreover, as might be predicted from some earlier remarks, some statement, P, might be absolutely false but relatively true. The former because the terms of its W fail to pick out natural kinds as portayed in P; the latter because the world is conventionally classifiable as portrayed in P. This last point emphasises that, though there is an associated categorial web for absolute as well as for relative truth, the crucial difference is the ontological status of that W. But, having noted this, it might seem strange that CRT but not CAT was worded as a three place relation. Appearances deceive, however. Siegel remarked of Meiland's three place relation suggestion that, to be genuinely three place, the three putative relata must be genuinely distinct (1987: 12). Construed as a web of artefactual classifications, W satisfies this requirement but would not were it to form a web of natural kind terms. I see the possibility of rival W's in the former case, but not the latter, as the key point here. In any event, whatever the detail of the conceptualisation of absolute truth might turn out to be, it is not my central concern here, relative truth is.

[32] I would emphasise that CRT is not able to be the vehicle for one common truth-relativist view. In its terms it is not possible to have the one statement, P, true-relative-to one conceptual web, W1, and false-relative-to another, W2. P will have only one associated W (the same as that of not-P, see earlier). Chris Swoyer has called such a common relativistic view: "strong relativism" (1982: 92). I take strong truth relativism to be a hopeless cause{22} and CRT is offered instead as an explication of what Swoyer called "weak relativism". The idea with weak truth relativism is that a statement could be true relative-to one W, and simply unstatable relative to, or in the terms of, another.

[33] Clearly the interest of the weak neo-Meilandian concept of relative truth, CRT, as a vehicle for at least some of the cognitive insights/tendencies of relativists would be dependent on the possibility of radically different conceptual webs. It might well be that we have two statements, P1 and P2, say, with two associated conceptual webs, W1 and W2, which, while different on the face of it, constituted mere verbal variations of the same web. That is, each might have the conceptual resources to be inter-translatable with the other and thus P1 and P2 be able to stand in some logical relations with one another. Were this to be the only sort of variation of conceptual web associated with statements that (in some sense that I can leave unexplored) engage with the same phenomena, then the interest of CRT would be minimal. What is clearly wanted is that the same stuff be radically differently conceived of by two W's and that there be the possibility of the world's stuff accepting each classification. Moreover, although many in the dialogue would allow that this might, to some extent, obtain (see earlier) what will be of importance will be the extent to which the world's description was only able to be done in terms of "competing", incommensurable, and but relatively true, statements.{23} But here we are going beyond stating and clarifying the mere idea of relative truth on offer and beginning to think of interesting theses that might be formulable in terms of CRT. Let us then proceed explicitly to that task.

The Thesis of Truth Relativism

[34] What, then, might be a subtantive thesis of relative truth which incorporates such a concept of relative truth? I shall state an extreme thesis and then examine it to see if it suffers the usual problems of viability. So, consider the thesis of radical truth relativism (Th. RTR).
Th. RTR: "All true{24} statements are truths-relative-to the conceptual web of conventional classifications associated with the statement in question".

[35] That is, for any P to be true is for it to be but true-relative-to its W. Or, even more briefly, there are no absolute truths.

[36] On the view I'm exploring, the motivation for saying this would be an ontological thesis, that no natural kinds exist but only conventional or artefactual classifications.

[37] The traditional challenge to relativistic theses about truth has been to query the status of the thesis itself. If it is not itself but relatively true, then, if true, it refutes itself by constituting its own counterexample.{25} If, on the other hand, a thesis of relative truth is taken to speak of itself, then, in virtue of being, at best, but relatively true, it might seem to parochialise its interest, to render itself irrelevant to those for whom relative truth is relativised to another conceptual web.

[38] I confess to considerable uncertainty as to the direction that a truth-relativist might be advised to go here. I have, in the past (1991), argued that perhaps a relativist is unwise to express her thesis too unrestrictedly and ought restrict it to statements made about the stuff of the universe, to, what I called then, object-linguistic claims. Theses of relativism about truth (including a suitably toned down, or less sweeping, version of Th. RTR) would thus not be included in their own remarks because metalinguistic. This might seem to be a particularly apt move in the case of the present version of relativism about truth, for the concept of relative truth upon which it was built itself had its genesis in an ontological thesis about the categorisation of stuff. That is, the focus seemed to be on statements about the stuff of the universe, not on meta-level statements about these statements. As will emerge, I now have qualms about the usefulness of this object-statement/meta-statement distinction in this context; however let's explore it. As a first approximation of such a restricted thesis of relative truth about object-linguistic statements, consider Th. O-LTR:

Th. O-LTR: "All true object-linguistic statements are truths-relative-to the conceptual web of conventional classifications associated with the statement in question."

[39] On the face of it, Th.O-LTR captures what we were after, for, being a meta-linguistic statement, it manages to avoid self-reference and thus self-refutation via providing its own counter-example. However, it is not as apt a restriction/clarification of the truth relativism thesis as it seems. One of our aims in the ploy was indeed to have the truth relativism thesis not refute itself, and Th. O-LTR manages that, but it doesn't mesh as well with the rationalisation offered as it might. Unless we construe the restricted thesis as an ad hoc move undertaken simply out of fear of truth-relativism refuting itself, we will want to be taking the object-language/meta-language rationalisation seriously in any revised statement of the thesis. Yet, for all that Th. O-LTR says to the contrary, non-object-language statements, like itself , might also be but relatively true. A more satisfactory restricted thesis would thus be:

Th. O-LTR*: "All and only true object-linguistic statements are truths-relative-to the conceptual web of conventional classifications associated with the statement in question."

[40] So, how would this avoid self-reference? Obviously, "because it is not an object-linguistic statement" would be the answer offered. And, at a superficial level of examination, Th. O-LTR* is not object-linguidtic; it talks, not of the world's stuff, but of statements talking about the world's stuff. Or so it seems; however I am not at all sure that this stands up to deeper examination. After all, the ploy's success depends a lot on what statements are and on what classification concepts are. In short, it is not clear that the sort of object-language/meta-language distinction that is commonly spoken of can bear the ontological weight that is being placed upon it here.

[41] I have no wish to pursue the chapter and verse of this as a topic in its own right, for whatever the details of a satisfactory analysis might turn out to be is beside our present purpose; for that purpose, suffice it to say there seem to be three types of possible theory. First, a statement might be taken to be some, perhaps quite complex, social/behavioural phenomenon. Second, it might be held to be something in the head of a cognitive agent (or in those of some community of such agents) or, perhaps, in some non-corporeal mind(s). Third, statements might be held to be denizens of some abstract realm (to be propositions).

[42] If I'm right, and any theory of what statements are fall into one or other of these categories, then it is not at all clear to me that it is inappropriate to ask whether statements form natural kinds or are but conventionally classified.

[43] Consider the first ontological hypothesis about statements, that they are some sort of social/behavioural phenomena. Whatever this amounts to on various theory-variants, statements would seem to most plausibly be phenomena of the stuff of the universe in some way or other.{26} Similarly for brains. Non-corporeal minds and their goings-on might seem more difficult to accommodate but I don't see why. Presumably minds are part (albeit a "ghostly" part) of the stuff of the universe{27} and thus, as much as matters of brain or social/behavioural practices, fair game for questions about the ontological status of categorisations of their goings-on. Finally, should there be an abstract realm of statements (of propositions, I suppose) as has, on occasion, been suggested, then it, too, is either a part of the universe, however queer a part, or, if some other realm, or "universe" why wouldn't just the same questions be raised about the status of such abstract stuff's categorisations as would be in the case of more common or garden stuff?

[44] In short, the crucial issue is that of the ontological status of categorisations and it doesn't seem of significance, in this context, to draw a distinction between statements employing concepts carving up stuff of a straightforward sort and meta-statements about such statements.

[45] But, if this is so, then the hope that I and others have had of avoiding relativism's prospect of self-refutation by restricting its domain of application, as in Th. O-LTR*, is misguided.

[46] So, if one reverts to Th. RTR, if true, is it absolutely true or true-relative-to its associated conceptual web of conventional classifications? Obviously this is dependent upon whether the categorisation concepts associated with that thesis pick out natural kinds or not. But look at the thesis, what are the associated concepts?

[47] An obvious list would include "statement", "true statement", "conceptual scheme", "conventional classification" etc., plus, in conceptual contrast role, such things as "natural kind". It's hard to see why one ought deem the reference any of these to form a natural kind, not even "natural kind". Even if one did not share our relativist's ontological conventionalism and thought there to be natural kinds (say gold, water etc.) though that would mean that "gold" etc. picked out stuff that formed a natural kind, it wouldn't mean that what "natural kind" picked out (the kind: gold, the kind: water etc.) themselves formed a natural kind.

[48] My judgement here might be wrong but all of this seems to point to Th. RTR being, if true, true-relative-to its associated conceptual web rather than absolutely true. Certainly it would seem that this is how the relativist would be inclined to view her thesis. Such merely relative truth for Th. RTR does not seem problematic to me. The "parochialism" concern is mainly that relativism, if but relatively true, will be true only for the relativist's conceptual web; and absolutism might be true for the absolutist's. But whatever the difficulties of this sort for other conceptions of truth relativism, Th.RTR is obviously not vulnerable to such objections. Note, first, that the denial of Th. RTR employs the same associated conceptual web. That is, the absolutist is not, on the above account, being distinguished from the relativist in virtue of employing a different conceptual web; rather, he employs the same web to make different substantive claims. Thus the absolutist's claim that some true statements{28} are absolutely true would, as this terminology has been set up above, be judged by the relativist to be false relative to that claim's associated web. And, odd though it might seem to those forgetful of the above definitions, the absolutist should, contrarily, conceive of his claim as but relatively true. This does not detract from its (putative) truth status but just reflects the plausible status of the claim's associated categorisation notions as but artefactual.

[49] In similar vein, focusing upon that core ontological issue is interesting. The relativist would subscribe to and hold to be true-relative-to its associated conceptual web, the statement:

Th. CC: "No natural kinds exist, only conventional classifications of the stuff of the universe are possible."
The absolutist would subscribe to and hold to be true, the statement:
Th. NK: "Natural kinds exist".
It's clear that the relativist's view of Th. NK is that it's relatively false but, again, even an advocate of Th. NK playing by the relativist's terminological rules and tying conceptions of relative and absolute truth to the ontological status of the associated web of categorisations, should deem Th. NK to be but relatively true.

[50] I fail to see any of this as problematic. What is important for the absolutist, I would have thought, is the status of at least some statements about the world, notably some of those made by science, not the status of any old statement about the world; and that the thesis of absolutism itself is but relatively true (at best) while it might be a disturbing thought on other, more radical, conceptions of relative truth (ones that are, for instance, more open than CRT to the accusation that they collapse the belief/truth distinction) does not seem disturbing as relative truth is here construed.

[51] I take it, then, that CRT and the thesis Th. RTR employing it are not incoherent or self-refuting, that Th. RTR is best thought of as included in its own scope and is thus (at best) true-relative-to its associated conceptual web. As to whether Th. RTR is relatively true or relatively false, the most important determinant of this is the fate of the dispute between Th. CC and Th. NK, and that question is beyond my paper's scope. All that I was concerned to do was investigate whether a concept of relative truth could be explicated in terms of which to advance a coherent truth-relativism thesis that escaped standard critisms. CRT and Th. RTR achieve this. But even if such truth-relativism "gets off the ground", is it worth fighting for?{29}

[52] I think so; it seems to me that Th. RTR captures a lot of what relativists have intuitively grasped for. Though the weak neo-Meilandian relativism advanced gives the world the power to falsify the relativists' deepest beliefs, this is the price of correspondence relativism about truth. It would be an objection to the thesis were it to collapse truth and belief; this would, to use Siegel's turns of phrase, "undermine the very idea of rightness" (1987: 4) and transgress the proper restraint that "necessarily some beliefs are false" (1987: 6). Yet, though Th. RTR preserves a welcome objectivity, it does so without being open to the earlier noted Husserlian concern, as reported by Meiland, that "relative truth must be either nothing at all or else a variety of absolute truth" (Meiland 1977: 572). CRT is not a variety of absolute truth. But though the world is seen on the above view as, in large part, the arbiter of truth, the determiner of fact, cognitive agents (human or alien) are the determiners of "the form of factuality".{30} And, if sufficiently radically "other"{31} conceptual webs can exist, the forms of factuality might allow for reality being carved up in wildly different ways without rejection. In short, this paper advances what looks like a coherent and genuine, though minimal, thesis of truth relativism. It might not satisfy all of the subjectivist yearnings that seem, on occasion, to manifest themselves in relativist tracts, but then, if nothing coherent can capture such relativism, perhaps relativists ought be satisfied that something seems salvageable; a sound fig leaf{32} is preferable to a self-destroying overcoat!{33}


Beach, E. (1984) "The Paradox of Cognitive Relativism, Revisited: A Reply to Jack W. Meiland." Metaphilosophy 15, 157-171.

Davson-Galle, P. (1988) "Relativism about Truth." Metascience 6, 54-62.

Davson-Galle, P. (1989a) "Truth and White 'Lies' -- A Frank Response." Metascience 7, 6-7.

Davson-Galle, P. (1989b) "The Strong Programme and Reflexive Incoherence." Metascience 7, 99-100.

Davson-Galle, P. (1991) "Can Relativism About Truth Avoid Self-refutation?" Metaphilosophy 22, Nos. 1 and 2, pp.175-178.

Davson-Galle, P. (1993) "Young and Old Arguments about Global Anti-Realist Relativism." Indian Philosophical Quarterly XX, No.4, pp.415-426.

Davson-Galle, P. (1994a) "(Relative) Truth and Whyte 'Lies'." Cogito 8, No.2, (forthcoming).

Davson-Galle, P. (1994b) "Relatively Speaking" (Review of M. Krausz, ed., Relativism: Interpretation and Confrontation and S. Edwards, Relativism, Conceptual Schemes and Categorial Frameworks.). Metascience, (forthcoming).

Davidson, D. (1973) "On the very idea of a conceptual scheme." Proceedings of the American Philosophical Association, 1973-74, pp. 5-20.

Devitt, M. (1984) Realism and Truth. Oxford: Blackwell.

Edwards, S. (1990) Relativism, Conceptual Schemes and Categorial Frameworks. Aldershot: Avebury.

Galle, P. (1983) "Kordig's Paradox Objection to Radical Meaning Variance Theories." Philosophy of Science 50, 494-497.

Goodman, N. (1978) Ways of Worldmaking. Sussex: Harvester Press.

Hollis, M. and S. Lukes (eds.) (1982) Rationality and Relativism. Oxford: Basil Blackwell.

MacIntyre, A. (1989) "Relativism, Power, and Philosophy." In M. Krausz (ed.), Relativism: Interpretation and Confrontation. Notre Dame: University of Notre Dame Press, pp. 182-204.

Meiland, J. (1977) "Concepts of Relative Truth." The Monist 60, pp. 568-582.

Meiland, J. (1980) "On the Paradox of Cognitive Relativism." Metaphilosophy 11, pp. 115-126.

Meiland, J. and M. Krausz (eds.) (1982) Relativism: Cognitive and Moral. Notre Dame: University of Notre Dame Press.

Newton-Smith, W. (1981) The Rationality of Science. Boston: Routledge and Kegan Paul.

Newton-Smith, W. (1982) "Relativism and the Possibility of Interpretation." In (Hollis and Lukes 1982: 106-122).

Nola, R. (ed.) (1988) Relativism and Realism in Science. Dordrecht: Kluwer Academic Publishers.

Okrent, M. (1984) "Relativism, Context, and Truth." The Monist 67, pp. 341-357.

Putnam, H. (1981) Reason, Truth and History. Cambridge: Cambridge University Press. Root, M. (1986) "Davidson and Social Science." In E. LePore (ed.), Truth and Interpretation, pp. 272-304. Oxford: Basil Blackwell.

Sankey, H. (1990) "In Defence of Untranslatability." Australasian Journal of Philosophy 68, pp. 1-21.

Scheffler, I. (1967) Science and Subjectivity. Indianapolis: Bobbs-Merrill.

Scheffler, I. (1980) "The Wonderful Worlds of Goodman." Synthese 45, pp. 201-209.

Siegel, H. (1987) Relativism Refuted: A Critique of Contemporary Epistemological Relativism. Dordrecht: Reidel.

Swoyer, C. (1982) "True for." In (Meiland and M. Krausz 1982: 84-108).

White, F. (1986) "On a Proposed Refutation of Relativism." Australasian Journal of Philosophy 64, pp. 331-334.

White, F. (1989) "Comments on Relativism and Truth." Metascience 7, pp. 2-5.

Whyte, J. (1993) "Relativism is Absolutely False." Cogito, Summer, pp. 112-118.


{1} For criticism of Whyte's paper see my (1994a). Return

{2} A number of people have noted this, see: Nola (1988: 1-36), Hollis and Lukes (1982 1-20), Newton-Smith (1981 and 1982) and my (1988). Return

{3} Or family of theses. Varieties of varieties of relativism about truth on different parameters are possible but will mostly be ignored here. See Swoyer (1982), my (1988) and Meiland (1977). Return

{4} By Swoyer (1982: 84). Though Socrates' main target was truth relativised to an individual cognitive agent, which won't concern us here. Return

{5} As I briefly mentioned in my (1988) and have pursued in my (1993). Return

{6} I remarked in my 1988 paper that "the only apparent hope for relativism about truth (within a correspondence theory was) the three term relation suggestion of Meiland's" (61). I also remarked that the suggestion had not been explored and that to do this "would be a paper in itself" (58). This is an attempt at that paper. Return

{7} See Edwards (1990). Edwards accepts and tries to work within Davidson's rejection (in Davidson 1984) of a conceptual scheme/substantive content distinction. Like Swoyer (1982: 99) and many other commentators, my tendency is to dismiss Davidson's views as overly verificationist, certainly for a paper trying to stay within a correspondence truth framework. Return

{8} Siegel's discussion (17) comes later in his chapter than his purported establishment of the collapse of Meiland's putative three-term relation into a two-term one. Siegel does address relativising truth to conceptual schemes, frameworks of belief and so forth in the second chapter of his book (32-44). Regrettably, though, the variant of conceptual scheme relativism considered here is not discussed. Siegel is more concerned with epistemological relativism than truth relativism (and in my view blurs them too much in his discussion) and with notions of "conceptual scheme" rather more closely related to a substantial framework of belief than here discussed. Return

{9} To use Siegel's turn of phrase (1987: 17). Return

{10} For simplicity's sake, I speak here in an implausibly sharp-edged bivalent way. Return

{11} Clearly if P is something like: "W does not have application", then this will not apply. Return

{12} Which will make P (absolutely) false or, maybe, neither true nor false if one is a Strawsonian about failed reference. Return

{13} For some consideration of this radical path of taking seriously some writers' talk of theorists living in different worlds, see my (1988: 59-62). Return

{14} Actually, there are difficulties associated with the concept of logical truth but I take them to be neutral to the present dispute and shall bypass them. Return

{15} I shall avoid issues arising out of the dispute between Strawson and Russell about failed reference. Return

{16} And I take these to be considerable. Return

{17} A possible example is the pre-modern (and whale including) term: "fish". It is also worth noting that a putatively artefactual categorisation term might manage to pick out a natural kind. Return

{18} I here ignore complications caused by purportedly conventional classification terms managing to pick out a natural kind. Return

{19} Or purported distinction; as we shall see, the truth-relativist will suggest that, as there are no natural kinds, the distinction is but a notional one. Return

{20} Though it shares a rejection of natural kinds with it, I take such conventionalism to involve a distinct notion of truth to that associated with Putnam's internal realism. Putnam's notion is a form of coherence theory. See for instance, Putnam (1981: 49-74). Further, though sharing with Goodman a rejection of natural kinds, the conventionalism sketched above seems, on some extreme "versions" of Goodman's views anyway, less radical. See Goodman (1978); good critical commentaries of Goodman are Siegel (1987: ch.7) and Scheffler (1980). Return

{21} Those interested in the detail of this are referred to chapter two of Schleffler (1967) and chapter two of Edwards (1990). Edwards' book is, among other things, a good treatment of Stefan Korner's views. I review it in my (1994b). Return

{22} Briefly, the difficulty lies in the tension between this being meant to be a rendition of correspondence truth relativism (and thus be giving some truth making role to the world) and the role of the W. The difficulty is dramatised when one thinks about the falsity making powers of the world. Concerning this, a sequenced discussion occurs in Newton-Smith (1981: 34 ff.); White (1986); my (1988); White (1989); and my (1989b). The situation is even worse for the view that it is impossible for there to be statements whose truth is invariant from conceptual scheme to conceptual scheme (a view I once called "the thesis of radical truth variance" -- see my (1983: 495)). Return

{23} It is by no means clear that the incommensurability literature provides much optimism for the existence of rival W's as wanted here (see Putnam (1981: 115) and Davidson (1984: 184); and, in criticism of them, (Sankey 1990)). It would make no difference for my present purposes to have to resort to appealing to the possibility of alien conceptual webs to get a suitably radically distinct W to our own. Return

{24} In the GCCT sense. Return

{25} It seems to me that the charge of "self-refutation" is too lightly bandied about; see my (1989b) and (1990), also see a nice exchange between Meiland (1980) and Edward Beach (1984). Siegel's various writings On this are mostly collected in Siegel (1987). Return

{26} Perhaps this seems objectionably reductionist but I think that a closer look will show that it isn't reductionist at all. Return

{27} I made this an explicit assumption earlier and don't see how relativists wishing to run a language stratification line like that in my 1991 paper could expect challenging it to help their case. Return

{28} Meaning, remember, just GCCT. Return

{29} The phrase is Nelson Goodman's (1978: 20). An issue not addressed in this paper is that of fighting for relativism. Shorn of its terminological proposals, the main substantive thesis is Th. CC. I have held that this thesis can, without awkwardness be held to be but relatively true. An issue not here pursued is whether it can be argued for in a way of interest to someone initially not sharing it. See Siegel (1987: 44) on such problems. Return

{30} To appropriate Mark Okrent's nice phrase; see his (1984: 347). Return

{31} The turn of phrase is adapted from Michael Root (1986: 272). Return

{32} The "fig leaf" metaphor is borrowed from Michael Devitt (1984: 15) who uses it of a minimalist realism. Curiously the minimal (in some ways, though not in scope) relativism discussed above has assumed "fig-leaf realism"! I am aware that my proposal has various unaddressed difficulties, this is, however, a paper, not a book, and at least it seems worth further exploration by any relativistically inclined philosopher who wishes to nonetheless have the world play some sort of truth making role. Return

{33} For their comments on various versions of the above, I am indebted to one of this journal's anonymous referees and to Jack Meiland (especially), Kathy Bohsted, Howard Sankey, Edgar Sleinis, Harvey Siegel and those present at its presentations at the 1991 conference of The Australasian Association of Philosophy and the 1992 conference of The Australasian Association for the History, Philosophy and Social Studies of Science. Hopefully the paper's present appearance will add to this list! Return

Peter Davson-Galle
University of Tasmania

The Electronic Journal of Analytic Philosophy 2:1 (1994)
ISSN 1071-5800
Copyright 1994