What’s Wrong with Global Cognitive Relativism?
Texte intégral
Introduction
1Global Cognitive Relativism (henceforth: GCR) is the view that all truths are true only relative to a frame of reference (a theory, a language game, a conceptual scheme).1 A common opinion among critics of GCR is that it is self-refuting. What I will here call the ‘standard’ argument to this effect was first formulated by Plato (cf. his Theaetetus) and can be found repeated somewhat variously expressed throughout the subsequent history of philosophy. According to the standard argument, holding hat truths are true only relative to a frame of reference leads to an impossible dilemma. To see why, consider the following proposition:
R All truths are true only relative to a frame of reference
2Why does the relativist believe R? So long as she isn’t willing to accept that if R is true, it is false, she cannot answer that she believes R because R is true (absolutely). So she must answer that she believes R because R is true relative to a frame of reference she accepts. But this answer leaves the relativist with no reason to believe R over not-R. For that R is true relative to a frame of reference the relativist accepts is perfectly compatible with R being false absolutely (or to speak in relativist terms: it is compatible with R being false relative to a frame of reference accepted by the absolutist). In Nagel’s words:
…the claim “Everything is relative” must be nonsense, for it would itself have to be either relative or absolute. But it can’t be absolute, since in that case it would be false if true. And it can’t be relative, because then it would not rule out any absolute claim, including the claim that it is absolutely false2. (Nagel, 1997, 15)
3Is the standard argument a good argument against relativism? Boghossian (not a relativist) doesn’t seem to think it is. For, he notes, just because the relativist cannot rule out absolutism using absolutist arguments, doesn’t mean that she cannot do so with reference to pragmatics: even if R is compatible with not-R, it may still be preferable to it on pragmatic grounds. In Boghossian’s words:
…it is not clear that it follows from the concession that relativism is itself to be true only relative to a theory, that it is just a report of what the relativist “finds it agreeable to say.” Perhaps relativism is true relative to a theory that it pays for us all to accept, relativists and non-relativists alike. (Boghossian, 2006, 54)
4Suppose for now (we’ll have reason to return to this verdict towards the end of this paper) that Boghossian is right: the relativist could in principle defend her view against that of the absolutist with reference to pragmatic reasons that it pays for all of us to accept. Then the standard argument does not succeed in showing that GCR is self-refuting. Does this mean that GCR is not self-refuting? Not according to Boghossian. After rejecting the standard argument against GCR, Boghossian proposes a new argument for the same conclusion. In this paper, I argue that Boghossian’s new argument against GCR fails.
1. A New Argument Against GCR
5In Fear of Knowledge (2006), Boghossian presents and defends what he takes to be a better argument for the conclusion that GCR is self-refuting. To set up his argument, Boghossian asks us to consider any proposition we take to be true. To illustrate, consider:
(1) There have been dinosaurs.
6If you’re a relativist, (1) is true only relative to some frame of reference. Hence, says Boghossian, to claim (1) amounts to claiming that…
(2) According to a frame of reference that we accept, there have been dinosaurs.
7What is the status of (2)? If it is true absolutely, GCR is refuted. If it is not true absolutely it must be true relative to some frame of reference. This is in line with GCR. However, says Boghossian, it is in line with GCR because when we claim (1) we are not only claiming (2) but, thereby:
(3) According to a frame of reference that we accept, there is a frame of reference that we accept, according to which there have been dinosaurs.
8It is easily seen that things do not, indeed cannot, stop here. Rather, when we claim (1) (i.e., any proposition), we in effect claim:
(∞) According to a frame of reference that we accept, there is a frame of reference that we accept, and according to this frame of reference that we accept there is a frame of reference that we accept and …∞… there have been dinosaurs.
9In other words: we’re off on an infinite regress. Why is this a problem? According to Boghossian, it is a problem because:
…it is absurd to propose that for our utterances to have any prospect of being true, what we must mean by them are infinitary propositions that we could not express nor understand. (ibid, 56)
10The real dilemma facing the proponent of GCR is therefore this: either some truths are true absolutely, or our utterances express infinitary propositions that we can neither express nor understand. They cannot be true absolutely, as this contradicts what GCR says, and they cannot mean something that we can neither express nor understand (for obvious reasons). Therefore, GCR ought to be rejected.
2. Boghossian’s Argument – A Critique
11In this section I argue that, because the infinite regress Boghossian takes GCR to give rise to is most likely not vicious, his new argument against GCR fails.
2.1. Truth not Meaning
12According to Boghossian, the regress that results from relativizing the truth of any proposition to a frame of reference is vicious because it turns that proposition into one “that we can neither express nor understand”. To see why Boghossian thinks it does, consider his argument in a bit more detail:
1. GCR (assumption)
2. <p> is true (for arbitrary p) (assumption)
3. If <p> is true, then what <p> means is that <According to a T, that we accept, p> (1,2)
4. If <According to a T, that we accept, p> is true then what that proposition means is that <According to T, that we accept, there is a T, that we accept, and according to this T, p> (1,2,3)
5. …etc. ad infinitum
6. Hence: If GCR then what <p> must mean is <According to T, that we accept, there is a T, that we accept, and according to this T, that we accept, there is a T, that we accept, and …∞… p> (1-5)
7. Hence: <p> is really an infinitary proposition (6)
8. Hence: since infinitary propositions are propositions that we can neither express nor understand, GCR must be rejected (reductio of 1)
13The crucial step in Boghossian’s argument, we can now see, appears already in line 3: If <p> is true, then <p> must mean <According to a T, that we accept, p>. It is this claim that eventually leads Boghossian to conclude that <p> is really an infinitary proposition, and it is this conclusion that makes him reject GCR for reasons of absurdity.
14Why does Boghossian think that, just because, according to the proponent of GCR, <p> can be true only relative to a frame of reference, what p must mean is that <According to a T that we accept, p>? The only reason I can think of is that Boghossian accepts a view of meaning according to which meaning is truth-condition (cf. e.g., Davidson 1967). If this is the reason, however, the argument is seriously compromised. The truth-conditional theory of meaning is one that far from everyone accepts, and it is certainly not one that the proponent of GCR must accept: in light of the problems accepting a semantics along these lines would seem to land her in, the proponent of GCR is therefore free to interpret the argument above as a reductio, not against GCR, but against understanding meaning in truth-conditional terms.
2.2. External not Internal
15According to Boghossian, if GCR is accepted, and <p> is true, <p>’s truth requires the truth of <According to a T, that we accept, p>. As we have just seen, Boghossian interprets this as saying that <p> must mean <According to a T, that we accept, p>. Suppose we reject this. Then the reason why the regress is vicious cannot be that it entails a proposition which meaning we can neither express nor understand. But if it cannot, (why) is this regress vicious?
16In order to answer this question, we must first decide how the regress to which GCR commits us is structured. Infinite regresses can be structured in one of two ways: ‘internalistically’ or ‘externalistically’. If we take the GCR-regress to be internalistically structured, it comprises an infinite succession of failed attempts to, in this case, ‘nail down’ what it is that we posit in the regress’s first step. Because the regress is infinite, it will feature infinitely many such (failed) attempts, the ‘last’ of which will be infinitely complex.
17If the regress is externalistic, on the other hand, <p> has its meaning independently of the regress, yet <p> is true only if <According to a T, that we accept, p>. If the regress is externalistic, therefore, it is a truth- not a meaning-regress. If the regress is externalistic, moreover, rather than at each step of the regress, what came before is replaced with something more accurate, at each step of the regress, something distinct and different from, something over and above, what came before exists. Which means that, rather than featuring just one (infinitely complex) proposition, the regress will feature infinitely many.
18If the regress is externalistic, the meaning of <p> is finite and, therefore, perfectly graspable. This means that those who, like Boghossian, think that the regress is vicious because it forces us to accept that the meaning of <p> is infinitely complex, must regard the regress as internalistic. But that the regress is vicious for this reason is, as we have seen, highly doubtful. So, perhaps the regress isn’t internalistic after all? Ironically, that it isn’t seems to something with which Boghossian agrees. For, even though he insists that the regress is vicious because it makes <p> infinitely complex, he also says things like:
If it isn’t simply true that we accept a theory according to which there have been dinosaurs, then that must be because that fact itself obtains only relative to a theory that we accept. (ibid., 55-56, italics added)
19And:
At each stage of the looming regress, the relativist will have to deny that the claim at that stage can be simply true and will have to insist that it itself is true only relative to a theory that we accept. (ibid., 56)
20Those statements make it sound as if what GCR commits its proponents to, is that the truth of any <p> (however complex) requires the existence of some (distinct and different) truth of the form <According to a T that we accept…p> to guarantee it. This is hence a regress of the following—externalistic—form (where what appears at each stage of the regress is to be understood as distinct and different from what appears at the stage before as well as after):
<p>
<T1(p)>
<T2(T1(p))>
<T3(T2(T1(p)))>
∞
21Is this a vicious regress? In the next section, I argue that it most likely isn’t.
2.3. Benign not Vicious
22If the regress to which Boghossian directs our attention is an externalistic regress we don’t end up with an infinitely complex proposition, and so the reason why the regress is vicious cannot be because of that. To make matters worse, quite a few philosophers (cf. e.g., Segelberg 1999 [1945]; Orilia 2006; Gaskin 2008; Morganti 2014) have argued that externalistically structured infinite regresses are not vicious. To justify the claim that the existence of the regress poses a threat to GCR, therefore, we must come up with a (new) reason to think that the regress is vicious, one that either applies to all regresses of the externalistic kind (in which case we must demonstrate that and why those philosophers who deem such regresses benign are wrong) or one that applies exclusively to the GCR-regress, externalistically understood (in which case we must be able to explain what sets this regress apart from the others).
23In what follows, I will consider two influential reasons against thinking that externalistic regresses are benign: (i) that, because the regress is infinite, whatever feature it is we want to attribute to what exists in the regress’s first step (be it justification, existence, or truth) can never start getting ‘passed down’ to it; (ii) that, because what exists at each new step in the regress (except for the first) both guarantees and depends upon what exists in the step before, for the regress to get going, that what exists in the regress’s first step is justified/exists/is true must be presupposed, and so cannot be explained. With respect to each of these reasons I will argue that it does not provide us with a reason to deem the GCR-regress vicious.
2.3.1 Why the No-Starting-Point Reason is a Non-Starter
24In an externalistic infinite regress, something is F (is justified, exists, is true) in virtue of something distinct from it (that is also F). And that thing is F in virtue of something distinct from it (that is also F). Etc. ad infinitum. In an externalistic infinite regress, therefore, nothing is F fundamentally. The problem with this, some philosophers have argued, is that, for anything to be F, something must be F fundamentally. Therefore, the externalistic regress is vicious.
25(Why) must we accept that for anything to be F, something must be F fundamentally? To answer that question, I suggest we take a closer look at this demand as it features in discussions of two types of F: justification and existence.
26Take justification first. According to the so-called infinitist (cf. e.g., Klein 2007), justification requires ‘appropriately structured chains of reasons.’ Unlike foundationalists (who claim that appropriately structured chains of reasons are non-repeating finite chains of reasons) and coherentists (who claim that they are repeating finite chains of reasons), the infinitist holds that appropriately structured chains of reasons are non-repeating infinite chains of reasons. According to the infinitist, in other words, that the chain of justification is without end, does not hinder what exists in its first step from being justified. Quite the opposite: that it is, is what justification requires.
27Infinitism has met with a lot of criticism. According to e.g., Fumerton, to accept infinitism is to accept that justification is inferential. Yet if justification is inferential, infinitism cannot be true. Why not? Because:
…[f]inite minds cannot complete an infinitely long chain of reasoning. (1995, 57)
28And:
…if all justification were inferential we would have no justification for believing anything. (ibid.)
29It is unclear if Fumerton thinks that the two reasons cited above are really one and the same. In any case, according to Klein (1998, 920), they should be distinguished. Taken at face value, what the first reason says is that, for a belief to be justified for someone, that person must have actually completed the process of reasoning to that belief. This is a very (indeed an unreasonably) strong demand to put on justification: according to Klein (ibid.), whether we are foundationalists or infinitists about justification, it leaves most of our beliefs unjustified. Much more plausible (and from the point of view of the infinitist: challenging) is therefore the second reason. According to Klein, to say that, if all justification were inferential, we would have no justification for believing anything, is to point to a distinction between, on the one hand, having a justification and, on the other, being justified, and saying that, just because p has a justification (e.g., q) doesn’t mean that p is justified. It is to say that, for p to be justified, something besides q is required. What? According to Fumerton:
We might want to require in addition that there be some legitimate chain of reasoning leading from non-inferentially justified beliefs to the conclusion in question, the existence of which is causally sustaining the belief. (ibid, 106)
30Suppose Fumerton is right: for p to be justified something besides q is needed and this something extra is a series of beliefs causally sustaining p. Does it follow that this series must have a ‘beginning’ in some non-inferentially justified beliefs? Not according to Klein. For, as he points out, so long as we do not (implausibly!) require that the person whose belief is in question has actually completed the infinitely long process of reasoning to the belief:
…unless Fumerton can produce an argument to show that either we never do (or never can) have an infinite, non-repeating set of beliefs that serve as reasons for and causes of the succeeding beliefs in the chain of reasons, the causal requirement will not aid in the rejection of infinitism. (ibid, 921-22)
31Is there some (other) reason to think that the chain of reasons (or, more generally, the chain of whatever-we-are-for-the-moment-debating) leading up to what exists in a regress’s first step must have a ‘beginning’? To answer that question, I suggest we turn our attention to another debate concerning the viciousness of another externalistic regress: the existence-regress.
32In its most straightforward form, an existence-regress is one that is generated on the assumption that something (anything!) exists, together with the assumption that, for anything to exist, something must ground its existence. The result is a regress where, at each step exists a (grounding-)fact that is distinct and different from the fact that exists at the step before. The existence regress in its most straightforward form is hence an externalistic regress.
33Other forms of the existence-regress are generated on the assumption that something specific exists. One famous example of an existence-regress of this kind is the so-called Bradleyan regress (named after F. H. Bradley 1893). The Bradleyan regress is generated on the assumption that some things sometimes form a complex unity (i.e., that complex unities possibly exist), together with the assumption that, for anything to form a complex unity, those things must be related. This regress is most commonly (and rightly, I think, but for the purposes of this paper this is neither here nor there) understood as internalistic. Thus understood, what exists at each step of the regress is a more accurate rendering of what exists at the first: the relata of the supposed relation and an ever-increasing number of relations but, crucially, never a complex whole:
a b
a R b
a R R R b
a R R R R R b
Etc. ad infinitum
34According to some philosophers, however, the Bradleyan regress is externalistic. On this view, what exists at each step of the regress is distinct from what came before it. The result is a regress along something like the following lines:
(ab)
R(ab)
R(R(ab))
R(R(R(ab)))
Etc. ad infinitum
35This regress, its proponents suggest, is not vicious. Why not? Because unlike what was true of the regress understood internalistically, already in the regress’s first step does there exist a complex whole. On this view, in other words, since what exists at each step of the regress is distinct from what exists at any other step of the regress, and since what exists at each step of the regress, after the first, is there to guarantee the existence of what exists at the step before it, rather than no complex whole, the regress gives us an infinite number of them, each guaranteeing the existence of the complex whole that came before it.
36Just like the justification regress had to face its completion challenge, those who insist that the existence regress is benign must face the challenge of explaining how, if the chain of beings is infinite, existence can start getting passed down through it. That it cannot has been argued by several philosophers. A version of this objection can be found already in Leibniz (Letters to Arnauld). More recently, it has been proposed by Schaffer and (in a weaker form) by Cameron. According to Schaffer, first, if the chain of beings is infinite:
[b]eing would be infinitely deferred, never achieved. (2010, 620)
37And according to Cameron:
…[if] there are infinitely many levels of facts, the obtaining of each depending on the facts at the next level /…/ it is hard to see how things could get off the ground in the first place. (2008, 3)
38Why think that the fact that dependence goes on to infinity would have this effect? According to Cameron, this is because of something like the following principle (what he calls ‘an intuition’):
…when there is an infinite chain of entities e1, e2, e3, …, or an infinite chain of facts f1, f2, f3, …, then while e2 may ontologically depend on e1, and e3 on e2, etc., and while f2 may obtain in virtue of f1 and f3 in virtue of f2, etc., it is impossible for e1 to be ontologically dependent on e2, and e2 ontologically dependent on e3, etc., or for f1 to obtain in virtue of f2 and f2 in virtue of f3, etc. There must be a metaphysical ground, a realm of ontologically independent objects, and a realm of basic facts which provide the ultimate grounding for all derivative facts. (Cameron 2008, 8)
39Call this principle the principle of ontological well-foundedness (WF). (Why) must we accept (WF)?
40Besides the fact that many take (WF) to be a highly intuitive principle, the main-reason why people seem to think that it is a principle we must accept is that, unless we do, we’re forced to admit that there can be explanation in the absence of a ground or source of explanation. What does existence have to do with explanation? One plausible answer is that existence is intimately connected to explanation because ontological dependence is grounding-dependence, and grounding is an ‘explanatory relation’.
41Here’s a reason not to think that the fact that grounding is an explanatory relation gives you any special reason to accept (WF): grounding is explanatory, but only in the sense that grounding backs (metaphysical) explanation. Grounding is a worldly relation. Explanation is not. Although grounding is an explanatory relation, therefore, that explanation must have a source does not give us any special reason to think that grounding does too.
42Here’s another reason: even if grounding is and not just backs (metaphysical) explanation, it is most likely not the sort of explanation that needs a source. Why not? Because the reason why most seem to think that explanation must have a source is that they think that unless it does, the role understanding plays in explanation will be compromised. In Orilia’s words:
...if in an attempt to explain [the existence of] P I begin an explanatory task wherein at every stage I must presuppose a succeeding stage, then there is no increase [of understanding]. For any such increase is an approximation to the final stage and if there is no such stage, then there is no explanation. (2006, 232)
43That explanation is intimately tied to (increased) understanding is uncontroversial. However, if grounding is a relation in mind-independent reality, and if grounding, thus understood, is ‘explanatory’, the sort of explanation grounding is does not seem to be the sort of explanation we have in mind when we say things like explanation is intimately tied to (increased) understanding. And if it isn’t, then that ‘ordinary’ explanation is tied to (increased) understanding does not mean that grounding must be as well.
44And here’s a third: Suppose that grounding, in being explanatory, is intimately tied to (increased) understanding. Then there’s reason to believe that this does not mean that grounding/explanation must have a source. This is the view Orilia ends up defending. For, as he points out:
… [t]hat at any given stage we can continue the explanatory task does not show that no knowledge or no understanding is provided at any stage. It merely shows that at no stage do we know/understand everything that there is to know/understand about the explicandum which gives rise to the explanatory chain. And noting that the explanandum in question gives rise to such an infinite chain may be considered part of our understanding of it. (ibid.)
45Finally, even if (WF) is a principle that sometimes regulates theorizing, it does not seem to be a fundamental principle in the sense that it is a necessary principle. According to Cameron, (WF) is at its strongest a ‘fundamental methodological rule of thumb’. For, he points out:
…there is no justification for the claim that the intuition is necessarily true. The status I am attributing to the intuition is like that enjoyed by Ockham’s razor: we should accept it because if it is true the theories we arrive at give a better explanation of the phenomena to be explained, and hence are more likely to be true. But such principles of theory-choice do not appear necessary; it is not as if the world is necessarily such that the simplest explanation is the right one—we just hope that our world is like this. (Cameron, ibid: 13)
46If Cameron is right, then (WF) need not be accepted. In which case the proponent of GCR (or any of the other regress-invoking positions) need not accept it. In the absence of any other, stronger, reason for (the analogue of) (WF), therefore, it does not seem as if the fact that in an externalist regress, existence, justification, or truth is ‘infinitely deferred’ must mean that it is ‘never achieved’.
2.3.2 Why the Symmetry Reason Doesn’t Generalize
47Another reason to think that the externalistic regress is vicious is that it instantiates a pattern of (ontological) dependence that makes it vicious. To see what I mean by that, consider the existence regress in its Bradleyan guise. If that regress is externalistic, then in the regress’ first step exists a unified whole (ab). An infinite regress is then generated because (ab) exists in virtue of R(ab) (which exists in virtue of R(R(ab)), etc. ad infinitum). In other words: the existence of (ab) depends on the existence of R(ab) (and so on). So far, no surprises. However, just as (ab) depends for its existence on what comes next in the regress, there is a sense in which, what comes next in the regress depends for its existence on (ab). This is the sense in which the existence of the regress requires the existence of that thing. To see why, suppose that what exists in the first step of the regress is not a complex unity (ab). Suppose, instead, that what exists at that step is the (disunified) aggregate of a and b. In that case, what exists in the regress’s next step cannot be something which operates on a complex unity, thereby ‘guaranteeing’ its existence. Rather, what exists in the regress’ next step is an R, added to the aggregate in an attempt to glue its members together. Which would make the regress internalistic, not externalistic. Therefore, in an externalistic regress, there is a pattern of symmetric existential dependence between what exists in the regress’s different steps.
48(Why) is this a problem? I can think of at least two reasons. The first, comparatively speaking weak, reason is that, if relations of existential dependence are explanatory, then if A and B symmetrically existentially depend on each other, A explains B and B explains A. Which, given that one also thinks that there can be no symmetrical (metaphysical) explanation, is bad.
49This is a weak reason for thinking that the regress is vicious because it depends on a prior notion of what is and what isn’t an explanation, a notion that the proponent of benign externalistic regresses can choose to ignore. It is a weak reason, moreover, because even if we were to agree that symmetrical explanations are non-explanations, we need not agree that if A and B depend on each other for their existence, A explains B and B explains A. This is because, even if (metaphysical) explanation does require existential dependence, most agree that for there to be explanation, something more or besides existential dependence is needed (explanation, but not ontological dependence is hyperintensional).
50The second reason one might think that the mutual dependence of what exists at the different steps of the regress gives rise to viciousness is that that this in some way prevents, not explanation, but explanation of the right kind. To illustrate, consider the Bradleyan regress. On at least one interpretation, this regress is generated in our attempt to explain how possibly there can be complex unity. The ontological infinitist thinks that how possibly there can be complex unity is explained by what exists after what exists in the first step of the regress. However, although what exists after the first step of the regress does explain why what exists in the regress’s first step exists, it does not seem as if it explains how possibly it exist. Why not? Because, in order to explain how possibly a complex unity exists, we must explain how that complex unity, and not just the aggregate of its constituents (can) exist(s). It is unclear if the infinitist can explain this, however. Since existential dependence is symmetrical, what exists in the step(s) after the first step of the regress will depend on the existence of what exists in the regress’s first step. And what exists in that first step is a complex unity. Which means that, for there to be a regress able to guarantee the existence of what we find in the regress’ first step (i.e., a complex unity), that complex unities (possibly) exist must be taken for granted (for more of this argument, cf. e.g., my 2013).
51Does the fact (assuming it is a fact) that this makes the Bradleyan regress vicious generalize? Here’s a reason to think that it doesn’t: in a GCR-regress, what is at stake is truth. In the regress’s first step exists a proposition, <p>, which truth is guaranteed by what exists in the regress’s next step, etc. ad infinitum. Although this regress and the Bradleyan one are both externalistic, therefore, they are still pretty different. For one, even though <p> depends for its truth on what comes next in the regress, it doesn’t seem as if it depends for its existence on that thing. <p> may very well exist (since it’s a proposition, perhaps it must exist) yet be false. If there is symmetrical dependence, therefore, it is not of the existential kind.
52And there doesn’t seem to be symmetrical truth-dependence either. Yes, <p> is true in virtue of <Tp>, but the dependence is one-way: <Tp> is not true in virtue of <p>. The truth of <Tp> does depend on there being a meaningful p for it to guarantee the truth of. But that is not the same as to say that its truth depends on the truth of <p>. If it did, that would be a problem. For then, to be able to account for the relativistic truth of any <p>, that in terms of which we give an account would, in turn, have its truth depend on the (absolute) truth of <p>. But we have no reason whatsoever to think that it does. Therefore, even if we judge externalistic existence regresses vicious for reason of symmetrical dependence, we shouldn’t do the same with the GCR-regress.
53I’m sure there are other reasons why one might think that externalistic regresses and, in particular, the externalistic GCR-regress, is vicious. I cannot think of one for the moment, however. And if there isn’t, that there is this regress is not in itself a reason to reject the view from which it came.
3. All Is Not Lost: Pragmatics to the Rescue!
54Is all lost for those of us itching to reject GCR? Not necessarily. Recall that to Boghossian, the problem with the traditional argument against GCR was that it overlooked the fact that, though the relativist cannot distinguish R from not-R with reference to absolute truth, she might be able to do so with reference to pragmatic betterness.
55According to van Inwagen (2015), however, even if we concede that pragmatic considerations could (even should) play a role in sorting propositions into good and bad piles, they fail to do so exactly when it comes to the proposition of relevance to the outcome of the traditional argument. For, as he points out, though it is easy to see the benefits of holding that lions are carnivorous over its opposite, it is hard to see what practical difference a belief in R over its opposite could make. This holds even if we take into account also “theoretical” betterness, he adds. Though we can certainly see the “betterness” of a conclusion drawn validly from true (or, in this case, pragmatically good) premises over one drawn invalidly and from false (or pragmatically bad) premises, it does not follow from this that R is in this sense better than not-R. No knock-down argument exists (at least on the side of R) and if a poll were to be made among professional philosophers (a rather dubious way of deciding on matters of theoretical betterness to be sure…) it is very likely that not-R would win. But then, even if Boghossian fails to strengthen the traditional argument against GCR, all is not lost. For, perhaps that argument didn’t need strengthening in the first place3.
Bibliographie
Boghossian, P. (2006). Fear of Knowledge. Oxford University Press.
Bradley, F. H. (1893). Appearance and Reality. Oxford University Press.
Cameron, R. (2008). Turtles All the Way Down: Regress, Priority and Fundamentality. Philosophical Quarterly, 58(230), 1-14.
Davidson, D. (1967). Truth and Meaning. Synthese, 17(1), 304-323.
Fumerton, R. (1995). Metaepistemology and Skepticism. Rowman & Littlefield.
Gaskin, R. (2008). The Unity of the Proposition. Oxford University Press.
Klein, P. (2007). Human Knowledge and the Infinite Progress of Reasoning. Philosophical Studies, 134(1), 1-17.
Klein, P. (1998). Foundationalism and the Infinite Regress of Reasons. Philosophy and Phenomenological Research, 59(4), 919-925.
Maurin, A.-S. (2013). Exemplification as Explanation. Axiomathes, 23(2), 401-417.
Morganti, M. (2014). Metaphysical Infinitism and the Regress of Being. Metaphilosophy, 45(2), 232-244.
Nagel, T. (1997). The Last Word. Oxford University Press.
Orilia, F. (2006). States of Affairs: Bradley vs. Meinong. In V. Raspa (Ed.), Meinongian Issues in Contemporary Italian Philosophy (pp. 213-238). Ontas Verlag.
Rorty, R. (1998). Truth and Progress. Cambridge University Press.
Schaffer, J. (2010). Monism: The Priority of the Whole. Philosophical Review, 119(1), 31-76.
Segelberg, I. (1999[1945]). Three Essays in Phenomenology and Ontology. Thales.
Tiercelin, C. (2013). No Pragmatism Without Realism. Metascience, 22, 659-665.
Tiercelin, C. (2017). Pourquoi le Pragmatisme Implique le Réalisme. Cahiers Philosophiques, 3(150), 11-34.
Tiercelin, C. (2018). The Economy of Research and the Proper Defense of Knowledge and Intellectual Virtues. Transactions of the Charles S. Peirce Society, 54(2), 183-207.
Van Inwagen, P. (2015). Metaphysics. Westview Press.
Notes de bas de page
1You could read this as saying that truths are true in virtue of our imposing our theories, language games, or conceptual schemes on reality. But saying this assumes the existence of something absolute: reality! Taking the ‘global’ in global cognitive relativism seriously means not presupposing that there is some way things are independently of our ways of theorizing about them. It means accepting a view on which (Boghossian 2006, 35; cf. also Rorty 1998, section I): “[n]o sense can be made of the idea that reality is a certain way in and of itself. And no sense can be made of the idea that the mind causes the world to be a certain way, through its use of descriptions.” It is relativism in this sense that is the topic of this text.
2Quoted from Boghossian (2006, 53): where Nagel uses ‘subjective’ and ‘objective’, Boghossian uses ‘relative’ and ‘absolute’. Since Boghossian’s terminology fits better with mine, I’ve opted for sticking to it!
3For a full-blown defense of the view that pragmatism (especially pragmatism of the kind defended by C. S. Peirce) should be combined with realism, not nominalism or, for that matter, relativism, cf. Tiercelin (2011, 2017, 2018).
Auteur
-
Anna-Sofia Maurin
Université de Gothenburg
Le texte seul est utilisable sous licence Licence OpenEdition Books. Les autres éléments (illustrations, fichiers annexes importés) sont « Tous droits réservés », sauf mention contraire.
Dans le labyrinthe : nécessité, contingence et liberté chez Leibniz
Cours 2009 et 2010
Jacques Bouveresse Jean-Matthias Fleury (éd.)
2013
Dans le labyrinthe : nécessité, contingence et liberté chez Leibniz
Cours 2009 et 2010
Jacques Bouveresse Jean-Matthias Fleury (éd.)
2013
