Abstract

Contemporary views of truth-directedness endorse what I shall call the Common-Element Argument. According to this argument, there is something in common between judgment and other attitudes like assumption and imagination: they all regard their contents as true. Since this regarding-as-true feature is not distinctive of judgment - the argument goes - it can’t explain its truth-directedness. On this ground, theorists have been motivated to endorse an inflationary view that tries to capture truth-directedness by appealing to some further feature: intentions, second-order representations, sub-personal mechanisms, or subjugation to norms are the most discussed candidates for fulfilling this role. In this paper I will argue that the Common-Element Argument is unsound. It rests on a false premise, namely that there is some common element such as a regarding-as-true component between judgment and other cognitive attitudes. I shall reject Velleman’s and Railton’s defenses of the Common-Element-Argument. Then I will discuss three influential inflationary accounts of truth-directedness: Railton’s account, Velleman’s teleological account, and Shah and Velleman’s conceptualist account. I shall argue that they all face a phenomenological and an explanatory challenge. Finally, I shall sketch a deflationary view of truth-directedness that evades these challenges.

Keywords:
Aim of belief; Judgment; Ethics of belief; Cognitive phenomenology

It seems that judgment - the conscious act of affirming the truth of a proposition - bears a special relationship with truth. Although all attitudes bear some relationship with truth - I desire something to be true, I imagine something to be true, I suppose something to be true, etc. - judgment seems to be special in this respect. Call truth-directness the property (or set thereof) that captures the distinctive way in which judgment relates to the truth.

What is truth-directedness? The contemporary debate offers several different answers to this question¹ 1 For an introduction to the debate see Fassio (2015). . Most answers are motivated by what I shall call the Common-Element Argument (hereafter CEA)² 2 As far as I can tell, CEA is accepted by everyone in the debate. In fact, CEA is rarely presented as a significant debatable move in the debate. It is rather presented as a set of platitudinous remarks that purport to set the stage for a discussion on truth-directedness and related phenomena. See Velleman (2000), Vahid (2006), Steglich-Petersen (2006), Humberstone (1992), Railton (1997). In this paper I shall discuss Velleman and Railton’s versions of the argument (§2 and §3 respectively). . According to this argument, there is something in common between judgment and other attitudes. To illustrate, it is often noticed that both judgment and supposition regard their content as true. But if they have this common element - the argument goes - then we should look for some further ingredient that explains the truth-directedness of judgment.

Theorists of truth-directedness disagree on what this further ingredient exactly is. Intentions, second-order representations, sub-personal mechanisms, or subjugation to norms are the most discussed candidates for fulfilling this role. Despite this disagreement, all these theorists endorse what I shall call an inflationary view of truth-directedness, because they hold that the fact that in judgment we regard the content judged as true is not enough to capture its truth-directedness.

In this paper I will argue that CEA is unsound. It rests on a false premise, namely that there is some common element such as a regarding-as-true component between judgment and other cognitive attitudes. On this ground, I shall open the way to a deflationary account of truth-directedness according to which there is no need to posit some special ingredient like the ones listed above in order to explain the truth-directness of judgment.

The plan of the paper is as follows. In §1 I explain why phenomenology is important in the debate on truth-directedness. In §§2-3 I discuss two influential versions of CEA. I take it that once CEA is defused, then the deflationary view looks like a natural position to take in a theory of truth-directedness. This is why in the rest of the paper (§§4-6) I shall show that inflationary accounts that are motivated by CEA run into problems that can be easily avoided if one rejects CEA. In §7, I conclude by providing a sketch of a particular deflationary account of truth-directedness.

§1 The Phenomenology of Truth-Directedness

The debate on truth-directedness is often presented as a debate on the nature of belief. So I need to clarify why I think it is important to focus on judgment rather than belief.

By 'judgment' I will refer to the conscious act that consists in affirming the truth of a proposition. In this paper I am interested in the topic of truth-directedness as it relates to judgment, and since judgment is a conscious phenomenon, my focus is on the phenomenology of judgment.

Belief is typically understood as a dispositional mental state or a disposition to form judgments³ 3 People who endorse a dispositionalist account of belief according to which if one believes that p, then one has the tendency to consciously affirm that p if one is asked whether p include Cohen (1992), Alston (1996), Schwitzgebel (2002), and Smithies (2012). , and since it is not obvious whether dispositions can be reduced to their conscious manifestations⁴ 4 According to Crane (2014) belief understood as a dispositional state can't have both a conscious and an unconscious life. See Pitt (2016) for discussion. , in this paper I prefer to speak of truth-directedness as it relates to judgment only, rather than belief, for the nature of judgment (as I use the term here and as it is often used in the literature) is exhausted, as it were, by its phenomenology. Whether these considerations can be extended to beliefs depends on the correct account of the relationship between belief and judgment, a topic which goes beyond the scope of the present paper and which requires, anyway, a prior independent characterization of the phenomenon of judgment itself.

The focus on the phenomenology of inquiry is of special importance for the debate on truth-directedness. For one thing, when we inquire about the way in which truth matters for our inquiry, we are interested in understanding something that we do. This process has both a conscious and an unconscious life and the two sides of the coin are related and interconnected - or so we think. But we mostly care about the conscious aspect of inquiry, for this is what we live, this is what we are aware of when we look at the way in which inquiry unfolds.

Second, and crucially, the phenomenology of judgment constitutes a subject matter that enjoys some sort of methodological unity. The phenomenon of inquiry - which includes the acts of judging, questioning, the act of taking some ground as a basis for judging, and many other acts - can be investigated in isolation from considerations about the unconscious life of the mind. To make things vivid, for all I know, I can’t exclude with absolute certainty that what I am living is just a dream, and yet, even if it were only a dream, I could attempt a phenomenological description of the dream and try to uncover its structural features. In particular, even if this experience were some sort of vivid hallucination, I would still form judgments - like the judgment that this might be a vivid hallucination - and I could still attempt an inquiry about the constitutive and distinctive features of the phenomenon of judgment itself. Therefore, I can provide a characterization of the nature of judgment and other inquiry-related phenomena in a way that is independent from an appeal to features that are not given in experience. Moreover, such a phenomenological account is necessary if we want to understand the relationship between conscious and unconscious life (and so, for instance, the relationship between judgment and belief), for if we lack an account of the conscious life we can’t understand the details of its relationship, if any, with the unconscious aspects of the mind.

Finally, in what follows I shall argue that all other existing accounts of truth-directedness either fail to capture the phenomenology of judgment or fail to take it seriously enough, as I think they should. In particular, I shall argue that other accounts of truth-directedness need a phenomenological characterization of judgment in order to explain some of the features that they want to posit as distinctive of the way in which judgment relates to truth. But this is to anticipate too much. First of all, we need to discuss what is seen as the main motivation for holding an inflationary account of truth-directedness.

§2 The Common-Element Argument I: Velleman

It is natural to start one’s reflections on truth-directedness by noticing that in a judgment we judge something to be true. Let’s call this feature of judgment the to-be-true feature. Here is how Velleman comments on this apparently innocuous thought:

Velleman’s point is that the to-be-true feature of judgment is in fact shared with other attitudes - like desire, wishing and hoping. Therefore this quality can’t be distinctive of judgment and can’t capture its truth-directedness.

First of all, in order to assess Velleman’s argument, we must distinguish two interpretations of the claim he is making: in one interpretation, he is claiming that our concept of judgment and other attitudes ascribes to them some common to-be-true component; in the second interpretation, he is claiming that the nature of judgment and other attitudes is such that they have some common to-be-true component. Either way, what is Vellemans’ evidence in favor of his claim?

On the face of it, he seems to be justifying his claim by observing the way in which we speak about such attitudes. However, these conceptual and metaphysical claims can’t be justified on such thin ground. I might say that I raise my hands, that I raise the standards, that I raise money, that I raise the volume, and that I raise my children, but this by itself doesn’t show that there is a common element in all these actions, beside the fact that we use a single analogical expressions for them. Unfortunately, Velleman doesn't give us anything beside this linguistic consideration as an argument for establishing something about the nature (or the concept) of judgment. How can we move beyond linguistic considerations in order to reach conceptual and metaphysical conclusions to the effect that there is some common to-be-true element between judgment and other attitudes?

Let us start with the metaphysical claim. Since we are wondering about the nature of judgment and other attitudes, the natural and obvious way of proceeding is to wonder whether the to-be-true feature that is posited as common between judgment and desire tracks some particular feature in the very experience that involves these attitudes. So, the question is whether judging something to be true is a phenomenon that shares something - the to-be-true feature - with desiring something to be true. For my part, I don't see any such common feature. The phenomenon of judging something to be true is different from the phenomenon of desiring something to be true. In one case, p is taken as being the case, whereas in the case of desire, one is taking p to be a desirable state of affairs in a way which is compatible with awareness that p is in fact false. So, even if it might be natural for us to say that judging is judging something to be true and that desiring is desiring something to be true, the phenomena of judging and desiring themselves do not share any recognizable common to-be-true component in their respective phenomenology⁵ 5 If this phenomenological claim is sound, then why do we speak as if there were some common element between judgment, desire, wishing, and hoping? We say that in judging we judge something to be true, but also in desidering we desire something to be true. One reader of this paper argued that my view might be committed to take “true” as ambiguous, which does not seem to be the case. My view is that there seems to be an ambiguity in the expression “to be”. When we say that judging is judging p to be true, we are saying that p is really true. It is true in being, as it were. But when we say that desiring is desiring p to be true, here we are not saying that p is true in being, but rather that we would like the world to be such that p is the case. Thus, in rejecting CEA we are not committed to take 'true' and its cognates expressions as ambiguous. .

This point about phenomenology suggests that there is no metaphysical ground (in this case, phenomenological ground, since we are wondering about the nature of judgment, namely the conscious act of taking a proposition to be true) for taking the linguistic “to-be-true” component as tracking some metaphysical to-be-true component. The same point is also true, I feel, about the concept of judgment. What I think when I think that judging is judging something to be true is significantly different from what I think when I think that desiring is desiring something to be true. It is not the case that the to-be-true component is thought of in the same way in the two thoughts. I am just thinking about two different things, and not about two ways in which the same thing - namely the to-be-true component - can occur. Analogously, when I think the thought that I am raising my children and the thought that I am raising the volume I am not thinking two thoughts that are composed by the same concept - the concept of raising. I am just thinking about two very different things. Or, at any rate, to think otherwise needs more argument than what Velleman provides us with.

Notice that Velleman's point about the presence of this common to-be-true component in all propositional attitudes was used in order to argue that since the attitudes differ despite their having this common component, there must be some further element that distinguishes them. But if we reject the existence of this common to-be-true component, then we have no pressure to think that there must be some further component that distinguishes these attitudes, and so we lose the main motivation for proposing an inflationary account of truth-directedness. However, let’s concede, for the sake of argument, Velleman’s point, and let's see how he moves forward in the development of his view, for this will prove to be very instructive.

This is how Velleman marks the difference between judgment and other attitudes like desire:

Interestingly, Velleman doesn't take that remark as casting doubts on his claim that when we speak of judging and hoping as judging and hoping something to be true we are speaking about some common element between them. He might have taken the discovery that judgment differs from desire in the way just quoted as evidence for refraining from taking the common way of speaking about all propositional attitudes as evidence that there is such a common element. Instead of taking this natural road, he says that judging and desire have further different entailments. In the case of judgment, judging something to be true entails regarding it as something that is true, whereas the same entailment doesn't exist in the case of desire.

In the case of judgment, according to Velleman, there is a regarding-as-true component which is absent in the case of desire. Yet, after having made that remark, he immediately notices that this component doesn't suffice for distinguishing judgment from other attitudes.

From this discussion Velleman concludes that the difference between these attitudes must be in the aim one has while having them. Using his technical vocabulary, the difference between these attitudes depends on the aim with which they accept a proposition, or regard it as true.

Before coming to the details of his inflationary account (see §5), let's evaluate his reason for thinking that there is something as a regarding-as-true component which is shared by judgment and by the other attitudes he mentions.

Here the argument is analogous to the one we saw about the to-be-true component. A linguistic pattern is noticed, and from this it is inferred that judgment has something in common with other attitudes, either at the level of concepts, or at the level of metaphysics, or at both levels. However, speaking of this common taking, or this common regarding-as-true, or acceptance, seems to be just a useful linguistic shortcut. Again, we should not be led to think that there really is this common component just because linguistic expression suggests that there is such a component. Noticing the linguistic point doesn't by itself tell us anything about the phenomena themselves. Surely, there is more similarity between judging and assuming, say, than between judging and desiring. But if we look at the phenomena themselves we don’t see anything substantial like a regarding-as-true component which is shared by assumptions, judgments, and acts of imagination. To regard a proposition as true in the case of a judgment is to take it as really true, whereas to regard a proposition as true in the case of assumption is merely to pretend that the proposition is true - it is to proceed as if it were true, that is, as if one were effectively judging it to be true. It is not as if the phenomenology of judgment and imagining, say, is identical in one component (or set thereof), and then there is some further component that distinguishes them. It is very hard to think in this way because mental attitudes do not look like the sort of things that can easily be treated in whole-parts terms. It is not that the whole judgment includes, as part, the feature to-regard-a-proposition-as-true, and that somewhere else, in the whole judgment there is the further for-the-sake-of-truth component⁷ 7 This way of thinking about attitudes gets even more problematic if we wonder how to understand the fact that beside the regarding-as-true component judgment must also include the to-be-true component that judgment shares with desires and other attitudes. . These are things that we can distinguish linguistically and conceptually for the sake of clarity. But to give a phenomenological weight to these expressions on linguistic grounds alone is methodologically unsound, and unjustified given the phenomenology itself.

The same point seems to hold at the conceptual level. When I think that in judging I regard a proposition as true I am thinking something very different from what I am thinking when I think that by assuming a proposition I regard it as true. It doesn't seem to me that there is a single concept involved here - the concept of regarding something as true - which picks out a single property. To see the point, notice that Velleman might have made his argument by speaking of a common taking-as-true component, instead of a common regarding-as-true component. But “to take” is an expression that can be used in order to express many different concepts, and this makes clear that the linguistic commonality shouldn't be taken as indicating any conceptual common component. To say the least, further argument is needed here to prove the point that there is something in common between judgment and other cognitive attitudes like assuming⁸ 8 If one doesn't like arguments relying on introspection of thought (though notice that Velleman should rely on his own experience of understanding of the relevant expression in the same way in which I did), we can still make a different point. Even if we were to discover that our concepts of attitudes such as judgment, imagining, etc., are such that they presuppose that there is some common-element between them, we should still ask whether our concepts represent reality in the right way. Metaphysical investigation might prove that our folk psychology is wrong. And in this case phenomenology does not seem to indicate that there is a single regarding-as-true component that is shared by cognitive attitudes. .

§3 The Common-Element Argument II: Railton

Railton (1997RAILTON, P. “On the hypothetical and non-hypothetical in reasoning about belief and action”. In G. CULLITY and B. GAUT (eds.) (1997) pp. 53-79.)⁹ 9 He puts forward a similar argument in Railton (1994). I focus on Railton (1997) since here the argument is more developed. also motivates his account of truth-directedness by appealing to the Common-Element Argument. Railton’s starting point for an analysis of belief as it relates to the aim of truth is Moore’s paradox. He starts by noticing the oddness of

and

Then he asks:

Like Velleman's version of CEA, Railton starts by offering a putative distinctive quality of judgment that would set it apart from other propositional attitudes; then he notices that this quality is also present in other attitudes; yet, judgment differs from other attitudes in that it gives rise to Moore's paradox (which is indicative of the truth-directedness of judgment), whereas some other attitudes don't. This suggests that we should look for some further feature that would distinguish judgment from other attitudes and explain the truth-directedness of judgment.

The structure of Velleman's and Railton’s arguments is the same. They differ in some interesting details though. Velleman speaks of judging/pretending that p as entailing judging/pretending p to be true (to-be-true component) and as entailing regarding p as true (acceptance or regarding-as-true component). Railton doesn't put the point in terms of entailment. He says that “belief is a propositional attitude partly characterized by its representation of its object as true”. Thus, his claim seems to be about the metaphysics of belief itself, rather than about our ways of speaking (and thinking) about belief.

The question is of course whether he is right in claiming that judgment and pretension are both partly characterized by a representation of their objects as true. Interestingly, he doesn't offer any argument for this claim. So, presumably, even if he doesn't explicitly make his point by anchoring it to the way in which we speak about propositional attitudes, he would appeal to similar linguistic considerations in order to back up his observation.

Here the same objections that we leveled against Velleman’s use of CEA apply to Railton’s argument mutatis mutandis. Beside the methodological points discussed above, the most important point to notice is that if we look at the phenomenology of judging it is clear that the way in which judgment represents its object as true is different from the way in which pretension represents its objects as true.

§4 Truth-Directedness as a Second-Order Representation

CEA has had a huge role in shaping philosophical reflection on truth-directedness. By thinking that judgment and other attitudes share some common element like a regarding-as-true component, theorists have been led to posit some further mental ingredient whose role is to make judgment the sort of thing that is distinctively directed at the truth. In what follows I shall discuss some of the most debated inflationary accounts of truth-directedness. Although they significantly differ in their details, I shall show that they all face one or both of the following challenges: a phenomenological challenge to the effect that the inflationary account fails to capture the phenomenology of judgment; an explanatory challenge to the effect that the inflationary account fails to explain why its preferred further mental ingredient is constitutive of judgment.

Let’s start with Railton’s view. Having concluded that both judgment and pretension are partly characterized by a representation of their objects as true, he goes on to say what distinguishes judgment from pretension.

This is surely right. Yet, this point alone doesn't explain in which sense judgment differs from other attitudes. What we are trying to do here is to offer a characterization of the phenomenon of judgment and of related phenomena. The characterization must be such as to allow us to individuate what distinguishes judgment from other attitudes. But mere subjugation to a norm, or mere capacity to be evaluated by a standard of correctness is not by itself a feature that needs to be evinced in the phenomenology itself. My present judgment that p might be evaluated as correct and incorrect according to some prudential standard, say. But the fact that we can evaluate our mental performances in this fashion need not be something to which our mental performances themselves are sensitive in such a way that this sensitivity is detectable in the experience. If the truth is the evaluative standard of correctness of judgment in this way - that is, in such a way that its being sensitive to the standard of truth is not detectable in the experience of judgment - then this doesn't make any difference to the phenomenology and so it doesn't distinguish judgment from other attitudes in the required way.

Railton says that by mastering the distinction between judgment and other attitudes we are understanding that judgments, but not other attitudes, are evaluated as correct if and only if they are true. Surely, this feature about how we evaluate judgment seems to be central in our concept of judgment. But unless this bit of understanding is somehow reflected in our experience of judgment and doxastic deliberation more broadly, this point alone does not suffice to distinguish judgment from other attitudes. This might be a point about concept, but as such it doesn't suffice to conclude anything about the metaphysics of judgment¹⁰ 10 One here might say that all we want to do when we inquire about truth-directedness is to make a point about our concept of judgment. I will discuss and reject this view in §6. .

Railton has something more to say that goes in the desired direction. Though he is not very explicit about this issue, he seems to think that aiming at the truth involves some cognitive effort on behalf of the agent, even a very minimal and quasi-automatic effort, some sort of second-order thought or representation to the effect that in judging one is holding one's judgment as accountable to the truth only. Here are the relevant passages in which he expresses this idea:

The first quotation might be read as suggesting that for something to be a judgment, instead of something else, it (the judgment itself) must represent itself as accountable to truth or evidence only¹¹ 11 The condition for believing that he actually presents in the first and second passage is negative: one should not represent one's own mental attitude as unaccountable to truth-relevant considerations. Yet, from what he says in the third and fourth passage, it is constitutive of belief not only, negatively, the absence of some representation, but also, positively, the presence of some second-order representation. This is why I read the passage as suggesting, however implicitly, the need for the relevant second-order representation. . Taken literally, I don't think it makes much sense, at least it doesn't make much sense on phenomenological grounds: judgments are representational mental acts, but they do not always (also) represent something about themselves.

In the second and third passage he locates the relevant representational component at the personal level: it is the believer who somehow represents her attitudes as being accountable to the truth. Since Railton thinks that the existence of this representation is constitutive of judgment, we should give a very strong reading to this claim: that is, if no such representation is present, one can't even have a judgment. But this is plainly false. Beside the fact that it is unclear what this personal representation is supposed to be, it is surely not the case that when I form judgments I also have a further recognizable representation to the effect that I am forming a mental state accountable to the truth. If this second-order representation is not meant to be phenomenologically salient, then I don't see why we should believe in its existence and, more fundamentally, if it is not phenomenologically salient, then it can't be constitutive of the phenomenon of judgment.

Putting phenomenology aside, the existence of these second-order representations is suspicious for other reasons. It is just unclear what sort of mental acts they are. Surely, they can't be judgments. If they were, judging would be impossible, because for any judgment to be a judgment it will require the presence of a further judgment whose content is somehow the representation of the first-order judgment as being accountable to truth, and this latter judgment will require a further judgment about the second-order judgment, and so on. So, the second-order representation can’t be a judgment.

But if it is not a judgment, then what is it? Since the candidate second-order representations have to represent that the judgment as accountable to truth, it seems that it has to represent the judgment in such a way that it regards as true that that first-order judgment is accountable to truth. Suppose that instead of so representing the mental state it merely amounts to the supposition (or assumption, or imagination, etc.) that it is accountable to truth. In this case, one would not be seriously taking one's state as a judgment, for I might suppose that p while taking p to be false.

So, if the second-order representation can't be a judgment because of the vicious regress problem, it must at least be some sort of proto-judgment or quasi-perceptual state. Let’s even suppose that we have identified a good candidate for fulfilling this role. Yet, in order to do its work, this second-order representation must represent its object as true. But then it is like judgment, by Railton’s own lights, by sharing with it this common element - the representing-as-true component. So, by Railton’s own lights, we need to postulate a further mental state which has the role of distinguishing judgment from the second-order representation. There must be some third-order representation whose role is to represent the second-order representation not as a judgment (that is, not as something that aims at the truth in the relevant sense), but as a quasi-judgment, or something of the sort. Beside the fact that all this seems implausible on phenomenological grounds alone, it seems that we would be led to another even more contorted vicious regress that would make judging impossible.

Instead of postulating mysterious second-order representations, the deflationary account simply notices that in judging we regard the content judged as true in a particular way that is transparent in the phenomenology of judgment itself. The deflationist can then explain why judgment is accountable to truth by appealing to the fact that in judgment we regard the content as true in a particularly committed way that is distinctive of judgment (see §7 below).

Railton’s account also faces an explanatory challenge that, as we will see, is common among inflationary views - though it takes different forms depending on the details of the view under consideration. Railton looks for some further ingredient that would explain the difference between judgment and other attitudes like supposition. This ingredient is some sort of second-order representation that says that judgment is accountable to truth. But why is this second-order representation true of judgment? Moreover, why isn’t this second-order representation also true of other attitudes like supposition? The natural answer would be to say that it is in the nature of judgment itself - but not in the nature of supposition, say - to be such that it is correct (that is, true) to represent it as being accountable to truth. But this answer is not available to Railton: he adheres to CEA and so he thinks there is nothing in the phenomenon of judgment itself that would explain why judgment is subject to the truth-norm (for the sole candidate for this role is thought to be common to other attitudes). The truth-norm, on Railton’s account, is then something that the subject imposes on judgment through some sort of second-order representation, but we have no explanation why this representation is true of judgment. On this score, the deflationary account does not face this explanatory burden: for it recognizes that judgment itself is different from other attitudes and on this ground it might then explain why some norms apply to it but not to other acts.

§5 Truth-Directedness as Aiming

According to Velleman¹² 12 Here I am discussing the view defended in Velleman (2000). But he has then changed his view and followed Shah's normativist account (see §6). , what distinguishes judgment from other acceptance-involving attitudes is the sake for which one is accepting a proposition as true.

What is the sake for which we accept propositions as true when we make judgments?

What does it mean to aim in Velleman's picture?

To accept a proposition as true with the aim of accepting it only if it is really true is the mark of judgment. This aiming is ensured either at the personal level (“the subject’s intentions”) or at the sub-personal (“other mechanisms”), where the two might be thought of as end points of a spectrum of possibilities in which aiming might be realized. This is a form of mixed account about the aim of judgment which endorses a teleological account at both the personal and sub-personal level¹³ 13 A similar mixed account is also defended in Millar (2004, pp. 56ff). .

Given the scope of the present paper, in this paragraph we focus on the view that aiming at truth is ensured only by the presence of some form of conscious intention¹⁴ 14 Bird (2007), McHugh (2012b) Yamada (2010) pursue strategies that purport to explain the truth-directedness of judgment by appealing to some sub-personal mechanisms. But since here we are concerned with judgment - that is, a conscious act - these views can’t explain its truth-directedness, because the posited unconscious feature can’t be constitutive of judgment, whereas truth-directedness is supposed to be a constitutive feature of judgment itself. . The account under discussion here sometimes goes under the name of the teleological account of the aim of judgment¹⁵ 15 There are many authors who endorse such teleological accounts of the aim of belief. See Velleman (2000), Steglich-Petersen (2006), Vahid (2006), Hieronimy (2006) and Whiting (2012). See Owen (2003), Shah (2003) and Kelly (2003) for objections to the teleological account on the ground that aims can be weighted, whereas truth can't. See Steglich-Petersen (2009) for an answer to this objection and McHugh (2012a) for a reply to Steglich-Petersen (2009). .

For the sake of defending the plausibility of a deflationary approach that locates truth-directedness in the phenomenology of judgment, I will here focus on objections that are related to phenomenology. The objections emerge if we ask ourselves, with an eye on experience, what the relationship between acceptance, intention and judgment is exactly supposed to be.

One possible picture is the following. There can be free floating acts of acceptance and it is only when the relevant intention is in place that these acts will become acts like judgment, supposition, and so on. This view is implausible, however, for there are no free floating acts of acceptances that are indiscriminate between judgments and other attitudes. Either I judge that p, or I suppose that p, or..., but it is not the case that I accept that p and wait to form the relevant intention in order to decide whether to judge that p or just to suppose that p, say¹⁶ 16 Remember that Velleman himself recognizes that his notion of acceptance is technical and differs from the one discussed in Bratman (1992). There might be states of acceptances of the sort Bratman describes, but there can’t be free-floating acceptances of the sort Velleman describes. .

Another picture denies the existence of free floating acceptances and countenances the existence of cognitive acts that are formed by the act of acceptance and the relevant accompanying intention. On this view there can’t be a judgment unless I have the right sort of intention while I accept that p. However, as many have noticed¹⁷ 17 See Coté-Bouchard (2016) for further critical discussion of the idea that some intention or desire systematically accompanies our beliefs. , it is not the case that there always are intentions when we form judgments (or other cognitive acts, for that matter). We all know what it means to have intentions. And the sort of intention that I have when I decide to go to the supermarket just does not seem to be the sort of mental state that I have when I form judgments¹⁸ 18 Another way of raising a problem to the teleological view will be to phrase the phenomenological considerations as raising a problem about self-knowledge. If there is nothing distinctive to the phenomenology of judgment itself that distinguishes it from other forms of acceptances, then how do we know when we are judging rather than assuming, say? We can't appeal to the simple fact that consciousness is self-consciousness, for even if by judging one was also conscious of being judging, judging by itself is like supposition in that it is an acceptance of some content as true, and thus they would have to be the same phenomena. We must then think that self-knowledge is here achieved by knowing whether one has the relevant intention. So, I know whether I am judging if I know that I am intending to accept a proposition as true for the particular sake that distinguishes the acceptance of a judgment from the acceptance of a supposition. But this doesn't have any phenomenological plausibility. When I judge I need not know whether I intend to get the truth. This is true regardless of the model of self-knowledge we want to embrace. I do not have to observe the presence of the relevant intention, nor do I have to make up my mind so as to have the relevant intention (see Moran (2000) for this distinction). . Moreover, and crucially, in the experience of judging we can’t distinguish an act of acceptance on the one hand and an intention on the other. We find judgments and suppositions, we do not find pairs of acceptances and intentions.

The two pictures discussed so far agree that for something to be a judgment there must be the relevant intention that turns the acceptance into a judging form of acceptance. A third picture would deny this and would say that there are judgments independently of the relevant intentions, though it would also insist that the intention is necessary in order to make judgment (or its acceptance-component) aiming at truth. Beside the phenomenological problems already mentioned, this view will have the consequence of abandoning the claim that aiming at truth is constitutive of judgment. For a judgment by itself would not be the sort of thing that aims at truth. Simply, it is the sort of thing that is systematically accompanied by an intention that makes it aiming at truth. But this is to abandon the view that aiming at truth is constitutive of judgment.

There is a further puzzling feature in Velleman's account, namely the fact that he seems to be more concerned with our concept of judgment rather than with judgment itself. Here is what he says, commenting on the discovery that our concept of judgment is such that judgment is correct only if true, and incorrect if false.

This passage clearly shows that Velleman does not regard truth as the constitutive aim of judgment. But if this is so, one has failed to provide a phenomenological characterization that elucidates in which sense judgment itself aims at truth.

Ironically, Velleman points to the right sort of evidence that would highlight one important sense in which truth plays a special normative role for judgment - namely that only evidential considerations are able to shape our doxastic deliberations (this feature is called exclusivity). Commenting on the possibility that the attitudes we conceive of as judgment might not be truth-regulated, he says:

But precisely because it is impossible to think of such evidence, we should not make the aim of judgment parasitic on our concepts or conceptions. The truth-directedness of judgment should be built in the nature of judgment itself, but this can be done only if it is built in the phenomenology of judgment itself, for anything that goes beyond the phenomenology (like the fact that we happen to have certain concepts or conceptions) is contingently related to the phenomenon of judgment itself. In order to articulate this objection, it is useful to turn to an explicit conceptualist view about truth-directedness, on which I am now going to focus.

§6 Normativism

Normativists hold that speaking of truth as the aim of judgment is to be taken as a metaphor²⁰ 20 Lynch (2004, p. 499); Wedgwood, (2002, p. 267). . Truth is rather the norm of judgment. This norm is typically taken as a norm of correctness according to which to judge that p is correct (if and) only if p is true. Correctness can then be understood in different ways, as a deontic property, as an evaluative one, as an ideal, or as a sui generis kind of normative property that can't be reduced to more familiar ones²¹ 21 Deontic interpretations: Wedgwood (2002), Boghossian (2003), Shah (2003), Gibbard (2003); evaluative interpretations: Sosa (2007), Lynch (2009, pp. 79-82), Fassio (2011), Jarvis (2012); ideal interpretation: Engel (2013); sui generis interpretations: McHugh (2014). Ferrari (2021) holds a pluralist view according to which truth is normative for belief in a plurality of ways. .

Normativism takes two forms, regardless of the particular content of the norm and the normative vocabulary that is taken to be needed in order to articulate it. Conceptualists²² 22 Boghossian (2003); Engel (2004); Shah (2003), Shah & Velleman (2005). take it that it is a constitutive feature of our concept of judgment that a judgment is correct (if and) only if its content is true. Essentialists²³ 23 Wedgwood (2002), (2007). take it that it is part of the nature of judgment itself that a judgment is correct (if and) only if it is true. Both versions of the view face difficulties that can be solved if we reject the need to move to an inflationary account of truth-directedness. I shall mostly focus on conceptualism, though, for it better serves the purpose of highlighting the merits of a deflationary approach.

According to a conceptualism modeled on Shah (2003SHAH, N. “How Truth Governs Belief”, Philosophical Review, 112: 447-482, 2003.) and Shah & Velleman (2005SHAH, N., Velleman, D. “Doxastic Deliberation”. The Philosophical Review 114, pp. 497-534, 2005. ), all it takes for one's judgment to be aimed at truth is for it to be conceptualized as a judgment by the subject in doxastic deliberation. Crucially, by so conceptualizing one's mental performance one is conceiving of one's attitude as being subject to the truth-norm of correctness. This is what is meant to explain doxastic transparency - the fact that to wonder whether to judge that p is transparent to a wonder as to whether p is true - and exclusivity - the fact that only evidential considerations count as grounds for settling the question of what to believe.

A first difficulty for the view is that when we deliberate we do not seem to bring into the deliberative process the concept of judgment (or the concept of belief). It is rarely the case that I begin a deliberation by asking: what should I believe? Rather, I directly start paying attention to the issue itself. Consider this doxastic deliberation:

This particular doxastic deliberation does not seem to involve - at least not explicitly - the concepts of judgment and belief, nor the concept of truth, for that matter. Yet, it is a paradigmatic case of doxastic deliberation. If these concepts are not involved when we deliberate, how is it that they are supposed to explain the fact that judging aims at truth?²⁴ 24 Shah (2003) claims that one need not have explicitly before one's mind the question 'what should I believe' in order to have one's stream of thoughts framed by the question 'what should I believe?'. But he doesn't explain what it takes to have one's own question so implicitly framed.

Even if we put this problem aside, I think that the view is objectionable for another simple reason. Let us grant to the conceptualist that it is a matter of conceptual necessity that we understand judging as aiming at the truth in such a way that a judgment is correct if and only if it is true²⁵ 25 See Street (2009) for criticisms of Shah's account of the concept of belief. . Let us also concede that it is in virtue of the deployment (however implicit) of that concept in deliberation that our judgments turn out to be responsive to evidential considerations only and more generally to be aimed at truth. However, the fact that we possess this concept of judgment seems to be contingent. And if the possession of this concept - and its deployment in doxastic deliberation - is what guarantees truth-directedness, then truth-directedness turns out to be a contingent feature of judgment itself. Or, which is worse, if truth-directedness is in fact constitutive of judgment itself, then the conceptualist view seems to entail that if we lose our concept of judgment then we lose the capacity to judge. But judgments (and other attitudes) do not seem the sort of things whose existence depends upon the possession of particular concepts.

Consider a person - perhaps a philosopher - who becomes strongly persuaded that whether a judgment is correct can be evaluated only by measuring the way in which a judgment satisfies non-epistemic aims, like moral ones²⁶ 26 See Street (2009) where she discusses the mental life of an agent who takes prudential considerations as reasons for belief. . This is not incredible. She is a philosopher who thinks that action is of a single kind, say, and that mental actions like judgment belong to that single kind. On this ground, she thinks that when deciding what to do one should weigh all possible aims, and eventually she thinks that moral considerations are the ones that end up having authority for us. We might dogmatically suppose that this person has a wrong conception of judgment (for exclusivity holds, say, and only evidential considerations count as grounds for judging) and we might even suppose that as a result she ends up having the wrong concept of judgment. Yet, this person keeps judging all the same.

Now, the conceptualist might want to reply that this person simply has a wrong conception of judgment, while still having the right concept of judgment and while unwittingly and unwillingly applying that concept to her doxastic deliberation. But this line needs argument. It is not absurd to think that by starting to think that judgment is not responsive to evidential considerations only one comes to apply other concepts in one's doxastic deliberation. Yet, the end point of a deliberation, and each intermediate step, will still be an instance of judging that aims at truth as judging does, namely by regarding its content as true in the particular committal way that is characteristic of judgment.

The conceptualist might perhaps concede the point and say that even though it is necessary for a judgment to be subject to the truth norm of correctness according to our actual concept of judgment, it is not necessary to have that concept. As a result, she will grant that were we to have different concepts, the way in which our current cognition is structured would change as well. Perhaps, in a mind animated by a different folk psychology, mental states will display different necessities (or we would have different mental states): to illustrate, it would no longer be the case that only evidential considerations count in doxastic deliberations. If the conceptualist view is open to that possibility, then it fails to capture the basic sense in which judgment aims at truth, namely by being an act that represents its content as being true in the particular committal way that is distinctive of judgment.

From a deflationary standpoint, the conceptualist is turning upside down the order of explanation. A conceptualist wants to explain the truth-directedness of judgment by appealing to our concept of judgment. But this approach does not have a plausible answer to the question why this concept is true of judgment. The conceptualist cannot appeal to the normative nature of judgment itself in order to explain why its concept is true of it, for according to the conceptualist the normative nature of judgment is parasitic on our actual concept of judgment. As a result the conceptualist is bound to accept the view that our concept of judgment is true of judgment because the nature of judgment itself is shaped by our concept of judgment. But this is implausible for we can easily think of a judger who has a different concept of judgment from the one that the conceptualist is describing.

The deflationary view has a more plausible account to offer. Instead of making the nature of judgment parasitic upon our concept of judgment, it says that what counts as the correct concept of judgment depends upon the nature of judgment itself. That is, it is the metaphysics (in this case, the phenomenology) that decides whether our folk psychology is true. The deflationist will have to isolate in the phenomenology of judgment itself some feature that justifies the claim that truth is normative for judgment. Part of this story should of course mention the fact that judgment regards its content as true in the particular committal way that distinguishes judgment from other propositional attitudes. A fuller story shall also include a phenomenological description of other aspects of doxastic deliberation that together explain why it is true that truth plays a special normative role for judgment²⁷ 27 In this paper I cannot provide a full explanation of why truth is normative for judgment. My view is that the phenomenology of questioning is fundamental for explaining why the truth-norm is constitutive for judgment. In short, truth is the standard of correctness of judgment because we answer our questions by forming judgments, and to ask a question is to desire to receive a true answer. The aim is not to be located at the level of intention, nor at the level of some second-order representation, nor or at the level of the concepts we deploy in doxastic deliberation. The aim-talk makes sense if we look at the nature of the act of questioning. For this deflationary phenomenologically grounded account of the truth-norm see §7 below, and Zanetti (2018), (2021), (2023). .

Deflationism has also an advantage over an essentialist inflationary view. The essentialist holds that it is constitutive of judgment itself that the judgment is correct (if and) only if it is true. But the essentialist has the same explanatory burden that the conceptualist is facing. What is it that explains why the truth-norm is constitutive (and distinctive) of judgment? And here it would be natural to explain why the truth-norm applies to judgment by mentioning some features of judgment itself (and perhaps some features that belong more generally to the cognitive acts involved in doxastic deliberation). But this can be provided only by a deflationary account that identifies in the phenomenology of judgment and inquiry some features that will eventually justify the fact that the truth-norm is constitutive for judgment.

§7 The Sketch of a Deflationary Account

When we judge we regard the content judged as true. The Common Element Argument (CEA) says that this feature is common between judgment and other cognitive attitudes. On this ground one is then led to think that we need to posit the existence of some further mental feature in order to explain the particular way in which a judgment regards its content as true. And so one moves to an inflationary view of the truth-directedness of judgment like those we have discussed above.

I have argued that CEA fails to establish its conclusion because it rests on a false premise: namely that there is something in common - a regarding-as-true component - between judgment and other cognitive attitudes. Although we can say that in a supposition that p we regard p as true, this way of regarding p as true is different from the way in which a judgment that p regards p as true. It seems that we are attributing to judgments and suppositions the same property - the fact of regarding their content as true - but this is a linguistic illusion.

Once we have rejected CEA, we can explain why there is a difference between a judgment and a supposition in a very natural way: a judgment regards its content as true in the form of judgment - that is, as really true, as true in being - whereas a supposition regards its content as true in the form of supposition - that is, as hypothetically true. The difference lies in the attitudes themselves, and not in some further mental ingredient (e.g., second-order representations, intentions, conceptual deployment).

This much is not enough, though, to characterize the truth-directedness of judgment. The truth-directedness of judgment hinges on a plurality of truth-related features that seem to be constitutive and distinctive of judgment. One such feature is, as we just noticed, the fact that a judgment regards its content as true in a particular way that is distinctive of judgment. Another important feature that we have discussed in this paper is that truth seems to be a constitutive and distinctive norm of judgment. Railton and conceptualists explain this feature by saying that we represent or conceptualize judgment as being something that is correct only if it is true. Teleologists explain this feature by saying that a judgment is an acceptance done with the aim of getting the truth. How can a deflationist explain this feature?

The deflationist might explain the distinctive way in which truth is normative for judgment by noticing that a judgment regards its content as true in its own distinctive way. A judgment that p is a sort of commitment to p’s being true. As such, it posits truth as its constitutive standard of correctness. This commitment is distinctive of judgment. In a supposition that p there is no commitment to p’s being true, and this is why truth is not its constitutive standard of correctness.

The distinctive normative role that truth plays for judgment can be further highlighted if we explore the interplay between questioning and judgment²⁸ 28 The view sketched in the main text is based on Zanetti (2018), (2021), (2023). .

When we ask a question we want to receive an answer. This is why when we take ourselves to have the answer to our question we stop asking it. The question disappears because we have satisfied the desire that it expresses. But when we ask a question we don’t want to receive any answer whatsoever. We want the correct answer, and the correct answer is the one that is true. In this sense, a question is a form of desire for the truth, for in asking a question we want to receive the true answer to it.

Now, the way in which we answer our questions is by forming judgments. And this is no accident. To answer a question is to take a stance as to how things are. And we take a stance as to how things are by forming a judgment. In other cognitive attitudes we do not take a stance as to how things are. In supposing that p I am not regarding p as true in a way that will be able to count as an answer to the question whether p is true.

In asking a question we are aiming to possess a true answer. In this way, a question posits truth as the standard of correctness for its answer. But we answer our questions by forming judgments. Therefore, our judgments are correct only if they are true. It is in the nature of questioning itself to posit truth as the standard of correctness for judgments, and it is distinctive of judgment to be the sort of cognitive attitude that can answer a question.

This is just a sketch of a deflationary account of truth-directedness. A deflationary account will also have to explain (or explain away) other truth-related features (e.g., the impossibility of judging at will, exclusivity, and the absurdity of having Moore-paradoxical judgments) that seem prima facie constitutive and distinctive of judgment. A fuller account of this sort has however to wait for another occasion.

Acknowledgments

An earlier draft of this paper has been presented at COGITO Research Centre, University of Bologna, and at a meeting of the European PhD Network in Epistemology, Language and Mind, at the University of Bergen. I am grateful to all people in attendance for helpful comments and suggestions. Special thanks are due to Mattie Eklund, Filippo Ferrari, Ole Hjortland, Sebastiano Moruzzi, Tommaso Piazza, Daniele Sgaravatti, Robert Stern, and Giorgio Volpe for their feedback on previous versions of this paper.

References

Publication Dates

History