Extended Rationality

A Hinge Epistemology

A. Coliva

• Publisher: Palgrave Macmillan
• Year: 2015
• ISBN:
  9781137501899
• Topic: Philosophy
• Subtopic: Epistemology in Philosophy

Extended Rationality provides a novel account of the structure of epistemic justification. Its central claim builds upon Wittgenstein's idea that epistemic justifications hinge on some basic assumptions and that epistemic rationality extends to these very hinges. It exploits these ideas to address problems such as scepticism and relativism.

1

Moderatism about Perceptual Warrants

We do think of having perceptual justifications for our ordinary empirical beliefs. Hence, we think that, when we look at our hand while holding it up in front of us, we do have a justification based on that perceptual experience for a belief such as “Here is a hand”, taken to be about a mind-independent object. When we consider how such a warrant can be had, however, things become incredibly complex and it is difficult to see exactly what is needed to achieve that end.

The aim of this chapter is to present an account of the epistemic structure of our empirical judgments based on the deliverance of our perceptual systems, which I dub moderatism. Accordingly, perceptual experiences can constitute a defeasible warrant for corresponding beliefs about mid-size objects only thanks to some very general assumptions. Chief among them is “There is an external world”. Such an assumption is needed in order to surpass what we may call our cognitive locality. The intuitive idea is that experiences can be subjectively just the same irrespective of whether or not they are brought about by a causal interaction with physical objects. In order to take them to bear onto a universe populated by material objects, it ought to be assumed that there is such a universe with which we are, at least mostly, in contact. Such an assumption, however, does not prevent one from occasionally going wrong, due to possibly unfavorable conditions. Therefore, the perceptual warrant one obtains through perception is defeasible. Yet, the relevant assumption allows subjects to bring their experiences to bear on beliefs about material objects.

As I characterize it in the first instance, moderatism is in fact a family of views, for the details of what it means to assume “There is an external world” can be spelled out in several ways, which would give rise to different types of moderatism – that is, to different species of the same genus. Thus, even though I end up favoring one possible variant, I think the moderate camp extends beyond that. I present some examples of views propounded in the history of philosophy, which belong to it. Equally, I defend one particular conception of the nature and content of perceptual experience and the conditions in which it can play a justificatory role with respect to the corresponding empirical beliefs. Yet again, this will be but one possible conception of it. Hence, although this is the view of perceptual content I favor, I think moderatism is compatible with those different conceptions.

The discussion shows how moderatism fares better than some of its main competitors in the internalist camp.¹ A first, quite intuitive, view, called “liberal” and put forward mainly by Jim Pryor, holds that just by having a certain experience, with a given conscious representational content, that P, absent reasons to doubt that it is being formed in normal conditions, one would thereby acquire an empirical warrant for the corresponding belief. This position is presented and criticized in §1. Its main fault is that it is very difficult to see how it can help us overcome our cognitive locality. For our experiences could subjectively be just the same had they been produced in unfavorable conditions. Yet, absent reasons to doubt that such uncongenial conditions obtain, we would be ipso facto and implausibly justified in believing “Here is my hand” and in disbelieving “I am dreaming of having a hand here” or “I am hallucinating having a hand”.

The liberal view can be seen as occupying one extreme in the range of possible (internalist) theories regarding the structure of empirical warrants. Namely, the one that is the least demanding regarding the conditions that need to be satisfied in order to possess such a warrant. At the other extreme, and still in the internalist camp, lies the, so-called, conservative view of the structure of empirical warrant, largely due to Crispin Wright.² This is a much more demanding theory for it requires that, besides having a certain experience, certain collateral factors are in play, such as further information, which, in order to be epistemically in good standing, must be justified in its turn. The intuitive motivation for the conservative view is that these collateral factors seem to be necessary in order for one’s experience to be brought to bear on the intended class of beliefs – i.e. beliefs about mind-independent, material objects in our surroundings. Arguably, however, the requirements imposed by the conservative view for the attainment of perceptual warrant are too demanding (§2).

Such a discussion paves the way to an alternative, intermediate position which, quite naturally, can be dubbed “moderate”, since it avoids both previous extremes. Accordingly, an empirical perceptual warrant is seen as depending on more than a mere experience, absent reasons to doubt that conditions are normal, but on less than having, besides such an experience, a warrant for some general propositions. In particular, in the moderate view the mere assumption of these general propositions is seen as providing the needed informational setting within which one’s experience can be brought to bear on propositions about material objects. The moderate position is presented in §3.

Although the moderate label is new, the theory is not, or so I will argue. In fact, it can be seen at work, at least implicitly, in a number of theories regarding the structure of empirical warrants that have been proposed throughout the history of philosophy. The reasons behind such widespread acceptance are presented and discussed in §4. To borrow the Wittgensteinian metaphor, the idea is that in order for justifications for ordinary empirical propositions to obtain, some background presuppositions must stay fixed, like hinges that have to stay put for the door to be able to turn.

Finally, this chapter addresses the issue of how we should think of the content of perceptual experience for it to play a justificatory role, together with some collateral assumptions, with respect to our ordinary beliefs about specific empirical objects. When dealing with the content of perception, two issues must be kept apart. The first is establishing the metaphysical nature of perception and hence the conditions in which a creature can be said to perceptually represent its environment. The second, instead, focuses on the epistemic problem of determining the requirements that need to be satisfied for a subject not only to have a given perception, but also to be in a position to avail himself of that perception in the course of obtaining a warrant for his empirical beliefs. There is no obvious reason to think that these two issues should be solved by imposing exactly the same set of conditions on perception. In this vein, it is argued (in §5) that while non- or a-conceptual creatures can have quite coarse-grained or basic perceptions, only creatures endowed with a conceptual repertoire can use their perceptions as part of their justification for specific empirical beliefs.

1 The architecture of perceptual warrants 1: the liberal view

Nothing seems more intuitive than the idea that just by having an experience with the representational content of a red table one would thereby acquire a warrant to believe that there is a red table where one seems to see it. Of course, we all know that sometimes things can be a little tricky: one may be in a room with mirrors and the table may actually be behind one’s back; or the lighting conditions may not be optimal and the table may look red while in fact it is white and bathed in red light. Yet, if there is no reason to think that these vagaries obtain, just by having the relevant experience one would have a warrant to believe that there is a red table in front of one. This is the gist of the, so-called, liberal view regarding the architecture of empirical warrant, put forward and defended in depth by Jim Pryor in several papers in the early 2000s.³

Liberal account of perceptual justification: a belief about specific material objects that P is perceptually justified iff, absent defeaters, one has the appropriate course of experience (typically an experience with representational content that P).⁴

Now, it has to be pointed out that the notion of justification (or warrant) under discussion here is that of an evidential, internalist kind. That is to say, the role of warrant is played by the internal state of a subject – namely his perception with a certain available content – that is to say, a conscious representational content that P, of which the subject is conscious. It is also a state that, for conceptually sophisticated enough creatures, can be articulated: conceptually endowed subjects could actually respond to the question “How do you know that there is a red table in front of you?” by saying “Because I see it”.⁵

Furthermore, here we are concerned with, so-called, propositional warrants. Namely, those warrants there are, in the abstract space of reasons, for a given proposition. Propositional warrants, therefore, exist independently of whether a subject forms the corresponding belief. Suppose a subject is absent-minded. He is in front of a red table and sees it. Yet, he has no time to, or interest in, forming such a belief. Nevertheless, we can actually say that, since he satisfies the condition imposed by the liberal view to have a perceptual warrant, the proposition “There is a red table” is justified, as would be the belief that there is such a table were he to form it on that basis, precisely by virtue of there being such a propositional justification. Indeed, were he to entertain the relevant belief on the basis of his perception, he would be doxastically justified in holding it. For, so-called, doxastic warrants are those propositional warrants which attach (or fail to attach) to one’s actually held beliefs, once the latter are entertained on the basis of the former.

Recent dissatisfaction has been expressed with the ordinary account of doxastic warrants, which sees them as a subspecies of propositional ones.⁶ Let me therefore speak to this general concern first. In my view, the idea of propositional justification is captured nicely by what happens in mathematics, where there are warrants for propositions, namely proofs of the relevant theorems. Furthermore, mathematical proofs are accessible, at least to some human beings who exercise their knowledge and cognitive faculties at their best. This is enough to make sense of the idea that those mathematical propositions can be justified for all of us and, in particular, for those who do not have the means either to grasp those justifications or to entertain the relevant beliefs, namely those containing germane theorems.⁷ Alternatively, we can think of the case presented in the previous paragraph, where there is a propositional warrant for the proposition “There is a red table” if a subject has an appropriate course of experience (taking for granted, for the sake of argument, that the liberal account is correct), even if he actually fails to form the corresponding belief.

Notice that there is a third notion that might be in play but it needs to be carefully distinguished from those of propositional and doxastic warrant. Namely, what might be called a rationally available warrant. Now, suppose our subject is convinced that sometimes he is given a pill that makes him hallucinate red tables; then suppose he actually sees a red table and forms the corresponding belief. His belief would be justified, yet he could not avail himself of that warrant, given his collateral beliefs. We could therefore say that that propositional, and even doxastic, warrant is rationally unavailable to him or, equivalently, that it is “rationally obstructed” from him. It has to be stressed that these collateral beliefs do not actually destroy one’s warrant, because there are no real defeaters in this case. Rather, we may see them as “hypothetically undermining” it.⁸ In sum, warrants are always propositional ones, yet sometimes they attach to propositions that are also the content of subjects’ actual beliefs (formed on the relevant bases), and may or may not be rationally available to them, depending on their further beliefs and cognitive situations at large.

Given these specifications, it should now be clear why the liberal view is so appealing and, indeed, so “liberal”. For, it easily makes sense of the fact that even creatures who are not conceptually sophisticated and are thus unable to articulate their warrants, could nevertheless be justified, propositionally, and even doxastically, were they to form the relevant beliefs. For, to that effect, it would be enough for them, in the normal run of cases, to have the appropriate kind of conscious perceptual experience.

A word of caution is needed though. Several factors make the liberal view different from the well known position put forward by Tyler Burge regarding, what he labels, “perceptual entitlements”.⁹ According to Burge, entitlements are warrants a subject may have even if he is not in a position to articulate them. For instance, by having the relevant kind of perception, absent defeaters, a child who formed the belief that there is a red table in front of him, would have an entitlement for it, even if, ex hypothesi, he could not articulate it (say because he does not have the concept of seeing yet). So much is common between entitlements and Pryor’s liberal conception of perceptual warrant. However, there are several important differences. First, the liberal view is not backed by an externalist story as to why the relevant experiences would provide subjects with a warrant for their beliefs, in the way Burge’s view of entitlements does. In particular, it is not part of the explanation of the nature of these warrants that they satisfy the natural norm that is characteristic of perceptual representations – roughly, to represent the environment around a perceiver correctly, at least with respect to the kind of surroundings within which the perceptual faculty evolved to produce correct representations of it. Furthermore, the role played by conscious experience is much more relevant in Pryor’s account than in Burge’s. While for the former it is a necessary condition for experience to play a justificatory role, for the latter it is not – or, at the very least, it is not clear that it is.¹⁰ Indeed, Pryor goes so far as to maintain that the mere phenomenal content of an experience – the specific way in which things feel to us when we have certain experiences – could be enough for it play a justificatory role with respect to at least some classes of belief, such as “I have a headache”.¹¹ Nothing like this can be found in Burge, who is very clear about the fact that the phenomenal does not amount to the representational and that only representational perceptual content, which (probably) need not even be conscious, could play an entitling role.¹²

The reader may wonder which notion of perceptual warrant is to be preferred. Now, I am skeptical of the existence of natural norms characteristic of perceptual representation. It seems dubious that the notion of representation carries with it a priori, as Burge has it, a commitment to its veridicality. In particular, the notion of representation seems simply to involve the idea of providing a potential layout of the world around a perceiver, which can be correct or incorrect, depending on whether it corresponds, or fails to correspond, to that particular environment. Yet, were it incorrect, and had it always been incorrect, the representation would be a representation nonetheless. Nor does the very notion of representation seem to me to entail that one must be, or have been, at least in the past, in causal relations with the object it allegedly represents. There can be representations of unicorns even if they never existed. Nor does it seem to me to be a natural norm that perceptual representations are factive and therefore entail the correctness of their content. To hold that much seems a conventional aspect, perhaps fostered by the use of English, which does not hold in other languages – such as Italian – or at least not as clearly as in English (but of course, such a use could be stipulated in either language). Now, without recourse to natural norms characteristic of perception, Burge’s position does not seem to be substantially different from forms of reliabilism that have been extensively criticized, at least as theories of justification, and from which he wishes to steer away. Furthermore, I am suspicious of a notion of warrant that is compatible with the possibility that subjects affected by blind-sight – and more generally by perceptual representations they are not conscious...

Table of contents

• Cover
• Title
• Introduction
• 1  Moderatism about Perceptual Warrants
• 2  Further Motivation for Moderatism
• 3  The Bearing of the Moderate View: Transmission Failures, Closure, Easy Knowledge, and Bootstrapping
• 4  The Extended Rationality View
• 5  The Extended Rationality View Extended
• Notes
• Bibliography
• Name Index