36

7 An intuitional construal of taking It might be thought that an intuitional construal of taking will do a lot better than any of the doxastic construals that we have been considering. While I think there is something to this claim, the intuitional proposal ultimately succumbs to some of the same objections we have been developing (and to some other ones as well). By an intuition I mean an ‘intellectual seeming.’ I won’t take a stand here about whether such a state has a distinctive phenomenology. Following Ernest Sosa, we may also think of an intuition as a ‘temptation to believe.’ We experience such an intuition when, for example, it seems to us that there must be more whole numbers than even ones, a seeming that seems to persist long after we have ceased to be disposed to believe that there must be more whole numbers than even ones.6 Now, if we think of the intentional state that connects premises with conclusions along intuitional lines,we can evade the first of the problems listed above. We would evade the problem of having to say how to tell the difference between the enabling background condition and the premise beliefs. An intuition, not being a belief, would not engender such a problem. We would also do better with at least part of the problem about inferential justification. An intuition, like a perception, is not subject to epistemic assessment—it is beyond justification. So we can rest content with saying that an inference is justified provided that its taking intuition is true; we wouldn’t face the problem of saying how the taking state could be justified in some non-inferential way. But even if all of this were so, it seems to me that the intuitional view would still clearly succumb to the final objection that we described for the doxastic view. For even if we switch to thinking of the connecting state as an intuition, it is still impossible to see how my intuition that (FORI) If, It rained last night and If it rained last night, then the streets are wet, then, the streets are wet. could non-circularly explain how I am justified in performing the (Rain) inference. 8 The counterfactual proposal Let us take stock. I have argued that a thinker doesn’t count as reasoning unless, in some appropriate sense, he takes his premises to support his conclusion. And I have looked at all the major ways of understanding this taking in terms of some occurrent intentional state. Of these ways, the most promising, I believe, is the intuitional proposal, though even it faces severe difficulties. The question arises, therefore, whether there is some other way of satisfying the Taking Condition, other than through an occurrent intentional state. I can think of two. The first attempts to account for the nature of inference counterfactually.7 What is it for S to infer (3) from (1) and (2)? According to the Counterfactual Proposal, the answer is: His beliefs in (1) and (2) cause S to believe (3). And, were he to be asked why he believe (3), he would offer (1) and (2) as his reasons for believing (3). And it is the truth of this counterfactual—that S would offer (1) and (2) as his reasons for believing (3)—that makes them his reasons for believing (3). There need be no occurrent intentional state contemporaneous with the 1-2-3 transition that constitutes S’s taking the premises to justify his conclusion. Rather, the difference between a mere transition and an inference consists in the holding of the counterfactual: Were S to be asked (or perhaps were he just to ask himself) why he believes (3), he would cite (1) and (2) as his reasons. I don’t believe that this proposal can be correct. For one thing, it seems to run into the old problem of demanding an unreasonable degree of sophistication from reasoners. But its main difficulty is that it treats the property of being an inference as though it were a responsedependent property, which I believe it cannot be. A property is response-dependent, roughly, if our disposition under certain circumstances to say that something has that property analytically entails that it has the property in question. Hume’s account of beauty is such a theory. According to Hume, what makes it the case that something is beautiful is that a preponderance of suitably ideal judges would judge it to be beautiful. Another favorite candidate for response-dependent treatment is color. There are many difficulties making sense of response-dependent properties in general (for some discussion see Boghossian 1989a , b). But at least such accounts have a certain face plausibility in the cases of color and beauty. However, the property of a person’s thinking something for a reason is not response-dependent. To say that R was S’s reason for A’ing implies that S took R to support his A’ing at the time that he A’ed, and that his so taking it led to his A’ing. I don’t see how S’s being disposed to say that R was his reason for A’ing could make it the case that he took R to support his A’ing and that this taking had a certain causal impact. His saying it might be very good evidence that R was his reason. But saying that R was his reason can’t be constitutive of R’s being his reason. Causation is part of the idea of R’s being his reason—and causation can’t be a responsedependent property. 9 The rule-following proposal: motivations The second main idea for avoiding an occurrent, intentional state account of inference involves deploying the idea of following a rule. The idea is that we can capture the notion of inference by thinking of our thought transitions as guided by inference rules. As we shall see, this view has great appeal. I have myself often assumed it in the past. Of course, I was aware that there are deep issues, raised by Wittgenstein and others, notably Kripke, about how rulefollowing is so much as possible. But I assumed that those problems could be resolved, with the result being a deeper account of rule-following rather than an abandonment of it. John Broome (ms), it turns out, has been thinking independently along very similar lines; here is an interesting statement of the rule-following view of inference from him: I have arrived at necessary and sufficient conditions for a process to be one of reasoning. Reasoning is a process in which you say to yourself the contents of your premise-attitudes, you operate on them by applying a rule to construct a conclusion, which is the content of a new attitude of yours that you acquire in the process. To put it briefly and not quite accurately: reasoning is an operation on contents. This account makes the whole of the process of active reasoning something you do. It is the activity of operating on the contents of your attitudes. For a contrast, compare the jogging model I described…That model limits your activity to saying certain propositions to yourself; you leave the rest of your reasoning to automatic processes. In my account, you are active right through to the conclusion. (p. 235) I would distance myself from Broome’s talk of reasoning as inevitably involving “saying to yourself the contents of your premise-attitudes.” This seems to me to involve an overly restrictive conception of what it is to reason with beliefs. But putting that optional restriction to one side, what could be said in favor of such a rule-guided conception of reasoning? Broome emphasizes the respects in which such a picture can capture the active character of reasoning, the respects in which inference is a mental action. I agree that this is important. But there are at least three further points that we can make in its favor. First, and most important, the rule-following picture seems to be able to effortlessly accommodate the Taking Condition. Consider, for example, the following mundane rule that someone might wish to follow as a matter of policy: (Email Rule) Answer any email that calls for an answer immediately upon receipt! It seems right to say that anyone following this rule would be treating the receipt of an email as a reason for composing an answer to it immediately. S’s acting on the Email Rule explains and rationalizes S’s answering the email immediately upon receiving it. If S did follow the Email Rule in answering the email, it would be natural to say that he took his receipt of the email to be a reason to answer it. (See Boghossian 2008a for further discussion.) So the notion of following a rule incorporates the very idea of taking a circumstance to be a reason for believing something that I have insisted is essential to inference. If we construe inference as a matter of following a rule of inference, then, at a minimum, we will not run afoul of the Taking Condition. Second, the rule-following picture enables us to capture something that seems very central to our inferential abilities, and that is that they are both general and productive. In addition to the (Rain) inference, I am disposed to perform many other inferences that, as we like to say, are of a “similar logical form,” for example: (Space-Time) If x is a Malament-Hogarth space-time, then it has no Cauchy surface. x is a Malament-Hogarth space-time. Therefore, x doesn’t have a Cauchy surface. I may know very little about the concepts ingredient in the (Space–Time) inference. But I will be very confident that, if its premises are true, its conclusion will be true as well. What’s more, it is quite clear that I did not acquire (Rain) and (Space–Time) individually, but rather all together. What explanation could there be for this except that I somehow came to accept a general Modus Ponens rule that now governs my thoughts? Finally, there is another aspect of inference that the rule-following picture seems to get right and that is that, in an inference, we might be hard pressed to say exactly why, or in what respects, we take these premises to justify this conclusion. We take the premises to justify the conclusion and we might have a rough sense of how they do so. But we might be hard pressed to say exactly how they do so. This is another aspect of reasoning that the rule-following picture seems to be able to capture, for we do tend to think that our thoughts can be under the influence of rules, even if we have not explicitly formulated those rules to ourselves, and would be unable to do so with great precision if we tried. These, then, are three further strong considerations in favor of the rule-following picture of inference. 10 The rule-following proposal: the intentional view Unfortunately, though, I believe that there is a difficulty seeing how inference in particular could consist in guidance by rules, a difficulty that I am now inclined to think is at the heart of Wittgenstein’s discussion of following a rule. (The argument of this section is laid out in more detail in my 2008a .) This problem is different from Kripke’s way of reading the moral of Wittgenstein’s discussion (see Kripke 1982 ). Kripke’s treatment was based on what we might call the impossibility of our being able determinately to internalize a particular rule, given that it is supposed to cover a potential infinity of cases. What, asks Kripke’s Wittgenstein, would determinately make it the case that it is Modus Ponens that I am following, as opposed to some bent rule that deviates from Modus Ponens at some point past my computational abilities? That is not the problem I have in mind. Kripke’s problem arises against the backdrop of a naturalistic outlook relative to which it is difficult to see how there could be determinate facts about which infinitary rule I have internalized. But I am willing to waive naturalistic constraints. Let it be that there is as determinate a fact as you might want about which inference rule it is that someone has internalized, and let’s allow such a fact to be a primitive of our ontology. The problem is to understand how such a rule could guide a person’s inferences. To see what the problem is, let us look closely at an especially clear case in which we would say that my thinking involved the active application of a rule. Suppose I receive an email and that I answer it immediately. When would we say that this behavior was a case of following the: (Email Rule) Answer any email that calls for an answer immediately upon receipt! as opposed to just being something that I happened to do that was in conformity with that rule? Clearly, the answer is that it would be correct to say that I was following the Email Rule in replying to the email, rather than just conforming to it, when it is because of the Email Rule that I reply immediately. Of course, the Email rule, qua rule, is just an abstract object. So, if it is to have any hope of explaining my behavior it must be by virtue of the fact that my behavior is to be explained via some state of mine that represents or encodes that rule. We seem to need, then, to postulate the existence of an intentional state that represents the rule. Let’s call this the Intentional View of rule-following.8 Now, how should we conceive of this intentional state as guiding the conduct that constitutes my following the rule it encodes? Well, I have grasped the rule, and so am aware of its requirements. It calls on me to answer any email that I receive immediately. I am aware of having received an email and so recognize that the antecedent of the rule has been satisfied. I know that the rule requires me to answer any email immediately and so conclude that I shall answer this one immediately. On this Intentional construal of rule-following, then, my actively applying a rule can only be understood as a matter of my grasping what the rule requires, forming a view to the effect that its trigger conditions are satisfied, and drawing the conclusion that I must now perform the act required by its consequent. In other words, on the Intentional view of rule-following, rule-following requires inference. Of course, all of this grasping and inferring can take place very quickly. Indeed, the entire procedure can be “automated” to such an extent that it can seem as though no thoughts at all intervene between receiving the email and answering it, just as one can so thoroughly master the route one takes from home to work that one can drive it without giving it a second thought. But just because one can automate an inference doesn’t mean that it wasn’t an inference to begin with. So, now we face a problem. On the one hand, we have the Intentional View of rule-following, according to which applying a rule always involves inference. On the other hand, we have the RuleFollowing picture of inference according to which inference is always a form of rule-following. These two views, however, can’t be true together. Combining the two views would lead us to conclude that following any rule requires embarking upon a vicious infinite regress in which we succeed in following no rule (for further discussion see Boghossian 2008a ). Call this the Inference Problem for Intentional views of rule-following. 11 Rule-following without intentional states: dispositions Is there any way to rescue the rule-following picture of inference from these considerations? We can try one of two tacks. The first and most radical would be to rid our picture of rule-following of any rule-encoding intentional state. The other would be to allow a rule-encoding intentional state into our picture, but to find a way of maintaining that it can exert its control over our thought and behavior in a way that’s non-inferential. Let us begin with the first idea. A number of philosophers have thought that we can explicate what it is for a rule R to guide our thoughts merely in terms of our dispositions to conform to R under appropriate circumstances. For S to follow rule R is for S to be disposed to conform to R under appropriately ideal conditions. If this sort of account could be made to work, then clearly there would be no Inference Problem. Unfortunately, I think that Kripke (see also Fodor 2008 and Boghossian 1989a , 2008a ) was perfectly right to say that any such construal would not be an explication of rule-following, but rather an abandonment of it. There are at least two major reasons. First, our dispositions are finite, whereas the rules we follow are infinitary. Second, the rules we follow are supposed to guide our behavior; but this element would appear to be missing from a Dispositional View. We can see this clearly if we observe that the rules we follow explain why we have the dispositions we have: for example, my following the Modus Ponens rule explains why I am disposed to perform the (Rain) inference. No such thing could be true on a Dispositional View, though, since our dispositions can’t explain themselves. I believe, therefore, that no Dispositional account of rule-following can hope to succeed. However, even without arguing against a dispositional account of rule-following in general, I think we are in a position to see why no such idea can help us with our present task of explaining inference, and specifically in helping us to explain how the Taking Condition might be accommodated. Recall that we are trying to explain what it is for a thinker to take (1) and (2) to justify (3). And our idea is that this taking can be understood as a matter of the thinker applying a rule that he accepts, Modus Ponens, to the contents (1) and (2), deriving from them the conclusion (3) and so coming to believe it. Now, if all we mean by a thinker’s applying the rule MP to the contents (1) and (2) is that the thinker is disposed, when considering such contents, to form the conclusion (3), we have clearly lost any prospect of respecting the Taking Condition through the deployment of the notion of following a rule. This will just look like regular causation of some thoughts by others, without the element of taking that Frege rightly saw to be essential to inference. If I am genuinely applying the rule Modus Ponens, then my applying that rule explains and rationalizes my concluding q from p and ‘if p, then q’. Nothing like that would be true on a merely dispositional account. 12 Applying a rule as sub-personal What about an intermediate picture, according to which, while there is a rule-encoding intentional state involved in following a rule, such a state is not consciously accessible to the thinker and is not something that he consults in figuring out what follows from the contents he believes? To flesh this out a bit more, suppose that what happens is something along these lines. I consider (1) and (2). I do so with the aim of figuring out what follows from these propositions, what proposition they support. A sub-personal mechanism within me ‘recognizes’ the premises to have a certain logical form. This activates some sub-personal state that encodes the MP rule which then puts into place various automatic, sub-personal processes that issue in my believing (3). In “Epistemic Rules,” (Boghossian 2008a ) I argued that going sub-personal in this way would not help with the regress problem: In the present context, going sub-personal presumably means identifying rule-acceptance…not with some person-level state, such as an intention, but with some sub-personal state… Let us say that [such a state] is some sub-personal intentional [i.e., representational] state in which the rule’s requirements are explicitly encoded. Then, once again, it would appear that some inference (now sub-personal) will be required to figure out what the rule calls for under the circumstances. And at this point the regress will recur. (498) Although I believe that this argument is importantly correct, I can imagine an opponent who insists that nothing deserving the name of ‘inference’—and certainly not a process that is characterized by the Taking Condition—would be needed at the sub-personal level. All that would be needed, such an opponent might claim, is a representational state that ensures, via some causal mechanism or other, and ceteris paribus, that a belief in (3) will result from judging (1) and (2). Such a picture deserves consideration. And it may well be that something like it perfectly characterizes what I earlier called, following Kahneman, System 1 reasoning—the sort of reasoning that is automatic, quick, sub-personal and done with little sense of voluntary control. Such a picture would be similar to what Broome calls the ‘jogging model of reasoning,’ according to which the only active thing that you do in reasoning is say certain propositions you believe to yourself and the rest is left up to subpersonal automatic processes. Such a jogging model may well adequately characterize System 1 reasoning. But that is not the sort of reasoning that this paper is about—rather, it is about person-level reasoning, reasoning as a mental action that a person performs, in which he is either aware, or can become aware, of why he is moving from some beliefs to others. No such process of reasoning can be captured by a picture in which (a) reasoning is a matter of following rules with respect to the contents of our attitudes and (b) our following rules with respect to the contents of our attitudes is a matter of automatic, subconscious, sub-personal processes moving us from certain premises to certain conclusions. 13 Why isn’t system 1 reasoning enough? At this point it might be worth considering why we are so sure that there is such person-level reasoning. Or, to put the matter differently, why are we so sure that all reasoning cannot be fully captured by something along the lines of the jogging model? A reasonable first response is simply that it is obvious that there we are able to engage in reasoning in which we are active all the way from premises to conclusion. But I think there is a stronger point in the offing, one that is elusive but important: and that is that our ability to think of ourselves as rational agents depends upon there being the sort of reasoning that I am trying to describe. Suppose some reasoning was such that, having rehearsed the premises, some conclusion simply came to you (accompanied perhaps by the feeling that it is ‘right’), but not accompanied by any awareness of the process leading up to it. Full rationality would require that one ask oneself whether to endorse the conclusion that has simply come to you in this way. And this in turn would require that you lay bare the reasoning process by which the premises are supposed to have led to the conclusion. In other words, full rationality requires that a self-aware process of reasoning, one with no blind spots in it, vet the deliverances of a System 1 process and rule on their correctness. And it is that sort of reasoning that I am trying to capture.14 Where do we stand? Here is where we have arrived. Inference requires Taking. The most obvious way to accommodate Taking would be via an account that postulates an intentional state whose content would concern the relation between the premises of the inference and its conclusion. But this leads nowhere palatable. The rule-following picture of reasoning seems to accommodate Taking naturally enough. But if we try to analyze rule-following, as it seems we ought to be able to do, we run smack into the Inference Problem for following a rule, which makes it impossible to see how inference in general could be a rule-governed phenomenon. Where can we go from here? If these considerations are correct, we face a stark choice between attempting to account for our mental lives without something that looks like the traditional notion of person-level reasoning, on the one hand, and being willing to take the notion of following a rule as an unanalyzable primitive, on the other. Since I have no idea how to dispense with the notion of reasoning, and since there are independent considerations that favor thinking that rule-following cannot be analyzed, I incline firmly towards the latter option. If this is correct, then inference is essentially a matter of following a rule of inference in one’s thought; and we can have no expectation that we will be able to give a non-circular analysis of what following a rule of inference amounts to. 15 Speculative conclusion By way of conclusion let me speculate briefly on what implications such a picture of reasoning would have for some of the topics that have interested philosophers. One consequence is for so-called ‘inferential-role semantics.’ Some philosophers, myself included, have allowed themselves incautiously to say things like this: it is in virtue of our inferring with Modus Ponens that we come to have the concept if. And some of us have gone further to say that it is partly in virtue of Modus Ponens’ having such a concept-constituting role that we are entitled to reason according to it. (This is presented, of course, as part of a broader strategy of explaining conceptpossession, at least in the case of the logical constants, in terms of ‘inferential role,’ and of explaining a priori entitlement in terms of concept possession.) One consequence of emphasizing the Taking Condition on inference is that it draws our attention to something that should have been obvious, but that is often lost sight of, including by me (though not by Broome): and that is that reasoning is an operation on thought contents and not on symbols (that have content). That immediately implies that the usual ways of presenting programs of ‘inferential role semantics’ are confused—a logical constant’s role in inference must be explained by its content; its content cannot be explained by its role in inference. Of course, it is always open to an ‘inferential role’ theorist to give up on the claim that concept possession arises out of the inferential manipulation of symbols, and to insist, rather, that both inference and concepts arise simultaneously out of some pre-cognitive operations on symbols. But it is not easy to see how to flesh out such a view in a plausible way. A second possible lesson of the present discussion concerns the question of the generality of reasons. Particularism, the view that we act out of reasons that are particular rather than general, is increasingly influential in many quarters (see Dancy 2004 for the moral case). The rule-following approach to reasoning would seem to militate against Particularism by claiming that our best prospects for making sense of reasoning sees us as guided by general rules of reasoning. I am not saying that this consideration by itself defeats Particularism in any specific domain, just that it poses a further challenge for it to overcome. Finally, there is a possible implication for Naturalism. The phenomenon of consciousness has, of course, always seemed to pose a deep challenge to naturalistic or physicalist views of the world. But the same has not generally been thought to hold true of cognition. If the present account of reasoning is along the right lines, it opens up the possibility that reasoning poses as much of a challenge to a naturalistic worldview as does consciousness. It makes it difficult to see what naturalistic process inference could consist in.