Reply to Holmen on Williamson

In his review of Jennifer Hornsby’s “Belief and Reasons for Acting”, Holmen briefly considers Williamson’s example of the burglar who persists in searching for a diamond in a house at great risk of being caught. Holmen takes the burglar example to establish the following claim:

[K]nowledge sometimes must figure in the best explanation for why some agent F-d. According to him [Williamson], attributions of knowledge may be a better predictor for determining someone’s actions by lending more probability to a certain way of conduct.

Roughly, the idea is supposed to be that the burglar’s persistence can only be satisfactorily explained by attributing him with knowledge. Consequently, it is alleged that knowledge attribution allows us to better explain or predict what the burglar would do.

In my previous post, I expressed reservations about whether Williamson’s burglar example establishes the conclusion Holmen claims it does. I consider the burglar’s psychophysical doppelgänger, who displays the same level of psychological conviction and engages in the same exact behaviour, but who nevertheless lacks knowledge because his belief has been rendered false by a highly improbably quantum occurrence. I maintain that so long as we keep his level of psychological conviction the same as that of the original burglar, we are equally able to explain (i.e., make sense of) and predict his behaviour. The upshot is that knowledge attribution is not necessary for a satisfactory explanation of the burglar’s conduct. What knowledge attribution adds (if anything at all) is not explanatory or predictive power, but rather a justificatory dimension—it helps us to understand why the burglar who has knowledge was justified in his conduct.

In his follow-up post, "Knowledge in Explanation: A reply to Avery Archer", Holmen attempts to sidestep my criticism by arguing that the fact that we can explain an agent’s conduct even when she lacks knowledge does not undermine the claim that knowledge is sometimes necessary for explaining or predicting an agent’s conduct. His point seems to be that, at best, my argument shows that something other than knowledge can explain or cause the burglar’s conduct. Even so, I have not shown that knowledge is not the explanation or cause of the agent’s behaviour in the particular example Williamson posits. Holmen summarises what is problematic about my argument as follows:

Avery's case is more like saying that "Shakespeare did not write Macbeth since there is a doppelgänger case where Macbeth gets written but by Marlowe." And even granted the presence of such a case Shakespeare surely wrote Macbeth and is the cause of Macbeth's existence. Likewise, there could be dozens of cases where the burglar acts in the same way but for other reasons than that he knew P. They say nothing whatsoever about whether Kp is the cause in our case for his behaviour.

However, it seems to me that Holmen mischaracterises the dialectical context in which my objection to the burglar example is articulated. He depicts my argument as attempting to establish a positive conclusion when the conclusion it seeks to establish is purely negative. This point may be illustrated by considering the following single-premise argument:

(1) The biological complexity we observe can only be satisfactorily explained by the existence of a Divine designer

(2) Therefore, there is a Divine designer

We may effectively refute (1)-(2) if it could be shown that:

(1*) The biological complexity we observe can be satisfactorily explained by natural selection.

If true, (1*) impugns (1)-(2) by revealing (1) to be false. However, (1*) only establishes the negative conclusion that (1)-(2) fails to establish its conclusion. It is not (nor should it be taken to be) a positive argument showing that (2) is false. For example, the following is clearly a bad argument:

(1*) The biological complexity we observe can be satisfactorily explained by natural selection.

(2*) Therefore, there is no Divine designer.

The lesson is that the same consideration can be effective in a negative argument (i.e., one intended to show that a specific argument fails to establish its conclusion) but ineffective in a positive argument against that same conclusion. It is therefore important to get clear on the aim of an argument (whether it is negative or positive) before one can evaluate its cogency. Since my objection only targets the efficacy of Williamson’s burglar argument it is clearly a negative argument.

Now Holmen’s point seems to be that my claim that beliefs are equally able to explain or predict an agent’s conduct does not entail that knowledge is not a cause. This is of course true if one assumes that I am offering a positive argument in which the conclusion “knowledge is not a cause” is supposed to directly follow from the claim that “things besides knowledge may also be causes”. But to construe my argument along these lines would be to overlook the dialectical fact that I am simply challenging one of the premises in Williamson’s argument (i.e., a purely negative endeavor). To see that this is so, we may summarise Williamson’s burglar example in terms of the following single-premise argument:

(3) The burglar’s conduct can only be satisfactorily explained by attributing him with knowledge.

(4) Therefore, knowledge is sometimes necessary for a satisfactory explanation of an agent’s conduct.

What Williamson’s burglar example alleges to show is that there are certain kinds of conduct (such as the burglar's persistence in the face of countervailing evidence) that simply cannot be satisfactorily explained without attributing knowledge to the relevant agent. (In this regard, his argument is analogous to the creationist who insists that a Divine designer is the only satisfactory explanation of the observed biological complexity.) Construed along these lines, all one needs to do in order to impugn Williamson’s argument is to show that the very same conduct can be satisfactorily explained without attributing knowledge. Since the agent’s conduct is one of the things that is held fixed between the burglar and his doppelgänger, then our ability to explain the doppelgänger’s conduct without attributing knowledge to him shows that knowledge attribution is not necessary to satisfactorily explain or predict the original burglar’s conduct. Thus, we arrive at my counterclaim:

(3*) The burglar’s conduct can be satisfactorily explained by attributing to him a belief with the appropriate level of psychological conviction.

If true, (3*) shows that (3)-(4) fails to establish its conclusion by revealing (3) to be false.

Why-Questions and Motive-Explanations

In this post, I wish to begin constructing a theory of action that builds on the foundation I set up in my previous post. I suggested that the most general purpose of a minimal causal account for some event e2 is to specify which event, from a number of actual or hypothetical alternatives, would feature in a robust causal account of e2, were such an account to be given. I believe this generalises to why-questions of the following form:

(Q) Why did the cat scratch at the door?

One possible because-answers to (Q) is:

(A) Because it wanted to get into the room

The because-answers given in (A) specifies which event would feature in a robust causal account of the action described in (Q). According to (A), it is the cat’s desire to get into the room—rather than, say, a desire to sharpen its claws—that would feature saliently in a robust causal account of the cat scratching at the door.

Unlike the because-answers given in response to the why-question: "Why did the window break?", the because-answer given in response to why-question (Q) implicates a desire - namely, the cat’s desire to get into the room. I hold that to desire X is for X to appear to be good. By ‘good’ I mean that which is beneficial. We may give the concept of the beneficial a Darwinian gloss by defining it in terms of that which promotes biological fitness. To say that a cognitive mechanism represents something as beneficial does not mean that the animal possessing that cognitive mechanism must have a concept of the “beneficial”. Consider, we may say that a cat’s visual system represent something as being true (for example, that there is a mouse over there) without attributing to the cat the concept of truth. Likewise, we may say that a particular cognitive mechanism of an animal aims after biological fitness, even though the animal in question has no such concept and is aware of no such aim. Thus, my claim that to desire something is for that thing to appear to promote biological fitness is meant to be definitional; it applies to all animals that have desires, whether or not they possess the concept of the good, the beneficial or of biological fitness.

In the proceeding discussion I will use the ‘good’ as a shorthand for the beneficial or that which promotes biological fitness. In the light of the above considerations, I take the following conditional to express a necessary condition for desiring X:

(D1) If S desires X, then X appears to be good to S.

This claim should be distinguished from the stronger claim that to desire something is to judge that it is good.

(D2) If S desires X, then S judges that X is good.

The difference between X appearing to be good and judging that X is good may be illustrated with an analogy from theoretical reasoning. To desire X is for X to appear to be good in the same sense that seeing a partially submerged stick as bent is for it to appear to be true that the stick is bent. Often, having an appearance that things are a certain way is enough to prompt one to judge that things are that way. Similarly, having a desire for X is often enough to prompt an “unconditional judgement” in favour of the relevant course of action. However, in the case of agents equipped with the appropriate deliberative capacities, having it appear that something is true may fail to yield the judgement that it is so. I take the same to be true of having an appearance that something is good. Thus, my claim that to desire something is for that thing to appear to be good stands in contrast to the stronger claim that to desire something is to judge that it is good.

Significantly, a single subject may have conflicting sets of appearances. For example, a submerged stick may appear bent when one looks at it, yet feel straight when one touches it. We may say that the stick appears bent, relative to the visual sensory modality, and appears straight, relative to the tactile sensory modality. I believe the same holds with respect to desires. The same thing may appear good relative to the desire for an ice-cream, let us say, but appear bad relative to the desire to remain faithful to one’s diet. Thus, to desire something is always for that thing to appear to be good from a certain perspective.

(D3) If S desires X, then X appears to be good, from a certain perspective

The claim that the cat’s desire to get into the room explains its scratching at the door presupposes that the cat has the belief that scratching at the door would somehow facilitate its getting into the room. Thus, the because-answer offered in response to (ii) takes the form of a belief-desire explanation. Such belief-desire explanations are limited to actions carried out by animals (such as cats, dogs and people) to which we may correctly attribute beliefs and desires. Thus, belief-desire explanations represent a special subset of because-answers. The desire provides us with an agent's motivation for acting and the belief provides us with the explanation of why the particular action was performed. Thus, the agent’s desires and beliefs combine to provide a motive-explanation. Contra Davidson, I believe that motive-explanations (or what he calls "reason-explanations") are equally applicable to rational and pre-rational animals. In this respect, I am in complete agreement with Mary Midgley:

In discussing the central importance of motives, I shall make no special distinction between man and other species, because I think the problem is the same for both. There is nothing anthropomorphic in speaking of the motivation of animals. It is anthropomorphic to call the lion the King of Beasts, but not to talk of him as moved, now by fear, now by curiosity, now by territorial anger. These are not the names of hypothetical inner states, but of major patterns in anyone’s life, the signs of which are regular and visible. Anyone who has to deal with lions learns to read such signs, and survives by doing so. Both with animals and with men, we respond to the feelings and intentions we read in an action, not to the action itself. (Midgley (1978), Beast and Man., pp. 105-6.)

One attractive feature of the above definition of motivation is that it allows for a natural continuity between doxastic and action motivation. On this picture, just as the true is the most abstract characterisation of the aim of belief, so the good is the most abstract characterisation of the aim of desire.

In conclusion, I hold that an action counts as intentional if and only if it possible to give a correct belief-desire or motive-explanation of that action. Hence, any action or event for which a belief-desire explanation cannot be correctly given fails to count as intentional. On this view, because-answers that implicate motive-explanations serve two purposes. First, like because-answers in general, they specify which events would feature in a robust causal account of the event, were such an account to be given. Second, the applicability of motive-explanations determines if an action is intentional or unintentional. On the present account, cats and dogs constitute intentional agents, but not rational ones. (I will have more to say about this later.) Thus, my position differs from that of Davidson since I do not see intentionality as co-extensive with rationality.

Why-Questions and Minimal Causal Accounts

Ordinarily, why-questions are calls for a certain type of explanation—namely, explanations that take the form of a because-answer. For example, if we ask “why did the window break?” then possible explanations (or because-answers) include: “because it was struck by a stone”, “because of an earthquake”, or “because of a strong gust of wind”. The primary aim of a because-answer is to specify the cause (in a minimal sense to be described at present) of an event. According to this minimal account, to say that “e1 caused e2” or that “e2 took place because of e1”, is to say that there is some explanation to be had for e2 in which e1 figures importantly.

A minimal causal account may tell us that e1 explains e2, but it does not tell us how e1 explains e2. Thus, even if one has given an accurate (and perhaps completely satisfactory) because-answer to a why-question, it is still possible to pose a further how-question in response that because-answer. In this way minimal causal accounts stand proxy for more robust causal accounts in which the underlying physical laws and relevant initial conditions are specified. Such robust causal accounts provide an answer to any further how-questions and are the proper domain of the physical sciences.

I take the minimal sense of cause described above to track our quotidian, pre-theoretical use of the word ‘cause’. It should hardly come as a surprise that our quotidian use of the word ‘cause’ (or minimal causal accounts) fails to specify the underlying physical laws and relevant initial conditions that explain how one event causes another. If being able to answer the how-question were necessary for responding to everyday why-questions, then a satisfactory reply to why-questions would be unavailable to the vast majority of people who lack the technical expertise necessary to answer the how-question. To put the matter rather patronisingly, the minimal causal accounts that feature in because-answers implicate the ‘cause’ of the hoi polloi.

But if the purpose of minimal causal accounts is not to explain how one event causes another, then what other purpose could it possible serve? On the present view, the most general purpose of minimal causal accounts is to specify which event, among any number of actual or hypothetical alternatives, would actually feature in a robust causal account, were such an account to be given. For example, a because-answer would specify that it was being struck by a rock, rather than an earth quake or strong gust of wind, which will ultimately feature in a robust causal account of why the window broke. However, the minimal causal account does not say how being struck by a rock—i.e., the relevant physical laws etc.—brings it about that the glass breaks.

(For a detailed discussion of many of the issues raised in this post, see G. F. Shueler (2003), Reasons and Purposes.)

Causal vs. Rational Explanations (Part 3)

In my post Causal vs Rational Explanations (Part1) I suggest that the 'Fanatical Tom' objection seems to pose a problem of causal theorists regarding justification. There are two strategies, available to the causal theorist, for responding to this objection :

STRATEGY1: Accept that the Tom has a justified belief, and tweak the causal requirement to reflect this fact.

STRATEGY2: Reject the assumption that Tom has a justified belief.

I believe the considerations presented in my previous post on this topic, Causal vs. Rational Explanations (Part 2), represent a genuine difficulty for attempts to defend causal theories along the lines of strategy1. More specifically, I regard all attempts to argue that the spyware causally sustains Alvin’s believe via an appeal to the notion of a counterfactual cause are moribund.

I now wish to propose an alternative reply, along the lines of STRATEGY2, for arguing that Tom’s belief is not justified. We begin with a few distinctions. First, we must differentiate between the justification of a proposition in the abstract, whether or not anyone holds it, and the justification of a particular subject’s, S’s, belief of a specific proposition p. The first refers to justification of a certain type, what we may call impersonal justification, and the second is justification of a token belief, or personal justification. Impersonal justification does not require any particular believer of p, nor does it presuppose that anyone holds the belief or has expressed it. Thus, what impersonally justifies a belief need not itself be believed by a certain subject. The distinction between impersonal and personal justification is that between general justifiability of the belief that p and S’s being justified in believing that p on some given occasion.
Second, there are several kinds of reasons one may have for holding a proposition as true. Here are four:

(R1) Objective reason is a reason there is for holding a proposition as true. This is a reason whose status as a reason is independent of any subject being aware of it or of its epistemic status. For example, the fact that the thermometer says that temperature of Bob’s son is above normal’ is a reason (in this sense) to believe that ‘Bob’s son has a fever’, whether or not anyone has bothered to consult the thermometer.

(R2) Person-relative objective reason is a reason there is for S, the subject in question, to believe that p. For example, the fact that the thermometer reads 200 is a reason for Bob, but not for Bob’s two year old son who is too young to understand thermometers—to believe that he has a fever. Again, Bob need not actually be aware of this reason.

(R3) Subjective reason is a reason the subject actually possesses for believing that p. Bob possesses a reason to believe his son has a fever when he takes in the fact that the thermometer says his son’s temperature is above normal.

(R4) Operative reason is the reason on the basis of which S believes that p. For example, if Bob came to believe his son had a fever because of his burning forehead, and only then bothered to consult the thermometer, then it would be his son’s burning forehead rather than his thermometer that constitutes his grounding reason.

The first two kinds of reasons provide a subject with propositional justification, or makes some proposition p justifiable. Some proposition or piece of evidence need not be known by anyone in particular in order to constitute a reason in the sense of (R1) and (R2). Although (R1) and (R2) fall short of having a reason, they are both prerequisites for having a reason. (R3) introduces the requirement that the reason be subjectively available and represent the minimal sense in which one may have a reason. (R4) takes this idea one step further by stipulating that the subject base his belief on the reason in question. Signficantly, (R4) captures what is known as doxastic justification, and I believe it is Tom’s failure to fulfil (R4) that accounts for his belief being unjustified.

As the thermometer example illustrates, one and the same reason can instantiate all of (R1)-(R4). Moreover, (R1)-(R4) seems to represent a plausible requirement for a reason to constitute a justifying reason for believing that p. Tom clearly has a reason for believing SPY in the sense of (R1)-(R3). Given the existence of the spyware, there is a normative reason for Tom to believe SPY. The spyware is also a reason for Tom, since he recognises that the spyware constitutes conclusive evidence for the proposition, ‘the NSA is spying on US citizens’. Third, Tom possesses the spyware evidence in the sense that he has it at his disposal.

However, Tom does not have a reason in the sense of (R4). Tom’s belief is not based on the spyware evidence. In short, Tom’s belief is not doxastically justified. This point can also be put in terms of the distinction between impersonal and personal justification described earlier. Tom’s belief that SPY is impersonally justified or justifiable. However, at present we are not merely concerned with whether his belief is justifiable. Rather, we want to ascertain if Tom’s belief is actually justified—to wit, whether Tom is justified in holding his belief. In short, we want to know if Tom is personally justified. Given the doxastically inappropriate basing of his belief, the answer seems to be no.

Causal vs. Rational Explanations (Part 2)

In my previous post on this topic, I briefly touched on a common reply to Lehrer’s Gypsy-Lawyer type cases—namely, the suggestion that a belief need not have its ‘causal origin’ in a reliable process, so long as it is ‘causally sustained’ by such a process. On this more moderate view, Tom’s belief can be seen as ‘causally sustained’ by the discovery of the spyware in the sense that if Tom were to give up his fanatical religious views, he would still continue to believe that ‘the NSA is spying on US citizens’ (henceforth ‘SPY’).

In this post I will examine one way of unpacking the concept of what it means for a belief to be causally sustained by some bit of evidence (whether propositionally or non-propositionally understood)—namely, Marshall Swain’s notion of a counterfactual cause. We may define a counterfactual in terms of possible world semantics. Following Lewis let us say that the truth of the statement A > B consists in the fact that, among possible worlds where A holds, there is at least one world where B holds and which is more similar to our world than any world where A holds but B does not. (For the sake of simplicity, I will just speak in terms of the closest possible world.)

Roughly, the idea is that a belief, B, is based on a reason, R, if R is either the non-deviant cause or counterfactual cause of B. R counterfactually causes B if in the closest possible world where the actual cause, C, is missing, R causes B. Thus, causal theorists maintain that the underlying intuition behind the causal theory can be preserved by insisting that if the justifying evidence does not actually cause belief, it must counterfactually cause belief. Applied to the ‘Fanatical Tom’ counterexample, we may say that in closest possible world in which Tom’s belief that SPY is not caused by his fanatical religious views, it is caused by the spyware evidence. Thus, the spyware evidence is the counterfactual cause of Tom’s belief.

We may add one important restriction to the idea of a counterfactual cause. First, it is crucially important that we limit ourselves the closest possible world. Otherwise, the notion of a counterfactual cause would become much too broad; allowing beliefs to be based on reasons when it is obvious that they are not so based in the actual world. (Presumably, it is such considerations that prompt Swain to limit the application of the concept of a counterfactual cause to the reason next in line to the reason that actually causes the belief.) Thus, Swain’s view may be put as follows:

S1: Subject S bases her belief that p on R IFF S’s having R is either (i) a cause of S’s believing that p in the actual world, or (ii) S’s having R is a counterfactual cause, such that R causes S’s believing that p in the closest possible world in which the actual cause, C, is absent.

If we accept the restriction on what constitutes a counterfactual cause imposed by S1, we seem able to adjust the ‘Fanatical Tom’ counterexample so that the spyware is no longer the counterfactual cause of Tom’s belief. For example, we could imagine that Tom is not only a religious fanatic, but that he is also a member and ardent believer in a webring devoted to conspiracy theories. Thus, holding all other parts of the ‘Fanatical Tom’ counterexample constant, we may stipulate that in addition to his fanatical religious views, he has also read about the NSA’s surveillance activities on the internet, before actually discovering the spyware evidence in Fred’s home. On this revised version of ‘Fanatical Tom’ case the closest possible world in which Tom’s belief is not based on his religious views, it would actually be based on his reading the conspiracy theory webring. Thus, according to S1, the spyware no longer constitutes a counterfactual cause for Tom’s believing that SPY. Since in the re-described case, Tom’s belief seems no less justified than it is the original case, the introduction of counterfactual causes fails to resolve the present difficulty.

But perhaps Swain is being unnecessarily conservative when he restricts the notion of a counterfactual cause to the closest possible world. Perhaps it would be sufficient to restrict ourselves to nearby worlds in which Tom both finds the spyware and believes SPY. On the present proposal, we might be able to rescue Swain’s theory by widening the scope of what counts as counterfactual cause to all worlds sufficiently close to the actual world. Thus modified, Swain’s counterfactual cause may be summarised thus:

S2: Subject S bases her belief that p on R IFF S’s having R is either (i) a cause of S’s believing that p in the actual world, or (ii) for some set, C, of possible counterfactual causes of S’s belief that p, R is a member of C.

On S2, R need not be the next in line to the actual cause, C, such that in the closest possible world in which C is not the actual cause R causes S’s belief that p. It is sufficient that R be the cause of S’s belief that p in some nearby world. As with most talk of nearby worlds, there is an obvious difficulty regarding how one delineates which worlds count as nearby and which do not. Or, to put the problem differently, how do we determine (in a non-arbitrary manner) which reasons, out of the maximal set of possible reasons, ought to qualify as counterfactual reasons.

One plausible criterion would be to restrict ourselves to reasons that are occurrent to the subject in the actual world. Thus, we may add to S2 the clause that for some possible reason, R, to be included in the set, C, of possible counterfactual causes, it must be occurrent to S in the actual world. Assuming that this provides a plausible and sufficient restriction on which worlds ought to count as nearby, we seem to have a working definition of a counterfactual cause. S2 provides us with a reply to the ‘Fanatical Tom’ example since it allows both the conspiracy webring and the spyware (along with any other potential reasons occurrent to Tom in the actual world) to be considered a counterfactual cause of his belief.

Unfortunately, this still is not sufficient for rescuing the notion of a counterfactual cause. An even more decisive objection to the Swainesque reply to the ‘Fanatical Tom’ problem is that it allows the spyware to be the counterfactual cause of Tom’s belief even when it seems obvious that it is not justified. For example, suppose that Tom’s religion teaches him to reject all evidential relations and that only beliefs based upon ‘faith’ are justified. However, Tom realises that his neighbour, Fred, still buys into ‘the whole evidence thing’ and appeals to the spyware solely for the purpose of converting Fred. The idea is that Tom does not see the spyware as genuine evidence himself. In such a case, Tom’s belief would clearly be unjustified. However, in such a scenario, the spyware would still be the counterfactual cause of Tom’s belief by Swain’s lights. In the closest possible world in which Tom no longer has his religious beliefs, and in which he still believed SPY, his belief would be based on the spyware. This is because in that world, the religious belief would no longer prevent Tom from recognising the spyware as a reason for believing SPY. Thus, Swain’s notion of a counterfactual cause allows the spyware to causally sustain a belief even when it is clear that the belief is not justified.

I believe the above analysis suggests that all attempts to cash out the idea of a belief being causally sustained by evidence in terms of counterfactual causation is moribund. In my next post on this topic I will present what I believe is a successful reply to Lehrer Gypsy-Lawyer type scenarios.

McDowell’s Direct Realist Reply to Scepticism

McDowell’s claim that in veridical cases of perception objects are “immediately present to the mind” has earned him the title of direct realist. Direct realism is the thesis that what we perceive are not merely appearances (insert: sense data, veil or ideas or whatever else suits your fancy) but rather objects themselves.

Direct realism can be used against arguments for external world scepticism (EWS) of of the following form:

P1: Our EW beliefs are based on sense data beliefs
P2: Sense data beliefs cannot justify EW beliefs
EWS: Our EW beliefs are not justified

The direct realist resists EWS by rejecting P1. There are two senses in which the direct realist can maintain that our EW beliefs are not based on sense data beliefs. First, they can deny that there are such things as sense data beliefs. This is certainly true in McDowell’s case since his rejection of the highest common factor view, a fortiori, includes the eschewal of any such mediating beliefs. Second, if we interpret the locution ‘based on’ in a strong inferential sense, then a direct realist may insist that our EW beliefs are not inferred from any other beliefs at all but are rather immediately grasped by the perceiving subject.

However, the locution ‘based on’ need not be taken in this strong inferential sense. A sceptic may insist that there can be a weak sense of ‘based on’ that is non-inferential. For example, we may interpret the phrase ‘based on’ in causal terms. Putatively, it seems undeniable that our EW beliefs are causally based on appearances or sense impressions. According to an empirical-psychological account of vision, the act of seeing an object involves several complex physiological processes. Moreover, we never see objects themselves, but merely the light bundles reflected off an object. Thus, there are several causal intermediaries between our EW beliefs and the objects in the eternal world. Thus, our sceptical argument may be reformulated as follows:

P1*: Our EW beliefs are (causally) based on sensory appearances
P2*: Sensory appearances cannot justify EW beliefs
EWS: Our EW beliefs are not justified

Significantly, P1* does not assert that sensory appearances are the intentional object of our perceptual experience. For instance, by P1*’s lights we are free to view sensory appearances as the means rather than object of our perceptions. In other words, P1* merely asserts that sensory appearances form part of the causal chain linking our beliefs to the EW that gives rise to our beliefs. Thus, P1* can be construed along purely direct realist lines—to wit, we may hold P1* while simultaneously maintaining that objects themselves (and not some mental intermediary) are the intentional objects of our perceptions. Thus, even if we grant direct realism’s denial of an inferential intermediary, P1* still stands.

Unfortunately for the sceptic, this reply fails in the case of McDowellian direct realism. McDowell may gladly concede that the causal descriptions (with all its various intermediary steps) of empirical psychology are true. However, McDowell maintains that the empirical-psychological account is not only one available. In fact, McDowell draws on the Sellarsian idea of there being a sui generis “space of reasons” for precisely this reason. When McDowell describes perception as being direct, he is speaking in terms of the logical space of reasons, not in terms of the logical space in which the subject matter of empirical science falls. Things like photons or electrical impulses are  not of the right sort to act as a reason for one's beliefs. Trading on this ambiguity in the locution ‘based on’, McDowell may accept P1*, but point out that P1* cannot plausibly be combined with P2* in order to generate EWS.  P1* (explicitly) exploits the epistemically insignificant notion of being ‘causally’ based on, while P2* (implicitly) exploits the  epistemically significant notion of being ‘justificationally’ based on.  This equivocation in the explicit notion of "based on" in P1* and the implicit notion of "based on" in P2* prevents the inference from going through.

In sum, McDowellian direct realism is effective against sceptical arguments of a certain form. Specifically, sceptical arguments that ride on there being an inferential or causal intermediary between our EW beliefs and the objects we perceive are easily defeated by direct realist considerations. Of course,  arguments for EWS may take other forms as well.

Causal vs. Rational Explanations (Part 1)

According to many J-externalists, a belief is only justified if it is caused in the right sort of way (for example, by a reliable process). Thus, we arrive at (J-Ext[C]):

(J-Ext[C]) For any agent S, S’s belief that p is justified IFF it is caused in the appropriate way by the fact that p.

In this post I will like to argue against this claim. I begin by distinguishing between two types of explanations of why someone might hold a certain belief—namely, causal and rational. Consider the following example:

The ‘Crazy Jack’ Counterexample:
Jack is convinced that alien creatures from a distant galaxy regularly abduct and perform bizarre experiments on him. We ask Jack’s psychiatrist why he has this belief, and the psychiatrist responds ‘because he has sush-and-such physical defect in his brain.’ There are two ways of interpreting the psychiatrist’s answer. First, we may view it as giving a causal explanation of Jack’s belief. On this interpretation, Jack need not be aware that he suffers from any such mental defect for it to count as the “reason” for his having the belief he has. It is simply a fact about jack that his belief is so caused. Alternatively, we could interpret the psychiatrist as giving Jack’s own reason for believing that he has been subjected to alien experimentation. Perhaps Jack has learned that he has a brain defect, and he takes it to be evidence that aliens have been tampering with his body. In that case, the psychiatrist’s explanation would actually be Jack’s reason for believing what he does.

Most J-externalists view justification along the lines of the first explanation. More specifically, J-externalists typically maintain that a belief is only justified if it is causally connected, in the right sort of way, to the state of affairs that makes it true. By contrast, the J-internalist insists that it is reasons, not causes, which are necessary for justification. More specifically, the J-internalist rejects the idea that the causal aetiology of a belief determines its epistemic standing. By her lights, this would be to confuse reasons and causes. Admittedly, there is some ambiguity in the term 'reasoning' itself. On the one hand, it may indicate a logical argument that, if valid, supports the truth of a belief. On the other hand, it may refer to the process of drawing a conclusion from premises. The J-internalist maintains that it is not the process but the validity that is epistemically relevant. We reason from our evaluation system to defend what we accept, and this has a causal explanation. Nonetheless, the validity of the reasoning is not to be explained causally but in terms of the inferential relation between the premises and what is concluded.

As the “crazy Jack” example illustrates, one and the same thing—namely, the physical defect in Jack’s brain—might be both the cause and evidential basis of Jack’s belief. Thus, the J-externalist points out that the mere fact that something is a causal explanation of why you believe what you do, does not entail that it cannot also be the reason for your belief. For example, in Alvin Goldman's causal account of knowledge, while it is ventured that “inference is a causal process”, all that seems to be meant by this is that, “if a chain of inferences is ‘added’ to a causal chain, then the entire chain is causal”. The J-externalist need not deny that validity cannot be explained causally. What he may claim is that when a belief is produced appropriately, the causal process by which it is produced just is the relation of evidential support. In that case, the J-internalist’s task is to show that the relation of evidential support is really distinct from whatever causal process produces a belief.

One way this may be done is by describing a case where a subject’s reasons justify her belief, even though they have no causal influence¹ on the formation and acceptance of the belief. The following example seems to fit the bill:

The ‘Fanatical Tom’ Counterexample:
Tom is a member of a religious cult that teaches that the director of the National Security Agency, Keith Alexander, is the anti-Christ. Based on this teaching, Tom comes to believe that the NSA is engaged in the illegal warrantless surveillance of US citizens. Let us assume for the sake of argument that Tom’s beliefs lack any credible evidence. However, it just so happens that the NSA really is illegally spying on US citizens. Although Tom’s belief is true, we would still deny that he has knowledge. Similarly, we would regard his belief as unjustified. In short, Tom has a true belief, but not a justified one. Now let us suppose further that Tom is determined to convert his neighbour, Fred, who is very sceptical about Tom’s claims. Tom sets out to find evidence that would convince Fred by taking apart every electronic appliance in Fred’s home. In the process, Tom discovers a mini camera in Fred’s smoke detector and a small microphone in his telephone receiver, both with NSA serial numbers. Given this newly acquired evidence we would now hold that Tom’s belief is justified and may even constitute knowledge. However, since Tom believed that the NSA was engaged in illegal surveillance even before he discovered the spyware, his belief was clearly not caused by this evidence. To wit, if Tom hadn’t discovered the spyware, he still would have the belief in question. Now the question that confronts us is this: is the questionable causal aetiology of Tom’s belief sufficient to denying that it is justified once he has acquired the new evidence? Intuitively, the answer seems to be no. We would regard Tom’s belief as justified despite its untoward causal origins. Therefore, it seems that it is the evidence one has for a belief, rather than its causal aetiology, that determines whether or not it is justified.

Perhaps the unpropitious consequences of the ‘fanatical Tom’ counterexample might be avoided by weakening the J-externalist causal requirement. For example, Goldman and others² have suggested that a belief need not have its ‘causal origin’ in a reliable process, but only that it be ‘causally sustained’ by such a process. On this more moderate view, Tom’s belief can be seen as ‘causally sustained’ by the discovery of the spyware in the sense that if Tom were to give up his fanatical religious views, he would still continue to believe that the NSA is spying on US citizens. While it may be true that beliefs are generally ‘causally sustained’ by their evidence in this way, this need not be the case. For example, the ‘fanatical Tom’ counterexample can be adjusted so that Tom’s belief would not be ‘causally sustained’ by the spyware evidence. We may suppose that if Tom were to lose his religious conviction this would cause his entire world-view to collapse, to the extant that he would lose all his other beliefs. However, this fact does not seem to affect our intuition that Tom’s belief is justified. Thus, even the weaker requirement that a belief be ‘causally sustained’ by a reliable process does not appear to be a necessary condition for justification.³

Endnotes:

1 This argumentative strategy can be broadened to apply to other nomonological J-externalist accounts by substituting “causal aetiology” with “history of the belief” [see: Lehrer 1990, ch. 8]. The "Fanatical Tom" counterexample represents what is known as a Gypsy-Lawyer type scenario.

2 Gilbert Harmon [1970, 1973, 1986] and Marshall Swain [1981] have argued that a belief is only based on evidence if the evidence conditionally or partially explains the belief.

3 For an argument that no revision of Swain’s account can succeed, see Kvanvig 1985.

References:
Goldman, A., (1967) "A Causal Theory of Knowing," Journal of Philosophy 64 (widely reprinted).

Harman, G., (1970) ‘Knowledge, Reasons, and Causes’, Journal of Philosophy 67: 844-55

Harman, G. (1973) Thought. Princeton, NJ: Princeton University Press.

Harman, G. (1986) Change in View. Cambridge, MA: MIT Press.

Kvanvig, J. (1985) ‘Swain on the Basing Relation’, Analysis 45: 153-58.

Swain, M. (1981) Reason and Knowledge. Ithaca, NY: Cornell University Press.

Looking for a Causal Criterion for Vision

Below, I attempt to derive (what may be described as) the standard causal criterion for vision [see (3)*** below] from a critical examination of a non-causal definition of vision. By so doing I hope to make explicit some of the motivating intuitions that lie behind the causal account. The non-causal definition of vision with which I will like to begin may be schematically expressed as follows:

S sees an object O before him only if

(1) S has an experience E such that it looks just as if O is before him
(2) There is an object O before S

However, this definition hardly seems sufficient for saying that S sees O before him. For example, we may construct something akin to a Gettier-type counterexample to (1) and (2). (I say that the following example is merely ‘akin’ to a Gettier case because unlike the genuine Gettier counter-example mine is not concerned with the question of justification.) Imagine S has his eyes closed and is hallucinating that there is an object O before him. Suppose further that there actually is an object O before him. In such a case, the requirements of both (1) and (2) have been met, and yet we would not wish to say that S sees O before him since S would continue to have the experience E even if O was not present. This observation has motivated a causal account of perception that stipulates that S only counts as seeing O if E is caused by O. Thus we arrive at:

(3) S’s experience E is caused by O.

As it stands, the stipulation that E is caused by O does not seem sufficient since we can imagine instances when such causal dependence holds but in which we would still not count S as seeing O. Grice presents us with two such scenarios [Grice 1961, 141-5]. The first case is where S has his eyes closed and another person, P, describes the object O to S, which then prompts S to have a hallucination of O. In such a case, S’s experience E is caused by O, albeit indirectly via P’s description. Since E is ultimately causally dependent on O, the demands of (3) have been met. Yet, we would still be unwilling to say that in such a scenario S sees O. Thus, as it stands, the addition of (3) is still not sufficient for defining a case of vision.

One possible reply to Grice’s first example would be to point out that it is highly unlikely that P’s description of O could give rise to a hallucination in S that perfectly matched O. Thus, if we added a clause to (3) requiring that E perfectly match O we would (it is supposed) exclude cases of hallucinations based on the description of an object. Thus we arrive at something like:

(3)* S’s experience E is caused by O and E perfectly matches O.

At least three points can be made with regards to the above reply to Grice’s first example. First, it is not clear that it is impossible for P to describe O in enough detail for S to have a hallucination of O that fulfilled the requirement of (3)*. Admittedly, this may be difficult to do depending on how complex O happens to be, but it certainly isn’t (strictly speaking) impossible. In fact, it may even turn out that the perfect match between E and O may simply be a matter of luck, this would still be sufficient to fulfill the requirement of (3)*, so long as E is (indirectly) caused by O. Moreover, we could imagine a case in which O is a simple geometric shape such as a triangle or square. In such a case it wouldn’t seem all that difficult for P to describe O in such a way that S’s hallucination of O perfectly matched O.

Second, (3)* does not seem like a plausible necessary condition for vision since there are many cases in which we may be willing to say that S sees O although E does not perfectly match O. For example, let us suppose that S sees a white shaggy dog, which he mistakes for a sheep. In such a case, S’s experience E may not perfectly match O and yet it is not clear that we would wish to deny that S sees O. Other examples include misreading the letters (or words) on a street sign and certain cases of blurred vision. Moreover, there is at least one reading of ‘perfectly matches’ in which it wouldn’t make sense to say any experience E perfectly matches (or in any way matches) O. On such a reading experiences may be taken to be a very different kind of thing from the objects that give rise to them. While I’m not particularly keen on pushing for such an interpretation of experience vis-à-vis objects, it does seem to be a viable position one may hold.

Our third and final objection to (3)* comes from Grice’s second counterexample. Grice imagines a scientist who looks at an object O and stimulates S’s brain in exactly the way necessary for S to have an experience E that matches O. If instead of a scientist we substituted a sophisticated computer, then the possibility of E perfectly matching O perfectly (at least as perfect as we could reasonably expect in a normal case of vision) becomes quite plausible. Given this fact, along with the fact that E remains causally dependent on O, the demands of (3)* have once again been met. Nevertheless, we would still be unwilling to say that in such a scenario S sees O. Thus, the addition of (3)* still leaves us with an insufficient causal criterion for vision.

Another way one may object to Grice’s examples is by pointing out that in both the causal dependence of E on O is indirect. If we stipulate that the causal dependence must be direct we arrive at:

(3)** S’s experience E is caused by O directly.

The word ‘directly’, as used in (3)** suggests that there cannot be any mediating factors between O and E. However, this does not seem like a plausible requirement for vision since there are cases where O may cause E indirectly where we would nevertheless still wish to insist that S sees O. For example, imagine that S is looking at bacteria on a slide with the aid of a powerful microscope. In such a case, E’s causal dependence on O is mediated (i.e., by the microscope) and yet we may still wish to insist that S sees O.

Although the requirement that O directly cause E seems problematic, it does hint at a valuable intuition: namely, that there are a number of possible causal scenarios linking E to O that would be unacceptable as far as our putative understanding of vision is concerned. According to this line of reasoning our concept of vision includes the notion that there should be a certain type of causal relationship between perceptual experiences and their objects. Thus, for S to see O, his perceptual experiences must be caused in the right sort of way, where this is understood as excluding all causal chains that deviate from ‘the norm’. What constitutes ‘the norm’ can be arrived at easily enough by examining unproblematic instances (i.e., paradigm cases) where the locution ‘S sees O’ correctly applies. Examples, such as Grice’s discussed above, represent causal chains that deviate from the norm and therefore fail to be causally relevant in the right way. If we add the requirement excluding deviant causal chains to (3) we arrive at:

(3)*** S’s experience E is caused by O in the right sort of way.

Admittedly, what the expression ‘right sort of way’ entails needs to be spelt on in greater detail. Nevertheless, (3)*** is regarded by many to reflect the canonical statement of the causal requirement of vision and is taken, along with (1) and (2) to be sufficient for establishing that S sees O.

The causal criterion we have derived represents merely the first step, rather than the end, of the debate regarding whether the causal description of vision is the best available. For example, many believe disjunctivism represents an attractive alternative to the causal picture, though it remains unclear whether the two approaches are necessarily incompatible. Moreover, while (1), (2) and (3)*** are here taken to be sufficient for vision, another question worth considering is whether they also represent a necessary condition.