Lecture Supplement of Alston’s The Reliability of Sense Perception Chapter 4
Copyright © 2009 Bruce W. Hauptli
Chapter 4. Empirical Arguments for the Reliability of SP
i. The Explanation of Sensory Experience:
60 Alston notes that as we are seeking non-circular arguments, as we have seen, simple “track record” empirical arguments for the reliability or SP will not work. “Indeed,” he notes, “at first glance it seems impossible to mount any a posteriori argument for the reliability of SP without relying on its outputs to do so. For where else would we get the empirical premises we need for such an argument? On pp. 62-63 he cites and discusses Locke’s argument which is circular.
“There is at least one other such of knowledge, however that deserves to be called ‘empirical’, namely introspection, our awareness of our own conscious states. If we could find some way to reason from facts about our sensory experience to external facts putatively perceived via sense experiences, without the conclusion depending on an appeal to our perception of any external facts, that might enable us to give a cogent noncircular argument for the reliability of SP.”
61 Since “introspective facts” are a posteriori, but not attained through sense perception, appeal to them would not appear to raise circularity concerns!
Here we must keep in mind that throughout the chapter and all of its moves, if we are to be true to the core hope that this sort of “empirical argument” is to represent, the phenomena we are going to talk about have to be our introspective awareness of our conscious states, and the awareness here is going to have to be an introspective awareness of the regularity of our phenomenal experience which we ordinarily call “sensory.”
But how, we must ask ourselves, are the results of introspection going to deliver truths about the world which can be appealed to as one tries to legitimate sense perceptual practice? Alston will consider contemporary attempts to mount what are often called hypothetico-deductive arguments or arguments to the best explanation:
61 “when high-level scientific theories are established…it is not by inductive correlations [of sensory experiences to directly and independently observed things]….Consider a theory of chemical compounds in terms of constituent molecular structure. We cannot establish by induction that whenever we have something with the surface properties of salt, the molecular structure is thus-and-so; for apart from this or some other equally high level theory, we have no way of knowing what kind of molecular structure we have there. We have no cognitive access to molecular structure that is independent of a theory….But this difficulty by no means leads scientists to throw up their hands in despair….Rather than seeking some inductive support for the theory, they try to show that it constitutes the best explanation of the empirical facts in question, in this case, inductively established lower-level lawlike generalizations.”
63 Such arguments will not work if they are not general—they can not consider “bits of sense experience.”
64 But the simplest presumptions are too weak: “a minimal explanatory gesture in the direction of an external cause would be that some (relatively) permanent and stable entity or entities is causally responsible for our sensory experience. But even if that were true, it would fail to show that these entities are as they perceptually seem to us. That explanation is compatible with the cause being God, a demon, Kantian things-in-themselves (with the ban on causal efficacy lifted), or Leibnizian or Whitehedian centers of psychic force.”
65 If we enrich the explanation so that it holds that our experience is caused pretty much in the way we believe, we must ask whether it is strong enough to support reliability (without begging the question).
65-69 Consider color attributions: secondary qualities—are the perceptual experiences “true” here? Normally we maintain that there are regularities behind the experiences, and, so, the attributions are “true” in an indirect way. In the case of primary qualities, however, the connection with truth is supposed to be direct. The “explanation” will need to be elaborated so that we don’t simply get reliable sensory attributions but reliable beliefs about things in the world as long as our phenomenal beliefs are arrived at under normal circumstances.
He will call the best explanatory story we could come up with the “standard explanation”—it will be our “normal” view that that our phenomenal [sensory] experiences are reliably caused by a physical world about which we gain true beliefs reliably through such experiences. This view will be predicated on several assumptions:
67 “our perceptual belief forming mechanisms are themselves sufficiently and approximately sensitive to differences in their experiential input”—that is, we get the same sorts on “belief outputs” from modestly different phenomenal inputs, and
“we are usually in normal situations in which noting is throwing off the tendency of the total apparatus to produce mostly true beliefs….
Critical Comment: in what he says up to this point, there appears to be several factors he appends to the discussion which are illegitimate. If he is trying to rely upon an “a posteriori” argument which relies only on introspective inputs, then all that he builds into it would seem to have to arise from an individual’s first-personal introspective awareness of that individual’s own phenomenal experience. That is, the third personal discussion throughout the whole chapter about “our experience” would appear illegitimate. Of course, this is not, really, his problem, as he is going to reject this sort of argument anyway. Nonetheless, we should keep this in mind as we examine the next chapter. On p. 101 Alston clearly recognizes this, but this factor seems to me to lessen seriously the attractiveness of all the views he criticizes in this chapter.
ii. The Explanation of Patterns in Sense Experience:
70 Alston notes that this “standard explanation” can not be attained in the manner of normal scientific theory development because there aren’t, really, any phenomenal law-like regularities which we can generalize from. Unlike chemical theory, we can’t build upon regularities in the evidential base to legitimate theoretical accounts of how these come about (regular experiments on particular things in the world which lead us to “salts,” “acids,” to a theory of the chemical bond. That is, if one truly limits oneself to introspection, is there regularity (even if we add in memory)?
71 If we are go get “regularities” we will need to add a theory about the movements, capabilities, etc., of the perceiving subject! And he will allow these to be added so that we get some “regularities” that then can be “explained” by the “standard explanation.” This will provide the basis for the “a posteriori argument which is going to claim that the “standard explanation” is better than any of the “alternatives,” and, thus, that the beliefs it produces are likely to be true. He will ultimately reject this argument because:
97 “at bottom the reason that so many acute philosophers have failed to do the job [of establishing that the “standard explanation” is the “best”] is the ban on epistemic circularity. When we are precluded from making use of anything we take ourselves to have learned from SP and whatever is based on that, we have little to go on in deciding which explanation of experience is more likely to be true. The various nonevidential criteria of explanations may be useful in certain choices between scientific theories, but it the present case they fail to pick a winner.”
74 Thus suppose that we do an excellent job of enriching the “standard explanation” by adding an account of the experiencer. Imagine, further, that we have filled in all the holes. In short, we have a “theory” which accounts for the phenomena of the regularity of our phenomenal introspective experiences. It holds that they are as they are because they are caused by a regular world in such a manner that the beliefs produced are reliably true. Here, then, we have an “explanatory” argument which claims that the pattern of phenomenal experience is explained by the existence and workings of a regular physical world which reliably produces true introspective phenomenal beliefs in us. This will be the “best explanation,” however, only if it is better than the other explanations—and the pattern of argument we are examining is that of that of “inference to the best explanation. The “other explanatory accounts” he will discuss (none of which will be developed even as “fully” as the “standard account are:
that the regularity we find in our introspective phenomenal world is the result of the work of a “Cartesian deceiver,”
that the regularity is the result of a Berkeleyan deity (a non-physical power which causes regular phenomenal introspective experience in us which lead us, if we are not careful, to presume there is a physical world,
the “phenomenalists’ view” which holds that all we can legitimately claim is that there are regular introspective phenomenal experiences,
75 that a Leibnizian or Whiteheadian view might be right—according to both there is no such thing as (completely) dead matter, and there is an overall regularity which results from a conscious origin of this universal consciousness. Every existing entity engages in unified perception (Leibniz) or feelings (Whitehead), though this rises to a level of self-awareness only in such high-grade existents as human beings,
that the phenomenal introspective regularity is simply generated by me myself.
For the next 20 pages Alston will argue, first generally, and then by discussion of a number of specific authors, that the only way to establish the superiority of any one of the explanations is by introducing epistemic circularity: [97] “…when we are precluded from making use of anything .we take ourselves to have learned from SP and whatever is based on that, we have little to go on in deciding which explanation of experience is more likely to be true.”
iii. Attempts to Support the Standard Explanation:
78 “Contemporary discussion of this problem typically assume that all these competitors are empirically equivalent in that they all explain the same range of experiential data. The task of choosing a winner then boils down to determining which candidate best satisfies other criteria for the comparative evaluation of explanations. Those most commonly cited are economy and simplicity. But it is not clear that the alternatives all score worse in these respects.”
78-79 Discussions of economy (number of entities or “basic” entities”) and simplicity (nature of the connections between explained phenomena and phenomena doing the explanation) are very complex, and it isn’t clear who “wins” here.
79 We may, following Alan Goldman, also want to add talk about explanatory depth in hopes that this provides us with a better rationale for a choice. Here we would discuss the reach of the explanatory ties (how much they explain, and how far their explanations can stretch beyond what we are currently endeavoring to explain):
80-81 “…while the standard theory is enormously rich and complex, the alternatives, as thus far developed, consist merely in bare suggestions as to how an explanation might be developed in detail. Thus, in comparing actually existing theories, the standard one is miles ahead on richness and detail….But what bearing does it have on the question of which explanation is the true one? Obviously, the reason the other alternatives have not been significantly developed is that we all accept the standard explanation and are quite satisfied with it. No one is motivated to carry out the enormous labor that would be involved in developing one of the alternatives to anything like the same extent. But to build this superiority of the standard explanation into an argument for its truth, we would have to be satisfied in supposing that (a) none of the other alternative could be developed into something equally rich and detailed, and (b) of the two explanations the richer and more detailed is more likely to be true, ceteris paribus.”
82-85 Suppose we take the “standard explanation” and add it to the Cartesian deceiver hypothesis—the deceiver works in the manner of the ordinary model! Now it is true that this is not at “simple” as the normal hypothesis, and it is not as “original as it is, but it is not clear that it loses the competition for “best explanation,” so “…we have no reason worthy of the name for supposing that the alternative explanations could not be worked out with a richness of detail equal to that of the standard explanation.”
85-87 Alston considers an argument by Laurence Bonjour against the “demon hypothesis:” that its a priori probability is very low, and, thus, we should prefer the “ordinary hypothesis[1]:
The probability of the world being as it is given the existence of such a “demon” is not very high (why should the demon construct the world in this fashion?). Using an analogy to another human being’s desires and wishes, on p. 87 Alston says “…given that you are capable of having indefinitely many sets of desires and purposes, the antecedent likelihood of your having this set is infinitesimally small, and the hypothesis collapses before leaving the starting gate.”
87-88 Alston claims the same is true for the standard explanation, and it is unclear that any such consideration can favor the one sort of explanation over the other. In fact, he maintains, 88 “…it would be child’s play to build something into the demon hypothesis as well that would explain why the demon prefers to produce sense experiences in the way we have them, rather than in some other way and rather than not producing any at all. Any appearance of a superiority of the “correspondence hypothesis” [or standard explanation] on this point stems from surreptitiously supposing ourselves to have independent grounds for accepting the evolutionary account [standard explanation].”
89-92 Alston next considers an argument by Michael Slote which tries to distinguish the standard and demon explanations by talking about “explanations which are “inquiry-limiting:”
an hypothesis is inquiry-limiting…just in case if S at t accepts that hypothesis and holds it to be the best and completest explanation of those phenomena available at t….he ensures the impossibility of his coming to have rationally justified or warranted belief in more and more true explanations of various aspects of or facts about the phenomena in question….[2]
Slote argues for a principle of unlimited inquiry (PUI) which is to stop us from accepting such explanations [89] “…on the grounds that science is essentially an attempt to explain as much as possible; and therefore any hypothesis that would imply that no further explanation is forthcoming will, by that fact, receive a black mark.” He also claims that the “demon explanation violates PUI!