It is common to invoke the idea--I do it myself--that theism entails that basic human cognitive faculties are not radically unreliable. However, it does not seem to me that theism entails that basic human cognitive faculties are completely accurate, even if used responsibly. Descartes does address this question, but not quite enough I think. The Scholastics also addressed the question Whence error? but it's hard for me to hook up their thoughts to my contemporary epistemology.
One sort of response I'd like to bracket are appeals to the Fall (at least not in any literalistic kind of way). Partly I just don't want to open that can of worms, but also, I'm interested in the issue whether there is any *other* account of allowable error. A bit more set up below the fold.
So one issue is what *are* our basic cognitive faculties. Let's just start with the standard list of the five senses plus memory and rational insight. So even though deduction is truth-preserving and induction is reliable, we might err on what we take to be basic logical principles of reasoning and we might err in our perceptual judgements and our perceptual faculties and memory might just malfunction. So let's start a list of the sources of error (for responsible inquirers, lets bracket error due to irresponsibility for now, for I'm interested in what the sources of error would be for such beings).
1. Error of rational insight. Misjudging the validity or cogency of a form of reasoning or principle of inference.
2. Mis-remembering information so that we have mistaken premises.
3. Malfunctioning memory which gives clear but false testimony and thus mistaken premises or basic beliefs.
3. Mis-judgeing perceptual experiences so that we have mistaken premises.
3. Malfunctioning perceptual faculties which give us clear but false impressions of the world and thus mistaken premises or basic beliefs.
I'm not sure this extends easily to testimony, but I'm more interested in the sources for error that one could have if one were the only one existing, at least to start.
I'm not sure how to analyze mis-judgeing in a way that doesn't collapse into either irresponsibility or dysfunction, but it seems that there is such a category between the two. If there weren't then it rather looks as if the question resolves into the question: just how much unreliability would God build into/allow into our basic cognitive mechanisms. I say not too much, but how reliable is reliable enough?
We should now ask *why* God would allow/build in unreliability at all. I suppose like any adversity it is an opportunity to triumph, to take on a worthy project and do what we can with it in the face of adversity. I think Swinburne says something like this in Providence and the Problem of Evil, but I don't know if its really addressing the present question.
I'm going to go find my Descartes and see again what he says, but in the mean time I'd love to hear ideas.
I'm not quite clear on how you are taking error as opposed to unreliability. The two are very different. An extremely reliable means of knowing can err under rare circumstances; and, indeed, reliability is always reliability under particular kinds of conditions. For instance, one might suggest there's a problem with the fact that we are often misled about what we are seeing when we go walking about with nothing but moonlight to see by; but that would be an odd sort of 'problem', since the obvious solution for the unreliability of vision at night is to take a light. Nor would it be easy to see why anyone should take seriously the idea that there's a problem with God's allowing the visual system to be unreliable under conditions in which it is receiving inadequate stimulus.
Descartes clearly has something like this in mind, I think; one could say his primary concern is to argue that, given God, all our errors are corrigible, and all our unreliabilities can be compensated for. And I think the sort of questions you are raising would require us to be more clear about why, in particular, there would need to be more than this. (There may well need to be more than this; but I don't think any answers can be developed to the questions you are asking without first developing an account of why there needs to be more than this.)
According to Descartes (at least, by my recollection), the question "Why does God allow our cognitive faculties to err?" is a more specific instance of the question "Why does God allow anything bad happen?" In other words, the problem of erring cognitive faculties is a specific instance of the problem of evil. I think Descartes's musings on this issue are in the 4th Meditation.
The method you propose is to go from possible sources of error to an explanation why those sources err. But it might be fruitful to consider paradigm cases of error.
On the one hand are cases like illusion (i.e., the bent oar, the rubber pencil, etc.) In these cases, it seems that the environment "outstrips" (in some sense) our cognitive abilities. Then you have cases involving "pure" cognitive malfunction, like hallucinations. Here, the machinery of cognition (literally) breaks down, due to physiological problems.
I'd be interested to hear what people think about these cases. What explains our erring in cases of illusion or hallucination?
Brandon, as a Latinist I'm using "error" in a broad way to mean missing the mark. The more ones faculties miss the mark, the less reliable they are. I think the questions Why any unreliability (modulo responsibility) at all? and How much is the most God would allow? are interesting questions.
I said from the outset that I the responsibility modulus was in place: "even if used responsibly." I think its pretty clear that our basic cognitive faculties do malfunction even when operating in the invironments (mini- and maxi-) for which they were designed. I said, and Alex seems to agree, that in some way it is just one instance of the Problem of Evil, but that doesn't make it any less interesting as a species.
It's an interesting question whether to include our ability to evaluate modal claims (i.e. our "modal faculty") as basic. Indeed, it's often neglected in discussions of (what Feldman calls) the "Standard View", but at least I'm inclined to consider it on par with other purportedly basic cognitive faculties (i.e. senses, memory, and rational insight). The question of the extent of divinely permissible error with respect to our modal faculties is especially intriguing (e.g. van Inwagen's "modal skepticism" and its implications for standard philosophical theorizing).
On some interpretations of the fall, not having the knowledge of good and evil before the fall would constitute one kind of incompleteness in human understanding of a very important subject matter. I'm not sure if that would fall under error as you're using the term, but it might be a serious limitation. The idea is that people would have had to rely on God for knowledge of moral matters. So it may be that the fall couldn't be the explanation anyway.
Of course, if Brandon is right then this may be just like the case of not bringing a flashlight at night. If there is the means to know, and someone doesn't avail themselves of it, it may not be a problem, even if that means isn't innate.
Jeremy, I think that greater limitation in some pre-Fall state would be explained by God's having a progressive plan of revelation and his close contact. Our problem is that though God surely does have a progressive plan he seems not to give much direct direction and that our cognitive faculties can fail just where they should succeed, even through no fault of our own.
As I said from the outset, I'm inclined to think that this is just part of the Problem of Evil. However, that explains the bare existence of imperfection, it doesn't give us much guidance as to how much dysfunction God would allow. So Let's put a point on it. Suppose I assert the following Divine Reliability Thesis.
(DRTa) If God exists, then the basic cognitive faculties of the vast majority of humans will be very accurate when used responsibly in the correct environment.
I think Plantinga assumes something like DRTa and I'm inclined to believe it myself. It seems intuitive, but I wonder what can be said for and against it.
A similar principle also interests me.
(DRTa) God would not let the reliability of the cognitive faculties of the average person be too close to 50% (or lower).
where the quotient n/m is too close to 50% (roughly) just in case if we knew that a process was accurate only n/m of the time, we would not be rationally justified in believing its output. (Two things: I'm not offering an anaylsis, it's subjunctive and that usually doesn't work, but you get the point; 2. I'm enough of a Bayesian that I'm not committed to the notion of belief as being coherent, so "too close to 50%" might just be equal to 50%).
Daniel, I think some kinds of modal intuitions are very core to rational insight which a sine qua non of the rational life. However (and this will have to remain cryptic for now) I think what we need out of rational insight is knowledge of necessities (which I think will be conceptual) and what PvI objects to for the most part are knowledge of possibilities (indeed he denies their existence).
Trent - I'm enjoying the thread, so I hope this doesn't derail things. What do you mean when you say that PvI "objects to for the most part are knowledge of possibilities (indeed he denies their existence)." I'm mostly concerned with the later. Van Inwagen in his paper "Modal Epistemology" says that he thinks we can and do have knowledge of a great many possibilities that are useful for everyday life, science, and even philosophy. However he does think there is an upper bound on when conceivability is a guide to possibility. His skepticism isn't a thoroughgoing skepticism.
Trent,
I'm still not wholly clear about the issue here. From faculty to faculty the mark (that will be missed in error) will differ, so I'm not sure there is any interesting general question here (which is not to say that there might not be interesting specific versions, of course, and I think there probably are -- but I think the questions need to be domain-specific). Not all our cognitive faculties are concerned with abstract truth (in fact, most of them are clearly not). Some, for instance, are primarily there for action -- the Cartesians, for instance, are big on self-preservation, but even there the cognitive faculty doesn't target self-preservation but simply the information useful for it.
Take, for instance, the visual system. If I look at the horizon and see a bluish water-like patch on the sand, is my visual system really malfunctioning; or is it, instead, functioning properly but, under these conditions, registering a mirage. Clearly if my eyes are designed to see things precisely as they are, they are malfunctioning. But there's excellent reason to think that this is not what they are designed to do; their proximate end is simply to bring in whatever visual stimuli they can, and their remote end is to do so in such a way that my brain can have a heads up about the environment. Which are both happening here, despite the incidental byproduct of the mirage.
Things get even more complicated when we move from the external senses to the 'internal senses' -- memory, for instance. What is the mark that memory can miss? Things will be very different if I am deliberately trying to remember something than if I am just recalling something in response to something I've seen; because when I'm deliberating trying to remember something, I've set up an artificial mark. But in ordinary circumstances where I'm just recalling something automatically, what does it mean for memory to malfunction or go astray? I don't think it's so clear as we tend to think. We have a tendency to dissolve everything into the intellect and will; my eyes miss the mark because they don't see what I want them to see, my memory fails because it doesn't recall what I want it to recall. But some room has to be left for the automaton (as Pascal calls it somewhere): the fact that these things weren't made to do just whatever I want them, so they aren't missing their natural mark or going astray when they fail to meet that standard. So it isn't obviously clear what reliability would consist in; far from it, it is very obscure. And it doesn't seem helpful to ask 'Why unreliability?' if we can't pin down what this unreliability would really be.
For similar reasons, I think the 'modulo responsibility' complicates rather than simplifies things -- it makes less clear, not more clear, what would even count as unreliability or error in our basic cognitive faculties.
I'm pretty sure that, for Descartes, error in judgment doesn't follow from any limitation in our cognitive faculties. It's rather the combination of limited faculties with an extensive freedom to "affirm or deny" propositions or, to make judgments. Error is the result, says RD, of failing to restrain our freedom to judge within the limits of what we know. We could avoid error altogether by making judgment suitably tentative: "I seem to feel cold", that sort of thing.
Given Descartes' analysis of error in terms of freedom, you generate an FWD. Just suppose there is some W in which every creaturely essence is contingently transworld inaccurate. For all worlds, W* that are epistemically ideal for any agent, God could not strongly actualize T* of W* without every agent freely misjudging. And since misjudging seems not to be an intrinsically major evil, the exercise of our freedom to judge might be worth considerable misjudgment.
This will have to be quick, I just picked up a he-uge batch of papers to grade.
Matthew,
I'm referring to the delightfully direct statements which can be found in his Ontology, Identity, and Modality (recently reviewed by Jon in Philosophia Christi ). For example, on p. 148, 1st full paragraph he denigrates the "widespread adherence of philosophers to the nonsensical idea of 'logical possibility'". Then on p. 247, last new paragraph, he avows "But there is no such thing as logical possibility--not, at least, if it is really supposed to be a species of possibility." He continues, "Belief in the reality of 'logical possibility' may be based, at least in part, on a faulty inference from the reality of logical *impossibility*, which is real enough."
Brandon,
I tried to make it clear that I was interested in the faculties when they are aimed at truth. I now explicitly sat that. By assigning a variable to the "marks" as you put it we can generalize over the individual faculties and as long as you admit that there are specific instances all I need to do is to abstract what's common to them to make the general statement. And I don't think my eyes "want" to see anything so I have no idea what's going on in your final paragraph. In general I don't think things are nearly so complicated as you seem to think. On any admissible precisification our basic cognitive faculties are not fully reliable. I'm interested in (DRTa) and (DRTb) (which I now see I accidentally labeled as "a" as well).
Mike,
My reading of RD is like unto yours. I take it that his thesis
(RD1) All error is the result of failing to restrain our freedom to judge within the limits of what we know.
is implausible. At any rate I now assert that it is implausible to see what can be said to defend it.
I tend to think of all I write as being within the parentheses of an "It seems to me that..." operator, but I have some worries about this strategy.
1. Is the cost of limiting all knowledge to seemings to high? That's a pretty wide-spread skepticism.
2. Given Williamson's anti-luminosity arguments, do we really have a sure foundation in such states anyway?
On the plus side, I don't think Williamson's arguments show nearly what he things they show and w.r.t. 1 we could understand a tacit "...or so it seems to me" pragmatically tacked on to all our utterances (supposedly at the end of C.S. Lewis's BBC broadcasts, after taping, Lewis said "or so it seems to me"). B. It would make a nice little pastoral point about philosophy teaching humility (which I'm all for, though I generally don't just try to come to conclusions for homiletical reasons!).
Trent,
Most of this seems right to me. The point about humility (cognitive humility, I'm guessing)seems especially relevant here. Williamson's anti-luminosity stuff occurred to me, but I'm pretty sure a luminous state S (for him) is such that, an agent is in S only he is in a position to know it. That is, we have some "cognitive home" among these states. But I didn't mean to take a position on that view. I meant only that we can say "it seems to me that p" without it ever being true that p or asserting that p. I didn't mean to say, for instance, that 'it seems to me that I am in pain' entails 'I am in pain'. It entails only that it seems to me that I am. I think that's consistent with taking no stand on the luminosity of pains.
Anyway, I'm not entirely sure why you doubt (RD1). I can certainly see why you'd not let (RD1) keep you from making the best judgments you can.
Trent,
I suppose I'll just have to bow out, because I still don't know what you are going for; I don't know how cognitive faculties are 'aimed at truth', if we aren't talking about purely intellectual capacities in particular; and it seems to me that there is no non-controvertible precisification on which our basic cognitive faculties are not fully reliable for that to which they are most reasonably said to be disposed or designed (setting aside, perhaps, issues like disease). Our visual system, for instance, is certainly a cognitive faculty as currently understood in cognitive science, and certainly counts as a basic one, and yet finding a precisification that would make it unreliable under normal conditions and that is not itself dubious seems very difficult; and similarly with memory. And without the right precisification it seems to me impossible to answer questions like "How much (of this sort of) unreliability is allowable?" because this would seem to vary depending on the type.
But I fully concede that I might be missing something, so I'll leave it at that. (I'm not at all sure where you got the part about the eyes 'wanting' anything, though, because I can't find anything in my comment that remotely suggests anything like this.)
Mike, we're on the same page here I think.
My worries about (RD1) could be summed up by saying I have non-trivial credence in the following thises:
(TDH1) We shouldn't have to retreat to a ubiquitous "or so it seems to me" caveat in order to count as sufficiently epistemically humble.
Maybe that's just wrong though.
Brandon,
Either I misread your sentence or mistyped mine. At any rate, here's the target sentence "my eyes miss the mark because they don't see what I want them to see" and here's the target comment. I don't know what *that* means. None of the options I could think of made sense in context. I could have this general desire to see "what's there" but that can't be what you mean, or I could have a desire to see "that thing" but I already see it. The object of my perceptual desire could be "the details" I suppose, but I think that' still too general. But when you make it specific enough at the descriptive level the scenarios in which I would have *that* desire don't seem plausible or helpful. It's just a failure of understanding on my part.
'Want' was perhaps a misleading term; the original point I was trying to make was that it isn't clear what mark is being missed when (for instance) I see a body of water that isn't there; it's not an error to see a mirage, just a byproduct of the particular conditions. Insofar as the visual system is a basic cognitive faculty, it's hard to define error for it, unless we take it to have an object or mark (like, for instance, seeing things exactly as they are) that seems debatable. The only way error and unreliability clearly become relevant is when we are talking about the faculty of judgment (intellect and will, as I called it), or when we are implicitly dragging it in. And, in particular, it wouldn't really be useful to say that, because our visual system doesn't meet the standard our intellectual faculties aim at, that they err or are unreliable, because that's not the mark they're organized to meet. It's not a function they are fit to fulfill.
And this seems quite general; it's difficult to see how any of our basic cognitive faculties are prone to error or unreliability except for whatever cognitive faculty involves intellectual judgment. As the Cartesians noted, most of our cognitive faculties aren't speculative but practical -- your eyes have failed you not when you fail to see things exactly as they are but when you don't have the visual information you need to function in your environment; your recognition memory has failed you not when you don't remember everything completely accurately in precise detail, but when your ability to handle basic survival tasks is hampered from your failure to recognize things with any regularity; and so forth. The object is preservation of the body rather than truth or the way things are. So it's an important question to ask: which of our cognitive faculties actually are aimed at truth? And the answer seems to be that most of them very clearly aren't, at least not without considerable qualification. The faculty of judgment arguably is; but we'd need some account of what's going on when, despite being aimed at truth, and despite operating normally under normal conditions, it errs or is unreliable. If Descartes is right that this is purely due to free will, then, as Mike notes above, we are in FWD territory. If something else is going on, FWD wouldn't work; and we'd need to know what is going on to know what would work.