Cartesian Demons and Infallibilist Justification

This post is a continuation of a few recent exchanges I’ve had with Steve Hays. My last post was here, his most recent post is here.

Most epistemologists and Christian philosophers don't think that our ability to falsify the Cartesian deceiver should be a condition of knowledge. If we are unable to disprove the hypothetical Cartesian deceiver, that's not a good reason to doubt our beliefs. That doesn't cast reasonable doubt on our beliefs. Indeed, it would be unwarranted to take that thought-experiment too seriously.

I agree with Steve here no matter which contemporary definition of “knowledge” is being used. Whether “knowledge” refers to “true beliefs,” “externally justified beliefs,” or “internally justified beliefs,” there is no need to disprove or falsify a Cartesian demon to “know” something. In the first place, proof or falsification pertains to internalist justification. And even on internalism – whether fallibilist or infallibilist – proof and falsification per se aren’t requirements for knowledge, let alone with reference to something like a Cartesian demon. Not all knowledge is inferentially justified, and not all inferentially justified beliefs need be thought of as the result of deductive reasoning. A Cartesian demon hypothesis needn’t cast doubt on knowledge; it just depends on what one’s theory of knowledge is. I don’t think it’s just my modified epistemology that can avoid this “distinctive problem.” It seems to me garden-variety Scripturalism can avoid it as well.

Where disproving a Cartesian demon would be more relevant is in the realm of apologetics. It’s one thing for me to know there is no Cartesian demon, it’s another to be able to show to someone else how I can rule that out. Do I need to be able to show there is no Cartesian demon to know it? No. But if I can show it, and if I can further make arguments which select for theism in general and Christianity in particular, that’s beneficial. There’s use for that. And in any case, while there are limits to what we can show, this only exhibits the limitations of apologetics, not knowledge.

Scripturalism is a theory of knowledge. More specifically, I’ve argued it ought to be formulated as a theory about a specific kind of knowledge: “any knowledge which is both internally justified and infallible must be founded on divine revelation which, in our case, is coextensive with Scripture” (link). With reference to apologetics, however (which I view as a more pragmatic enterprise), Scripturalists should feel free to utilize all sorts of arguments, not merely those which would constrain all knowledge to refer to internally and infallibly justified beliefs. Frankly, I admit many Scripturalists seem to be a ways off from understanding that.

Given all of this, there is one final question about the Cartesian demon which Steve’s post provokes. Can the Cartesian demon actually be disproved or falsified? Let’s see:

Take internalism. Suppose you have introspective access to your reasons. They seem to be good reasons. But how is that a check against self-delusion? Like LSD, the Cartesian deceiver is persuading you to mistake bad reasons for good reasons. You can't help but find these reasons to be convincing, even though they are deceptive reasons. 

For starters, a hypothesis of “deception” necessitates a distinction between truth and error. There is something about which we are being deceived; that is, we are deceived into erroneous rather than true beliefs. This idea in turn necessitates certain categories of logic and language. What is truth such that we can be said to be deceived with respect to it? This line of thought leads to further interesting questions. “Deception” also necessitates there being at least one thinking entity, and in the case of the Cartesian demon, two. 

Steve at least in principle agrees TAGs are good (link). The initial point, then, is that the Cartesian demon cannot be as omnipotent as many skeptics would frame it. There are some things about which a Cartesian demon can’t deceive us. But then is the Cartesian demon the same as the sort Descartes had in mind? In his outlining of the hypothesis, Steve similarly notes: “[The deceiver] needn't be omniscient or omnipotent. A fallible deceiver could be the source of fallible beliefs, if our beliefs are dependent on that erratic source.”

Okay, but if there are necessary truths which we can recognize as such and show others, then in what sense is this sort of Cartesian demon a problem? We would seem to have internally and infallible justified beliefs after all. Naturally, this would be limited to a subset of our beliefs, and being able to show which of these beliefs qualify may further depend on a TAG or TAGs showing that it would be inconsistent to deny, say, a good, omniscient, self-authenticating communicator, but I think this is possible (link, link, link).

Now, Cartesian skeptics could reply that the demon could be deceiving us as to the necessity of all of these conclusions. But skeptics can say or ask a lot of things. Who cares? Bare assertions or repetitious questioning doesn’t constitute a substantial reply. And obviously, it’s not as if skeptics have privileged, objective, third-person access to what is and isn’t possible. They operate on their own sets of assumptions just as Christians do – only Christians can acknowledge their assumptions and remain what they profess themselves to be. Cartesian skeptics want to get away with climbing the ladder they hope to throw over. They can’t even ask questions without contradicting methodological doubt.

So much for Cartesian demons. Moving on to a different subject, I had written:

Sensations are neither true nor false and so cannot function as premises by which our beliefs are inferentially justified.

Steve replied:

I think that's too crude or overstated. There are different kinds of sensory information. The sound of breakers isn't true or false. But the spoken word (a sentence) can be true or false.    

a) The spoken word is structured sensation that uses sound waves to encode and communicate ideas or propositions.  

b) Likewise, although sensations alone are neither true nor false, sensory input, in combination with ideas, can generate true or false beliefs.  

If I see a red rose, I can rightly infer that I saw a colored object. If every red object is a colored object, then that's a valid deduction.     

Now, it may not be possible to derive the principle that every red object is a colored object from sensory perception or induction. That principle may be intuitive or innate. That must already be in mind for me to draw inferences about the rose. But seeing the rose, in combination with that a priori truism or analytical truth, yields a new and true belief. 

I'm not so sure Steve and I are very far apart here, if at all. The first sentence of b) seems to capture the essence of my position. Our senses were designed to be secondary causes by which we form true beliefs. This causative process isn't arbitrary. It's not as if any old belief would normally be caused by a given sensation. Rather, sensations themselves are the product of interaction with our surrounding environment. That stimulus and our physical, divinely-created processing equipment yield non-arbitrary beliefs. 

My only point was that sensations qua sensations aren’t logical justifiers. We can say spoken sentences are true or false, but I would consider that a kind of metonymy (or some such literary device) which substitutes, for the intended meaning of the physical expression or manifestation, the actual meaning, which isn't physical.

Finally, I had written:

Our senses can cause numerous false beliefs. Sense knowledge is fallible.

To which Steve replied:

True, but the same can be said for reason and memory. Scripturalists need to get down from their high horse and join the rest of us at ground level. They stipulate an inhumane standard of knowledge. Finite creatures can't satisfy those godlike conditions. But why should we?

Not that I don’t appreciate the concern that we keep our heads out of the clouds, but if Steve would admit his demand is only fallibly justified from an internalist perspective, he must also have to admit the demand could itself be unreasonable, right? At the very least, the statement “finite creatures can’t” have beliefs which are internally and infallibly justified is itself, from Steve’s perspective, possibly false. Although that may not have been what Steve meant, I’m not sure what else “godlike conditions” could refer to. And if that is what he meant to refer to, why not continually look to see if that’s [demonstrably] not the case? If, for all we can know, it’s possible, what’s the harm in continually searching? I think the potential windfall exceeds the drawbacks.

Furthermore, I think there’s a relevant disanalogy between the fallibility of sensation and the fallibility of reason and memory. In a recent comment, I wrote:

…if infallibilist, internalist justification is possible in principle, and all our beliefs rely on memory - or even just beliefs relevant to forming infallibly, internally justified beliefs - I would suppose a transcendental argument could be constructed to defend against the idea all memorial beliefs are fallible in respect to justification. This could be a way in which memorial and sensory beliefs are relevantly disanalogous. That's why I'm interested in whether the actual believing of a proposition always, in human cases, requires memory. But this is just a suspicion.

To expand on this, how is it that our reasoning and memory could be, in every case, fallible? I can see how we could have two different sense experiences, or how we could have a different sense experience from someone else, which yield contradictory beliefs and therefore leave us unable to ascertain which of the two beliefs is true.

But I don’t see how this could apply to reason or memory across the board. If any beliefs we have are in some sense memorial insofar as our thoughts either reference memories or themselves occur over a span of time rather than an instant, and if our thoughts in every case depend on our implicitly, if not explicitly, following certain logical structures, then I think there is a path to internally and infallibly justified beliefs which isn’t logically founded on sense knowledge.