Imagine you start your framework with some facts.

Now, you then derive some statement. Can you hold the entire chain of inference in your mind simultaneously? If not, then that statement is dependent on the soundness of your memories claiming that each step in the chain was indeed correct.

So, it seems the only solution is to take memory to be infallible as an axiom of the framework. That could easily be refuted, however, and suddenly, the framework contains a contradiction.

It seems we can only ever know facts that are, right then, being held in one's mind simultaneously as the basic truths they come from (in the event of foundationalism), or facts that exist within a loop which is, in its entirety, being held in one's mind (in the event of coherentism).

This severely restricts the framework. How can one deal with this?


I just began reading this SEP article on the matter; hopefully it shows the issue being dealt with.


The linked-to article definitely acknowledges this as a fundamental issue (see the last chapter, The Reliability of Memory). As for solutions though, I found it lacking. I have read through it quite quickly, but my first read-through makes it seem like the different responses are mostly begging the question or allowing the problem.

  • Memory is irrelevant here. Even if we could hold all of that in our mind simultaneously (I am not even sure what that means) we couldn't reliably process it all. So we know that everything in our frameworks is fallible and impermanent. How does it prevent complexity and where is the contradiction? The Earth is limited in resources and impermanent too, it grew plenty complex over those billions of years it managed to stick around nonetheless.
    – Conifold
    Mar 9, 2023 at 5:09
  • @Conifold Consider this idea: a square with a circle in it. If you are not aphantasic, you can likely visualize it in its entirety, thus keeping the constituent ideas in your mind simultaneously. Without visualization, it might be possible too, though I am not so sure; we can non-visually read the ideas out in our head, but thus seemingly not achieve a simultaneous apprehension of them. Now, reasoning is just a sequence of ideas, and unless one can apprehend the entire sequence simultaneously, then memory because a potentially fallible middle-man. [1/2] Mar 9, 2023 at 5:30
  • [2/2] And if memory is unavoidable for sufficiently complex ideas/chains of reasons, and if memory is fallible, then building a framework with any considerable level of complexity becomes impossible; and one only knows the parts of the framework one actually apprehends; the rest of the unapprehended framework is just remembered as true. If one accepts this, without offering any solution, then one is accepting Pyrrhonist skepticism for whenever one isn't apprehending one's framework. Mar 9, 2023 at 5:34
  • 1
    First of all, a gold star for what seems to be an original argument for skepticism. Second, you need to unapack recall fallibility if memory serves that is. 😅 Mar 9, 2023 at 6:22
  • 1
    @user1113719, yes and it looks quite complicated from where I'm at. Mar 9, 2023 at 7:19


You must log in to answer this question.

Browse other questions tagged .