I'm trying to get a grasp on coherentism and what is proposes is the epistemological justification for knowledge. From what I've taken so far, coherentism relies on what is commonly referred to as "circular reasoning" in that knowledge B could be the justification for knowledge A, and knowledge C could be the justification for knowledge B, and knowledge A could be the justification for knowledge C. However, it shouldn't be referred to negatively as circular reasoning because we are wrong to presuppose that reasoning should be linear.

If I have this right so far, I'm understanding it, in theory at least. However, the question still remains for me. How do we gain that first piece of knowledge? Do we simply start with beliefs instead of knowledge, and those beliefs lead us to form knowledge about new things which in turn justifies our beliefs and makes our first belief knowledge?

Sorry if that sounds like a mouth full, but I'm having a hard time finding exactly what a coherentist would give as justification for the first piece of knowledge that one encounters in their life. If there is no formal response, would they say that pure or impure justification is not possible but that their system can explain the "bigger picture"?

Sorry if that seems like a lot of questions, I was really only trying to ask one, but was just trying to clarify. To sum it up: assuming that my understanding is correct, how does the purported system of justification explain the originating source of knowledge?


If I have this right so far, I'm understanding it, in theory at least. However, the question still remains for me. How do we gain that first piece of knowledge?

It sounds like you're talking about a form of foundationalism, not (pure) coherentism. The idea that there are basic (i.e., foundational) beliefs that ground the less basic beliefs is the core idea of foundationalism (what is justified is the foundation and what follows from it).

Coherentism, by contrast, holds that beliefs might not be justified in isolation. Rather, a system of beliefs that "coheres" will result in the justification of the "cohering" beliefs.

What you seem to be worrying about is what is known as the "regress problem". One way to respond is to deny that it is a problem. Those familiar with some philosophy of mathematics and discussion of circularity in definition will know that not all circles are vicious, so this isn't as hopeless a reply as it might seem. So, you shouldn't assume that to describe reasoning as "circular" is always a negative thing (even though it might be in very many instances).

The other strategy, adopted by Bonjour, is to deny that all reasoning is linear. This is not to embrace circular reasoning, necessarily, but rather to block the assumption that seems to require foundational beliefs to serve as our "ultimate reasons".

You can read much more about the view here at the SEP article.

| improve this answer | |
  • Thank you so much for that helpful response. I think I had my understanding mixed up. So pure coherentism proposes that beliefs may not be justifiable in isolation but that the coherence of all the beliefs in a system justify them, and impure coherentism could be the coherentist's response to the regress problem? Sorry, that's the only thing I'm not quite clear on. – DanL Mar 27 '13 at 5:03
  • @danL that sounds about right. One thing discussed in the linked article is combining coherence with a weak foundationalism as a response to regress. That would seem to deserve the name impure. – Dennis Mar 27 '13 at 6:26
  • Do you have an example of a non-vicious circle in mathematics, or the philosophy thereof? – Mozibur Ullah Mar 27 '13 at 21:31
  • @MoziburUllah If you're not a predicativist, then any impredicative definition you accept will be circular in a sense (circular in quantification as the wiki puts it). See the vicious circle principle. – Dennis Mar 27 '13 at 22:18

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

Not the answer you're looking for? Browse other questions tagged or ask your own question.