The regress argument (or the argument of infinite regress) is a philosophical concern held in epistemology. The worry follows from the structure that we use to arrive at justified beliefs, namely that every conclusion requires a premise. In other words, when you ask someone how he/she arrived at a given conclusion you expect them to provide some rationale or evidence supporting it; if he/she tells you that there is absolutely nothing to support that conclusion then you would probably disregard it. This is more succinctly displayed in formal logic, i.e. if p then q, p so q; to merely say q so q would be begging the question, which means that you are presupposing the conclusion in its premises. Here is where the problem appears, if every conclusion requires a premise then every premise must itself require an earlier premise, given that one must conclude that a premise is true in order to use it reliably, so where does it end? It would seem that there is an infinite regress of premises, which ultimately undermines all of them. After all, how can we confidently draw any conclusions from premises that are themselves infinitely in question? This argument therefore threatens the supposition that we are capable of forming reliable beliefs through reasoning.
Responses to the Regress Argument
There are two candidate responses to this argument. One is foundationalism, which proposes that there are some fundamental bedrock beliefs that are self-justifying. These beliefs are typically considered to reside in introspection, for instance I can confidently believe that I perceive myself to be typing on a keyboard without any evidence other than that very perception itself. Foundationalists claim that the regress problem can be solved by supposing that all of our external beliefs are fundamentally resting upon introspective ones.
The second is coherentism, which claims that a belief is justified only if, and as long as, that belief fits into the overall system. Thus, as long as my beliefs all form a coherent belief-system, then it doesn’t matter if there is no bedrock, since every belief supports every other belief. In this view, if I adopt a new belief that is at odds with one that I already hold, then I must discard one of them or risk both of them.
Worries about these replies
It seems highly debatable that all of our beliefs are fundamentally formed from introspective observations, which renders foundationalism useless in legitimizing our beliefs about the external world. After all, when most people turn on a light in a room they don’t think “i see a light switch, thus there is a light switch, thus i can use it to operate the light”, rather it seems more direct than that. Many foundationalists will reply to this criticism by suggesting that there are introspective tacit beliefs in all belief formation; in other words, i merely fail to consciously perceive the full extent of the belief formation process in my mind. However, this criticism is supplemented by the concern that it seems difficult (if even possible) to reach beliefs about the external world from introspective beliefs, i.e. the fact that i see a keyboard doesn’t necessitate the existence of one.
The chief worry about coherentism can be seen as the reverse of that supplied for foundationalism. Where the requirements of forming justified beliefs seem too stringent in foundationalism, they seem too loose in coherentism. Consider the requirements of coherentism for a moment (as long as all of ones beliefs are coherent with each other then they are justified), someone can have utterly absurd beliefs as long as they are coherent with one another. For instance, I can become convinced that I am Napoleon and be justified in believing that as long as all of my other beliefs are distorted in such a way that they are coherent with it; perhaps I believe that it is the early 19th century, that I am in France, that people fear and admire me, despite all of (or at least some of :P) these facts being false, I would be justified in these beliefs according to the coherentist account. This clearly appears to be a wrong result, which gives coherentism the responsibility of specifying stricter requirements in their theory, as well as casting doubt on the validity of any coherentist answer to the infinite regress argument.
As you can see, the argument of infinite regress is a troubling one. Neither of these major responses are without great criticism, nor is the problem itself insignificant. I admit that when I first discovered this one I found it earth shattering (epistemically speaking), so hopefully I won’t be responsible for any one’s mental collapse by writing this :P. Until next time, examine, analyze, and adapt.