Being a responsible believer requires one to have reasons for one’s beliefs. In fact, it seems that having reasons for one’s beliefs is a requirement for seeing them as beliefs at all. Consider the conflict in thought that arises with assertions like the following:
I believe I live in Nebraska, but I have no idea why I believe that.
I hold firmly that there are jellybeans in that dish, but I have no reason for doing so.
I’m confident that it will not rain on the picnic, but I have no evidence for that.
I support a flat-tax system, but all of my information concerning economic matters is highly unreliable.
Statements like these are conflicted because in each the but-clause seems to retract the grounds for asserting what came before. To affirm, for example, that one lives in Nebraska is often to affirm also that that one has reasons that are sufficient to support that claim. Statements of the kind above, then, don’t look like they could be beliefs at all; they rather something else – perhaps a cognitive symptom, an obsession, a queer dogmatism.
We may say that beliefs are supposed to be not only reason-responsive, but reason-reflective. Our beliefs should be based on our evidence and proportioned to the force of our evidence. And so, when we hold beliefs, we take ourselves to be entitled to reason to and from them. So beliefs must be backed by reasons.
Reason-backing has a curious pattern, however. Each belief must be backed by reasons. But those backing reasons must themselves be backed by still further reasons. And so on. It seems, then, that every belief must be supported by a long chain of supporting reasons.
This is a point familiar to anyone who has spent time with children. Why? is a question that can be (and often is) asked indefinitely. The child’s game of incessantly asking why? may not be particularly serious, but it calls attention to the fact that, for every belief you hold, you ought to be able to say why you hold it.
These rough observations give rise to a deep problem, one that has been at the core of the philosophical sub-discipline of epistemology since its inception.
In contemporary parlance, it is called the regress problem. Briefly stated, the problem is this: Responsibly held beliefs are held on the basis of chains of reasoning. These chains of reasons may come in three forms: circles, finite chains, and infinite chains. Initially, none of these options seems to offer a fully satisfying account of responsible believing. Circular chains of reasons look patently fallacious; they “beg the question.” Finite chains of reasons terminate with beliefs which often look like arbitrary endpoints. If they are arbitrary endpoints for reasoning, then they aren’t really reasons after all. Yet if they are not arbitrary, then they enjoy the support of still further reasons, and thus they aren’t really endpoints. Infinite chains of reasons seem to forever defer the question of whether any belief is held responsibly; hence they defeat human understanding, and so are not viable options for finite creatures like ourselves.
When the ancient skeptic Sextus Empiricus posed this challenge to the project of responsible believing (and thereby, to knowledge, too), he found that it was best, given the obvious structural difficulties raised by the alternatives, to suspend judgment about all things. Notice that something ironic has happened. In order to avoid queer dogmatisms, we have ended up with a skepticism that appears equally queer. Indeed, most philosophers since have vehemently opposed skepticism as not only queer, but positively pernicious. And so the traditional road of philosophy veers back into dogmatism in the face of the regress problem. In more recent philosophy, the three options for chains of reasons have been refurbished. Coherentism is the view that reasons come as systems, and therefore circular chains of reasons can be virtuous. Foundationalism is the view that some reasons are self-verifying or “basic,” and therefore finite chains of reasons can terminate in non-arbitrary endpoints. Infinitism is the view that chains of reasons are in fact infinite, and this does not pose a problem for responsible belief at all. (This view is under-represented, but Aikin's recent book Epistemology and the Regress Problem makes a case.) In addition, some epistemologists propose that the requirement of reason-backing is sometimes suspended, and that certain kinds of beliefs – religious, cultural, or commonsensical beliefs – stand in need of no backing reasons at all.
Centuries of epistemological debate have been fueled by the felt need to resist skepticism at all costs. But we ask: Is skepticism really so awful? Is philosophical dogmatism really preferable? The skeptic is presented in the history of philosophy as a kind of pseudo-intellectual bully who wouldn’t dare to live by his professed doctrine. But this is surely an intentionally unflattering portrait fashioned by anti-skeptics for propagandistic purposes. A more evenhanded estimation has it that the central skeptical aspirations are intellectual integrity and intellectual self-control. Far from being a petty nay-sayer, the skeptic has the cognitive courage to admit when she does not know and when she has nothing more to say. Moreover, the skeptic is especially attuned to the easy counterfeits for intellectual responsibility that all too often pass for intellectual rigor, and is keen to maintain a strict policy of self-criticism. When seen in this light, we find that there is an undeniable kind of Socratic dignity in a skeptical life. Moreover, there is no doubt something shameful in acquiescing in comfortable dogmas, even those dressed up in the garb of philosophical sophistication. To be sure, this noble vision of the dignified skeptic is open to dispute. But we hold that if the going solutions to the regress problem do not pass muster, the skeptical consequences should be wholeheartedly embraced.