# The Regress Argument

The regress argument is one of the oldest and most straightforward arguments in favour of scepticism. It proceeds as follows. Imagine that I claim to know something, anything, which for simplicity we will call “X”. Imagine now that someone were to ask me “how do you know that X is true?” If knowledge is justified true belief, then in order for me to know something, I must not only believe it, but also have some justification for my belief. Suppose then I provide some justification, in the form “I know X because Y”, where Y is something else that I claim to know. My partner in conversation could then ask me “okay, but how do you know Y?” Again, if my belief that Y is true is to be counted as knowledge, it must be justified, and hence I must be able to say something like “I know Y because Z”. This pattern can be repeated indefinitely, thus leading to what is called an infinite regress. Whatever reasons I give, I can always be asked “but how do you know that”, and then I must provide yet further reasons.

In such a circumstance, there seem to be three possible outcomes, all of which are problematic:

1. The questioning could go on forever, with me providing an infinite number of justifications for why I know each new belief. This seems hard to accept, since it means that we never ‘get to the bottom’ of how we can know something – there is always one more round of “but how do you know that”? Furthermore, it seems that in real life we are never able to actually give an infinite number of justifications. Our brains can only hold so much knowledge, and so eventually it seems we must run of out reasons to give.
2. The justifications could ‘bottom out’ in some belief that has no justification. Eventually we reach some belief, call it “A”, which when called upon to justify (“how do you know that A is true?”), we respond by saying something like “I don’t have a justification, I just know”. A variation of this would be for A to be self-justifying, such that we respond “I know that A is true because A”. Many philosophers think that neither of these outcomes quite works. Surely beliefs need to be justified in order to count as knowledge, and nor does it seem that they can simply justify themselves – how can we know that the justification is correct if the justification is itself the very thing we are trying to justify our belief in? Isn’t that a case of the proverbial impossibility of ‘pulling onself up by one’s own bootstraps’?
3. The justification could become circular, such that we say something like “I know that X because Y, and I know that Y because Z, and I know that Z because X”. Of course, the circle could be much bigger than this simple example – it may take a hundred iterations of the question because we ended up back to the same belief we started with. Regardless of how long it takes, however, this outcome seems no better than the previous two. How can a set of beliefs somehow justify themselves in this way? How do we know that any of it is justified? Is this just not another, more complicated form of ‘self-bootstrapping’?

If these are the only three possibilities for justifying our beliefs – an infinite regress with no end, a terminal unjustified or self-justifying belief, or a chain of beliefs which justify each other – then, so the argument goes, it seems that ultimately there are no satisfactory answers to the question of how we can justify our beliefs. Taking this reasoning to the extreme, one could conclude that no beliefs can be justified, and hence become a global skeptic.

Very few philosophers take this extreme option. Rather, most think that the dilemma can be answered or avoided, though there is little agreement as to the best way of doing so. Nevertheless, the regress argument is still very important, because how one responds to the dilemma will shape what sort of beliefs one is likely to form, and affects the sort of reasons (or lack thereof) one gives for justifying these beliefs. As such, much work in epistemology focuses on how to best respond to this dilemma.