structure for 'Reason'    |     alphabetical list of themes    |     unexpand these ideas

2. Reason / F. Fallacies / 2. Infinite Regress

[hopeless explanations which go on forever]

4 ideas
Not everything can be proven, because that would lead to an infinite regress [Aristotle]
     Full Idea: It must be true, quite generally, that not everything can be proven, on pain of an infinite regress.
     From: Aristotle (Metaphysics [c.324 BCE], 1006a09)
     A reaction: Compare Idea 1672, where the possibility of a circular set of mutual proofs is considered. Aristotle seems committed to the present idea.
Infinite regresses have propositions made of propositions etc, with the key term reappearing [Russell]
     Full Idea: In the objectionable kind of infinite regress, some propositions join to constitute the meaning of some proposition, but one of them is similarly compounded, and so ad infinitum. This comes from circular definitions, where the term defined reappears.
     From: Bertrand Russell (The Principles of Mathematics [1903], §329)
Regresses are only vicious in the context of an explanation [McGinn]
     Full Idea: Regresses are only vicious in the context of some explanatory aim, not in themselves.
     From: Colin McGinn (Logical Properties [2000], Ch.2 n11)
     A reaction: A nice point. It is not quite clear how 'pure' reason could ever be vicious, or charming, or sycophantic. The problem about a vicious regress is precisely that it fails to explain anything. Now benign regresses are something else… (see Idea 2523)
Vicious regresses force you to another level; non-vicious imply another level [Baron/Miller]
     Full Idea: A regress is vicious if the problem at level n can only be solved at level n+1; it is non-vicious if it can be solved at n, but the solution forces another level n+1, where the problem can be reformulated.
     From: Baron,S/Miller,K (Intro to the Philosophy of Time [2019], 2.3.2)
     A reaction: So in a vicious regress you chase the apparent solution, but never attain it. In the non-vicious you solve it, but then find you have a new problem. I think.