| 2085 | Parts and wholes are either equally knowable or equally unknowable [Plato] |
| Full Idea: Either a syllable and its letters are equally knowable and expressible in a rational account, or they are both equally unknowable and inexpressible. | |
| From: Plato (Theaetetus [c.368 BCE], 205e) | |
| A reaction: Presumably you could explain the syllable by the letters, but not vice versa, but he must mean that the explanation is worthless without the letters being explained too. So all explanation is worthless? |
| 2091 | Without distinguishing marks, how do I know what my beliefs are about? [Plato] |
| Full Idea: If I only have beliefs about Theaetetus when I don't know his distinguishing mark, how on earth were my beliefs about you rather than anyone else? | |
| From: Plato (Theaetetus [c.368 BCE], 209b) | |
| A reaction: This is a rather intellectualist approach to mental activity. Presumably Theaetetus has lots of distinguishing marks, but they are not conscious. Must Socrates know everything? |
| 1671 | Sceptics say justification is an infinite regress, or it stops at the unknowable [Aristotle] |
| Full Idea: Sceptics say that there is either an infinite regress of ideas based on one another, or things come to a stop at primitives which are unknowable (because they can't be demonstrated). | |
| From: Aristotle (Posterior Analytics [c.327 BCE], 72b09) | |
| A reaction: This is one strand of what eventually becomes the classic Agrippa's Trilemma (Idea 8850). For Aristotle's view on this one, see Idea 562. |
| 8850 | Agrippa's Trilemma: justification is infinite, or ends arbitrarily, or is circular [Agrippa, by Williams,M] |
| Full Idea: Agrippa's Trilemma offers three possible outcomes for a regress of justification: the chain goes on for ever (infinite); or the chain stops at an unjustified proposition (arbitrary); or the chain eventually includes the original proposition (circular). | |
| From: report of Agrippa (fragments/reports [c.60], §2) by Michael Williams - Without Immediate Justification §2 | |
| A reaction: This summarises Ideas 1911, 1913 and 1914. Agrippa's Trilemma is now a standard starting point for modern discussions of foundations. Personally I reject 2, and am torn between 1 (+ social consensus) and 3 (with a benign, coherent circle). |
| 1816 | Sceptics say demonstration depends on self-demonstrating things, or indemonstrable things [Diog. Laertius] |
| Full Idea: Sceptics say that every demonstration depends on things which demonstrates themselves, or on things which can't be demonstrated. | |
| From: Diogenes Laertius (Lives of Eminent Philosophers [c.250], 9.Py.11) | |
| A reaction: This refers to two parts of Agrippa's Trilemma (the third being that demonstration could go on forever). He makes the first option sound very rationalist, rather than experiential. |
| 8840 | There are five possible responses to the problem of infinite regress in justification [Cleve] |
| Full Idea: Sceptics respond to the regress problem by denying knowledge; Foundationalists accept justifications without reasons; Positists say reasons terminate is mere posits; Coherentists say mutual support is justification; Infinitists accept the regress. | |
| From: James Van Cleve (Why coherence is not enough [2005], I) | |
| A reaction: A nice map of the territory. The doubts of Scepticism are not strong enough for anyone to embrace the view; Foundationalist destroy knowledge (?), as do Positists; Infinitism is a version of Coherentism - which is the winner. |
| 8834 | Infinitism avoids a regress, circularity or arbitrariness, by saying warrant just increases [Klein,P] |
| Full Idea: Infinitism can solve the regress problem, because it endorses a warrant-emergent form of reasoning in which warrant increases as the series of reasons lengthens. The theory can avoid both circularity and arbitrariness. | |
| From: Peter Klein (Infinitism solution to regress problem [2005], 'Step 2') | |
| A reaction: It nicely avoids arbitrariness by offering a reason for absolutely every belief. I think the way to go may to combine individual Infinitism with a social account of where to set the bar of acceptable justification. |
| 2754 | Foundations are justified by non-beliefs, or circularly, or they need no justification [Dancy,J] |
| Full Idea: Foundationalism can get rid of the regress argument with one of three types of belief: those justified by something other than beliefs, those which justify themselves, or those which need no justification. | |
| From: Jonathan Dancy (Intro to Contemporary Epistemology [1985], 4.3) | |
| A reaction: A nice clear trilemma, and none of them will do, which is why foundationalism is false. I vote for Davidson's view, that only a belief can justify another belief. |
| 8851 | Coherentists say that regress problems are assuming 'linear' justification [Williams,M] |
| Full Idea: From the point of view of the coherentist, Agrippa's Dilemma fails because it presupposes a 'linear' conception of justifying inference. | |
| From: Michael Williams (Without Immediate Justification [2005], §2) | |
| A reaction: [He cites Bonjour 1985 for this view] Since a belief may have several justifications, and one belief could justify a host of others, there certainly isn't a simple line of justifications. I agree with the coherentist picture here. |
| 2731 | Justification is either unanchored (infinite or circular), or anchored (in knowledge or non-knowledge) [Audi,R] |
| Full Idea: There are four possible kinds of epistemic chain: infinite and unanchored, circular and unanchored, anchored in a belief which is not knowledge, and anchored in a belief which is bedrock knowledge. | |
| From: Robert Audi (Epistemology: contemporary introduction [1998], VII p.183) | |
| A reaction: About right, though I don't think 'chain' is the right word for what is proposed if justification is to be coherent. The justifications float like lilies in the pond of reason, and a Self (Monet?) seems needed to assess the picture |