PH29A Rejecting analysis
The current standing of the standard analysis of knowledge
We have briefly seen a snapshot of the current state of debate abut the analysis of the concept of knowledge. The traditional tripartite analysis was wrecked by Gettier. Since then, various proposals have been made either to patch it up or to replace it with alternative accounts that have emphasized external factors other than justification (causal origin, counterfactual variation, etc.). All of these responses have, however, continued to share certain assumptions:
In reviewing more radical approaches to our problem, I shall take these assumptions in reverse order.
The egocentric predicament
Descartes can take the blame here. The stress on the egocentric predicament derives from the reflective and self-critical nature of the Cartesian programme: the enquirer's need to consider how his or her beliefs are acquired. (One recent on-line discussion of knowledge takes off from Bernard Williams' famous account of the Cartesian project: see the fifth chapter of Papineau's book, Philosophical Naturalism, 1993.) However much people pay lip-service to the socially distributed nature of knowledge and justification, I think it is fair to say that most epistemology has continued to think mostly about my own case.
But knowledge is a we-affair, criticism likewise a collective product. For one thinker who took this seriously we can glance at Popper, for whom "objective knowledge" is an inhabitant of "world 3". World 1 is the physical universe; world 2 the mental, psychological realm. World 3 is "the world of intelligibles, or of ideas in the objective sense; it is the world of possible objects of thoughts; the world of theories in themselves; and their logical relations; of arguments in themselves; and of problem situations in themselves" (Objective Knowledge, 1972, p. 154). We can speak of knowledge in the objective sense as knowledge without a knower, without a knowing subject. Mackie summarises Popper's view as involving
1. world 3 is a human product (unlike Plato's Forms);
2. it is not to be reduced to world 2, it is somehow autonomous;
3. it is to be understood on its own terms, and this will throw light on world 2 — more light than attempts to go in the other direction;
4. worlds 2 and 3 interact;
5. worlds 1 and 3 do not interact directly but only via world 2 (Mackie, Logic and Knowledge, 1985, p. 119).
Mackie's clarifications (primarily his insistence on distinguishing between logical and
historical denizens of world 3) leaves us with a "stress on understanding the logic
of problem situations and on the importance of public, inter-personal, and multi-personal
traditions of knowledge and critical enquiry" (p. 127), and so on public matters
rather than what individuals happen to have in the heads.
There have been other non-first-person approaches to knowledge in more standard
epistemology. E.J. Craig (Knowledge and the State of Nature, 1990) sees
knowledge as "the state our informants need to be in, for us to avoid error, not the
state that we ourselves need to be in" (Papineau).
Craig attempts to show that the human condition is such that we need to be able to
distinguish among possible informants those who know and those who don't; that we need a
concept that makes the kinds of distinctions that know makes with respect to other
persons. Still, Craig's actual working out of his story covers much the same ground
as first-person accounts, and while he acknowledges the need to escape a rigid demand for
analysis in terms of necessary and sufficient conditions (p. 15) he continues to share the
other assumption, that knowledge is somehow a composite, that I mentioned.
Analysis
When we looked at the 'standard analysis' of knowledge we were looking at a typical example of philosophical analysis. A concept X, of some central concern for our thought about the world and our place in it, is analysed into a complex of other concepts, more precisely into a set of individually necessary and jointly sufficient conditions. To be an X is to be an (A + B + C), or something is X if and only if it is A & B & C. Many people have cast doubt on the feasibility of this general enterprise. It may work for a few obviously fabricated concepts like unicorn, but here it may well be the fabrication that makes it work. It doesn't seem to work for some of the standard and allegedly uncontroversial examples, like bachelor (? = unmarried man). Is the Pope a bachelor? Is an unmarried man who lives with the same partner for decades a bachelor? So why think it will work for cause, identity, intention, knowledge? What would be the point of having so many simply dissolvable concepts?
One area where analysis of the traditional type does seem to work is mathematics. It is a fact that many of the philosophers who did most to spread the belief in the importance and the possibility of analysis in the 20th century were stimulated by problems in the foundations of mathematics. It is possible they overlooked the type of gap that separates mathematical thinking from most of the rest of our concerns.
Knowledge as composite
Michael Williams quotes Brandom approvingly as describing knowledge as a kind of "complex, hybrid deontic status" (p. 17): complex because it involves commitment and entitlement; hybrid because it says things both about another's knowledge and our own. Timothy Williamson (Knowledge and its Limits, 2000, p. 258, ftn) refers to the same description as indicating a disagreement with his own far-reaching rejection of analytical assumptions. For Williamson, one major error of the tradition and its post-Gettier inheritors is the idea that knowledge is a complex composed of specifiable other elements: belief plus truth plus....
He acknowledges that knowledge is "factive": if Kap then p. So one could say that the truth of p is a necessary condition for anyone's knowing that p. But he denies that one can go on to assume that if there is one separately specifiable necessary condition for X then there are necessarily other separately specifiable non-trivial necessary conditions for X that are jointly sufficient. For Z to be red it is necessary that Z is coloured; but that gives one no reason to suppose that we can analyse being red as being coloured plus some other non-trivial condition.
Williamson might seem to claim that, while knowing that p is a mental state (a factive one) and believing that p is a non-factive mental state, believing truly that p is not a mental state but a genuine example of a complex hybrid state of affairs, with a mental element and an environmental element (in most cases, not when the p is about something mental itself). I qualify the attribution because he goes on to cast doubt on the tenability of the internal/external contrast here invoked, but he certainly wishes to maintain that believing truly is complex and decomposable in a way he thinks knowing is not.
One interesting suggestion Williamson makes is that one could characterise the traditional analysis of knowledge as getting the order of priority wrong, as it were: it assumes know should be broken down into believe plus other things, when perhaps we should start with know and derive believe etc. from them. (He suggests a rough account of believing p as treating p as if one knew that p, p. 47.) This is one conclusion you might try to extract from the point about how we typically express knowledge or affirm truths - simple knowing is the default, presumptive starting-pont, believing or self-consciously judging that one knows is a more sophisticated activity. A lot of the points about the differences between know and believe and the relations between them may still remain acceptable, but the theoretical perspective in which they are viewed will be different.
Another important reason for Williamson's rejection of the whole tradition to which the standard analysis belongs and his exploration of a third option in which knowledge is taken as unanalysable simply is the fact that both of the responses to Gettier that Williams mentions have led to increasing epicycles: a degenerating research programme in which initially attractive ideas get lost in intricate complexities.
Ed Brandon
Last revised September 24th, 2003
Back to PH29A Homepage.