Scots Philosophical AssociationUniversity of St. Andrews

Can Reliabilitists Believe in Subjective Probability?Author(s): Peter BaumannSource: The Philosophical Quarterly, Vol. 48, No. 191 (Apr., 1998), pp. 199-200Published by: Oxford University Press on behalf of the Scots Philosophical Association and theUniversity of St. AndrewsStable URL: http://www.jstor.org/stable/2660290 .

Accessed: 20/12/2014 17:02

Your use of the JSTOR archive indicates your acceptance of the Terms & Conditions of Use, available at .http://www.jstor.org/page/info/about/policies/terms.jsp

.JSTOR is a not-for-profit service that helps scholars, researchers, and students discover, use, and build upon a wide range ofcontent in a trusted digital archive. We use information technology and tools to increase productivity and facilitate new formsof scholarship. For more information about JSTOR, please contact support@jstor.org.

.

Oxford University Press, Scots Philosophical Association, University of St. Andrews are collaborating withJSTOR to digitize, preserve and extend access to The Philosophical Quarterly.

http://www.jstor.org

This content downloaded from 128.235.251.160 on Sat, 20 Dec 2014 17:02:48 PMAll use subject to JSTOR Terms and Conditions

The Philosophical Quartery, Vol. 48, Jo. 9ig April 998g ISSNoo3I-8o94

DISCUSSIONS

CAN RELIABILISTS BELIEVE IN SUBJECTIVE PROBABILITY?

BY PETER BAUMANN

I shall argue here that the answer to the above question is 'No': reliabilist

conceptions of knowledge do not (given very plausible assumptions) cohere with a

subjectivist conception of probability. Why not? Suppose that

I. If S knows that q, then Ss true belief that q is reliable.'

'Reliability' is an irreducibly probabilistic notion: a belief is reliable if and only if the conditional probability P of its truth, given that it was acquired in a certain way, is

very high, that is, higher than a certain value a. Hence

2. If S knows that q, then P > a.

Now suppose further that

3. All probability is subjective probability.2

According to adherents of (3), probability is relative to a given body of knowledge of a given person at a given time. Hence, given this purely epistemic notion of prob- ability, no sentence of the form 'P > a' as such has a truth-value; only sentences of the form 'P > a, relative to body of knowledge B of person A at time t' have truth- values. Thus, given (3), one should replace (2) by

4. If S knows that q, then P > a, relative to body of knowledge B, person A and time t.

Hence it is also true that

5. Whcther S knows that q varies with persons, times and bodies of knowledge. In other words

6. Whether S knows that q depends on the subjective probability P of the person considering the question at a given time t.

I See, e.g., F.P. Ramsey, 'Knowledge', in his Philosophical Papers, ed. D.H. Mellor (Cam- bridge UP, 1990), p. Io.

2 See, e.g., Bruno de Finetti, Theory of Probability (New York: Wiley, 1970), Vol. I, pp. x-xi.

? The Editors of 771e P/ilosophical Qtar/terlr, 1998. Ptblished by Blackwell Publishers, o18 Cowley Road, Oxford ox4 IJF, UK, and 350 Main Street, MIalden, MA 02148, USA.

This content downloaded from 128.235.251.160 on Sat, 20 Dec 2014 17:02:48 PMAll use subject to JSTOR Terms and Conditions

200 PETER BAUMANN

Here is an example. Maria has acquired the true belief that it is going to rain by using a particular - and contested - method of weather-forecasting M. According to Fred's subjective probabilities, M is pretty unreliable, whereas according to Jack's it is pretty reliable. How shall we answer the question whether Maria knows that it is

going to rain? Given (3), such a question cannot even be posed: whether Maria knows that it is going to rain depends on the subjective probabilities of the person who is considering the question. In our example, it depends on whether we listen to Fred or to Jack: if to Jack, then Maria may know that it is going to rain, but if to Fred, then Maria does not know that.

To be sure, whether S knows that q depends on objective facts. But - according to those who believe both in reliabilism and in exclusively subjective probability - it does not depend on objective facts about S alone but also on facts about another

person A: not only on S's epistemic state but also on A's epistemic state. In this

respect, knowledge would turn out to be irreducibly social. The result of all this is the following. Three things are incompatible:

I. a non-relativist view of knowledge 2. reliabilism

3. an essentially subjective view of probability.

Hence, if we start both with reliabilism and an essentially subjective view of prob- ability, then the concept of knowledge gets relativized. This gives us three options:

(a) accept a relativized notion of knowledge (b) not accept that and dump reliabilism (c) not accept that and dump an exclusively subjective view of probability.

If the above relativity of knowledge seems absurd, and if reliabilism is plausible, then one should give up a purely subjective view of probability. And that seems to be the right thing to do.

Universitdt Gittingen

RESEMBLANCE CANNOT BE PARTIAL IDENTITY

BY ARDA DENKEL

I intend to refute the analysis of resemblance as partial identity. By doing this I hope to render a service to the so-called 'resemblance thesis', a principal buttress of conceptualism. That some realists try to reduce resemblances to partial identities is well known. If their attempt is right, the conceptualist who explains away the

? The Editors of The Plilosophical Qugarterly, 1998

This content downloaded from 128.235.251.160 on Sat, 20 Dec 2014 17:02:48 PMAll use subject to JSTOR Terms and Conditions