Presented at the Philosophy of Science Associate meeting in San Diego (November 2012) with the final version to be published in a supplementary volume of Philosopy of Science.
There are two senses of `what scientists know': An individual sense in which scientists report their own opinions, and a collective sense in which one reports the state of the discipline. The latter is what is of interest for the purpose of policy and planning. Yet an expert, although she can report the former directly (her opinion on some question), can only report her considered opinion of the latter (the community opinion on the question). Formal judgement aggregation functions offer more rigorous frameworks for assessing the community opinion. They take the individual judgements of experts as inputs and yield a collective judgement as an output. This paper argues that scientific opinion is not effectively captured by a function of this kind. Inferential relationships are themselves matters to be judged by experts involving risks which must be weighed, and the significance of the risk depends on value judgements.
@ARTICLE(Magnus2013b,
AUTHOR = {P.D. Magnus},
TITLE = {What scientists know is not a function of what scientists know},
JOURNAL = {Philosophy of Science},
YEAR = {2013},
VOLUME = {80},
NUMBER = {5},
MONTH = dec,
URL = {http://www.jstor.org/stable/10.1086/673718}
)