|
|
|
 |
 National Science
Foundation Award #0549886 |
 |
|
|
Learning From Error: A Philosophy of Evidence and Inference |
| |
| Investigator(s): |
Deborah Mayo (PI)
|
| Sponsor: |
Virginia Polytechnic Institute and State University, VA 24060 5402315281
|
| Start Date/Expiration Date |
2006-02-01 to 2006-07-31 (amended 2006-01-31) |
| Awarded Amount to Date: |
$41,330 |
| Abstract: B. Project Summary
How do we obtain reliable knowledge about the world despite errors and uncertainties? Although science is traditionally viewed as a source of authority, knowledge, and progress, it has been notoriously difficult for philosophers of science to pinpoint what is distinctive in science as compared with other possible ways of knowing. During this grant period, the PI will conduct research and writing chapters for a book, Learning From Error: A Philosophy of Evidence and Inference based on the idea that the ways humans learn from error is the key to the growth of knowledge. Intellectual Merit: This research will yield significant progress on: traditional philosophical problems of evidence, induction, and reliability, and problems of methodology in scientific practice. It will respond to criticisms and gaps in earlier work and will serve to link several current approaches to philosophy of science: work on experiment, modeling, and naturalistic philosophies of science. In the philosophy of science being developed, the goal of inquiry is not avoiding anomaly and error but learning from error creatively. By examining how analogous uncertainties, biases, and limitations crop up in different fields, this research will show how effective ways of coping in one field gives insight into very different areas. This work will enable the PI to more effectively teach interdisciplinary courses in the departments of philosophy, economics, and STS. Broader Impacts: Failing to solve traditional problems of justifying science has influenced attitudes toward science and to responsible inquiry inside and outside academia. Since the classic strictures about testing and evidence set out by philosophers appear to be violated in the best scientific practices, many conclude that no general principles exist. The problem is that these principles have been appraised in a context-independent fashion rather than trying to understand what work they do and when they may be violated (which is very different from saying they may be ignored). The PI gets around this problem in two steps: first by articulating the value of reliably and severely probing for, and learning from, errors, and second, by wrestling with evidential principles that are often debated, discarded or used without clear justification. One of the controversial methodological questions considered is: Does it matter if the same data are used both to arrive at, as well as evaluate or test, a hypothesis or model? Inability to distinguish warranted cases of double counting, especially in the social sciences, leads some to question the epistemological credentials of all forms of searching and data-mining. Discovering the lack of any neat computational method to resolve this long-debated question, the PI takes this as a valuable invitation for synergistic work (between philosophers, statisticians, social and natural scientists). The problems being taken up, and how researchers in science studies resolve them, will influence where practitioners look for methodological guidance, as well as our ability to appreciate and understand the nature of responsible inquiry. Mayo s work in philosophy of science interrelates with issues in evidence-based policy, and this work will be applied directly to current methodological issues in the areas of environmental and medical risk assessment. A central theme is that adequately addressing issues about values in evidence-based policy calls for corresponding attention to issues about methods for collecting, interpreting, communicating, and evaluating uncertain evidence; (e.g., Acceptable Evidence, Science and Values in Risk Management, jointly edited with Rachelle Hollander.) In this project the PI is promoting, both by argument and illustration, the goal of a genuinely engaged interdisciplinary philosophy of science. |
|
| NSF Org: |
SES - Division of Social and Economic Sciences |
| Award Number: |
0549886 |
| Award Instrument: |
Standard Grant |
| Program Manager: |
Ronald Rainger
SES Division of Social and Economic Sciences
SBE Directorate for Social, Behavioral & Economic Sciences
|
| NSF Program(s): |
Hist & Philosophy of SET |
| Field Application(s): |
|
| Program Reference Code(s): |
UNASSIGNED, 0000 |
| Program Element Code(s): |
1353 |
|
|
| |
 |
|
|
|
|
|
|
|
