Epistemic Values and Information Policies

1
Epistemic Values and Information Policies

• Don Fallis
• Associate Professor
• School of Information Resources and Library
  Science
• University of Arizona

2
Some Information Policies

• Intellectual Property Law
• Speech Regulation
• Privacy Policies
• Digital Divide Policy
• Collection Management Policies
• Evaluation Guidelines
• Library Internet Policies

3
Components of Decision Analysis

• What are the available alternatives (i.e.,
  policies)?
• What are the likely consequences of each of these
  policies?
• What are good consequences? (In other words, what
  do we value?)

4
Consumer Health Information on the Internet

• Alternatives (author is a doctor, HONcode logo,
  many inlinks, no spelling mistakes, etc.)
• Consequences (accuracy, completeness, etc.)
• Values (avoid error, acquire relevant true
  beliefs, etc.)

5
Knowledge is an Important Value in Information
Policy Decisions

• U. S. Constitution on Intellectual Property Laws
• John Stuart Mill on Intellectual Freedom
• Guidelines for Evaluating Information on the
  Internet
• Accuracy in Reference Work

6
Necessary conditions for S to know that p

• S believes p.
• p is true.
• S is justified in believing p.

7
Some Value-Theoretic Approaches to Epistemology

• Why is knowledge more valuable than true belief?
  (Plato, Kvanvig, etc.)
• What are the epistemic utilities of scientists?
  (Levi, Maher, etc.)
• Goldmans theory of veritistic value.

8
Developing an Epistemic Value Theory

• Leveraging existing work in epistemology.
• Answering open questions about epistemic values.

9
Components of a Value Theory

• What things are valuable?
• What is the structure of these values?

10
Questions about Epistemic Values

• What things (e.g., properties of beliefs) are
  epistemically valuable?
• How are epistemic values distinguished from other
  values?
• How can we measure epistemic values?
• What do epistemic values apply to?
• Who do epistemic values apply to?

11
Some Epistemic Values

• Having many true beliefs
• Having few false beliefs
• Having justified beliefs
• Having reasons for ones beliefs
• Having consistent beliefs
• Having explanatory beliefs
• Proportioning ones belief (or confidence) to the
  evidence
• Being certain
• Having understanding
• Being wise

12
Some More Epistemic Values (Goldman 1987)

• Power having many true beliefs
• Speed acquiring true beliefs quickly
• Fecundity many people having true beliefs
• Reliability having relatively few false beliefs

13
Questions about the Structure of Epistemic Values

• What is valuable as a means to what?
• Why is knowledge, in particular, valuable?
• Are epistemic values interconnected?
• How do epistemic values change?
• How are conflicts between epistemic values
  resolved?

14
Types of Value Theories

• Descriptive what peoples epistemic values
  actually are.
• Normative what peoples epistemic values ought
  to be.

15
Conflicts between Epistemic Values

• More True Beliefs versus Fewer False Beliefs
  (i.e., Power versus Reliability)
• Speed versus Reliability
• Power versus Speed
• Power versus Fecundity
• Distribution over People
• Other conflicts

16
Dominance

• Outcome A

• Outcome B

17
Power versus Reliability

• Outcome C

• Outcome B

18
The Role of Non-epistemic Considerations

• Sometimes, non-epistemic considerations will
  trump our epistemic values.
• For example, some knowledge is too painful.
• More often, non-epistemic considerations will
  simply shape our epistemic values.
• For example, the importance of reliability can
  depend on what is at stake.

19
Speed

• Outcome E

• Outcome D

20
Power vs. Speed

• Outcome F
• (e.g., more interlibrary loan)

• Outcome D
• (e.g., more acquisitions)

21
Power vs. Fecundity

• Outcome H
• (e.g., strict intellectual property laws)

• Outcome G
• (e.g., weak intellectual property laws)

22
Worries about Epistemic Value Theory

• We are not just concerned with good consequences.
• We are not just concerned with good epistemic
  consequences.
• Good epistemic consequences are not always good
  consequences.

23
Conclusion

• In order to make information policy decisions, we
  need to be clear about our epistemic values.
• What things are epistemically valuable?
• How are these epistemic values structured?
• My project is to develop such an epistemic value
  theory,
• by leveraging existing work in epistemology,
• and by answering open questions about epistemic
  values.