Analysis of Kelly's Disagreement and Higher Order Evidence Theory

a) Abstract

In his article “Peer Disagreement and Higher Order Evidence,” Thomas Kelly offers his explanation of, and response to, the “normative significance of peer disagreement.”[1] This involves the question of the epistemically responsible course of action when two individuals, with access to the same evidence supporting a given proposition, who feel that the other has no advantage when it comes to assessing said evidence, and who possess no knowledge of relevant defeaters which would compromise the other party’s assessment abilities, come to incompatible conclusions.  Kelly criticizes one of the suggested responses to the problem of peer disagreement, the equal weight view, and offers his total evidence view as an alternate course of action.

This paper will critique Kelly’s four major criticisms of the equal weight view, identifying major shortcomings in each of Kelly’s efforts, including an unacceptably ambiguous use of the term reasonableness, a misconstruction of the argument for the equal weight view, and a theoretical argument that relies too much on omniscient or ex post facto information to be of normative use. It will also take issue with Kelly’s total evidence view, arguing that it is founded on an incorrect interpretation of the interplay between first and higher order evidence. This paper will, therefore, conclude that Kelly’s efforts are insufficient to advise abandoning the equal weight view in favour of his total evidence view.

b) Summary

Kelly begins by summarizing his understanding of the equal weight view and the nature of peer disagreement. He suggests that the existence of peer disagreement, when epistemic peers arrive at incompatible conclusions, despite access to the same evidence and comparable evaluation abilities, leads to a significant normative quandary. This quandary,   “how (if at all) [each party] should revise their original views.”[2] He offers the equal weight view as a common response, stating that it requires that “in cases of peer disagreement, one should give equal weight to the opinion of a peer and to one’s own opinion.”[3]

Get Help With Your Essay

If you need assistance with writing your essay, our professional essay writing service is here to help!
Find out more about our Essay Writing Service

Kelly offers four main critiques of the equal weight view. It is important to note that in these examples, Kelly uses the standard Bayesian conception of credence in some proposition as represented by numerical confidence levels, rather than just dichotomous or ambiguous categories.[4] His first critique involves an analogy to intrapersonal beliefs. He suggests that in cases in which an individual realizes they hold two inconsistent beliefs, the most logical course of action is not always to abandon both beliefs, as is the type of action recommended by the equal weight view in similar interpersonal peer disagreements.[5] He argues in these intrapersonal cases, which beliefs are abandoned depends on the total evidence supporting each belief, and that a similar course of action is best for interpersonal disagreements.

His second critique involves “implausibly easy bootstrapping.”[6] He introduces a case in which two individuals respond to a body of evidence E, which leads one individual to believe proposition H with a confidence of .8 (on a 0 to 1 Bayesian scale of confidence), while the other believes H with a confidence of 0.2. Based on the equal weight view, both parties split the difference and arrive at a confidence level of 0.5 in proposition H. However, one or both individuals arrive at the original confidence level by grossly misinterpreting evidence E. Kelly argues that the equal weight view allows us to arrive at a rational belief in confidence 0.5, despite irrational prior methods of evidence evaluation.

Kelly’s third critique involves “even easier and more implausible” bootstrapping in single person cases. This involves a distinction of two types of evidence, first the direct or first order evidence E that bears on some proposition, and second the psychological evidence that consists of a belief formed about some proposition from evidence E. Kelly argues that in cases in which peer disagreement occurs, only each peers psychological evidence is considered, namely the level of confidence each individual has in some proposition. It is in this level of confidence which the equal weight view has each party split the difference.[7] Kelly argues this completely neglects the original first order evidence. He further argues that in single person cases, proponents of the equal weight view would have to similarly privilege higher level psychological information. This allows individuals to form “reasonable” beliefs despite having themselves completely misevaluated the original body of evidence.

Kelly’s fourth critique involves the treatment of other individual’s opinions as a form of privileged information. He uses the analogy of a piece of litmus paper to demonstrate why he considers this fallacious. He states that we use litmus paper because it is generally a reliable indicator of whether a solution is acidic. This is similar to how one uses the opinion of a peer as a reliable indicator of some state of affairs.[8] However, if we possessed another body of evidence that strongly indicated that the litmus paper was faulty, for example, if you had recently poured 10 grams of lithium hydroxide in a 10 ml solution of water, you would not privilege the litmus paper’s opinion above the strong evidence that the solution is indeed basic. Similarly, Kelly argues we should not privilege a peer’s opinion above the wide array of non-psychological evidence we have access to.

This leads Kelly to introduce what he calls the total evidence view. This states that when considering some proposition H, each party has total evidence consisting of E*, namely, our original non-psychological evidence E, one party’s opinion that H is true and the other party’s opinion that H is false. Kelly argues that the two latter pieces provide some evidence towards agnosticism about H, but still must be weighed against the original evidence E for H. Therefore, the correct response to H is a somewhat weakened, but does not completely split the confidence in each party’s opinion.[9] Given the introduction of more peers, there may come an instance where the psychological evidence outweighs the original evidence E, but this would require a much larger batch of psychological evidence than required by the equal weight view.

Bibliography

Kelly, Thomas. “Peer Disagreement and Higher Order Evidence.” In Social Epistemology:  Essential Readings, edited by Alvin I. Goldman and Dennis Whitcomb, 183-217. New               York: Oxford University Press, 2011.

Schafer, K. “a Brief History of Rationality: Reason, Reasonableness, Rationality, and Reasons.”               Manuscrito 41, no. 4 (2018): 501-529.

This article provides a brief history of the development of the concepts of reasonableness and rationality. It will be used to identify the ways in which Kelly is conflating or ambiguously using the concepts of reasonableness and rationality throughout his critiques of the equal weight view.

Williamson, Timothy. Knowledge and Its Limits. Oxford: Oxford University Press, 2000.

This book engages with the nature of evidence, rationality, and reasonableness and whether rationality is only based on the quality of one’s response to evidence, or whether that response reflects the actual truth. This will be used to engage with Kelly’s idea that a failure to engage with first-order evidence in a way that reflects the objective truth is a defect in the reasonableness of the equal weight view.

Jeffrey, Richard C. Probability and the Art of Judgment. Cambridge [England]: Cambridge               University Press, 1992.

This book explores probabilistic forms of evidence, which revolve around the Bayesian concept that confidence in proposition H is based on the probability of a given piece of evidence supporting H. This will be used to critique Kelly’s use of confidence probabilities in support of his total evidence view.

[1] Thomas Kelly, “Peer Disagreement and Higher Order Evidence,” in Social Epistemology: Essential Readings, ed. Alvin I. Goldman and Dennis Whitcomb (New York: Oxford University Press, 2011), 183.

 [2] Thomas Kelly, “Peer Disagreement and Higher Order Evidence,” in Social Epistemology: Essential Readings, ed. Alvin I. Goldman and Dennis Whitcomb (New York: Oxford University Press, 2011), 184.

 [3] Thomas Kelly, “Peer Disagreement and Higher Order Evidence,” in Social Epistemology: Essential Readings, ed. Alvin I. Goldman and Dennis Whitcomb (New York: Oxford University Press, 2011), 184.

 [4] Thomas Kelly, “Peer Disagreement and Higher Order Evidence,” in Social Epistemology: Essential Readings, ed. Alvin I. Goldman and Dennis Whitcomb (New York: Oxford University Press, 2011), 187.

 [5] Thomas Kelly, “Peer Disagreement and Higher Order Evidence,” in Social Epistemology: Essential Readings, ed. Alvin I. Goldman and Dennis Whitcomb (New York: Oxford University Press, 2011), 192.

 [6] Thomas Kelly, “Peer Disagreement and Higher Order Evidence,” in Social Epistemology: Essential Readings, ed. Alvin I. Goldman and Dennis Whitcomb (New York: Oxford University Press, 2011), 193.

 [7] Thomas Kelly, “Peer Disagreement and Higher Order Evidence,” in Social Epistemology: Essential Readings, ed. Alvin I. Goldman and Dennis Whitcomb (New York: Oxford University Press, 2011), 185.

 [8] Thomas Kelly, “Peer Disagreement and Higher Order Evidence,” in Social Epistemology: Essential Readings, ed. Alvin I. Goldman and Dennis Whitcomb (New York: Oxford University Press, 2011), 197.

 [9] Thomas Kelly, “Peer Disagreement and Higher Order Evidence,” in Social Epistemology: Essential Readings, ed. Alvin I. Goldman and Dennis Whitcomb (New York: Oxford University Press, 2011), 203.