Justified True Belief Triggers False Recall of “Knowing”

gettierThis post is based on the paper presented at the 2013 Annual Conference of the of the Cognitive Science Society, “Justified True Belief Triggers False Recall of “Knowing”” by Derek Powell, Zachary Horne, Angel Pinillos, and Keith J. Holyoak. People’s beliefs are the primary drivers of their actions, yet these beliefs are often uncertain—the products of limited information about the world and interconnections between other (often uncertain) beliefs.

For this reason, a capacity for evaluating the status of different beliefs is important for individuals in directing their own rational behavior, and for predicting the behavior of others. Understanding these processes requires an analysis of the concept of knowledge:  the distinction between what is known versus what is merely believed, imagined, hoped for, or assumed.  According to the authors, there has been very little psychological research examining the self-assessment and attribution of knowledge. Making a decision about when to attribute knowledge, either to oneself or to another, hinges on one’s conception of knowledge: it is a decision about whether or not the concept applies in a particular instance. Although other factors may play into this decision process, understanding the exact nature of the concept itself is essential to understanding the overall process of knowledge attribution. Philosophers have
long contemplated the nature of knowledge, and have also developed a variety of methods for studying concepts. A common method involves using thought experiments. One of the most influential of these thought experiments was proposed by Edmund Gettier (1963). “Gettier cases” challenge the traditional conception of knowledge. Prior to the 1960s, most philosophers thought that knowledge should be analyzed as justified true belief. Today, many philosophers see Gettier cases as counterexamples to that analysis. Gettier cases are situations in which an agent has a true belief that is justified, but an element of luck is involved that disqualifies their cognitive state from being considered knowledge. To illustrate such a case, suppose that at 3:34pm an agent comes to believe it is 3:34pm by looking at her normally reliable watch. Suppose also that unbeknownst to the agent, her watch had been stopped for exactly 24 hours—she just happened to glance at her watch at the correct time. The agent’s belief is not only true, but is also justified (since looking at one’s normally reliable watch is a good way to form veridical beliefs about time of day). However, most philosophers judge that this agent does not know that it is 3:34pm, because her belief is true only by luck. If this judgment is correct, then this case is a counterexample to the traditional thesis that knowledge is justified true belief.

Do laypeople share Gettier’s conception of knowledge? Starmans and Friedman investigated lay people’s evaluations of Gettier cases by presenting participants with short vignettes that described agents forming beliefs under different circumstances. Three different versions of each scenario were created: the agents in the vignettes either formed a false belief, formed a justified true belief, or were “Gettiered”—the belief they formed was both justified and true, but was true only by luck. Starmans and Friedman then asked participants to judge whether the agents “knew” or “only believed” the proposition in question, and to rate how confident they were in their judgment.  Based on the results Starmans and Friedman concluded that laypeople view knowledge as justified true belief, in accord with the more traditional philosophical view.

Powell et al saw weaknesses in the survey methodology of Starman and Friedman and set up experiments using semantic integration.  The experiments although clever confirmed the Starman and Friedman findings. False recall of ‘knew’ was observed significantly more often in the Justified True Belief (42%) and Gettier (47%) conditions as compared with the False Belief (23%) condition.  Participants thus seemed to believe that agents in Gettier cases possess knowledge, and apparently drew no distinction between Gettier cases and non-Gettier cases of justified true belief.

Frankly, this paper seemed more interesting before I read it. It exposed me to the Gettier conditions of which I was completely ignorant.  It seems unsurprising that laypeople would not be great at distinguishing what would likely be a relatively rare situation, especially when 23% thought they “knew” what was clearly a false belief.

Powell, D., Horne, Z., Pinillos, A., Holyoak, K. (2013)  “Justified True Belief Triggers False Recall of “Knowing.””  Annual Conference of the of the Cognitive Science Society.