So good it has to be true: Wishful thinking in theory of mind

5 In standard decision theory, rational agents are objective, keeping their beliefs independent from their 6 desires. Such agents are the basis for current computational models of Theory of Mind (ToM), but the 7 accuracy of these models are unknown. Do people really think that others do not let their desires color 8 their beliefs? In two experiments we test whether people think that others engage in wishful thinking. We 9 find that participants do think others believe that desirable events are more likely to happen, and that 10 undesirable ones are less likely to happen. However, these beliefs are not well calibrated as people do not 11 let their desires influence their beliefs in the task. Whether accurate or not, thinking that others wishfully 12 think has consequences for reasoning about them. We find one such consequence—people learn more 13 from an informant who thinks an event will happen despite wishing it was otherwise. People’s ToM 14 therefore appears to be more nuanced than the current rational accounts in that it allows other’s desires to 15 directly affect their subjective probability of an event. 16 Whether thinking “I can change him/her” about a rocky relationship or the more benign “those clouds 17 will blow over” when at a picnic, people’s desires seem to color their beliefs. However such an 18 explanation presupposes a direct link between his desires and beliefs, a link that is currently absent in 19 normative behavioral models and current Theory of Mind (ToM) models. 20 Does a causal link between desires and beliefs actually exist?1 The evidence is mixed. There are a 21 number of compelling studies that find “wishful thinking,” or a “desirability bias” in both carefully 22 Corresponding author: Daniel Hawthorne-Madell, [email protected] 1 While the causal link between desires and beliefs may, in fact, be bi-directional, we will focus on the evidence for the a priori effect of desires on beliefs. Copyright ©2017 The Author(s). Published by MIT Press. All Rights Reserved. == D R A F T July 29, 2017 == Journal: OPEN MIND / Title: So good it has to be true: Wishful thinking in theory of mind Authors: Daniel Hawthorne-Madell and Noah D. Goodman controlled laboratory studies (Mayraz, 2011) and real world settings, such as the behavior of sport fans 23 (Babad, 1987; Babad & Katz, 1991), expert investors (Olsen, 1997), and voters (Redlawsk, 2002). 24 However, other researchers have failed to observe the effect—for example, Bar-Hillel et al.’s The elusive 25 wishful thinking effect (1995), have provided alternative accounts of previous experiments (Hahn & 26 Harris, 2014), and have argued that there is insufficient evidence for a systematic wishful thinking bias 27 (Hahn & Harris, 2014; Krizan & Windschitl, 2007). 28 Whether or not there actually is a direct effect of desires on beliefs, people might think that there is and 29 use this fact when reasoning about other people. That is to say, people’s ToM might incorporate the 30 wishful thinking link seen in Figure 1b. The direct influence of desires on beliefs is a departure from 31 classic belief-desire “folk” psychology in which beliefs and desires are independent and jointly cause 32 action (Figure 1a). Previous models of ToM formalize belief-desire psychology into probabilistic models 33 of action and belief formation. They show that inferring others’ beliefs (Baker, Saxe, & Tenenbaum, 34 2011), preferences (Jern, Lucas, & Kemp, 2011), and desires (Baker, Saxe, & Tenenbaum, 2009) can be 35 understood as Bayesian reasoning over these generative models. A fundamental assumption of these 36 models is that beliefs are formed on the basis of evidence, and a priori independent of desire. We will 37 call models that make this assumption rational theories of mind (rToM). We can contrast this rationally 38 motivated theory with one that incorporates the rose-colored lenses of a desire-belief link, an optimistic 39 ToM (oToM).2 We use their qualitative predictions to motivate two experiments into the presence (and 40 calibration) of wishful thinking in ToM and its impact on social reasoning. 41 In Experiment 1 we explore wishful thinking in both ToM and behavior. In the 3rd person point-of-view 49 (3-PoV) condition, we test whether people use a rToM or an oToM when reasoning about how others 50 play a simple game—will manipulating an agent’s desire for an outcome affect people’s judgments about 51 the agent’s belief in that outcome? In the first person point of view (1-PoV) condition we test whether 52 people actually exhibit wishful thinking when playing the game themselves. We carefully match the 53 (3-PoV) and (1-PoV) conditions and run them concurrently to have a clear test of whether people’s ToM 54 assumptions lead them to make appropriate inferences about people’s behavior in the game.3 Regardless 55 of its appropriateness, people’s ToM should have consequences for both how they reason about others’ 56 2 We formally describe Bayesian models of both rToM and oToM in the Supplement (citation of supplementary material here XXX. 3 Experiment 1 is a slightly modified replication of the two conditions previously run as separate experiments (see Supplement). –2– == D R A F T July 29, 2017 == Journal: OPEN MIND / Title: So good it has to be true: Wishful thinking in theory of mind Authors: Daniel Hawthorne-Madell and Noah D. Goodman Figure 1. Competing models of ToM. Causal models of (a) rational ToM based upon classic belief-desire psychology and (b) optimistic ToM that includes a direct “wishful thinking” link between desires and beliefs. 42