Do I have more free will than you do? more

The New School Psychology Bulletin 2011, Vol. 9, No.1 Copyright 2011 by The New School for Social Research Print ISSN: 1931-793X; Online ISSN: 1931-7948 Do I Have More Free Will Than You Do? An Unexpected Asymmetry in Intuitions About Personal Freedom Brian D. Earp Yale University The present research explores the relationship between moral evaluations and intuitions about the causes of human behavior, in particular freedom of the will. Two studies test for a self-serving bias in intuitions about free will. Study 1 explores whether individuals may seek to exculpate themselves from wrongdoing by denying free will, while justifying blame of others by endorsing free will. Study 2 explores whether individuals may justify personal failures by denying free will, while taking credit for personal successes by endorsing free will. In neither study do the data show the predicted and collapsing across conditions, it is shown that participants give greater endorsement of free will whenever actions are Key Words: free will, perspective, blame, self-serving, self-other According to common wisdom, people either believe in free will or they do not. That is, intuitions about free will are taken to be relatively robust and stable, not susceptible to major alteration experimental philosophy has shown that beliefs involving free instance, have found that people’s beliefs about the compatibility conception of free will is not an entirely new idea, but rather has a long history in philosophical thought. According to Nietzsche, The notion is also consistent with seminal work in social people pay more attention to others’ socially undesirable behavior than to their socially desirable behavior, and are likelier to make dispositional inferences in cases involving the former compared to more likely to think of them as personally responsible for their Whatever the case when it comes to judging others’ behavior, intuitions about freedom might differ if one is the actor rather than the observer of a morally-valenced action. If people may be motivated to assign blame to others, but to avoid blame for themselves, it would follow that individuals will endorse free violent, immoral act is described in gory detail, participants tend to hold the actor morally responsible, even if the actor is said to inhabit an explicitly deterministic universe. But if the same act, occurring in the same universe, is described in a more abstract or theoretical way, participants will deny that the actor could be a normative ethical perspective, such as the emotional salience of an action, may dramatically sway folk judgments about moral responsibility. If this is the case, what about people’s intuitions about free will itself? Recent research suggests that certain considerations, that action. For example, in one study, participants interpreted immoral actions as being less forced, and therefore more free, than In other words, the freedom that participants attributed to actors behaving in much the same way varied on the basis of whether the behavior elicited a negative moral reaction in the perceiver. as being blameworthy intuitions about freedom and responsibility in line with this initial judgment, whether consciously or unconsciously. According to for] observers . . . to attribute the same actions to stable personal self-serving inferences about the causes of their behavior, allowing them to take credit for the good things they do, while blaming the situation for the bad things they do. Applied directly to the concept of free will, these convergent blaming others by attributing more free will to them when their 21 EARP own immoral behavior by denying that they were free to have done otherwise. Accordingly, this paper’s theoretical expectations are as follows: when individuals perform a good action, they will but when they do something bad, they will attribute their behavior upbringing, and the immediate situation. But when individuals perceive others performing good and bad actions, their free will own free will, while their good actions are determined by what’s expected of them socially. Two studies test for such a self-serving bias in intuitions about free will. Study 1 explores whether individuals may seek to exculpate themselves from wrongdoing by denying free will, while blaming others by endorsing free will. Study 2 explores whether individuals may justify personal failures by denying free will, while taking credit for personal successes by endorsing free will. Study 1 belief in free will. In particular, it asks whether the direction of change varies depending upon whether one is the actor or observer of a moral versus immoral act. If belief in free will is selective and self-serving, participants made to feel as though they have done something immoral should respond that behavior is due more to genetic and situational causes, and less to free will. Conversely, Participants in the bad condition copied Participants in the and morally bad conditions copied identical passages except that they were written from the third-person perspective about a character named Gary. about the role of free will in human behavior, which was embedded and were asked to graphically indicate their response to the were considered ‘free’ only so that they might be considered in order to blame or condemn others, selective belief in free will might also function such that individuals can take credit for their own moral triumphs. Therefore, it is predicted that participants made to feel that they themselves have acted morally will also attribute more of human behavior to free will and less to genetic or environmental factors. Methods Participants The participants in this study were undergraduates at a N female, 18 male; ages 18-27, M= 19.36; SD Procedure The independent variable in this study is based on a priming were asked to copy out by hand a passage that described either a moral or immoral behavior, and that was written from either the in previous research as a way of inducing moral guilt and moral judgment in a controlled manner. Participants in the morally good condition copied out the following passage: perspective conditions. Based on these analyses, participants were no more likely to endorse free will as a cause of behavior M M t = .78, nor did participants attribute more to free will when writing about M M t = .84. The self-serving free will hypothesis predicted an interaction between the perspective were probed for knowledge of the hypothesis. Results Contrary to hypotheses, expressed belief in free will did not differ between conditions. An independent-samples t-test immoral act in the third person condition. Based on a univariate ANOVA, no such interaction was found: = .74. 22 DO I HAVE MORE FREE WILL THAN YOU DO? M M = 4.13, t regardless of the moral valence of the passage. Discussion There are several possible interpretations of these results. While the null result could be taken as evidence that belief in free will is in fact highly robust and consistent across situations, alternative explanations are more plausible. Because Study 1 did not include a manipulation check, it is unclear whether participants were indeed made to feel as though they had acted immorally. experiment were adapted likewise did not include a manipulation check; and the purpose here was to replicate, not validate, their materials. Copying the morally negative passage from a it is possible that participants continued to think about the agent in the passage as another individual. In other words, the pronoun manipulation may have failed to tap into the actor-observer distinction as intended. Also, it is possible that the moral valence of the passage was not salient enough to reveal the hypothesized effect. It is conceivable that standard course of action in the competitive environment described. A more dramatic difference between the valence of the moral and immoral passages might be necessary to obtain the expected shift use of a stronger manipulation, and the inclusion of manipulation as moral and immoral, and that participants internalize the related side note, it may be worth mentioning here that at least Procedure The experimental design for Study 2 was identical to that participants were assigned to a condition, and again into a Participants in the the following passage: or will condition. will condition copied or described behavior, intuitions about free will could plausibly be more sensitive to these manipulations. Intuitively, denial of free will would be a most effective exculpation strategy if the offense were directly rooted in personal will. It is therefore predicted that although the moral-immoral distinction did not generate the hypothesized shift in intuitions, manipulating success-of-will and failure-of-will may. Method Participants The participants in this study were undergraduates at a N= 41; 19 female, 22 male; ages 18-23, M= 19.72, SD locations on campus. Although in Study 1 participants did not appear to decrease attributions of free will after being made to feel immoral, it is possible that belief in free will could provide a self-serving function. Study 2 explores a different context in which free will could be selectively endorsed so as to exculpate oneself for negative actions or take credit for positive ones. In this study, participants are again primed with behaviors that are oppositely valenced across the two conditions, but rather than a moral and immoral act, they are an act of failed will versus the successful exercise of restraint. In = 0.136], directly implicated. Accordingly, Study 2 was designed to focus on actions more clearly involving the concept of will. Study 2 of the reasons just outlined. That is, the priming manipulation’s effectiveness in the original study, widely-cited though it is, may be inconsistently reproducible. Nevertheless, at the time the present research was conducted, there was no reason to think that the reputable journal , was at all dubious. As more labs clearer picture will emerge. In addition to these concerns about the effectiveness of the priming task, it is possible that the manipulation, even as intended, would not reveal a self-serving bias. While the action described in the morally bad condition may be perceived as a moral transgression, it is not necessarily a failure of personal will. Therefore, even if participants fully internalized the action by taking the perspective of the speaker, and were motivated to exculpate themselves, reduced endorsement of free will does not clearly accomplish this goal. The self-serving effect of free will may be restricted to contexts in which the concept of will is Participants in the will condition 23 EARP portion of human behavior is caused by free will, as distinct from this common endorsement of free will, participants’ intuitive and condition copied out identical passages except that they were written in the third-person. Results & Discussion As before, the hypotheses are not supported by the data. Contrary to predictions, participants were no more likely to endorse free will as a cause of behavior when they wrote about exercising restraint M wro M t = .89, nor did participants attribute more to free will when writing about failed restraint from a third-person M M t = .99. In this study, there were no clear intuitions about whether there would be an actor-observer asymmetry. However, it was predicted that there should be a difference based on the valence of the action, which was not found. Again, there was a moderate trending towards a M M = 4.09, t = .107], regardless of the moral valence of the passage. While Study 2 failed to support the stated hypothesis, there are again several possible interpretations of these results. Failure to internalize the actions of the passage is again a plausible methodological problem. As with Study 1, a manipulation check was not included: both studies were conducted partially concurrently, and both were modeled strictly on the design of the study materials were adapted. Further, given that existing research reveals that differing values generate different folk judgments seems plausible that a diversity of attitudes towards dieting may create variance in judgments about free will, swamping out a potential effect. Again, a manipulation check should be used the valence of the passages as predicted and that they take on the also signal a problem with the experimental design. Since the is selective and self-serving, it may be problematic that participants have such diverse conceptions of what it means to be free. However, the dependent measure seems to tap into a libertarian free will concept, resembling something like agent causation. It is interesting that nearly all participants express the belief that some percentage of human behavior is caused by something distinct from genetic and environmental factors, early learning, and so on. A further potential limitation concerns the scope of the This broad phrasing was chosen on purpose, to see whether the hypothesized motivations to blame others or to excuse one’s own behavior might have a global effect on one’s intuitions about free will , as opposed to a narrowly-tailored effect on one’s The reasoning was that if an interaction effect were obtained in the broad case, one could expect to see it, , in the narrow case. One interpretation of the data, then, is that the broad effect simply does not obtain. However, this would not rule out the possibility of an interaction between perspective and valence conditions for would do well to test the narrow, and perhaps more conservative, hypothesis. Despite failing to elicit the predicted response, one could argue that the dependent measure used in this study represents intuitions about free will. First, by asking about the to which free will is a cause of behavior rather than measuring binary or scaled endorsement of belief in free will, we create room for greater variance in responses. Given that ceiling effects are likely to occur for a binary measure of belief in free will, the variable used here may prove useful in future studies in that it may be sensitive to more nuanced or subtle shifts in folk intuitions. Moving beyond limitations, however, the most interesting point that emerges from these two studies is that, collapsing across the moral-immoral and strength-weakness of will conditions, belief in free will. Although this had not been the primary interest that intuitions about free will are not stable, but rather vary depending on motivation, context, and framing. Further research Again, participants in the from the larger sample size, the amount of behavior attributed to M 2.249, following section. General Discussion Although the data fail to show the predicted effects, the obtained results raise several points of interest. First, the responses to the dependent measure used in this study are interesting in their own some M = 4.11, t the kinds of information we have about ourselves and our own behavior compared to others. Together with her colleagues, Pronin 24 DO I HAVE MORE FREE WILL THAN YOU DO? has found, for example, that we perceive others as susceptible to access to our own beliefs and motivations causes us to perceive on our everyday practices of praise and blame; and it is hoped that the two studies presented in the current paper add a small drop to the form of their behavior. It is possible that an awareness of the different thoughts, feelings, deliberations, considered choices, and so on, within our own heads, gives rise to a feeling of personal we lack awareness of their internal states and have access only to actual behavior, giving an impression of greater determinacy. At present, it can only be said that the main effect observed, although was designed to test, does support the general prediction that actor and observer perspectives generate different free will attributions. the speculative explanation we have provided. between conditions along both the moral-immoral and personal failure-success dimensions lends itself to several conceivable interpretations. It is possible that belief in free will is simply not rooted in, or sensitive to, the desire to blame or take credit for actions. It is also plausible that the primary function of belief in free will is indeed to facilitate responsibility attributions, but that motivations to assign blame or receive praise. Belief in free will might serve as a basis for responsibility judgments but nonetheless given the prevalence of motivated reasoning in a variety of related Amsterdam: Elsevier. plausible. Existing research has documented that people locally differentiate their concepts of freedom and constraint based on the research should continue to explore whether such selectivity exists in more global intuitions about freedom. Conclusions intuitionist approach to moral judgment. , 814-34. Much of the agenda for contemporary philosophical work ‘Freedom and Resentment.’ In that essay, Strawson suggests that we focus not so much on metaphysical speculation about the nature of freedom and determinism as on understanding the actual practices surrounding the assignment of praise regarding how this practice works and what role it serves in people’s lives, it is hoped, progress can be made on the apparent philosophical paradoxes surrounding the notion of moral responsibility. It is worth saluting this shift in emphasis from the millennia-old, seemingly intractable debates about free will to current philosophical and psychological work on the empirical The attribution process in social psychology. In L. Berkowitz New York: Academic Press, pp. 219-266. Divergent perceptions of the causes of the behavior. In E. E. 21–30. of the Macbeth Effect with a Spanish sample. studies have introduced a useful tool in measuring folk beliefs about free will, as well as opened a door for future study designs which may avoid some of the pitfalls and limitations of the present research. The emphasis has been on actor-observer discrepancies in general endorsements of free will, which were hypothesized While these hypothesized differences between conditions were not supported by the data, there was an unexpected general effect of perspective on free will endorsements collapsed across studies, more of typical human behavior to free will. This tentative effect merits further exploration. References blame. in free will, and how it may be Morristown, NJ: General Learning Press, pp. 79-94. Unpublished manuscript, Yale University. The EARP Oxford: Oxford University Press. , 480-498. determinism: The cognitive science of folk intuition. . Translated from Penguin Books: London. about freedom. , 1177-1180. sheep: Asymmetric perceptions of conformity and their roots in an introspection illusion. moral values produce different judgments of intentional action. Unpublished Manuscript, University of California, Irvine. Threatened morality and physical cleansing. , , 30-36. or: The motivational bias is alive and well in attributional theory. 26 DO I HAVE MORE FREE WILL THAN YOU DO? Appendix Figure 1. Prompt and circular chart used to record participant responses. Figure 2. Participants were shown written into the chart. Figure 2 27