Unintended, but still blameworthy: the roles of awareness, desire, and anger in negligence, restitution, and punishment

ABSTRACT

Two experiments (Experiment 1 N = 149, Experiment 2 N = 141) investigated how two mental states that underlie how perceivers reason about intentional action (awareness of action and desire for an outcome) influence blame and punishment for unintended (i.e., negligent) harms, and the role of anger in this process. Specifically, this research explores how the presence of awareness (of risk in acting, or simply of acting) and/or desire in an acting agent's mental states influences perceptions of negligence, judgements that the acting agent owes restitution to a victim, and the desire to punish the agent, mediated by anger. In both experiments, awareness and desire led to increased anger at the agent and increased perception of negligence. Anger mediated the effect of awareness and desire on negligence rather than negligence mediating the effect of mental states on anger. Anger also mediated punishment, and negligence mediated the effects of anger on restitution. We discuss how perceivers consider mental states such as awareness, desire, and knowledge when reasoning about blame and punishment for unintended harms, and the role of anger in this process.

KEYWORDS:

• Awareness
• desire
• anger
• negligence
• punishment
• restitution
• intentionality
• blame
• social cognition

Acknowledgements

The authors would like to thank Roger Giner-Sorolla and two anonymous reviewers for their comments on an earlier version of this manuscript. We would also like to thank Jennifer Gray for helping with data collection for Experiment 1 and her comments on an early draft of this manuscript.

Disclosure statement

No potential conflict of interest was reported by the authors.

Notes

¹Nuñez et al. (2014) also describe how negligence can involve agents not possessing knowledge they reasonably should have possessed or lacking awareness when they reasonably should have been aware.

²It should be noted that some research has shown that the tendency to hold others accountable for their immoral thoughts may vary as a function of cultural or religious background. For example, Jews may focus more on observable behaviours when assigning blame, while Protestants are more likely to view immoral thoughts as equally harmful (e.g., Cohen, 2015; Cohen & Rozin, 2001).

³Ten participants were removed for failing a check question regarding the names of the agents in the presented scenarios (original N = 159).

⁴All stimulus materials for Experiments 1 and 2 are available in the appendix.

⁵In addition to the variables reported here, two variables related to accidental causation and intentionality were collected using the same 6-point scale, primarily as respective checks of whether (a) in the no awareness/no desire cell, participants perceived the killing of the pet as reasonably accidental, and (b) intentionality ratings were not unreasonably high in the presence of either awareness or desire. Analyses of these variables confirmed that accidental cause ratings were reasonably high in the appropriate cell (M = 3.71, SD = 1.64), and that ratings of intentionality in the presence of awareness, desire, or both, were not unreasonably high (awareness present, desire absent, M = 2.03, SD = 1.06; desire present, awareness absent, M = 3.22, SD = 1.60; awareness and desire both present, M = 4.03, SD = 1.47). Because our interest was in ratings of negligence and not in perception of accident or intentional action, no further discussion of these variables is provided. Two variables also asked the extent to which perceivers were and “most people would be” disgusted by Mr Clark's actions. These variables were meant to index annoyance rather than bodily disgust (e.g., Nabi, 2002) and were strongly correlated with anger ratings, but were dropped because our interest was primarily in anger. However, analyses were not substantively different from those reported when using a composite “negative moral emotion” variable (i.e., anger and disgust, combined) rather than the reported anger variable.

⁶Because the effects of perceived awareness were unexpectedly affected by the manipulation of desire, ANCOVAs (analysis of covariances) examining the effects of awareness while controlling for perceived desire were also conducted. The significant effects of awareness on anger and negligence remained significant controlling for perceived desire (Fs > 6.93, ps < .01).

⁷The identical effect of awareness and desire on negligence, while surprising, is simply coincidental.

⁸In addition to the variables presented here, three items measured whether the agent intentionally acted in such a way that harm would come to the victim (e.g., “Ben's actions were intentionally done in order to hurt Adam.”). As in Experiment 1, inclusion of these variables was simply to insure that ratings of intentionality were not unreasonably high in the presence of awareness, desire, or both. Neither awareness nor desire significantly affected intentionality ratings, and the highest rated intentionality (when awareness and desire were both present) was low (M = 3.12, SD = 2.09, using the same 9-point scale as other items).

⁹Although these items were not strongly correlated—most likely because the manipulation and manipulation check of awareness involved two yoked actions that were independently assessed—results from analyses using the items independently did not differ substantively from analyses using the composite item and led to the same conclusions.

¹⁰The correlation between own and others’ anger was high (r = .82) and results were not substantively different using either variable independently. Thus, we combined all items to form our measure of anger.

¹¹A marginally significant interaction between awareness and desire on perceived desire emerged because the effect of desire was stronger when awareness was absent (no desire M = 2.12, desire M = 5.49, d = 1.86) rather than present (no desire M = 2.22, desire M = 4.54, d = 1.28). The simple main effect of desire was significant at both levels of awareness (ps < .001).

¹²Two additional path analyses were conducted, estimating the same model described here (see Figure 2), but separately testing the simple effect of (a) awareness with desire absent and (b) desire with awareness absent as exogenous variables. The model using only awareness fit the data adequately well, χ²(5, N = 74) = 8.18, p = .15, CFI = .97, RMSEA = .09, p-close = .23, ps < .005 for all direct and indirect paths. The model using desire also fit the data well χ²(5, N = 67) = 6.61, p = .25, CFI = 0.98, RMSEA = .07, p-close = .34, ps < .002 for all direct and indirect paths except the direct paths from anger to restitution (p = .07) and the indirect path from desire, through anger, to restitution (i.e., not additionally mediated by negligence, p = .05).