Skip to main content
NCBI home page
NIHPA Author Manuscripts logoLink to NIHPA Author Manuscripts
. Author manuscript; available in PMC: 2020 Dec 29.
Published in final edited form as: Cortex. 2019 Apr 11;119:33–45. doi: 10.1016/j.cortex.2019.03.024

Soft on crime: Patients with ventromedial prefrontal cortex damage allocate reduced third-party punishment to violent criminals

Erik W Asp 1,2,3,*,†, James T Gullickson 4,†, Kelsey A Warner 1,5, Timothy R Koscik 6, Natalie L Denburg 1, Daniel Tranel 1,7
PMCID: PMC7771036  NIHMSID: NIHMS1653672  PMID: 31071555

Abstract

The human impulse to punish those who have unjustly harmed others (i.e., third-party punishment) is critical for stable, cooperative societies. Punishment selection is influenced by both harm outcome and the intent of the moral agent (i.e., the offender’s knowledge of wrongdoing and desire that the prohibited consequence occur). We allocate severe punishments to those who commit violent crimes and milder punishments to those who commit non-violent crimes; and we allocate severe punishments to criminals who have malicious intent and milder punishments to criminals who lack malicious intent. Prior research has indicated that aversive, emotional responses of third-party judges may influence punishment allocation, as increased negative emotion correlates with more punitive punishments. Here, we show that patients with damage to the ventromedial prefrontal cortex (vmPFC; a region necessary for the normal generation of emotion), compared to other neurological patients and healthy adult participants, allocate more lenient third-party punishment to criminals who commit emotionally-evocative, violent crimes. By contrast, patients with vmPFC damage did not differ from comparison participants on punishment allocation for non-emotional, non-violent crimes. These results demonstrate the necessity of the vmPFC for the integration of emotion into third-party punishment decisions, and indicate that negative emotion influences third-party punishment allocation particularly for scenarios involving physical harm to another.

Keywords: moral judgment, prefrontal cortex, harm, third-party punishment, social cognition

1. Introduction

“Passion…is the soul of punishment.”

Emile Durkheim (1893/2003, p. 193)

Stable societies depend upon the enforcement of social norms and the punishment of norm-breakers by impartial third-parties (Fehr & Fischbacher, 2004), a distinctively human process (e.g., Riedl, Jensen, Call, & Tomasello, 2012) referred to as third-party punishment (TPP). While it may seem desirable for these impartial third-parties to judge with a disinterested, calculated rationality—untainted by Durkheimian passion—numerous empirical studies of the mechanisms of TPP have implicated affective processes in punishment allocation (e.g., Fehr & Gachter, 2002). Of central interest in recent psychological and neuroscientific investigations is whether emotions indeed play a causal role in TPP—directly impacting punishment—and, by extension, how putative emotion-related neural structures contribute to TPP.

Individuals tend to experience an intuitive emotional reaction to perceived injustices towards others (Haidt, 2003). Witnessing violence, especially unjustified violence, elicits negative emotional responses with marked psychophysiological correlates (Koukounas & McCabe, 2001; Samson & Potter, 2016). In an experimental TPP paradigm, emotions elicited by the description of the crime can influence subsequent decision making, serving as a heuristic to third parties for how to punish (Buckholtz & Marois, 2012; Darley, 2009; Feigenson & Park, 2006). Behavioral evidence is congruent with this perspective as higher self-reported affective arousal in third-party judges is associated with increased punishment (Bellucci et al., 2016; Buckholtz et al., 2008; Krueger & Hoffman, 2016). Moreover, studies that manipulate third-parties’ emotions produce altered TPP. For example, increasing anger and guilt by altering situational parameters or the inclusion of emotionally-evocative gruesome evidence can sway third parties to be more punitive towards criminals (Bright & Goodman-Delahunty, 2006; Goldberg, Lerner, & Tetlock, 1999; Nelissen & Zeelenberg, 2009; Salerno & Bottoms, 2009; Treadway et al., 2014). Moreover, compelling third parties to consider mitigating circumstances of a crime (as a way to increase empathy for the criminal) can reduce punishment severity (Haegerich & Bottoms, 2000; Yamada et al., 2012). This perspective suggests emotional responses of third-party judges influence punishment allocation following the presentation of crime parameters (such as criminal intent and graphic detail).

Neuroimaging studies consistently show both putative emotion-related and mentalizing-related brain areas activated during TPP. The salience network (anterior insula, amygdala, and the dorsal anterior cingulate cortex) has been implicated in the generation of the emotional experience during TPP (Krueger & Hoffman, 2016). TPP-induced amygdala activity is thought to reflect negative affective arousal which codes harm to guide punishment selection (Buckholtz et al., 2008; Treadway et al., 2014). Ventromedial prefrontal cortex (vmPFC) activity has been associated with the decision to punish crimes and the evaluation of overall harm caused by the crime as activity correlated with punishment magnitude and third-party arousal ratings (Buckholtz et al., 2008; Ginther et al., 2016). Given its putative role in mentalizing, default mode network (medial prefrontal cortex, temporal-parietal junction, and posterior cingulate) activity has been implicated in mental state integration during TPP (e.g., criminal intent; Bellucci et al., 2016; Ginther et al., 2016; Waytz & Mitchell, 2011; Young, Cushman, Hauser, & Saxe, 2007). Moreover, punishment selection has been associated with the central executive network and dorsolateral prefrontal cortex activity (Buckholtz et al., 2008; Ginther et al., 2016; Krueger & Hoffman, 2016). However, TPP neuroimaging studies do not adjudicate whether TPP-associated activations produce or are themselves produced from TPP. And of course, behavioral studies of healthy individuals do not address the neural basis of TPP. Thus, there remains a critical gap in the literature relating TPP, emotion, and the brain.

Investigating TPP in patients with vmPFC damage may provide a decisive test. The vmPFC has reciprocal connections with several regions that putatively mediate emotion including the amygdala, basal forebrain, hypothalamus, and brainstem areas (Neafsey, 1990; Öngür & Price, 2000; Petrides & Pandya, 2002). Neurons within the vmPFC encode expected value, reward outcome, and the emotional value of stimuli (Grabenhorst & Rolls, 2011; Rolls, 2000), while vmPFC activations are evoked during negative emotional regulation (Etkin, Egner, & Kalisch, 2011). Human lesion studies of patients with damage to the vmPFC note emotional dysfunction and blunting following damage (Anderson, Barrash, Bechara, & Tranel, 2006), deficiencies in psychophysiological responses to emotionally-evocative stimuli (Damasio, Tranel, & Damasio, 1990), and deficits in social emotions (such as empathy and guilt; e.g, Beadle, Paradiso, & Tranel, 2018; Krajbich, Adolphs, Tranel, Denburg, & Camerer, 2009).

The first direct examination of TPP in human patients with brain lesions implicated the right vmPFC as critical for normative TPP. Glass and colleagues (2016) conducted a whole-brain voxel-based lesion-symptom mapping study which found that patients with right vmPFC damage (as well as patients with damage to other regions including dorsomedial prefrontal cortex, dorsolateral prefrontal cortex, and the right intraparietal sulcus) produced atypical TPP rankings relative to comparisons. However, due to the methodological constraint of using a rank-order TPP task (where participants needed to rank-order crimes to reflect the relative degree of punishment the moral agent deserved), a clear directional effect (increased or decreased punishment) could not be determined for these patients.

The necessity of the vmPFC for normative moral judgment also informs its potential role in TPP. Models of TPP suggest that prior to punishment selection, the third party must form a moral evaluation based on overall harm and intent of the causal agent (e.g., Buckholtz & Marois, 2012). Thus, TPP adds a behavioral component (i.e., punishment allocation) to the mere appraisal component of classical moral judgment paradigms. The distinguishing feature of this decision-making process is the assessment and selection of a punishment that is appropriate to the relative moral blameworthiness of a criminal and to the relative severity of the criminal offense (Robinson, 1997). Empirical studies investigating reasoning during TPP have demonstrated that individuals punish in direct proportion to the moral wrongfulness of an offender’s action (“just deserts” motivation; Alter et al., 2007; Carlsmith et al., 2002). That is, individuals in a TPP scenario use their moral judgment of how wrongful the offense feels to them as a heuristic for an appropriate criminal sentence (Alter et al., 2007).

Neuropsychological research robustly demonstrates the critical role of the vmPFC in the formation of normative moral judgments. In high-conflict moral dilemmas, where an aversive action is paired with a utilitarian outcome (e.g., the choice to smother your crying baby to save you and a group of villagers from enemy soldiers), patients with vmPFC lesions disproportionately endorse the utilitarian option (Ciaramelli, Muccioli, Ladavas, & Pellegrino, 2007; Koenigs et al., 2007). Importantly, patients with vmPFC lesions have normative decisions on low-conflict moral and non-moral dilemmas (Koenigs et al., 2007), suggesting basic moral rules and decision-making abilities are intact. Recent research has shown that the utilitarian choice in these dilemmas is driven by reduced harm aversion (Cushman, Gray, Gaffey, & Mendes, 2012; Patil, 2015). Taken together, this body of research suggests that vmPFC damage impairs negative emotion induction following the presentation of emotionally-salient moral dilemmas. Theoretically, this vmPFC-lesion-derived emotional vacancy then produces increased utilitarian decisions. If this hypothesis is correct, then vmPFC damage should impair negative emotion induction for other emotionally-salient moral scenarios, including specific TPP scenarios where emotion influences the punishment decision. Since increased negative emotion results in more punitive third-party punishments in healthy individuals, a lack of induced negative emotion in patients with vmPFC lesions should produce more lenient third-party punishments relative to comparison participants. This is the central hypothesis this study sought to investigate.

Subsequent moral judgment research has provided evidence consistent with the notion that patients with vmPFC lesions do not experience the negative emotions that normally arise from understanding than one person intended to harm another (Young et al., 2010). Two independent studies found that patients with vmPFC lesions rated a failed attempt to harm another person as more morally permissible than comparison participants (Ciaramelli, Braghittoni, & Pellegrino, 2012; Young et al., 2010). Both studies indicate that the vmPFC is necessary for incorporating malicious intent of moral agents to outcome information for normative moral judgments. However, there is a discrepancy between these studies for situations where the moral agent accidentally harms another. Here, a lack of malicious intent must be incorporated into outcome information for normative moral judgments. While Young and colleagues (2010) showed that patients with vmPFC lesions were normative for accidental harm moral judgments, Ciaramelli and colleagues (2012) found patients with vmPFC lesions rated accidental harms as less permissible than comparison participants. Given abnormal moral judgments for patients with vmPFC lesions in both attempted and accidental harm scenarios, Ciaramelli and colleagues (2012) suggested the mechanistic explanation of a general deficit attributing intentions to moral agents. By contrast, Young and colleagues (2010) proposed impaired negative emotion induction following malicious intent understanding to explain their findings. Therefore, we sought to clarify this discrepancy with a distinct task as the secondary aim of this study.

If the principle of moral proportionality (Carlsmith et al., 2002), i.e., a punishment must be proportional to the blameworthiness of the offense, is preserved in patients with vmPFC lesions, then an examination of TPP can inform this moral judgment discrepancy. Given the results of Ciaramelli and colleagues (2012), in a TPP scenario, patients with vmPFC lesions should be more punitive in situations where a harm to another is committed but malicious intent is reduced or absent. One method to reduce malicious intent (and decrease TPP) is to include descriptions of situational constraints (e.g., Yamada et al., 2012), or contextual rationale, for why a rational moral agent may be compelled to commit a crime (or is not a rational moral agent). For example, sentencing for a robbery is reduced with the contextual rationale that the moral agent is desperate to feed his starving children. Thus, third-party punishments by patients with vmPFC lesions for individuals who have committed crimes with high situational forces (i.e., where the environmental context reduces the malicious intent of the criminal) produce divergent predictions. In opposition to the punitive TPP predicted by the Ciaramelli et al. (2012) data, data from Young et al. (2010) would suggest patients with vmPFC lesions should have normative TPPs in scenarios with constrained malicious intent of criminals.

Although the neuropsychological moral judgment research indicates that the emotional vacancy of patients with vmPFC lesions should produce punishment leniency, two additional lines of evidence may suggest vmPFC-lesion-related TPP alternative outcomes: 1) patients with vmPFC lesions actually show increased punitive actions toward others when given unfair treatment in second-party punishment scenarios (Koenigs & Tranel, 2007), and 2) patients with vmPFC lesions are high in authoritarianism (Asp, Ramchandran, & Tranel, 2012), which is strongly correlated with third-party punishment allocation in healthy individuals (Altemeyer, 1996). Thus, it remains an open question as to how vmPFC damage may affect TPP.

In the current investigation, we had two primary aims. First, we sought to determine whether the vmPFC plays a causal role in TPP; and second, we sought to determine if the vmPFC is necessary for normative TPPs for crimes with reduced malicious intent. Using the human lesion method and a novel TPP task, we hypothesized that vmPFC damage would be associated with decreased negative emotion induction and specifically reduce punishments for emotionally-evocative crimes, such as those with a high level of violence. Moreover, we hypothesized that patients with vmPFC lesions would be normative in their punishment for non-emotionally-evocative, i.e. non-violent, crimes. Finally, we hypothesized patients with vmPFC lesions would be normative in their TPPs for crimes with constrained malicious intent consistent with Young et al. (2010).

2. Methods and Materials

2.1. Participants

Drawing from the Neurological Patient Registry of the Division of Neuropsychology and Cognitive Neuroscience at the University of Iowa, we studied 26 individuals with adult-onset, focal brain lesions. The etiologies of the lesions included cerebrovascular disease (n = 15), surgical resection for treatment of a meningioma or seizure control (n = 9), and focal contusions from trauma (n = 2; Table 1). Subsequent to enrollment in the registry, brain damaged patients have been extensively characterized neuropsychologically and neuroanatomically, using standard protocols of the Benton Neuropsychology Laboratory and the Laboratory of Brain Imaging and Cognitive Neuroscience (Tranel, 2007). The human lesion method precludes experimental randomization of our groups. Sample size was limited by the rarity of patients with focal brain lesions; however, we note that we obtained a relatively large sample for this type of study. Ten patients with bilateral damage to the vmPFC were classified into our vmPFC group (Figure 1; Table 1); while 16 patients had lesions outside the prefrontal cortex and were classified into our brain damaged comparison group (BDC; Table 1). Patients with damage to another putative emotion-related brain structure, i.e., the amygdala, were excluded from the analyses as they are the focus of a separate study. All neuropsychological and neuroanatomical data were collected in the chronic phase of recovery, at least 3 months post-lesion onset, and there were no significant differences in time since lesion onset between the lesion groups at the time of testing (chronicity; Table 2). We chose to sample exclusively patients with bilateral vmPFC (rather than unilateral) to minimize any potential contralateral compensation effects following damage (Price & Friston, 2002). However, this sampling created a mean difference in lesion size between the lesion groups, as the vmPFC group had larger lesions than the BDC group, t(11.1) = −2.80, p = .017 (Table 1). Several exploratory post-hoc analyses indicated that lesion size, per se, did not contribute to the TPP results (Supplementary Materials). We also included a healthy adult comparison group (HA; n = 57), which was comprised by individuals of similar age to our lesion groups. HA participants had slightly more years of education than our lesion groups, F(2, 80) = 10.5, p < .001; however, planned contrasts revealed there was no significant difference between the lesion groups, t(80) = 1.1, p = .287 (Table 2). Nevertheless, years of education was used as a covariate in the analyses. No patient had significant language impairments (defined as 2 standard deviations below the mean on the Boston Naming Test or the Token Test), significant reading deficits (defined as 2 standard deviations below the mean on the Iowa Chapman Reading Test), significant memory deficits (defined as 2 standard deviations below the mean on the Auditory Verbal Learning Test delayed recall), or significant visuoperceptual impairments (defined as 2 standard deviations below the mean on the Facial Recognition Test; Lezak, Howieson, Bigler, & Tranel, 2012). There were no significant differences between the vmPFC and BDC groups on various neuropsychological measures, including IQ, reading ability, memory, and executive function (Table 3). Patients with vmPFC lesions had impaired autonomic responses to emotionally charged pictures, as well as conspicuous deficits in empathy and guilt (Table 1). All participants were free from intellectual disability, learning disabilities, psychiatric disease, substance abuse, and dementia. Participants gave informed consent approved by the Institutional Review Board of the University of Iowa.

Table 1. Neuroanatomical, etiological, emotional, and social data for the lesion groups.

AVM, arteriovenous malformations; TLE, temporal lobe epilepsy. Numbers in parentheses in the Lesion Location column represent ratings of the percentage of damage to the vmPFC proper (to the left and right vmPFC, respectively). The rating used a four-point scale denoting damage to the vmPFC, where 0 = no damage, 1 = 1-25% damaged, 2 = 26-75% damaged, and 3 = 76%-100% damaged. Lesion Size is given in cubic centimeters. Autonomic Responses, skin conductance responses to emotionally charged socially significant stimuli (e.g., pictures of social disasters, mutilations, nudes), using methods previously described (Damasio et al., 1990). A clinical neuropsychologist rated each vmPFC patient’s demonstrated capacity for empathy and guilt in his or her personal life. The rating used a four-point scale denoting severity of impairment, where 0 = normal, 1 = mild, 2 = moderate, 3 = severe. Ratings were based on data derived from spouse or family member reports in the Iowa Scales of Personality Change (Barrash et al., 2011) and from clinical interview data.

Patient # Etiology Lesion Location (% L vmPFC, % R vmPFC) Lesion Size Autonomic Responses Empathy Guilt
vmPFC group
318 meningioma resection bilateral vmPFC (2, 3) 141.53 Impaired 3 3
1983 hemorrhagic stroke bilateral vmPFC (3, 3) 93.76 Impaired 3 3
2352 hemorrhagic stroke bilateral vmPFC (1, 1) 17.45 Impaired 2 1
2391 meningioma resection bilateral vmPFC (3, 3) 88.82 Impaired 3 3
2577 hemorrhagic stroke bilateral vmPFC (2, 2) 61.25 Impaired 3 3
2855 meningioma resection bilateral vmPFC (1, 1) 57.57 Impaired 3 3
3350 meningioma resection bilateral vmPFC (3, 3) 61.01 Impaired 3 3
3383 hemorrhagic stroke bilateral vmPFC (2, 1) 38.12 Impaired 3 3
3534 meningioma resection bilateral vmPFC (3, 3) 53.51 Impaired 3 3
3591 focal contusion bilateral vmPFC (3, 3) 200.57 Impaired 3 3
BDC group
1730 ischemic stroke left parietal (0, 0) 46.61 Unimpaired 0 0
2328 ischemic stroke right posterior frontal/parietal (0, 0) 74.68 Unimpaired 1 1
2355 AVM resection right parietal (0, 0) 44.18 Unimpaired 0 0
2710 hemorrhagic stroke left parietal (0, 0) 5.58 Unimpaired 0 0
2764 hemorrhagic stroke right temporal (0, 0) 20.80 Unimpaired 2 1
2894 focal contusion left temporal (0, 0) 11.87 Unimpaired 0 0
2994 hemorrhagic stroke left parietal (0, 0) 62.42 Unimpaired 0 0
3138 ischemic stroke left parietal (0, 0) 31.35 Unimpaired 0 0
3202 ischemic stroke left posterior frontal (0, 0) 15.53 Unimpaired 0 0
3353 ischemic stroke left posterior frontal (0, 0) 0.97 Unimpaired 0 0
3378 TLE resection left temporal (0, 0) 35.03 Unimpaired 0 0
3386 TLE resection right temporal (0, 0) 32.81 Unimpaired 0 0
3464 ischemic stroke left posterior frontal (0, 0) 10.23 Unimpaired 0 0
3567 hemorrhagic stroke right parietal (0, 0) 57.47 Unimpaired 0 1
3604 ischemic stroke right parietal (0, 0) 14.41 Unimpaired 1 1
3605* meningioma resection left parietal (0, 0) NA Unimpaired 0 0
Open in a new tab
*

Lesion size data were not available for patient 3605.

Figure 1. Lesion overlap of patients with ventromedial prefrontal cortex lesions.

Figure 1

Open in a new tab

The overlap map shows the lesions of patients with vmPFC damage in anterior/mesial views and coronal slice (a-f, with the right hemisphere on the left in the coronal views). The color bar indicates the number of overlapping lesions per voxel.

Table 2. Demographic data.

Age, education, and chronicity are presented in years. There were no significant differences between the groups for age, chronicity, and sex ratios.

vmPFC BDC HA
Number 10 16 57
Age (SD) 65.3 (7.3) 60.7 (13.2) 55.5 (19.0)
Education (SD)* 13.2 (1.9) 14.2 (2.5) 16.2 (2.3)
Sex 5 M; 5 F 7 M; 9 F 28 M; 29 F
Chronicity (SD) 11.8 (10.0) 10.1 (6.6) NA
Open in a new tab
*

HA participants had significantly more years of education than vmPFC and BDC patients.

Table 3. Neuropsychological data for the lesion groups.

WAIS, Wechsler Adult Intelligence Scale-III scores (FSIQ, Full-Scale IQ). WRAT, Wide Range Achievement Test-4 scores (Read, Reading Standard Score). AVLT, Auditory Verbal Learning Test scores (an index of memory function at 30 min), # recalled/15. CFT, Complex Figure Test recall scores (an index of memory function at 30 min), # recalled/30. TMT, Trail Making Test Part B scores, an index of divided attention and multi-tasking (time to completion, in seconds). There were no significant differences between the groups for any of the neuropsychological indices.

vmPFC BDC
WAIS -FSIQ (SD)* 107.7 (17.0) 102.0 (11.0)
WRAT -Read (SD)** 101.4 (11.6) 98.7 (14.6)
AVLT −30 min recall (SD) 9.9 (3.0) 9.3 (3.3)
CFT −30 min recall (SD) 19.4 (7.8) 17.4 (5.7)
TMT -Part B (SD) 77.5 (37.1) 97.4 (51.1)
Open in a new tab
*

WAIS-IV-FSIQ scores were used for 2 patients with vmPFC damage and 2 patients in the BDC group; 1 patient in the BDC group did not take the WAIS.

**

WRAT-R Read scores were used for 1 patient with vmPFC damage; 1 patient in the BDC group did not take the WRAT.

2.2. Stimuli and procedure

All tasks were performed in isolation in a testing room at the University of Iowa Hospitals and Clinics. Participants read 32 crime vignettes (4-7 sentences in length) in a novel TPP task presented on a computer screen (see Supplementary Materials for complete stimuli set). Each vignette had a unique male criminal as the moral agent and gave a brief account of the agent committing an ostensibly real crime. Participants received the instructions verbally and visually on the computer screen: This task will require you to make a variety of judgments. You will be presented with 32 crime reports. Your job is to read the reports and give an appropriate prison sentence for the crime. Participants then read and punished a criminal in a practice vignette that was not analyzed (Supplementary Materials). Participants gave third-party punishments in the form of prison sentences from a numerical button press, denoted as such: 1 (<1 year of prison time); 2 (1-2 years of prison time); 3 (3-5 years of prison time); 4 (6-10 years of prison time); 5 (11-20 years of prison time); 6 (21-30 years of prison time); 7 (31-40 years of prison time); 8 (41-50 years of prison time); and 9 (>50 years of prison time). Participants had unlimited time to read each vignette and make a punishment decision. A specific crime type (e.g., stabbing) was presented twice to the participants (one in the 1st half of the set presentation order and the other in the second half). Thus, each vignette was paired with another via the exact nature of the crime (totaling 16 crime vignette pairs). One item in a pair was designed with high situational forces (for constrained malicious intent) while the other item in the pair was designed with low situational forces (for high malicious intent). Crime vignettes with high situational forces offered contextual rationale for the crime occurrence, such as the desperate need to feed one’s children or an inability to determine right from wrong due to psychiatric disease (see Murder and Robbery-Charles vignette and Mutilation-Dave vignette, respectively, in the Supplementary Materials). Crime vignettes with low situational forces gave biographical details of criminals but did not offer contextual rationale for the crime occurrence. Crimes that explicitly described a victim injury were labeled as violent (10 crimes total), and those that did not were labeled non-violent (6 crimes total). Our categorization of violent crime vignettes contained crimes (i.e., mutilation, killing children, school shooting, rape, murder and robbery, shooting, stabbing, kidnapping, robbery with injuries, and assault) that were objectively more serious with longer criminal justice allotted incarceration periods than non-violent crime vignettes (i.e., robbery without injures, evading police, dealing drugs, prostitution, embezzlement, and tax fraud). Crime vignettes were presented in a pseudorandom order (see Supplementary Materials for exact ordering).

Following the TPP task, participants performed additional judgment vignette tasks, where they needed to read vignettes and rate unique protagonists on several attributes. The Healthy Lifestyle Judgment Vignette Task (HLJVT) had the participants rate 10 vignettes on healthy lifestyle (from 1-Not at all healthy to 9-Extremely healthy), and the Moral Judgment Vignette Task (MJVT) had the participants rate 10 vignettes on morality (from 1-Not at all moral to 9-Extremely moral; see Supplementary Materials for complete stimuli set). The moral judgment vignettes ranged from “moral” protagonists (e.g., volunteering at local soup kitchens, see Moral-Victor) to “immoral” protagonists (e.g., bullying others, see Moral-Sam). However, “immoral” actions in the moral judgment vignettes were decidedly less malevolent (and putatively less emotionally upsetting) than the violent crimes of the crime vignette task. These vignette tasks were designed to examine potential group differences of protagonist ratings in vignettes in the non-moral domain (i.e., Healthy Lifestyle Judgment Vignette Task) and in the moral non-emotional domain (i.e., Moral Judgment Vignette Task).

Finally, a paper-and-pencil authoritarianism scale (AS) was completed by participants following the computer tasks. The AS has been used extensively and is psychometrically well established (Altemeyer & Hunsberger, 1992). It defines authoritarianism as the covariation of 3 attitudinal clusters: submission, aggression, and conventionalism. Several studies have shown that the AS correlates with TPP in healthy individuals (e.g., Altemeyer, 1996).

To obtain objective data on level of violence and negative emotional impact of the crime vignettes in the TPP task on a naïve reader, we conducted a separate pilot study with 56 healthy adults who rated each vignette on how violent each crime is (from 1-Not violent at all to 4-Extremely violent), how emotionally upsetting the crime was (from 1-Not emotionally upsetting at all to 4-Extremely emotionally upsetting), what level of empathy was felt for the criminal (from 1-No empathy at all to 4-Strong empathy), how common in one’s community the crime was (from 1-Not common at all to 4-Extremely common), and for the crimes with identifiable victims, what level of empathy was felt for the victim (from 1-No empathy at all to 4-Strong empathy). Empathy for the criminal was included as a malicious intent manipulation proxy; empathy for the victim was included as a crime severity manipulation proxy; and the commonality rating was included to determine the perception of the crime occurrence rate for the pilot participants. See Supplementary Materials for average ratings for each vignette. These healthy adult raters were not part of the 83 participants of the study and did not confer punishments to the criminals.

2.3. Neuroanatomical analysis

The neuroanatomical analysis of the lesion patients was based on magnetic resonance imaging data for 18 patients and on computerized tomography imaging data for 8 patients. Brain image data were examined in native space and a patient was categorized into the vmPFC group if there was evidence of damage to the vmPFC proper. We define the vmPFC as having the following bounds: the dorsal border, which extends on a horizontal plane at the anterior-most point of the genu of the corpus callosum to the anterior-most aspect of the frontal pole; and two lateral borders, which extend on sagittal planes at the pits of the orbital fossae. This region encompasses the medial orbital gyri, the gyrus rectus, the subcallosal area, the ventral anterior cingulate cortex, and structures directly inferior to the frontal pole. This definition has been used previously (e.g., Koenigs et al., 2007; Taber-Thomas et al., 2014). Using Brainvox (Frank, Damasio, & Grabowski, 1997), each patient’s lesion was reconstructed in three dimensions. The lesion contour for each patient in was manually warped into a normal template brain using the MAP-3 method. The overlap of lesions in this volume, calculated by the sum of n lesions overlapping at any single voxel, is color-coded in Figure 1. As Figure 1 shows, the greatest overlap of patients with vmPFC lesions is in the mesial orbital region, especially the anterior half of the gyrus rectus. No BDC lesions included the vmPFC or the amygdala (Table 1).

2.4. Statistical analyses

For all tests, α = 0.05. To address our hypotheses, we modeled TPP with two linear mixed-effects models using R (version 3.5.1); effect significance for linear mixed-effects models were examined using Type III Wald Chi-squared tests. In the first linear mixed-effects model predicting TPP, we entered group, violence of crime vignette (rating taken from pilot study), empathy for the criminal of crime vignette (rating taken from pilot study), and age as fixed effects (Model 1-Pilot ratings). Intercepts and slopes for participants were added as random effects. Sex, years of education, healthy lifestyle ratings (from the HLJVT), moral judgment ratings (from the MJVT), and authoritarianism were entered as covariates due to their potential relationship with TPP. Manipulation checks included the main effect of violence of crime vignette on TPP and the main effect of empathy for the criminal of crime vignette on TPP (i.e., a proxy for malicious intent manipulation); and the effects of interest included the violence of crime vignette by group interaction.

We chose to follow this analysis up with a second linear mixed-effects model which utilized a categorical 2 (crime type: violent vs. non-violent) by 2 (crime context: high situational factors vs. low situational factors) design. Here group, crime type, and crime context were entered as fixed effects (Model 2-Categorical vignettes). Again, intercepts and slopes for participants were added as random effects. Manipulation checks included the main effect of crime type on TPP and the main effect of crime context on TPP; and the effects of interest included the crime type by group interaction and the crime context by group interaction.

2.5. Data availability

The data that support the findings of this study are available on request from the corresponding author. The data are not publicly available due to potential loss of confidentiality of neurological patients.

3. Results

There were no group differences on the Healthy Lifestyle Judgment Vignette Task, F(2, 80) = 1.84, p = .167, and the Moral Judgment Vignette Task, F(2, 80) = .21, p = .808, indicating the vmPFC group was normative in judging protagonists in the non-moral and non-emotionally moral domains. Our pilot data with healthy adult raters showed a strong correlation between the violence and emotionally upsetting ratings (r = .91, p < .001), suggesting more self-reported negative emotional reactions were associated with the assessment of objectively more violent crimes. As Model 1-Pilot ratings (Table 4) shows TPP was not predicted by group, χ2(2) = 3.02, p = .220, nor age, χ2(1) = 1.06, p = .302, but TPP was predicted by violence of crime vignette, χ2(1) = 1256.73, p = 2.20 × 10−16, empathy for the criminal of crime vignette, χ2(1) = 60.77, p = 6.41 × 10−15, and the interaction of violence of the crime vignette and group, χ2(2) = 13.93, p = 9.45 × 10−4. Interaction of empathy for the criminal of crime vignette and group did not predict TPP, χ2(2) = 0.69, p = .708, nor was there an age by group interaction, χ2(2) = 4.06, p = .131. There was a slight trend for patients with vmPFC lesions to punish more harshly with age (see Table 4), but this is controlled for in the main effect. Covariates sex, χ2(1) = 1.82, p = .177, education, χ2(1) = 5.02, p = .479, healthy lifestyle ratings, χ2(1) = .13, p = .723, moral judgment ratings, χ2(1) = 2.54, p = .111, and authoritarianism, χ2(1) = 1.16, p = .282 did not predict TPP. Thus, our paradigm manipulation successfully elicited increased punishment for more violent crime vignettes relative to less violent crime vignettes, and increased punishment for vignettes with higher empathy for the criminal relative to lower empathy for the criminal. Pilot ratings of empathy for the criminal were different between our nominal low and high situational factor crime vignettes, t(30) = −10.68, p < .001. Indeed, as every high situational (low malicious intent) crime vignette had a greater empathy for the criminal rating than its low situational (high malicious intent) crime vignette pair (see Supplementary Materials), the empathy for the criminal generally indexed malicious intent. As hypothesized, the interaction of violence of the crime vignette and group was driven by the vmPFC group who were more lenient in their punishments on violent crimes, t(2567) = 3.70, p = 2.23 × 10−4, whereas the BDC group was not t(2567) = .16, p = .871, relative to the HA group. However, vmPFC group TPP did not differ from the HA and BDC groups on non-violent crimes (Figure 2, Table 4).

Table 4. TPP linear mixed effects model results.

Coefficients for both linear mixed models of TPP was created using R. The group by violence of crime vignette effect was the key test for differential crime type TPP by group and was statistically significant in both models. A crime context by group interaction failed to emerge as significant in Model 2. SE, standard error; DF, degrees of freedom.

Model 1: Pilot ratings
Effects Estimate SE DF t value p value
Intercept 0.76 1.09 76.93 0.70 0.487
Group: BDC 0.31 1.02 104.10 0.31 0.761
Group: vmPFC −3.75 2.24 80.05 −1.68 0.098
Violence of crime vignette 1.61 0.05 2567.00 35.44 2.00 × 10−16
Empathy for the criminal of crime vignette −4.89 0.06 2567.00 −7.80 9.24 × 10−15
Age −5.2 × 10−3 5.0 × 10−3 72.00 −1.03 0.310
Sex: Male 0.22 0.16 72.00 1.35 0.182
Education 0.03 0.04 72.00 0.71 0.481
Healthy lifestyle rating 0.05 0.13 72.00 0.36 0.724
Moral judgment rating −0.18 0.11 72.00 −1.59 0.116
Authoritarianism 0.13 0.12 72.00 1.08 0.286
Group: BDC × Violence of crime vignette −0.02 0.10 2567.00 −0.16 0.871
Group: vmPFC × Violence of crime vignette −0.44 0.12 2567.00 −3.70 2.23 × 10−4
Group: BDC × Empathy for the criminal of crime vignette 0.05 0.13 2567.00 0.39 0.695
Group: vmPFC × Empathy for the criminal of crime vignette 0.13 0.16 2567.00 0.79 0.429
Group: BDC × Age −9.9 × 10−5 0.02 72.00 −0.01 0.995
Group: vmPFC × Age 0.07 0.03 72.00 2.01 0.049
Model 2: Categorical vignettes
Effects Estimate SE DF t value p value
Intercept 3.20 0.13 244.70 25.17 2.00 × 10−16
Group: BDC 0.28 0.27 244.70 1.03 0.303
Group: vmPFC 0.06 0.33 244.70 0.17 0.867
Violent category of vignette 2.70 0.11 2567.00 25.49 2.00 × 10−16
High situational factor category of vignette −0.96 0.10 2567.00 −9.34 2.00 × 10−16
Group: BDC × Violence category of vignette −5.3 × 10−4 0.23 2567.00 −2.0 × 10−3 0.998
Group: vmPFC × Violence category of vignette −0.93 0.27 2567.00 −3.39 0.001
Group: BDC × High situational factor category of vignette 0.14 0.22 2567.00 0.64 0.523
Group: vmPFC × High situational factor category of vignette 0.29 0.27 2567.00 1.08 0.278
Open in a new tab

Figure 2. Predicted TPP by violence of crime vignette.

Figure 2

Open in a new tab

Group differences in predicted third-party punishment by violence of crime vignettes ratings. Y-axis scale is 1-9 punishment allocation (see Methods for years in prison labels). X-axis scale is 1 (Not at all violent) to 4 (Extremely violent) pilot data ratings of the crime vignettes. All 3 groups increase punishment with increasing violence. However, the vmPFC group is more lenient to criminals as crime vignette violence is increased. Color surround indicates 95% confidence limits.

In a subsequent linear mixed-effect model, we sought to directly address our secondary hypothesis that patients with vmPFC lesions would be normative when given high situational factors (and low malicious intent) crime vignettes. Although there was a main effect of empathy for the criminal of crime vignette (proxy for malicious intent) but no interaction between empathy for the criminal of crime vignette and group as demonstrated by Model 1, Model 2 examined categorical differences of our vignettes: crime type (violent vs. non-violent) and crime context (high situational forces vs. low situational forces). Model 2-Categorical vignettes showed TPP was not predicted by group, χ2(2) = 1.07, p = .587, but TPP was predicted by crime type, χ2(1) = 659.64, p = 2.20 × 10−16, crime context, χ2(1) = 87.20, p = 2.20 × 10−16, and a group by crime type interaction, χ2(2) = 11.85, p = .003 (Table 4). Our secondary analysis showed that the paradigm manipulation successfully elicited increased punishment for violent crime vignettes relative to non-violent crime vignettes, and high situational (low malicious intent) crime vignettes elicited reduced punishment than low situational (high malicious intent) crime vignettes. However, critically, a group by crime context interaction did not emerge as a predictor for TPP in Model 2, χ2(2) = 1.38, p = .502. Thus, consistent with Young et al. (2010), the vmPFC group was normative in their punishments from crimes with low malicious intent rather than the more punitive punishments predicted from the Ciaramelli et al. (2012) data (Figure 3).1

Figure 3. Predicted TPP by crime context.

Figure 3

Open in a new tab

Group differences in predicted third-party punishment by crime context. Y-axis scale is 1-9 punishment allocation (see Methods for years in prison labels). X-axis is the nominal crime categorization based on either high situational factors (i.e., low malicious intent) or low situational factors (i.e., high malicious intent). While not statistically different than the HA group and the BDC group, the vmPFC group trended toward leniency in both vignette types; a finding inconsistent with moral judgment data suggesting increased punitive punishments in scenarios with low malicious intent. Error bars indicate 95% confidence intervals.

4. Discussion

The central aim of this study was to determine if the vmPFC plays a causal role in third-party punishment. On the premise that negative emotion is critical for influencing punishment selection, and given that patients with vmPFC lesions have deficits in emotional processing, we hypothesized that the vmPFC group would be more lenient in their TPP than comparison groups on more emotional, violent crimes, and that the vmPFC group would be normative in their TPP on non-emotional, non-violent crimes. The results were consistent with these hypotheses as demonstrated by both mixed-effects models: patients with vmPFC lesions were lenient only to criminals that committed violent crimes, but were normative to criminals that committed non-violent crimes. To be sure, patients with vmPFC lesions did punish criminals that committed violent crimes more harshly than criminals that committed non-violent crimes; however, the healthy adult and brain damaged groups punished criminals who committed violent crimes with, on average, 6.15 more years of prison time than the vmPFC group2.

Our secondary aim was to examine TPP in patients with vmPFC lesions when malicious intent was varied. We theorized that, consistent with Ciaramelli et al.’s (2012) study, we would find increased punitive TPP from the vmPFC group in low malicious intent (and high situational factor) crimes; whereas findings from the Young et al.’s (2010) study would predict normative TPP from the vmPFC group in low malicious intent crimes. Consistent with Young et al. (2010), a crime context by group interaction did not emerge as a predictor of Model 2-Crime vignettes, and follow-up analyses revealed a vmPFC group trend toward leniency in both low and high malicious intent crimes (Figure 3, Supplementary Figure 1). This trend is due to the inclusion of violent crimes in each crime context variable, to which patients with vmPFC damage punished at a reduced rate. There was a main effect of crime context on TPP such that low malicious intent (high situational factors) received reduced punishment relative to high malicious intent (low situational factors) crimes; yet there was no crime context by group interaction and there was no evidence that patients with vmPFC damage were more punitive than comparisons in low malicious intent contexts.

However, we note two caveats to this conclusion. First, there is overlap in the patient samples used in the current study and in Young et al. (2010): 6 patients with vmPFC lesions participated in both studies. This could limit external validity. Replications with distinct patients are warranted. Second, the low malicious intent TPP conclusion in patients with vmPFC damage relies on the validity of the preservation of moral proportionality assumption. Since we did not directly measure moral culpability attributions in our participants, it is possible that patients with vmPFC damage would rate low malicious intent vignettes as less permissible (i.e., a moral judgment) than comparisons, in accordance with Ciaramelli et al. (2012), and still produce normative TPP. However, we consider a vmPFC-lesion-derived impairment of moral proportionality unlikely as TPP rankings of the 32 vignettes were generally consistent between the groups. That is, vignettes with high violence and high malicious intent received the harshest punishments and vignettes with low violence and low malicious intent received the most lenient punishments across all of the groups (Supplementary Figure 1). A theoretical decoupling of blameworthiness and punishment in patients with vmPFC damage would be expected to produce conspicuous deviations in TPP ranking relative to comparisons, and we simply did not observe this TPP pattern.

While our and Young et al.’s (2010) data tend to show a leniency pattern in patients with vmPFC lesions, general punishment leniency is an inadequate explanation for our results. That is, the vmPFC group was normative in TPP for criminals that committed non-violent crimes. Moreover, patients with vmPFC lesions have been shown to be more punitive than comparisons in second-party punishment scenarios (Koenigs & Tranel, 2007), in a study that included 5 patients with vmPFC lesions that participated in the present study. Thus, these data suggest that the vmPFC does not mediate punishment per se, but instead, mediates another factor that influences punishment allocation.

Rather, we suggest the pattern of results observed in this study is due to impaired emotional processing in patients with vmPFC lesions. Specifically, the results are consistent with the possibility that when patients with vmPFC lesions encounter a violent crime scenario, they do not experience the aversive, negative emotions that normally arise from understanding that one person has harmed another. Indeed, our pilot results indicated a strong association between crime vignette violence and emotional distress in healthy adults. This vmPFC-lesion-derived lack of emotional response then results in a reduced punishment selection relative to individuals that experience the aversive emotional state. The patients with vmPFC lesions in this study had severe deficits in trait empathy and reduced autonomic responses to socially significant and emotionally competent stimuli (see Table 1). Moreover, patients with damage to the vmPFC tend to show reduced empathetic behavior toward others, i.e., when given a chance to help a conspicuously suffering individual, they do not (Beadle et al., 2018). The mode of aversive emotion elicitation is distinct in second-party punishment scenarios, where the punishing agent is directly harmed by those whom they punish. Clear evidence of harm or unfairness to oneself does elicit an aversive emotional response in patients with vmPFC lesions, i.e., primary emotion induction (Bechara & Damasio, 2005), which results in a punitive retribution decision (Koenigs & Tranel, 2007). However, the vmPFC is a critical neural substrate for secondary emotion induction, i.e., triggering emotions in response to recalled, inferred, abstract, or imagined events (Bechara & Damasio, 2005). We suggest that patients with vmPFC lesions are lenient to criminals that commit violent crimes because they fail to generate a normative emotional response to our abstract and imagined crime vignettes. Via this rationale, the vmPFC group is normative in their TPP to non-violent criminals because an aversive emotional response does not drive punishment allocation in healthy individuals. In this case, we suggest patients with vmPFC lesions use intact social knowledge and moral rules (Saver & Damasio, 1991) to guide normative punishment allocation in the sentence adjudication of criminals that commit non-emotional, non-violent crimes. While a specific deficit in experiencing emotion triggered from abstract and imagined scenarios is the most parsimonious explanation for our results, here we consider two alternative hypotheses for the pattern of TPP we observed.

First, patients with vmPFC lesions may have produced abnormal TPP due to deficits in domain-general cognitive abilities as opposed to social-emotional deficits. This hypothesis appears unlikely for several reasons. Our vmPFC group had preserved intelligence, memory, and executive function (Table 3) and other vmPFC human lesion studies indicate preserved declarative knowledge of social and moral norms (Burgess et al., 2006; Saver & Damasio, 1991). Moreover, the vmPFC group demonstrated an intact ability to judge vignette protagonists in non-moral and non-emotionally moral domains. These patients also selectively gave lenient TPP to criminals who committed violent crimes, but gave normative TPP to criminals who committed non-violent crimes, an asymmetry which suggests it is improbable that a generic cognitive deficit drove the results.

Second, vmPFC group performance may have been produced by a basic theory of mind or mentalizing deficit. That is, lesions to the vmPFC could have resulted in an impairment in attributing intentions across all the vignettes. Again, we find this explanation unlikely as Model 2-Categorical vignettes did not find a crime context by group interaction when we specifically varied malicious intent. Crime vignettes with reduced malicious intent were punished normatively by the vmPFC group indicating an intact ability to incorporate a paucity of malicious intent to outcome information for punishment allocation. This indicates our vmPFC group’s selective TPP deficit on criminals for violent crimes cannot be due to a failure in representing the content of a belief or intention (see also Young et al., 2010). Therefore, while we did not measure mentalizing nor explicit intention understanding during this task, the complete pattern of results suggest a mentalizing deficit did not lead to the pattern of results observed.

We also note the vmPFC group’s TPP leniency to criminals who commit violent crimes in the context of high authoritarianism (Asp et al., 2012). Strong correlations in healthy adults have been reported between authoritarianism and TPP (Altemeyer, 1996; Wylie & Forest, 1992); i.e., individuals high in authoritarianism tend allocate more punitive TPP. However, in the current study, authoritarianism did not emerge as a predictor of TPP in Model 1-Pilot ratings. The data indicate that authoritarianism, per se, does not directly produce punitive TPP; but rather a third factor likely mediates both authoritarianism and TPP. We speculate that this factor is secondary emotion induction as healthy individuals high in authoritarianism report increased levels of disgust and aversion to crime vignettes relative to healthy individuals low in authoritarianism (Altemeyer, 1996). However, more research is needed to fully explore this hypothesis.

To our knowledge, this study is the first demonstration of TPP leniency toward criminals who committed violent crimes in patients with focal ventromedial prefrontal cortex damage. Glass et al. (2016) reported right vmPFC damage produced an atypical ranking on a putative TPP task, where participants needed to rank crime vignettes on the degree of punishment the protagonist deserved. However, this methodology obscures the nature of these group differences both quantitatively (i.e., rankings techniques lose power due to a lack of raw TPP scores) and directionally (i.e., an inability to determine whether patients with vmPFC lesions are more punitive or lenient than comparisons). Moreover, we note two further potential limitations in Glass and colleagues’ (2016) design. First, their rank-order TPP task did not specify the punishment to be given to the criminal. That is, there was a lack of punishment details (e.g., years to be spent in prison) that may have created an additional individual difference variance as participants were unsure what type of punishment is to be allocated. Second, Glass and colleagues’ (2016) used one moral agent name (i.e., John) for all the vignettes. If patients with vmPFC lesions conceptualized the same individual in each of Glass and colleagues’ (2016) vignettes, TPP ranking deviations relative to comparisons could be due to impaired moral updating (Croft et al., 2010). Therefore, with our contrasting TPP methodology, we consider the current study an important and novel extension of Glass and colleagues (2016) work.

Our results indicate the vmPFC is a critical structure when allocating third-party punishment for emotionally-evocative scenarios. This is consistent with neuroimaging models wherein the vmPFC acts as an internetwork hub between the salience network, which putatively generates the aversive experience during TPP, and the default mode network, which putatively integrates intent and harm during TPP (Ginther et al., 2016; Krueger & Hoffman, 2016). However, these data argue against the vmPFC as a direct mediator of criminal intent as patients with vmPFC lesions normatively adjusted for situational variables (i.e., lower criminal intent) from low to high situational TPP vignettes (Figure 3). Rather, in concurrence with Young et al. (2010), we suggest the vmPFC directly mediates secondary emotion induction, which is critical for the aversive emotional state when perceiving that an individual intends to harm another. Patients with vmPFC damage, then, do not experience this emotional state and therefore judge malicious attempts of harm as more permissible. This hypothesis is consistent with evidence that suggests the vmPFC is critical for processing affective facets of another’s mental states (e.g., Shamay-Tsoory & Aharon-Peretz, 2007).

One limitation to this study was a lack of self-reported ratings of violence, emotions, and empathy from the participants. We chose not to assess this due to demand characteristic concerns, but one might expect group differences on emotion and empathy ratings if examined. However, we note that patients with vmPFC damage often demonstrate preserved social and emotional knowledge, in off-line or laboratory-based contexts (e.g., Feinstein, 2012; Saver & Damasio, 1991). Indeed, patients with vmPFC lesions produce normative cognitive and emotional empathetic ratings to questionnaires, but fail to act empathetically in the context of suffering individual and the ability to help (Beadle et al., 2018). This discrepancy may explain why a group by empathy for the criminal interaction did not emerge as a predictor in Model 1, as patients with vmPFC damage could logically reason that high situational factors reduce the moral agent’s culpability in the crime’s occurrence.

There are several other limitations that are important to note. First, the lesion method relies on natural brain damage, which often results in large overlapping regions of damage for a region of interest (such as in our vmPFC group, Figure 1). Since all patients in the vmPFC group had bilateral damage (Table 1), potential TPP laterality differences could not be examined. Future work with more subjects (including patients with unilateral vmPFC damage) will be needed to see if unique regions of the prefrontal cortex serve differential aspects of TPP beyond secondary emotion induction. Second, this study focused on only patients with vmPFC lesions which limits conclusions about broader network involvement. Third, neuroplasticity and recovery of function after lesion may have impacted TPP (Rorden & Karnath, 2004) as our methodology did not allow us to obtain pre-lesion data. Finally, there remains the possibility that real-world third-party punishment decisions would produce different results as the participants understood they were not actually inflicting punishment on the criminals in our vignettes.

The current results suggest that an aversive, negative emotional state does influence TPP in scenarios where one person physically harms another. Thus, TPP is reduced when this negative emotion is not experienced such as in patients with vmPFC lesions. It is well known that affective biases can influence judgment and decision-making (Finucane, Alhakami, Slovic, & Johnson, 2000; Lerner, Li, Valdesolo, & Kassam, 2015). Alteration of affective physiological states can change moral judgment; as alcohol intoxication predicts utilitarian responses on moral dilemmas (Duke & Begue, 2015) and selective serotonin reuptake inhibitors act to enhance harm aversion and lower utilitarian responses (Crockett, Clark, Hauser, & Robbins, 2010). Moreover, in real-world third-party punishment decisions, trial judges’ sentencing can be altered by factors that should have no bearing on a rational punishment allocation; such as current physiological state (e.g., Danziger, Levav, & Avnaim-Pesso, 2011) and the putative emotion level of the trial (Krueger & Hoffman, 2016; Steffensmeier & Britt, 2001). Taken together, this work confirms that an emotionally-motivated, retributive sentiment, that individuals who commit violent crimes deserve punishment drives TPP (Aharoni & Fridlund, 2012; Carlsmith & Darley, 2008), and uniquely demonstrates that the ventromedial prefrontal cortex is a necessary neural structure mediating this aversive experience during third-party punishment.

Supplementary Material

supplementary data

Acknowledgements

We thank Alexandra Fleszar and Matthew Sutterer for help creating the stimuli, and Joel Bruss for assistance with the neuroanatomical figure. We would also like to thank the University of Iowa Libraries for their excellent support of research.

Funding

This work was supported by NIH P50 MH0942581, NIH U01 NS103780, the Kiwanis International Neuroscience Research Foundation, and a generous personal donation by Dr. Jerry Artz.

Footnotes

The authors declare no competing interests.

1

An exploratory post-hoc linear mixed-effects model showed that patients with vmPFC damage did not differ in TPP relative to comparisons in a crime type by crime context by group interaction (see Supplementary Materials).

2

This calculation was made by substituting median, discrete prison year values for each response: 1 (0.5 years of prison time), 2 (1.5 years of prison time), 3 (4 years of prison time), 4 (8 years of prison time), 5 (15.5 years of prison time), 6 (25.5 years of prison time), 7 (35.5 years of prison time), 8 (45.5 years of prison time), and 9 (55.5 years of prison time).

Supplementary Material

Supplementary material is available at Cortex online.

References

  1. Aharoni E, & Fridlund AJ (2012). Punishment without reason: Isolating retribution in lay punishment of criminal offenders. Psychology, Public Policy, and Law, 18, 599–625. [Google Scholar]
  2. Altemeyer B (1996). The Authoritarian Specter. Cambridge, MA: Harvard University Press. [Google Scholar]
  3. Altemeyer B, & Hunsberger B (1992). Authoritarianism, religious fundamentalism, quest, and prejudice. The International Journal for the Psychology of Religion, 2, 113–133. [Google Scholar]
  4. Alter AL, Kernochan J, Darley JM (2007). Transgression wrongfulness outweighs its harmfulness as a determinant of sentence severity. Law and Human Behavior, 3, 319–335. [DOI] [PubMed] [Google Scholar]
  5. Anderson SW, Barrash J, Bechara A, & Tranel D (2006). Impairments of emotion and real-world complex behavior following childhood- or adult-onset damage to ventromedial prefrontal cortex. Journal of the International Neuropsychological Society, 12, 224–235. [DOI] [PubMed] [Google Scholar]
  6. Asp EW, Ramchandran K, & Tranel D (2012). Authoritarianism, religious fundamentalism, and the human prefrontal cortex. Neuropsychology, 26, 414–421. [DOI] [PMC free article] [PubMed] [Google Scholar]
  7. Barrash J, Asp EW, Markon K, Manzel K, Anderson SW, & Tranel D (2011). Dimensions of personality disturbance after focal brain damage: Investigation with the Iowa Scales of Personality Change. Journal of Clinical and Experimental Neuropsychology, 33, 833–852. [DOI] [PMC free article] [PubMed] [Google Scholar]
  8. Beadle J, Paradiso S, & Tranel D (2018). Ventromedial prefrontal cortex is critical for helping others who are suffering. Frontiers in Neurology, 9, 1–13. [DOI] [PMC free article] [PubMed] [Google Scholar]
  9. Bechara A, & Damasio A (2005). The somatic marker hypothesis: A neural theory of economic decision. Games and Economic Behavior, 52, 336–372. [Google Scholar]
  10. Bellucci G, Chernyak S, Hoffman M, Deshpande G, Monte O, Knutson K, … Krueger F (2016). Effective connectivity of brain regions underlying third-party punishment: Functional MRI and Granger causality evidence. Social Neuroscience, 12, 1–11. [DOI] [PubMed] [Google Scholar]
  11. Bright DA, & Goodman-Delahunty J (2006). Gruesome evidence and emotion: Anger, blame, and jury decision-making. Law and Human Behavior, 30, 183–202. [DOI] [PubMed] [Google Scholar]
  12. Buckholtz JW, Asplund CL, Dux PE, Zald DH, Gore JC, Jones OD, & Marois R (2008). The neural correlates of third-party punishment. Neuron, 60, 930–940. [DOI] [PubMed] [Google Scholar]
  13. Buckholtz JW, & Marois R (2012). The roots of modern justice: Cognitive and neural foundations of social norms and their enforcement. Nature Neuroscience, 15, 655–661. [DOI] [PubMed] [Google Scholar]
  14. Burgess PW, Alderman N, Forbes C, Costello A, Coates L, Dawson DR, … Channon S (2006). The case for the development and use of “ecologically valid” measures of executive function in experimental and clinical neuropsychology. Journal of the International Neuropsychological Society, 12, 194–209. [DOI] [PubMed] [Google Scholar]
  15. Carlsmith KM, & Darley JM (2008). Psychological aspects of retributive justice. Advances in Experimental Social Psychology, 40, 193–236. [Google Scholar]
  16. Carlsmith KM, Darley JM, Robinson PH (2002). Why do we punish? Deterrence and just deserts as motives for punishment. Journal of Personality and Social Psychology, 83, 284–299. [DOI] [PubMed] [Google Scholar]
  17. Ciaramelli E, Braghittoni D, & Pellegrino G (2012). It is the outcome that counts! Damage to the ventromedial prefrontal cortex disrupts the integration of outcome and belief information for moral judgment. Journal of the International Neuropsychological Society, 18, 962–971. [DOI] [PubMed] [Google Scholar]
  18. Ciaramelli E, Muccioli M, Ladavas E, & Pellegrino G (2007). Selective deficit in personal moral judgment following damage to ventromedial prefrontal cortex. Social Cognitive and Affective Neuroscience, 2, 84–92. [DOI] [PMC free article] [PubMed] [Google Scholar]
  19. Crockett M, Clark L, Hauser M, & Robbins TW (2010). Serotonin selectively influences moral judgment and behavior through effect on harm aversion. Proceedings of the National Academy of Sciences, 107, 17433–17438. [DOI] [PMC free article] [PubMed] [Google Scholar]
  20. Croft K, Duff M, Kovach C, Anderson SW, Adolphs R, & Tranel D (2010). Detestable or marvelous? Neuroanatomical correlates of character judgments. Neuropsychologia, 48, 1789–1801. [DOI] [PMC free article] [PubMed] [Google Scholar]
  21. Cushman F, Gray K, Gaffey A, & Mendes W (2012). Simulating murder: The aversion to harmful action. Emotion, 12, 2–7. [DOI] [PubMed] [Google Scholar]
  22. Damasio AR, Tranel D, & Damasio H (1990). Individuals with sociopathic behavior caused by frontal damage fail to respond autonomically to social stimuli. Behavioural Brain Research, 41, 81–94. [DOI] [PubMed] [Google Scholar]
  23. Danziger S, Levav J, & Avnaim-Pesso L (2011). Extraneous factors in judicial decisions. Proceedings of the National Academy of Sciences, 108, 6889–6892. [DOI] [PMC free article] [PubMed] [Google Scholar]
  24. Darley JM (2009). Morality in the law: The psychological foundations of citizens’ desires to punish transgressions. Annual Review of Law and Social Science, 5, 1–23. [Google Scholar]
  25. Duke AA, & Begue L (2015). The drunk utilitarian: Blood alcohol concentration predicts utilitarian responses in moral dilemmas. Cognition, 134, 121–127. [DOI] [PubMed] [Google Scholar]
  26. Durkheim E (2003). Division of labor in society: Mechanical solidarity In Bean P (Ed.), Crime: Critical Concepts in Sociology (Vol. 3, pp. 193–200). New York: Routledge. [Google Scholar]
  27. Etkin A, Egner T, & Kalisch R (2011). Emotional processing in anterior cingulate and medial prefrontal cortex. Trends in Cognitive Sciences, 15, 85–93. [DOI] [PMC free article] [PubMed] [Google Scholar]
  28. Fehr E, & Fischbacher U (2004). Third-party punishment and social norms. Evolution and Human Behavior, 25, 63–87. [Google Scholar]
  29. Fehr E, & Gachter S (2002). Altruistic punishment in humans. Nature, 415, 137–140. [DOI] [PubMed] [Google Scholar]
  30. Feigenson N, & Park J (2006). Emotions and attributions of legal responsibility and blame: A research review. Law and Human Behavior, 30, 143–161. [DOI] [PubMed] [Google Scholar]
  31. Feinstein J (2012). Examination of the limbic system’s role in emotional experience using a human lesion model. University of Iowa, Iowa City, IA, USA. [Google Scholar]
  32. Finucane M, Alhakami A, Slovic P, & Johnson SM (2000). The affect heuristic in judgments of risks and benefits. Journal of Behavioral Decision Making, 13, 1–17. [Google Scholar]
  33. Frank RJ, Damasio H, & Grabowski TJ (1997). Brainvox: An interactive, multimodal visualization and analysis system for neuroanatomical imaging. Neuroimage, 5, 13–30. [DOI] [PubMed] [Google Scholar]
  34. Ginther MR, Bonnie RJ, Hoffman M, Shen FX, Simons KW, Jones OD, & Marois R (2016). Parsing the behavioral and brain mechanisms of third-party punishment. The Journal of Neuroscience, 36, 9420–9434. [DOI] [PMC free article] [PubMed] [Google Scholar]
  35. Glass L, Moody L, Grafman J, & Krueger F (2016). Neural signatures of third-party punishment: Evidence from penetrating traumatic brain injury. Social Cognitive and Affective Neuroscience, 11, 253–262. [DOI] [PMC free article] [PubMed] [Google Scholar]
  36. Goldberg JH, Lerner JS, & Tetlock PE (1999). Rage and reason: The psychology of the intuitive prosecutor. European Journal of Social Psychology, 29, 781–795. [Google Scholar]
  37. Grabenhorst F, & Rolls ET (2011). Value, pleasure and choice in the ventral prefrontal cortex. Trends in Cognitive Sciences, 15, 55–67. [DOI] [PubMed] [Google Scholar]
  38. Haegerich TM, & Bottoms BL (2000). Empathy and jurors’ decisions in patricide trials involving child sexual assault allegations. Law and Human Behavior, 24, 421–448. [DOI] [PubMed] [Google Scholar]
  39. Haidt J (2003). The moral emotions In Davidson RJ, Scherer KR, & Goldsmith HH (Eds.), Handbook of Affective Sciences (pp. 852–870). Oxford: Oxford University Press. [Google Scholar]
  40. Koenigs M, & Tranel D (2007). Irrational economic decision-making after ventromedial prefrontal damage: Evidence from the ultimatum game. The Journal of Neuroscience, 27, 951–956. [DOI] [PMC free article] [PubMed] [Google Scholar]
  41. Koenigs M, Young L, Adolphs R, Tranel D, Cushman F, Hauser M, & Damasio A (2007). Damage to the prefrontal cortex increases utilitarian moral judgements. Nature, 446, 908–911. [DOI] [PMC free article] [PubMed] [Google Scholar]
  42. Koukounas E, & McCabe MP (2001). Emotional responses to filmed violence and the eye blink startle response. Journal of Interpersonal Violence, 16, 476–488. [Google Scholar]
  43. Krajbich I, Adolphs R, Tranel D, Denburg NL, & Camerer CF (2009). Economic games quantify diminished sense of guilt in patients with damage to the prefrontal cortex. The Journal of Neuroscience, 29, 2188–2192. [DOI] [PMC free article] [PubMed] [Google Scholar]
  44. Krueger F, & Hoffman M (2016). The emerging neuroscience of third-party punishment. Trends in Neurosciences, 39, 499–501. [DOI] [PubMed] [Google Scholar]
  45. Lerner JS, Li Y, Valdesolo P, & Kassam KS (2015). Emotion and decision making. Annual Review of Psychology, 66, 799–823. [DOI] [PubMed] [Google Scholar]
  46. Lezak M, Howieson DB, Bigler ED, & Tranel D (2012). Neuropsychological Assessment (5th ed.). New York: Oxford University Press. [Google Scholar]
  47. Neafsey EJ (1990). Prefrontal cortical control of the autonomic nervous system: Anatomical and physiological observations. Progress in Brain Research, 85, 147–166. [DOI] [PubMed] [Google Scholar]
  48. Nelissen R, & Zeelenberg M (2009). Moral emotions as determinants of third-party punishment: Anger, guilt, and the functions of altruistic sanctions. Judgment and Decision Making, 4, 543–553. [Google Scholar]
  49. Öngür D, & Price JL (2000). The organization of networks within the orbital and medial prefrontal cortex of rats, monkeys and humans. Cerebral Cortex, 10, 206–219. [DOI] [PubMed] [Google Scholar]
  50. Patil I (2015). Trait psychopathy and utilitarian moral judgement: The mediating role of action aversion. Journal of Cognitive Psychology, 27, 349–366. [Google Scholar]
  51. Petrides M, & Pandya DN (2002). Association pathways of the prefrontal cortex and functional observations In Stuss DT & Knight RT (Eds.), Principles of Frontal Lobe Function (pp. 31–50). New York: Oxford University Press. [Google Scholar]
  52. Price CJ, & Friston KJ (2002). Degeneracy and cognitive anatomy. Trends in Cognitive Sciences, 6, 416–421. [DOI] [PubMed] [Google Scholar]
  53. Riedl K, Jensen K, Call J, & Tomasello M (2012). No third-party punishment in chimpanzees. Proceedings of the National Academy of Sciences, 109, 14824–14829. [DOI] [PMC free article] [PubMed] [Google Scholar]
  54. Robinson P (1997). Structure and Function in Criminal Law. Oxford: Clarendon Press. [Google Scholar]
  55. Rolls ET (2000). The orbitofrontal cortex and reward. Cerebral Cortex, 10, 284–294. [DOI] [PubMed] [Google Scholar]
  56. Rorden C, & Karnath H (2004). Using human brain lesions to infer function: A relic from a past era in the fMRI age? Nature Reviews Neuroscience, 5, 813–819. [DOI] [PubMed] [Google Scholar]
  57. Salerno J, & Bottoms BL (2009). Emotional evidence and jurors’ judgments: The promise of neuroscience for informing psychology and law. Behavioral Sciences and the Law, 27, 273–296. [DOI] [PubMed] [Google Scholar]
  58. Samson L, & Potter RF (2016). Empathizing and systemizing (un)justified mediated violence: Psychophysiological indicators of emotional response. Media Psychology, 19, 156–180. [Google Scholar]
  59. Saver JL, & Damasio AR (1991). Preserved access and processing of social knowledge in a patient with acquired sociopathy due to ventromedial frontal damage. Neuropsychologia, 29, 1241–1249. [DOI] [PubMed] [Google Scholar]
  60. Shamay-Tsoory S, & Aharon-Peretz J (2007). Dissociable prefrontal networks for cognitive and affective theory of mind: A lesion study. Neuropsychologia, 45, 3054–3067. [DOI] [PubMed] [Google Scholar]
  61. Steffensmeier D, & Britt CL (2001). Judges’ race and judicial decision making: Do black judges sentence differently? Social Science Quarterly, 82, 749–764. [Google Scholar]
  62. Taber-Thomas B, Asp EW, Koenigs M, Sutterer M, Anderson SW, & Tranel D (2014). Arrested development: Early prefrontal lesions impair the maturation of moral judgement. Brain, 137, 1254–1261. [DOI] [PMC free article] [PubMed] [Google Scholar]
  63. Tranel D (2007). Theories of clinical neuropsychology and brain-behavior relationships: Luria and beyond In Morgan JE & Ricker JH (Eds.), Textbook of Clinical Neuropsychology (pp. 27–39). New York: Taylor and Francis. [Google Scholar]
  64. Treadway MT, Buckholtz JW, Martin JW, Jan K, Asplund CL, Ginther MR, … Marois R (2014). Corticolimbic gating of emotion-driven punishment. Nature Neuroscience, 17, 1270–1275. [DOI] [PubMed] [Google Scholar]
  65. Waytz A, & Mitchell JP (2011). Two mechanisms for simulating other minds: Dissociations between mirroring and self-projection. Current Directions in Psychological Science, 20, 197–200. [Google Scholar]
  66. Wylie L, & Forest J (1992). Religious fundamentalism, right-wing authoritarianism and prejudice. Psychological Reports, 71, 1291–1298. [Google Scholar]
  67. Yamada M, Camerer CF, Fujie S, Kato M, Matsuda T, Takano H, … Takahashi H (2012). Neural circuits in the brain that are activated when mitigating criminal sentences. Nature Communications, 3, 1–6. [DOI] [PMC free article] [PubMed] [Google Scholar]
  68. Young L, Bechara A, Tranel D, Damasio H, Hauser M, & Damasio A (2010). Damage to ventromedial prefrontal cortex impairs judgment of harmful intent. Neuron, 65, 845–851. [DOI] [PMC free article] [PubMed] [Google Scholar]
  69. Young L, Cushman F, Hauser M, & Saxe R (2007). The neural basis of the interaction between theory of mind and moral judgment. Proceedings of the National Academy of Sciences, 104, 8235–8240. [DOI] [PMC free article] [PubMed] [Google Scholar]

Associated Data

This section collects any data citations, data availability statements, or supplementary materials included in this article.

Supplementary Materials

supplementary data
NIHMS1653672-supplement-supplementary_data.docx (1.2MB, docx)

Data Availability Statement

The data that support the findings of this study are available on request from the corresponding author. The data are not publicly available due to potential loss of confidentiality of neurological patients.

RESOURCES