, Volume 11, Issue 1, pp 55–65 | Cite as

Psychopathy, Executive Functions, and Neuropsychological Data: a Response to Sifferd and Hirstein

  • Marko Jurjako
  • Luca Malatesti
Original Paper


Katrina Sifferd and William Hirstein, in their paper ‘On the criminal culpability of successful and unsuccessful psychopaths’, argue that neuropsychological data show that unsuccessful psychopaths have diminished mental capacities that warrant a partial or diminished responsibility defence. We respond that the currently available neuropsychological evidence does not warrant their conclusion that unsuccessful psychopaths should not be deemed completely legally responsible. Instead, we maintain that the current state of this type of research suggests that psychopaths might be suffering very specific cognitive impairments. However, the impact that these impairments might have on the specific criminal behaviours that courts have to assess is far from clear.


Executive function Legal responsibility Neuropsychological evidence Psychopathy Sifferd Hirstein 

not broken, just bent ….



Katrina Sifferd and William Hirstein in their paper ‘On the criminal culpability of successful and unsuccessful psychopaths’ [1] have argued that neuropsychological data show that unsuccessful psychopaths have impairments in their executive functions (from now on EF) that would warrant a partial or diminished responsibility defence. Generally, successful psychopaths have standard psychopathic personality traits: they are unemotional, callous, glib, manipulative, pathological liars, etc. However, they usually have no criminal records and rarely come into contact with the criminal law. Unsuccessful psychopaths, in addition, have criminal records and often are entangled with legal systems [2, ch. 7]. EF include a class of processes such as attention, decision-making, reasoning, problem solving, memory, and inhibition of action that are at the focus of much current neurocognitive research [3].

In this paper, we object to their conclusion. We recognise that their proposal has several methodological merits. We argue, however, that it ultimately fails because the current neuropsychological literature on the EF of psychopaths does not support the conclusion that unsuccessful psychopaths are impaired so severely to lack or have diminished criminal responsibility. In fact, we show that current neuropsychological studies on EF in psychopaths offer mixed results. Moreover, we argue that none of the most plausible interpretations of these inconsistences, that are advanced by experts in the field, support, at least without further argument, a line of reasoning of the type offered by Sifferd and Hirstein.

In the remaining sections, we proceed as follows. Firstly, we outline Sifferd and Hirstein’s argument and clarify its central notions. In the next section, we argue that the kind of neuropsychological evidence concerning the EF of psychopaths that they offer, given mixed results, cannot support their case. In the further section, we show that two explanations of these mixed results, similarly, do not support Sifferd and Hirstein’s reasoning. On the one hand, the hypothesis that mixed results mirror the distinction between successful and unsuccessful psychopaths is empirically unsubstantiated. On the other hand, it could be plausible to claim that the mixed results depend on the heterogeneity of the underlying factors of psychopathy. In fact, although there are contrasting empirical data, we cannot say that the position is empirically implausible. However, we argue that this further hypothesis cannot support, without more detailed arguments concerning open issues in the science of psychopathy, Sifferd and Hirstein’s line of reasoning. Finally, we consider a hypothesis for the explanation of the mixed results on the EF of psychopaths that has considerable currency amongst the experts. This is the assumption that there are at least two different types of executive function: hot EF and cool EF that are involved in the performance of different EF tasks and that are, probably, served by different neural mechanisms. Also in this case, we argue that this explanation does not allow interpreting the current neuropsychological evidence as supporting directly the conclusion that unsuccessful psychopaths suffer EF impairments so severe to completely undermine or substantially diminish their criminal culpability.

Sifferd and Hirstein’s Argument

Sifferd and Hirstein’s paper offers an important methodological contribution to the ongoing debate on the criminal responsibility of psychopaths for, at least, two principal reasons. Firstly, while many have focused on the impact that the emotional and empathic deficits associated with psychopathy have on their moral understanding and motivation (see for a critical survey of this wide literature [4]), they address the less investigated issue of the integrity of their rational capacity [5].

Secondly, by means of the notions of EF and capacity responsibility, Sifferd and Hirstein offer an important insight on how to relate the rational capacities required for criminal liability, that the law expresses in folk psychological terms, with current neuropsychological evidence [6]. Following H.L.A. Hart [7], Sifferd and Hirstein assume that capacity responsibility is at the core of the psychological preconditions for criminal responsibility. Capacity responsibility refers to:

understanding, reasoning, and control of conduct: the ability to understand what conduct legal and moral rules require, to deliberate and reach decisions concerning these requirements; and to conform to decisions when made [1, p. 131].

In other words, according to them, capacity responsibility refers to rational abilities that a person needs to have in order to be legally accountable. They maintain that the function of rationality is to “recognize and correct for abnormal intentional states” [1, p. 132]. In addition, they think that “proper executive function is the scientific equivalent of legal rationality” [1, p. 132]. Combining these two claims, they contend that EF can be construed as the capacity to “correct for emotional and perceptual deficits” [1, p. 136]. Therefore, properly functioning EF are, at least, necessary conditions for capacity responsibility and thus for criminal accountability.

The core of Sifferd and Hirstein’s reasoning is that criminal psychopaths, unlike successful ones, have diminished control over their action because they suffer EF impairments. For instance, they write:

Given the connections between executive function and basic human rationality, as well as the ability of executive function, when it exists, to correct for emotional and perceptual deficits, we would argue that those psychopaths who possess normal or above normal executive function bear full responsibility for their acts. Alternatively, psychopaths whose executive function is substantially compromised should be considered less responsible. In what follows, we will assume that those psychopaths with normal [or] above normal executive function are successful psychopaths, while those with below normal executive function are unsuccessful, as indicated by the Ishikawa et al. study.1 [1, p. 136].

Therefore, their argument could be formulated as follows:
  • P1: EF are necessary for capacity responsibility.

  • P2: EF are substantially impaired or dysfunctional in criminal or unsuccessful psychopaths.

  • C: Therefore, criminal or unsuccessful psychopaths have diminished capacity responsibility.

The overall conclusion is that unsuccessful psychopaths should be either granted the insanity defence or some form of the diminished responsibility defence. Sifferd and Hirstein opt for the second possibility [1, p. 137].

Given the general specification of EF offered by Sifferd and Hirstein, we grant the first premise of their argument. In fact, in most legal frameworks, the epistemic and the control requirement, which together comprise the capacity responsibility, are central in the insanity and diminished responsibility defences. The epistemic requirement is that the agent did not know the nature of the criminal action. The control condition is that the agent could not regulate one’s behaviour in the light of practical and/or moral demands and commitments when committing the unlawful act. For instance, the American Model Penal Code states that a person might be exculpated for her action if she:

at the time of such conduct as a result of mental disease or defect lacks substantial capacity either to appreciate the wrongfulness of [her] conduct or to conform [her] conduct to the requirements of the law. [8, 4.01.]

Most philosophical research on whether psychopaths should be exculpated on the basis of the epistemic condition has focused on whether psychopaths have appropriate moral understanding to “appreciate the wrongfulness of [their] conduct”. In a recent review paper, Schaich Borg and Sinnott-Armstrong seem to conclude that currently there is not enough evidence to think that psychopaths suffer from significant deficits in moral understanding (4; however, see, [9]).

Sifferd and Hirstein’s argument, instead, is plausibly construed as aiming to show that psychopaths to an important degree satisfy the exculpating control condition [1, pp. 137–138]. According to them, the function of EF is to regulate intentional states and actions [1, pp. 131–132]. However, the notion of control is not all or nothing. A person might be more or less in control of what she is doing and thinking. In addition, not every impairment in abilities that underlie control will warrant an exculpation. To illustrate this point, Sifferd and Hirstein offer the following analogy:

Tom […] is completely color blind. This makes his representations of the world abnormal in that they don’t contain information most people have. Imagine Tom is arrested for running a red light and hitting a pedestrian. He might argue that, because he was incapable of perceiving the traffic light normally, he shouldn’t be held responsible. The police and prosecutors will not consider this incapacity to be an excuse for Tom’s behavior and the harm he caused, however. Why? Because Tom has had ample opportunities, in the gaps between his intentional perceptual states (faulty as they may be) and his behavior, to correct for his incapacity. Tom should have noticed his problem and corrected for it, by memorizing the position of the red light versus the green. [1, p. 132].

Tom seems to be responsible for his behaviour because, despite his defective colour perception, he still has the rational capacity, by “exerting a bit of effort”, to “correct for the lack of relevant information” and react appropriately to available reasons [1, p. 133].

Determining in general the minimal degree of control that is necessary for criminal accountability is difficult. Sifferd and Hirstein do not explicitly address this potential problem. However, they reasonably contend that “psychopaths whose executive function is substantially compromised should be considered less responsible” [1, p. 136]. In addition, they think that neuropsychological evidence shows that unsuccessful psychopaths comprise a group that has substantially compromised EF.

In the remainder of this paper, we question the second premise of Sifferd and Hirstein’s argument, as it pertains to the control condition. We argue that the empirical evidence that they adduce does not support the conclusion that even unsuccessful psychopaths suffer impairments in EF so severe and of the type that should undermine their accountability.

Mixed Results on the Executive Functions of Unsuccessful Psychopaths

To assess Sifferd and Hirstein’s argument, we need to address the preliminary task of clarifying, in general, the kind of neuropsychological evidence that is relevant for the case at issue. They [1, pp. 133–134] adduce several studies concerning the brain differences or impairments that correlate with psychopathy. They maintain that current empirical research supports the view that:

Most executive processes reside in the prefrontal lobes, including the dorsolateral frontal lobes on the side of the brain, the ventrolateral frontal lobes below them, the medial prefrontal lobes on the inner surfaces of the two hemispheres, and the orbitofrontal lobes located on the under surface of the brain, just above the eye sockets. They function as parts of larger cortical networks containing sensory and mnemonic areas located in the posterior regions of the brain, supported by subcortical nuclei. [1, p. 132].

There are brain studies that appear to show that psychopaths have a lowered activity in the amygdala, and the broader paralimibic region of the brain [10], including the anterior cingulate cortex, the ventromedial prefrontal cortex, the orbitofrontal cortex and, especially, in the functional connection between the ventromedial prefrontal cortex and the amygdala [11, 12].

Studies concerning the anatomic and physiological differences in the brain of psychopaths cannot, without further argument, support the second premise in Sifferd and Hirstein’s argument. Even if we grant that these brain studies are robust and valid, they do not ipso facto show that psychopaths suffer impairments in EF that are relevant for the ascription of rational control and, thus, criminal responsibility. For example, people whose corpus callosum (brain part that connects the two hemispheres of the brain) was surgically removed tend to exhibit negative side-effects (such as difficulties in speaking and remembering things, paralysis, loss of sensation, etc.). People who were born without the corpus callosum, tend not to manifest any harmful behavioural effects [13, pp. 183–184]. Similarly, there are EF studies with psychopaths in which it was confirmed that children with psychopathic traits and conduct disorder, and criminal psychopaths show abnormal brain activations, but do not exhibit behavioural differences in task performance [14, 15]. Thus, relying solely on brain studies is not sufficient to conclusively argue that psychopaths’ EF are diminished in a way that is relevant for the ascription of criminal responsibility.2

The central issue is whether the neurological peculiarities of psychopaths cause or at least reliably indicate EF impairments that are significant for the ascription of criminal responsibility [16]. Thus, the most important source of evidence must be the performance of unsuccessful psychopaths in behavioural and cognitive tasks that measure EF that are relevant for the rational control of behaviour. For instance, Sifferd and Hirstein write that “unsuccessful psychopaths have reduced prefrontal and amygdala volumes and hippocampal abnormalities”, which suggests cognitive dysfunctions, such as “reduced executive functioning, including impaired decision-making” [1, p. 134]. However, we think that the evidence they offer does not support their line of reasoning.

As already noted, Sifferd and Hirstein refer to Ishikawa et al. study [17] to claim that unsuccessful psychopaths have impaired EF. This study is based on the Wisconsin Card Sorting Task (WCST). This task represents the most general way to measure EF, since the successful performance on the task utilizes attention, working memory, and inhibition. The task involves learning to appropriately match cards. Participants are not told how they should match the cards, but they are given positive or negative signals indicating what is the right or wrong way of proceeding. Usually the task is to learn a criterion according to which the cards should be matched (for instance, by the shape of the item, its colour or the number it displays). The task is to flexibly learn to change responses, since, for instance, every 10 trials the matching criterion is changed. Sifferd and Hirstein maintain that:

The Ishikawa et al. study also found that, compared with unsuccessful psychopaths (…), successful psychopaths had enhanced executive functioning as measured by the Wisconsin Card Sorting Task (WCST) [17] (…). Indeed, successful psychopaths showed significantly better performance on the WCST than non-psychopathic controls [17]. In contrast, unsuccessful psychopaths scored lower than the controls, even though the two psychopathic groups did not differ on full scale IQ compared with the controls [1, pp. 134-135] (Reference number adjusted and italics added).

Indeed, in that study the results showed that on the WCST successful psychopaths performed better than unsuccessful psychopaths and non-psychopathic control groups. However, Sifferd and Hirstein describe this study inaccurately.

The study by Ishikawa et al. does not show that unsuccessful psychopaths have impaired EF to an extent that would undermine their responsibility. In fact, Ishikawa et al. explicitly state that an “interesting finding was that unsuccessful psychopaths did not differ from controls on executive function” [17, p. 430].3 Thus, if Sifferd and Hirstein had interpreted Ishikawa et al. as giving a more nuanced view on the connection between unsuccessful psychopaths and the control, non-psychopathic, participants, they probably would not assume, without further argument, a tight connection between unsuccessful psychopaths and below normal EF capacities that would support exculpation. However, besides the issue of the correct interpretation of this specific study, we think that there are more general difficulties concerning the cogency of current neuropsychological evidence for the conclusion that unsuccessful psychopaths have impaired EF in a way that is relevant for their capacity responsibility.

A general problem for the debate on psychopaths’ responsibility is that studies on psychopathy and EF have produced mixed results. In the words of renowned psychopathy researchers, the “results are equivocal with regard to the relationship between deficits in EF and psychopathy” [18, p. 336]. In fact, some studies show that psychopathy negatively correlates with EF, some show that there is no general correlation, other still indicate that there is positive correlation between EF and psychopathy [19]. In the following sections, we show that a specific problem for Sifferd and Hirstein’s view is that the most plausible explanations for these conflicting results do not favour their conclusion.4

The Heterogeneity of Psychopathy

Siffred and Hirstein’s line of reasoning could be defended by suggesting that the inconsistency in the results of different studies might be accounted for by the different EF profiles of successful and unsuccessful psychopaths. Since unsuccessful psychopaths are supposed to have lower EF and the successful ones normal or higher EF, not distinguishing between the two groups in different studies might explain the mixed results. In fact, this is what Ishikawa et al. were trying to establish:

If null findings in prior research on failed, criminal psychopaths have, in fact, resulted because the controls were failed criminals themselves, then the unsuccessful psychopaths in the current study should have shown worse performance than the nonconvicted controls. However, this was not the case. [17, p. 430].

As we have seen in the previous section, they were not successful in supporting this hypothesis. In any case, considering the current literature, this explanation cannot account for the whole range of inconsistent or mixed results among different studies.5 In fact, most of the variability in results that we emphasize comes from studies on incarcerated psychopaths, since most of the neuropsychological studies are performed on prison populations [20]. In particular, as we already mentioned, there are studies that show that even in the population of criminal or unsuccessful psychopaths the behavioural results on certain EF tasks are not significantly different from control groups [14, 15, 17].6

A more plausible explanation of the inconsistency in results might be based on the assumption that the construct of psychopathy is not unitary, but it rather captures a heterogeneous collection of traits [21, p. 127]. For instance, it is often taken that the “golden standard” for measuring psychopathy in forensic settings, the Psychopathy Checklist-Revised (PCL-R) that consists of 20 diagnostic items, is divided into two factors [22].7 Factor 1 comprises interpersonal and affective traits, which are usually taken to denote the core psychopathic traits, while Factor 2 comprises antisocial and lifestyle traits.8 Here the idea is that, for instance, Factor 1 traits correlate more with normal EF, while Factor 2 traits correlate with subnormal or dysfunctional EF. Thus, on this view, the variations in the quality of EF performance depends on which factors of the psychopathy construct are more prominent in an individual.

Crystal Lantrip and colleagues [21], for instance, examined the correlations between psychopathy traits and items that characterize executive functions. As a measure of psychopathy they used the Psychopathic Personality Inventory (PPI). PPI is a self-report measure that was devised to assess psychopathic traits in noncriminal populations [23]. Lantrip et al. used a two factor model of PPI, which includes the Fearless-Dominance (FD) factor and the Antisocial-Impulsive (AI) factor.9 In order to test executive function, Lantrip et al. used a self-report measure of EF, the Behavior Rating Inventory of Executive Function (BRIEF-A) [see, for instance, [21], p. 128]. This inventory includes 75 items that are divided into two indexes that measure the subjects’ reports on their own behaviour regulation, such as cognitive flexibility, inhibition, emotional control, and self-monitoring. The second index measures items that are related to metacognition, such as working memory, planning and organization, and task monitoring. On this measure, subjects rate themselves on particular items, where higher scores indicate lower EF. In a sample of more than 500 participants from a general nonforensic population, Lantrip and colleagues found correlations between the total scores on the PPI and lower self-assessments of EF. In particular, they found that higher BRIEF-A scores correlate positively with Antisocial-Impulsive traits, while to a greater extent they correlate negatively with Fearless-Dominance traits. Thereby, providing support to the hypotheses that the mixed results on the relation between psychopathy and EF might be explained by the fact that psychopathy involves separable heterogeneous traits. In addition, Ross, Benning, and Adams [24] found similar positive and negative correlations by using different self-report measures of psychopathy and executive function in a sample of incarcerated individuals and students from a college population.

The explanation of the mixed results in EF studies on psychopaths, however, does not support Sifferd and Hirstein’s argument. Firstly, this explanation cuts across the distinction between successful and unsuccessful psychopaths. Namely, Lantrip et al. found correlations between different psychopathic traits and worse EF self-assessments in a general nonforensic population.

Secondly, there is also some indication that the mixed results of the studies that tested the performance of EF in psychopaths cannot be wholly accounted for in terms of different traits that are comprised in psychopathy. In fact, there are mixed results given this proposal as well. For instance, Mol and colleagues [25] tested the differential contribution of Factor 1 and 2 of PCL-R on incarcerated psychopaths by administering the Wisconsin Card Sorting Task. They report that after controlling for confounding variables, they did not find statistically significant correlations between the performance on WCST and Factor 1 or Factor 2 and non-psychopathic participants. In addition, Maes and Brazil [19, pp. 1271–1272] in their review study report that majority of studies failed to establish statistically significant positive correlation between interpersonal and affective traits (Factor 1) and EF or negative correlations between antisocial and impulsive traits (Factor 2) and EF.

The possible explanations for these inconsistencies might involve multiple factors. In the studies that we mentioned, for instance, Lantrip et al. [21] used self-report measures of psychopathy (PPI) and executive function, while Mol and colleagues [25] used PCL-R, which is administered by trained psychologists, and performance-based test for EF (WCST). This raises difficult questions about the methodology of assessing EF, which do not seem to be settled among the scientists working in this area. For instance, Lantrip et al. seem to claim that in the context of inconsistent results in the psychopathy research self-report measures of executive function could provide more ecologically valid results than performance-based measures [21, p. 130]. Given that substantially impaired EF would be related to notable dysfunctions in the brain’s frontal lobe, Ross, Benning, and Adams, on the contrary, claim that using a self-report measure of EF, although useful for performing large-scale studies, is “not optimal”, and that “the most epistemologically sound method for measuring frontal lobe dysfunction” is provided by neuroimaging [24, p. 387]. In a similar vein, Mol et al. suggest that progress on the question of the performance of EF in psychopaths could be made by using more diverse neuropsychological tasks that target more specific abilities that underlie EF [25, pp. 136–137]. This suggests another possible explanation of the inconsistent results, that is worth considering in the context of our discussion of the argument offered by Sifferd and Hirstein. We turn to it in the next section.

Hot and Cool Executive Function

A further plausible explanation of the inconsistent results concerning EF tasks is that in the studies different research teams have used different types of tasks and measures that target different types of executive functions. For instance, there is a consensus, that the term “executive function” is ambiguous [26, p. 46]. In fact, there are at least two conceptualizations of EF in the psychopathy literature:

Cool EF tasks are relatively abstract tasks without motivational or emotional relevance, largely mediated by lateral inferior and dorsolateral frontostriatal and frontoparietal networks (…). Instead, hot EF tasks are motivationally or emotionally salient and mediated by orbitomedial and ventromedial frontolimbic structures (…). [19, p. 1267].

Thus, Sifferd and Hirstein’s argument might be interpreted as claiming that unsuccessful psychopaths suffer from impairments in cool, hot or both types of EF that are significant enough to undermine their capacity responsibility.

It might be assumed that Sifferd and Hirstein claim that the relevant impairment in unsuccessful psychopaths concerns their cool EF. In fact, they seem to mostly rely on studies which tested psychopaths’ performance on Wisconsin Card Sorting Task [1, pp. 134–135, 137]. Most notably, cool EF are mediated by the dorsolateral prefrontal cortex (DLPFC) and are standardly measured by the aforementioned task.

Sifferd and Hirstein’s argument, however, cannot rely on the claim that unsuccessful psychopaths have impairments in cool EF as measured by WCST. Just to reiterate, already Ishikawa et al. [17], in their seminal study on successful and unsuccessful psychopaths, and Mol et al. [25] found that unsuccessful psychopaths do not perform differently on WCST than non-psychopathic control participants. In addition, in a recent study, Pera-Guardiola and colleagues [26] give evidence that prisoners with high psychopathic traits perform better on the WCST than other offenders with lower psychopathy scores. The results showed that individuals with antisocial personality disorder (ASPD) and low psychopathy scores (as measured by PCL-R, see [22]) performed worse than healthy controls and individuals with ASPD and high psychopathy scores. Despite some incongruent results on cool EF and psychopathy [see, [27]], Pera-Guardiola et al. [26] conclude that their study supports the claim that psychopaths do not show relevant deficits in cool EF and therefore do not show deficits in the DLPFC. On the basis of their studies, Bagshaw, Gray, and Snowden [28] and Snowden et al. [29] reach similar conclusions about functioning of DLPFC in psychopaths. Thus, it seems that different studies do not confirm the hypothesis that unsuccessful psychopaths suffer from substantial impairments in cool EF.

It might be suggested, alternatively, that Sifferd and Hirstein’s argument relies on the thesis that unsuccessful psychopaths have impaired hot EF. There is some evidence that psychopaths have abnormal brain areas underlying hot EF [30]. To repeat some data described in the third section of this paper, most notably, these areas are the amygdala, the ventromedial prefrontal cortex, and the orbitofrontal cortex [11]. Hot EF are standardly measured by instrumental learning tasks, such as passive avoidance, response-reversal, and gambling tasks.10 For example, in the Iowa gambling task, participants need to spontaneously learn to avoid responding to bad decks of cards and learn to respond to good decks [31]. In particular, a participant is presented with four decks of cards. Choosing from two decks, early in the game gives high rewards (earning money), but it also incurs higher costs (losing money). Choosing from the other two decks gives lesser rewards, but on average it constitutes a winning strategy (overall net gain). Thus, participants need to learn to avoid choosing from the first two decks and pick cards from the other two decks. In these types of tasks, it has been shown that psychopaths often perform worse than non-psychopathic control groups [30]. For example, in gambling tasks, psychopaths tend to be risk prone and keep picking cards from disadvantageous decks, and in response reversal tasks they tend to learn slower reinforcement contingencies of the stimuli.

Also interpreting Hirstein and Sifferd’s line of argument as concerning hot EF does not make it viable for two reasons. Firstly, if we plug in the notion of hot EF in Sifferd and Hirstein’s argument then the argument becomes inconclusive. In fact, it seems that hot EF performances do not distinguish significantly between successful and unsuccessful psychopaths (compared to non-psychopaths). In fact, in a recent review of the relevant data, Andrea Glenn and Adrian Raine [2, pp. 151–152] indicate that both types of psychopaths perform similarly on passive avoidance and gambling tasks, indicating that they both exhibit abnormalities in the function of the amygdala and orbitofrontal cortex, which underlie hot EF. In addition, Snowden et al. [29] found that students with higher psychopathic traits tend to exhibit similar cool and hot EF profiles to incarcerated psychopaths.

Secondly, the deficits exhibited on hot EF tasks do not seem to be significant enough to grant psychopaths the insanity or diminished responsibility defence without further argument. This type of exculpation would require that psychopaths are incapable to control their behaviour in a wide range of ordinary situations. However, the current neuropsychological investigations appear to testify deficits that are highly context depended and that are revealed by studies carried forward in very specific conditions [20, 32, 33]. Let us consider, for instance, passive avoidance tasks, where the participant has to spontaneously learn how to avoid negative feedback. For example, in a task of this type, a participant chooses from one deck of cards, where the chosen card indicates whether she gains or loses points. After a certain number of rounds, the losses start to pile up, so it becomes advantageous to stop playing the game. The goal of the task is to stop playing when further card picking becomes disadvantageous. If the task is performed under certain conditions, psychopaths keep playing the game even when this becomes disadvantageous. However, in conditions where they pay attention to their cumulative earnings and pause before making a decision whether to continue to play, psychopaths perform on the task similarly to healthy control groups [20, p. 96]. Thus, it seems that, at least when they are attentive to the stage and the goal of the task psychopaths’ performance on passive avoidance task is similar to the performance of the control groups [20, p. 96]. Similarly, Lösel and Schmucker [34] showed that psychopaths’ performance on gambling tasks, is predicted by their attention scores. Thus, if a person is less attentive, she will score less well than those who are more attentive. This variability or context dependency of psychopaths’ performance led Maes and Brazil [19; see, also, 18] to argue that there is no persuasive evidence for thinking that psychopathy is correlated with generalized or global EF deficits.

This context dependency is significant because it indicates that psychopaths have enough capacities to correct potential “emotional and perceptual deficits”. To remember, according to Sifferd and Hirstein the role of EF is to “correct for emotional and perceptual deficits” [1, p. 136]. As showed in section 2, they give the analogy with a colour-blind person. Just because someone is blind to colours and lives in a world where important cues are signalled by colours, it does not follow that one has substantially impaired capacity to control herself. With this example, Sifferd and Hirstein put emphasis on the top-down processes that are necessary for correcting “emotional and perceptual deficits”. A long line of research by Joseph Newman and his colleagues [for an overview, see, 20] gives significant evidence that the contextuality of psychopaths’ performance on instrumental learning tasks can be explained by how their intentional and sub personal processes are engaged. It seems that psychopaths can rectify their performance on instrumental learning tasks, which usually involve properly conditioned affective responses, by employing top-down attentional processes. In fact, this research plausibly indicates that psychopaths do not have substantially impaired capacities that are needed for correcting “emotional and perceptual deficits”.

For the reasons adduced above, we add that even the conjunctive construal of EF, as including both hot and cool EF, does not support Sifferd and Hirstein’s argument. However, we think that there are important issues concerning the culpability of psychopaths that are raised by our discussion.

It could be argued that the deficits in the affective mechanisms that are exhibited on hot EF tasks are significant enough by themselves to warrant less culpability, given how they influence psychopaths’ behaviour.11 For instance, non-responsiveness to threat cues and to punishing stimuli, as exhibited on many instrumental learning tasks, might warrant diminished legal accountability since it disables psychopaths from affective-based learning and possibly from being responsive to the repressive apparatus of a legal system in so far as it depends on having normal affective reactions (35; for an opposite opinion, see, [36]). In addition, and more importantly, it could be argued that the fact that in experimental conditions, psychopaths perform differently, depending on how their attention is focused, might not provide an ecologically valid measure of control. For instance, there is no guarantee that when they are in the real world they could utilize top-down attentional processes to control effectively their behaviour [21].

Although the context dependence of the deficits might reduce the strength of these complaints, in future discussions they should be more thoroughly addressed. In the context of the present debate, we can only offer few remarks or a gesture towards a possible answer. Concerning the ecological validity objection, we would suppose that in real life, given the multiple sources and forms of stimuli, an array of different capacities would normally be engaged, including attention, working memory, long-term memory, sub personal modules, domain-general mechanisms, and so on. Given this, it is not clear how to disentangle the amount and strength of the causal impact of any specific mechanism in ordinary life. For sure, ordinary life involves conditions that are wholly unlike those in which psychopaths’ affective and cognitive deficits, as they relate to the neuropsychological definition of EF, are discovered. They are established in experimental situations, where a specific task is designed to test a specific capacity or only its parts. Given all this, without further empirical evidence, we should not easily discard the idea that psychopaths’ ability to regulate their behaviour by top-down processes might be more important in daily life, and for judging their legal accountability, than the particular impairments that they exhibit in controlled experimental conditions.


In conclusion, we agree with Sifferd and Hirstein [1, p. 136] that the EF profile of successful psychopaths cannot justify deeming them less or wholly nonresponsible for their behaviour.12 However, we disagree with their claim that current neuropsychological evidence shows that unsuccessful psychopaths suffer EF impairments so severe that undermine their criminal accountability as it pertains to the control condition.

We would like to point out the limits of our discussion that, in responding to Siffred and Hirstein, focuses on the type of evidence they rely upon. Our view is arrived at by paying close attention to the specific experimental circumstances of the relevant neuropsychological studies on the EF of unsuccessful psychopaths and some interpretative issues that experts raise about them. Of course, we are not excluding that further empirical studies of this type or refinements on their interpretations could provide more conclusive evidence.

However, we would like to close by suggesting that relating legal practices of ascription of culpability to the crimes of those classified as being psychopaths might require also more ecologically oriented types of studies than those that are so far available in neuropsychology. The clinical diagnosis of psychopathy, as for instance regimented by the PCL-R, offers a suggestive description of psychopaths as impulsive individuals that appear to lack control, where the notion of impulsivity and control are cashed out in terms of descriptive operationalization of folk psychological notions. Perhaps, besides neurological hypotheses based on highly specific EF paradigms, an explanation and a validation of the generality of these behavioural features, as tendencies that might be relevant for legal practice, should be casted in more general psychological terms. This could suggest investigating the cognitive profile of psychopaths with qualitative or structured interviews and analyses, which are aimed at testing the integrity of general cognitive functions and ecologically individuated capacities.13 These capacities could be more readily individuated as necessary requirements for legal culpability.


  1. 1.

    In fact, as we show below, Sifferd and Hirstein are mistaken in grounding the correlation between unsuccessful psychopaths and below normal EF on Ishikawa et al.’s [17] study.

  2. 2.

    An explanation why people who are born without the corpus callosum do not exhibit behavioural deficits is the fact that during development their brains manage to learn how to transmit information form one hemisphere to the other by utilizing some alternative pathways [13, p. 185]. Similarly, there is evidence that core psychopathic traits are highly heritable [2, ch. 1], it could be that through development psychopaths’ brains learn to solve different EF tasks even though they show abnormal functioning in the brain areas that normally subserve EF.

  3. 3.

    A reviewer of this journal has interestingly noted that in the Ishikawa et al. study, there is some reported raw data according to which there is a positive correlation between lower performance on the WCST and unsuccessful psychopathy. However, in their study, as indicated by the quote above, Ishikawa et al. do not take these raw data as relevant since these differences vanish after the IQ scores of the participants have been taken into account. In any case, Sifferd and Hirstein do not appear to rely on these raw data. Instead, it seems that they couple their incorrect reading of the main conclusion of Ishikawa et al. study with the claim that this result is robust even when participants’ IQ’s have been matched.

  4. 4.

    Thus, we do not find the inconsistency in results significant because we think that there is something methodologically wrong with the studies on which Sifferd and Hirstein rely. We argue that once a plausible explanation for the mixed results is on the table, then we should see that they do not provide sufficient data for thinking that unsuccessful psychopaths, unlike successful ones, should be held less legally accountable. We thank an anonymous reviewer for nudging us to be more explicit about this point.

  5. 5.

    In fact, Roald Maes and Inti Brazil [19] in their review study argue that, despite the appearances, there is no plausible evidence that there is a group of psychopaths who have superior EF.

  6. 6.

    We thank an anonymous reviewer for pressing us to bring the claim in this paragraph to the foreground.

  7. 7.

    In fact, the issue whether the PCL-R should be divided into two or more factors is still debated among the psychopathy researchers [37]. For an overview, see, [38]. We do not take sides on this issue. However, it has to be noted that not all researchers agree that factorization of psychopathy measures implies disunity in the construct of psychopathy. For instance, Robert Hare [22, p. 50] contends that PCL-R, despite its factorization, measures a unitary construct. On the other hand, Jalava, Griffiths, and Maraun argue that the factorization of PCL-R indicates that “there is no single, real, entity called psychopathy” [39, p. 194, see also, pp. 194–207].

  8. 8.

    According to Hare [22, p. 79], Factor 1 includes: Glibness and superficial charm; Grandiose sense of self-worth; Pathological lying; Conning/manipulative; Lack of remorse or guilt; Shallow affect; Callous/lack of empathy; Failure to accept responsibility for own actions. Factor 2 includes: Need for stimulation/proneness to boredom; Parasitic life style; Poor behavioural controls; Early behavioural problems; Lack of realistic, long-term goals; Impulsivity; Irresponsibility; Juvenile delinquency; Revocation of conditional release.

  9. 9.

    According to the original factor analysis, PPI’s 187 items were divided into 8 factors. These included Machiavellian Egocentricity, Social Potency, Coldheartedness, Impulsive Nonconformity, Blame Externalization, Carefree Nonplanfulness, and Stress Immunity [23, p. 44]. Later statistical analyses indicated that PPI could be divided into two factors. Social Potency, Fearlessness, and Stress Immunity form a factor that is nowadays called Fearlessness-Dominance. This includes traits such as absence of anxiety, willingness to take risks and skill at influencing others. Machiavellian Egocentricity, Impulsive Nonconformity, Blame Externalization, and Carefree Nonplanfulness form Factor 2 or Antisocial-Impulsivity traits. These include lack of concern for the social norms, attitude of indifference toward the future, and “a ruthless willingness to manipulate and take advantage of others” [23, p. 44]. These analyses also indicated that FD contains items that correlate with PCL-R’s Factor 1, while AI correlates with Factor 2 of PCL-R [21, 23].

  10. 10.

    Instrumental learning tasks involve learning to connect the valenced (positive or negative) value of stimuli to an action that is appropriate for solving the task. For an overview of these tasks as they apply to psychopaths in the present context, see [40].

  11. 11.

    We thank Neil Levy for pressing us to address this worry.

  12. 12.

    Of course, in this occasion, we have to leave open the issue whether they have abnormal affective or other mechanisms that might undermine their responsibility, especially by affecting specific moral competencies.

  13. 13.

    In this respect, the "Socratic" interviews done by philosopher Jonathan Glover and psychiatrist Gwen Adshead with the inmates of the Broadmoor Hospital are very suggestive. See, for instance, Glover’s recent book [41].



Many thanks to Zdenka Brzović and Lovro Savić for reading and giving us valuable comments on different drafts of the paper. We would like to thank the two anonymous reviewers and Neil Levy, the editor of Neuroethics; their comments helped us to improve greatly this paper. In addition, we thank the organisers and participants of the following events, where parts of the paper were presented and discussed: Mente e medicina, Uniser, Pistoia, Italy 8/9/2016; Invited talk, Faculty of Media and Communications, Belgrade (Serbia), 9/6/2016; Conference: Ethical Issues: Theoretical and Applied, Bled (Slovenia) 6-10/6/2016; Invited talk, Donders Institute for Brain, Cognition and Behaviour, Nijmegen (Netherlands) 12/04/2016.

The Croatian Science Foundation (HRZZ) funds our research (project CEASCRO: grants n. 8017 and n. 9522).


  1. 1.
    Sifferd, K.L., and W. Hirstein. 2013. On the criminal culpability of successful and unsuccessful psychopaths. Neuroethics 6: 129–140.CrossRefGoogle Scholar
  2. 2.
    Glenn, A.L., and A. Raine. 2014. Psychopathy: an introduction to biological findings and their implications. New York: New York University Press.CrossRefGoogle Scholar
  3. 3.
    Otero, T.M., and L.A. Barker. 2014. The frontal lobes and executive functioning. In Handbook of executive functioning, ed. S. Goldstein and J.A. Naglieri, 29–44. New York: Springer.CrossRefGoogle Scholar
  4. 4.
    Schaich Borg, J., and W.P. Sinnott-Armstrong. 2013. Do psychopaths make moral judgements? In Handbook on psychopathy and law, ed. K.A. Kiehl and W.P. Sinnott-Armstrong, 107–128. Oxford: Oxford University Press.Google Scholar
  5. 5.
    Maibom, H. 2005. Moral unreason: the case of psychopathy. Mind & Language 20: 237–257.CrossRefGoogle Scholar
  6. 6.
    Hirstein, W., and K. Sifferd. 2010. The legal self: executive processes and legal theory. Consciousness and Cognition 20: 156–171.CrossRefGoogle Scholar
  7. 7.
    Hart, H.L.A. 1968. Punishment and responsibility: essays in the philosophy of law. Oxford: Clarendon Press.Google Scholar
  8. 8.
    Model penal code. 1962. Philadelphia: American Law Institute.Google Scholar
  9. 9.
    Levy, N. 2013. Psychopaths and blame: the argument from content. Philosophical Psychology 27: 351–367.CrossRefGoogle Scholar
  10. 10.
    Kiehl, K.A. 2006. A cognitive neuroscience perspective on psychopathy: evidence for paralimibic system dysfunction. Psychiatry Research 142: 107–128.CrossRefGoogle Scholar
  11. 11.
    Blair, J.R. 2008. The amygdala and ventromedial prefrontal cortex: functional contributions and dysfunction in psychopathy. Philosophical Transactions of the Royal Society B 363: 2557–2565.CrossRefGoogle Scholar
  12. 12.
    Motzkin, J.C., J.P. Newman, K.A. Kiehl, and M. Koenigs. 2011. Reduced prefrontal connectivity in psychopathy. The Journal of Neuroscience 31: 17348–17357.CrossRefGoogle Scholar
  13. 13.
    Jeeves, M.A. 1996. Callosal agenesis. In The Blackwell dictionary of neuropsychology, ed. G.J. Beaumont, P.M. Kenealy, and M.J.C. Rogers, 182–186. Oxford: Blackwell Publishers.Google Scholar
  14. 14.
    Finger, E.C., A.A. Marsh, D.G. Mitchell, M.E. Reid, C. Sims, S. Budhani, D.S. Kosson, G. Chen, K.E. Towbin, E. Leibenluft, D.S. Pine, and R.J.R. Blair. 2008. Abnormal ventromedial prefrontal cortex function in children with psychopathic traits during reversal learning. Archives of General Psychiatry 65: 586–594.CrossRefGoogle Scholar
  15. 15.
    Gregory, S., R.J.R. Blair, D. Ffytche, A. Simmons, V. Kumari, S. Hodgins, and N. Blackwood. 2015. Punishment and psychopathy: a case-control functional MRI investigation of reinforcement learning in violent antisocial personality disordered men. Lancet Psychiatry 2: 153–160.CrossRefGoogle Scholar
  16. 16.
    Sifferd, K.L. 2013. Translating scientific evidence into the language of the ‘folk’: executive function as capacity-responsibility. In Neuroscience and legal responsibility, ed. N.A. Vincent, 183–204. Oxford: Oxford University Press.CrossRefGoogle Scholar
  17. 17.
    Ishikawa, S.S., A. Raine, T. Lencz, S. Bihrle, and L. Lacasse. 2001. Autonomic stress reactivity and executive functions in successful and unsuccessful criminal psychopaths from the community. Journal of Abnormal Psychology 110: 423–432.CrossRefGoogle Scholar
  18. 18.
    Baskin-Sommers, A.R., I. Brazil, J. Ryan, N.J. Kohlenber, C.S. Neuman, and J.P. Newman. 2015. Mapping the association of global executive functioning onto diverse measures of psychopathic traits. Personality Disorders: Theory, Research and Treatment 6: 336–346.CrossRefGoogle Scholar
  19. 19.
    Maes, J.H.R., and I.A. Brazil. 2013. No clear evidence for a positive association between the interpersonal-affective aspects of psychopathy and executive functioning. Psychiatry Research 210: 1265–1274.CrossRefGoogle Scholar
  20. 20.
    Koenigs, M., and J.P. Newman. 2013. The decision making impairment in psychopathy: psychological and neurobiological mechanisms. In Handbook on psychopathy and law, ed. K.A. Kiehl and W.P. Sinnott-Armstrong, 93–106. Oxford: Oxford University Press.Google Scholar
  21. 21.
    Lantrip, C., S. Towns, R.M. Roth, and P.R. Giancola. 2016. Psychopathy traits are associated with self-report rating of executive functions in the everyday life of healthy adults. Personality and Individual Differences 101: 127–131.CrossRefGoogle Scholar
  22. 22.
    Hare, R. 2003. The psychopathy checklist-revised. Toronto: Multi-Health Systems.Google Scholar
  23. 23.
    Fowler, K.A., and S.O. Lilienfeld. 2013. Alternatives to the psychopathy checklist-revised. In Handbook on psychopathy and law, ed. K.A. Kiehl and W.P. Sinnot-Armstrong, 34–57. Oxford: Oxford University Press.Google Scholar
  24. 24.
    Ross, S.R., S.D. Benning, and Z. Adams. 2007. Symptoms of executive dysfunction are endemic to secondary psychopathy: an examination in criminal offenders and noninstitutionalized young adults. Journal of Personality Disorders 21: 384–399.CrossRefGoogle Scholar
  25. 25.
    Mol, B., P. Van Den Bos, Y. Derks, and J. Egger. 2009. Executive functioning and the two-factor model of psychopathy: no differential relation? The International Journal of Neuroscience 119: 124–140.CrossRefGoogle Scholar
  26. 26.
    Pera-Guardiola, V., I. Batalla, J. Bosque, D. Kosson, J. Piffare, R. Hernandez-Ribas, X. Goldberg, O. Contreras-Rodriguez, J.M. Menchon, C. Soriano-Mas, and N. Cardoner. 2016. Modulatory effects of psychopathy on Wisconsin card sorting test performance in male offenders with antisocial personality disorder. Psychiatry Research 235: 43–48.CrossRefGoogle Scholar
  27. 27.
    De Brito, S.A., E. Viding, V. Kumari, N. Blackwood, and S. Hodgins. 2013. Cool and hot executive function impairments in violent offenders with antisocial personality disorder with and without psychopathy. PloS One 8: 1–12.CrossRefGoogle Scholar
  28. 28.
    Bagshaw, R., N.S. Gray, and R.J. Snowden. 2014. Executive function in psychopathy: the tower of London, Brixton spatial anticipation and the Hayling sentence completion tests. Psychiatry Research 220: 483–489.CrossRefGoogle Scholar
  29. 29.
    Snowden, R.J., N.S. Gray, S. Pugh, and G. Atkinson. 2013. Executive function as a function of sub-clinical psychopathy. Personality and Individual Differences 55: 801–804.CrossRefGoogle Scholar
  30. 30.
    Blair, J.R., D.G. Mitchell, and K. Blair. 2005. The psychopath: emotion and the brain. Oxford: Blackwell Publishing.Google Scholar
  31. 31.
    Bechara, A., A. Damasio, and S. Anderson. 1994. Insensitivity to future consequences following damage to human prefrontal cortex. Cognition 50: 7–15.CrossRefGoogle Scholar
  32. 32.
    Brazil, I.A., J.H.R. Maes, I. Scheper, B.H. Bulten, R.P.C. Kessels, R.J.A. Verkes, and E.R. de Bruijn. 2013. Reversal deficits in individuals with psychopathy in explicit but not implicit learning conditions. Journal of Psychiatry and Neuroscience 38: E13–E20.CrossRefGoogle Scholar
  33. 33.
    Hamilton, R.K.B., K. Hiatt Racer, and J.P. Newman. 2015. Impaired integration in psychopathy: a unified theory of psychopathic dysfunction. Psychological Review 122: 770–791.CrossRefGoogle Scholar
  34. 34.
    Lösel, F., and M. Schmucker. 2004. Psychopathy, risk taking, and attention: a differentiated test of somatic marker hypothesis. Journal of Abnormal Psychology 113: 522–529.CrossRefGoogle Scholar
  35. 35.
    Glenn, A.L., A. Raine, and W.S. Laufer. 2011. Is it wrong to criminalize and punish psychopaths? Emotion Review 3: 302–304.CrossRefGoogle Scholar
  36. 36.
    Maibom, H. 2008. The mad, the bad, and the psychopath. Neuroethics 1: 167–184.CrossRefGoogle Scholar
  37. 37.
    Međedović, J., B. Petrović, D. Kujačić, J. Želeskov-Đorić, and M. Savić. 2015. What is the optimal number of traits to describe psychopathy? Primenjena Psihologija 8: 109–130.Google Scholar
  38. 38.
    Cooke, D.J., C. Michie, and Stephen D. Hart. 2006. Facets of clinical psychopathy: toward clearer measurment. In Handbook of psychopathy, ed. C.J. Patrick, 91–106. New York: The Guilford Press.Google Scholar
  39. 39.
    Jalava, J., S. Griffiths, and M. Maraun. 2015. The myth of the born criminal. Toronto: University of Toronto Press.Google Scholar
  40. 40.
    Blair, J.R. 2006. Subcortical brain systems in psychopathy: the amygdala. In Handbook of psychopathy, ed. C.J. Patrick, 296–312. New York: The Guilford Press.Google Scholar
  41. 41.
    Glover, J. 2014. Alien landscapes? Interpreting disordered minds. Cambridge: Harvard University Press.Google Scholar

Copyright information

© Springer Science+Business Media Dordrecht 2016

Authors and Affiliations

  1. 1.Department of Philosophy, Faculty of Humanities and Social Sciences, CEASCRO ProjectUniversity of RijekaRijekaCroatia

Personalised recommendations