Psychopathy is often associated with a lack of empathy; however, it cannot be a mere lack of empathy that defines the psychopath, since, as Jeanette Kennett put it, “another group of people, autistic people, who oven more conspicuously lack empathy…do in some cases seem capable of compensating for this deficit and becoming conscientious, though often clumsy, moral agents” (Kennett, “Autism,” pg. 345). The purported moral agency of autistic people aside, in order to parse out the difference, more must be made of empathy. Empathy is an active topic in the mindreading literature, and it is closely tied to simulation theory, which is associated with the slogan, that to simulate is to put oneself in the shoes of another. Indeed, it has occasionally been called empathy theory, but this is not to assert an identity between simulation and empathy. Simulation theory’s constant competitor is theory-theory, which generally claims that minds rely on an extensive folk-psychological theory in order to read one another. Of course, there are also hybrid views at play, which incorporate elements of both simulation and theorizing in accounting for mindreading. In understanding mindreading here, I endorse the modest claim that simulation is important for normal mindreading. Further, I identify empathy as a species of affective simulation. Now, it remains for us to consider psychopathy in relation to simulation theory, in order to see to what extent psychopaths do engage in simulative mindreading, and to what extent those simulations are affective. Lastly, we shall consider in what respects psychopaths can be held morally responsible for their behavior.
In order to mindread simulationally, one must be able to replicate the simulated states in certain relevant respects. This can involve anything from an automatic response with mirror-neurons to a complex line of reasoning. It is important to note that Gregory Currie and Ian Ravenscroft would identify the latter as an example of theorizing, but one which is simulated, for “when what is simulated is theorizing, the simulation is theorizing also; when it isn’t, it isn’t” (Recreative Minds, pg. 66). For our purposes, all simulation could be theorizing. The important part is that it is still in a sense simulative, and therefore need not be problematic. Now, with regard to the aforementioned replication, there may be a problem insofar as psychopaths tend to have an overall shallow affect (Hare PCL-R). If the psychopath tries to simulatively mindread another, would this not impact his simulation, and therefore his mindreading ability? Thus it seems that a tension arises between the psychopath’s shallow affect and his good mindreading ability. The latter is represented on the Hare PCL-R by such members as glib and superficial charm, as well as cunning and manipulativeness. This is echoed in Hare’s report elsewhere that
Intense eye contact, distracting body language, charm, and a knowledge of the listener’s vulnerabilities are all part of the psychopath’s armamentarium for dominating, controlling, and manipulating others. We pay less attention to what these individuals say than to how they say it—style over substance. (Hare “Psychopaths and Their Nature,” pg. 204)
How can psychopaths have both shallow affect and relatively normal mindreading abilities? It could be said that theory-theory gives a better account of psychopathic mindreading, since that does not depend on any ability to experience (in certain relevant respects) what the target is experiencing. This conclusion does not necessarily follow. Ordinary simulations of affections involve far less intensity than the target’s experience; otherwise, all we would have is contagion. So, it cannot be said that a psychopath would need a particularly deep affect in order to identify what another feels or thinks. To identify a particularly intense affection does not require that we be able to feel it ourselves any more than a psychopath needs affections of even normal intensity to identify them. The psychopath needs nothing more than such identification for effective manipulation, which is no doubt largely theoretical enterprise anyway, and if anything it would be made easier by a shallow affect. So a psychopath can engage in simulational mindreading in virtue of being able to replicate the target in the relevant respects, among which respects is not the ability to experience an affection as strongly as the target. The more interesting tension is between selective impairment and normal mindreading, since selective impairments are much stronger than mere shallowness.
First, let us see more precisely how the typical psychopath is impaired. Alvin Goldman reports several studies conducted by R. J. R. Blair and others that indicate that psychopaths are deficient in recognizing disgusted faces, angry faces, both fearful faces (Blair et al. 2004) and fearful voices (Blair et al. 2002), and in interpreting emotion stories have trouble attributing guilt, even though they have no trouble attributing happiness, sadness, and embarrassment (Blair et al. 1995) (Goldman, Simulating Minds, pg. 116-119). Victoria McGeer also cites Blair, this time reporting that psychopaths “have trouble recognizing some facial expressions of emotion, especially fear and sadness” and similarly do not understand how to react to them (“Varieties,” pg. 230). Taken as a whole, these reports indicate that in the case of sadness, facial recognition is a problem but discernment in a story is not. McGeer also cites Haidt (2001) saying that psychopaths are generally unaffected by remorse, sympathy, shame, embarrassment, love or grief (“Varieties,” pg. 230). Embarrassment can apparently be identified in a story, but it is not an affection that the psychopath experiences directly. From this it may be generalized that there are some affections that the psychopath does not grasp at all (at least not in an ordinary, direct way), such as fear, others that the psychopath can grasp indirectly, such as sadness and embarrassment, doing better when certain conditions are established (as in a story) for judging the emotion in question, and some that he can apparently grasp directly (albeit typically superficially), such as happiness.
The psychopath’s better discernment through narrative may be telling, since narrative provides the mindreader with more factual information of the sort that could be used in a theoretical inference. Of course, a facial expression is also subject to theoretical inference, but in general it is associated with automatic, mirroring activity, and it is among Goldman’s examples of “low-level” mindreading (Goldman, Simulating Minds). If it is granted that narrative provides a different kind of information than facial expression does, requiring a more reflective than automatic sort of mindreading, then it makes sense to conclude that for embarrassment and sadness, the psychopath is relying on more on theoretical mindreading than a normal mindreader would in the same situation. At any rate, there appears to be more than one way to engage in mindreading, and where psychopaths are unable to read some states in one way, they can compensate with another; however, sometimes not even that is effective. In order to understand such mindreading better, it will be helpful to examine simulation more closely.
A common problem in simulation theory is the question of how pretend beliefs, desires, and the like can be distinct from real beliefs, desires, and the like. For instance, how can one take on the beliefs of a target without actually adopting such beliefs? Further, how does one keep one’s own beliefs out of the simulation, especially when the target is significantly different from oneself? One common approach, endorsed by Currie and Ravenscroft (Reactive Minds) says that during simulation, action systems and the like are taken “off-line,” so that one does not flee from pretend bears, for example. Another account is Goldman’s hybrid theory, which includes that one’s own beliefs &c. must be “quarantined” when they are not consistent with those of the target (Goldman, Simulating Minds, pg. 29-30). Since our general simulation claim is amenable to myriad simulation models, let these be representative of purer and hybrid theories, respectively.
So given these models, where might the psychopath go wrong, specifically such that mindreading is selectively impaired and empathy is diminished? The psychopath’s mindreading problems might derive from the imputation (or projection—though Goldman uses “imputation” in a special sense that is not synonymous with projection) itself, in which case the other mind will be left unknown because the simulated conclusion is never reattributed to it. This does not seem right, however, because the psychopath would frequently find that he has no idea what his imaginings are about. Having gone through a mindreading routine, he would be left with a reading, but one that is not to be reattached to a mind. This does not sound promising at all. Also, to try to place the psychopath’s defect here would be a strike against his whole mindreading ability, which is unwarranted because psychopaths can and do mindread. The only way to avoid that conclusion is to say that imputation itself is selectively impaired, which would seem to require separate, dedicated imputation mechanisms, apart from whatever mechanisms are used for simulation or theorizing—this point holds for both theories—where it seems strange and uneconomical enough to posit dedicated imputation mechanisms at all. Therefore, the problem is poorly characterized as just a problem with imputation. If one wants to connect the problem to certain mechanisms, they are much better characterized as the mechanisms associated with the impaired states themselves, but we do not commit to this here.
Whence the selective impairment? Recalling the “off-line” view, a psychopath may lack empathy because he or she fails to take his or her own beliefs, desires, and so forth off-line. On the Goldman hybrid theory, the same follows from quarantine failure. This failure may either include or exclude pretend beliefs taken on for the sake of simulation. If the psychopath excludes such pretend beliefs, then he or she will simply impute to the target whatever he would think or do, depending on the specific simulation. This latter possibility is especially important if psychopaths are determined to be incapable of forming certain beliefs, feeling certain affections, and the like, since one cannot impute what one simply does not have. If the psychopath includes pretend beliefs &c. of the target alongside still on-line beliefs &c. of his or her own, then there will almost certainly be conflict, and this likely has implications for mindreading.
At this point, it is worthwhile to consider the idea of imaginative resistance. Shaun Nichols attributes this idea to Richard Moran, and it is basically that there are some things that are difficult or impossible to imagine. Moran emphasizes the emotional and moral imaginings that we resist, for example, that in fact murder is universally acceptable. Nichols amends this with non-affective beliefs, such as mathematical contradictions, that the imagination resists (Nichols “Imagining,” pg. 136). There are two potential applications of imaginative resistance to the psychopath. If the psychopath tries to simulate another mind without taking her own systems off-line, then the imaginative resistance would presumably ensue, in light of inconsistent beliefs, desires, &c. If something like this happens, then it is plausible to think that the psychopath’s simulative abilities are rather limited. Problematically, however, a general defect in simulative ability would indicate an overall defect in mindreading, and this does not characterize the psychopath. Psychopaths have selective mindreading defects, as noted above. If the psychopath’s problems result from a failure to take her own beliefs, desires, &c. offline, then the psychopath must actually experience the relevant states in order for them to conflict with pretend states. Further, we should only expect to see resistance occur when the target is significantly different from the psychopath, or else what conflict can there be? Imaginative resistance as a result of the inability to go off-line (or as a result of quarantine failure) does not adequately account for the psychopaths’s own lack of certain affections or the selectivity of the defects.
There is, however, another potential use of imaginative resistance in explaining the psychopath’s condition. This is that the psychopath experiences imaginative resistance precisely when required to take on those states for which she has a strong affective deficit. This is to be expected on a simulationist account, since pretend states are supposed to resemble the target’s states in certain relevant respects. When the possibility of significant resemblance is nonexistent, then imagination will be quite impossible; hence, there is what we can call imaginative resistance. This application amounts to a description of how, given a selective impairment, both affect and mindreading are altered in parallel. Importantly, the evidence does not suggest that the psychopath is utterly unable to simulate but is still a basically normal mindreader; rather the psychopath is selectively unable to simulate in those ways that would characterize a lack of empathy, and in a parallel, selective way she is unable to mindread, with some mindreading abilities apparently compensated in non-simulative ways, perhaps theorizing. So the psychopath’s affective deficits are apt to attract notice, especially when they manifest behaviorally, but the ability to read such deficits has considerably less impact on overall mindreading ability, first because much mindreading has little to nothing to do with affect, and also in light of probable compensation by non-simulative means. This, then, is our working characterization of the psychopath, in light of those claims stated at the beginning. Imaginative resistance is a proposal as to how impaired affect and impaired mindreading are linked, but this need not be accepted to use our operative conception of the typical psychopath, so long as it is granted that such a link is coherent (as virtually any simulationist should grant). Where exactly the selectivity comes from is a question for scientific investigation.
Now that we have an idea of what psychopathy is, we shall see what this could mean for the moral responsibility of psychopathic agents. Lloyd Fields argues that there is a certain subset of psychopaths that cannot be held morally responsible because they are unable to form other-regarding moral beliefs (Fields, “Psychopathy”). For Fields, having a moral belief involves being “motivated to act in accordance with it,” and moral reasons have priority over non-moral reasons in guiding action, and sentiments follow from having moral beliefs (Fields 1995, pg. 268). Fields says that his subset of psychopaths fails to exhibit any of these three conditions, and indeed they cannot exhibit them, when it comes to other-regarding moral beliefs. Fields thinks that from an inability to form other-regarding moral beliefs it follows that one is “an unsuitable subject of social pressure,” which in turn entails that one is not subject to scrutiny according to the social morality (Fields “Psychopathy,” pg. 273). Thus, any psychopaths in the relevant subset are not morally responsible agents with respect to the social morality.
Fields is right to point out the heterogeneity of those bearing the psychopathic label, a heterogeneity that is supported by our analysis of psychopathy above. Because the concept of the psychopath developed here does not specify precisely which impairments are in the selection (while naming some typical candidates), which no doubt vary to a degree from case to case, different types of psychopath can easily be accounted for. In fact, even different sorts and severities of empathy-related impairments can be accounted for. So, as far as our analysis is concerned, Fields points to the right question when he asks whether there is any type of psychopath that is not properly held to be morally responsible or legally responsible, rather than whether psychopaths simpliciter are subject to such responsibilities.
What sort of psychopath is not morally responsible? If Fields is right, it is precisely those incapable of forming other-regarding moral beliefs, but is this so? What does it mean to form other-regarding moral beliefs? Further, why should these beliefs be other-regarding in particular? Starting with the formation of moral beliefs, Fields says, “to have a moral belief is to accept a practical principle as a moral principle” (Fields 1995, pg. 268). So a psychopath must be unable to accept other-regarding practical principles as moral principles. The problem may arise either in relation to the acceptance of an other-regarding practical principle or in relation to the acceptance of one as a moral principle. The former is not plausible at all. Even if we were to wholeheartedly accept such a claim as Hare’s that a psychopath’s “‘conscience’ is only half-formed, consisting merely of an intellectual awareness of the rules of the game” (Hare, “Psychopaths,” pg. 205), we would be left with some rules by which the psychopath abides, be it for a purpose like manipulation or an attempt to pass as normal, and these rules are taken to count as practical principles. By way of illustration, psychopaths can play chess, though they might cheat. The very facts of psychopathic mindreading and manipulativeness alone are enough to amply evidence the psychopath’s ability to hold other-regarding practical principles. To see another or even a thing as a means to an end suggests that even the worst sort of psychopath holds some other-regarding practical principles. So the psychopath’s problem is not with practical principles.
If Fields is right, then the psychopath is left with a number of practical principles, undifferentiated between moral and non-moral, at least when it comes to other-regarding principles. We have already seen what Fields considers to be some marks of a moral principle: motivation, priority over the non-moral, and accompanying sentiment. Given the psychopath’s selective impairment, we should expect certain moral sentiments to be foreign to the psychopath and the rest may be fairly superficial. Unless one is willing to argue that a psychopath lacks all the sentiments relevant to other-regarding moral belief, then the concept is problematic. Take happiness, for example, with which the psychopath does not appear to have trouble (except possibly superficiality), as seen above. Can happiness (or any other non-impaired affection) not be an other-regarding moral sentiment? Can happiness not be involved in, say, charity? Can happiness not motivate one to act charitably? In order to hold that the psychopath fails in this way is to eliminate all affections that could possibly be involved in a moral principle. This alone is suggestive of the idea that the psychopath that is not morally responsible is hard to come by, but one of Fields’s conditions is still untouched.
Prioritization remains a problem. If the prioritization of principles is sentimentally guided, and psychopaths typically have shallow affect, how can they be expected to order their principles? For one, it is reasonable to think that the ordering of principles is often non-sentimental. In support of this, prioritization is often important in end-directed practice. Some psychopaths appear to be at once quite capable of end-directed activity and others quite inept at it. On the one hand, the typical psychopath can work toward ends, as the manipulation for which psychopaths are known is frequently end-directed (for example, passing as normal to gain release from prison); however, the Hare PCL-R includes impulsivity, irresponsibility, and lack of realistic long-term goals as other marks of the psychopath. To confound the matter further, the same Hare describes psychopaths as “calculating predators” (Hare, “Psychopaths,” pg. 205). Thus it seems that there are two kinds of psychopath being described here, one that is at least moderately adept at achieving his ends and another that is far too impulsive to consider any ends apart from those of the moment. In all probability, this constitutes a continuum, with the Machiavellian (who is probably not a psychopath) on one end and the most impulsive sort of psychopath on the other. Hence, it is reasonable to say that some psychopaths can prioritize principles, and these psychopaths can hold some moral beliefs, at least self-regarding ones. This also suggests that the more manipulative psychopaths are more likely to be more morally accountable. Thus the psychopath has at least a basic moral responsibility for those moral beliefs he does hold; however, it is important to remember that Fields is interested in whether the psychopath can be held morally responsible in the context of a social morality. This requires that we examine what it means for a moral belief to be other-regarding and how this relates to social morality.
Clearly the typical psychopath can at least form other-regarding beliefs. If not, we would be left saying that they are incapable of any normal mindreading, which simply is not true. Fields is not looking to classify his subset as those psychopaths that cannot mindread normally at all; indeed, that sounds more like autism. Recall that in normal mindreading, we have postulated that simulation plays an important part. So the psychopath is capable of simulative mindreading, except concerning a selective set of affective states, with some variance from case to case. The psychopath will be expected to be unresponsive to social pressure—Fields’s somewhat vague account of moral epistemology—concerning those affections about which she is impaired, since these the psychopath cannot apprehend by normal mindreading.
Unless a particular set of impaired affective states comprises the basis for forming moral beliefs simpliciter or forming other-regarding moral beliefs in particular, which the relevant kind of psychopath must be said to lack, the case against the moral responsibility of psychopaths is in trouble. We have seen that the typical psychopath is capable of holding some moral beliefs, though there could be a subset that cannot do so at all (a thesis stronger than even Fields is interested in supporting). Since the psychopath is capable of holding other-regarding beliefs as well as moral beliefs, do the sets at any point overlap? This seems inevitable concerning those affections for which psychopaths usually have no deficit, such as happiness. In addition, the possibility of mindreading through non-simulative means discussed above allows for other-regarding moral beliefs to be inculcated and moral behavior acted out on a basis that is non-simulative, non-sentimental, &c, but this is a vast and different line of criticism. If this is true, then we have defined the sort of psychopath to which Fields’s argument does not apply. Fields should not find such a set disagreeable, since his claim was a limited one, though it is one that is likely more limited than he should like, since it seems to require a very thorough affective deficit, much more so than the typical psychopath would seem to exhibit. For if empathy is a species of simulational mindreading, then it requires the ability to simulate certain empathetic affections, and Fields’s psychopaths would have to be missing all of them. Further, even granted that the psychopath is genuinely incapable of empathy in this sense, other forms of mindreading must not be able to compensate for the deficiency in a way that inculcates moral belief. So while there may turn out to be some psychopaths that are not good candidates for moral responsibility, there are certainly a great many that are given our broad conception of mindreading. Thus, not only have we tried to display conditions on which a psychopath is morally accountable, we have also attempted to define more clearly those conditions on which a psychopath might be said not to be a proper candidate for moral responsibility. First, the non-responsible psychopath must have a deficit including precisely those affections that comprise empathy, where empathy is a species of affective simulation consisting in the aforesaid affections (which must be enumerated, something we have not attempted here). Second, the psychopath must not be able to compensate for any of those impaired affections through a non-simulative form of mindreading. If this is right, and the basic simulationist and basically moral-epistemological sentimentalist picture is accepted, the best one could do is argue for moral immunity only with respect to those specific deficits that meet the stated conditions.
Bibliography
Currie, Gregory and Ian Ravenscroft. Recreative Minds. Imagination in Philosophy and Psychology. Oxford, New York: 2003.
Fields, Lloyd. “Psychopathy, Other-Regarding Moral Beliefs, and Responsibility.” Philosophy, Psychiatry, & Psychology, Vol. 3, No. 4, Dec. 1996. pp. 261-277.
Goldman, Alvin. Simulating Minds. The Philosophy, Psychology, and Neuroscience of Mindreading. Oxford, New York: 2006.
Hare, Robert. “Psychopaths and Their Nature: Implications for the Mental Health and Criminal Justice Systems.”
---. Revised Psychopathy Checklist.
Kennett, Jeanette. “Autism, Empathy, and Moral Agency.” The Philosophical Quarterly, Vol. 52, No. 208, 2002, pp. 340-357.
McGeer, Victoria. “Varieties of Moral Agency: Lessons from Autism (and Psychopathy).” Moral Psychology, Vol. 3: The Neuroscience of Morality: Emotion, Brain Disorders, and Development. Walter Sinnott-Armstrong, ed. MIT Press, Cambridge, MA: 2008.
Nichols, Shaun. “Imagining and Believing: The Promise of a Single Code.” The Journal of Aesthetics and Art Criticism, Vol. 62, No. 2, Spring 2004. pp. 129-139.
