Fearlessly Amoral: Psychopaths, Autistics and Learning with Emotion

In “The Emotional Construction of Morals”, Jesse J. Prinz suggests that the reason for the inability of psychopaths to learn the moral/conventional distinction is because morality is essentially emotion-based.  “If moral properties were not essentially emotion involving, there should be a way of drawing the moral/conventional distinction without appeal to emotions.  Psychopaths should be able to learn the difference. … The fact that they fail to master the moral/conventional distinction suggests that there may be no way to draw that distinction without averting to or experiencing emotional responses”[Prinz, The Emotional Construction of Morals, pg 46] Ultimately, Prinz notes psychopaths are generally of at least average IQ, and some are quite intelligent.  Moreover, psychopaths get much aid from others – including parents, teachers, and psychologists and psychiatrists – in learning morality.  They also seem to have an interest in learning these distinctions since much of the time their amorality lands them in prison or psychiatric hospitals.  That they continually fail to learn morality seems to indicate that they cannot, which may suggest that they cannot because they lack emotional mechanisms [ibid].

This essay will examine the question of whether it is simply or primarily the lack of emotions that makes it so that psychopaths fail to properly learn morality.  The main thrust of this examination will be comparing psychopaths to their classic foil in this area: autistics.  Autistics seem to have emotional deficiencies, and certainly share the lack of empathy that characterizes psychopaths, and yet they seem to perform significantly better on almost all moral measures than psychopaths.  Why would this be the case?  Also, this essay will tease out the distinction between learning about morals – and thus, for example, the moral/conventional distinction – and acting on them.  Psychopaths clearly do not perform well morally, but their failure to understand the moral/conventional distinction suggests a failure to learn basic concepts of morality.  Is their problem performance, learning or both?  Could the combination of performance and learning disabilities be responsible for their characteristic amorality?  Could it be a general learning disability that is causing the problem for psychopaths, one that autistics do not share and can work around?  If not, what exactly is the difference between psychopaths and autistics that allows autistics to be moral where psychopaths are not?

Before getting into these issues, though, it must be established that autistics really are moral and psychopaths really are not.  The latter seems to be well-established, but the former may be questionable.  However, in order to determine if autistics can be a valid case study for morality it must be established what should be used to determine what counts as, at least, moral enough.

The first thing to take into account is, in fact, moral performance.  As stated by Prinz in the quote above, the moral/conventional distinction is a key indicator of moral performance.  The moral/conventional distinction is the ability to tell a moral rule from one that is merely conventional.  The normal way to test this is to test for authority-independence.  Moral rules hold independently of any authority, while conventional rules do not.  Thus, asking someone if the action would be okay if the most directly relevant authority said it would be okay – like, say, a teacher in a classroom – is a good way to test for this distinction.  Psychopaths do not pass this test.  In fact, their responses demonstrate their complete lack of understanding of this distinction.  Prinz cites a study by James Blair that demonstrates that adult criminal psychopaths respond as if both moral and conventional rules are authority-independent, while psychopathic children answer as if neither are.  Prinz points out that Blair suggests that the psychopathic criminals are trying to make it appear like they understand what things are moral and understand why moral rules are bad, but fail miserably, while the psychopathic children are perhaps giving a more genuine reply as to how they view the moral and the conventional.  But in all cases, what is important is that neither can see any distinction between moral and conventional rules. [Prinz, The Emotional Construction of Morals, pg 44].

Autistics, however, do seem to perform relatively well with respect to the moral/conventional distinction.  Frederique de Vignemont and Uta Frith cite R.J.R.Blair’s 1996 study that demonstrates that autistics perform about as well as non-autistics on determinations of moral or conventional violations, despite – as they note – having sharp deficiencies in theory of mind [De Vignemont and Frith, Comment on McGeer, pg 276].  Thus, when we appeal to the moral/conventional distinction as a measure of moral performance, autistics – at least, high-functioning autistics – generally succeed, while psychopaths generally fail.  Thus, in terms of moral performance, autistics are far ahead of psychopaths.

De Vignemont and Frith, however, challenge the claim that autistics do perform better on that scale in an interesting way: by challenging if they really and truly understand the distinction.  They bring up two issues.  The first is that while autistics can identify faux pas, they may not properly understand them.  When asked why one should not commit faux pas, they appealed to rules instead of the pain of the victim.  The second issue asks if the autistics were really treating moral rules as authority-independent, or if they were just appealing to another authority than the one presented in the test (so, say, instead of it being dependent on the authority of the teacher it was dependent on the authority of their parents, even though they were not present) [De Vignemont and Frith, Comment on McGeer, pg 277].  Both of these challenges appeal to the basis of the judgement, as opposed to simply getting the answers right.   However, to suggest that having a different basis for their moral judgements must exclude autistics from being moral would be to beg the question against them, and  is outweighed by the fact that even getting that level of performance is beyond the capacities of psychopaths.  Even if the basis of moral behaviour in autistics is one that could be judged as not a proper moral basis, at least autistics present a fairly good facsimile of moral behaviour, a facsimile that one can only wish psychopaths could achieve.

However, the challenges do raise a second concern, which is that while we need not be overly concerned here about the basis of their moral performance, it does in fact matter why they report those distinctions.  Ultimately, the judgements need to be genuine, in the sense that we would have to believe that psychopaths and autistics really do think that the distinction really is a meaningful one.  It cannot be the case that our contenders are, in fact, faking their moral views just to make other people think better of them or to achieve some goal.  What is interesting is that, as seen earlier, when psychopaths try to fake performance on the moral/conventional distinction they still perform poorly.  They would sound more moral, but still clearly demonstrate that they do not understand the moral/conventional distinction.  Autistics, on the other hand, do seem to genuinely hold the distinction.  Temple Grandin in her book “Thinking in Pictures: My Life With Autism” learned in college that there were some rules defined as “sins of the system”, and that as long as those major rules were followed, minor ones would be ignored.  “Students got into trouble for smoking and having sex.  If a student could be totally trusted not to break those two rules, she could break some of the minor rules with no consequences.”[Grandin, Thinking in Pictures: My Life with Autism, pg 108].  Ultimately, this led her to hold three categories: “really bad”, “illegal but not bad”, and “sins of the system”.  For her, stealing, destroying property, and hurting others are “really bad”, slight speeding and illegal parking count as “illegal, but not bad”, while the “sins of the system” were for things that were punished harshly for what seemed like illogical reasons [ibid, pg 109].  Translating this to the moral/conventional distinction, the “really bad” category is the moral, while the latter two seem to be conventional.  And Temple Grandin seems genuine in these classifications; she does not seem to be just saying that for personal gain.   Thus, Grandin gives us proof that autistic performance on the moral/conventional distinction is, in fact, genuine; she really does hold that distinction in her own judgements.

The conclusion, then, is this: autistics are, in fact, good enough at morality to be our example.  They succeed, generally, on both performance and on genuineness, while psychopaths succeed at neither.  If psychopaths could learn to be even as good at morality as autistics that would be a massive improvement in their moral knowledge and behaviour.  However, they do not seem to even be able to get to that point.  Why is it that psychopaths cannot even achieve the level of autistics with respect to morality?

The first thing that can be eliminated is empathy.  While it seems that most people do rely in some way on empathy in their moral reasoning, both psychopaths [Blair, The Psychopath: Emotion and the Brain, pg 53-56] and autistics [Yirmaya, Empathy and Cognition in High-Functioning Children with Autism, pg 150-160] demonstrate a deficiency with respect to empathy.  So, while a popular explanation might be that psychopaths lack empathy and therefore lack the ability to be or learn morality properly, the fact that autistics lack empathy as well belies this.  Note that this seems to be in relation to a specific sort of empathy, that which involves actually feeling the emotions.  Yirmaya et al demonstrated that high-functioning autistics perform better on empathy tests as their IQ increases, suggesting that they are using a more cognitively based model for empathy than non-autistics do [ibid, pg 157].  This could, then, suggest that autistics are substituting a cognitively based empathy for the more emotionally based one, which allows them to learn morality where psychopaths do not.  However, this is belied by the fact that psychopaths are generally both manipulative and successful at manipulating others.  In order to manipulate people and succeed in social situations, psychopaths have to in some sense be able to recognize and manipulate the emotions that people are and will feel.  Simply having a good, solid theory about what people will feel in certain situations would provide this, and that is a more cognitively based approach to empathy.  Since they are better at such manipulation than autistics, it suggests that their cognitively based empathic abilities are at least as good as that of autistics, if not better.  Thus, it is unlikely that the reason for the improved performance of autistics with respect to morality is due to their superior ability at cognitive empathy.

It is worth taking a deeper look at the evidence that psychopaths lack empathy, to see if it might be the case that the most indicative studies detailing their lack of empathy might be focussing on emotional empathy, rather than cognitive empathy.  Blair outlines three different types of studies that are used to test empathy in psychopaths.  The first is to give the subjects a questionnaire, asking them to self-report on their empathy.  The second is to measure the subject’s autonomic responses to the distress of others.  The final type of study is based around identifying emotional expressions vocal effect [Blair, The Psychopath: Emotion and the Brain, pg 53-55].  Blair notes that psychopaths do poorly at the last two paradigms, and declines to discuss their results on the first because of potential confounds with verbal and linguistic capacity.  Specifically: “Those who wish to look more empathic can learn to parrot empathic verbal responses.” [ibid, pg 53].  However, it is interesting to note that the latter two cases explicitly measure emotional empathy, by either directly measuring the emotional affect of psychopaths or by asking the psychopaths to relate an expression with the emotion felt.  The first type, however,  could be answered using cognitive empathy as well as emotional empathy.  While more detailed analysis would be required, that the clearest cases of lack of empathy in psychopaths explicitly measure emotional empathy coupled with the earlier arguments about how psychopaths do seem to require at least a cognitive understanding of empathy in order to succeed at their manipulation makes it quite likely that psychopaths – like autistics —  lack emotional empathy but retain at least some cognitive empathy.

Thus, the lack of emotional empathy is common between autistics and psychopaths, and thus is not likely to be the explanation for the morality of autistics and amorality of psychopaths. And neither seem to lack cognitive empathy, so that being lacking in psychopaths does not seem to be the explanation either.   However, this would not rule out a difference in emotions being the cause.  As Prinz says: “Children with autism do not exhibit the characteristic signs of guilt, but they do exhibit aversive emotions when they violate rules … “[Prinz, The Emotional Construction of Morals, pg 117].  Could they be using a different emotion or set of emotions to achieve the same results?

It seems that the key might well be the aversive emotions.  Temple Grandin has stated that when she breaks a moral rule – like that of lying – she feels incredibly anxious [McGeer, Varieties of Moral Agency: Lessons from Autistics (and Psychopaths), pg 244].  A large part of this is because of the social uncertainty and expectations, which is referenced by Heidi Maibom: “It is not hard to see that having to function in a society that one does not comprehend can be anxiety inducing” [Maibom, The Will to Conform, pg 268].  Their anxiety over social situations gives them motivation to learn the social structures that others follow and that anxiety carries over into their behaviour.  As part of that, they learn the moral/conventional distinction.  After all, not being able to tell the difference can lead to many errors in the social sphere.  First, one might treat moral rules too leniently, and be criticized for not being good enough.  Second, one may take conventional rules too seriously and thus refuse to participate in social situations – eating or singing in class, for example – because of an over-extension of those conventional rules.  Autistics, then, are motivated to learn these distinctions properly and have aversive emotions that tie into their understanding of these motivations.  If anyone can learn the moral/conventional distinction, it seems that they can.  And they do.

Thus, autistics do indeed have aversive emotions, and at least one of them – anxiety – ties directly into not only their behaviour, but their motivation to learn proper behaviour including the moral/conventional distinction.   What can be said for psychopaths?  It turns out that the very emotion that autistics can rely on is not present in psychopaths.  While autistics clearly experience anxiety and may even in fact experience it more strongly than non-autistics (as evidenced by Grandin’s extreme anxiety when lying, which is above and beyond what most non-autistics experience) psychopaths seem the exact opposite: cool, calm, collected and unaffected.  On many measures, they seem to demonstrate reduced anxiety [Blair, The Psychopath: Emotion and the Brain, pg 47 – 50].  Thus, there is one critical difference between autistics and psychopaths: anxiety level.  So Prinz’s suggestion has some merit.  Autistics may be using a different aversive emotion as the underpinnings of their morality.  This would seem to be anxiety.  Psychopaths are missing anxiety.  Anxiety, then, might be the answer.

Or, it might not.  Or, at least, not the whole answer.  What is it about being anxious that allows autistics to learn the moral/conventional distinction and hold it genuinely while psychopaths do not?  Prinz’s initial question is still unanswered: if there is some way of telling which rules are “really bad”, “illegal but not bad”, or “sins of the system”, why are psychopaths unable to learn that, at least to the point of being able to convincingly answer therapists about what counts as each, even if they cannot legitimately hold those distinctions themselves?  Anxiety gets us the genuineness factor for autistics, and perhaps without that psychopaths need something else to get that genuineness factor.  But the problem is still one of general performance.  Since psychopaths can generally learn, why can they not learn this well enough to fake it?

Thus, we must turn to learning in the psychopath, and there we discover that they indeed have a learning disability that seems to be tied to their levels of anxiety: psychopaths have serious problems with aversive conditioning and instrumental learning.  Aversive conditioning is conditioning a fear response to a stimulus that the subject had not previously been afraid of before, by delivering a punishment when that stimulus is experienced.  Blair cites a study by David Lykken which paired a buzzer with an electric shock to try to see if psychopaths would start sweating in response to the buzzer, which is the normal response to the electric shock.  The reactions of psychopaths were impaired with respect to the sweating reaction [Blair, The Psychopath: Emotion and the Brain, pg 49].  More interestingly, they also seem to have greatly reduced anticipation of negative events.  If someone knows that they will receive an electric shock in five seconds, as that time approaches their anxiety level should increase.  This is not the case for psychopaths; the measurable qualities of emotional anxiety do not increase and certainly start far closer to the actual shock than they do for non-psychopaths [ibid].  This would seem to relate both to their reduced anxiety in general and perhaps to a deficiency in being able to anticipate that a shock will occur; at a basic level, they may not really believe that they will be shocked when the buzzer stops, even if they are intellectually aware of it.

Instrumental learning is also interesting.  “Instrumental learning involves learning to commit specific behavioral responses in order to gain reward or avoid punishment.” [ibid, pg 51].  Psychopaths have issues with specific forms of this, particularly passive avoidance and response reversal.  In passive avoidance, the subject must learn to avoid responding to thing that will give them punishments, while response reversal is when the subject must stop responding to a stimulus that was once a reward but now punishes.  The impairment of the first has been repeatedly demonstrated, while Blair uses the example of a card game developed by Joe Newman to demonstrate the second.  In that game, participants must decide whether to play a card or not.  At first, playing is always rewarding, but as the game goes on the probability of playing being rewarding decreases, and eventually it will be primarily punishing.  While most non-psychopaths do learn to stop playing once punishment becomes too likely, psychopaths do not, to the point of losing all of their points [ibid, pgs 51 – 52].

This also ties in nicely with the results from Antonio Damasio and his card game, applied to acquired sociopaths and psychopaths.  In that game, Damasio had decks of various risk/reward ratios, and participants were given a small amount of money to start.  Their goal was to finish the game with the highest amount of money they could, but they did not know when the game would end.  At the beginning, all of the subjects sampled all of the decks evenly.  However, once they had done so, eventually the non-sociopath/psychopath participants stayed away from the high risk deck, while the sociopaths and psychopaths continued to select from it.  And for them, the results were the same as in the experiment above: they ended up having to “borrow” from the experimenter to keep themselves in the game. What is particularly interesting about this is that Damasio had his participants hooked up to a skin capacitance test in order to measure their emotional responses.  The non-sociopaths/psychopaths had an emotional response immediately before selecting the higher risk decks, which explains their reluctance to select that deck, and their lower incidence of selecting that deck.  The sociopaths and psychopaths, however, were missing that emotional response, and so continued to select that deck. [Damasio, Descartes’ Error: Emotion, Reason and the Human Brain, pgs 212 – 216].

This, then, relates the deficiency in their instrumental learning back to the passive avoidance case, and ultimately back to the reduced anxiety in psychopaths.  It is quite likely that it is some sort of fear or anxiety response that is at work when most people select the higher risk decks, or when they start getting punished for playing cards, or when they start to hear the buzzer.  But for psychopaths, this reaction is either non-existent or at least greatly muted.  And this would contrast sharply with autistics, whose anxiety reactions –  it can be presumed –  are in no way impaired.  Moreover, those anxiety reactions are an important part of both how autistics act in the social sphere as well as their learning how to act in the social sphere; they are motivated to learn by their anxiety and anxiety holds them to the rules they learn.

So this seems an incredibly good candidate for what autistics have that psychopaths do not.  However, what Prinz noted about all the help psychopaths have in attempting to learn about morality still stands.  Arguably, they have far more help in that regard than autistics do.  Is this lack of anxiety enough to ensure that psychopaths do not learn the moral/conventional distinction?

To understand this, it is best to start with the normal practices for learning about morality as children.  Blair identifies two main ways that this can be taught to children.  The first, and the most effective, is to ask the children how they would feel if someone did the thing to be considered bad to them.  The second is to have it imposed in some way by an authority through punishment, where the authority simply punishes them when they do something wrong.  Blair argues that the second is not effective because it tends to associate to fear of the authority as opposed to any real moral learning [Blair, The Psychopath: Emotion and the Brain, pg 42].  It would be clear that that method, then, would not allow a proper learning of the moral/conventional distinction, as it would make morality tied to the authority of the punisher, and not to anything independent of authority.  That being said, even that would seem to be immensely better than what psychopaths currently have, especially if it was associated with a more general authority, like that of the police.  So, then, keeping in mind the deficiencies of psychopaths, can psychopaths use either of these methods to learn the moral/conventional distinction, and morality in general?

When trying to teach morality by asking the child to think about how they would feel if that was done to them, the first reaction would be to tie that method to empathy.  After all, it does seem very much related to understanding what the other person would be feeling and quite possibly feeling it yourself.  As such, the fact that psychopaths have a deficiency in empathy would close this option off from them; they cannot engage emotional empathy to be able to properly gather in how the other person would feel.  However, that analysis would leave autistics out as well.  Thus, perhaps it is not really empathy that is engaged in that method, but is instead some sort of appeal to the feelings and fears of the individual making the judgement, a pure simulation case.  Temple Grandin, in fact, deliberately simulates and puts herself in the position of the person in order to do what she calls empathy [Grandin, Thinking in Pictures: My Life with Autism, pgs 97 – 99].  In this case, then, the psychopathic lack of empathy cannot be a factor, since it is not for autistics.  However, the issues with instrumental learning may be an explanation here.  Are psychopaths able to anticipate harms or punishments to themselves before they occur?  Their issues with anticipation in aversive conditioning experiments and with instrumental learning seem to suggest that they do not.  The anticipation of future punishments does not seem to strike an emotional chord with them, and so they do not get the anxiety or fear responses that can be used to condition against those future harms.  In this case, then, they are unable to learn morality because they are unable to attach their feeling of their own harms to that of the action they are not to commit – if they even feel the harm at all.

The second is more direct.  It involves using a fear of punishment to condition children away from doing immoral things.  But psychopaths have problems applying a fear of punishment to learning what to do or what not to do.  In a real sense, they are not in fact afraid of punishment, even if they still find it punishing.  Because they are not afraid of punishment, they cannot learn not to do things because they will be punished for doing so.  Because they cannot learn not to do things because they will be punished for doing so, they will not even be able to learn by the less effective of the two main ways children are taught morality.  Our psychopaths, then, cannot succeed when taught the way most children are taught about morality.

This eliminates the help they get from parents and teachers.  However, the influence of professionals must still be considered.  Psychopaths, once discovered, are brought in and treated by psychiatrists, psychologists, and social workers.  While some may attempt to use the above methods, it is unlikely that those are the only tools available.  So, then, why do professional attempts fail to teach the psychopath morality?

The first thing to note is that any methods that rely on punishment or the realization of punishment if the psychopath does not either learn what morality is or applies morality in their decisions is not likely to work, for the reasons that have been discussed earlier.  Psychopaths do not properly learn from either punishments or even the anticipation of punishments.  However, there are two techniques that showed promise and yet did not actually result in successful treatment which shall be examined in detail here.

The first is therapeutic communities.  “Hare (1970) suggested that the reshaped socal millieu of a therapeutic community might alter the basic personality characteristics and social behaviour of psychopaths.  Although lacking comparative data for untreated psychopaths, there were several early positive reports … “ [Harris and Rice, Treatment of Psychopathy: A Review of  Empirical Findings, pg 556].  To this end, an intensive community was evaluated.  This community had been held in a maximum security prison for over a decade, and involved group therapy of up to 80 hours a week [ibid].  However, the results were poor.  “… treatment was associated with lower violent recidivism for nonpsychopaths but higher violent recidivism for psychopaths (emphasis in original)” [ibid].   Even in the program, “Psychopaths showed poorer adjustment in terms of problem behaviours while in the program, even though they were just as likely to … to achieve positions of trust and early recommendations for release.” [ibid].  Ultimately, the authors believed that while the psychopaths did indeed “ … learn more about the feelings of others, taking others’ perspective … and delaying gratification” [ibid] instead of using that to allow them to behave better  “ … the new skills emboldened them to manipulate and exploit others.” [ibid].  So, it seems that they learned the skills that would allow them to act more morally, but never bothered to use them for that purpose.

The other approach is cognitive-behavioural therapy, and just like therapeutic communities it showed initial promising results, and yet its usefulness was also challenged in very similar ways.  For example, in a study of sex offenders being treated with cognitive-behavioural therapy “High psychopathy offenders who were rated as showing the most improvement (as measured by conduct during the therapy sessions, quality of homework, and therapists’ ratings of motivation and change) were more likely to reoffend than other participants … “ [ibid, pg 557].  This result should look very familiar.  As with therapeutic communities, psychopaths seem to be able to represent themselves as picking up the skills and knowledge they need in order to overcome their psychopathy.  However, they do not seem to apply that in any situations outside of the therapist’s office.  In fact, studies suggest that not only are they no better off than psychopaths who are not treated with cognitive-behavioural therapy, they are worse than psychopaths who were not treated. [ibid].

These results, then, are very interesting.  While it must be noted that it is unlikely that the moral/conventional distinction is monitored in these studies – which, recall, is the main criteria for performance in this paper – it seems that psychopaths can indeed convince professionals that they are at least showing moral improvement in the treatment programs outlined above.  And in so doing, they get rewarded, with positions of authority and early release.  However, at that point it all goes askew and psychopaths reveal that while they did gain some knowledge from the treatments, they did not really learn a thing about morality.  Or, at least, nothing that led them to act more morally themselves.

So, what, if anything, did they learn?  Taking Temple Grandin’s categories, it seems that one might be able to learn how to classify things into her three categories rationally, perhaps by appealing to actual loss or some other criteria to outline the category of “really bad”.  So, for example, one could dispassionately note that all of the really bad categories involve actual loss from someone (even if that was not the subject), while the other two categories do not, and from there that the “sins of the system” are treated in the same manner as the “really bad” things while not containing any actual loss.  From there, one could easily imagine deriving the authority-independence angle, since what is “really bad” has to hold even if the system that has the sins or the legal authority goes away.  So, while it may be difficult to do, autistics do seem to demonstrate that at least some loose categorization system can get at least a decent understanding of the moral/conventional distinction.  Thus, it is possible that psychopaths could at least improve their performance with that distinction through a building up of categories in the same way, but without relying on emotion in any way.  And it is quite likely that the methods outlined above would, in fact, help them do so.  However, it is abundantly clear that if that occurred, the psychopath’s behaviour would not improve.  So, if their performance on the moral/conventional distinction would improve, but their moral behaviour would not, what are psychopaths still missing?

It appears that the answer is genuineness.  Recall that autistics seem to both understand the distinction and also adopt it as a guide to their behaviour.  With respect to the treatment methods outlined earlier, psychopaths seem to gain the knowledge to perform better, but only seem motivated to perform according to that knowledge when they see an explicit reward for doing so.  Under normal circumstances, psychopaths simply do not and cannot see why they should apply those behaviours, and so they do not.  Autistics, on the other hand, clearly see why they should apply those behaviours.  Why are they different?

And the answer seems to lead right back to instrumental learning: autistics not only can see how acting morally prevents them from accruing negative outcomes, they are also in a unique position where those negative outcomes are particularly salient for them.  Heidi Maibom appears to be correct when she argues that the driving force behind the autistic sense of morality is social concern [Maibom, The Will to Conform, pg 268].  Autistics are well aware that if they do not act according to the rules that everyone else follows, there will be negative consequences.  Autistics desperately want to ensure that they are, indeed, accepted into social groupings.  However, their mindreading abilities are deficient, and they thus quickly learn that they do not understand the rules, roles and conventions that will gain them social acceptance.  Therefore, they must strive to learn all of these distinctions in any way they can, and stick to them.  To do otherwise risks their exclusion, a negative consequence that they desperately wish to avoid.  This, then, is why Temple Grandin faces such anxiety whenever she wants to break even the smallest moral rule: every such break runs the risk of social ostracization, and it is only those rules that prevent that for her.

So, in order to ensure that they remain part of the social grouping, they have to learn all of the relevant behavioural rules and act on them at all times.  And one key distinction there is the moral/conventional distinction.  An autistic who protested that they should not eat in class during the school Christmas party would be laughed at and cut off – at least potentially – from the social grouping.  So the autistic person must learn when rules are simply based on authority and when they are really bad.  Since that is likely learnable, autistics at least learn a basic categorization of that distinction, which allows them to perform sufficiently when faced with that distinction.  But their acting on it is enforced by their understanding that there will be great negative consequences if they do not follow those rules.

Psychopaths, on the other hand, have none of this.  Even if it was the case that they had negative consequences from not following that distinction – which is not normally the case – they are not motivated to learn or act based on punishments applied to them.  Their issues with aversive conditioning and instrumental learning prove this.  However, when they can see a reward for the appropriate behaviour, they do indeed act appropriately.  This is why they can convince all of those professionals that they are, in fact, learning how to be better morally when in reality they are not improving at all.  In a sense,  psychopaths “fake” moral behaviour because they can do nothing else; they can only act morally when they get an explicit reward for acting morally.  Ultimately, they can only fake moral behaviour because there is no way for them to develop genuine moral behaviour.

Comparing, then, psychopaths and autistics to those who are not, we can see that autistics are far closer to the norm than psychopaths are.  While one may get a warm, fuzzy feeling from acting morally good, in general morality stops us from doing things that may harm ourselves or others.  Emotional empathy gives people a direct negative to guide their behaviour; if you take an action that might hurt someone, you may feel at least part of that hurt yourself through empathy.  That would be a negative reaction, and so you will be inclined to not take that action.  This also accounts for why we are more willing to help people we know and care about than strangers; the negative response is stronger in those cases, even as it likely still exists in the other cases.  Failing that, the negative consequences of having people dislike us or even the possibility of us being sent to prison are sufficient to moderate our behaviour, and make us act moral.  Autistics are clearly on the same side as non-autistics in the latter, even as they are generally incapable of the former.  Psychopaths, however, are missing both.

Ultimately, Prinz is right that there is something missing emotionally that contributes to the amoral behaviour of psychopaths.  The first is that they lack emotional empathy, but that in and of itself is not sufficient to cause their amorality.  The second is that they have a deficient fear/anxiety system, which makes aversive conditioning and instrumental learning difficult.  Even if psychopaths had emotional empathy, being unable to be driven by punishments would still be sufficient to cause their amorality.  At the end of the day, the problem with psychopaths is that they have a learning disability in instrumental learning caused by reduced anxiety/fear response that means that they simply cannot learn to avoid negative consequences, and so cannot learn to be moral on the basis that bad things will happen to them if they do not.

So, what does this mean for psychopaths?  Well, it means that clearly simply giving them the knowledge of moral rules like the moral/conventional distinction is insufficient.  Psychopaths need to genuinely adopt that knowledge as the way to best live their lives, but they cannot do so if the only reason for them to do so is only to avoid bad consequences.  Thus, the best approach to treating psychopathy might be to make acting morally directly rewarding, instead of indirectly rewarding by avoiding negative consequences.  This is quite difficult to do in the every day world, since in general morality works to stop us from achieving our maximum perceived rewards because of the negative consequences associated with being immoral.  Morality only matters when it stops someone from getting the most gain for themselves at the expense of others.  Thus, a new way of getting psychopaths to genuinely adopt moral rules is required, although there do not seem to be any particularly good candidates at the moment.

Now, Prinz introduces psychopaths in an attempt to show that emotionism is correct and rationalism is flawed.  At this point, it is clear that Prinz’s specific form of emotionism is not supported by psychopaths, because the difference between them and autistics is not one of moral emotions.  Anxiety for autistics is not a moral emotion in the way Prinz wants it to be.  But can that be used to argue that morality for autistics is still an emotional attitude?  This does not seem to be the case.  Autistics do indeed simply have a set of beliefs that they cleave their behaviour to in order to gain social acceptance, but emotion is only involved as a tool.  It motivates them to learn those moral beliefs and to always follow those moral rules, but what is moral to the autistic is not what makes them anxious; breaking any of the rules might make Temple Grandin anxious, but she still considers the “really bad” a distinct, identifiable category apart from the level of anxiety it fosters in her.

This might, however, lead one to question an internalist account of morality, by recognizing that psychopaths could very well have the knowledge of what is moral but would still lack the motivation to be moral, thus separating moral beliefs from moral action.  However, it seems more likely that there are still some beliefs missing.   Autistics genuinely believe that they should follow those rules; the rules apply to them and it is much better for them if they follow. The psychopath may be able to come to believe that when people talk about “moral” there are certain things they mean by that and that certain rules follow, but the psychopath seems unlikely to be able to come to believe that they, personally, should act morally and that those rules apply to them. What they lack is the understanding that it is better for them to follow those rules and/or that they ought to follow those rules.  Without that, they are not motivated to follow them, but without that it seems unlikely that anyone would have any motivations whatsoever.

Prinz is somewhat right to tie psychopathy to an emotional deficiency, but it is not the sort of deficiency that supports his case.  The difference between psychopaths and autistics is not in the emotions that Prinz would consider moral, but simply in fear and anxiety.  That autistics have functioning fear and anxiety responses allows them to see that they ought to be moral because there are bad consequences for them if they are not, which drives them genuinely adopting moral rules and conventions.  Psychopaths, on the other hand, do not genuinely adopt behaviour on that basis; for them, punishment is irrelevant, and only reward matters.  Since acting morally is not generally rewarding except in the sense that acting morally means avoiding punishments, psychopaths have no reason and no ability to adopt moral behaviour.  Ultimately, they are unable to learn through aversion a field that generally only generates aversive consequences for failure, and a field where violating its rules at least generally seemingly leads to the greater rewards.  In short, psychopaths are amoral because without considering the negatives acting immorally or amorally seems to pay off better.  That it generally does not when the negatives are considered explains why psychopaths so often end up in prison or therapy, with a much worse life than they could have had if they had only been able to overcome their amorality.

Blair, James, Derek Mitchell and Karina Blair.  The Psychopath: Emotion and the Brain.  Oxford:   Blackwell Publishing Ltd, 2005.

Damasio, Antonio.  Descartes’ Error: Emotion, Reason and the Human Brain.  New York: The Penguin Group, 1994.

De Vignemont, Frederique and Uta Frith.  Comment on McGeer in Sinott-Armstrong, Walter (ed).  Moral Psychology Volume 3: The Neuroscience of Morality: Emotion, Brain Disorders and Development.  Cambridge: The MIT Press, 2008.

Grandin, Temple.  Thinking in Pictures: My Life with Autism.  New York: Vintage Books, 2006.

Harris, Grant T. and Marnie E. Rice.  Treatment of Psychopathy: A Review of Empirical Findings.  in Patrick, Christopher J.  Handbook of Psychopath.  New York: The Guildford Press, 2006.

Maibom, Heidi. The Will to Conform  in Sinott-Armstrong, Walter (ed).  Moral Psychology Volume 3: The Neuroscience of Morality: Emotion, Brain Disorders and Development.  Cambridge: The MIT Press, 2008.

McGeer, Victoria. Varieties of Moral Agency: Lessons from Autistics (and Psychopaths)  in Sinott-Armstrong, Walter (ed).  Moral Psychology Volume 3: The Neuroscience of Morality: Emotion, Brain Disorders and Development.  Cambridge: The MIT Press, 2008.

Prinz, Jesse J.  The Emotional Construction of Morals.  New York: Oxford University Press, 2007.

Yirmaya, Nurit, Marian D. Sigman, Connie Kasari and Peter Mundy.  Empathy and Cognition in High Functioning Children with Autism.  Child Development, 63(1), 150-160.

61 Responses to “Fearlessly Amoral: Psychopaths, Autistics and Learning with Emotion”

  1. Learning Theology … « The Verbose Stoic Says:

    […] lots of examples of these, but if you want something accessible you could do worse than my essays on psychopaths, which is empirically minded moral philosophy and aesthetics, which is pure […]

  2. Autistics are Kantians; Objectivists are Hobbesian Egoists … « The Verbose Stoic Says:

    […] As seen in the essay here, people like Heidi Maibom argue that autistics tend to act like Kantians — follow the rules without exceptions — than Humeans — let emotions guide your morality — when it comes to morality. From this, you might get the sense that autistics lack empathy and even, in some cases, compassion. And they at least lack empathy. But that isn’t what Objectivists hold. And you might also argue that autistics have a tendency to be self-absorbed, in the sense that they can be unaware of the world or those that exist outside of their direct experience. But that’s not even “self-interested” as per the Objectivist view, let alone “selfish”. […]

  3. Julia Says:

    Autistics and and people with Asperger’s ARE able to empathize with people they just do not know how to express it or do not care to express it. This idea that we do not empathize is a MYTH. STOP spreading this myth and get it out of your special theories.Read articles on this site and dispel your myths http://www.autismandempathy.com/?p=391

    This site exists to undo the myths about autism and empathy that have stigmatized autistic people for so long.

    It features writing by autistic individuals, by autism parents and family members, by autism professionals, and by others who understand that autistic people, all along the spectrum, can experience the world in highly empathetic and sensitive ways. Telling our stories, describing our experiences, and speaking the truth in our own voices, we can break dehumanizing stereotypes and increase understanding.

    • verbosestoic Says:

      Well, first, I’ll be more than happy to read that site if you can tell me what I’ll find in there that I couldn’t find by reading Temple Grandin’s book and the studies and papers I referenced in this paper.

      Second, I find it odd that you’re associating me with promoting dehumanizing stereotypes when the crux of my paper is that autistics are, in fact, moral (to contrast them with psychopaths). Note that this was challenged in one of the sources I used, which I argued against.

      Finally, there is much evidence that autistics lack empathy as defined by being able to determine what other people are feeling. Autistics have issues with mind-reading in general (meaning the ability to predict and understand other people), and particularly with empathy (the Yirmaya paper is a good source). Additionally, it does appear that autistics have deficiencies in their mirror neurons, which are associated with empathy. That they are deficient in empathy — defined as knowing what other people feeling — seems pretty reasonable. Does this mean that they can’t do it at all? No, and I explicitly granted them cognitive empathy, reasoning out what people were thinking instead of feeling it themselves. Could it be the case that autistics can have emotional empathy for cases they understand well? Yes, and some things Grandin says suggest that. It’s not really relevant for my paper, though, and we would have to be careful to separate empathy from sympathy for the purposes of this paper. I say nothing about sympathy, just about empathy. I am totally willing to concede that what you say is true for sympathy, but not for empathy as I defined it above.

      • Andrea Says:

        You sound very defensive, as if your ego is contingent on your beliefs in what another person says. I have aspergers, and I am a walking mirror neuron receiver for other people’s emotional empathy. I cry when other people hold their grief in and deny there is a problem. Further, the only time I have a problem with cognitive empathy is when a psychopathic-like person or institution has singled me out for abuse. Yes, I have a hard time verbalizing my empathy for someone whose child has died; at the same time I find that a genuine look that says “I’m sorry” and a hug is much more meaningful and less harmful than giving the typical “neurotypical” response, “Well, she’s in a better place now.” What?!? Someone’s kid has just died and “normal” people, a lot of the time, give some superficial response because they can’t look in the bereaved eyes and feel their pain. People who are neurotypical armshair spectators should tend their own gardens and leave the pontificating up to the people who know from experiential learning. That Temple Grandin explained to you what it was like for HER to have autism, that doesn’t make YOU anymore of an expert on the subject.

        Walk a mile — then buckle down and walk ten more.

      • verbosestoic Says:

        It’s less defensiveness and ego and more annoyance over someone wandering over to the site, reading the paper, and accusing me of fostering obvious myths and insisting that I do research despite my having referenced the accounts of people who are autistic AND relatively current psychological research on the topic, and being accused of dehumanizing autistics DESPITE insisting — against counter-argument, you will note — that they are moral and genuinely so. Your comment is not much better, especially since you turn this into a “neurotypical vs neuroatypical” issue … despite you have no idea whether I am or am not neurotypical. (For the record, I don’t have any explicit diagnosis of anything but am absolutely NOT neurotypical, so thanks for that). The comments don’t seem productive, and seem to be nothing more than getting upset and angry over something that I probably didn’t really say.

        Let me try to turn this into a more productive discussion. I note that you seem to imply that I was saying that autistics have issues with cognitive empathy, by pointing out that you generally don’t have a problem with that except in certain cases. However, as I replied to Julia, in the paper I’m explicit that autistics do generally have cognitive empathy. That’s the point of the Yirmaya paper. What I say they lack is what I called “feeling” empathy, but perhaps I should have called it “simulated empathy”, where you end up feeling what the other person themselves are feeling because you use the background simulation method to figure out what they are thinking, and that method involves putting yourself in their place and running things forward, which would include the emotions at the time. If mirror neurons do anything, it is to facilitate this form of “mind reading”. However, autistics are indeed known to have problems with this sort of mind reading, and with determining what other people are feeling. It’s one of the diagnostic criteria, I think, or at least is very common. The Yirmaya paper showed that high functioning autistics do much better at this than those who aren’t, and that this ability scales with IQ which is NOT the case for those who are not autistic. Thus, it implies that autistics tend to reason out what other people are thinking, and so better reasoning ability matters for their ability to do that sort of empathy — ie knowing what others are thinking — which is not the case for those who are not autistic … and simulation/mirror neuron methods would not require as much reasoning, so this all makes sense.

        This does not mean that autistics don’t care about other people, and other people’s feelings. They do. I was a bit short in just saying that they care about others and instead linking it to a social fear, but arguably the concern for others ALWAYS comes from something like that, especially in light of how psychopaths completely lack that sort of concern. Essentially, autistics are moral because they genuinely care about the needs and desires and feelings of other people and psychopaths don’t, even as psychopaths are often better at knowing what those are (so that they can manipulate them).

        Taking your example, what you do is a perfectly wonderful expression of sympathy that will make people feel better. You want to make people feel better, so you do it. You also are likely fairly good at figuring out what situations will cause what emotions, and that sort of reasoning can indeed trigger emotions in you, often of sympathy: you may not feel grief yourself, but feel bad for them because you know that they are, indeed, grieving. After an incident in my past, a woman noted that while I didn’t grieve myself, I did seem to feel bad for her, who was grieving, because I didn’t know the person as well as she did. Simulated empathy would be feeling grief, and not just sympathy knowing that they are grieving. But you also don’t seem to understand that “Well, she’s in a better place now” is something that many people DO find comforting, child or no. They say it to THEMSELVES to comfort them, so they also say it to others. You find it superficial, and so won’t say it, but that is indeed something that is different from most people. In short, you get your empathy WRONG on that count for a number of people.

        To put it back in the context of the paper, you would not say that not because you don’t want to comfort people, but because you don’t know that that is actually comforting to many people. The psychopath won’t say that not because they don’t know that it’s actually comforting, but because without some other reward they don’t really care to comfort anyone. The autistic is definitely better in this situation.

        In short, the main points of the paper wrt autism are:

        Autistics are capable of morality.
        They are capable of morality because they genuinely care about how their behaviours impact others.

        I don’t see what is so bad about saying that.

  4. Scientism 101: Philosophism and Religionism … « The Verbose Stoic Says:

    […] concerns can help us tease out ethical concerns, then there’s no problem with doing that (My essay on psychopaths does just that). And if concerns about things like epiphenomenalism may raise issues for straight neuroscientific […]

  5. Controversial Controversies? « The Verbose Stoic Says:

    […] by my post about some complaints that seemed to not be about things that were unreasonable, my essay on the morality of psychopaths and autistics and rounding out the top 5 my response to calls asking theists what would convince them that God […]

  6. Héctor Muñoz Says:

    Psychopaths cause great social and personal harm because they will steal, lie and hurt other people whenever they think they can get away with it. They are also very manipulative and will exploit every person they can trough deception.

    I think the society needs to be warned about these individuals. Flagging psychopaths might be a way to control the consequences of their amoral way of being.

  7. Viktor Czernin Says:

    Unless you’re willing to offer a justification as to why moral facts objectively exist, it might be more intellectually precise to use neutral words like “internalize the moral/conventional distinction” rather than “learn…” That one can “learn” something implies that it exists, and if you’re not prepared to justify that, it would be dishonest to use the word.

    A sort of agnosticism to the existence of moral fact is probably the only stance that agrees with the precepts of empirical science. Especially if you’re going for a descriptive approach.

    Other than the linguistic nitpicks, stimulating stuff. Don’t agree with all of the conclusions, but they are compellingly argued.

    If we focus more on the normative aspects of moral philosophy, psychopathy poses a fair problem for eudaimonists and rational egoists. If the statement “their life would be improved, on the whole, by adherence to altruistic rules” doesn’t hold true for a given person, what can we say about morality for them? Do we conclude that altruism is immoral for psychopaths? Quite counterintuitive, especially in the context of the idea that moral judgment reinforces moral behavior which reinforces the individual and the social good. The relationship seems to break down into a muddle of inelegant (yet logically sound) deceptions. Example – it’s morally right for the psychopath to con the old widow, but it’s morally wrong for us to publicly announce this moral truth and thus reinforce the psychopath’s harmful (to us) behavior. On the other hand, it is morally right to generally maintain open and truthful communication with other people. Ugh.

    Needless to say, this particular mental deviance is a minefield for any purported rationally compelling morality. A fruitful minefield, yes, but a minefield nonetheless.

  8. verbosestoic Says:


    Sorry for not replying earlier, but things have been quite hectic the past few weeks, especially this past one. Things have settled down a bit now.


    I don’t need to prove objective morality here because my claim is not that psychopaths fail to learn any kind of “right” morality, but that they don’t even learn, naturally, the moral/conventional distinction that everyone else — including autistics — learn. Whether that division of rules in moral and conventional is arbitrary or whether those moral rules are any less conventional isn’t really a concern here. The key is that due to their inability to learn from negative conditioning psychopaths just don’t pick up the distinction, and moreover even if you teach them that distinction they never internalize it, which explains why they are amoral and autistics aren’t.

    As for morality, I am an objectivist but think that science has little to do with morality, mostly because I deny that morality is descriptive (I am firmly on the normative side of the line).

    As for the last part, you are right that psychopaths cause major problems for moral theories that are based on well-being. However, there are many moral theories that don’t do that. Kant, for example, would have no problem with psychopaths and would probably simply declare them incapable of being moral. The Stoics, as well, would have no issues with psychopaths, and both of these insisted that they were describing THE rationally compelling morality. Virtue Theories also won’t have too many issues with psychopaths beyond perhaps claiming that they are incapable of getting the virtues. So, if that minefield ends up blowing up well-being theories, we aren’t licked yet.

  9. I Lost on Empathy … « The Verbose Stoic Says:

    […] suggesting that people with an ASD don’t have empathy. This was also mentioned in a comment on my essay about morality, psychopath and autistics. However, that essay clearly establishes that science says that, yes, they do have problems with […]

  10. Phil Mathers Says:

    I have finally got professionals to acknowledge that my son has difficulties and they are intimating Asperger’s. His mother will not accept this. I always thought she was a psychopath but having read yours and other information sites I am now not sure if she is also autistic although she is very maniplulative and “evil”.

    Is it likely that a psychopath mother would have an autistic child. My son certaily seems more of a normal devcent person than his mother. Mother father and grandson all seem to have similar symptoms.

    How key is inheritance?


  11. verbosestoic Says:


    Well, I’m a philosopher and not a psychologist, so I might not be the best person to answer this question. You can look at some of the sources in the paper to see if they give better answers. That being said, it seems that autism and psychopathy have different causes, so it isn’t that likely that they are the same, and they both relate differently to moral questions, so that sort of heredity doesn’t seem to match much. We don’t know entirely what causes either psychopathy or autism, though, so we can’t say for certain.

    You can be manipulative and evil and not be a psychopath, though, so it’s more about the specific symptoms than just having bad behaviour. The “conventional/moral” distinction, it seems to me, is one of the key differences, because people who are truly “evil” would know that the actions are morally wrong but wouldn’t care. A lack of emotional empathy might help with this if one was never given a reason to care about others … but that would require them to be in a position to get what they want no matter how badly they treated others, which in and of itself can lead to that sort of manipulative and “evil” behaviour.

  12. Jesse Lykken Says:

    My father, David Lykken, believed that psychopaths could in fact be socialized through effective parenting. He would often illustrate this through his experiences raising bull terriers. One of his uncompleted projects was to develop a screening tool to identify psychopaths at an early age, and that once identified, utilize effective parenting techniques (house parents) to socialize these children. A socialized psychopath could make a wonderful soldier, astronaut, explorer or president (TR) for that matter.

    • verbosestoic Says:

      Part of that is covered where I talk about CBT and its impact on psychopaths. The problem is that it seemed to generate the right behaviour in the specific cases but outside of that it not only didn’t improve their chances of reoffending, but increased it. I posit that in those cases the psychopaths still don’t genuinely adopt the rules as real important rules, but adopt them as behaviours that it benefits them to adopt in those cases, to be forgotten outside of that. Autistics, on the other hand, adopt them as moral rules genuinely and so apply them in all situations … and, in fact, in cases where they DON’T apply. So I guess the difference is that psychopaths apply them because it seem to work for them in certain cases, while autistics apply them because they accept them as rules that must be followed.

  13. Jesse Lykken Says:

    I am David Lykken’s eldest son, and actually was a test subject of the aforementioned electric shock test. I also score quite high on the Psychopathic Scale (although am not diagnosed as a full psychopath) while my best friend is a psychologist and a diagnosed psychopath. Furthermore, I am the parent of three autistic children. Some quick points and observations. Re. the lowered anxiety on the shock test. I obviously knew when the shock was coming, but I knew the pain would not injure me, so why worry? It was just pain. As far a morality goes, the whole subject is so subjective. My friend and I have chosen our own individual codes of behavior and live by them quite strictly. These codes are by in large not “instinctive” but rather learned/copied from other individuals that we respect, such as our parents. Is that so different from what typical folks do? I do have an instinctive reaction to bullying behavior, which includes child/animal abuse. If I were to see somebody violently abusing an animal or a child, I would most likely kill them. Cheating/stealing/sexual mores are all copied from respected adults. Two of my own autistic children seem to have moral codes hard-wired in their brains. They never, ever, cheat, lie, steal. Just not an issue. The third child, the most profoundly autistic, is manipulative, will lie and steal, and basically do whatever she has to to obtain my doughnut (yes, she invariably gets it, being smarter than I). As far as a psychopath’s ability to “learn moral behavior”, learning is one thing, demonstrating it is another. Understanding how typical humans behave is just another tool in the toolbox. Replicating it is another matter. Think of acting on the stage. Some can, some cannot. Same with psychopaths aping human behavior.

    • verbosestoic Says:

      Welcome back.

      For some of your comments, you might want to read the Damasio reference. For example, in the card test he demonstrated that when picking the higher risk deck most people had an emotional reaction even though, presumably, they thought it was the right thing to do. This sort of reaction is, I think, what would be lacking in the shock test and something that most non-psychopaths would have, even if they used the same reasoning you did. Also, this seems to be the case for making decisions, as Damasio gives an example of a person who is emotionally impaired who cannot decide what day to schedule the next appointment for, even though they obviously don’t care about it. Moral dumbfounding also seems to indicate that most people decide what is and isn’t moral based not on a set of rules, but on a set of instinctive reactions, which also explains why autistic moral reasoning is far more rules-based than that of others.

      As for the last point, remember the main thesis of this paper: that the difference between autistics and psychopaths is that autistics genuinely adopt moral reasoning while psychopaths only adopt it as a means to an end. Thus, it’s only when psychopaths see acting morally as benefiting them that they do it, while autistics adopt it “wholeheartedly”. So reasoning ability and the like would play into being able to “fake” being moral better or worse, true, but that is in line with my thesis, I think.

  14. Phil Mathers Says:

    Jesse’s comments and my observations seem to indicate that psychopaths and autistics perhaps have the strength to say no or not comply with peer pressure when others go along with it to keep the peace. If so maybe they have a better outlook on life that the rest?

    Phil Mathers

    • verbosestoic Says:

      That doesn’t seem to be the case. Autistics seem to care about what other people think — which is the main reason people give in to peer pressure — but usually don’t know how to actually act according to the social rules. See Temple Grandin’s work for more on that, I think. As for psychopaths, they in general don’t seem to care about what other people think because they don’t, in general, seem to care about people. But that’s hardly a good outlook.

      Note that the underlying presumption here is that not complying with peer pressure is a GOOD thing, in general. The problem is that it isn’t necessarily so, because “giving in to peer pressure” is the other side of the coin to “follow the unspoken social rules”. And what we can see from psychopaths and autistics is that being unable to understand or follow those unspoken rules causes them much grief when the rules really do need to be followed. Surely you’d agree that while it is bad to jump off a bridge just because everyone else is doing it that not following the rules about, say, not picking your nose in public is, in general, a bad thing.

      (As an aside, the comment would have shown up earlier, but I thought it was already approved.)

  15. Peta Barber Says:

    Interesting. I’m a psychologist working with an adolescent who after being taken into Public Care for being out of control as a 9-year old was subsequently diagnosed with Aspergers. This has somehow never felt right to me and I have long experience of young people with ASD. Now aged 16, he is at high risk of offending, aggressively and/or sexually if not closely supervised. I m thinking I will explore your thesis and see if if I can assess his fear/anxiety response to moral behaviours. I am also considering genetic factors – his natural father has been diagnosed schizophrenic – and early developmental factors – a mother with learning difficulties who suffered from post-natal depression, so an Attachment Disorder might be in there as well. A complex case study.

  16. What Sam Harris Gets Wrong. | The Verbose Stoic Says:

    […] people side with the Utilitarians, but there is a group of people who don’t: autistics [see https://verbosestoic.wordpress.com/fearlessly-amoral-psychopaths-autistics-and-learning-with-emotion/ for details]. They tend to side with saying that it isn’t right for the parent to steal the […]

  17. Joshua White Says:

    Hi Verbose Stoic. I go by Brony on FTB and here is my reply to your comment on Ms. Benson’s blog,
    …for some reason I can’t get wordpress to let me use the name I use there.

    I apologize for the lateness (busy weekend and ADHD makes parsing a detailed blog post more challenging). I will also post this to your blog to make sure you see it. I tangented into autism a lot, I hope you don’t mind.

    Before getting into these issues, though, it must be established that autistics really are moral and psychopaths really are not.

    I agree in the colloquial sense. But My mental model of “moral” involves multiple elements and I might not be contrasting with you in my #9, but rather using a different way describing things. Either way I don’t mind the opportunity to get more accurate (thank you for that) and I will try to contrast what I think what you are saying and see what you think. I do need to get better at the proper jargon and rely on my own ways of describing what I see in the literature.
    I see “moral behavior” breaking down into:
    *The emotional felt states that allow accurate read-outs of a situation (context, persons…), and accurate targeting of responses.
    *The ability to mentally model oneself and other people in social situations.
    *The ability to paint the relevant feelings onto the mental model such that other actors are accurately felt as if they are moral proxies equivalent to oneself. What are called mirror neurons systems are important here because the self is constructed by mapping the body fully into interactions, and “theory of mind” is essentially being able to connect the “self map” to simulations of other people.
    The second one is where I think what I referred to as “strategic empathy” lies. Psychopaths can form mental models of social situations more accurately than autistics, but they lack the emotional targeting systems so they have limited accuracy. The more successful and harmful ones are that form the social stereotypes of the “master psychopathic manipulator”.
    Autistics I see as less able to model social situations, but they have the emotions and honestly “mean well”, but have to work harder to make social mindware. While a lot of what gets discussed around here involves intent not mattering as much as reality when it comes to emotions, I personally allow room for intent when it’s just me because of autistics and people like myself (it’s harder to balance in the larger social issues around here).
    That last one is where my own empathy issues can lie. The literature describes individuals with Tourette’s as having trouble inhibiting personal perspective. I can do it and quite successfully create very strong empathy, but it’s more deliberate than for most. It’s like we have an internal [selfother] dial that is turned back and forth contextually and I have to make switching it a familiar routine. Essentially where the “moral emotions” are present in autism, and absent in psychopathy, I believe they howl in my head like a hurricane which creates its own precision problems. Sometime that sort of “on/off” choice makes me concerned.

    Moral rules hold independently of any authority, while conventional rules do not. Thus, asking someone if the action would be okay if the most directly relevant authority said it would be okay – like, say, a teacher in a classroom – is a good way to test for this distinction. Psychopaths do not pass this test…
    …adult criminal psychopaths respond as if both moral and conventional rules are authority-independent, while psychopathic children answer as if neither are. Prinz points out that Blair suggests that the psychopathic criminals are trying to make it appear like they understand what things are moral and understand why moral rules are bad, but fail miserably, while the psychopathic children are perhaps giving a more genuine reply as to how they view the moral and the conventional.

    This and the surrounding seems to match up with what I was saying about missing the emotions that normally allow implicit understanding required for successful creation and targeting of moral social behavior. Since they have to try to create moral behavior purely intellectually it ends up being “strategic” and not “felt”.

    They bring up two issues. The first is that while autistics can identify faux pas, they may not properly understand them. When asked why one should not commit faux pas, they appealed to rules instead of the pain of the victim. The second issue asks if the autistics were really treating moral rules as authority-independent, or if they were just appealing to another authority than the one presented in the test (so, say, instead of it being dependent on the authority of the teacher it was dependent on the authority of their parents, even though they were not present) [De Vignemont and Frith, Comment on McGeer, pg 277].

    Which rules? Because one distinction that I have brought up is that there are multiple “rules”. In recent arguments there have been appeals to the rules of logic and rationality (learned patterns), but our emotional systems also run off of rules. It might be like an emotionally meaningful, but intellectually void waving at “rules”.
    Alternatively if they mean either moral/conventional rules, this could be a case where the appreciation of the “moral” thing to do is stuck in the implicit realm more strongly, and autistics have a problem moving moral mindware from the implicit to the explicit where it can become more versatile.
    Another possibility is that psychopaths are able to model physical behavior with intellectual effort so can “fake it” better in terms of sounding convincing in speech, but fail when being observed practicing moral behavior (so maybe my term should be “attempted strategic modeling of empathy”). Autistics on the other hand have the emotions but can’t model the behavior as well so they seem socially awkward, but get moral behavior more correct based on what they can model. Your aside on Temple Grandin (I really really like her) seems to support this as she discusses “really bad”, “illegal but not bad”, and “sins of the system”. Autistics are moral on the most critical matters, and are generally able to intellectualize through the rest of it unlike psychopaths.

    This could, then, suggest that autistics are substituting a cognitively based empathy for the more emotionally based one, which allows them to learn morality where psychopaths do not. However, this is belied by the fact that psychopaths are generally both manipulative and successful at manipulating others.

    I think this and the following is also related to the feeling and intellectualizing of morals but poor modeling of interactions of autism (emotional empathy), contrasted with absent feeling of morals and intellectualizing on interactions in psychopathy (cognitive empathy). We could be looking at two parts (with their own systems for contextual adjustment) of the how the mirror neuron system connects the external simulation to the self with application of emotions for proper formation, selection, and targeting of moral mindware. I disagree that autistics lack the emotional empathy, I just think they have problems functionally connecting it to cognitive empathy (any autistic is now encouraged give me their perspective).

    Could they be using a different emotion or set of emotions to achieve the same results?

    Same emotions but different ability to attach them to models of social reality and the self is I think a better interpretation up to this point in your post.
    But this does not preclude there being actual differences in what emotions are present or relative proportions (I need to get into the literature on the anatomical correlates of emotions more). The anecdote you provide from Temple Grandin bears more consideration, but it still speaks of feeling socially contextualized emotional responses (anxiety about breaking rules in this case). If psychopaths don’t feel that socially contextualized anxiety, they are not driven by more than intellectual desire to learn the rules and instead try to act empathetic and moral in a socially strategic sense.

    But the problem is still one of general performance. Since psychopaths can generally learn, why can they not learn this well enough to fake it?

    Because they don’t feel it and feeling drives learning if one is learning on their own. I imagine most psychopaths are very isolated when having to act morally and might carefully limit learning moral rules to situations where they can pick it up but might not have to use it. This argues for increased socialization of psychopaths, but you refer to “therapeutic communities” and CBT being ineffective later. That is unfortunate and the example also highlights my view that teaching people a social pattern makes it a tool that can be used for may purposes. Psychopaths need a proxy for the emotional signals in social contexts, and more focus on reward instead of punishment as you suggest later in your piece (this might annoy some people but if the science suggests a solution it is what it is). What you refer to as “genuineness” is the appearance of the emotional felt states in the stream of consciousness. Autistics feel them but have trouble functionally actualizing them.

    Blair cites a study by David Lykken which paired a buzzer with an electric shock to try to see if psychopaths would start sweating in response to the buzzer, which is the normal response to the electric shock. The reactions of psychopaths were impaired with respect to the sweating reaction [Blair, The Psychopath: Emotion and the Brain, pg 49].

    “Embodied cognition” comes to mind. I explain a bit about how emotions are thought to be elaborations of felt body states here. “Self Comes to Mind” by Damasio is a good source.
    The reduced anticipation of negative events speaks to a lack of mindware from the lack of emotions to me. Similar for instrumental learning. This is interesting in the context of four ways of modifying behavior (positive/negative punishment/reward. Note: positive= addition of something to environment, negative =removal of something from environment, punishment= decrease in specific behavior, reward= increase in specific behavior) and I do think that different cognitive shapes are differently affectable by them.

    Temple Grandin, in fact, deliberately simulates and puts herself in the position of the person in order to do what she calls empathy [Grandin, Thinking in Pictures: My Life with Autism, pgs 97 – 99].

    If empathy is a phenomena with more than one component, I don’t see how this is not empathy, or at least a form of it. If she can successfully create the ability to see another as herself then I think we need more complex language for discussing empathy.

    • verbosestoic Says:

      Since you’re referencing more of the essay, I’ll reply here instead of at BaW. I’ll also hop around a bit to try to get a handle on what you’re arguing here.

      I disagree that autistics lack the emotional empathy, I just think they have problems functionally connecting it to cognitive empathy (any autistic is now encouraged give me their perspective).

      Well, the point of my referencing the Yirmaya paper was to demonstrate that, as it showed that their ability to simply predict the emotions of others scaled with IQ when it didn’t for non-autistics, which seems to me to suggest that in general we use simulation to get empathy (which wouldn’t strongly scale with IQ), but autistics can’t, so they substitute it with theory, which DOES scale with IQ and isn’t as good at it. Add in that autistics do seem to have mirror neuron deficiencies and problems with pretend-play and we have pretty good reason to think that they’d have issues with simulation, and simulation seems to be the way we’d get to really FEEL someone else’s emotions, and therefore have emotional empathy. So it seems reasonable to say that autistics have cognitive empathy but don’t have emotional empathy … but then psychopaths ALSO seem to have cognitive empathy so having that and lacking emotional empathy doesn’t seem to be the reason why autistics are capable of morality and psychopaths aren’t.

      This and the surrounding seems to match up with what I was saying about missing the emotions that normally allow implicit understanding required for successful creation and targeting of moral social behavior. Since they have to try to create moral behavior purely intellectually it ends up being “strategic” and not “felt”.

      Of course, that seems to be as true for autistics as for psychopaths, given the claim that they tend to be more Kantian than Humean — as referenced by Maibom — and thus will say that stealing is wrong even if it is to feed their children, and also that they appeal to moral rules instead of harm to others in justifying a moral conclusion. But none of this would imply that they aren’t moral, and we’d still need to explain why psychopaths can’t learn that as well. That’s why I argued that the difference is that autistics care about getting the rules right and about others, and psychopaths don’t.

      I see “moral behavior” breaking down into:
      *The emotional felt states that allow accurate read-outs of a situation (context, persons…), and accurate targeting of responses.
      *The ability to mentally model oneself and other people in social situations.
      *The ability to paint the relevant feelings onto the mental model such that other actors are accurately felt as if they are moral proxies equivalent to oneself. What are called mirror neurons systems are important here because the self is constructed by mapping the body fully into interactions, and “theory of mind” is essentially being able to connect the “self map” to simulations of other people.

      This sounds like a claim that moral behaviour just IS empathetic behaviour, which I don’t think you’ve supported and, in fact, strongly disagree with. I don’t think these are necessarily necessary to moral behaviour, let alone identical to it.

      Because they don’t feel it and feeling drives learning if one is learning on their own.

      But why is learning morality such a problem while learning, say, physics isn’t for psychopaths?

      What you refer to as “genuineness” is the appearance of the emotional felt states in the stream of consciousness. Autistics feel them but have trouble functionally actualizing them.

      I disagree. I don’t think they feel the emotions of others at all, or at least generally. I think the motivating emotion is social anxiety, which explains why autistics tend to often be even MORE moralistic than others, always following what they identify as the moral rules when others will bend or break them as circumstances suggest. Psychopaths might feel the negative emotions, but they don’t learn from negative emotions and don’t alter their behaviour based on them, because of their inability to feel anticipatory fear.

      The reduced anticipation of negative events speaks to a lack of mindware from the lack of emotions to me.

      Well, one emotion in particular: fear. Which autistics still possess.

      If empathy is a phenomena with more than one component, I don’t see how this is not empathy, or at least a form of it. If she can successfully create the ability to see another as herself then I think we need more complex language for discussing empathy.

      The key is that it’s the other way around: she can see herself as someone else but that doesn’t mean that she can see another person as herself. The idea in that paragraph was to demonstrate that she can potentially generate the emotions in herself when she deliberately and consciously reasons it out, but lacks the ability to do it automatically. In short, she HAS to do it consciously because she doesn’t do it automatically. But it is that automatic mechanism that I call “emotional empathy”, and so what she does is not feel the emotions of others but instead tries to explicitly generate an emotion in herself to do the work for her, which may not be the emotion they are feeling at all, nor is it attached to it directly.

      The summary is this: while both psychopaths and autistics seem to be capable of cognitive empathy and deficient in emotional empathy, autistics can still be moral and psychopaths can’t. So lacking emotional empathy doesn’t seem to make you immoral and possessing cognitive empathy doesn’t make you moral. Thus, there’s nothing left that can be called empathy — at least as a psychological mechanism; you can have a philosophical view that uses that — that makes someone moral or immoral. Thus, morality is not about empathy.

  18. The Professors Who Most Influenced Me Philosophically. | The Verbose Stoic Says:

    […] to hold the exact opposite view than me), and was instrumental in my discovering the link between simulation and theory with regards to empathy. Ultimately, being Stoic-leaning and distrusting empathy-based moralities started with her, for […]

  19. Facts, Values, Opinions and Morality. | The Verbose Stoic Says:

    […] Presumably, this is a rule that Coyne would impose on any class he teaches: you cannot copy assignments. But if there isn’t a moral value backing to the rule, then all Coyne would be saying is that the rules say that you can’t copy assignments, not that it is a bad thing to do, even if he tries to argue reasons for it. So someone who copies assignments has merely broken a conventional rule of the class, one that is only imposed on the class by the overwhelming authority of the professor. Essentially, this reduces all of the rules that we consider moral to be merely conventional … and as we’ve seen, the main way that psychopaths differ from others is that they consider all mora… […]

  20. A Criminal Mind … | The Verbose Stoic Says:

    […] but has a brain condition that means that they get an uncontrollable urge to steal. 2) A psychopath who is incapable of understanding that stealing is wrong. This maps to Coyne’s example of the sociopath, but the sociopath is someone who can know that […]

  21. dollhouse scientist Says:

    Thank you for sharing your thoughts here. I’ve read through this with much interest.

    I have my own informed views here, but first I wish to make sure I understand your position. Are you maintaining that people on the autism spectrum don’t experience affective empathy any more than psychopaths (if by a different mechanism), but act in a (conventionally) ethical manner because of strong anxiety barriers raised at the thought of violating poorly intuited social rules?

    If I’m misunderstanding your position, then my apologies, and may I please ask for correction or clarification?

    If I do correctly understand your position, I’d like to dialogue, as I’ve spent several years researching closely related issues and would greatly appreciate the opportunity to compare notes.

    • verbosestoic Says:

      Essentially, yes. Those who are not autistic and not psychopathic learn moral rules and hold them genuinely because of the negative feedback from affective empathy. Autistics lack affective empathy, but learn moral rules and hold them genuinely because of the negative feedback from the violations of social rules. Psychopaths do not learn them and do not hold them genuinely because, in general, they can’t learn from negative feedback.

  22. dollhouse scientist Says:

    Okay, thank you. I’ve been reading this stuff for years and I’m also far from certain that the mainline theories on how autism and psychopathy compare are accurate. Just to start with, too much research on autism generalizes from children and too much research on psychopathy generalizes from prisoners.

    On autism and affective empathy. The mainline story is that autistics possess weak cognitive empathy but a normal distribution of affective empathy. But I’ve seen the claim that autistic people experience *elevated* affective empathy (http://www.thedailybeast.com/articles/2009/05/11/a-radical-new-autism-theory.html) based on an “intense world” model of autism. And I’ve seen views like yours, that autism spectrum people have reduced cognitive *and* affective empathy. Simon Baron-Cohen seems to hold this view.

    Do have any specific arguments as to why the mainline or intense world theses are wrong? I’m personally unsure people are correctly drawing the line between the subjective experience of moral emotions (guilt, shame, sympathy, love, compassion) and the experience of these emotions in involuntary simulation of others’ emotional states. Do you think autism spectrum people get hit with guilt and shame when they violate (1) social rules or (2) their own conscious ethics?

    • verbosestoic Says:

      Well, I have to start from the idea of mind-reading. There were two main theories of how we go about figuring out what other people were thinking. The first is simulation theory, where we essentially put ourselves into that situation and run it forward to see what WE’D think in that situation, which we then apply to other people. The second is theory theory, where we reason out what we think the other person would be thinking. Note that this could all be done subconsciously.

      In the referenced Yirmaya paper, it was noted that while in non-autistics there was no correlation between IQ and empathy, there WAS a correlation for autistics, where the higher the IQ was, the better they were at it. While simulation doesn’t require a lot of reasoning — you need some to set-up the parameters, but beyond that it’s just imagination — for theory the better you are at reasoning the better you are at it. This indicates, to me, that non-autistics simulate, while autistics can’t simulate and so rely on theory to do empathy (meaning, to come to know what others are feeling). This ties in with their lack of pretend play, mirror neuron deficiencies, and even their failure at conservation tasks (as that requires you to imagine differing situations and compare them). Psychopaths, in general, don’t show these specific deficiencies and behaviours and correlations.

      So, to apply this directly to the question, we need to know what we mean by affective and cognitive empathy, which I myself used too loosely in the essay. It is quite possible that AFFECT is stronger in autistics than in those that are not, which means that then once they do get triggered to feel what someone else is feeling it strikes them quite hard, but if they aren’t simulating they won’t get that trigger as often as those who are not autistic, who ought to get it every time they simulate (ie if you put yourself in that situation and run it forward, you’ll likely feel the emotions as well). So, if you mean affective empathy to simply mean that feeling, then there’s no real disagreement here, but that would imply that cognitive empathy is merely knowing that other people are feeling, which then is where autistics fail.

      Essentially, the data seems to show that autistics have a harder time figuring out what other people are feeling, and thus coming to know it. I’d potentially argue that psychopaths are at least as good at that as autistics are, because they’re so manipulative, and being manipulative requires that … but maybe autistics COULD be that manipulative except they’re afraid of the consequences if the manipulation fails, and psychopaths, again, don’t seem to learn from negative stimuli.

      So, to answer the last question, I’d say “both”. But they try very hard to learn which of 1) are really violations and which aren’t. Also, research with the moral/conventional distinction shows that it’s not that easy to say that an internal conscious ethic isn’t also derived from social rules.

  23. dollhouse scientist Says:

    Greatly appreciate the complement of taking the time to explain your position in detail. This is helpful.

    Yes, the above model matches both my experience and understanding of the science. We’re on the same page here. The only caveat I’d make is to emphasize that all forms of human neurodivergence are matters of degree. That is, autism spectrum people are only *relatively* incapable of cognitive empathy; few are completely incapable of simulating the emotions of others if it closely matches a familiar experience. Conversely, neurotypical people are only *relatively* capable of cognitive empathy. Neurotypical people don’t *automatically* simulate – consider the degree to which men and women routinely fail at cognitive empathy with each other. In my experience human empathy in general is cognitively weak outside familiar social circumstances and affectively weak outside the in-group. We’re all pretty autistic and rather psychopathic towards the “other”.

    Now, on a related issue, where I’m not sure if we’re on the same page. You suggest that the primary reason autistic people don’t act (conventionally) morally is because of anxiety raised at the thought of violating poorly perceived social boundaries. Now, it certainly fits my experience this is *a* phenomenon, and a significant one, in autistic moral behavior. Yes, the confusion and fear is there, and there’s definitely corrective over-politeness and scrupulosity caused by anticipatory anxiety. However, I’m pretty sure that this isn’t the only thing going on. If it was, autistic people would probably act like neurotic and paranoid psychopaths, Hobbesian egoists kept rigidly in check by confusion and anxiety. But as a general rule they don’t, at least as far as I’ve known.

    May I ask what your personal experience with autism spectrum people has been like? I’ve known at least twenty or so variously autistic people in significant depth, including personal relationships with at least two on the high-functioning end. My impression is people deeply motivated by positive moral passions, including a willingness to direct transgress social norms perceived as unethical. Several have been extremely expressive about the emotional value of caring for friends and society in general. A former partner of mine once “adopted” a homeless man. I think he desperately wanted someone to care for, so the consequence of his difficulty in successfully making connections was to attach to someone who needed him too badly to say no. He *visibly* twinged in agony in the presence of the guy suffering. Or: on another occasion, the same ex was offered a position as shift manager at his workplace. He agonized for days over the issue, as taking even a minor position of authority over his coworkers violated his anti-capitalist convictions.

    I could easily provide further examples, but I have a high degree of certainty that this is moral behavior and it’s not motivated by anticipatory social anxiety. Obviously, I’m not claiming that *all* autistic people are highly affectively empathetic, but those I’ve known seem to average at least normal in subjective social emotional affect in a manner which is directly applied to their practical moral behavior. Is this data consistent with your theory?

    • verbosestoic Says:

      Neurotypical people don’t *automatically* simulate – consider the degree to which men and women routinely fail at cognitive empathy with each other

      Actually, to me that’s an argument FOR people automatically simulating, because simulation fails when the other person is sufficiently different from yourself that you can’t set up the parameters properly. That simulation and therefore generally empathy fails in these cases is one of the reasons I oppose empathy-based moralities.

      You suggest that the primary reason autistic people don’t act (conventionally) morally is because of anxiety raised at the thought of violating poorly perceived social boundaries.

      Actually, I think that autistics do, in fact, act morally and DO, to a large extent, get conventional morality. Where they differ is that they tend to do morality based on rules rather than on strict conscience, but that’s related to the simulation vs theory part, not the social anxiety. In LEARNING morality, non-autistics (who are not also psychopaths) tend to learn it, I argue, from the negative feedback of the simulation, meaning that they consider doing something, simulate responses, and feel BAD if it would make the other person feel bad. Autistics, on the other hand, learn it either from social backlash or from the anxiety caused by considering that there will be social backlash, which is also a negative stimulus. Whatever psychopaths have, they can’t learn from negative stimuli regardless.

      Once autistics learn the rules, then they insist on following them and so feel very anxious when those rules are violated. This is, in fact, completely in line with your examples and, in fact, with the studies that show that autistics, in answering questions like “Is it moral to steal bread to feed your starving children?” tend to respond in a more Kantian or deontological way by saying that it’s still wrong because it breaks the rules, while non-autistics tend to allow it, likely because they empathize more with the parent than with the store owner … which is ANOTHER reason why I oppose empathy-based moralities, because the emotions of someone can be wrong, and so using that as your basis either leads you to immoral outcomes or else is something you have to constantly check with something else — typically reason — anyway.

      So, again, once the rules are learned autistics react to those moral rules pretty much the same way everyone else does, except that they stick stronger to the rules than most people. In short, their moral consciences are more strongly based on rules than others, whose consciences are based more on affective empathy and simulation.

      • dollhouse scientist Says:

        “Actually, I think that autistics do, in fact, act morally and DO, to a large extent, get conventional morality.”

        Yes, I meant “do”, not “don’t”. My bad. Apologies for the typo and the resulting confusion.

        I completely agree with you that empathy is overrated as a source or standard of value. Empathy and social emotions generally are adaptions to regulate individual behavior to promote group success in social organisms. Empathy does not reveal truth, respect individuals, reach far outside the tribe, or scale effectively to the needs of a global civilization. It’s enjoyable and a good foundation for sustainable friendships, but an unreliable guide to personal or public policy.

        May I please ask what your personal life experience with autism spectrum people has been like? I’m actively working to verify my own data. Texted an aspie friend at length this afternoon to ask precisely how his inner life and motivation work. I have two other friends I could reach in a couple of days. I could also float the issue with an autism spectrum support group which meets on Monday. For that matter there’s a local aspie women’s group I’ve been meaning to check out.

      • verbosestoic Says:

        It’s somewhat limited, which is why I rely on the accounts of others in the paper and the actual psychological research. But, then, I’m primarily a philosopher, so don’t really cotton to that whole “empirical evidence” thing [grin].

      • dollhouse scientist Says:

        Thank you for clarifying. It’s an inherently slippery subject, because everyone’s is in an awful epistemic standpoint with regard to autistic empathy. Because whatever autism spectrum people feel, other people aren’t very good at understanding it: autistic people are often inexpressive, and the “theory of mind” problems of understanding people of other neurotypes go both ways. Meanwhile, autism spectrum people are quite frequently alexithymic and may have low consciousness of their own emotions. And, unfortunately, even in the sciences human beings generally don’t have a great record on understanding people unlike themselves.

        It is lovely to meet someone pursuing the philosophic life in the world today. There is so little recognition. But I think it was Epicurus, wasn’t it, who said the philosopher was the most happy and fulfilled of all men? 🙂

  24. dollhouse scientist Says:

    Hello again. I finished my research on this issue and I have high confidence I understand what’s going on.

    Autistics qua autistics do have affective empathy. Go talk on autism forums and you’ll run into plenty of auties and aspies who feel normal or even excessive affective empathy; it just only gets triggered when cognitive empathy picks up on someone else’s emotions. Cognitive and affective empathy are indeed two different vectors and autistics as such only lack the former. This confusion is caused by neurotypicals not seeing emotional expression in autistics and autistics failing to give affective responses at “appropriate” times due to cognitive empathy failures.

    However, it’s still true that autistics as a group skew towards the low affective empathy end of the spectrum with a distinct cluster approaching zero affective empathy. Why? Because autism is 50% comorbid with high alexithymia, which is the subclinical (functional) condition of not being able to understand, identify, perceive, or in extreme cases even feel one’s emotions. It’s a spectrum, and the emotions are still there and have the usual physiological components, but an extreme alexithymic may have so little consciousness of their emotions that they aren’t there for practical purposes. And that includes affective empathy, and the combination of poor cognitive empathy and high alexithymia can kill affective empathy for all practical purposes.

    High alexithymics do look a lot more like psychopaths than autistics as such. (And 10% of the general population is alexithymic, and it’s not just autists or aspies). Alexithymics have little cognitive or affective empathy, feel empty inside, have few human attachments, and even use thrills and drugs to feel like psychos. The extreme cases are actually more purely rational than psychopaths who may feel anger, excitement, lust, and pride as intensely as neurotypicals. High alexithymics tend to have highly rationalistic ethics they follow passionlessly and conform to society in their personal lives because its easier and they don’t have much intense inner motivation to do otherwise. They’re sometimes unimaginative and hyposexual and often feel uncomfortable in emotional and sexual situations. Temple Grandin is an example of an alexithymic autistic. Your paper perfectly describes extremely alexithymic autistic psychology but only only 50% of autists are alexes and only some alexes are autists. It’s correct as such but you’re misdefining your target population in ways that will make highly affective autistics want to throw rocks at you..

    tl;dr: Autistics have affective empathy like everyone else but half of them have trouble perceiving their own emotions and these autistics have affective empathy issues not as autists but as alexithymics. Extreme alexithymics lack affective empathy like psychopaths but they aren’t the same thing.

    Obviously I haven’t given you evidence but this is the end product of months to years of obsessive research, personal contact with dozens of autistics, polling, and interviews. A week’s research on alexithymia should confirm most of the points I make above.

    • verbosestoic Says:

      Thanks for the reply. I would have replied earlier, but things have been busy the past few weeks.

      Anyway, while as I said above I wasn’t as careful as I might have liked about the terms “Cognitive Empathy” and “Affective Empathy”, it’s best to start from what I think is the simplest and clearest way to distinguish them: cognitive empathy is about knowing what people are feeling, and affective empathy is about FEELING what people are feeling.

      The main hypothesis/theory/wild guess that underlies my view here is that non-autistics use simulation to do mind-reading, and that gives them affective empathy first, and then they get the cognitive empathy by reading off the feelings from the affective empathy. Autistics, on the other hand, don’t simulate — or at least not very well — and so don’t get their cognitive empathy from affective empathy, but instead directly from theorizing about it.

      So when you say this:

      Go talk on autism forums and you’ll run into plenty of auties and aspies who feel normal or even excessive affective empathy; it just only gets triggered when cognitive empathy picks up on someone else’s emotions.

      That seems to be perfectly consistent with my viewpoint, as it suggests that they get cognitive empathy first, and affective empathy follows from that, whereas for non-autistics it’s the other way around. I didn’t explicitly mention that aspect in the paper, though, so it makes sense to look to see how that impacts my view on morality. So, starting from the top again:

      My theory is that we learn morality primarily through negative stimuli. Those who are non-autistic and are not psychopaths — the term “neurotypical” irritates me, for some reason — learn it primarily due to the negative affect caused by the simulation; they feel the other person’s negative emotions, which is a negative stimulus (for most people [grin]) and so learn to avoid doing that. That’s why they are more Humean than Kantian; they base their moral reasoning on the feelings of the people involved, and not as rules for their own sake.

      Psychopaths, on the other hand, get the negative affect from the simulation … but can’t learn from negative stimuli. So they are incapable of learning moral rules.

      Now, my theory is that autistics do not get the negative affect from simulation, and so can’t learn morality that way. However, they still learn morality. Given, then, that it still seems like they get their cognitive empathy by theory, and given that they tend to be Kantian in their outlook — these are the rules that must be followed — the theory is that they learn the rules through some other kind of negative stimuli, and I proposed anxiety since it is fairly typical of autistics to be anxious in social settings, and of breaking social rules.

      Now, as you yourself say, that would describe fairly well 50% of autistics … but that then would skew the results. So an interesting question arises: can those autistics who get a strong affective empathetic reaction use that negative reaction to learn empathy instead of something else? To test this, we’d want to assess the ones who have strong affective empathetic reactions to their cognitive empathy and the ones who don’t and run the former through the morality tests, to see if the former act more Humean than Kantian. If they do, then my theory holds: we learn morality through negative stimuli and negative affective empathy is a very strong stimuli for that. If they don’t … then my theory still holds, because it is then likely that they start learning by other methods first and so in general their morality works out to be Kantian. Yes, it sounds like my theory works no matter what — I love it when a non-plan comes together [grin] — but that’s only because the key is the INITIAL lack of affective empathy formed naturally by simulation … and it looks like even your research has confirmed that that’s the case.

      Anyway, what you said was interesting, and thanks for returning to share it. If I didn’t have a day job, I’d probably be doing more research and putting more thought into this. And please correct me if I’ve interpreted something wrong [grin].

    • dollhouse scientist Says:

      Thanks for your reply. No, I don’t believe you’ve misinterpreted me. I don’t have a strong opinion as to the truth of your primary thesis. Morality in the conventional sense you describe may well be largely tied to negative affect. I can’t say. I can offer something useful on a few very different points.

      1) What about sociopaths or secondary psychopaths? Using the terms synonymously, a sociopath is a person with partially or wholly reduced moral affect due to primarily environmental factors (there’s likely a genetic component, as psychopathy is generally thought to involve environmental epigenetic triggers in early infancy or in utero). A sociopath shares a psychopath’s poverty of moral affect and social emotions, but sociopaths as a population show heightened anxiety, as sociopaths as produced by adverse environments, usually in childhood through early adolescence. Psychologically speaking, a sociopath has failed to form or aborted an emotional identification with a species experienced as unsupportive and hostile. Morally speaking, sociopaths are almost as amoral as psychopaths — merely more impulsive and erratic, more flamboyant, less proficient at manipulation. Sociopaths are sometimes able to form social bonds exclusively with specific individuals or social groups, and may have emotional allegiance to idiosyncratic moralities without foundation in conventional moral principles. The more extreme cases, however, are functionally equivalent to psychopathy in moral respects, especially as the loss of effective empathy tends to be progressive and all but irreversible. With regards to your theory, the trouble is that a sociopath can be normal or high in anxiety yet completely amoral.

      2) I’ve had empirically significant experience with at least some primary and secondary psychopaths, and I believe I can assure you with a reasonable degree of certainly that you misunderstand psychopathic subjective experience. Psychopathic emotions are generally present, but shallow and fleeting, especially in the primary variety. But psychopaths do not experience a negative affect from conventional moral violations. To the degree someone is psychopathic remorse, guilt, and shame are simply not experienced, at least not as a result of moral violations and interpersonal harms. Anticipatory fear deficits may be a factor in psychopathic behaviour but this is a severable issue. Psychopaths simply don’t have an affective response to human suffering and no sense of pity, kindness, or compassion. A psychopath can steal, rape, murder, and torture and feel no negative moral affect whatsoever (and, in primary cases, furthermore no fear or anxiety). This can easily be verified by reading online psychopathic writers. I suggest Jay Jones and Athena Walker on quora, M.E. Thomas at sociopathworld.com and Jessica Kelly at psychogendered.com.

      3) This objection is more properly philosophical. You speak of morality as if it’s an objective reality to be discovered, in the manner of Newtownian laws of motion. Now, I’m not particularly motivated to dispute your ethics, but i do think one should recognise one’s subjective perspective or standpoint. To wit: when we describe something as, e.g., “sexy” or “delicious”, we’re generally aware that the qualities we ascribe to an object are experientially relative to the tastes of the observer. A straight woman may find a given man attractive, but a straight man may not agree, and a straight iguana almost certainly won’t. Similarly, your impression of the status of an object as “delicious” would differ were you a vulture, and would differ even further if were you a latrine fly. The quality “moral” is parallel. One may say metaphorically that a lesbian is “blind” to male eroticism, but cosmopolitan people manage to keep the subjectivity in mental context, and are aware that her evaluation is more a projection of her erotic tastes than a recognition of an objective quality. Similarly, from a psychopathic or sociopathic perspective, conventional morality is a set of emotional responses experienced the human majority which appears in stark relief as a subjective imposition onto a value-neutral reality. Reality itself has no more inherent moral texture than there exists a nonperspectival sexiness or deliciousness.

      • verbosestoic Says:

        Just a couple of quick comments on this one. It may take me a little bit to get to the latest comment.

        I think, from re-reading it, that 1) and 2) will be covered by replies to the latest.

        On 3) I have a series of posts starting this Friday on objective morality, so if you’re interested in objective morality, that would be the best place to discuss that.

  25. dollhouse scientist Says:

    I’d be very curious to hear your response.

    I’ve never known an autism spectrum person with an apparent affective empathy deficit which didn’t resolve into superficial problems with reading and displaying emotions, poor perception of emotion (alexithymia), repression, depression—or a comorbid condition or personality disorder. In other words, either the emotion’s there and doing all its work, but unacknowledged or stuffed down (a phenomenon continuous and hard to distinguish from the mentally unhealthy habits of many conservative and low confidence men)—or it’s not there for reasons not traceable to autism. I’ve seen autism spectrum psychopaths, sociopaths, narcissists, schizoids, and clinical depressive cases. It’s certainly likely in some of these cases that these conditions were coping strategies and defence mechanisms against negative affect resulting from living with autism, but I think the view that autistics as such are unaffective is disinformation.

    I just got done spending a few hours with a diagnosed autistic I used to know, heading off every escapist attempt to dodge the conversation into theory and refocusing the discussion on his actual feelings. They were there. Mainly a ball of anxiety, anger, shame, guilt, and depression. He was dying inside for lack of success and love.

    • verbosestoic Says:

      Sorry, I meant to reply to the other comment, but I’m INSANELY busy at the moment. But as a quick response to this one, remember that I’m not claiming that autistic, in general, don’t FEEL emotions. I’m claiming that they have simulation failures, not emotional failures.

  26. Dollhouse Scientist Says:

    I read something recently which made me think of your theory and our earlier debate.

    According to a study by Tove Lugnegård at the University of Gothenberg, 70% of young adults (mean age 27) with Asperger’s syndrome reported suffering from a depressive episode, with 50% reporting repeated episodes. I believe this is relevant here.

    My primary contention had been is that autism per se does not involve low affect and thus does not imply lowered affective empathy. That is to say, autistics qua autistics feel just as much as anyone else, and feel as much towards others; it’s merely that simulation deficits result in fewer situations which trigger affective empathy (and also that non-autistics tend to confuse autistic facial inexpressiveness for a lack of emotion). Now, a great deal of research seems to support this position, as do many vocal autism spectrum activists and my own personal experience with many autistic individuals. Nevertheless, I’ve run into too many cases of autism spectrum people who appear genuinely low in affective empathy to close the books there.

    I earlier suggested autism’s 50% comorbity with alexithymia as a culprit. Comorbidity with depression could do the same thing. Repeated depressive episodes highly suggests a some level of chronic depression in a person, so with some imprecision it’s fair to say that 50% of aspies may be depressed as well as alexithymic. And depression is characterised by flattened affect (as well as anhedonia and a lack of interest in social situations). Therefore I want to suggest the idea that low affective empathy in autistic people could often be an epiphenomenon of depression rather than autism as such. Especially given that while there is some (in my opinion not very strong) evidence linking autism to genetic depression, it’s overwhemingly likely that this depression stems primarily from the environmental suffering autistic people often to experience due to social rejection, both directly and in terms of a lifetime of socioeconomic and personal consequences.

    This thought came to me while reading the work of Edwin Rutsch. Rutsch’s personal project is his Center of Building a Culture of Empathy (http://cultureofempathy.com/Projects/Bios/EdwinRutsch/). What struck me is that he’s obviously autistic, and he’s not merely a nakedly passionate person devoting a large portion of his life to empathy as a passionate cause. It’s much more than that: he’s plaintively *begging* for affective connection. His entire work is painfully animated by a need to reach out and a tragic frustration that all the writing, thinking, and activism in the world *about* closeness isn’t getting him closeness himself. I’ve been watching videos and sadness is painfully melting off his eyes; his voice and posture is pleading for a hug. He’s depressed, horribly depressed. And this despite a quite impressive and probably comfortable life as a successful IT professional, a published writer, a world-traveler, and a prolific online activist. He’s a perfect case study for what high affect and simulation deficits look like together as well as depression as a common consequence of an autistic’s unfulfilled affective needs.

    Here’s a video which illustrates a contrast. It’s an interview between Edwin Rutsch and Sam Vaknin discussing the subject of empathy. Rutsch is a high-functioning autistic with high affective empathy; Vaknin is a psychopathic narcissist with zero affective empathy. A low affective empathy autistic with narcissistic or psychopathic tendencies might share some characteristics of both parties.


    (Incidentally, Vaknin, who is philosophically educated, also directly contradicts your theory that psychopaths get a negative affect from simulation. Non-autistic psychopaths develop simulation abilities to read emotions like most people and typically develop cognitive empathy consciously building on this foundation; it’s just that this simulation will trigger little or nothing in the way of social emotions such as guilt or shame which grant intuitive validity to conventional morality [and primary psychopaths won’t get anxiety either]. Vaknin terms the result “cold empathy”, which I think is a helpful descriptor.)

    • verbosestoic Says:

      On another note, I was sure that I had replied to the other comment, and it looks like I didn’t. Sorry about that. I’ll try to reply to both in the near future. It reflects a problem where I think things out long in advance, even as if they were replies, but then I can’t remember if I actually wrote it down or not.

    • Dollhouse Scientist Says:

      Well, thank you. No obligations implied, of course. Your blog, your convenience. I know I personally overthink everything and compose spontaneous paragraphs in real time but find it almost impossible to force myself to write under social demand.

      I really do recommend the Rutsch/Vaknin videos if this subject interests you. Rutsch is more empathic than your average human (autistic or otherwise), and Vaknin’s narcissism gives him a social motive (recognition) which blurs his otherwise extreme psychopathy, but otherwise the contrast is near-perfect and palpable. They’ve done lots of interviews together which I’m presently watching through with fascination.

      • verbosestoic Says:

        A couple of points on empathy, since I strongly disagree with empathy being used as a basis of morality. I think there are two points where Rutsch and Vaknin talk about it that are important:

        1) Vaknin claims that even if we can figure out that someone is feeling a certain thing, we can’t know what that sadness is “like” for that person, and if it is the same as ours. This isn’t particularly important, unless you are some variety of Utilitarian, because you couldn’t judge the DEGREE of emotion there. You might think it is merely mild sadness, while for them it is devastating. On the flip side, you could consider an action devastating while they would find it mild. Thus, calculating the total “harm” using empathy would risk getting it wrong a lot of the time.

        2) More importantly, Vaknin is right to be suspicious that we can know what other people are feeling. Rutsch tries to use simulation there, but simulation will fail for someone that is sufficiently different from you that you don’t understand very well, and we should note that empathy fails in the precise same way. So if used as morality it could be an excuse to act in certain ways on the basis of “empathy” and ignore that that isn’t what they’re feeling

        As someone who is Stoic-leaning, I think that ANY emotionally-based assessment has to be verified using reason, for multiple reasons. Given that, trying to develop empathy here seems pointless or even harmful, since either your reason would already tell you that it is the right answer, or else it will tell you the wrong answer that you’ll need to overcome when reason tells you that it’s the wrong answer. Our simulation and emotions fails far too often to be considered reliable, and the impressions they give us are often very hard to overcome.

    • verbosestoic Says:

      I promised I’d get back to this, and I’m finally able to get around to it.

      My primary contention had been is that autism per se does not involve low affect and thus does not imply lowered affective empathy. That is to say, autistics qua autistics feel just as much as anyone else, and feel as much towards others; it’s merely that simulation deficits result in fewer situations which trigger affective empathy (and also that non-autistics tend to confuse autistic facial inexpressiveness for a lack of emotion).

      Hey, that’s MY contention [grin].

      Let’s start by outlining the terms again:

      Affective empathy: Feeling what other people are feeling.
      Cognitive empathy: Knowing what other people are feeling.

      And another term that I don’t have a good name for yet, but that we might call for starters “moral empathy”, which is CARING about what other people are feeling as something more than just a means to get what you want from them.

      My contention is that humans typically get affective empathy from simulation by — consciously or unconsciously — putting ourselves in their position, rolling things forward, and thus get both the predicted actions AND the predicted emotions from that. Autistics however, have simulation deficiencies, and so that process doesn’t work. For them, I propose, they figure out what a person is feeling cognitively first — ie they come to know it — and THEN that might trigger an affective emotion in them that gets them affective empathy. Or, it might not. I don’t claim that autistics have diminished affect, and also claim that autistics, unlike psychopaths, have moral empathy: they clearly CARE about what other people are feeling. That being said, they also COULD have affective issues as well, and some of that could be caused by the inversion of causation; some of them don’t feel as much because their mechanisms don’t kick in as often, which can have an impact on neural development.

      Okay, so what about psychopaths? On re-reading the essay, I noted that nothing in the essay says that psychopaths in fact HAVE affective empathy. In fact, it presumes that they DO have empathy issues similar to that of autistics and insists that even if that is true that does not explain their lack of morality (to which you can add moral empathy). So why, then, in the comments am I arguing that they have affective empathy? Well, the reasons I posited for autistics not having affective empathy is that they have simulation failures — for example, clear mirror neuron deficiencies — and that their cognitive empathy abilities correlate with IQ. As far as I know, neither of these are typically the case for psychopaths. We know that psychopaths do not LEARN from negative stimuli, but that’s not the same thing as saying that they don’t, in fact, HAVE negative stimuli. Thus, it is possible that they DO experience the negative emotions and simulate properly, but don’t ever learn from them.

      However, there are several ways to explain that data for psychopaths:

      1) They DO show that correlation with IQ, but no one has ever bothered to LOOK for it.

      2) The tests on autistics had a wide range of IQs represented, as even high functioning autistics are often on the low end of the IQ scale. Psychopathy might correlate with higher IQs, and once you get over a certain IQ the difference is mostly indistinguishable.

      3) In general, autistics might in fact attempt to do things “the way everyone else does”, impeding their ability to predict emotions, while psychopaths — especially ones who are narcissistic — might instead develop and focus on what they can do well ignoring what others do, meaning that they just get better at it through simply directly using theory and reasoning (compatible with Vaknin’s self-assessment of how he actually DOES predict other people).

      So, I’m not wedded to the idea that psychopaths really have affective empathy. In fact, the amygdala damage suggests that they might indeed have problems with at least negative emotions, or feeling them. That being said, we need to know why it is that they are so good at predicting the emotions of other people so as to manipulate them so well, typically, although that might be driven more by inducing positive emotions in them than in recognizing negative ones.

      I’m not convinced by Vaknin’s comment that narcissism and psychopathy ought to be considered the same thing. What we’d need to do is test the moral/conventional distinction, because narcissists ought to be CAPABLE of knowing the difference, but typically won’t care, while as I argued psychopaths seem incapable of telling the difference.

      • Dollhouse Scientist Says:

        Thank you most kindly for taking the time to reply, especially in depth and seriousness; it is greatly appreciated. My apologies if it’s taken me some time to answer in turn; I noticed your response about a week ago and finally have a break from my work life.

        First of three part reply. On Vaknin and narcissism:

        “I’m not convinced by Vaknin’s comment that narcissism and psychopathy ought to be considered the same thing.”

        I completely agree. Antisocial personalities are about a lack of affective emotion. Narcissistic personalities are about dependency on a sense of superiority. I’m undecided as to whether the two are etiologically entirely independent, or whether they are both positions on a superspectrum of what psychology calls the Cluster B or “dramatic” personality disorders (and what moral philosophy or ordinary language might simply term varieties of selfish character). Regardless, they’re not the same thing, though they can certainly coexist in the same person and their expressions share convergent characteristics.

        I brought up Sam Vaknin because those videos provided a good side-by-side contrast of two different types of lack of empathy. Beyond that I don’t trust him. His self-promotion as the man who wrote the literal book on narcissism does seem justified by a depth of understanding which could only derive from sincere self-reflection and self-insight. He’s also a convicted con artist who feels relatively emotionally silent so believe he’s genuinely psychopathic. But his equation of the two isn’t justified (even if others have made the same conflation) and I’m highly skeptical of his motives. I completely agree that non-psychopathic narcissists possess a normal range theoretical capacity for affective empathy; they’re just so busy with their own superiority and so little concerned with others’ needs they will give little and what they do give will be self-centred. I’m willing to let Vaknin speak for psychopaths because he does seem to show a base lack of effective feeling as well as narcissistic compensatory grandiosity and his walkthrough of how psychopaths experience others squares with what both researchers and antisocial writers have written.

        I do agree that empathy is generally a poor tool to measure wise conduct, including for altruistic ends. Empathy comes loaded with cognitive biases, from salience to in-group favoritism. It favors fulfilling the feelings of the empathizer over effective action and crowds out awareness of opportunity costs and unintended consequences. Oxytocin encourages trust and care only to simultaneously heighten ethnocentrism and distrust of outsiders. And in the form of pity it inherently enlists feelings of superiority.

        Vaknin’s point that we can’t “really” know if we’re feeling the same thing as another is not terribly relevant. Personally I think he’s mainly using bringing up a theoretical point as cover to slip in and attack Rutsch’s most likely vulnerabilities as an autism spectrum person with an intense need for human connection. Partially because Vaknin wants to establish dominance (psychopath-style instrumental social aggression) and partially just to be a dick (narcissistic maintenance of inflated confidence by lowering others).

      • Dollhouse Scientist Says:

        On affective, cognitive, and moral empathy.

        (now that I’ve written this out, my apologies for the length. I’ll hold off on the last point I was going to make on moral knowledge and foundations)

        Okay, I do now think we’re on roughly the same page as regards the mechanics of simulation as a trigger for affective empathy to express, and the relative absence of simulation in autistic individuals (sometimes) causing relative de facto absences of empathic response. My initial objection was that many researchers and many voices from the autistic community have strongly expressed that autistic people experience normal or even elevated empathic emotion. Many autistic people strongly reject and feel silenced by a popular perception that autism implies affective deficits and (quite reasonably, to my mind) ascribe the issue to non-autistics reading a lack of emotion into a lack of reading or displaying emotional cues. i don’t think autism per se involves affective deficits. At the same time, as mentioned I’ve come across too many autism spectrum people with what look like genuine affective empathy issues not to believe something important is going on here.

        I’m beginning to think that just as some psychology writers and tendencies in popular opinion are dismissing the experiences of high affective empathy autistics, others may be invested in minimizing the experiences of autistics who experience less affective empathy. It does seem reasonable to describe someone whose affective empathic capacity is rarely triggered due to simulation deficits as low affective empathy. The same might be said of individuals with high alexithymia.

        You’re also right that the cumulative effect of these mechanisms not kicking in as often could influence neural development. Here however the question gets murkier. Part of me wants to say that a person whose baseline affective empathy underdevelops is to that degree by definition a sociopath or pscyhopath — whether or not the lack of reinforcement of developing neural pathways is mediated by autism or some other factor. Especially given that hostile childhood environments, lack of bonding with caregivers, complete social rejection, etc. are the standard conditions for developing secondary psychopathy and even known epigenetic triggers for MAOA genes typically associated with primary psychopathy. In other words, if an autistic child doesn’t simulate enough to experience love and care, is that the same thing as a non-autistic child who has the same subjective experience for different reasons?

        I think this could be quite logical, but it’s unlikely to be true. The reason is precisely the main concern of your essay: autistics as a group show no deficits in pro-social motivations and in fact seem to be more conventionally moral than the general population. So if low affective empathy autistics are a significant part of autistic population (which I think they are), then they can’t have the same kind of lack of empathy as antisocials while producing this outcome. Even alexithymia can’t explain it, because while extreme alexithymics (autistic or otherwise) do start bleeding over into amorality and other characteristics suggesting possible continuity with psychopathy, autistics are, again, and despite high rates of alexithymia, as or more conventionally moral as the general population.

        So my present theory is to make a distinction between *moral affect* and *affective empathy*. What I’m implying is that conventionally moral behavior is built off empathic or social emotion (love, kindness, care, guilt, shame, remorse, etc.) This is what psychopaths and sociopaths don’t have or barely have, but autistics and the general population do. This is distinct from affective empathy, which is the “distribution” (like in mathematics) of this emotion in immediate response to real-time social signals. You can have high or normal empathic *emotion* while still having low or rare affective empathy as a result of a high relative absence of simulation. The result would be (I think) generalized emotional concern for the welfare of others but low emotional response in others’ emotional states.

        I think of a song I ran across titled “We’ll Get By (the Autism Song)” by Johnny Orr Band. There’s this line which I’d never been able to make sense of before: “Share your joy but only in my mind.” I think this might describe moral affect without affective empathy. (My apologies for the overall tone of the song, which I think is overly depressing about autistic people and says more about the musician’s internalized cultural attitudes towards autism than autism itself.)

        Maybe this is what you describe as “moral empathy?” You define this as “CARING about what other people are feeling as something more than just a means to get what you want from them”, which may be the same thing for which psychological literature uses the term “compassion”. I’m picturing the same basic emotional “stuff” as affective empathy, just not applied in the specific mode of empathic response. When you care about the suffering of people even if you have no idea or emotional impression of what it’s like for them, which the general population does as one mode of caring, and may be the dominant kind of caring in autistic people with low simulation. (This might explain why autistic people are overrepresented among moral and political idealists where abstract concern for humanity takes precedence over personal emotional bonds)

        I’m not sure if my theory is right here. Primarily because I’m working from abstract concepts and I’m not at all sure I know what “moral empathy” is. For that reason, could I please trouble you to explain what moral empathy is like? I don’t mean so much as an idea or theory, which I think I get well enough. I mean in terms of phenomenology or qualia: what does care or compassion in this sense feel like subjectively to an individual? I imagine something warm, something attractive, immersive like a liquid. I don’t feel I know what the idea (and I think others have suggested the distinction you’re making) is trying to capture.

        Let me present a foil against the picture of a certain shade of autistic I’m trying to describe above. You seem to be suggesting someone who is morally empathic but low in affective empathizing due to simulation deficits. Okay, I totally believe this describes a significant number of autistic persons. I want to contrast this with what I think I would call an autistic sociopath. What touched on this to me in your article was the suggestion that autistic people learn (conventional) morality by anxiety produced by unintentionally overstepping social/moral boundaries, which teach pro-social behavior by trial and error. When I read this, my mind immediately went to someone who doesn’t have a sense of right or wrong, but does feel anxiety/fear, and who thus comes to observe social limits from anxiety in the absence of guilt, remorse, or similar negative moral emotions. That looks like very neurotic sociopathy to me, but I don’t think that’s what you’re describing. I think (I may be wrong), you’re discussing someone with all the normal emotional feelings, but who accidentally walks over interpersonal boundaries from lack of simulation or empathy and goes through a learning process from anxiety as a result. I’m picturing someone with sincere moral motives, whose accidental trespasses don’t trigger guilt due to a combination of purity of intent and lack of simulation feedback of others’ emotional pain. (And, yes, I think the result is absolutely more likely to encourage a Kantian rather than Humean psychology).

        By contrast, here’s what I want to call an autistic sociopath (drawn, like most of my theorizing here, from a composite of research and personal experience). First of all, I’ll simply stipulate this person is autistic (with the usual array of pattern-focused thinking, cognitive empathy and executive dysfunction issues, social difficulties, sensory issues, etc). Now, in addition, she shows a complete or near-complete lack of emotional concern for others, relating to others purely for advantage or stimulation, extreme selfishness, consistent lack of respect for “moral” and social limits and no intuition as to why hurting others is a problem (unless we count game- theoretical or self-interested reasons for behaving pro-socially). She checks most of the standard psychopathic boxes: extreme pride, lack of remorse or guilt, shallow affect, callousness, proneness to boredom, impulsivity, promiscuity, parasitic lifestyle, varied antisocial behavior, etc. Masking, lying, and manipulation are meaningfully present but inconsistently competent and developmentally delayed. Anxiety, however, is normal to high (along with stress responsiveness and threat reactance), and there’s an intuitive attraction to theory, reflection, abstraction, and coherence atypical for antisocials.

        Let’s also stipulate there isn’t any other pervasive condition (such as narcissism or depression or drug addiction) curtailing moral or affective empathy.

        Is this a categorically different case than your low simulation autistic person who learns moral rules by anxiety? On a related note, is there any way the empathic deficits of any shade of autism as you understand it can explain or generate this type of low empathy? Is this way of experiencing the world in reflective or continuous with your understanding of how autistic people experience the world/other people, or decisively different in kind? If there is an essential difference between a low affective empathy autistic and a sociopathic autistic, what is it?

      • verbosestoic Says:

        I’m not sure how you’re monitoring this thread, and my reply is delayed again, but let me address your example.

        The first thing that needs to be made clear is how simulation works. Simulation is you putting yourself in the place of the other person and running that forward to see what you would do or how you would react in those cases. This may be subconscious, which then could allow for this happening without you knowing or without any experienced result at all; you come out of it just “knowing” what the other person would do. My theory — not expressed here in detail — is that doing that you’d also get what the person would be FEELING, because you’d get what YOU would be feeling in their place automatically. This is NOT true for a theory approach, because you don’t automatically get the emotional reaction from the reasoning process (although you potentially could). And, again, that feeling might be subconscious.

        But what this means is that you aren’t necessarily going to get guilt directly from the simulation unless the person themselves would feel guilty, or rather that you yourself would feel guilty if you were in their shoes. So with simulation the link is not “Simulate and get guilt” but “Simulate, get emotions/reaction, and then something ELSE triggers guilt in you”. And so we have to ask what triggers an emotion of guilt. And, in general, it seems to me that at a very basic level what triggers guilt is a feeling that you are doing or are contemplating doing something wrong, and seriously wrong. You’re breaking a serious rule, and often a moral one.

        Thus, my theory is that in general we teach moral rules — the ones that are the most likely to trigger guilt — using aversive conditioning, using negative emotions. For those who are not autistic and not psychopathic, that comes from feeling the negative emotions of the person that you would be hurting, which explains why, in general, they tend to be Humean, as they learn that the common factor is that the other person feels bad, and so their moral “rules” are based primarily on that criteria. Autistics don’t get that, so they learn the rules and learn which rules are particularly bad to break due to the anxiety they feel when they ponder how people will react to them breaking them, which makes them very attached to the rules, and so Kantian. Psychopaths can’t learn from aversive conditioning, and so never learn the rules at all, as evidenced by their issues with the moral/conventional distinction.

        But just having the rules in your head doesn’t mean that you will feel guilty if you break them. You have to GENUINELY adopt them as rules that denote what is right and what is wrong. When you have that, then you will feel guilty when you break or even when you contemplate breaking the rules. So once you learn the rules and come to genuinely accept them as denoting what is right or what is wrong, then you will feel guilty if you break them and thus will genuinely act in some sense morally.

        There’s also a place in here for how you think of other people; if you care about them, then you will feel guilty if you hurt them, arguably even if the rules say you have to.

        Thus, the difference between those two cases is that in the first case the autistic person learned the rules through their anxiety reaction, but genuinely accepts that they denote what is right or wrong and so will feel guilty when they break or contemplate breaking them. In the second case, either the person never learned the rules or has them as abstract conditions in their head, but hasn’t adopted them as denoting what is right or wrong. And so, having no notion of right and wrong, they don’t feel guilty when they do wrong things or contemplate doing wrong things.

        So it’s not that, under my theory, being anxious makes autistics act morally. Anxiety gives them the rules, and they genuinely adopt them as their idea of what is right and what is wrong, and so breaking the rules triggers guilt in them. Thus, there’s a separation between how one learns the rules and if one accepts the rules, and thus between the anxiety and the guilt reaction.

  27. Philipse on the Predictive Power of Theism | The Verbose Stoic Says:

    […] make a sharp distinction between social conventions and moral claims, and this rather famously is something that psychopaths fail at. So taking that route to eliminate objective morality doesn’t seem all that plausible […]

  28. Walk (Up) of Life … | The Verbose Stoic Says:

    […] school shootings explicitly mentioned psychopathy as the explanatory cause for one of them, and the last time I checked psychopathy was, in fact, a mental […]

  29. Andrew Says:

    I find it interesting that Prinz (as described by your account) believes some form of emotional connection is necessary for morality. There are plenty of classical moral systems that are based on external “rules” for right and wrong. Even the so-called “Golden Rule” – “Do as you would be done by” – is applied morality (how to be moral) rather than morality itself (why be moral).

    Fundamentally, a moral position of “do these things so that other people will like you” is no more moral than a psychopath – you’re just more effective at fitting in. “Right / good” is subjective, not objective.

    One could even argue that normals are *less* moral than autistics. Autistics want to know what is “right” and keep the rules, whereas normals are better at knowing when they can get away with breaking the rules.

    I see you address this in the comments:

    which is ANOTHER reason why I oppose empathy-based moralities, because the emotions of someone can be wrong, and so using that as your basis either leads you to immoral outcomes or else is something you have to constantly check with something else — typically reason — anyway

    I find Dollhouse’s comments very interesting, especially talking about different symptoms of autism. Let me give another 1st person perspective…

    I’m diagnosed with “Asperger’s syndrome”, which these days is considered an “Autism Spectrum Disorder”. Take that with however many grains of salt you consider wise.

    Emotionally, I personally feel high adrenaline emotions: fear, excitement, anger for example. From observing other people’s language, I perceive that there’s at least one other category of emotions, that I refer to as “soft” emotions, that I don’t experience. For example, I have no personal experience of “warm fuzzies”, but I’ve observed enough to be sure about that. Similarly, I experience “comfort” as a pleasant external physical sensation(e.g. a hug) and a lack of pain or threat or anxiety, but there’s no other psycho-somatic response.

    I completely lack effective empathy. I can cognitively put myself into someone else’s shoes, and perhaps even stimulate “hard” emotions in myself by modelling my own emotional reaction to the situation. But any sort of reading or reacting to another person’s emotions is entirely driven by cognitive processing.

    I think you’re right in putting “caring” in a different category. Broadly speaking, I want good for other people. This is a combination of an objective morality and that I like making other people happy. Note that “like making other people happy” is “pleasure in job well done” not an emotional sympathetic response.

    Similarly, my desire for “connection” is expressed in “we understand each other’s thought” and “we like making each other happy”.

    The other ASD aspect of my personality is anxiety with lack of control and of being easily overwhelmed by organisational tasks. I also get overwhelmed by excessive stimuli.

    Socially, this puts me at an interesting place from an Introverted / Extroverted perspective. I love getting to know strangers. But I hate crowds. When I am paying attention, I can read the thoughts and emotions of a single person very well. But groups are very difficult, because there are too many people to read.

    My children all display similar symptoms, in one form or another. The eldest thrives on over-stimulation (and then crashes), while the younger two generally avoid it. The elder and younger easily get into conflict situations, while the middle child is conflict-averse. The elder child cannot handle breaking rules; the middle child is sometimes like this, and sometimes redefines reality in her head to fit whatever she thinks the asker wants to hear.

    This last is interesting. She is certainly capable of lying and deception, but sometimes just has a flexible view of reality and terrible memory for organisational details. For example:

    “Did you clean your room?”

    thinks: clean my room? Cleaning my room would be a good thing, so I must have done it. Yes, I remember doing it now. “Yes”

    Without being conscious of doing so, she will create memories in order to fit her desires for the situation.

    • verbosestoic Says:

      I find it interesting that Prinz (as described by your account) believes some form of emotional connection is necessary for morality. There are plenty of classical moral systems that are based on external “rules” for right and wrong. Even the so-called “Golden Rule” – “Do as you would be done by” – is applied morality (how to be moral) rather than morality itself (why be moral).

      I might have been a little misleading there, because Prinz thinks that morality JUST IS an emotional reaction of the right type, not that it requires an emotional connection per se. Thus, for rules-based moralities he would say that, yes, they have come up with systems but they aren’t accurately describing what morality is. He argues against this using “moral dumbfounding”, arguing that you can refute all of the arguments someone is using to call something immoral and they will likely STILL consider it immoral. His example is of incest, where if you can show that the two are of age, there is no power differential, and they are using birth control so no child will result people STILL tend to call it wrong, despite all the common reasons for calling it immoral being taken into account.

      Obviously, I disagree with him on this, but that’s the take he has on morality.

      Your account of your own experience is interesting, as it somewhat confirms my view on the difference between simulation theory and theory-theory when it comes to emotions and the like.

      Socially, this puts me at an interesting place from an Introverted / Extroverted perspective. I love getting to know strangers. But I hate crowds. When I am paying attention, I can read the thoughts and emotions of a single person very well. But groups are very difficult, because there are too many people to read.

      This is compatible with introversion, though, as they tend to do well in small groups but not with large ones. I wonder how much introversion comes from these sorts of social concerns: maybe the reason introverts are tired by social situations is that they have to dedicate more mental resources to them than extroverts, which overcomes the typical energy boost that people can get from “having fun” in a social situation.

      This last is interesting. She is certainly capable of lying and deception, but sometimes just has a flexible view of reality and terrible memory for organisational details. For example:
      “Did you clean your room?”
      thinks: clean my room? Cleaning my room would be a good thing, so I must have done it. Yes, I remember doing it now. “Yes”
      Without being conscious of doing so, she will create memories in order to fit her desires for the situation.

      I wonder if this is less “creating memories” and more just not being able to remember and so coming to a conclusion about it. Both myself and my father tend to have a specific sort of behaviour in certain situations where we lock the door or close the window or whatever, but because it’s such a routine task we don’t necessarily remember if we did it. So then something reminds us that we needed to do it — for me, recently, we had very dark clouds that looked like VERY imminent rain — and can’t remember if we did. So, instead of worrying about it, we go check, because otherwise we’ll keep wondering if we did. Your daughter might be doing the same sort of thing, except that while we can’t remember and so reason that maybe we didn’t and should check, she instead reasons out that it sounds like something that she would have done, and so then concludes that she must have, even if she can’t remember it explicitly. One of my tricks for remembering these things is in fact to do something memorable so I can say “Well, I did this today, so I must have actually done it”, even if the explicit overall memory of doing it isn’t there. So she might have less a created memory and more a conclusion.

      Anyway, interesting examples and comments here.

  30. The Joker’s Wild | The Verbose Stoic Says:

    […] about psychopaths? One of the more interesting results about psychopaths is that they seem to be amoral, as they consistently fail to properly make the moral/conventional distinction. What this means is that they don’t seem to be able to distinguish moral actions from […]

  31. Bright Colors, Dark Times | The Verbose Stoic Says:

    […] because they are incapable of understanding what it means for an action to be moral or immoral (my essay on psychopaths argues that they are amoral in that precise way, due to them not grasping th…). The other way is for them to be able to understand what is moral or immoral, but be incapable of […]

  32. Adam Lee’s Universal Utilitarianism (Part 2) | The Verbose Stoic Says:

    […] wouldn’t have that. This runs into problems with humans who have different ideas of morality, such as sociopaths, psychopaths and, on the positive side, autistics. The first two are more amoral, while autistics tend to follow a more rules-based morality. But […]

  33. Bob Seidensticker on Objective Morality | The Verbose Stoic Says:

    […] hold across the species.  I’ve already mentioned psychopaths and sociopaths, but those on the autistic spectrum are noted for not making the same moral judgements as others.  Seidensticker can make a reasonable argument that psychopaths and sociopaths are immoral or at […]

  34. Should Vampires Be Held Accountable For Their Bloodthirsty Behavior? | The Verbose Stoic Says:

    […] understanding what is and isn’t moral and choosing to do otherwise (which distinguishes them, in my view, from psychopaths who have great difficulty learning that distinction).  So if a vampire is just […]

  35. “The Bloody Connection Between Vampires and Vegetarians” | The Verbose Stoic Says:

    […] is this sort of goal-directing reasoning is not sufficient to make someone moral.  As I noted in my essay on psychopaths, while they may not make the right choices they certainly seem capable of some sort of goal-based […]

  36. “Who Trusts the Watchmen?” | The Verbose Stoic Says:

    […] or doing anything moral.  He, then, is an awful lot like how I considered psychopaths to be in my essay on them:  he doesn’t care enough about morality to be anything other than […]

Leave a Reply

Fill in your details below or click an icon to log in:

WordPress.com Logo

You are commenting using your WordPress.com account. Log Out /  Change )

Twitter picture

You are commenting using your Twitter account. Log Out /  Change )

Facebook photo

You are commenting using your Facebook account. Log Out /  Change )

Connecting to %s

%d bloggers like this: