This post is by Marko Jurjako (pictured above) and Luca Malatesti (pictured below), collaborators on the project Classification and explanations of antisocial personality disorder and moral and legal responsibility in the context of the Croatian mental health and care law (CEASCRO), funded by the Croatian Science Foundation and based in the Department of Philosophy of the Faculty of Humanities and Social Sciences in Rijeka (Croatia). Marko is a junior researcher on the project and his interests lie in moral psychology and evolutionary approaches to metaethics. Luca is assistant professor of philosophy and works mainly in philosophy of mind and philosophy of psychiatry. In this post they summarise their paper ‘Instrumental Rationality in Psychopathy: Implications from Learning Tasks’, forthcoming in Philosophical Psychology.
Psychopathy is a personality disorder that involves traits such as pathological lying, manipulativeness, superficial charm, no or little concern for the interests of others, a grandiose sense of self, and, usually, a long history of offences and encounters with justice. In the last twenty years this condition has attracted a mounting interest by philosophers (see the online bibliography PhilPapers: psychopathy).
Some debate the significance of psychopathy for adjudicating between different accounts of our moral psychology. While sentimentalists maintain that empathic and emotional deficits explain the limited moral understanding and motivation of psychopaths (Nichols 2004; Prinz 2006), rationalists explain these limits in terms of rational deficits. Specifically, amongst the rationalists, Heidi Maibom in her seminal paper ‘Moral Unreason: The Case of Psychopathy’ (2005) argues that empirical studies on the instrumental learning of psychopaths show that they are irrational insofar they fail to use the means at their disposal to accomplish their goals.
There is some empirical evidence that show that psychopaths learn slower than controls in instrumental learning tasks. These tasks involve establishing associations between a set of stimuli, the response of subject and a rewarding or punishing stimulus in a certain context. One of the paradigmatic examples of this kind of task, where psychopaths show learning impairments, is the so-called response reversal task. In tasks of this type, in a first stage, participants have to learn to choose between two stimuli presented to them simultaneously on a screen. If they choose correctly they get rewarded (they earn points), if they do not they get punished (they lose points). The correct stimuli are often specified by one of their dimensions, for instance their shape. In the second stage of the task, participants need to learn to change (reverse) their previously learned responses. For example, if in the first stage choosing the round stimuli was rewarded and choosing the square stimuli was punished, in the second stage the reverse is the case (choosing the square is rewarded and choosing the circle is punished).
In our paper we argue that the empirical evidence adduced by Maibom and newer results on the instrumental learning of psychopaths leave the issue of their instrumental rationality undecided.
Maibom thinks that psychopaths violate the following familiar Kantian principle of instrumental rationality:
(i) who wills the end, also wills the means that are indispensably necessary to his actions and that lie in his power.
We maintain that evaluating the significance of the studies on the instrumental learning of psychopaths for the issue of their rationality requires spelling out the requirement in (i) that ‘a mean lies in the power of an agent’. Following the currently dominant interpretation, we assume that this condition requires that a mean is epistemically accessible to an agent (Kolodny and Brunero 2013). Bernard Williams’s famous gin and tonic example illustrate this reading. A person who drinks a glass containing petroleum in a bar thinking that the glass contains gin and tonic is not instrumentally irrational because the information about the content of the glass is not relevantly accessible to her.
We argue that when (i) contains the requirement of epistemic accessibility, the empirical data on instrumental learning do not show that psychopaths are irrational. We offer several arguments for this conclusion, given that there are different plausible ways in which (i) can be understood. However, a similitude might illustrate the gist of our reasoning. Colour-blind persons are not irrational because they would do worse than controls on the instrumental learning tasks where the relevant stimuli include colours that they cannot perceive. Their performance would not show that they do not will the necessary means for their ends. It would indicate, instead, that the means are not accessible to them. Similarly, we argue that current empirical evidence concerning instrumental learning in psychopaths, and some of the leading explanatory hypotheses of these results, authorise the conclusion that psychopaths might have a deficiency that undermine the accessibility of certain information to them.
However, there are suggestive experiments with ventromedial prefrontal cortex patients, that involve self-reports about what is their experience or awareness of the relevant information while they perform certain instrumental learning task (Rolls et al. 1994). We suggest that similar experiments if conducted with psychopaths might help to address the issue of their instrumental rationality.