The relationship between psychotic illness and criminal culpability is complex. This much is largely agreed upon: a defendant (D)’s experience of psychosis at time t1 has the potential to be relevant in some way to their culpability for acts performed at t1. The scope and nature of that relevance is where things get messy. One popular notion is that there is (or can sometimes be) something especially exculpatory about psychotic illness that goes beyond its evidentiary relevance to standard principles of culpability (mens rea, voluntariness, mistaken justification, etc.). A common version of this idea is that D’s psychosis makes them someone to whom those standard principles do not apply – in other words, as someone who is not merely exculpated on the facts, but actually non-responsible in a way that renders them an inappropriate object of sanction by the criminal law. On this kind of understanding, psychosis of a specified nature and/or degree operates like childhood does, identifying the individual as someone whose actions cannot give rise to criminal liability.
I’m concerned with a specific and popular way of advancing this kind of argument: one that understands the special exculpatory function of psychotic illness in terms of rational incapacity. Many accounts of this nature exist, but the broadly unifying idea is that an attribution of criminal responsibility to D in respect of an act a1 at a time t1 presumes that D had certain rational capacities at time t1, and these may have been absent if D was experiencing psychosis. In such cases, therefore, the best analysis for D’s non-liability is that, with respect to act a1 at time t1, they failed to meet a necessary capacity threshold for criminal responsibility.
My aim in this post is to introduce a note of caution. To be clear: I do not intend to argue that it is never appropriate to analyse the non-culpability of a psychotic defendant in terms of rational incapacity. Cases may exist in which D’s rational failures are of such kind and/or degree that it becomes impossible to hold him to a standard of criminal responsibility. What I do intend, however, is to advise against an overreliance on this analysis as a theoretical basis for legal approaches to mental condition-based exculpation. To make this case, I will present a class of cases involving intuitively non-culpable delusional defendants in which a rational capacity analysis simply doesn’t work very well, leaving us unable to a) tell the full story as to why the relevant delusion gives rise to non-culpability, and b) adequately exclude similar cases which appear not to give rise to non-culpability in the same way. My objective in presenting these cases is to advance the idea that while rational capacity models may form part of a proper understanding of mental condition-based exculpation, they need to be supplemented. We must commit to the task of working out additional avenues of mental condition-based exculpation that do not rely on findings of incapacity, but instead allow for an understanding of D’s actions in terms of the interplay between their rational failures and their rational, moral or valuational successes. What these avenues might be is a question for another time.
“Doing the Right Thing”: R-Type Cases
I’m going to look at a loose class of cases which I will designate R-Type Cases: cases in which it can be said of a delusional defendant that they believed themselves to be “doing the right thing” – the thing morally necessary in the circumstances. Consider the following hypothetical:
R: Ruth is at home with her young child when she experiences intense auditory and visual hallucinations that her block of flats is on fire. Believing that there is no other way out, she wraps the child in blankets and jumps out of her third-floor window. Though she tries to absorb the impact herself, the child sustains injuries.
This strikes me as a strong example of a case where one’s intuitions are likely to be that a) the defendant is not criminally culpable, and b) the presence of psychotic symptoms is relevant to that non-culpability. But what exactly is the nature of that relevance? On a rational-capacity analysis, the exculpatory narrative would run something like this: Ruth’s behaviour is motivated by an irrationally held belief that the building is on fire. Her behaviour, though coherent in light of that belief, is a case of applying functional means-end reasoning to an irrational premise. As such, she fails to meet the minimum standard of rational capacity necessary for criminal responsibility and must be regarded as non-culpable.
This narrative gets a lot right – it recognises the impact of Ruth’s experience of psychosis on her decision-making, and leads us to the right conclusion (non-culpability). There is, however, a sense in which it fails to tell the whole story. There’s a dimension to Ruth’s intuitive exculpability that’s missed on this analysis, for what makes Ruth such a sympathetic example is not simply the degree to which her decision-making is impaired by her condition, but how much she gets right, even so. Specifically: being properly motivated by appropriate care and concern for her child, she does the best she can think to do in the situation as she perceives it to be in order to protect her child from coming to greater harm. This seems to me a crucial dimension of the exculpatory story, here, and it’s one that a rational capacity model is unable to capture. Understanding Ruth’s non-culpability in terms of rational incapacity puts the focus firmly on her rational failures – failures which, if significant enough, write the whole decision-making process off as insensitive to moral and/or legal analysis. There’s no room, on such an understanding, to factor in any moral or valuational successes. But what the Ruth example shows is that these successes can be as much a part of the exculpatory story as the rational failures.
At this point, the rational capacity theorist may wish to say something along the following lines: It may be true that the Ruth example is poorly served by a rational capacity analysis, but this is no surprise, since Ruth’s problem is not a lack of rational capacity. Rather, it is that she is faultlessly subject to a number of major visual and auditory misperceptions, in light of which her motivating mistaken belief (that the building is on fire) is not an irrational one. Outside of these specific misperceptions, furthermore, her rational faculties appear to be in perfect working order. All the example shows, therefore, is that a) Ruth may not be a typical example of a person experiencing psychotic symptoms, and b) Ruth isn’t the kind of defendant to whom a rational incapacity analysis was ever intended to apply. Rather, her exculpation would properly proceed from the ordinary application of culpability principles around faultless mistake. In response to this potential objection, I would first emphasise that my aim was never to claim that no cases exist in which the rational capacity analysis is the right analysis, but merely to challenge it as the sole or primary analysis of exculpation in cases involving psychotic illness. In this respect, the Ruth example serves its purpose. In partial concession to the point, however, I think it is possible to frame a less artificial example, involving more comprehensive rational failures, that nonetheless gives rise to the same concerns:
R1: During an episode of psychosis, Rachel suffers from paranoid delusions that her neighbours are planning to kill their infant child. She attempts to contact both the police and social services on multiple occasions, but her concerns are consistently dismissed. Eventually, hopeless and terrified, she breaks into her neighbours’ house and abducts the child, absconding to what she regards as a safe location. She and the unharmed infant are shortly discovered by police.
This case is less easily dismissed as a matter of misperception affecting an otherwise functional rational process. Unlike Ruth, there is no interpretation on which Rachel’s motivating belief is anything other than unfounded and irrational. Yet, there remains a sense in which an analysis of Rachel’s non-culpability in terms of incapacity fails to satisfy. As with Ruth, there’s a sense that while Rachel’s decision-making faculties are undeniably impaired, appeal to that impairment in isolation cannot provide a full exculpatory narrative – for that, we need to factor in Rachel’s admirable protective impulse, her concern for an unrelated child’s wellbeing, her willingness to put herself at risk to secure the safety of that child, her understandable desperation following the repeated dismissal of her very real fears, and so on.
There are several potential ways of framing the missed element, here. Perhaps we could say that Rachel’s conduct, when considered in light of her delusional beliefs, isn’t culpable because it doesn’t display insufficient concern or culpable indifference, or that she engaged in something like uncorrupted deliberation with respect to her values and the weight she attaches to various guiding principles. However we prefer to analyse it, the point is that although there’s a lot going wrong with Rachel’s decision making here, there’s also a fair amount that’s going right – and this feels as integral to the story of her non-culpability as presence of delusions.
For those still unconvinced, let’s introduce a contrasting example. So far, we’ve been looking at R-Type cases in which the likely impulse is to exculpate. Not all R-Type cases give rise to similarly sympathetic intuitions. Consider the following case:
R2: During an episode of psychosis, Roland suffers from paranoid delusions that his wife is having an affair. In the course of a heated argument in which she repeatedly denies his accusations, he beats her, causing minor injuries. Although he feels some regret for the harm he caused, he maintains that his actions were justified and necessary, as it is a husband’s right and his duty to discipline his wife for infidelity.
Roland is (I would posit) far less intuitively exculpable than Rachel, and this disparity is not attributable to any difference in either the nature or the degree of their respective rational impairments. Rather, it is because of everything that goes into their decision-making processes outside of those impairments: their values, what and who they care about, the moral principles they hold, and so forth. If exculpation is understood in terms of a lack of rational capacity, however, then the two cases stand and fall together: either Ruth and Roland meet the capacity threshold, or they do not. The factors which might be regarded as differentiating the two cases necessarily fall out of consideration on an incapacity analysis; if Rachel and Roland lack the rational capacity required for legal responsibility, we can no more regard the moral and valuational dimensions of their decision-making as legally significant than we could the values and motivations of a child.
A few caveats. Firstly, there’s an important distinction to be made here between rational capacity and rational impairment. Clearly, rational impairment may be relevant to the story of why a Ruth or a Rachel is non-culpable on the facts. This doesn’t mean, however, that it’s appropriate to tell that story in terms of rational capacity, for this implies an insensitivity to rational and/or moral analysis that may leave us unable to arrive at the right exculpatory account.
Secondly and relatedly, the position advanced here is not that the behaviour of a Ruth or a Rachel can be interpreted as rational once a specific delusional mistake is granted. The behaviour of a defendant experiencing psychosis is likely to be understandable only in light of quite extensive impairments to their reality testing at the relevant time, and factoring in these wider impairments may be crucial to returning non-culpability-inferring answers to questions about their conduct (Why didn’t you take further steps to confirm your beliefs before acting? Why did you ignore obvious and available alternative courses of action? Etc.). It is important, therefore, that neither example relies on the claim that the behaviour in question can be interpreted as rational (or rational given a particular irrational belief). The force of the Ruth and Rachel examples comes instead from the challenge they pose to the common-sense framing of those rational impairments in terms of incapacity – a framing which leaves us unable to account for the exculpatory relevance of any decision-making successes.
Legal Implications and Final Thoughts
I have kept the legal implications of my argument deliberately vague, firstly to keep the focus on the theoretical question, and secondly because this discussion really requires separate and detailed treatment. Still, there are a few things to say before closing. If my argument is accepted, the task before us is to figure out how the complex relationship between psychosis and culpability can be modelled within legal culpability tests outside of appeals to incapacity or non-responsibility. This task is not an easy one, and may require some rethinking of our general approach to mens rea and doctrines of excusing mistake. Certainly, it seems as though concepts of mistaken justification and/or necessity are likely to play a role here, though whether these are capable in their current construction of satisfactorily accounting for delusional mistakes is a serious question. Recent years have seen courts grappling with the idea that a delusional mistake may be capable of grounding a defence other than insanity – though they have not, as yet, shown any real willingness to take this line of thought all the way. Of the relevant cases, Taj may be of particular interest on its facts: while in a state of drug-induced psychosis, D formed the belief that the victim (V) was a terrorist with an explosive device in his vehicle. He subsequently attacked V with a tyre lever and an aerosol can, his intention being (in his words) “to incapacitate [V] or disable him from going and detonating that bomb”. Taking D’s claim at face value (and disregarding any complicating elements introduced by his intoxication), the case presents us with a non-hypothetical example of an R-Type case. Given this, the Court of Appeal’s consideration (and subsequent rejection) of mistaken self-defence here is both interesting and potentially instructive. Discussion of the judgement and its implications is beyond my present scope, however; my aim at this stage has not been to propose any specific alternative legal approaches, but simply to prepare the ground for such debates by illustrating the limits of appeals to incapacity as the theoretical basis of mental condition-based exculpation. To conclude, I would reiterate that nothing in the above argument should be taken to suggest that there exist no cases where a defendant’s non-culpability can only be understood in terms of the absence of rational or decision-making capacity. If, however, we take this kind of analysis to be the fundamental theoretical basis of mental condition-based exculpation, we will struggle to account for cases where some presumption of rational or moral capacity is required in order to tell the right exculpatory story, and will have to either a) wilfully misinterpret those cases, or b) exclude them altogether. What the R-Type cases demonstrate, then, is not that a psychotic defendant can never lack rational capacity to the degree that he cannot be held to legal standards of culpability, but rather that reducing the impact of psychosis to a kind of incapacity may come at the expense of a full and satisfactory analysis of a defendant’s culpability.
 See e.g.: S Morse, ‘Crazy Behaviour, Morals and Science’ (1978) Southern California Law Review, 51, 529; M S Moore, Law and Psychiatry: Rethinking the Relationship (Cambridge Paperback Library 1982); V Tadros, ‘Insanity and the Capacity for Criminal Responsibility’ (2000) Edinburgh Law Review, 5 (3). pp. 325-354
 Not here, at any rate – although I’ll admit that I have certain reservations.
 I focus here on cases where D really regarded their actions as morally required, as opposed to just not morally prohibited. I think these cases give rise to stronger intuitions. But it is conceivable that my argument, or a version of it, could be repurposed to apply to cases in which D regarded their actions as not morally prohibited.
 L Alexander, ‘Insufficient Concern: A Unified Conception of Criminal Culpability’ (2000) California Law Review 88, 931-954
 K W Simons ‘Punishment and Blame for Culpable Indifference’ (2015) Inquiry: An Interdisciplinary Journal of Philosophy, 58:2, 143-167
 G Yaffe, ‘Excusing Mistakes of Law’ (2009) Philosophers’ Imprint, 9:1-22
 R v Taj  EWCA Crim 1743; R v Oye  EWCA Crim 1725. Both were cases in which the court were tasked to consider the relationship between self-defence and delusional honest mistake. See also R v Martin (Anthony)  EWCA Crim 2245 and R v Canns  EWCA Crim 2264.
  EWCA Crim 1743
 R v Taj , at 11