Skip to content

Information for students, faculty and staff regarding COVID-19. (Updated: 26 February 2021)


Responsibility and disability

Research project The philosophical literature on moral responsibility traditionally assumes an either-or-view: A person is either a morally responsible agent, or not at all responsible for what they do. The purpose of this project is to discuss how we ought to relate to all those in the grey area in between – including ourselves, when this applies to us. I focus on everyday interactions and friendship, rather than on forensic or medical issues.

When an agent with a psychiatric or neuropsychiatric disability wrongs us, how are we to respond? Should we blame her or not? Theories of moral responsibility often have trouble with borderline cases – i.e. agents who have the relevant capacities, but to a diminished extent. According to an influential philosophical tradition, people can be divided into two groups; those we can have normal relationships with, including blaming them when they do wrong, and those we regard as objects to be handled. It is also popular to distinguish between different kinds of moral responsibility, and argue that slightly impaired agents can be responsible in one sense but not another. But this kind of theory, too, relies on the assumption that sharp distinctions can be made. I will develop a theory about when and how to hold agents with psychiatric and neuropsychiatric disabilities responsible which takes into account nuances and grey areas. My focus will be on everyday interactions and friendships, rather than on clinical or forensic settings.

Head of project

Sofia Jeppsson
Associate professor

Project overview

Project period:

2019-01-01 2021-06-30


This project receives funding from the Swedish Research Council.

Participating departments and units at Umeå University

Department of Historical, Philosophical and Religious studies

Research subject

Philosophy and philosophy of science

Project description


When an agent with a psychiatric or neuropsychiatric disability wrongs us, how are we to respond? Should we blame her or exempt her from blame? Moral responsibility theories often have no obvious way to handle borderline cases –i.e. agents who have the moral responsibility grounding capacities, but only to a diminished extent. According to an influential philosophical tradition beginning with Peter Strawson (1962/2013), we can either have normal relationships with people, which include blaming them when they do wrong, or we can regard them as objects to be handled and managed, and exempt them. This tradition rests on the assumption that people can be neatly divided into “the normal”, with whom we have normal human relationships, and “the insane”, whom we can merely try to “manage” in the best possible way. It ignores all those agents who fall somewhere in between fully normal on the one hand and completely insane on the other.

It is also popular to distinguish between different kinds of moral responsibility, and argue that slightly impaired agents can be responsible in one sense but not in another. But this kind of theory, too, relies on the assumption that sharp distinctions can be made, between kinds of responsibility and kinds of agents.

Proponents of both theories, furthermore, tend to assume that we normally have good knowledge about whether someone is normal or insane, or lack this or that morally relevant capacity.

I will develop a theory about when and how to hold agents with psychiatric and neuropsychiatric disabilities responsible, which takes into account the nuances and grey areas that reality contains. My theory will acknowledge the following conditions:

  1. It is of vital importance to people to feel included and to be treated with respect.
  2. We often do not know precisely which capacities a psychiatrically or neuropsychiatrically disabled person has.
  3. Capacities come in degrees. It is common to suffer difficulties and to have slightly compromised capacities, but still not completely lack any morally relevant ones.

My focus will be on everyday interactions and friendships, rather than on clinical or forensic settings. I will furthermore mostly focus on the issue of exemption from responsibility, rather than on excuses –that is, I will mostly focus on the question of whether a wrongdoer was the kind of agent that can be morally responsible for what she does. Since many of us live with psychiatric or neuropsychiatric disabilities, and even more people have friends or family with this kind of disability, this project is highly significant for our lives.


I will develop a theory about when and how to hold agents with psychiatric and neuropsychiatric disabilities (henceforth, PND agents) responsible, which takes into account the nuances and grey areas that reality contains. My focus will be on everyday interactions and friendships, something which deeply concerns everyone who either has this kind of disability herself or is friends or family with one, rather than on clinical or forensic settings.

In what follows, I will outline the main steps in the project. Each such step corresponds to a paper and is marked with “P” below. Two papers (P1and P3) will be longer and take more time and effort to write than the others (P2andP4), which will build on the former. Papers 1 and 2 will be written during 2019, and 3 and 4 during 2020. I intend to write single papers to be published in international philosophical journals, but also to publish popular versions of these papers in, e.g., psychology and culture magazines.

Paper 1. PND agents of whose responsibility we are uncertain: How to err on the side of caution.

In the first part of the project, I will determine which attitude to take up and how to treat a PND wrongdoer of whose responsibility we are uncertain.

Uncertainty about her moral responsibility is common when faced with a PND wrongdoer. Her mental impairments might seem to affect moral responsibility, whilst not completely undermining it. When we feel uncertain in this way about the moral responsibility of a PND friend who has wronged us, we face a problem: should we blame her or not? We cannot solve this problem by simply suspending judgment about her moral responsibility, because there is no neutral way to treat her that would correspond to such suspension. If I do not react to her wrongdoing at all – I do not blame her, I do not express anger or sadness and so on – I have, in effect, exempted her. Thus, even if I privately suspend judgment, I must still choose whether to hold her responsible or not.

Common sense dictates that in situations of uncertainty, we ought to err on the side of caution. But which side is the cautious one? In order to answer that question, I must firstly determine how many distinct attitudes we can take up towards the wrongdoer and the different options we have for how to treat her, and secondly what their respective advantages and drawbacks are. In the literature so far, no one has (to my knowledge) tried to list and compare all the different attitudes written about. They are at least five in number: the participant, the objective, the nurturing, a version of the participant one where sadness replaces anger, and a let-go-attitude.

Peter Strawson (1962/2013) famously distinguished a participant attitude from an objective one. When interacting with other normal adults, we normally take up a participant attitude. We care about whether their actions express a good, ill or indifferent will towards us. When their will is malicious or indifferent, we respond in turn with resentment or anger – we blame the wrongdoer. “Hopeless schizophrenics” and other people with whom we cannot have normal adult relationships, on the other hand, are regarded with an objective attitude. With such people, we want to know how they think and behave merely in order to manage, handle, cure or train them, but we do not really care about how they feel about us and we do not blame them if they lash out against us. Strawson leaves the question open of what to do with people who fall somewhere in between “normal” and “insane”.

However, it hardly seems appealing to be regarded “objectively” –the objective attitude is an isolating one, shutting people off from normal relationships (Wolf 1981; Shabo 2012; Nichols 2016). Since we ought to treat other people with respect, we might therefore think that as unpleasant as it is to be angrily blamed, we should take up a participant attitude towards all but the most extreme PND agents. It is not obvious, though, that there are only two attitudes we can take towards wrongdoers.  

Hannah Pickard (2013) writes about how clinical staff in real life hold mental patients responsible for their behaviour, even though they abstain from angrily blaming them. Daphne Brandenburg (2017) builds on Pickard in her theory about a separate nurturing stance that parents can take towards little children as well as clinical staff towards mental patients. Derk Pereboom (2014), in a discussion about moral responsibility scepticism, argues that we ought to hold each other responsible in a sad rather than angry way, and Per-ErikMilam (2016) writes about the possibility of having an adult relationship without holding people responsible for their wrongdoings at all.

I will explore all the different attitudes or stances we can take towards a wrongdoer. I will determine which the elements are of which an attitude or stance is constituted, and which element or elements differ between the stances described by Brandenburg, Pereboom, Milam and others. They might differ, for instance, in whether they entail responding to the wrongdoer with anger, sadness or merely understanding, and whether they entail seeing the wrongdoer as one’s equal or as someone to be nurtured and helped up to one’s own level. When these elements are identified, it is also possible that further separate attitudes or stances can be distinguished, attitudes or stances hitherto not described in the literature. When these different attitudes and stances have all been identified and described, it will be possible to determine which their respective advantages and drawbacks are, and thus which is the cautious side (or sides) on which to err when we are faced with a PND wrongdoer of whose moral responsibility we are uncertain.

Paper 2. Backward-looking vs forward-looking reasons to hold responsible.

In the first paper written during this project, I will determine which drawbacks and advantages various attitudes that we can take up towards PND wrongdoers have. Without knowing, yet, which they are, it is still clear that drawbacks and advantages can be broadly divided into forward-looking or consequentialist ones, and backward-looking or deontological ones. The second part of the project will determine how to weigh forward-and backward-looking reasons to hold or not to hold someone responsible.

Some moral responsibility theories heavily stress the consequentialist advantages of holding responsible. J. J. C. Smart (1961) argues that the only real reason we have for labelling people good or bad, virtuous or vicious, and to praise and blame them for what they do, is that such practices are useful and can serve to encourage good behaviour and discourage bad (see also Dennett 1984). Lately, Manuel Vargas (2013) has provided a more sophisticated consequentialist theory of moral responsibility according to which our practices of praise and blame are broadly justified since they can help foster more responsible agency.

However, even if holding someone responsible has good consequences, it seems problematic to do so insofar as the responsibility holding is unpleasant for the wrongdoer and she does not really deserve it (Pereboom 2014: 129-130). Nelkin (2016) and Wolf (1990: 86-87) argue that if it is very difficult for an agent to do the right thing, she is less blameworthy for doing wrong; possibly, if her difficulties are extreme enough, not blameworthy at all. It might be hard for a PND agent to do the right thing for a number of reasons, depending on her precise disability; it might be difficult for her to wrap her head around the fact that her behaviour is hurtful or offensive to others, she might have very bad impulse control and so on. If so, we have a backward-looking or deontological reason not to blame her (or at least not to blame her much) for doing wrong. This reason comes into play even if we merely suspect, but do not know, that the PND wrongdoer has diminished responsibility for what she did due to her impairment; insofar as being blamed is unpleasant, we should be fairly certain that the agent is blameworthy before we blame her.

How are we, then, to weigh our forward-looking reasons to blame against our backward-looking reasons not to blame PND wrongdoers who might be somewhat but not fully responsible for what they do, or of whose responsibility we are fundamentally uncertain? This question is not only relevant when we discuss angry blame. Since it can be profoundly unpleasant to be exposed to other people’s sadness or disappointment over what one has done (Jeppsson 2016), it is also relevant for sad or disappointed blame, and perhaps further kinds as well. Determining how to weigh these reasons is the purpose of the second paper I will write during this project.

Paper 3. PND agents and different kinds of moral responsibility.

David Shoemaker (2015) argues that PND agents can be morally responsible in one or two senses whilst not in another. I will critically assess the premises on which he builds his argument, and investigate whether there are alternative explanations for our mixed feelings for PND wrongdoers.

Some philosophers slice moral responsibility up into different kinds, and argue that PND agents might be barred from having one or two kinds of moral responsibility by their impairments whilst still being responsible for what they do in another sense. Since Gary Watson’s 1996 paper “two faces of responsibility”, it has been popular to divide moral responsibility into attributability and accountability, and argue that some impaired agents who cannot be accountable for what they do can be responsible in the attributability sense (see also Fischer and Tognazzini 2011). David Shoemaker (2015) adds a third kind of responsibility, answerability, to the two previous ones. He discusses different kinds of psychiatric and neuropsychiatric disabilities and which kinds of moral responsibility they do and do not preclude at length, attempting to build a theory that can make sense of the mixed feelings we often have upon encountering PND wrongdoers. According to Shoemaker, we are often ambivalent rather than uncertain about the moral responsibility of PND wrongdoers, and only his tripartite theory can make sense of this ambivalence. I will question whether we really are ambivalent. It is not obvious that we can always distinguish between ambivalence and uncertainty through introspection. If ambivalence is not necessarily what we feel when confronted with PND wrongdoers, one motivation for the tripartite theory is undermined.

Another important premise of Shoemaker’s is that PND agents frequently completely or almost completely lack one or more morally relevant capacities; therefore, they lack certain kinds of moral responsibility, even if they have other kinds. Shoemaker does add the caveat that capacities often come in degrees (Shoemaker 2015: 120), but his focus throughout the book is still quite consistently on extreme cases who completely lack one capacity or other. If we take seriously the insight that capacities often come in degrees, it becomes less plausible that our mixed feelings depend on us intuitively realizing that the wrongdoer has, say, attributability responsibility for what she did whilst completely lacking accountability responsibility.

Even if the tripartite theory of moral responsibility is argued for from questionable premises, however, we still want some theory that explains the mixed feelings we often have for PND wrongdoers. Some interesting avenues to explore includes the possibility that their difficulties might diminish without extinguishing their moral responsibility (even if we think of moral responsibility as one single thing) (Nelkin 2016; Wolf 1990: 86-87), and that conditions such as autism might render some PND agents non-culpably ignorant about the fact that they did something hurtful. The thesis that non-culpable ignorance excuses is fairly uncontroversial, and does not require slicing moral responsibility up into different kinds.

However, it remains to be seen how far alternative solutions can take us. It is possible that we still need something like the tripartite theory in order to account for our moral responsibility judgments of some PND agents.

Paper 4. The importance of accountability.

In the final part of the project, I will assume for the sake of argument that some PND agents (e.g., autists) completely lack accountability responsibility, and then investigate what the implications are for friendships and other relationships.

Peter Strawson argued that we ought to take up an objective attitude towards the insane. However, the objective attitude isolates and shuts one out from ordinary relationships. Shoemaker argues that, among others, autists and psychopaths cannot have accountability responsibility for what they do. They are therefore not eligible for angry blame, one of the elements of Strawson’s participant attitude. Shoemaker does not, however, believe that we can only regard autists and psychopaths as objects to be handled, managed, cured or trained, since he argues that we can hold them responsible in some respects; we can label them cruel or kind and grade their judgment as better or worse. Still, how isolating is it to be shut off from accountability? How close are we to regarding a PND agent with Strawson’s objective attitude, if we regard her as someone who cannot be accountability responsible for what she does?

Even if we assume for the sake of argument that the tripartite theory is largely correct, and that some categories of PND agents lack accountability for their actions, we might have strong reasons to treat them as if they were accountable if the alternative is cold and isolating, and even precludes friendships and normal adult relationships in general. Whether this is so, depends on how big a role answerability and attributability responsibility can fill in friendships and other relationships.

According to Shabo (2012), we cannot have deep personal relationships unless we take it personally when people hurt us, and he argues that taking things personally occasionally requires getting angry. However, the debate in which Shabo argues for this thesis is concerned with holding responsible versus not holding responsible at all. Whether holding people attributability and/or answerability responsible for what they do would suffice for deep and personal relationships is an avenue that largely remains unexplored