John Gardner at Home

'Why Blame?', forthcoming in a volume edited by Iyiola Solanke in honour of Nicola Lacey.
Download the preprint here.

Why Blame?

1. Blame scepticism

The uneducated person blames others for their failures; those who have just begun to be instructed blame themselves; those whose learning is complete blame neither others nor themselves.[1]

So says Epictetus, spelling out one tenet of Stoic thought: that blame, whether of oneself or another, has no place in a life wisely lived. To blame is unhealthy and dispensable. This tenet long endeared me to Stoicism. For I was, for many years, what Peter Graham calls a ‘blame sceptic’.[2] That is not to say that I resiled from blaming. Rather, I blamed and then reproached myself for doing so. Since reproaching entails blaming, I thereby compounded my felony. And then, reproaching myself for compounding my felony, I compounded it some more.

Blame sceptics come in many stripes. Numerous and notorious among them are responsibility sceptics. They doubt whether anybody is ever responsible for anything. Rightly thinking that being responsible is a necessary condition of being blameworthy, they doubt whether blame is ever in order. Yet that is only the start of their troubles. Responsibility matters for so much more than blame. If I lack responsibility, what I do is no special concern of mine. Why, then, should I pay special attention to what I do? In particular, why should I pay special attention to the advice of responsibility sceptics and stop blaming? If I lack responsibility, it is no special concern of mine whether I blame or not. As you can see already, it is doubtful whether responsibility scepticism is a coherent position. Realizing this, some responsibility sceptics restrict their scepticism to ‘moral responsibility’, supposedly a special kind of responsibility that, if it existed, would be the one to open the way to blame. What the word ‘moral’ adds here is generally left unclear. Does it mean that our blame-scepticism should be restricted to moral blame, and should not extend to the cases in which, e.g. we only have ourselves to blame for missing the last train home? Be that as it may, this ‘moral responsibility’ is very commonly portrayed as a magical property of persons, possessed under conditions so ratcheted up by philosophical fantasists that scepticism about it is unavoidable. The lesson of their work is not that one should be a responsibility sceptic. Rather one should be a sceptic about philosophical fantasies according to which responsibility, moral or otherwise, requires possession of some magical property such as ‘contracausal freedom’.[3]

Responsibility scepticism will not concern us further here. It is a blind alley. Blame-sceptics of a more interesting stripe – I will call them judgment sceptics – are sceptics about standards, or some kinds of standards. Blamers, they say, purport to apply standards of rectitude that exist, and apply to those who are blamed, independently of the invocation, application, acceptance, endorsement (etc.) of those standards by the one who is blamed. No such independent standards exist, says the judgment sceptic, or at least none that are suitable to vindicate judgments of blameworthiness. Thus Bernard Williams writes:

[W]e can blame a man (we may think) for neglecting his wife even though he has no motivation to be concerned about his wife. So if blame is necessarily connected with reasons, it seems to be necessarily connected with external reasons [reasons that exist independently of the agent’s motivations]. Therefore, if there are no such things as external reasons for action, there is something suspect about blame.[4]

Here are two possible answers to Williams, each of which strikes me as both true and dispositive: (a) there are external reasons for action and (b) in judging someone blameworthy, one need not invoke external reasons for action. I will simply assume the truth of (a) here. Meanwhile, we already have an example of (b) to hand. ‘You only have yourself to blame for missing the last train home’ is commonly uttered to someone who was indeed motivated to catch the train. Blame is not necessarily connected to external reasons for action. At least some blame, then, seems to escape the Williams dragnet even if you agree with him that there are no such things as external reasons for action.

Judgment sceptics join responsibility sceptics in doubting whether anyone is ever blameworthy. Standing-to-blame sceptics differ in allowing that people can indeed be blameworthy. But they doubt the suitability of any of us to be the one who actually does the blaming. ‘He that is without sin among you,’ cautions Jesus in a standing-sceptical vein, ‘let him cast the first stone.’[5] This is just one of several flavours of blame-scepticism in which problems are supposed to begin, not with the soundness of our judgments, but with the attitudes, emotions, or actions to which those judgments give rise. The underlying thought here is that actually blaming someone – whether that involves taking a certain attitude to her, feeling a certain way about her, or addressing or treating her in a certain way – involves optional further steps beyond a mere judgment of his or her blameworthiness. This formulation is already problematic. For perhaps judging someone blameworthy is also an optional further step beyond her merely being blameworthy? Could one not instead suspend judgment? ‘Don’t judge me,’ says the attention-seeker on social media, nervous that he may have crossed the limits of style or taste. Does this mean ‘pay no attention’ (an unlikely position for a social media attention-seeker!) or ‘don’t form any opinions’ (how could I not?) or ‘give me the benefit of the doubt’ (isn’t that judging?) ... or what? We will have cause to return to this ‘don’t judge me’ problem below. For now let’s keep the making of bare judgments of blameworthiness out of the picture. Blaming, let’s say for now, means going a bit further than that.

But how much further? Nicola Lacey and Hanna Pickard are efficacy sceptics about blame. Like the standing-to-blame sceptics they allow for blameworthiness but they caution against the move to blame itself. That is because they think that blaming is typically counterproductive. To say this, they have to assess it against some aim or aims. In the main they have responsibility-accepting and (thereby?) behaviour-improving aims in mind. But never mind that for now. For now our interest is in what, for them counts as the blaming that we should be efficacy-sceptical about. They call it ‘affective blaming’:

Affective blame, as we define it, is the range of hostile, negative attitudes and emotions that are typical human responses to blameworthiness. It can include, for instance, hatred, anger, resentment, indignation, disgust, disapproval, contempt and scorn, and can be manifest in any number of ways, including seeking retaliation, retribution, and vengeance.[6]

Here are three queries arising out of this passage.

(i) Lacey and Pickard lump ‘attitudes and emotions’ together. ‘Affective’ seems to be a good label for negative emotions of the blamer but a bad label for negative attitudes. In the light of what someone did, I can (dispassionately) lose faith in her or (dispassionately) lose admiration or respect for her.[7] Then I take an attitude towards her which lacks any affective ingredient. She is lowered in my estimation but I have no feelings on the subject. Here I am not merely judging her blameworthy. I am actually blaming her. But in what sense is the blame affective?

(ii) True, we may be disappointed in, or despair of, or despise those whom we blame. Then there is something affective going on. But are these really ways of blaming, or are they just possible emotional consequences of blaming? Couldn’t I say ‘I’ll always blame him for what happened and that’s why I’ll always hate him’? Isn’t blaming best thought of, then, as an intermediate (attitudinal) step between judging someone blameworthy and harbouring hostile feelings towards that someone? (Note that this is compatible with there being some emotions, such as remorse or vengefulness, that are partly constituted by the blaming attitude; something can exist in consequence of one of its constituents.[8])

(iii) The harsh ways of treating the person blamed that are listed at the end of the passage (‘retaliation, retribution, and vengeance’) are certainly ‘typical ... manifestations’ (and typical consequences) of the ‘affective blame’ described. However the blamer can in principle avoid the typical manifestations, unlike the typical emotions, by deciding to avoid them. One can be a blamer who is full of resentment, disgust, or scorn and yet not do anything to reveal or convey how one feels, or even that one blames, to the person blamed (or to anyone). True, concealment is easier if one does not have to fight against strong feelings. Still, one can make an effort to hide how one feels and often one succeeds. Myself, I often make the effort (e.g. when dealing with a cold caller) and sometimes but not always I succeed.

Of these, query (iii) is the most pressing. Perhaps what one conveys to the person blamed (or indeed to others) should be judged by its efficacy relative to some aim or aims. Efficacy is a suitable standard for judging actions, including expressive ones. But should the attitudes and emotions that one expresses be judged in the same way? Surely attitudes and emotions are to be judged by how they befit their objects, not by what they achieve?[9] One should love the loveable, trust the trustworthy, admire the admirable, regret the regrettable, be disgusted by the disgusting, give credit for the creditable, and (by the same token) blame the blameworthy. The case of belief is similar: one should believe the believable (a.k.a. the credible). Of course it might be that a certain belief is giving one a lot of trouble, exposing one to ridicule, causing one to lose sleep, sending one off the rails, etc. But is that strictly speaking a reason against the belief, or is it only a reason to keep quiet about believing it, develop strategies to avoid losing sleep over it, etc.? This choice is too stark. There is an intermediate possibility. That a certain belief is troublesome is not a reason against the belief. Yet it is a reason to take steps to rid oneself of the belief – not merely to manage its negative effects. Even though the belief is fully in order qua belief, one has reasons to expunge it. Likewise, it seems, with troublesome attitudes and emotions, including blame and its cognates. Imagine that one’s disposition to blame the blameworthy (or, say, to resent them or despise them) is interfering with one’s work as a therapist or probation officer or social worker or schoolteacher. One option is to find a new line of work. But first one might try to contain one’s disposition to blame, say by taking a closer interest in the blameworthy people one deals with, and learning to focus on something about them other than their blameworthiness.[10] True, one needs some such indirect strategy to contain attitudes and emotions that one still regards as befitting their objects. The unwelcome attitudes and emotions do not disappear merely because one realizes that it would be better if they did. Nor can one lose them simply by deciding to lose them. Yet they can at least sometimes be tackled, and the mere fact that they still befit their objects is not enough to make a conclusive case for holding onto them.

It is a long way, however, from here to the thesis that blame, even affective blame, is to be judged by its efficacy, where this means assigning some aim or aims to it. Attitudes and emotions do not have aims.[11] What has an aim, or often has an aim, is the expression or communication of an attitude or emotion. When Lacey and Pickard worry about the inefficacy of blame, then, they are not exactly blame-sceptics, or even affective-blame-sceptics. They are something more like reproach-sceptics, where reproach is an expression of blame conveyed to the person who is blamed.[12] Reproach is the right kind of thing to be judged by its efficacy; blame, by contrast, is not. One could be a reproach-sceptic even though one is an enthusiast for blaming. In fact one could be a reproach-sceptic because one is an enthusiast for blaming. One might think that reproaching, being inefficacious, gives blaming an undeservedly bad name.

2. The dispensability of blame

I have considerable sympathy with Lacey and Pickard’s reproach-scepticism, and with the ‘counterproductivity’ case they make for it. But in this paper I want to develop and test a second thesis that they advance less conspicuously, one that (if sound) would support a far-reaching blame-scepticism. They say this:

In keeping with the justice model, [our] model judges patients responsible ... for wrongful or harmful conduct to the extent that they possess the relevant cognitive and volitional capacities in relation to it. But in contrast, it resists any corresponding tendency towards affective blame. Put simply, according to [our] model, blameworthiness, understood as responsibility ... for wrongdoing, does not entail the ‘worthiness’ of affective blame.[13]

Although there is a lot going on in this passage, the core proposition seems to me to be this: Inasmuch as blaming is not to be judged by its efficacy (but instead e.g. by its justice) there is nothing that blame gives us that responsibility without blame would not also give us. Once we are willing to hold ourselves and each other responsible, blaming can be dispensed with. As Epictetus thought, it has no place in a mature moral outlook.

Possibly I am reading Lacey and Pickard wishfully,[14] for the core proposition that I have just distilled captures my own longstanding qualm about blame, the one that originally attracted me to Epictetus’ precept. One way in which I am clearly laundering the Lacey-Pickard position is this. Pace Lacey and Pickard, the dispensability objection (as I will call it) is not merely an objection to attitudinal blame and its associated emotions. It extends to judgments of blameworthiness too. The last sentence of the quoted passage, with its curious scare-quotes around ‘worthiness’, shows why. Being blameworthy just is being worthy of blame. If responsibility already gives us everything that we need, then not only our tendency to blame but our concern with blameworthiness must be superfluous. We can manage without the whole apparatus of blame-related thought, feeling, and action. Lacey and Pickard avoid this conclusion by equating blameworthiness with responsibility for wrongdoing. But this equation is a mistake. As they acknowledge in other remarks,[15] I may be responsible for wrongdoing without being blameworthy. Most obviously, I may have an excuse for what I did. It is only if I am responsible for wrongdoing that the question of whether I have an excuse for it arises. If I am not responsible for it, then how come the wrongdoing is mine to excuse?

It is tempting to associate the dispensability objection with the question ‘what is blame for?’ If one were an evolutionary psychologist, that might be a well-formed question.[16] But not for us. What we seek here is not just any explanation for the fact that people blame, but an explanation that brings out the rational intelligibility of their doing so (if there is any rational intelligibility to be brought out). ‘What is blame for?’ is a bad way to ask for that kind of explanation. We already know why. Attitudes and emotions can be reasonable or unreasonable; they answer to reasons. But qua attitudes and emotions they are not to be assessed in terms of the value of having them. The reasons to which they answer are facts about their objects that make them befit those objects: one should despise the despicable, love the loveable, etc. If blaming the blameworthy is superfluous, then, that is not because there is no value in blaming. (Blaming might be, for example, cathartic or fun.) Rather, it is because blameworthiness is a pseudo-property of its objects that is conjured up only to give blame something to befit. The whole apparatus, according to the dispensability objection, is rationally mysterious. If only we could get past our psychological attachment to blaming, we would see: it does not matter who is blamed and therefore it does not matter who is blameworthy.

Now I am associating the dispensability objection with a different question: ‘why does blame matter?’ You may think that the answer to this question is obvious, and that the dispensability objection is therefore easily dismissed. Blame matters because adverse consequences for the blamed person, such as reproach and punishment, are attached to it by the blamed person and/or by others. I should care whether I am being blamed because, if I am blamed, the next thing you know I will be accused, admonished, guilt-tripped, prosecuted, denounced, shunned, etc. That answer, however, gets the problem back to front. We already need to know why blame matters apart from such adverse consequences if we are to find out why such adverse consequences should ever be attached to it. It cuts no ice that such adverse consequences will be attached to blame whether they should be or not. Recall: we are interested in the rational intelligibility, not the empirical psychology, of blame and its associated practices. Why should adverse consequences be attached to something that, apart from those adverse consequences, does not matter? If I shouldn’t care whether I blame you, or for that matter whether you are blameworthy, why should I care to reproach you or punish you or otherwise convey that I do or might blame you? Why is it not enough, in particular, just to acknowledge and mark your responsibility without judging you blameworthy, without blaming you, and without attaching any blame-conveying consequences?

This, it seems to me, is the challenge[17] posed by Bernard Williams in the following famous passage:

The lorry driver who, through no fault of his, runs over a child, will feel differently from any spectator, even a spectator next to him in the cab, except perhaps to the extent that the spectator takes on the thought that he himself might have prevented it, an agent’s thought. Doubtless, and rightly, people will try, in comforting him, to move the driver from this state of feeling, move him indeed from where he is to something more like the place of a spectator, but it is important that this is seen as something that should need to be done, and indeed some doubt would be felt about a driver who too blandly or readily moved to that position. We feel sorry for the driver, but that sentiment co-exists with, indeed presupposes, that there is something special about his relation to this happening, something which cannot merely be eliminated by the consideration that it was not his fault.[18]

‘It was not his fault’ is a synonym for ‘he was not blameworthy in respect of it’ (or, equivalently, ‘he was not to blame for it’). The ‘something special’ that nevertheless connects him to the child’s fate is the driver’s responsibility. As already noted, some people like to say that, since the driver was not to blame, he was not morally responsible. But that, as Williams goes on to explain,[19] is a tendentious way to talk. Why should we (or the driver) care whether his responsibility is ‘moral’? If ‘moral responsibility’ just is blameworthiness, then we are back at our original question: Why should we (or the driver) care whether or not he was to blame? The question is pressing because, as Williams explains, the fact that the driver was responsible for what happened to the child already leaves its mark on his life, blameworthy or not. The various justifications and excuses that stand between him and blame are cold comfort. He may offer them. He may be bound to offer them. They may be accepted, agreed, endorsed. But, except inasmuch as they help to protect him from certain adverse consequences, what is the significance of these justifications and excuses? And if they have no significance apart from their ability to protect him from certain adverse consequences, then how do they serve to protect him against those consequences? Why would anyone care about anyone’s (including their own) justifications and excuses?

Here Williams gives us the dispensability objection in a nutshell, at least the version of it that always resonated with me. When we are responsible agents, and our actions have bad outcomes, that is bad enough already. Our lives are already blemished, sometimes even blighted. Various adverse consequences are already in the offing. Some relationships will never be the same; some debts will never be repaid; some negative feelings about what we did are bound to afflict us, and rightly so; some burdensome actions now have to be undertaken.[20] How does our blameworthiness make it any worse, or even any different? And if it doesn’t, why blame?

3. How we relate to reasons

As these remarks reveal, any attempt to meet the dispensability objection has to begin not with blame but with blameworthiness. The key thing we need to establish is that, quite apart from whether we are actually blamed, it matters whether we are blameworthy. Once that is established, it will be a relatively short step to defending blame, for blame will be the attitude that befits the blameworthy. This formulation may seem to suggest, implausibly, that blameworthiness has logical priority over blame. But that is not what I have in mind. To understand what blameworthiness is, I agree, we already have to grasp what it is to blame someone. All that I am adding is that, to make the resort to blame rationally intelligible, we first have to grasp why it matters whether someone is blameworthy. In short: while blame has logical priority, blameworthiness has justificatory priority.

Although other currencies might serve equally well, I will attempt to explain why blameworthiness matters in the currency of reasons. At this point I have in mind not our reasons to blame, but rather the ways in which we fall short relative to reasons when we are blameworthy. And although our own actions are not the only things in respect of which we can be blameworthy, I will simplify by focusing on our blameworthiness in respect of our own actions. This already reveals an important necessary truth about blame and blameworthiness. Both are of someone for something. If out of the blue I say ‘I blame Joe’, an essential piece of information is missing. What do I blame Joe for? Similarly with ‘Joe is blameworthy’ and ‘Joe is to blame’. For what? There is no such thing as being blameworthy ‘in the air’, to adapt a famous phrase from Pollock’s The Law of Torts.[21] If I am asked what I blame Joe for, and I say ‘oh, nothing in particular, I just blame him’, then I am making no sense. Maybe I am making an esoteric joke. Contrast ‘I trust Joe’, ‘Joe is to be trusted’, ‘Joe is trustworthy’. These already express complete thoughts. Part of the mystery of blameworthiness and blame is why they are, so to speak, essentially localized in this way: why I cannot be blamed, or be to blame, if not for something. I hope that thinking about why blameworthiness matters in the currency of reasons, focusing on reasons for action, will help to unravel this mystery.

Consider, then, the three ways in which responsible agents might fall short in their actions, relative to the reasons that militate for and against their performing those actions.

(a) Failure to conform to a reason

The most obvious failure relative to a reason for action is failure to do what the reason is a reason to do. There is a reason r1 to j or to c or to Y (let’s say, doing any one of those things would please my aunt) but I do none of those things. Clearly this alone does not make me blameworthy. For everything that there is a reason for me to do, there is invariably at least one reason for me not to do it. That is not a conceptual truth but it is a fact of the human condition. To find me blameworthy we need to begin by thinking about how competing applicable reasons fare in their competition with each other. We need to ask whether reason r1, which I did not conform to, was defeated by reason r2, which I did conform to. I did nothing that pleased my aunt, but I did something (incompatible with pleasing my aunt) that saved a life. Here we might say that what I did was justifiable. As we’ll see under sub-heading (b) below, justifiability alone does not rule out blameworthiness. Nor, as we’ll see under sub-heading (c), does unjustifiability guarantee blameworthiness. Nevertheless justifiability is the place to start in exploring the subject.

My cursory example of a competition between reasons for action (aunt-pleasing vs. life-saving) is deliberately ecumenical. My verdict in the example (life-saving defeats aunt-pleasing) is compatible with various thoughts about how one reason defeats another. I should emphasise that very often there is no defeat. There are undefeated reasons on both sides and one is left with a choice of justifiable actions.[22] But when one reason does defeat another, we may be inclined to think at first in terms of weight. All else being equal I have a weightier reason to see a more beautiful city (say, Rome) than a less beautiful one (say, Detroit): the more beautiful city will bring more beauty into my life. But this is not the only way in which one reason may be defeated by another. Sometimes one reason holds a trumping or pre-empting force, such that it is capable of defeating another independently of weight. I promised to be a commentator at a conference in Detroit over that May weekend, so now unfortunately I have to miss my cousin’s wedding in Rome. Could a bare promise really add enough weight to tip the balance against a May family wedding in Rome, in favour of a mere conference comment in Detroit? Unlikely. More likely, the promise excludes the joys of the Rome trip, or some of them, from the balance.[23] The point of promising was to bind myself, to give myself an obligation to go to Detroit. It is not an absolute obligation; if my son is suddenly hospitalized with meningitis on the eve of departure it would surely be justifiable to pull out. But just to take up a later invitation to a more beautiful city for a more fun event? That does not cut the mustard in the same way. My existing obligation to the Detroit folks rules out my accepting the Rome invitation, never mind how attractive the Rome trip is made. Feel free to adjust the example. Make Rome better; make Detroit worse. At some point you realise that weight relative to competing reasons isn’t the only factor bearing on whether a reason is defeated, and hence whether nonconformity with it is justifiable.

Some people associate blameworthiness with breach of an obligation. But that is too narrow. Recall from section 1: ‘You only have yourself to blame for missing the last train home’. You had no obligation to catch the last train or to get home at all. Nobody is waiting for you. But staying that extra few minutes at the pub was a stupid mistake. The only reason to stay (one last drink) was easily outweighed, hence defeated, by the main reason for leaving (one last train). That puts you, so to speak, in blame’s way. True, there might well be doubts about my standing to blame you, when you tell me the story at work the next day. Yet you clearly do have standing to blame yourself: not for breaching any obligation to yourself, I hasten to add, but just for making such a stupid mistake. So although it’s not for me to blame you, I might say, you only have yourself to blame.

(b) Failure to act for a reason

Sometimes what I do is justifiable, yet unjustified. I do what an undefeated reason would have me do, but I do not do it for that reason. I do it for some other (defeated) reason. A classic example is found in the English criminal case of R v Dadson.[24] Under the law of the time, a police constable was permitted to use lethal force against a felon in flagrante delicto, but not against the perpetrator of a mere misdemeanour. Stealing wood was treated as a felony only upon a third offence. Constable Dadson shot and killed a wood thief who, unbeknown to him, was a third-time wood-thief. He pleaded the lethal force rule. The court denied its benefit to him. Why? Because one acts with justification, not only in English law but also in ordinary life, only if one meets what lawyers call a ‘subjective’ condition. Not only must there be an undefeated reason to do as one does; one must also do it for that reason. Here’s another example, this time not from the law.[25] An arms dealer or an estate agent says: ‘If I didn’t do these terrible things, others would quickly fill the gap that I leave in the market and do even worse things. You’re lucky to have me. I’m actually protecting you against the really bad guys out there.’ We should reply: ‘Is that why you’re doing these terrible things? Are you doing them in order to stop other people doing worse things? If not, you are not offering a valid justification. You are just confirming that you are the villain we originally took you to be, and on top of that, a bare-faced evader of blame.’[26]

Justified action is action performed for an undefeated reason, not merely action that conforms to an undefeated reason.[27] In saying this we presuppose that ordinarily one’s failure to act for a reason that applies to one, even an undefeated reason, is unobjectionable. Barring exceptional cases, not acting for a given reason is not a failure relative to that reason, for a reason calls only for conformity.[28] That justified action is action performed for an undefeated reason reflects an independent principle of rationality (a.k.a. reasonableness). Those who conform to this principle act blamelessly. As noted already, justifiability does not rule out blameworthiness, but justification does.

(c) Failure of sensitivity to a reason

Our rationality is pervasively vulnerable to error. Frequently we treat defeated reasons as undefeated, or nonexistent reasons as existent. We take actions to be justifiable, or justified, when they are not, and we perform them accordingly. When our mistakes in such matters are themselves reasonable, we have excuses for what we do. An excused action, like a justified action, is blameless. But why? The quick answer is that when we are excused, we are duly sensitive to reasons and their force, even though we fail relative to them in other ways. The classic examples, also marked in the law, are those in which we act without justification owing to reasonable mistakes of fact, or reasonable but powerful emotions such as fear and anger. We act for what we have undefeated reasons to treat as undefeated reasons for what we do, but which in fact are not. In short, we do what we are justified in treating as justified.[29]

Some people extend the word ‘excuse’ beyond such cases to include a very different class of cases in which we are not responsible for our actions, e.g. actions borne of severe mental illnesses. When we are not responsible for our actions, we are not expected to be sensitive to reasons, or to act for them, or even to conform to them. We are exempt from rational assessment. The only objection I have to calling such an exemption an ‘excuse’ is that it is so far removed from the cases of excuse that interest us here. Here we are assuming responsible agency and considering those failures relative to reasons that are possible within the scope of responsible agency. One distinct type is a failure of sensitivity: one was not attentive or receptive enough to certain applicable reasons and made an error beyond the limits of excuse.

If Williams’ lorry driver writes to the parents of the injured child, or visits the child in hospital, or sends flowers or toys, he may be exhibiting his sensitivity to some of the reasons that, alas, he did not conform to when he knocked the child down. The explanation for his nonconformity, he thereby asserts, was not his lack of sensitivity. That this point can be made afterwards reveals that sensitivity to reasons, unlike conformity to reasons, is conceived as a property of agents, not a property of individual actions. One might of course in one solitary action show a momentary insensitivity to some reason, constituting a puzzling aberration in an otherwise admirably sustained record of sensitivity to reasons of the same kind. The point is that (barring special cases where responsibility is lacking) such a momentary insensitivity still lies on the record of the agent even when it is an aberration. In such a case one can certainly say ‘that wasn’t like me’ but one does not so easily get to say the responsibility-denying ‘that wasn’t me’.[30] For one did reveal, alas, what one is capable of, and thereby left a stain on one’s reputation (or at any rate on the reputation that one deserves to have).

A fairly common view has it that blaming targets, as it were, the agent as opposed to the action. Blameworthiness on this view is insensitivity to reasons, never mind conformity with them.[31] But this is a simplification. As we noted already, an agent is to blame for an action. The action is being judged too. In what way? In the way that we have just been exploring. To be blameworthy in respect of a given action, a responsible agent must (a) have failed to conform to a reason in performing that action, and (b) not have been justified in so failing, and (c) not have been excused in so failing. The question of the agent’s sensitivity to reasons arises, strictly speaking, only at stage (c). If by good fortune she conforms to undefeated reasons and also by good fortune acts for such reasons then her insensitivity to those or other reasons does not make her blameworthy. For in such a case there is nothing to blame her for. So judging someone to be blameworthy is not a simple case of judging someone negatively. Rather, it is judging that something (in the cases that concern us here, an action) bears negatively on how someone is to be judged. In a familiar idiom, we judge that the action ‘reflects badly’ on the agent. Do we now judge her disloyal, dishonest, unreliable, mean-spirited, or what? That remains to be seen. All that is clear is that ‘blameworthy’ is not a further entry on the same list. Blameworthiness is not a negative trait of the agent, but rather the condition under which one or more negative traits of the agent are manifested (exhibited, displayed) in her actions.

We return to the social media attention-seeker who says ‘don’t judge me’. Could the request be to judge the faux-pas if one must, but not to judge the maker of it in the light of it? It is hard to see how that request could be honoured. It is not the same as a request to be forgiving or merciful or lenient, which one honours by judging someone to be blameworthy but then declining to blame or reproach or punish (as much as one otherwise would). ‘Don’t judge me’ seems to be a request not to judge the maker of the faux-pas to be blameworthy. To honour that request one must grant, at least arguendo, that the faux-pas is justified or excused. As a rule, however, ‘don’t judge me’ is uttered without any suggestion of a justification or excuse for the faux-pas. That being so, how can we avoid the conclusion that the faux-pas reflects badly on its maker? We may be reminded here of the Augustinian advice ‘hate the sin, not the sinner’.[32] Fully generalized as a precept for life, I tend to regard that advice as an invitation to hypocrisy, but it can at least be implemented by the cultivation of a certain mindset. Suppose, however, that it were replaced with ‘judge the sin, not the sinner’. It is not clear how that advice could be implemented, leaving aside cases in which there is a valid justification or excuse for the sin. And in such cases the sin does not reflect badly on the sinner anyway, so the advice is otiose. It becomes: ‘when the sinner cannot be judged by the sin, judge the sin, not the sinner (obviously).’

4. Blame and character

This necessarily abbreviated discussion may well make you think of what Lacey at one time called ‘the character conception of responsibility’[33] but which, in the light of her work with Pickard, might better be renamed ‘the character conception of blameworthiness.’ As I just explained, being blameworthy is not a negative trait of the agent, but it is the condition under which one or more negative traits of the agent are manifested (exhibited, displayed) in her actions. When we say that someone is ‘at fault’ in their actions we mean that in their actions they manifest one or more of their faults (a.k.a. vices, flaws, deficiencies, shortcomings). Blameworthy people, we might then summarize, are people of pro tanto bad character.

It may strike you that this summary is too crude. To be blameworthy one must manifest one’s fault; but surely one has the fault irrespective of whether one manifests it? Bad character, unlike blameworthiness, is a matter of sheer insensitivity to reasons; nonconformity does not come into it. But that is a mistaken view of character. Character traits exist only inasmuch as they are manifested. The action partly constitutes the trait.[34] Someone who so far never had the opportunity to act (dis)honestly is not yet (dis)honest, never mind how they might have acted had the opportunity for (dis)honesty presented itself. That is how a court is able to say that a criminal defendant, just found guilty of an offence, is of ‘previous good character’. The picture is not that there was a hidden vice lurking there all the time, one that we only now got to see. The picture, which is indeed the right picture, is that until it was manifested, there was no vice: until one was insensitive to reasons in failing to conform to them, one’s character was still unblemished. (It must be recalled here that we have artificially restricted ourselves throughout to thinking about blameworthy actions, which manifest practical faults. One’s epistemic faults are manifested in one’s beliefs, one’s intellectual faults are manifested in one’s reasoning, and so on. Still, each exists only inasmuch as it is manifested.)

This ‘no character without manifestation’ thesis is the flip side of the thesis, already advocated, that even a momentary nonconformity with reason, if unjustified and unexcused, counts against an agent’s character. When someone says of such a lapse that it was ‘out of character’ we should not hear them to say that the lapse somehow does not count against the agent’s character. On the contrary: the lapse shows what the agent is capable of, and her character is, in the relevant sense, a set of capacities. One may be interested for other reasons, of course, in whether the action was ‘out of character’. If one is considering one’s future relationship with the agent, for example, one might be tempted to forgive what one holds to be an unjustified and unexcused aberration. That can be true even with quite serious lapses. One infidelity might be forgivable with a bit of effort; two or three, starting to form a pattern, might be impossible to forgive. Forgiving, however, is not excusing. Excusing negates the judgment of blameworthiness, whereas forgiving presupposes the judgment of blameworthiness.[35] The forgiven action reflected badly on the agent, to be sure, but one manages not to hold it against him. Or put in other words: his character did not come out smelling of roses but one can still reconcile oneself, in one’s attitude towards him, to what one discovered about his flaws.

Lacey’s longstanding worries about the institutionalization in the law of the character conception stem principally from the evidence, which she has done so much to document, that thinking in terms of character can lead the law astray on these very points. Too easily, she thinks, the law may end up

‘exhibit[ing] what we might call ‘character essentialism’ and ‘character determinism’. In other words, it proceeds from a view of human character ... and of identity as fixed, or at least as relatively stable; and it regards character as determining conduct.[36]

Lacey’s ‘character essentialism’ is the mistaken thesis that character is necessarily temporally extended. It follows that there is no room for one-off aberrant actions that reflect badly upon my character, and so no possible case for showing forgiveness or mercy towards me in respect of blameworthy actions on the ground that are ‘not like me’. That is because, for the character essentialist, there can be no such actions. Meanwhile, Lacey’s ‘character determinism’ is a behaviourist exaggeration of the ‘no character without manifestation’ thesis. We might call it the ‘character as sheer manifestation’ thesis. It treats nonconformity with reasons as sufficient, not merely necessary, for bad character.[37] Hence it gives no quarter to excuses (and possibly not even to justifications). Sometimes Lacey gives the impression that the ‘character conception’ of culpability is at odds with what she calls the ‘capacity conception’.[38] But once we rid ourselves of the twin errors of character determinism and character essentialism, it is doubtful whether this is true. The capacity conception seems to be just one attempt – personally, I am not enamoured of it[39] – to explain the difference between an action that exhibits bad character and one that does not. It attempts to do so by attempting to explain the difference between an excused action and an unexcused action in terms of the capacities and opportunities of the agent to avoid performing each.

Lacey tends to worry most about the potential of the character conception, in comparison with the capacity conception, to spread blame too widely and too harshly. But in one way the character conception is too restrictive. Our actions may reflect badly on us by manifesting our want of skill as well as our bad character. As Aristotle explains:

[I]t is from the same causes and by the same means that every virtue is both produced and destroyed, and similarly every art; for it is from playing the lyre that both good and bad lyre-players are produced. And the corresponding statement is true of builders and of all the rest; men will be good or bad builders as a result of building well or badly. For if this were not so, there would have been no need of a teacher, but all men would have been born good or bad at their craft. This, then, is the case with the virtues also; by doing the acts that we do in our transactions with other men we become just or unjust, and by doing the acts that we do in the presence of danger, and being habituated to feel fear or confidence, we become brave or cowardly.[40]

Here we see Aristotle articulate the ‘no character without manifestation’ thesis, as well as the thesis that actions may manifest traits of character without those traits being settled or stabilized in the person who manifests them. The same is true, as he notes, of skills and their absence. We are interested here in the negative cases, the cases of actions that went awry. For those are the cases in which the question of blameworthiness arises. Those, to put it another way, are the cases in which the question arises of whether the action reflected badly on the agent, such as to count as a manifestation of a fault, now understood as extending beyond her character flaws to include her technical failings, such as being a bad builder or a bad lyre-player.

5. Saving blame from dispensability

We judge human actions, or more generally the lives to which those actions belong. We also judge human agents, who are the people leading those lives. The two classes of judgments are not interchangeable. That you are leading a bad life (or acting badly) does not automatically make you a bad person (even pro tanto). It does so only inasmuch as the way you are leading your life reflects badly upon you. As soon as the question arises of whether the way you are leading your life (or anything else) reflects badly upon you, we are unavoidably concerned with blameworthiness. It follows that if people are to be among the possible objects of judgment, the dispensability objection to blame fails. You might think that people could be among the possible objects of judgment without judging that anything they do (or think or feel etc.) reflects badly upon them. But inasmuch as judging people means judging them in respect of their character or skill, that is false. For character and skill do not exist without actions (etc.) that manifest them, and actions manifest character or skill only if they reflect well or badly upon their agent. Accordingly we have no escape from judgments of blameworthiness, short of a comprehensive refusal to judge people as being of bad character or deficient skill. Such a refusal would clearly be a tall order. Would it even be possible? Can human beings implement ‘don’t judge me’ as an across-the-board self-denying ordinance? Not for as long as they seek relationships with people who are worth relating to, and try to be worthy of such relationships themselves. Not for as long as they engage with history, literature, and politics. Arguably, not for as long as they think of other people (and themselves) as people. These are all things that human beings, on at least some occasions, have reason to do.

Understanding this much should make us (and certainly has made me, over the years) less profoundly blame-sceptical. Once we see that judgments of blameworthiness play an unavoidable role in the rational architecture of human life, it is a short step to rehabilitating blame. Blaming someone is taking the attitude towards her that befits her blameworthiness. It is losing faith in her, or losing admiration for her, or taking a dim view of her, or something like that. The move from judging someone blameworthy to blaming is subtle, but the possibility of forgiveness shows that it exists. The move from blame to blaming emotions is also short and subtle. The possibility of someone’s being diminished in one’s eyes, but only dispassionately, shows that it is there. The blaming emotions may be fitting but, as Lacey and Pickard point out, it does not follow that they are welcome. In some contexts and on some occasions they are better tamed. One of the reasons, emphasized by Lacey and Pickard, is that these emotions can motivate counterproductive or otherwise unjustified blaming actions, such as excessive reproaching or punishing. Such excessive actions, it seems to me, tend to give blame itself a bad name. The blame does not lie with blame, exactly, but with the ease with which people are driven to express blame in harsh and hostile ways. After all of these years I still cannot see the case for all this harshness and hostility. That is what is left of my blame-scepticism. That one feels like coming down hard on somebody is no reason to do it. Even if such a hostile emotion is fitting, the fittingness of the emotion only makes the reaction excusable, not justifiable. It follows that the rest of us, including politicians, prosecutors, journalists, and so on, should not support the expression of the emotion but rather manage it.[41]

The problem, in short, is not so much that we live in a ‘blame culture’ but that we live in a ‘call-out culture’ where accusation, reproach, censure, and punishment run wildly out of control and can no longer be managed back into proportion by mediating institutions such as the criminal courts.

Niki Lacey’s first book, State Punishment,[42] was written when I was lucky enough to be her undergraduate student in the mid-1980s. I was privy to some of the process. In this early work Niki was already grappling with many of the themes that I have traversed, all too cursorily, in the preceding pages. I was an enthralled and broadly sympathetic reader of State Punishment when it came out, and the book did more than any other to set the thematic direction of my own academic work for the following twenty years (even though Niki and I always disagreed about various questions of approach). As Niki once said to me, we tend to work on the things that we find most difficult in our own lives, and for me the relevant difficulty, laid bare in her work, was always the difficulty with blame: I was a blamer myself, and in particular a self-blamer, but I couldn’t see a good place for blame in the rational architecture of human life.

I struggled for those 20 years to tackle this worry, as Niki did, through critical engagement with the criminal law. I tried to find a rationale for all the blaming that we all do by starting from the way in which blame is institutionally expressed, captured, and managed, in the criminal justice system. I remained a blame-sceptic throughout, moved mostly by what I have been calling the dispensability objection, which gave me a kind of abolitionist predisposition in respect of criminal punishment. That was healthy: every time I found something in the criminal law that I thought I could make sense of, I was on my guard. Was I being duped? But I later came to think that my blame-scepticism was artificially nourished by my abolitionist predisposition, in a feedback loop that was by no means so healthy. Contrary to what Epictetus thought, blame itself is not the main problem, the main immaturity. Blame has an important place in the architecture of rational human life after all; just not where I had been looking for it. The main problem is with the expression of blame, and in particular with the Janus-faced way in which the criminal law must always seem to be endorsing harsh and hostile public blaming emotions, by giving expression to them, if it is going to be able to manage them. Up to a point, the law must cosy up with the mob if it is to enjoy any sustainable control over the mob’s excesses.[43] In the last few years this problem has been overtaken, to some extent, by technological developments. Now the mob can bypass the criminal justice system with virtual impunity by use of Twitter and other online instruments of populist reproach and punishment, via a kind of Orwellian ‘group hate’. Even in top form, the criminal justice system would be virtually powerless to manage any of this and, like other mediating institutions of the state, it is anyway limping along in a ramshackle condition owing to the hyperdependency of the state on the robber barons of high finance. A book called State Punishment could still usefully be written. But now the title would surely carry a more ironic and despairing ring.


[1] Epictetus, Enchiridion, ch 5.

[2] ‘The Standard Argument for Blame Incompatibilism’, Noûs 42 (2008), 697.

[3] The best diagnosis of the slippage between ideas that draws people into this philosophical fantasy is probably Harry Frankfurt, ‘Alternate Possibilities and Moral Responsibility’, Journal of Philosophy 66 (1969), 829.

[4] ‘Internal Reasons and the Obscurity of Blame’, in Williams, Making Sense of Humanity (Cambridge 1995), 41.

[5] As G.A. Cohen says, ‘the question, “Who can say what to whom?,” goes largely unexplored in contemporary moral philosophy.’ Cohen, ‘Casting the First Stone: Who Can, and Who Can’t, Condemn the Terrorists?’, in his posthumously published collection Finding Oneself in the Other (Princeton: PUP 2013) 115 at 119. This paper by Cohen, and his follow-up ‘Ways of Silencing Critics’ in the same volume, are welcome exceptions. A more recent exception that repays study is Macalester Bell, ‘The Standing to Blame: A Critique’ in D. Justin Coates and Neal A. Tognazzini (eds), Blame: Its Nature and Norms (Oxford: OUP 2013).

[6] Lacey and Pickard, ‘From the Consulting Room to the Court Room? Taking the Clinical Model of Responsibility Without Blame into the Legal Realm’, Oxford Journal of Legal Studies 33 (2013), 1.

[7] See George Sher, In Praise of Blame (Oxford: OUP 2006), 88-9.

[8] Compare R. Jay Wallace, Responsibility and the Moral Sentiments (Cambridge, Mass.: HUP 1994), 74. Wallace seems to run together the question of whether certain emotions are part-constituted by blame with the question of whether it would be odd to blame without experiencing at least one of them.

[9] Classic discussions of the question include Justin D’Arms and Daniel Jacobson, ‘The Moralistic Fallacy: On the “Appropriateness” of Emotions’, Philosophy and Phenomenological Research 61 (2000), 65 and Wlodek Rabinowicz and Toni Rønnow-Rasmussen, ‘The Strike of the Demon: On Fitting Pro-Attitudes and Value’, Ethics 114 (2004), 391.

[10] For some more specific suggestions on how to contain a disposition to blame, see Lacey and Pickard, above note 6, 22-4.

[11] The same point is made with respect to beliefs in David Owens, ‘Does Belief Have an Aim?’, Philosophical Studies 115 (2003), 283.

[12] They may also support punishment-scepticism if, as I suspect, one cannot punish without reproaching. As it happens, Lacey and Pickard do not agree that one cannot punish without reproaching. They aim to ‘challenge[ ] the strong association between punishment and blame’ (Lacey and Pickard, above note 6, 2). So in a surprising twist they aim to mount a rescue, not a critique, of punishment. Here I will not focus on this aspect of their thinking.

[13] ibid, 3.

[14] Here is one piece of evidence that suggests that I am. In the second major instalment of their joint work on the subject, Lacey and Pickard contrast blaming with forgiving, and speak up for forgiving: ‘To Blame or to Forgive? Reconciling Punishment and Forgiveness in Criminal Justice’, Oxford Journal of Legal Studies 35 (2015), 665. But forgiving, for Lacey and Pickard, is curtailing or forestalling blame. In a world in which blame is dispensed with, in the way that I have in mind, there is no role for forgiveness so understood. Both belong to the same redundant apparatus. Possibly, however, forgiving is not quite so closely connected with blaming as Lacey and Pickard suggest, and so could survive the end of blame. See Julia Driver, ‘Wronging, Blame, and Forgiveness’, Oxford Studies in Agency and Responsibility 4 (2017), 206.

[15] e.g. Lacey and Pickard, above note 6, 18: ‘When a person is responsible for harm and has no excuse, they are blameworthy’ (emphasis added).

[16] A ‘naturalistic’ treatment of the subject that makes sense of this question – at the price of a certain measure of critical abstention – is Victoria McGeer’s ‘Civilizing Blame’ in the Coates and Tognazzini volume, above note 5.

[17] This challenge does not depend on, does not support, and does not resemble Williams’ judgment-sceptical challenge to blame outlined above (see text at note 4). In the passage quoted here Williams is not being even slightly judgment-sceptical.

[18] Williams, ‘Moral Luck’, in Williams, Moral Luck: Philosophical Papers 1973-1980 (Cambridge 1981), 20 at 28.

[19] Ibid, 39.

[20] This is the main subject of my book From Personal Life to Private Law (Oxford: OUP 2018).

[21] Sir Frederick Pollock, The Law of Torts (11th edn, London: Stevens 1920), 455. Pollock is talking about negligence, not blameworthiness.

[22] For some this happens only when the reasons have equal force; for others, it also extends to cases of incommensurability. For a good overview of the debate, see Timothy Macklem, ‘Choice and Value’, Legal Theory 7 (2001), 1.


[23] This way of explaining what is going on is owed to Joseph Raz’s famous analysis in Practical Reason and Norms (London: Hutchinson 1975), 35-48.

[24] (1850) 4 Cox CC 358.

[25] The most exacting discussion of these cases is Jonathan Glover, ‘It Makes No Difference Whether or Not I Do It’, Proceedings of the Aristotelian Society, Supplementary Volume 49 (1975), 171.

[26] For further explanation see my Offences and Defences (OUP: Oxford 2007), chs 3 and 5.

[27] This proposition requires some interpretation to apply it to cases in which people act for multiple independent reasons. These need not detain us here.

[28] Joseph Raz, Practical Reason and Norms (2nd ed. Princeton: PUP 1990), 178-82.

[29] For further explanation see Gardner, above note 26, ch 6.

[30] I am not sure that Nicola Lacey upholds the difference when she writes, in her State Punishment: Political Principles and Community Values (London: Routledge 1988), at 71, that the criminal justice system should ‘respond punitively only to actions which are in a real sense [the agent’s] own.’

[31] A leading version of this view equates blameworthiness to ‘subjective wrongness’, following Derek Parfit’s proposal in his Reasons and Persons (Oxford: OUP 1984), at 25.

[32] St Augustine, Letters 211-270 (trans Teske, ed Ramsay, New York: New City Press 2005) at 25 (‘with love for the persons and a hatred for their vices’).

[33] Lacey, above note 30, e.g. at 73.

[34] Compare Michael S. Moore, ‘Choice, Character, and Excuse’, Social Philosophy and Policy 7 (1990), 29 at 47. Moore’s objection to the constitutive view depends on the ‘partly’ being implicitly replaced with ‘wholly’ to yield what he calls ‘the behavioural view’, which is obviously false.

[35] A valuable discussion of the point is Lucy Allais, ‘Wiping the Slate Clean: The Heart of Forgiveness’, Philosophy and Public Affairs 36 (2008), 33.

[36] Lacey, In Search of Criminal Responsibility: Ideas, Interests, and Institutions (Oxford: OUP 2016), 35.

[37] Moore collapses these two theses in the passage cited above note 34.

[38] e.g. Lacey, above note 36, at 59.

[39] See Gardner, above note 26, ch 6.

[40] Nicomachean Ethics 1103b7ff.

[41] For wider application of this point see David Enoch, ‘The Masses and the Elites: Political Philosophy for the Age of Brexit, Trump and Netanyahu’, Jurisprudence 8 (2107), 1.

[42] Above note 30.

[43] See the reflections in Gardner, above note 26, ch 11.