What impact do cases of moral ignorance have on the notion of moral responsibility?

By: James Regan, 4th year undergraduate 2015, University of Edinburgh


I will argue that cases of moral ignorance reveal that the notion of moral responsibility is inadequate. These inadequacies are highlighted by an analysis of both Gary Watson’s and Elinor Mason’s conceptions of moral responsibility. I recommend the concept be further revised.

Some cases of putative ignorance result in a feeling of ambivalence, suggesting that it is inappropriate to either fully ascribe responsibility to agents or fully exculpate them.[1] Consider Robert Harris. Harris was a vicious and brutal murderer that displayed complete disregard and contempt for others. On the one hand, we feel antipathy towards Harris due to knowledge of his malevolence; this encourages us to ascribe responsibility. Yet this may be mitigated, if only partially, because of the sympathy aroused when considering Harris’ formative circumstances – his childhood was marred by undeniably cruel and gross mistreatment (Watson, 2004, pp.235-251). Thus, we also have reason to withhold responsibility. However, there is a problem with exonerating agents like Harris because we no longer capture the sense that such agents are, at least to some extent, still responsible. As Mason puts it, “They seem blameworthy because they seem bad [my emphasis]” (forthcoming, p.10). As a result of this ambivalence, a nuanced account of moral responsibility is required. Extreme positions on moral ignorance such as Matthew Talbert’s (2012) and Gideon Rosen’s (2003) are unable to facilitate this ambivalence because they fully assign and withhold responsibility respectively. I will demonstrate that attempts to facilitate opposing sentiments, as in the case above, reveal weaknesses in the underlying notion of moral responsibility.

Different Conceptions of Moral Responsibility

I begin by illustrating why Watson’s account of responsibility is weakened by a case of ignorance. Watson distinguishes two kinds of responsibility (2004, pp. 260-88): attributability and accountability.[2] Harris would be considered responsible in the attributable sense because of the moral faults exhibited by his ill attitude towards others. This is applicable regardless of additional considerations such as avoidability, history, and fairness (Watson, 2004, pp. 273-4). However, if ignorance is exculpatory then Harris would not be held accountable. This is because accountability is sensitive to additional considerations. Ignorance must therefore be reflected in this face of responsibility. By withholding accountability, an ignorant agent is no longer fully responsible, but Watson’s position does confer some degree of responsibility since the agent’s actions remain attributable to them. Hence, the strength of Watson’s position is that responsibility can vary in accordance with the aforementioned ambivalence. Nevertheless, I contest that this conception of responsibility is altogether inadequate because it is incapable of handling the subtleties of more complex cases of ignorance. Consider Harris* whom, I stipulate, committed identical crimes to Harris and did so as a result of the very same ignorance. He is evil and morally ignorant, but there is a difference in that Harris* was not subjected to an abusive childhood. Accordingly, there is a sense in which Harris* is somehow more responsible than Harris because his ignorance is less responsibility-undermining. Watson’s account cannot explain this. Since they are both acting from ignorance they are both absolved of accountability. Thus, only attributability remains appropriate. However, attributability cannot differentiate these two agents because it is rigid in its application; a moral fault either is or is not (in theory) identifiable in an agent so they are both responsible in the attributable sense or not. This cannot vary by degrees because it is “stable in the face of considerations of history” (Mason, forthcoming, p.17). In Mason’s words, this position cannot account for “the complexity of our hard cases (…) and the different reasons that we might want to grant some clemency” (forthcoming, p.11). I now look to Mason’s conception of moral responsibility. If her account can make sense of cases like this one then ignorance may not have the same damaging impact on responsibility as it first appears to.

Like Watson, Mason divides responsibility broadly in two. Ordinary blame isn’t applicable to Harris. This is because Harris’ ignorance indicates a lack of the capacities required to be a member of the moral community whose constituents are subject to this type of blame (Mason, forthcoming, p.12).[3] Since ordinary blame is inappropriate, Mason retreats to objective blame instead. Objective blame focuses on the character of an agent, apt when an agent is a “bad-act maker” (Mason, forthcoming, p.13). To the limited extent that this is similar to Watson’s conception, ordinary blame is more akin to accountability, and objective blame more akin to attributability. However, there is a key difference – illustrated by the handling of the Harris/Harris* case – that enables Mason to distinguish the Harris’ in a way that Watson was unable to. Whereas attributability is rigid, objective blame is adjustable on the basis of ‘damage’ (forthcoming, pp. 17-18). Damage, caused by non-natural factors such as a harmful upbringing, forces us to view individuals as less of an agent and more of a subject. This position is the truly objective stance, which represents the minimal ascription of moral responsibility possible. Mason would retreat to this stance with Harris but not with Harris* because he is not emotionally damaged in the way Harris is. Hence, Mason is able to differentiate these two agents by appropriately adjusting objective blame according to damage. Therefore, this conception of moral responsibility is not rendered inadequate by the particular case of ignorance I am considering. An additional benefit of Mason’s account is it can explain how Harris is partially responsible – albeit minimally – using the truly objective stance. Watson, on the other hand, left no room for any degree of responsibility if attributability was undermined. However, I will now question a key assumption that facilitates Mason’s ability to handle this case; if the assumption proves false – as I will argue it does – this conception may also be inadequate.

A Critique of Mason’s Account

From the outset I have permitted the assumption that Harris is morally ignorant. Mason also makes this assumption, suggesting that “he probably cannot grasp what morality requires” (forthcoming, p.17). Presumably this is because Harris appears to be oblivious to “the rational force of moral considerations” (Rosen, 2004, p.305). With this assumption in place, I used the Harris case to expose weaknesses in Watson’s conception of responsibility and shed favourable light on Mason’s account because of its ability to handle the case. However, I will now argue that the assumption is false. There is evidence suggesting that although Harris is “morally defective in some important way” (Mason, forthcoming, p.1), that defect is not indicative of ignorance. His actions – malignantly opposed to moral considerations – are performed because he doesn’t care about those considerations, not in ignorance of them. As Watson says, “he exhibits an inversion of moral concern, not a lack of understanding” (2004, p.239). This is well illustrated by comparing the Harris case with a more obvious case of ignorance. Whereas slaveholders were genuinely unaware that their behaviour was wrong, Harris is aware of the immorality of his actions but chooses to dismiss this due to a lack of compassion. Hence, there is reason to believe that Harris is not ignorant. If the assumption is false, Mason cannot differentiate Harris and Harris*, as I will now explain. For Mason, excusing conditions such as ignorance undermine ordinary blame. According to my previous analysis, it is not clear that Harris is ignorant. Since it was stipulated that Harris* is identical to Harris in all respects except for his formative circumstances, it is not clear that Harris* is ignorant either. Accordingly, neither Harris nor Harris* would be excused from ordinary blame, so Mason could not retreat to objective blame. The problem is that the subsequent adjustment of objective blame, (varied with damage), allowed Mason to differentiate the two Harris’. Now that Mason cannot retreat to this adjustable form of blame, both agents are blameworthy in the ordinary sense which does not have the flexibility to accommodate the apparent difference between them. This calls into question Mason’s conception of responsibility because it was this very ability that previously afforded such credence to her position. Whilst this case is no longer one of ignorance, the fact that Mason’s account of responsibility cannot explain it still reveals inadequacy. I will now evaluate some counter-arguments to this objection, but I ultimately find that Mason’s position is severely weakened.

It may be contested that Mason isn’t vulnerable to this objection because she can explain the Harris/Harris* case (using damage and objective blame) without presuming their ignorance. Arguably she shifts the locus of explanation away from ignorance towards the notion of being “relevantly like us”, (Mason, forthcoming, p.13).[4] If Mason can prove both that this latter concept can be used to undermine ordinary blame instead of ignorance, and that Harris is not relevantly like us, then her account would be unaffected by my criticism. I take no issue with the first point; however, I do contest the second which suggests that Harris is not relevantly like us. I argue that in order to support this Mason is forced to apply a certain requirement that must be met for someone to qualify as relevantly like us. The requirement must be that an agent is only relevantly like us when their practical application of moral considerations is the same as ours.

The reason Mason is forced into using this requirement is because the other obvious option – that Harris does not have an understanding of morality – is precluded by her analysis, as evidenced above. Harris does have an understanding of morality, so this cannot be used as the requirement or else Harris would qualify as relevantly like us. Of the remaining differences that might be used to distinguish Harris, thereby deeming him not relevantly like us and undermining ordinary blame, the practical application of moral considerations is the most obvious.[5] However, I argue that this requirement is not only inappropriate, but also raises a number of questions, the answers of which are quite unclear. Must agents always act similarly in order to be relevantly like one another? How similar must practical application be for agents to be relevantly like one another? The absurdity of this requirement can be illustrated with a simple thought-experiment: Bob – a sane and morally-conscientious individual – has an identical grip on morality as another like-minded individual, Barbara. Their moral considerations are identical and they both act on those considerations in morally acceptable, though different, ways. Does this mean they are no longer relevantly like one another? If it does – which the requirement implies – then the population of ‘us’ that are relevantly like one another according to the standards Mason is forced to set is severely diminished. This is because not being relevantly like one another would apply to anyone acting on the same considerations but in different ways. This consequence represents how inappropriate the requirement is. Though agents may act differently, they can remain relevantly like each other, at least on my understanding. My understanding is that what is relevant is the capacity for rational deliberation and a proper understanding of moral considerations. Although Harris may be sadistic, I still believe he is relevantly like us because he can understand morality in an important way. This can be illustrated by contrasting him with JoJo who is clearly not relevantly like us (Wolf, 2003, p.379). JoJo cannot understand why killing inanely is wrong, but Harris is not like that. Harris is a different case; a case that I have shown to depict neither ignorance nor not being relevantly like us. Consequently, the success of Mason’s account relied on an assumption that is conceivably false. If Mason’s account is unable to handle these complex cases, I am left wondering whether or not an adequate conception of moral responsibility exists.

Final Considerations

Having discussed various cases, I now consider which notion of moral responsibility, if any, remains plausible. If Harris is no longer considered ignorant, there may be reason to reconsider the merits of Watson’s account. My dismissal of his account was based on its inability to differentiate the Harris’; however, with ignorance potentially dismissed, Harris and Harris* are distinguishable using the two faces of responsibility. On this interpretation, Harris remains unaccountable for his actions – not due to ignorance, but due to his unfortunate formative circumstances. In contrast, it is fair to hold Harris* accountable since his former circumstances were not at all unfortunate. Seemingly, it is only the presumption of ignorance that prevented Watson differentiating Harris from Harris*. However, I do not endorse Watson’s position as the most plausible. Even if Harris is not deemed ignorant, Watson remains vulnerable to genuine cases of ignorance. Furthermore, there are other concerns with his account (see Mason, forthcoming, pp.15-16). Whilst I do not consider these here, they are cause for enough concern to reconsider the concept of moral responsibility altogether.

I now consider Mason’s account. It must first be noted that interpretation of these cases is entirely subjective, rendering the implications for moral responsibility extremely contentious. For instance, there is evidence in favour of the assumption that Harris is ignorant (or, rather, not relevantly like us).[6] Proponents of this assumption would likely support Mason’s account, arguing that the notion of moral responsibility is perfectly adequate. However, even if Harris is not found to be ignorant, there are other concerns with Mason’s account. For example, we still consider Harris to be blameworthy in an important way, but the minimalistic notion of the truly objective stance arguably does not reflect that. In addition, there is another case that Mason cannot deal with. In this case, two agents are only objectively blameworthy because they are not relevantly like us, but neither is damaged. It may be stipulated that one agent takes greater pleasure in their sadism than the other. For this reason, we consider that individual more blameworthy, as we did with Harris*. To account for this, objective blame needs to be altered but damage is the only mechanism Mason possesses to do this. Since neither is damaged, objective blame cannot be varied and Mason is unable to distinguish between the two.


Mason’s conception of moral responsibility is a sophisticated one that offers greater explanatory power than the likes of Watson’s. However, as is endemic to many of these accounts, there are problematic cases which highlight weaknesses in that conception. I have acknowledged that how one interprets these cases has significant implications for any evaluation of the positions discussed, but I have endeavoured to highlight that on either interpretation of the Harris case – ignorant or otherwise – there are problems with our conception of moral responsibility.

The overwhelmingly apparent theme emerging from my discussion is that a variety of complex cases, including ones of moral ignorance, reveal inadequacies in the notion of moral responsibility. At the very least, this proves the need for future revision of the concept.



[1] In much of the literature, ‘blame’ is discussed instead of ‘responsibility’. Blame is relevant to my discussion because the preconditions for blame, i.e. being ‘blameworthy’, insinuate some sense of responsibility. For this reason, I will temporarily suspend differences in meaning and use blameworthiness and responsibility fairly interchangeably.

[2] Scanlon (1998 & 2008) offers something similar.

[3] As above, blameworthiness confers responsibility.

[4] ‘Us’ presumably represents some arbitrary majority that share a common morality, i.e. the ‘moral community’.

[5] The inversion of moral sentiments is another possibility, though I find this option troubling because it might also suggest a lack of moral understanding.

[6]Mason (forthcoming) endorses this; Watson (2004, p.242) describes Harris as incapacitated for ordinary interpersonal relationships; and Rosen says “an objectionable attitude towards others is often constituted by a form of moral ignorance” (2003, p.72). However, on balance, the evidence suggested to me Harris is not ignorant.



Fischer, J. M. and Ravizza, M. (1998) Responsibility and Control: A Theory of Moral Responsibility, Cambridge: Cambridge University Press. Chs. 1-3 & 9

Harman, E. (2011) ‘Does Moral Ignorance Exculpate?’ in Ratio [e-journal], December Vol. 24 (4), pp. 443–468. Available at: <https://www.ebscohost.com&gt; [accessed 7th April 2014 through University of Edinburgh library website]

Hieronymi, P. (2004) ‘The Force and Fairness of Blame’ in Philosophical Perspectives [e-journal], Vol. 18 (1), pp. 115-148. Available at: <https://www.ebscohost.com&gt; [accessed 5th March 2014 through University of Edinburgh library website]

Mason, E. (2005) ‘Moral Responsibility’ in Philosophical Books [e-journal], October Vol. 46 (4), pp. 343-353. Available at: <https://www.ebscohost.com&gt; [accessed 28th January 2014 through University of Edinburgh library website]

Mason, E. (forthcoming) ‘Responsibility for Moral Ignorance’ in Philosophical Studies. [Accessed 6th April 2014 via e-mail]

Rosen, G. (2003) ‘Culpability and Ignorance’ in Proceedings of the Aristotelian Society [e-journal], New Series, Vol. 103, pp. 61-84. Available at: <http://www.jstor.org/stable/4545386&gt; [accessed 11th March 2014]

Rosen, G. (2004) ‘Skepticism about Moral Responsibility’ in Philosophical Perspectives [e-journal], Vol. 18 (1), pp. 295–313. Available at: <https://www.ebscohost.com&gt; [accessed 7th April 2014 through University of Edinburgh library website]

Scanlon, T.M. (1998) ‘Responsibility’ in What We Owe to Each Other, Cambridge: Harvard University Press. Ch. 6, pp. 248-294

Scanlon, T.M. (2008) ‘Blame’ in Moral Dimensions: Permissibility, Meaning, Blame, Cambridge: Harvard University Press. Ch. 4, pp. 122-214

Smith, A. (2008) ‘Control, responsibility, and moral assessment’ in Philosophical Studies [e-journal], April Vol. 138 (3), pp. 367-392. Available at: <https://www.ebscohost.com&gt; [accessed 7th April 2014 through University of Edinburgh library website]

Talbert, M. (2012) ‘Moral Competence, Moral Blame, and Protest’ in Journal of Ethics [e-journal], Vol. 16 (1), pp. 89-109. Available at: <https://www.ebscohost.com&gt; [accessed 17th March 2014 through University of Edinburgh library website]

Watson, G. (2004) Agency and Answerability: Selected Essays, Oxford: Clarendon Press. Chs. 8-9, pp. 219-288

Watson, G. (2011) ‘The Trouble with Psychopaths’ in Wallace, R.J., Kumar, R., and Freeman, S., Reasons and Recognition: Essays on the Philosophy of T.M. Scanlon [e-resource], ch. 13, pp. 307-324. Available at: <http://www.oxfordscholarship.com/view/10.1093/acprof:oso/9780199753673.001.0001/acprof-9780199753673-chapter-13&gt; [Accessed 21st April 2014]


Wolf, S. (1993) ‘Problems with the Real Self View’ in Freedom Within Reason, Oxford: Oxford University Press. Ch. 2, pp. 35-45

Wolf, S. (2003) ‘Sanity and the Metaphysics of Responsibility’ in Watson, G. (2003) Free Will (2nd ed.), Oxford: Oxford University Press. Ch. 19, pp. 372-87

Share your thoughts on this!

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

WordPress.com Logo

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

Google+ photo

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

Twitter picture

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

Facebook photo

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

Connecting to %s