Responsibility

Eve Garrard has a good post at Normblog on the subject of blame. She considers, and rejects, the view that the individual agent is always entirely responsible for actions (e.g. terrorist bombings):

“…we don't have to buy into all the anodyne exculpations of some root-causes talk in order to accept that sometimes circumstances make a difference to the degree of responsibility a person carries for her misdeeds. We think that the person who has been hideously abused or oppressed, or appallingly impoverished or indoctrinated, doesn't carry the same weight of responsibility for at least some kinds of wrongdoing”.

She also rejects the view that we can blame the initiator of some chain of effects for all the bad consequences that result, whoever carries these out:

“If we view the initiators of a war as the only blameworthy agents in the context of that war, then not only do we deny responsibility to all the other participants (thus reducing them to moral puppets), we also place a limitless burden of responsibility on the initiators. Indeed this principle gives carte blanche to all the other parties to a military conflict to behave as atrociously as they please, since all the blame will be attributable to the initiators.”

But on top of this, she argues that a third view – under which responsibility can be shared among anyone who has (foreseeably) played a part in producing some negative outcome:

“A great many people will make some causal contribution or other to the atrocities, since causal chains ramify so fast; and there seems no reason to exclude any of them from sharing in the responsibility, once we allow that it can be shared by more than the direct agents, and especially once we allow that it can be shared by people (such as the initiators of the war) who may have neither wanted nor aimed at the horrific outcomes. We'll rapidly get to the stage where it's easy to say that we're all responsible for the horrors, which is of course tantamount to saying that no-one is really responsible.”

All of which sounds fair enough. So what do we do? She suggests that we might “distinguish between those who are primarily responsible, and those who bear some real, but lesser, secondary, responsibility”. Those directly and intentionally carrying out the action will generally bear primary responsibility, and those who have contributed in some way to its happening will be liable to secondary responsibility.

For want of an elegant segue: here are some of my thoughts.

One way of taking on board the insight that contextual factors matter, while maintaining that there is a unique type of responsibility for what one directly does, is to stipulate that we are discussing actions rather than mere physical behaviour. What I’m getting at here is that actions – things we do in specific situations for particular reasons – are inherently contextual. (One could define behaviour as simple bodily motion – so throwing a grenade would be the same as throwing a ball – but I’m using a little more expansive notion that covers the causally relevant immediate surroundings. This probably needs more fleshing out.)

Different intentions and different circumstances can make different actions out of the same basic behaviour. For instance, shooting someone dead could well be an act of revenge, of racial hatred, of accidental clumsiness, of mercenary greed, of self-defence, of incitement to riot, of unthinking panic… Different levels of culpability will attach to these different actions.

The fact that we are often inclined to heap less blame on the shoulders of people who are insane, traumatised, brainwashed or coerced in some way can be covered by this idea: if the action I’m responsible for is assaulting a tourist and smashing their camera while experiencing paranoid delusions, then that is far more forgivable than similar behaviour motivated by a (sane) personal dislike.

On this view, it’s not so much that I have diminished responsibility for my behaviour as that my action (which is defined to encompass my mental state) carries with it a diminished level of blameworthiness.

Blameworthiness is determined by intention and by consequences. Both are necessary (manslaughter isn’t as bad as murder; attempted murder is bad even if no harm results).

In terms of consequences, the most relevant factor is: which consequences could the agent, at the time of acting, have reasonably been able to predict as likely? This is another notion that need more finessing, but something like this formulation seems right. If my seemingly innocuous action produces an utterly shocking disaster, then the action may well be regretted – but if I couldn’t possibly have known, then I can’t really fairly be held responsible.

But if my well-intentioned rashness produces a disaster that I could have predicted with a little forethought, that’s different. Negligence may be less reprehensible than malice, but reprehensible it surely is.

The fact that ill intent adds to culpability is obvious. This doesn’t mean, though, that good intentions always protect innocence. The issue of predictable yet unwanted side-effects is complex (see this discussion at Tom H’s for a taster), but if I expect my action to result in some good things and one bad thing, then I can only take the credit for the good as much as the blame for the bad.

Tying this back in to my suggestion that we consider actions as inherently contextual, it’s clear that important parts of the context are the agent’s intentions in acting and ability to judge likely consequences.

Relating this to some of what Eve was saying, there are cases when some third party – with intentions and motives very different from my original ones – goes on to commit some atrocity within a context that I have created. It may be that the number of different people’s actions that lie between my own action and some outcome is a good measure of the difference between ‘primary’ and ‘secondary’ (and tertiary…) responsibility, which she uses in her analysis. But the mere presence of some such intervening decisions doesn’t in itself insulate me from some responsibility for the outcome.

Say a lax parole board takes an ill-considered decision to grant early release to a serially violent criminal who shows little sign of remorse or reform. He gets out, and attacks some more people. Obviously, he committed these acts, and he’s responsible for deciding to do them. But, equally obviously, the board members have some responsibility, as they gave him the opportunity to do so.

Or say a terrorist kills some civilians in Britain, accompanied by a video containing angry statements about Iraq and Afghanistan, among other things. It’s his fault; he chose to do it. But, accepting that his motives are what he says they are, if the invasions hadn’t happened, he would probably have been less inclined to set the bomb off. To the extent that such a consequence was foreseeable, the invaders are partly responsible for his increased likelihood of doing it, as is anyone else who has done anything that predictably contributed to getting him into the position he was in.

And the issue of predictability does come in here: the more different people’s decisions are in the chain leading to the outcome, the less scope there is for the original agent to predict the outcome in question. As such, the level of responsibility decreases.

So if I do something that can reasonably be expected to increase the motive and/or opportunity of someone else to do something bad, and then they do it, then yes, it’s entirely their fault for doing it, but I also have some of the responsibility for making it likelier to be done. Given the contextuality of actions, then the other person’s action has to be viewed within the context that I predictably contributed to.

One final thought: there may be scope for drawing a distinction between responsibility and blame. With blame, there is an automatic acknowledgement of personal wrongness. Not necessarily so with responsibility: for instance, I’m responsible for the fact that I’m currently sipping coffee rather than tea. I’m the deliberate author of this state of affairs, but there’s no moral evaluative aspect to my decision.

Maybe there are grounds for applying this distinction in cases where one person plays a role in making someone else predictably likelier to do something bad, but the first person (a) doesn’t want the bad outcome to happen, (b) has made their decision as the lesser of two evils, and (c) tries (if possible) to reduce the likelihood of this outcome. In this case, we might agree that there is some responsibility for the bad outcome but that it’s reasonable to exempt their action from blame, given the context.