Elizabeth Harman (2016) – Morally Permissible Moral Mistakes

By Aart van Gils

Today, we discussed Elizabeth Harman’s (2016) “Morally Permissible Moral Mistakes”. Euan had proposed this paper to us, and he was so kind to present the key points of the paper and lead the subsequent discussion. There were several interesting points made in the discussion, but for the sake of clarity and brevity only the most significant will be mentioned.

The first thing we noted from Harman’s rather dense paper was that it is not entirely clear what the background assumptions are for trying to argue for this new category of “morally permissible moral mistakes”. Fortunately, Euan’s expertise came to our rescue here. Put briefly, this new category is of particular interest to advocates of the view that ‘ought’ can be explained in terms of reasons. This is because one of the fundamental problems advocates of that view face is that they seem unable to account for supererogatory actions. To show this, we can consider a general form of one of the many examples Harman uses.

Suppose that you promised your friend to pick him up from the airport, because he is a nervous flyer and seeing a familiar, friendly face would do him a lot of good. Unfortunately, on your way to the airport your car breaks down and you can either pay a very expensive bill to have it ‘instantly’ repaired, or decide to have it repaired later, which will save you some money but will also make it the case that you will not be able to pick up your friend from the airport and so you fail to keep your promise. Harman wants to say that failing to pick up your friend is a “morally permissible moral mistake”.

The explanation for this is that the original moral requirement that stems from your promise to pick up your friend is nullified by the unexpected great burden that you now face – the hefty repairs bill – if you keep your promise to pick him up. Nevertheless, the keeping of the promise remains present as a moral reason upon which you can act. Therefore, we have a balance of reasons between at least the moral reason to still pick up your friend and the conflicting prudential reason to not pay the extra expenses of having your car repaired on the spot.

Harman then relies on the separation of the idea of moral requirement from the all-things-considered ought. She would argue that the overall balance of reasons in this scenario, falls on the side of the moral reason to still pick up your friend. This means that all things considered, you ought to pick up your friend. However, given that there is no moral requirement involved, this means that we can still characterize this act as supererogatory. Crucially, failing to act upon this moral reason or choosing a different action than the one that the overall balance of reasons points to, is a morally permissible moral mistake. The result would be that those who advocate the view that ‘ought’ can be defined in terms of reasons, are in fact able to account for the supererogatory.

Criticisms were raised, however, that Harman heavily relies on rather technical understandings of ‘mistake’ and ‘moral mistake’, which seem to depart quite a bit from the common-sense understanding of when someone makes a ‘mistake’. Furthermore, with respect to the example mentioned here, she argues that when the agent decides to still pick their friend up by saying to themselves ‘I should do it!’ that the agent is “right”. But to say that the agent is “right” seems to suggest that we are no longer speaking about a supererogatory (i.e. merely permissible) act, but rather something that she is required to do.

A final interesting point that Harman makes, and which was unanimously accepted is that those who accept the possibility of performing supererogatory actions, are also committed to the possibility of suberogatory actions – actions which are bad but not wrong to do.

All in all, Harman’s “Morally Permissible Moral Mistakes” was a rich and interesting read which prompted both questions about metaethical categories, as well as the plausibility of a hotly-debated view on reasons and oughts. We thank everyone who was there for contributing to a very fruitful discussion, and we thank Euan in particular for bringing this interesting article to our attention.

Robert E. Goodin (1989) – Theories of Compensation

By Aart van Gils

Today, we discussed Robert E. Goodin’s (1989) “Theories of Compensation”. The paper was chosen by Josh, who was so kind to introduce the paper and its key points to us. There were several interesting points made in the discussion, but for the sake of clarity and brevity only the most significant will be mentioned.

First, Goodin distinguishes between two types of compensation, namely “means-replacing-compensation” (compensation1) and “ends-displacing-compensation” (compensation2), and he argues for the superiority of compensation1 over compensation2. Some worries were raised regarding this supposed superiority of compensation1. To name one in particular, there was the worry that the value of certain ends seems to derive in great part from the particular means that someone may use in order to achieve his or her end. Therefore, solely compensating people such that they would be able to achieve the same end, without considering whether the particular compensatory means provided were similar enough to the original means, seemed problematic.

In Goodin’s defence, however, it was pointed out that this objection relied on blurring the distinction between means and ends, and that if certain means were considered as necessary and/or invaluable, it seemed hard not to slot them under the heading of ‘ends’, which would require compensation. But Goodin could accept this point by broadening compensation2 and claim that this is how we compensate for the necessary and/or invaluable aims.

Second, Goodin seemed to waver between two different understandings of a Pareto improvement in an interpersonal distribution. Put simply, it was unclear whether Goodin’s view strictly required that people were in fact compensated for their losses or whether ‘virtual’ compensation (i.e. that the total gains would theoretically outweigh the total losses) would be sufficient.

It was acknowledged that Goodin’s view could be improved by a stricter definition of Pareto improvement. However, it was eventually agreed that this point was not part of the main argument and so, if necessary, Goodin could easily concede this point all the way, if pressed on it.

Third and final, there was disagreement regarding the strength of Goodin’s conclusion. The conclusion seemed to be that even if individuals were to have legitimate complaints for not being (sufficiently/appropriately) compensated, then it might be that the policy could nevertheless be adopted. The worry here was that a second level of compensation was introduced, meaning that besides the question of how people might be compensated for their losses, there was the option that the impersonal gains of a particular policy could outweigh the losses done to several individuals.

There was unanimous agreement that Goodin had more work to do here and that his conclusion moved too quickly from one step to the other. But again, in his defence, it was pointed out that actually Goodin had made the correct move by arguing for the permissibility of public policies, even if these policies would have the effect that some people would suffer losses which could not be compensated. To set compensation as a requirement on public policy, disallowing it when compensation was impossible or insufficient, would lead to an implausibly restrictive and/or conservative attitude in political decision-making.

All in all, Goodin’s “Theories of Compensation” raised several interesting issues, ranging from the theoretical distinction between different types of compensation to the practical application to several concrete cases. We thank everyone who was there for contributing to a very fruitful discussion, and we thank Josh in particular for bringing this interesting article to our attention.