Article Summary: “The Supervenience of the Ethical on the Descriptive” by Frank Jackson


Jackson argues that ethical properties are merely descriptive properties – there are no properties above and beyond the descriptive. Jackson develops his thesis in light of the a priori truth that ethical properties supervene, or depend upon, descriptive properties. Then, Jackson handles two objections to his view.

Jackson begins by clarifying what ethical supervenience means, as opposed to psychological supervenience. In the case of psychological supervenience, physicalism (a metaphysical doctrine) holds that a full account of the physical world will entail a full account of the psychological world. The psychological depends upon the physical. But this is an intraworld, contingent supervenience thesis, rather than a global one. Another way of putting the point is that a full account of the psychological world does not entail a full account of the physical world: because the psychological is multiply realizable, there could be many different physical accounts which entail the psychological world. So the fact that the physical entails the mental in the way it does in our world is a contingent reality; it could have been otherwise.

Ethical supervenience, meanwhile, is a global supervenience thesis: it holds across all possible worlds. The descriptive entails the ethical in virtue of the meaning of these words; it cannot be otherwise. Suppose, for example, that two situations are exactly alike descriptively. Is it possible for us to condemn an actor in one situation and not the other? It seems not. The morality of the situation is determined by what occurs descriptively: an actor’s motivations, what the actor did, the outcome of the act, etc. It cannot be otherwise.

Now Jackson presents his argument: for any sentence ‘x’ about ethical nature, that sentence will be fully entailed by the world’s descriptive nature (this is a reiteration of the global supervenience thesis described above). Since the ethical depends upon the descriptive, it can be expressed fully in descriptive terms. So, any sentence ‘x’ about ethical nature ultimately reduces to merely descriptive terms, and ethical properties are merely descriptive properties.

Jackson then aims at clarifying his thesis: he says, for example, that it does not necessitate that an account of the ethical world is symmetrical with an account of the descriptive world. This is because a full account of the ethical world is consistent with infinitely many accounts of the descriptive world: one which includes electron E, for example, and another that excludes it.

Further, Jackson insists that his thesis does not imply that it is acceptable to do away with ethical vocabulary. If a given ethical property is an infinitely large descriptive disjunction, for example, the only way to express it will be to use an ethical term as shorthand. So too with ‘baldness’: baldness is really just referring to an infinitely large disjunction of possible hair distributions, but we point to a paradigmatic example of a bald person and use the word as shorthand out of necessity. But this does not mean ‘baldness’ is some extra feature of the world; so too, ethical terms, despite their necessity, do not imply extra features of the world.

Jackson then deals with two variations of an objection to the idea that ethical properties are merely descriptive properties. The general objection is that logically equivalent predicates may nonetheless point out distinct properties. In the case of ethical predicates, while they may be logically equivalent to descriptive sentences and necessarily co-extensive with them, it is still possible that they are pointing to distinct properties besides merely descriptive ones. The specific example in the first variation of the objection is the case of triangles: there seems to be a property of equiangularity and a property of equilaterality, despite their always occurring alongside one another.

Jackson thinks that this case conflates how we single out properties from how we distinguish how many properties there are. While it is true that equiangularity and equilaterality are distinct, this is so because they are ways in which we single out a property. The property itself, understood as a feature of the world, is singular in nature. So, we are singling out one property in two different ways.

Jackson also attempts to diffuse a more elaborate variation of the above objection: suppose we construct a machine that can detect the equilaterality of a triangle, but not the equiangularity. The machine has a light which blinks when this feature of the triangle is detected. This would seems to demonstrate that these two features of a triangle are distinct properties, since one is causally efficacious (it causes the machine’s light to shine) while the other is not. Jackson thinks that the force of this example derives from the machine’s operating in a segmented fashion: it first measures the sides of the triangle, then blinks if they are equal. It never gets a chance to measure its angles. In any case, Jackson thinks the example pertains merely to whether angles and sides are equivalent, not whether being an equilateral triangle and being an equiangular triangle are distinct properties or not.

Jackson finishes his rebuttal by deploying Mackie’s ‘argument from queerness’ (see my summary of Mackie on the ‘metaethics’ page). He does not understand what an ethical property could possibly be if not a descriptive one: what could a person mean when they say that something is ‘right’ if not the merely descriptive facts of the thing being picked out? What else could possibly be picked out? To Jackson, it is better to suppose such non-natural properties do not exist than to suppose they exist in some strange, unknown way.

Jackson concludes by stating one important implication of his argument: all  moral realists must express ethical properties in descriptive terms. Since his thesis is modal and not metaphysical, it does not claim such properties exist. It merely says that if indeed they do exist (or if one is a moral realist), then they must be expressed as descriptive, natural features of the world, regardless of one’s specific metaphysical allegiances.

Logical Outline

Primary Argument: Ethical properties = Descriptive properties

1. Ethical nature depends upon descriptive nature (global supervenience thesis).
2. Descriptive nature can be expressed in full as a sentence with merely
descriptive terms.
3. Therefore, ethical nature can be expressed in full as a sentence with
merely descriptive terms. [1,2]
4. Therefore, when one expresses something about ethical nature, one is
merely expressing something about descriptive nature. [from 3]
5. Therefore, ethical properties are merely descriptive. [from 4]


Leave a comment

Filed under Article Summaries

Leave a Reply

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

You are commenting using your 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 )

Google+ photo

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

Connecting to %s