DOI: 10.4324/9780415249126-N057-1
Version: v1,  Published online: 1998
Retrieved October 20, 2019, from

1. Varieties of supervenience

The basic instinct behind a supervenience claim is that although we talk in terms of two levels of fact, one of them is fundamental in that once it is fixed then so is the other. It is as if God had only to decide how the basic level is arranged and the other follows on automatically, just as a composer need only decide on the notes and the tempo and then the melody emerges without anything further being done. In this kind of case the relationship is simply one of composition, and there is perhaps nothing mysterious in the way that the melody supervenes upon the notes and tempo, any more than there is anything mysterious in a house supervening on an arrangement of bricks and mortar. But in other cases we are not talking of the supervenience of one thing upon other things, but of one kind of fact upon other kinds of fact, and the relationship seems not to be so simple. In the philosophy of mind, for example, physicalists will certainly want to say that the fact that a creature is alive or conscious supervenes upon complex neurophysiological facts. God needs only to create beings of sufficient neurophysiological complexity and thereby will have created living or conscious beings; no further act of breathing life into the being is necessary.

But it seems only metaphorical to say that life is ‘composed’ of neurophysiological complexity. The relationship is not simply that of a building to its building blocks. Similarly, one might think that the fact that a computer is running a particular programme supervenes upon microphysical facts about its components, yet it is hard to make sense of the idea that its running the programme is literally composed of some arrangement of hardware. The term ‘supervenience’ was itself originally applied to the way ethical properties relate to natural ones, and in this case too it is hard to see how a person’s virtue can literally be composed out of their natural dispositions, although here it may be more appropriate to think in analogous terms, since at least their virtue may be said to have components, such as charity or honesty. In any event, a general definition is needed.

If the root idea is that once the basic level is fixed, nothing more need be done to fix the upper level, then at least it should be true that if two things are identical in respect of their basic properties, then they must also be identical in respect of the upper-level properties. The other way round this is not necessarily so, for two things may be identical in the upper-level respect without being identical all the way down. For example, two computers might be running the same programme but with rather different configurations of hardware, or two people may be equally virtuous but because of somewhat different characters. This is frequently referred to as the ‘variable realization’ of the upper-level properties, and is compatible with their supervenience on the lower level. Returning to the root idea, we can now frame it as an impossibility claim:

  • (S) It is not possible that two things should be identical in respect of their lower-level properties without also being identical in respect of their upper-level properties.

Various questions arise. First, and most obviously, any such claim only makes sense given a background distinction between the two levels of property. If we cannot genuinely separate the basic or subjacent (sometimes called subvenient) level from the upper or supervening level, then no such claim can be formulated. And in interesting cases, such as the supervenience of the ethical on the natural, there will be philosophers who deny that such a separation can be achieved, seeing it as dependent on a doubtful fact–value distinction.

Similarly, there may be disputes about priority: some philosophers may hold that the semantic elements of a language supervene upon the intentions with which its speakers use its elements; others hold that the content of peoples’ intentions supervene upon the semantics of the language in which they would be disposed to express them (this kind of dispute also generates the battle between methodological individualism and various kinds of holism). There may also be disputes about the extent of the basic class. Thus, to some philosophers it is important that the present content of someone’s sayings should supervene only on their present brain states; to others it is important that it supervene upon the nature of the historically extended entire social and physical environment in which they live.

A second range of problems arises because we might worry about the strength of impossibility in any particular case. Is it a matter of logical impossibility, or metaphysical impossibility, or is it simply a question of the laws of nature? One might need to advance claims in different strengths, such as:

  • (E) It is not possible, as a conceptual matter, that two things should be identical in respect of their natural properties without being identical in respect of their ethical properties.

  • (M) It is not possible, as a metaphysical matter, that two things should be identical in respect of their physical properties without being identical in respect of their mental properties.

  • (P) It is not possible, given the laws of physics as they are, that two things should be identical in respect of their microphysical properties without being identical in respect of their macro-physical properties (such as weight or tensile strength).

Such formulations probably do not exhaust the intended significance of supervenience claims. For example, they do not capture the asymmetrical dependency, whereby the upper arises ‘in virtue of’ the lower (after all, denying variable realization, someone might think it impossible that two things should be identical mentally without being identical physically, but this would not mean that the physical supervenes on the mental). Nevertheless, they form the essential core.

The third kind of claim is perhaps the easiest to understand: it only demands that the laws of physics, as they actually obtain, connect the two levels in such a way that identity in one respect delivers identity in the other. In different possible worlds obeying different physical laws, the connection may be severed, with the weight or tensile strength of objects depending on other factors also (in fact weight, in our world, actually depends on other factors, such as the distribution of matter around the universe). The only problem with this kind of supervenience claim is that it inherits the difficulties surrounding the concept of a law of nature.

A claim such as (M) raises the stakes. Here we want to say that there is a more intimate connection between the physical and the mental. It is not just in the world as we have it that mental properties emerge out of physical properties. Rather, in any world we can properly describe, the mental must emerge out of the physical. There is simply no metaphysical possibility of a ‘ghost stuff’ or separate source or underlying basis of mentality. It has to occur in creatures of particular kinds of physical complexity. It offends against metaphysics, not just physics as we have it, to postulate two physically identical creatures, one of whom is conscious, the other a zombie. Anybody denying this would be thinking in terms of a dualism of some metaphysically improper stamp. Notice that although many philosophers will subscribe to this thought, it is not altogether easy to formulate it. One distinction frequently made is between ‘strong’ supervenience, whereby the claim applies to any two things in any possible worlds, and ‘weak’ supervenience, whereby it only governs things within the same world. The difference is that across worlds we might imagine different laws of nature, and perhaps different constituent elements of nature, whereas one world must be governed by one set of laws and contain only whatever elements it has.

Finally according to (E) anyone denying the supervenience is in conceptual error, betraying some kind of incompetence with the concepts involved. This goes further than (M), unless we think that metaphysics is no more than a display of what is implicit in the concepts with which we think. It is, at least, clear that someone disposed to value two things differently, although admitting that they are identical in every other respect, is somehow confused or out of line, or ignoring the entire purpose of the notion of valuing a thing, compared with someone who denies (M), perhaps because of lingering Cartesian dualism (see Dualism).

Citing this article:
Blackburn, Simon. Varieties of supervenience. Supervenience, 1998, doi:10.4324/9780415249126-N057-1. Routledge Encyclopedia of Philosophy, Taylor and Francis,
Copyright © 1998-2019 Routledge.

Related Articles