PHIL2206 Philosophy of Mind
Week 13: Reductive and Nonreductive Physicalism

Substance physicalism

  1. Early in the course we looked at substance dualism - the view that there are substances of two kinds, physical and non-physical, and that minds are non-physical substances. We saw that one big problem for this view is explaining how mental causation is possible.
  2. Most philosophers of mind today are substance physicalists - they accept that all substances are physical. So mental substances, if any, are physical substances.
  3. The main issue is whether or not mental properties are physical properties, or at least reduce to physical properties (so they are 'nothing over and above' physical properties).
  4. According to reductive physicalists, mental properties do reduce to physical properties - they are nothing over and above physical properties.
  5. According to nonreductive physicalists, mental properties do not reduce to physical properties - they are something over and above physical properties.

Reduction

  1. What is it for properties of one kind to reduce to properties of another kind? Above we have used the slogan, 'nothing over and above', but can we say more?
  2. There is perhaps a sense in which the property of being a witch reduces to the property of being a psychotic woman: the latter property was mistaken for the former property; really there are no witches, there are just psychotic women. This would be an example of eliminative reduction.
  3. Another kind of eliminative reduction is when we lose interest in properties of one kind in favor of properties of another kind.
  4. But most philosophers of mind agree that there are such things as mental properties and that they are interesting properties. The interest is in whether they reduce to physical properties in some non-eliminative sense.
  5. There are at least three ways in which properties of one kind, K, might count (in a non-eliminative sense) as reducing to properties of another kind, k.

Bridge-law reduction

  1. Suppose we have lawful correlations between properties of kind K and properties of kind k:

    For each property P of kind K there is a property p of kind k such that it is a law that: for all things x and times t: x has P at t iff x has p at t.

    (These are also called bridge laws, hence the name of this kind of reduction.)

    Suppose also that we can derive all the laws about properties of kind K from the laws about properties of kind k by using these lawful correlations.

    Then there is a sense in which properties of kind K reduce to properties of kind k. This is bridge-law reduction.

  2. Example. Suppose that we have one law about mental properties, one law about brain properties, and two lawful correlations between mental properties and brain properties:

    Then we can derive the law about mental properties from the law about brain properties using the two lawful correlations, and mental properties bridge-law reduce to brain properties.

  3. But this is not a very satisfying kind of reduction of mental properties:

Identity reduction

  1. Suppose we have identities between properties of kind K and properties of kind k:

    For each property P of kind K there is a property p of kind k such that P = p.

    Then there is a sense in which properties of kind K reduce to properties of kind k. This is identity reduction.

  2. Example. Suppose that there is just one mental property, being in pain, and it is identical to the brain property of having one's c-fibres firing. Then mental properties identity reduce to brain properties.
  3. Note some nice things about this kind of reduction of mental properties:
  4. But:

Functional reduction

  1. Suppose we have functional characterisations of properties of kind K:

    For each property P of kind K there is a role C such that to have P is to have some property p that plays role C.

    Suppose also that for each property P of kind K the role associated with P is played by a property of kind k.

    Then there is a sense in which properties of kind K reduce to properties of kind k. This is functional reduction.

  2. Example. Suppose that there is just one mental property, being in pain. Suppose that to be in pain is to have some property p such that p is caused by tissue damage and causes wincing and groaning. Suppose that, in humans, it is the firing of c-fibres that is caused by tissue damage and causes wincing and groaning. Then mental properties (in humans) functionally reduce to brain properties.
  3. Note some nice things about this kind of reduction of mental properties:

Reductive physicalism

  1. We have looked at three substance physicalist theories of mind according to which mental properties reduce to physical properties:
  2. According to logical behaviourism, mental properties are behavioural properties (which are a special kind of physical property). If this is right, then mental properties identity reduce to physical properties.
  3. According to the identity theory, mental properties are brain properties (which are a special kind of physical property). If this is right, then mental properties identity reduce to physical properties.
  4. According functionalism, mental properties are functional roles, and they are played (in humans) by physical properties. If this is right, then mental properties functionally reduce to physical properties.
  5. But we have seen problems for each of these theories. For example:
  6. These days, the prospects for reducing mental properties to physical properties are thought to be dim, and most embrace non-reductive physicalism.

Nonreductive physicalism

  1. Nonreductive physicalists hold the following two views:
  2. But they typically hold the following two views as well:
  3. Does it make sense to suggest that mental properties supervene on physical properties, but do not reduce to physical properties? How does this work?
  4. According to one view, mental properties emerge from sufficiently complex systems of physical substances. They are determined by the phenomena from which they emerge, but they are 'new' properties that do not exist at the 'lower' level.

    Transparency might be like this. It doesn't make sense to talk about the transparency of a single water molecule, or of a small number of water molecules - the property doesn't really exist at this level. But gather together sufficiently many water molecules and it does start to make sense - the property exists at this higher, more complex level.

    This is unlike the property of having mass, which is sometimes called a resultant or additive property.

    We might employ the following slogan to capture the idea of emergence: the whole can be more than the some of the parts.

  5. Does it make sense to suggest that mental properties supervene on physical properties, do not reduce to physical properties, but can causally interact with physical properties?
  6. We have seen some problems for this idea:
    1. The exclusion argument. The idea runs afoul of the principle of the causal closure of the physical domain.
    2. The supervenience argument. M causes M*, but must do so by causing P*. But that requires mental-to-physical causation, which is under pressure from the exclusion argument.
  7. So nonredeuctive physicalism faces plenty of problems.