Functionalism and property dualism are both physicalist theories of the mind in that they don't admit any substances other than physical substance.

Property dualism holds that mental states are non-reducible properties of physical brains. Functionalism says that mental states should be defined in terms of their functional roles as opposed to being identified with physical brain states.

I can't see any difference between these positions other than they are looking at slightly different aspects of the mind-body problem: Property dualism is looking at it from an ontological point of view, while functionalism is addressing the issue of causal relationships between metal states, brain states and behavior. In fact, it seems to me that functionalism is just a refinement of property dualism: mental states are properties irreducible to brain states, and these properties can be described by their functional relationship to other brain mental states and to behavior.

In both cases, the key point is: mental states can't be reduced to brain states and behavior, they are fundamentally anti-reductionist. Once ones dismisses reductive physicalism (i.e. identity theories of the mind), whether it is a dualist theory, a functionalist theory or some form of emergentism seems a matter of semantics.

Yet in the literature, functionalism and property dualism are frequently described as being in opposition to each other: Daniel Dennett for example is described as functionalist who is opposed to dualism (presumably including property dualism).

My questions:

  1. What is the difference between property dualism and functionalism?
  2. Are there any specific points on which property dualism and functionalism are incompatible?
  3. More generally, how can we ever differentiate non-reductive physicalist theories and dualist theories?
  • why are you describing functionalism as non reductive? is a running program irreducible to the physical computer running it?
    – nir
    Commented Nov 4, 2015 at 19:02
  • @nir , inherently, there's no reason why it shouldn't be. But several courses, or lectures I've looked at mention it as an alternative to type identity theory. Multiple realizability is mentioned as the reason why type identity doesn't hold and functionalism does. It seems to me that any form of reductionism would imply type identity, since reducing the mental state to a physical state pretty much amounts to identifying it as such, no? Commented Nov 4, 2015 at 19:35
  • I don't see why; given the complete state of a physical system, and the laws which govern it, may you not deduce its future evolution? if so, in what way is functionalism anything more than a high level story of the dynamics of the physical system? e.g. this collection of atoms functions as a spring, or an engine, or a computer, etc...
    – nir
    Commented Nov 4, 2015 at 20:16
  • @nir It is not the high-level story -- it is the insistence there are levels of stories, that there is a high-level story, and a low-level story, and some medium-level stories, and that having only the single lowest-level one does not give you the whole story in any case. Facts at one level of realization may not be traceable to those at a lower level -- what is a recession, in terms of minting coins?
    – user9166
    Commented Nov 4, 2015 at 20:45
  • If the lower level always explains the upper one, there should be a chain of causation from physical money (or in a reserve system from requests-to-lend) to market patterns -- and there just isn't, or if there is, it flows the other direction.
    – user9166
    Commented Nov 4, 2015 at 20:49

2 Answers 2


The main difference is that functionalism is not an ontological doctrine, although it imposes some constraints on ontology, while property dualism is. The point of functionalism is to reduce consciousness to its manifestations in terms of its functional role in behavior, leading to the idea that it is implementation independent. This is compatible with most forms of physicalism, except the hardcore type identity, which identifies each mental state with a specific physical entity, rather than a functional pattern. Any weaker materialism, e.g. functional properties are physical properties, or are composed of such, is compatible with functionalism. Lewis is both a functionalist and a reductionist.

Property dualism is in principle also compatible with functionalism, but most property dualists make additional claims that exclude it. The best known is Searle (although he disputes being a property dualist), who denies implementation independence by insisting that biochemical processes have special properties that produce intentionality. Davidson's epistemology also precludes fully functionalist description of the mental, according to him normative relations among intentional states can not correspond to empirical relations among internal states and sensory input/output due to indeterminacy of translation. But his theory of meaning is a kind of holistic pseudo-functionalism. The root of it is that most current property dualists postulate a weaker relation between physical and mental (something like supervenience) than is minimally required for a functionalist description. But that is over and above property dualism as such.

  • "This is compatible with most forms of physicalism, except the hardcore type identity, which identifies each mental state with a specific physical entity, rather than a functional pattern." This is the part I don't get. I can identify a human leg with a specific physical entity, the configuration of bones and tissue that form a human leg, and still speak about a broad functional category of legs in humans, spiders and robots. If functionalism is to really be non compatible with type identity, as opposed to being just a semantics game, then it has to preclude reducibility. Commented Nov 4, 2015 at 22:34
  • 3
    @Alexander S King Perhaps you underestimate how strong the type identity of Smart, etc. was. They literally claimed that pain and C-fiber stimulation are interchangeable in all contexts, identical types, almost no one holds this anymore. Lewis showed how rather robust reductionism can be combined with functionalism, this is now called analytic functionalism, perhaps type identity you have in mind is something like that. plato.stanford.edu/entries/functionalism/index.html#AnaFun
    – Conifold
    Commented Nov 5, 2015 at 1:39

Most forms of functionalism don't insist on irreducibility.

So the one, functionalism, is a perspective on what is appropriate and productive, and the other, property dualism, is a theory of what is and is not possible.

Property dualism is, to my mind, fully compatible with 'emergentist' form of functionalism -- that mind is an emergent product of biology and it is counterproductive to reason about emergent properties primarily in terms of their lower-level causes.

Emergentism says this is not only because

1) attempts to do so are often misleading, (e.g. it is tempting to ascribe Brownian motion to quantum forces, instead of complex largely-Newtonian dynamics, but it is not reasonable). This is one version of the standard notion of functionalism.

But also because

2) actual derivation of such a reduction is often impossible (e.g. in the previous example the mathematics descends into the chaotic dynamics of feedback, at which point it can only be measured, not predicted from more detailed observations).

And the latter assertion of not just pointlessness, but likely impossibility makes it a property dualism.

  • How is functionalism that accepts reducibility different from type identity? Commented Nov 4, 2015 at 19:40
  • The facts of chemistry and physics do not constitute different 'types' in the sense of 'type identity'. But they do represent different functional perspectives -- chemistry is reducible to physics, but it is not productive to do most of your chemistry as physics. Functionalists who accept reducibility apply the same distinction there that they do to brain and mind. Propery dualists would not.
    – user9166
    Commented Nov 4, 2015 at 19:54
  • Maybe I'm misunderstanding, but you seem to be confirming my point: functionalism + reducibility is the same as type identity (hydrogen atom -> one electron plus a nucleus with one proton and one or two neutrons) Commented Nov 4, 2015 at 20:05
  • Functionalism is a perspective, reducibility is a physical theory, and type identity is a logical category. No such math is possible.
    – user9166
    Commented Nov 4, 2015 at 20:10
  • 1
    The vast majority of hydrogen atoms are tokens of that type. But if you are doing chemistry, that is usually irrelevant, and the ones that aren't (deuterium, e.g.) behave very much like the ones that are (they form water with oxygen, e.g.) -- so 'hydrogen', chemically is not really about that. Functionally hydrogen has distinct types for chemistry and nuclear physics. What really does not matter, functionally when you are talking about mind, is whether the thoughts 'are' or 'arise from' neurological processes. Thoughts may have both these types, or only one of them.
    – user9166
    Commented Nov 4, 2015 at 20:21

You must log in to answer this question.

Not the answer you're looking for? Browse other questions tagged .