Functionalism and the principle of multiple realizability

Photo by bruno neurath-wilson on Unsplash

Introduction

     After discussing substance dualism (more specifically, Cartesian interactionism), analytical behaviorism, and the identity theory of mind, this day’s introduction will be regarding functionalism. As substance dualism, analytical behaviorism, and the identity theory, functionalism is a theory of mind in philosophy.

     George Rey, in Houdé’s (2004) dictionary of cognitive science, describes functionalism as “the name of a popular philosophical strategy with regard to the proper analysis (or definition) of mental phenomena (mental terms, concepts, properties).” Rey continues, stating that functionalism is based upon a simple idea: the fact that many things in the world are what they are, not particularly by virtue of what they are made of, but by virtue of what function, or role, they serve in a system.

     Rey continues his explanation using money as an analogy. Rey states that something is money by virtue of its being the kind of thing that serves in a certain way to exchange commodities. Hilary Putnam was a prominent philosopher who pioneered this way of thinking about mental properties. Since Putnam, many philosophers of mind have argued that mental phenomena ought to be understood in this way (Houdé, 2004).

The beginning of functionalism

     Heil (2013) affirms that the identity theory of mind enjoyed but a surprisingly brief period of favor among philosophers. The reason why identity theory was not favored for a long time was a consequence of the rise of functionalism. Functionalism is not a materialist theory per se, but it can be understood as compatible with the aims of materialism. Most functionalists, states Heil, would regard themselves as materialists of one sort of another.

     Chalmers (2002) explains that, broadly speaking, functionalism holds that mental states correspond to functional states. This means, states of playing a certain role within the cognitive system (compare with Rey). Chalmers, in his book on classical and contemporary readings on philosophy of mind, identifies two different forms of functionalism:

1. Machine functionalism: According to Putnam, says Chalmers, mental states are functional states of a computational machine. This makes mental states more abstract than any particular biological state, and so allows the possibility of multiple realization. Also, it allows a loose tie between mental states and behavior, without an absolute tie.
2. Analytic functionalism: Armstrong, contrasting Putnam, held a different form of functionalism in terms of the general idea that mental states are defined in terms of their causal role. He held that the concept of a mental state is the concept of a state that is apt to be the cause of certain effects or apt to be the effect of certain causes.

     On the other hand, Block (1980) has stated that three functionalisms have been enormously influential in philosophy of mind and psychology: (1) functional analysis, (2) computational-representation functionalism, and (3) metaphysical functionalism. We care, nonetheless, about metaphysical functionalism only. Block states that metaphysical functionalism is a theory of the nature of the mind, rather than a theory of psychological explanation. Metaphysical functionalists are basically concerned with what mental states are, not with how they account for behavior. Functionalists believe that mental states are functional states (see also Heil, 2013; Houdé, 2004; Chalmers, 2002; Braddon-Mitchel & Jackson, 2007).

     The main concern of metaphysical functionalism, adds Bock, is the same as that of behaviorism and physicalism. All three doctrines address questions such as “What is pain?” or, put simply, “What is pain qua pain (hence, what is there common to all pains in virtue of which they are pains?)?” It must be borne in mind that functionalism is concerned with mental states types, not mental states tokens. This means that, using pain as an example, functionalism is concerned with pain (a type of mental state), and not with particular pains (different tokens of the same mental state, in this case, pain in general). This is why Block stated before that functionalism addresses questions such as “What is pain qua pain?” and not “What is a particular pain like?”

Abstracting from neurological processes

     When we consider mental states as functional states, it is the same when we consider computational operations in a computing machine, in which we ‘abstract’ from the machine’s hardware (Heil, 2013). Two very differently constructed mechanisms can be said to perform the same computations, or run the same programs. Imagine a regular calculator and an iPhone. Both can do basic mathematical operations and yet both constructed mechanisms are quite different.

      Heil also adds that functionalists think of computational processes not as material processes, but as ‘realized’ in material systems. The material system that realizes a given computational operation in a normal calculator differs from material processes that realize it in a more modern way, such as smartphones. If there are immaterial substances, the very same processes could well have an immaterial realization (Heil, 2013) too. This is what functionalists call the principle of multiple realizability.

     “Suppose you thought of minds in roughly the way you might think of computing machines. A mind is a device capable of performing particular sorts of operation. States of mind resemble computational states, at least to the extent that they could occur, in principle, in many different kinds of material (and perhaps immaterial, a qualification I shall henceforth omit) system. To talk of minds and mental operations, is to abstract from whatever ‘realizes’ them: to talk at a ‘higher level’.”

Heil (2013)

     Basically, minds are capable of performing certain types of operations. In this manner, states of mind resemble computational states, since they can occur, at least in principle, in different kinds of material systems. When we are talking about mental operations, we are abstracting from the material system that realizes these operations. What we are doing is talking at the ‘software’ level, instead of the ‘hardware’ level.

The principle of multiple realizability

     Braddon-Mitchell & Jackson (2007) explain that there is a distinction to be drawn between the functional role and what occupies or fills it. Imagine thermostats; some are bimetallic strips and some are more complex electronic devices. We see that the functional role of a system, in Heil’s terms, is different from the system that realizes it. Braddon-Mitchell & Jackson state that functionalists distinguish the functional roles specified by the input, output and internal role clauses from what occupies them, and insist that “what matter for being in one or another mental state are the roles that are occupied, not what occupies them,” which is, at least to me, more or less just a way to escape from inter-theoretic reduction in cognitive science.

      Braddon-Mitchell & Jackson also suggest that multiple realizability is appealing for a number of reasons, namely, (1) we do not know what realizes mental states, (2) we can imagine beings unlike us but with mental states (appealing to modality), (3) human brains may be very diverse, (4) if a different part of the brain takes over a job, we do not mind, and (5) we might replace part of our brains with artificial aids.

     Heil (2013) provides a figure explaining multiple realizability, which is meant to represent schematically a case of a multiply realizable property. Let’s imagine that M₁ is a mental property —such as being in pain— and that P₁, P₂, P₃, P₄ are physical realizers of M₁. The idea is that M₁ is a ‘higher-level’ property, distinct from, but dependent on, its several realizing properties.

     From Heil, John (2013). Philosophy of mind. A contemporary introduction. 3rd edition.

Summary

     Although this is not a comprehensive introduction to functionalism, it will serve its purpose as a brief summary of its most important tenets. For more information on functionalism, you can check out the references below. Before doing so, nonetheless, we shall enumerate the main tenets of functionalism discussed in this short article:

1. Many things in the world are what they are, not particularly by virtue of what they are made of, but by virtue of what function, or role, they serve in a system.
2. There are, according to Chalmers, to main forms of functionalism: machine functionalism and analytical functionalism.
3. States of mind are not neurological processes, but are ‘realized’ in neurological processes like computational processes are ‘realized in material systems.
4. Hence, mental states are functional states.
5. Functional states, or properties, are multiply realizable in different material (or even immaterial) systems.
6. In principle, thus, mental states are multiply realizable in different material systems.

References

Block, Ned (1980). Readings in philosophy of psychology, Volume I

Braddon-Mitchell, D. & Jackson, F. (2007). The philosophy of mind and cognition. 2nd edition

Chalmers, David J. (2002). Philosophy of mind. Classical and contemporary readings

Heil, John (2013). Philosophy of mind. A contemporary introduction. 3rd edition

Houdé, Olivier (2004). Dictionary of cognitive science. Neuroscience, psychology, artificial intelligence, linguistics, and philosophy