Functionalism, in philosophy, is a theory about the essential nature of minds—a metaphysical theory with possibly dramatic consequences for our society in the near future. Because one of the strongest motivations for functionalism, among its supporters, is its implication that artificial intelligence could indeed be conscious. According to most versions of functionalism, any machine that can do what a human mind can do, must be considered to have mental states, just as much as a human being.
This is the basic idea of the Turing Test, which you may have heard of. The Turing Test was proposed in 1950 by Alan Turing, the British scientist who created the modern theory of computing and built the first computers in order to break Nazi codes during WWII. Turing suggested that if a machine can fool a human interviewer into believing that it is human, through dialogue (through a terminal of course) then it must be considered intelligent in the same sense as a human being; that is to say, we must admit that it thinks.
It is not clear, according to Turing’s argument, whether this would mean that the machine has that seemingly magical quality that human beings call “consciousness” or whether it means that there is no such quality in human beings either. But in either case, it does mean that “being a mental state” is not dependent on having a human-like brain, or any specific physical qualities; any physical system that can do what a mind does is good enough.
Not all functionalists are overly concerned with A.I.; the primary motivation for the philosophy is to explain the nature of mind, considering that previous explanations—dualism, behaviorism, and “identity theory,” have been rejected by almost all scientists and philosophers—and functionalism remains the most popular understanding of mind among scientists today. We’ll tell that story in history section.
There are also many other “functionalism”s – such as in the study of art, literature, and anthropology, for example — that don’t have a lot to do with philosophical functionalism, and we shall not discuss them here.
The greatest controversies in functionalism have been debates over several related attempts to disprove the basic idea; these include John Searle’s Chinese Room, Ned Block’s China-Brain, and the “zombie” thought experiments. They all present imaginary scenarios intended to demonstrate that some system functionally equivalent to a human brain would not have mental-states.
The Chinese Room argument asks you to imagine a person inside a locked room, communicating with the outside world only by passing messages written in Chinese through the door. The person in the room has no understanding of Chinese. She or he receives a message in Chinese from the outside, and then follows a book of instructions (written in English) in order to put together a sensible response in Chinese. The instructions just tell the person that when they see such and such Chinese characters in such and such an order, to respond with such and such other characters, without the person actually knowing what any of the Chinese means. Searle’s point is, that from the point of view of people outside the room, the room “speaks Chinese” perfectly. Therefore, according to functionalism, the room has mental-states. Since, according to Searle, the room obviously does not have mental-states, functionalism is incorrect.
Searle’s argument has been countered mainly in two ways. Some functionalists simply reply, that indeed, the room does have mental-states; they’re just not observable from the point of view of the person in the room. Another, perhaps more sophisticated response, is that Searle’s Chinese Room isn’t possible; that, in order to process language as well as a human being, other processes are necessary, which do not require the biology of human brains, but are more similar to brain-activity than the process described by Searle.
The China-Brain and zombie thought experiments attempt to make up for the short-comings of the Chinese Room. The China-Brain would be a mind made out of the activity of billions of Chinese people, each one performing the function of a single neuron in a brain (sending signals to each other according to certain protocols)—so that their collective activity would be functionally equivalent to a human mind. For those of us who don’t believe the China-Brain would have consciousness, this argues against functionalism. And the zombie thought experiment simply says that if it is possible to create a being that behaves just like a human, but without consciousness, then functionalism should be wrong.
Since the question of whether any of these experiments could actually reproduce human functionality without consciousness has not been settled, these debates rage on!
III. Famous Quotes
“One can imagine a computer simulation of the action of peptides in the hypothalamus that is accurate down to the last synapse. But equally one can imagine a computer simulation of the oxidation of hydrocarbons in a car engine or the action of digestive processes in a stomach when it is digesting pizza. And the simulation is no more the real thing in the case of the brain than it is in the case of the car or the stomach. Barring miracles, you could not run your car by doing a computer simulation of the oxidation of gasoline, and you could not digest pizza by running the program that simulates such digestion. It seems obvious that a simulation of cognition will similarly not produce the effects of the neurobiology of cognition.” — John Searle, “Is the Brain’s Mind a Computer Program?”, Scientific American (January 1990).
This quote conveys Searle’s objections to functionalism most clearly. Searle has argued many times that the mind is a biological function, dependent on biological processes, or something similar. Although this quote is compelling, Searle could be accused of using false analogies here. The relationship between digestion and a simulation of digestion is different from the relationship between cognition and its simulation – because a computer’s “simulation” of cognition, such as figuring out the answer to a math problem, does perform that function, whereas a simulation of digestion does not perform digestion!
“Are zombies possible? They’re not just possible, they’re actual. We’re all zombies.* Nobody is conscious — not in the systematically mysterious way that supports such doctrines as epiphenomenalism. (*It would be an act of desperate intellectual dishonesty to quote this assertion out of context!)” – Daniel Dennett
Daniel Dennett, one of the most influential supporters of functionalism today, gives us his interpretation of the zombie thought experiment – the scenario in which artificial beings that behave exactly like human beings, are created, and have no consciousness. A functionalist must believe that such zombies and human beings equally do or do not possess consciousness. While many functionalists, especially computer programmers, like to believe that such machines would have consciousness, Dennett’s position – that human beings don’t actually have consciousness either – is also a defense of functionalism. There has been much loud criticism of Dennett’s position; Dennett responds that none of us understand what he’s saying!
Hilary Putnam’s 1950 version of functionalism, where mind is thought to be made of states and the rules governing transitions between those states, which is also the abstract definition of a Turing machine — any digital computing device.
This theory, associated with Jerry Fodor, is similar to structuralism, regarding the elements of mind as psychological units, however not necessarily just the ones we’re conscious of, but including the elements of unconscious cognition.
This is a theory about how we should define mental states – e.g. whether “belief” is such-and-such going on in the neurons, or whether it is better defined through a functional analysis – i.e. “believe” means “to mentally represent some state of affairs, and that it is true.”
Functionalism didn’t gain its name until the 20th century, but the earliest theory that could be considered a kind of functionalism was Aristotle’s theory of the soul. Unlike most of his contemporaries, Aristotle did not believe that the soul existed independently of the body, nor that it was identical to the material of the body or brain. He seemed to consider the soul identical to what we call “consciousness” and to be made of its functional capacities, such as its ability to think. Aristotle’s view may conflict with some versions of functionalism because he believed that an organ’s functionality was created by its physical form. This is consistent with some modern versions of functionalism.
Thomas Hobbes, in the 17th century, voiced a point of view remarkably like modern computational functionalism. He claimed that mental activity was made out of computations, and an entirely mechanical process, even though he had never heard of a computer.
It’s poetic that Hobbes published this theory in 1650, the same year that Renee Descartes died. Because points of view like Hobbes’ were motivated by widespread dissatisfaction with Descartes theory of mind-body dualism. Dualism is the idea that mind and matter are fundamentally different “substances” (whatever that means), and owes a lot to religious thought. Dualism is favorable towards the idea of an incorporeal soul. But dualists were never able to say what the “thinking substance” of the mind is; and unable to explain how mind and body could interact if they were made of completely different stuff. Functionalism seems to solve these problems.
VI. Functionalism versus Structuralism
As with functionalism, there are several other “structuralisms,” such as in theories about society, that have nothing to do with philosophy of mind. In relation to functionalism, structuralism was a psychological theory popular around the beginning of the twentieth century that mind should be understood as the structure of inter-related psychological elements, like beliefs and desires. This may have been the first theory of psychology that did not reduce mind to something else, such as physics, behavior, or soul, and in this way structuralism prepared the way for functionalism.
Functionalism is not structuralism, because it defines mind in terms of what it does, such as forming images and making inferences. Structuralism implies that mind is made of psychological experiences. It’s a subtle distinction but crucial.
VII. Functionalism in Pop Culture
Example #1: The Imitation Game
In this clip, Benedict Cumberbatch, playing Alan Turing, the father of computing, presents a functionalist point of view when he says that digital computers should be accepted as truly thinking machines. It is functionalist because Turing assumes that such machines will be based on the kind of circuitry we are familiar with today, which is not much like the human brain. Most recent stories about A.I., such as Westworld, are not necessarily functionalist because they don’t tell us much about the machines’ “brains”; if the machines’ “brains” are made of neural networks sufficiently similar to human brains, then it’s not necessarily a functionalist perspective; the consciousness of the machines might depend on mimicking human brain function.
Example #2: Tron
You have to go back a few years to find sci-fi where the kind of computers we use today (or in the 1980s) are thought capable of hosting conscious software. Tron, in this clip from the original 1982 version, goes one better. Not only is it full of conscious “programs,” but a young Jeff Bridges is fully transferred into the computer (very un-dude!). This scenario depends on the functionalist assumption that mental-states do not necessarily require brain-like materials or activity.