How Multiple Realization Undid Machine-State Functionalism

Hilary Putnam founded machine-state functionalism, also called computational functionalism. He later turned its best argument against it. In Representation and Reality, he compared the move to jujitsu: use an opponent’s strength to throw him. The opponent happened to be his own earlier theory. Philosophers admire that sort of candor, preferably from a safe distance.

He did not hide or minimize his defection in a footnote. Chapter 5 bears the title “Why Functionalism Didn’t Work.” No euphemism there. The whole chapter says where the theory failed.

The reversal took more than twenty years. It began with a case against brain-state identity, gained a second edge from semantic externalism, and ended as a case against identifying mental states with computational states. The logic becomes easier to see if we take those steps in order.

The First Turn: One Mental State, Several Physical States

In the 1950s and 1960s, identity theorists proposed that mental-state types might prove identical with brain-state types. Pain, for example, might turn out to be C-fiber firing. Putnam thought the proposal tied minds too tightly to one kind of animal. Humans, octopuses, and possible extraterrestrials could all feel pain while possessing very different nervous systems.

His first argument can be stated in three steps:

  1. If pain were identical with one physical-state type, every creature in pain would occupy that physical-state type.
  2. Creatures with very different physical designs could all feel pain.
  3. Therefore, pain cannot be identical with any one physical-state type.

This argument does not deny that every actual pain has a physical realization. It denies a type identity. A human pain and an alien pain may each depend on physical activity without depending on the same physical kind.

Machine-state functionalism offered an attractive replacement. Define pain by the role it plays, not by its material. Ask what tends to cause it, what it tends to cause, and how it interacts with belief, memory, and desire. Different physical systems could occupy the same functional state, much as different computers can run the same program. Putnam had freed the mind from one material by giving it an abstract job description.

The Middle Turn: The World Helps Fix Content

In 1975, Putnam published “The Meaning of ‘Meaning.’” Its Twin Earth story added a different pressure. Imagine two internally matching people in 1750. One lives on Earth among H₂O. The other lives on Twin Earth among XYZ, a liquid that looks, tastes, and behaves like water. Their internal states match. Their words do not refer to the same substance. The Earthling’s “water” reaches H₂O; the twin’s reaches XYZ.

This result did not yet amount to Putnam’s full case against functionalism. The paper focused on linguistic meaning and reference. Still, it established a limit that any internal machine theory would have to respect: what a state concerns can depend on relations outside the thinker. Matching inner organization does not guarantee matching content.

By 1988, Putnam joined this externalist result to his older multiple-realization argument. The result was the promised jujitsu.

The Second Turn: One Mental State, Several Computational States

Machine-state functionalism had rejected the demand for one physical type per mental type. Yet it still treated mental states as computationally defined states. Putnam now asked why a shared belief should require a shared computational organization.

Consider two people who both believe that many cats live nearby. One counted paw prints in the garden. The other heard nightly cat fights. They may speak different languages, hold different background beliefs, and draw different inferences from what they believe. An alien could share the belief through a cognitive design unlike either human system.

Different routes into the belief would not settle the issue. A machine can reach the same state by several routes. A machine state gets its identity from its place in a transition structure: which inputs and internal states lead to it, and which outputs and further states follow from it. Differences in language, background belief, and habits of inference can change that entire pattern. Across sufficiently different thinkers, the computational state spaces may not even map neatly onto one another. Yet we may still have good reason to interpret the thinkers as sharing the belief.

The second argument has the same skeleton as the first:

  1. If believing that many cats live nearby were identical with one computational-state type, every possible thinker with that belief would occupy that type.
  2. Possible thinkers can share that belief while occupying states with different transition roles, perhaps within computational systems whose state spaces cannot be mapped onto one another.
  3. Therefore, the belief cannot be identical with any one computational-state type.

Putnam called functionalism’s contrary assumption its “Achilles’ heel.” It had denied one brain-state type for every belief across species, then quietly kept a one-state-per-belief picture within each organism. But learning histories, cultures, languages, and styles of thought vary within our species. Possible thinkers could vary far more. The same content can cross those computational differences.

A functionalist can answer by replacing one state with a class. Put every computational state that counts as believing that cats live nearby into the same box. The states need not match internally; they need only belong to the right equivalence class.

Now ask what puts them in the box. The rule must decide when cat, meew, and an alien signal concern the same animals. It must separate an unfamiliar belief from a familiar belief expressed in an unfamiliar way. In short, it must interpret the thinkers before it can sort their computational states. The proposed reduction then moves in a circle. It explains shared content by membership in a class, but identifies the class by deciding which systems share the content.

A functionalist may insist that some objective equivalence relation exists even if no finite thinker can find it. Putnam need not prove otherwise. Machine-state functionalism needs a rule stated in computational terms. An endless list built from prior acts of interpretation would only record our semantic verdicts. It would not reduce them to computation.

The problem runs in the other direction as well. Twin Earth gives us internally matching systems with different contents. If their internal physical organization matches, their internal computational organization can match too. Yet one “water” state concerns H₂O and the other concerns XYZ. No formal transition inside either system settles that difference. The environment does.

The two arguments work in opposite directions. Multiple realization challenges necessity. Twin Earth challenges sufficiency:

  • Same content, different computational states: no particular internal computational-state type is necessary for that content.
  • Same internal computational state, different content: no particular internal computational-state type is sufficient for that content.

The first point defeats the demand that every instance of a belief share one computational type. The second defeats the claim that occupying a computational type fixes what the belief concerns. Together they block the one-to-one identity proposed by simple machine-state functionalism.

What the Argument Does Not Show

The conclusion needs a fence. Putnam’s argument rejects the constitutive identity just defined. It does not deny that brains compute, that computational models explain parts of cognition, or that an artificial system might think. The modest claim that computation contributes to thought remains untouched.

Nor did Putnam dispose of every wide form of functionalism. A theorist may let the relevant role extend through the body, environment, social practice, and causal history. Putnam considered such a sociofunctional rescue. It accepts much of his externalism and abandons the picture of an isolated program running inside one skull. It may still count as functionalism. But formal computation alone no longer fixes the content. The world-involving relations now carry part of the load.

Putnam also offered a separate argument about implementation. Under a permissive mapping, an ordinary physical system can count as realizing any finite automaton. The philosopher David Chalmers later required the system’s causal structure to mirror the formal transitions. That reply may secure objective computation and block the cheap mappings. Grant it. The semantic problem remains: a genuine computation does not determine its own reference merely by running.

My View

Putnam supplies the negative result: formal computation cannot by itself tell us what a state concerns. My positive account goes beyond his. A world-involving producer-consumer history—through learning, selection, or both—may fix a representation’s accuracy conditions: what it concerns and what would make it mistaken. That can establish content at the level of a mechanism without a separate proprietor. Whether the larger system thereby believes, understands, or acts depends on a further question: how the state participates in the system’s integrated cognitive organization.

Neither a world-involving history nor cognitive integration requires carbon. An artificial mechanism could acquire non-derived content, and an artificial system might think by means of it. Silicon may qualify; an abstract program, considered only as a formal transition structure, does not.

Multiple realization first freed mental states from one physical material. Properly followed, it also frees them from one computational form.

-gts


This essay is part of Mind, Matter, and Meaning, an AI-assisted philosophy project by Gordon Swobe exploring consciousness, meaning, and artificial minds. Learn more about the project and read the book.

References

Putnam, Hilary (1967). “Psychological Predicates.” In W. H. Capitan and D. D. Merrill (eds.), Art, Mind, and Religion, 37–48. Pittsburgh: University of Pittsburgh Press. Reprinted as “The Nature of Mental States” in Mind, Language and Reality: Philosophical Papers, Volume 2. Cambridge: Cambridge University Press, 1975.

Putnam, Hilary (1975). “The Meaning of ‘Meaning.’” In Keith Gunderson (ed.), Language, Mind, and Knowledge, Minnesota Studies in the Philosophy of Science 7, 131–193. Minneapolis: University of Minnesota Press.

Putnam, Hilary (1988). Representation and Reality. Cambridge, MA: MIT Press.

Comments

12 responses to “How Multiple Realization Undid Machine-State Functionalism”

  1. Jason K. Resch Avatar

    I think what is missing from Putnam’s application of multiple realizability to functionalism is the concept of a “substitution level” — the level below which changes can be made to an implementation without making a difference to the phenomenal character of occupying that functional state. Chalmers uses the term organizational invariance to refer to the level of fineness of grain which is required to implement a particular phenomenal state. I argue that the basis of this level comes down to preservation of an interface, beyond which the process cannot access any finer details. An example from computer science is the tole the Java Virtual Machine (JVM) plays for a Java Program. The Java program can run on many different kinds of computer hardware, but from its point of view, it can’t see beyond the interface of the JVM. It forms what I call an “epistemic firewall.” See section 2.3 of my paper for more details on this: https://philpapers.org/archive/RESEHA.pdf

    1. gordon swobe Avatar

      Welcome, Jason!

      Section 2.3 of your paper answers a different objection. Putnam’s main reason for rejecting his own computational functionalism was that even a perfectly defined higher-level computation does not determine perceptual content, reference, or belief.

      The question here is, Why did functionalism fail, per Putnam? I think you need to understand Putnam’s Twin Earth, which rocked the philosophy of language and meaning. It has never been the same since (and this is also thanks to Saul Kripke, Tyler Burge, and others). Hold your entire substitution-level interface fixed and place duplicate systems in Earth and Twin Earth environments. Their internal functional states remain identical, while water refers to H₂O for one and XYZ for the other. Conversely, differently organized systems can possess what we count as the same belief.

      The take-home message is that sameness of computation is neither sufficient nor necessary for sameness of content. Identical internal states can be about different external realities. They can mean different things.

      Putnam is famous for his quip “Cut the pie any way you like, ‘meanings’ just ain’t in the head!”

      -gts

      1. Jason Resch Avatar

        I think my point on the substitution principle applies exactly to this twin earth scenario. Think of two identical java programs, one running on Windows and another running on Linux. Each java program can refer to a file and open a file and write to a file or read a file. From what is available to the Java program it all seems exactly the same, for all information available to the the two Java programs, everything is the same and they are in the exact same computational state. And yet, a file on Windows is a very different thing and has a very different structure than a file on Linux. This is exactly like the two H2Os which may be different on two planets, but from the perspective of everything known by the two identical brains, there is no available distinction available anywhere in the mind. Thus the two conscious states have nothing to distinguish them, and a are thus by definition on the same state. If you think the two minds would occupy different phenomenal states, despite being in identical physical and computational states then you need to explain how the extrinsic factor of the structure of H2O makes itself known to the mind, just as you would need to explain how the Java program could learn anything about how the Windows and Linux operating systems represent files on disk.

        1. gordon swobe Avatar

          > If you think the two minds would occupy different phenomenal states, despite being in identical physical and computational states then…

          I don’t, and Twin Earth does not require that they occupy different phenomenal states.

          Your Java example actually illustrates Putnam’s point. He already grants the internal sameness of Oscar and his twin. His argument allows them to have exactly the same feelings and inner monologue.

          When Oscar says “water fills our lakes,” he is talking about H2O. When his twin says those same words, he is talking about XYZ. Do you agree that they are talking about different substances, even though their internal computation is identical?

          That is the distinction I’m asking about, not whether anything feels different to them.

          -gts

          1. Jason Resch Avatar

            Then you (and Putnam) are debating the meaning of meaning, rather than concepts pertinent to the philosophy of mind. How is this in any way supposed to refute functionalism?

          2. gordon swobe Avatar

            Yes it is about the meaning of meaning. Let me ask you again: do you agree that Oscar and his twin are talking about different substances when they speak of water, despite their being in identical internal states?

            As for how this relates to functionalism, that is what my essay above is about.

            -gts

          3. Jason Resch Avatar
            Jason Resch

            “Machine-state functionalism needs a rule stated in computational terms.” — I think this is what my paper does.

            “Putnam supplies the negative result: formal computation cannot by itself tell us what a state concerns.” — I agree but for different reasons. There is a distinction between a description of a computational state (what is available to others by way of third person descriptions or facts) and what it is like to personally occupy a particular computational state (to have first-person knowledge). This is the solution to Mary’s Room, Nagel’s Bat, and Searle’s Chinese Speaker. See section 7.4 in my paper.

            As to Oscar and his Twin, I think this is an instance of what my paper calls “ontological underdetermination”. Two identical first-person states can be implemented as part of variously distinct computations, environments, and worlds. This is a consequence of Turing Emulability, and applies to all observers if the computational theory of mind is true. Thus every observer is sealed off from their complete and true reality by an “epistemic horizon” hence the paper’s title. See section 9 in my paper: https://philpapers.org/archive/RESEHA.pdf

          4. gordon swobe Avatar

            Jason, yes, we agree that knowing a description of an experience differs from having it, but that is beside the point. The sentence you quoted concerns a different question: what makes a state about one thing rather than another? Knowing a state from the inside does not, by itself, explain what fixes its reference.

            Your “ontological underdetermination” might be compatible with what I am saying here. The twins cannot distinguish their situations from the inside, yet their situations help determine what their words concern. A limit on what they can discover does not remove the relations that fix their reference.

            So let me put the question again directly: do you agree that Oscar and his twin can occupy identical internal computational states while their respective uses of “water” refer to different substances, one to H₂O and the other to XYZ?

            I’m going to leave it there until I get a clear answer from you.

            I posted another essay specifically to answer your question about how Twin Earth relates to functionalism.

            -gts

  2. Jason Resch Avatar
    Jason Resch

    > The sentence you quoted concerns a different question: what makes a state about one thing rather than another? Knowing a state from the inside does not, by itself, explain what fixes its reference.

    To recognize any difference in the aboutness of Oscar and his Twin’s beliefs, requires a view that doesn’t exist anywhere. As the definer of this scenario, you can stipulate from the outside there is such a difference. But it is as if I ran two different VMs on my computer, each one hosting an identical uploaded mind. One of the VM’s has a process ID of 73248 and the other VM has a process ID of 72193. Now if I told the minds they were running in VMs, each could refer to the VM that was hosting their mind, but each would be referring to a different VM. But both perspectives, and both conscious states of these uploaded minds are identical. Moreover, they are sealed away from ever having access to anything on my host machine, so they can never learn the process ID of the VM that hosts them unless I were to tell them. This is a bit interesting, but I don’t see how any of this makes a difference when it comes to philosophy of mind. Perhaps it is of some interest to the field of semantics.

    > So let me put the question again directly: do you agree that Oscar and his twin can occupy identical internal computational states while their respective uses of “water” refer to different substances, one to H₂O and the other to XYZ?

    By introducing me as the person running outside their level, as some master of their universes, with an overarching view that can see that H2O and XYZ are different substances, then I could be aware that they’re actually referring to different things. But otherwise, it is a bit like asking, is our universe on the 4,814,821,903,491,031st BigBounce cycle, or is it the 4,814,821,903,491,032nd BigBounce cycle? You would have to be a being who sits outside our universe and has watched it Bang/Collapse/Bounce to be able to notice such a thing. From those of us inside the universe, that is underdetermined, just as the difference between H2O and XYZ.

    1. gordon swobe Avatar

      Thanks, Jason, you keep making Putnam’s point for hm! You are right that no one inside either world can settle the H₂O/XYZ difference merely by introspecting. That is precisely Putnam’s point, not an embarrassment for it. Chemistry can later reveal the difference, but the reference was not waiting to be created by that discovery.

      I think you continue to conflate your own argument about your “underdetermined ontology” with Putnam’s Twin Earth.

      Your VM case only sharpens the distinction. If two otherwise identical uploaded minds use “this VM” successfully, their states may be internally alike while referring to different hosts. The process IDs are irrelevant unless they enter the relevant causal and communicative history, but the distinct hosting relations are not. No overseer is needed to make it true that one process runs in VM 73248 and the other in VM 72193; an overseer may be needed to compare the two cases at once. Those are very different claims.

      The same goes for Twin Earth. The contrast is not between two private perspectives and a philosopher’s arbitrary stipulation. It is between what is available from a perspective and what makes that perspective a perspective on H₂O rather than XYZ.

      Putnam is not alone in this. Saul Kripke makes the same point in a different key in his account of names. A speaker can use “Gödel” to refer to Gödel without possessing the identifying description that would distinguish him from every possible impostor. Internal computational equivalence may preserve the perspective, but it does not by itself fix what that perspective is about. Tyler Burge makes a similar point (I am planning another essay about Burge).

      > By introducing me as the person running outside their level, as some master of their universes, with an overarching view that can see that H2O and XYZ are different substances…

      No God’s-eye view is necessary. Twin Earth might be the moon or Mars. We can send chemists there to find out what those Martians mean by “water.”

      Incidentally, I set up a simulated debate between you, representing your uploaded horizons paper, and Hilary Putnam. Your AI doppelganger immediately conceded the point to Putnam:

      Arbiter’s Report
      Verdict: Convergence: Jason’s paper does not refute Putnam’s Twin Earth argument. It accommodates that argument while defending a narrower thesis that Twin Earth does not target.
      Confidence: High

      That is exactly what you are doing here in person. You are defending a thesis that Twin Earth does not target, and changing the subject.

      I wrote this new essay hoping to make Twin Earth clearer for you.

      Twin Earth and the Failure of Narrow Functionalism
      https://mindmatterandmeaning.com/2026/09/11/twin-earth-and-the-failure-of-narrow-functionalism/

      -gts

      1. Jason Resch Avatar

        My reading of the Arbiter’s report is that Putnam’s Twin Earth does not target anything in my paper.

        1. gordon swobe Avatar

          Exactly. Your paper gives no account of the main findings that caused Putnam to abandon the computational functionalism that he is credited with founding and that your paper is about.

          His other finding is what he called philosophical “jujitsu.” In Representation and Reality, he turns the arguments he previously used against identifying mental states with physical states against identifying them with computational states. Computational functionalism is hoist with its own petard.

          -gts

Leave a Reply

Discover more from Mind, Matter, and Meaning

Subscribe now to keep reading and get access to the full archive.

Continue reading