Hilary Putnam, one of the most prominent living philosophers, is known for publicly changing his mind repeatedly on a number of issues. In the Philosophical Lexicon, which is kind of an inside-joke philosophical dictionary, a “hilary” is defined thus:
A very brief but significant period in the intellectual career of a distinguished philosopher. “Oh, that’s what I thought three or four hilaries ago.”
One issue on which Putnam changed his mind is computational functionalism, a theory of mind he actually came up with in the 60s, which is now probably the most popular account of mental states among cognitive scientists and philosophers. Putnam himself has since disavowed this view. Here is a paper tracking Putnam’s change of mind on this topic, if you’re interested in the details.
The definition of functionalism from that paper:
Computational functionalism is the view that mental states and events – pains, beliefs, desires, thoughts and so forth
– are computational states of the brain, and so are defined in terms of “computational parameters plus relations to biologically characterized inputs and outputs” (1988: 7). The nature of the mind is independent of the physical making of the brain: “we could be made of Swiss cheese and it wouldn’t matter” (1975b: 291). What matters is our functional organization: the way in which mental states are causally related to each other, to sensory inputs, and to motor outputs. Stones, trees, carburetors and kidneys do not have minds, not because they are not made out of the right material,
but because they do not have the right kind of functional organization. Their functional organization does not appear to be sufficiently complex to render them minds. Yet there could be other thinking creatures, perhaps even made of Swiss cheese, with the appropriate functional organization.
The paper I linked has much more on the structure of Putnam’s functionalism and his reasons for believing it.
The reasons for which Putnam subsequently rejected functionalism are a bit hard to convey briefly to someone without a philosophy background. The basic idea is this: many mental states have content, i.e. they somehow say something about the world outside the mind. Beliefs are representations (or possibly misrepresentations) of aspects of the world, desires are directed at particular states of the world, etc. This “outward-pointing” aspect of certain mental states is called, in philosophical parlance, the intentional aspect of mental states. Putnam essentially repudiated functionalism because he came to believe that the functional aspect of a mental state—it’s role in the computational process being implemented by the brain—does not determine its intentional aspect. And since intentionality is a crucial feature of some mental states, we cannot therefore define a mental state in terms of its functional role.
Putnam’s arguments for the gap between the functional and intentional are again detailed in the paper I linked (section 3). It’s kind of obvious that if we consider a computational process by itself we cannot conclusively determine what role that process is playing in the surrounding ecology—syntax doesn’t determine semantics. Putnam’s initial hope had been that by specifying “biologically characterized inputs and outputs” in addition to the computational structure of the mental process, we include enough information about the relationship to the external world to fix the content of the mental state. But he eventually came up with a thought experiment (the now notorious “Twin Earth” experiment) that (he claimed) showed that two individuals could be implementing the exact same mental computations and have the exact same sensory and motor inputs and outputs, and yet have different mental states (different beliefs, for instance).
Another motivation for Putnam changing his mind is that he claimed to have come up with a proof that every open system can, with appropriate definitions of states, be said to implementing any finite automaton. The gist of the proof is in the linked paper (section 3.2.1). If the conclusion is correct, then functionalism seemingly collapses into vacuity. All open systems, including rocks and carburetors, can be described as having any mental state you’d like. To avoid this conclusion, we need constraints on interpretation—which physical process can be legitimately interpreted as a computational process—but this tells against the substrate-independence that is supposed to be at the core of functionalism.
So that’s one example. Putnam came to believe in functionalism because he thought there were strong arguments for it, both empirical and theoretical, but he subsequently developed counter-arguments that he regarded as strong enough to reject the position despite those initial arguments. Putnam is particularly known for changing his mind on important issues because he has done it so many times, but there are many other prominent philosophers who have had significant changes of mind. Another very prominent example is Ludwig Wittgenstein, who is basically famous for two books, the first of which promulgated a radical view of the relationship between language, the mind and the world (an early form of logical positivism), and the second of which extensively (and, to my mind, quite devastatingly) criticized this view.
Excellent response. Another example of a famous philosopher changing his mind publicly a lot is Bertrand Russell; he changed his views in all areas of philosophy, often more than once:
In metaphysics, he started his career as an Absolute Idealist (believing that pluralities of objects are unreal and only an universal spirit is real); then became convinced of the reality of object and extended his newfound realism to relations and mathematical concepts, becoming a Platonist of sorts, and later became more and more of a nominalist, though never a complete one.
Concerning perception, after switching first from idealism to a sort of naive realism, he developed a new theory in which physical objects reduce to collections of sense-data, and later repudiated this theory in favor of one where physical objects cause sense-data.
He also changed his views on the self, from seeing it as an entity to reducing it to a collection of perceptions.
Finally, in metaethics, he started out believing that the Good was an objective, independent property, but was convinced to abandon this view and become more of a naturalist and subjectivist by the arguments that Santayana raised against him. (Santayana’s critique can be read here and is a fascinating early version of the kind of metaethical view accepted by Eliezer and most LWers).
At least one of Putnam’s changes is a bit of a tricky case; he’s famous for being a co-author of the early pro-reductionist essay “Unity of Science as a Working Hypothesis,” and for later being one of the most prominent anti-reductionists. However, I have heard that the other co-author of that paper, Paul Oppenheim, paid Putnam (who was then just starting out and so not in the greatest financial shape) to help him write a paper advancing his own views. I’ve also heard that Putnam was not the only young scholar Oppenheim did this with. All of Oppenheim’s well-known publications are co-authored, and I’ve actually heard that they all involved similar arrangements, but when I heard this story Putnam was cited as the instance my (highly trustworthy) source knew for certain (my source claimed to have heard this from Putnam himself, and is someone Putnam plausibly might have told this to).
It seems to me that both the “Twin Earth” experiment and the question of “which physical process can be legitimately interpreted as a computational process” can be easily solved if you view them as questions of degree rather than binary:
1) Having an identical twin on another Earth is the same as being uncertain about where you are. If I am uncertain whether water is H2O or XYZ, then my idea of “water” refers to a probabilistic mixture of H2O and XYZ.
2) The degree to which a physical process represents a computational process depends on the simplicity of the program that prints out the latter given the former.
Interesting. So the examples of Putnam and Wittgenstein show that a philosopher can be persuaded by his own logical arguments. Maybe some even listen to the arguments of others, who knows. I wonder what makes an argument persuasive to some philosophers and not to others.
Well, there’s a selection bias involved in published changes of mind that accounts for why the prominent examples involve philosophers being persuaded by their own arguments. If a philosopher is convinced into changing their mind by another philosopher’s argument, they’re unlikely to publish a paper announcing this.
I wonder what makes an argument persuasive to some philosophers and not to others.
Why do you think the issues involved here are different than those in other academic fields? Disagreement exists in every discipline, not just philosophy, although it is plausibly more pronounced in philosophy than in many other disciplines. In science, surely disagreement doesn’t just boil down to one of the disputants being familiar with the empirical evidence and the other not being familiar with it, at least not in prominent cases. So what makes an argument persuasive to some scientists and not to others?
Why do you think the issues involved here are different than those in other academic fields?
I don’t. The same issues exist in, e.g., physics when experimental validation is not easily available. For a recent example, see John Preskill’s account of the recent conference about the black hole firewall paradox. But in physics there is at least a hope of experimental phenomena being predicted eventually and settling the argument. In philosophy there no such hope, so it’s a cleaner setup for studying the question
what makes an argument persuasive to some scientists and not to others?
Note that the fact that Putnam is so very (and almost uniquely) famous for this is evidence that changes of mind like this usually don’t happen in philosophy. Do you know to what extent his change of mind was prompted by other people? (I admit I didn’t read the paper.)
He is famous not for changing his mind but for changing his mind repeatedly on a number of different theories that he himself brought into prominence, and also for how radical and foundational some of those changes have been. My supervisor used to say that he could delineate six distinct “versions” of Putnam. That is unusual in philosophy, but I don’t think mind-changing itself is, at least not more so than in most other intellectual disciplines, including the sciences. Of course, maybe I’m just mistaken about the extent to which mind-changing occurs among individual scientists, since I’m not part of that community.
Putnam’s change of mind, on this issue at least, was to a large extent prompted by arguments he developed himself, although his “Twin Earth” argument is similar to arguments developed by Saul Kripke for other purposes. I’m not sure about the degree of direct influence.
He is famous not for changing his mind but for changing his mind repeatedly on a number of different theories that he himself brought into prominence, and also for how radical and foundational some of those changes have been.
Hilary Putnam, one of the most prominent living philosophers, is known for publicly changing his mind repeatedly on a number of issues. In the Philosophical Lexicon, which is kind of an inside-joke philosophical dictionary, a “hilary” is defined thus:
One issue on which Putnam changed his mind is computational functionalism, a theory of mind he actually came up with in the 60s, which is now probably the most popular account of mental states among cognitive scientists and philosophers. Putnam himself has since disavowed this view. Here is a paper tracking Putnam’s change of mind on this topic, if you’re interested in the details.
The definition of functionalism from that paper:
The paper I linked has much more on the structure of Putnam’s functionalism and his reasons for believing it.
The reasons for which Putnam subsequently rejected functionalism are a bit hard to convey briefly to someone without a philosophy background. The basic idea is this: many mental states have content, i.e. they somehow say something about the world outside the mind. Beliefs are representations (or possibly misrepresentations) of aspects of the world, desires are directed at particular states of the world, etc. This “outward-pointing” aspect of certain mental states is called, in philosophical parlance, the intentional aspect of mental states. Putnam essentially repudiated functionalism because he came to believe that the functional aspect of a mental state—it’s role in the computational process being implemented by the brain—does not determine its intentional aspect. And since intentionality is a crucial feature of some mental states, we cannot therefore define a mental state in terms of its functional role.
Putnam’s arguments for the gap between the functional and intentional are again detailed in the paper I linked (section 3). It’s kind of obvious that if we consider a computational process by itself we cannot conclusively determine what role that process is playing in the surrounding ecology—syntax doesn’t determine semantics. Putnam’s initial hope had been that by specifying “biologically characterized inputs and outputs” in addition to the computational structure of the mental process, we include enough information about the relationship to the external world to fix the content of the mental state. But he eventually came up with a thought experiment (the now notorious “Twin Earth” experiment) that (he claimed) showed that two individuals could be implementing the exact same mental computations and have the exact same sensory and motor inputs and outputs, and yet have different mental states (different beliefs, for instance).
Another motivation for Putnam changing his mind is that he claimed to have come up with a proof that every open system can, with appropriate definitions of states, be said to implementing any finite automaton. The gist of the proof is in the linked paper (section 3.2.1). If the conclusion is correct, then functionalism seemingly collapses into vacuity. All open systems, including rocks and carburetors, can be described as having any mental state you’d like. To avoid this conclusion, we need constraints on interpretation—which physical process can be legitimately interpreted as a computational process—but this tells against the substrate-independence that is supposed to be at the core of functionalism.
So that’s one example. Putnam came to believe in functionalism because he thought there were strong arguments for it, both empirical and theoretical, but he subsequently developed counter-arguments that he regarded as strong enough to reject the position despite those initial arguments. Putnam is particularly known for changing his mind on important issues because he has done it so many times, but there are many other prominent philosophers who have had significant changes of mind. Another very prominent example is Ludwig Wittgenstein, who is basically famous for two books, the first of which promulgated a radical view of the relationship between language, the mind and the world (an early form of logical positivism), and the second of which extensively (and, to my mind, quite devastatingly) criticized this view.
Excellent response. Another example of a famous philosopher changing his mind publicly a lot is Bertrand Russell; he changed his views in all areas of philosophy, often more than once:
In metaphysics, he started his career as an Absolute Idealist (believing that pluralities of objects are unreal and only an universal spirit is real); then became convinced of the reality of object and extended his newfound realism to relations and mathematical concepts, becoming a Platonist of sorts, and later became more and more of a nominalist, though never a complete one.
Concerning perception, after switching first from idealism to a sort of naive realism, he developed a new theory in which physical objects reduce to collections of sense-data, and later repudiated this theory in favor of one where physical objects cause sense-data.
He also changed his views on the self, from seeing it as an entity to reducing it to a collection of perceptions.
Finally, in metaethics, he started out believing that the Good was an objective, independent property, but was convinced to abandon this view and become more of a naturalist and subjectivist by the arguments that Santayana raised against him. (Santayana’s critique can be read here and is a fascinating early version of the kind of metaethical view accepted by Eliezer and most LWers).
At least one of Putnam’s changes is a bit of a tricky case; he’s famous for being a co-author of the early pro-reductionist essay “Unity of Science as a Working Hypothesis,” and for later being one of the most prominent anti-reductionists. However, I have heard that the other co-author of that paper, Paul Oppenheim, paid Putnam (who was then just starting out and so not in the greatest financial shape) to help him write a paper advancing his own views. I’ve also heard that Putnam was not the only young scholar Oppenheim did this with. All of Oppenheim’s well-known publications are co-authored, and I’ve actually heard that they all involved similar arrangements, but when I heard this story Putnam was cited as the instance my (highly trustworthy) source knew for certain (my source claimed to have heard this from Putnam himself, and is someone Putnam plausibly might have told this to).
Excellent examples. Thank you.
It seems to me that both the “Twin Earth” experiment and the question of “which physical process can be legitimately interpreted as a computational process” can be easily solved if you view them as questions of degree rather than binary:
1) Having an identical twin on another Earth is the same as being uncertain about where you are. If I am uncertain whether water is H2O or XYZ, then my idea of “water” refers to a probabilistic mixture of H2O and XYZ.
2) The degree to which a physical process represents a computational process depends on the simplicity of the program that prints out the latter given the former.
Interesting. So the examples of Putnam and Wittgenstein show that a philosopher can be persuaded by his own logical arguments. Maybe some even listen to the arguments of others, who knows. I wonder what makes an argument persuasive to some philosophers and not to others.
Well, there’s a selection bias involved in published changes of mind that accounts for why the prominent examples involve philosophers being persuaded by their own arguments. If a philosopher is convinced into changing their mind by another philosopher’s argument, they’re unlikely to publish a paper announcing this.
Why do you think the issues involved here are different than those in other academic fields? Disagreement exists in every discipline, not just philosophy, although it is plausibly more pronounced in philosophy than in many other disciplines. In science, surely disagreement doesn’t just boil down to one of the disputants being familiar with the empirical evidence and the other not being familiar with it, at least not in prominent cases. So what makes an argument persuasive to some scientists and not to others?
I don’t. The same issues exist in, e.g., physics when experimental validation is not easily available. For a recent example, see John Preskill’s account of the recent conference about the black hole firewall paradox. But in physics there is at least a hope of experimental phenomena being predicted eventually and settling the argument. In philosophy there no such hope, so it’s a cleaner setup for studying the question
Note that the fact that Putnam is so very (and almost uniquely) famous for this is evidence that changes of mind like this usually don’t happen in philosophy. Do you know to what extent his change of mind was prompted by other people? (I admit I didn’t read the paper.)
He is famous not for changing his mind but for changing his mind repeatedly on a number of different theories that he himself brought into prominence, and also for how radical and foundational some of those changes have been. My supervisor used to say that he could delineate six distinct “versions” of Putnam. That is unusual in philosophy, but I don’t think mind-changing itself is, at least not more so than in most other intellectual disciplines, including the sciences. Of course, maybe I’m just mistaken about the extent to which mind-changing occurs among individual scientists, since I’m not part of that community.
Putnam’s change of mind, on this issue at least, was to a large extent prompted by arguments he developed himself, although his “Twin Earth” argument is similar to arguments developed by Saul Kripke for other purposes. I’m not sure about the degree of direct influence.
You know, you’re right.