Sorry for the late comment, but had some thoughts I wanted to share. To start with, I take psychophysical harmony to be about the harmony between phenomenal states and psychological states, not between phenomenal states and behavior. What explains the harmony between psychological states and behavior are the laws of physics and biology. So those world where people smelt strawberries whenever they stubbed their toes and continued stubbing their toes are actually psychophysically (a bit of a misnomer yeah) harmonious worlds, since the phenomenal (good smell) is aligned with the psychological (desire to engage in action which elicits good smell). But notice that in such worlds such people are at an evolutionary disadvantage, since obviously toe stubbing behavior is not good for organisms. So it’s no surprise that psychological states and behavior are well aligned in our universe, as we would expect them to be well aligned in any universe with natural selection pressures regardless of the psychophysical laws.
Cutter and co. will often abbreviate the argument as being about the alignment between behavior and phenomenology, because it’s assumed that psychological states will always be well aligned with behavior, as this is explained by evolution. But really they mean that it’s a coincidence that our psychology is well aligned with our phenomenology.
So when you say that the psychological states in D1 don’t explain the actions of said people, and hence disharmonious worlds are actually incredibly coincidental, I think this is a misunderstanding. The people in D1 should have the exact same psychological and neural profiles as ordinary people, it’s only their phenomenology which is weird.
Also, it’s the descriptive components of D1 and H1 people’s psychological states which explains their behavior, not their normative affects. For example, we wouldn’t say that the actions of a person who saved a puppy was explained by the fact that saving puppies was good, or that actions of a person who drowned a puppy was explained by the fact that saving puppies is bad etc…
Rather, such actions are explained by descriptive facts about the person (e.g. they believed that drowning puppies is good, they are disposed to sadism etc…) irrespective of the normative facts. Hence there’s no reason to assume that D1 actions are unexplained; we would presumably explain their actions the same way we would explain the actions of people in our universe. That they believed that receiving a ticket should be met with a verbal response and so forth (or if you think beliefs are grounded in phenomenology, that they were disposed to give a verbal response). It’s just that their phenomenology associated with their beliefs and actions is weird. But their beliefs (grounded in their dispositions or cognitive architectures) are still the same as ours.
Now this gets a bit tricky if you think that cognition is grounded in phenomenology. Then you would technically have to say that people in disharmonious worlds don’t believe the same things as us, but a puzzle would still remain as to why our beliefs are correct (and why they are appropriately aligned with our psychological dispositions), whereas they are incorrect in disharmonious worlds. And even if that were the case, it wouldn’t be coincidental that D1 people have their psychology aligned with their behavior, since again their behavior is explained by their dispositions and neural makeup, which is the same as ours.
Great reply, thank you for the comments! I agree that Cutter and Crummett and I are both abbreviating the link between phenomenal states and functional/behavioral states a bit, when in reality there’s a middle step involving psychological states where the impact of disharmony is unclear. But I’m not sure it matters, because I’m very skeptical it’s even possible for agents to have identical cognitive processes (and thus identical behavior) with radically distinct phenomenal states unless you accept a sort of epiphenomenalism that we have independent reasons to reject. You said, for example, that the person receiving the ticket believed a verbal response was warranted. But how could that belief form in response to a phenomenal experience that doesn’t carry any information about receiving the ticket? If phenomenal states actually play a role in our cognition in some way – which seems necessary to avoid epiphenomenalism – then they must be what agents “rely on” for at least some cognitive processes. And if that’s the case, then it does seem like only lucky coincidence could explain why sometimes those cognitive processes (and the behavior that results) hit their mark despite not having access to what would normally provide that target. And if the argument does require epiphenomenalism or overdetermined dualism, then IMO the whole thing just becomes a good argument against those two views.
(Also, another issue arises if you do think cognition can be entirely separate from phenomenology, which is that then it becomes impossible for us to know if we're even harmonious at all - the disharmonious agents also believe they have harmonious states, right?)
Things definitely become a little more complicated on the interactionist kind of dualism, but the basic argument still goes through I think. One key operating assumption here that should be pointed out (and that Cutter and Crummett are definitely adopting) is that phenomenal properties are supposed to be intrinsic/categorical as opposed to purely structural/dispositional. So what that means is that on interactionist dualism, you can have a phenomenal state A which plays some causal role in the brain, and another phenomenal state B which plays the exact same causal role. Crucially, these two phenomenal states can be different, even though they share the same causal role, and have the same dispositional tendencies, so long as their intrinsic properties are different. It’s their differences in their intrinsic properties which makes a difference to their phenomenology, even though the “outside” causal structure is essentially the same.
If you buy into this metaphysical picture (which I personally do not) then it’s easier to see how the idea of psychophysical harmony can be made to work for the interactionist. The interactionist would say that the causal “information” is the same for the different phenomenal states, and it is in virtue of this that the physical behavior is explained in both cases.
Another assumption worth pointing out that I think Cutter and co. are making is something akin to the notion that certain phenomenal states are essentially representational. E.g. a visual experience of a round object is essentially about a round object, and not a triangle or a pain in your foot. If this was not the case, then it makes little sense to suppose that there’s any harmony between our perceptual psychological and phenomenal states, or between our epistemic brain states and the cognitive phenomenology accompanying it. Somehow, it’s just supposed to be obvious that “round object” phenomenology is about “round object” in the environment, irrespective of the functional role that the round object phenomenal state plays. I personally think this assumption is somewhat dubious too (though I agree it’s more intuitive for primary qualities. But for secondary qualities like color- it seems to me quite doubtful that, say, the color blue is essentially about a certain EM wavelength, irrespective of its functional
role).
If you deny representationalism the problem of psychophysical harmony is greatly ameliorated, although it’s still an issue with respect to hedonic harmony, since presumably we all agree that pain is bad and pleasure is good, irrespective of the functionality of these phenomenal states.
Lastly, regarding your point about cognition being separate from phenomenology- yeah that’s a well known problem going back to Chalmers which is termed “the paradox of phenomenal judgement” in the literature. I don’t think it’s applicable here since the authors think that cognition is grounded in phenomenology at least to some extent (which they in fact use to construct an additional argument for the harmony of our epistemic beliefs).
Yeah, I’m very sympathetic to illusionism myself and arguments like this ultimately just lead me to think the basic concepts being utilized are incoherent. But even if I put myself in the mindset of someone who is a more mainstream qualia theorist, I struggle to understand how the experience of hearing "I would like a ticket for X" leads someone to form a desire to print that ticket without the information necessary for generating that desire being communicated through the hearing itself. If I could have a completely distinct experience of hearing something else entirely and it would still lead me to print the ticket, then it just seems to me like what I'm hearing isn't actually doing anything, right? It might be that “the presence” of those sounds causes me to print the ticket, due to some arbitrary (or divinely guided!) law, but the sounds themselves aren't transferring any information; somehow, the law itself is structured to "get that information across" without any phenomenal aspect. But if that’s the position they’re taking – which I think you’re right that it is – then I think my general point still stands: A world in which those laws are individual instances of a broader normative harmony are simpler and contain fewer coincidences than disharmonious laws, which would have to be specified arbitrarily in every case.
In order to hold that it’s inconceivable for a cause to have different effects at different times, are you committed to a principle that says that all changes in an object’s causal profile must be explained by some feature of the object?
I'm not sure what you mean by some feature of the object - I definitely accept that a cause can have different effects at different times, but the difference in effect must be in virtue of *something.* And I'm very skeptical that you could actually lay out a coherent framework to explain why (using the example I gave) the smell of freshly baked bread would be able to cause every single human action. The only plausible grounding would be other facts about the world (time of day, temperature, whatever) and I don't think those are sufficiently fine-grained enough to account for all human mentation.
Interesting post. It reminds me of a possible response to the Fine-Tuning Argument from Order/Coherence; why does our world look orderly/coherent instead of being a chatoic mess? One of the responses offered to such an argument is that the existence of any physical universe is going to have order as a necessary state of affairs as C.M. Lorkowski points out:
"First, it helps to consider whether there is a viable alternative to an ordered universe, presumably a universe with no order, that is, a completely chaotic universe. But while we can talk casually about such a notion, many have argued that the concept of a completely chaotic universe is incoherent. This is because any coherent picture requires kinds to serve as sortals, at minimum, the most basic kinds such as “matter” and “space.” But to have any such kind requires that there are rules, definitions, natures, essences, etc., something that makes it a member of that kind rather than something else and thereby allows us to conceptually distinguish that kind. This is true of even the broadest kinds such as “matter.” A truly chaotic universe would therefore have to be a universe without any kinds, but such a state is inconceivable. A universe could certainly have a different order, perhaps even significantly less order, but it is far from clear that it could have no order at all. [10] The very existence of the universe seems to entail some order, at least enough to have rules governing basic kinds, [11] but once we have an ordered universe, it is not clear that there is anything left to explain, save perhaps how it is ordered." (Atheism Considered, pg. 81)
I think in a similar sort of way, your argument points out that across the modal space, even disharmonious worlds are going to contain the same datum in terms of the "lucky conincidences" that the original psychophysical harmony tries to highlight as datum pointing to Theism, thus it seems the relevant datum here is just going to be a function of the relevant modal space, just as "order" is a necessary function of any possible world that contains a physical universe.
“the fact that he prints the ticket can’t possibly be explained by the fact that he felt excruciating pain”
Him printing the tickets is explained by the conjunction of his pain and the psychophysical laws. I think it would be wrong to infer that this situation is inconceivable. I can conceive it just fine if I imagine the psychophysical laws are different. Your suggestion seems to be that most sets of psychophysical laws are inconceivable because the psychological state cannot explain the action? That seems wrong because it is always the conjunction of the state and the laws that explains the action.
To illustrate, the mass of a large object explains why objects fall towards it only in conjunction with facts about the law of gravity. In a hypothetical situation where objects are repelled away from the large object, this can be explained by the mass of the object and the reversed law of gravity. The reversed law of gravity is not inconceivable.
The difference, imo, is that there is no normatively appropriate role for gravity, whereas there are islands of normative harmony that *do* have a privileged explanatory status. So if there was some sort of privileged way that gravity "ought to be," such that it made optimal sense as an explanation for how objects move, then I think it would be reasonable to say gravity causing other sorts of movements would be plausibly inconceivable. But that is itself a hard situation to conceive of!
There seems to be a set of possible worlds that are disharmonious but not rational. For example, the experience of random static for a second causes agent x to yell, and the experience of a blue piano causes agent x to say “I smell coffee”. Surely this represents the vast majority of possible worlds, so it’s still fortunate that we’re not in these worlds, so the original argument can still work.
Tally of imaginable wacky worlds means nothing in terms of fortune without weighting their likelihoods to arise. Imaginability does not entail even shots at accessibility.
Hey Both Sides,
Sorry for the late comment, but had some thoughts I wanted to share. To start with, I take psychophysical harmony to be about the harmony between phenomenal states and psychological states, not between phenomenal states and behavior. What explains the harmony between psychological states and behavior are the laws of physics and biology. So those world where people smelt strawberries whenever they stubbed their toes and continued stubbing their toes are actually psychophysically (a bit of a misnomer yeah) harmonious worlds, since the phenomenal (good smell) is aligned with the psychological (desire to engage in action which elicits good smell). But notice that in such worlds such people are at an evolutionary disadvantage, since obviously toe stubbing behavior is not good for organisms. So it’s no surprise that psychological states and behavior are well aligned in our universe, as we would expect them to be well aligned in any universe with natural selection pressures regardless of the psychophysical laws.
Cutter and co. will often abbreviate the argument as being about the alignment between behavior and phenomenology, because it’s assumed that psychological states will always be well aligned with behavior, as this is explained by evolution. But really they mean that it’s a coincidence that our psychology is well aligned with our phenomenology.
So when you say that the psychological states in D1 don’t explain the actions of said people, and hence disharmonious worlds are actually incredibly coincidental, I think this is a misunderstanding. The people in D1 should have the exact same psychological and neural profiles as ordinary people, it’s only their phenomenology which is weird.
Also, it’s the descriptive components of D1 and H1 people’s psychological states which explains their behavior, not their normative affects. For example, we wouldn’t say that the actions of a person who saved a puppy was explained by the fact that saving puppies was good, or that actions of a person who drowned a puppy was explained by the fact that saving puppies is bad etc…
Rather, such actions are explained by descriptive facts about the person (e.g. they believed that drowning puppies is good, they are disposed to sadism etc…) irrespective of the normative facts. Hence there’s no reason to assume that D1 actions are unexplained; we would presumably explain their actions the same way we would explain the actions of people in our universe. That they believed that receiving a ticket should be met with a verbal response and so forth (or if you think beliefs are grounded in phenomenology, that they were disposed to give a verbal response). It’s just that their phenomenology associated with their beliefs and actions is weird. But their beliefs (grounded in their dispositions or cognitive architectures) are still the same as ours.
Now this gets a bit tricky if you think that cognition is grounded in phenomenology. Then you would technically have to say that people in disharmonious worlds don’t believe the same things as us, but a puzzle would still remain as to why our beliefs are correct (and why they are appropriately aligned with our psychological dispositions), whereas they are incorrect in disharmonious worlds. And even if that were the case, it wouldn’t be coincidental that D1 people have their psychology aligned with their behavior, since again their behavior is explained by their dispositions and neural makeup, which is the same as ours.
Great reply, thank you for the comments! I agree that Cutter and Crummett and I are both abbreviating the link between phenomenal states and functional/behavioral states a bit, when in reality there’s a middle step involving psychological states where the impact of disharmony is unclear. But I’m not sure it matters, because I’m very skeptical it’s even possible for agents to have identical cognitive processes (and thus identical behavior) with radically distinct phenomenal states unless you accept a sort of epiphenomenalism that we have independent reasons to reject. You said, for example, that the person receiving the ticket believed a verbal response was warranted. But how could that belief form in response to a phenomenal experience that doesn’t carry any information about receiving the ticket? If phenomenal states actually play a role in our cognition in some way – which seems necessary to avoid epiphenomenalism – then they must be what agents “rely on” for at least some cognitive processes. And if that’s the case, then it does seem like only lucky coincidence could explain why sometimes those cognitive processes (and the behavior that results) hit their mark despite not having access to what would normally provide that target. And if the argument does require epiphenomenalism or overdetermined dualism, then IMO the whole thing just becomes a good argument against those two views.
(Also, another issue arises if you do think cognition can be entirely separate from phenomenology, which is that then it becomes impossible for us to know if we're even harmonious at all - the disharmonious agents also believe they have harmonious states, right?)
Things definitely become a little more complicated on the interactionist kind of dualism, but the basic argument still goes through I think. One key operating assumption here that should be pointed out (and that Cutter and Crummett are definitely adopting) is that phenomenal properties are supposed to be intrinsic/categorical as opposed to purely structural/dispositional. So what that means is that on interactionist dualism, you can have a phenomenal state A which plays some causal role in the brain, and another phenomenal state B which plays the exact same causal role. Crucially, these two phenomenal states can be different, even though they share the same causal role, and have the same dispositional tendencies, so long as their intrinsic properties are different. It’s their differences in their intrinsic properties which makes a difference to their phenomenology, even though the “outside” causal structure is essentially the same.
If you buy into this metaphysical picture (which I personally do not) then it’s easier to see how the idea of psychophysical harmony can be made to work for the interactionist. The interactionist would say that the causal “information” is the same for the different phenomenal states, and it is in virtue of this that the physical behavior is explained in both cases.
Another assumption worth pointing out that I think Cutter and co. are making is something akin to the notion that certain phenomenal states are essentially representational. E.g. a visual experience of a round object is essentially about a round object, and not a triangle or a pain in your foot. If this was not the case, then it makes little sense to suppose that there’s any harmony between our perceptual psychological and phenomenal states, or between our epistemic brain states and the cognitive phenomenology accompanying it. Somehow, it’s just supposed to be obvious that “round object” phenomenology is about “round object” in the environment, irrespective of the functional role that the round object phenomenal state plays. I personally think this assumption is somewhat dubious too (though I agree it’s more intuitive for primary qualities. But for secondary qualities like color- it seems to me quite doubtful that, say, the color blue is essentially about a certain EM wavelength, irrespective of its functional
role).
If you deny representationalism the problem of psychophysical harmony is greatly ameliorated, although it’s still an issue with respect to hedonic harmony, since presumably we all agree that pain is bad and pleasure is good, irrespective of the functionality of these phenomenal states.
Lastly, regarding your point about cognition being separate from phenomenology- yeah that’s a well known problem going back to Chalmers which is termed “the paradox of phenomenal judgement” in the literature. I don’t think it’s applicable here since the authors think that cognition is grounded in phenomenology at least to some extent (which they in fact use to construct an additional argument for the harmony of our epistemic beliefs).
Yeah, I’m very sympathetic to illusionism myself and arguments like this ultimately just lead me to think the basic concepts being utilized are incoherent. But even if I put myself in the mindset of someone who is a more mainstream qualia theorist, I struggle to understand how the experience of hearing "I would like a ticket for X" leads someone to form a desire to print that ticket without the information necessary for generating that desire being communicated through the hearing itself. If I could have a completely distinct experience of hearing something else entirely and it would still lead me to print the ticket, then it just seems to me like what I'm hearing isn't actually doing anything, right? It might be that “the presence” of those sounds causes me to print the ticket, due to some arbitrary (or divinely guided!) law, but the sounds themselves aren't transferring any information; somehow, the law itself is structured to "get that information across" without any phenomenal aspect. But if that’s the position they’re taking – which I think you’re right that it is – then I think my general point still stands: A world in which those laws are individual instances of a broader normative harmony are simpler and contain fewer coincidences than disharmonious laws, which would have to be specified arbitrarily in every case.
Very nicely put. Imaginability does not mean accessibility, and sometimes accessibility is the material issue, not a side issue.
A quick vid from a couple years back when this was first going viral:
https://twitter.com/StanRockPatton/status/1567287203616935936
In order to hold that it’s inconceivable for a cause to have different effects at different times, are you committed to a principle that says that all changes in an object’s causal profile must be explained by some feature of the object?
I'm not sure what you mean by some feature of the object - I definitely accept that a cause can have different effects at different times, but the difference in effect must be in virtue of *something.* And I'm very skeptical that you could actually lay out a coherent framework to explain why (using the example I gave) the smell of freshly baked bread would be able to cause every single human action. The only plausible grounding would be other facts about the world (time of day, temperature, whatever) and I don't think those are sufficiently fine-grained enough to account for all human mentation.
Interesting post. It reminds me of a possible response to the Fine-Tuning Argument from Order/Coherence; why does our world look orderly/coherent instead of being a chatoic mess? One of the responses offered to such an argument is that the existence of any physical universe is going to have order as a necessary state of affairs as C.M. Lorkowski points out:
"First, it helps to consider whether there is a viable alternative to an ordered universe, presumably a universe with no order, that is, a completely chaotic universe. But while we can talk casually about such a notion, many have argued that the concept of a completely chaotic universe is incoherent. This is because any coherent picture requires kinds to serve as sortals, at minimum, the most basic kinds such as “matter” and “space.” But to have any such kind requires that there are rules, definitions, natures, essences, etc., something that makes it a member of that kind rather than something else and thereby allows us to conceptually distinguish that kind. This is true of even the broadest kinds such as “matter.” A truly chaotic universe would therefore have to be a universe without any kinds, but such a state is inconceivable. A universe could certainly have a different order, perhaps even significantly less order, but it is far from clear that it could have no order at all. [10] The very existence of the universe seems to entail some order, at least enough to have rules governing basic kinds, [11] but once we have an ordered universe, it is not clear that there is anything left to explain, save perhaps how it is ordered." (Atheism Considered, pg. 81)
I think in a similar sort of way, your argument points out that across the modal space, even disharmonious worlds are going to contain the same datum in terms of the "lucky conincidences" that the original psychophysical harmony tries to highlight as datum pointing to Theism, thus it seems the relevant datum here is just going to be a function of the relevant modal space, just as "order" is a necessary function of any possible world that contains a physical universe.
Really interesting connection, thank you for making it! I've had similar thoughts in the past but this expresses the intuition very well.
“the fact that he prints the ticket can’t possibly be explained by the fact that he felt excruciating pain”
Him printing the tickets is explained by the conjunction of his pain and the psychophysical laws. I think it would be wrong to infer that this situation is inconceivable. I can conceive it just fine if I imagine the psychophysical laws are different. Your suggestion seems to be that most sets of psychophysical laws are inconceivable because the psychological state cannot explain the action? That seems wrong because it is always the conjunction of the state and the laws that explains the action.
To illustrate, the mass of a large object explains why objects fall towards it only in conjunction with facts about the law of gravity. In a hypothetical situation where objects are repelled away from the large object, this can be explained by the mass of the object and the reversed law of gravity. The reversed law of gravity is not inconceivable.
The difference, imo, is that there is no normatively appropriate role for gravity, whereas there are islands of normative harmony that *do* have a privileged explanatory status. So if there was some sort of privileged way that gravity "ought to be," such that it made optimal sense as an explanation for how objects move, then I think it would be reasonable to say gravity causing other sorts of movements would be plausibly inconceivable. But that is itself a hard situation to conceive of!
There seems to be a set of possible worlds that are disharmonious but not rational. For example, the experience of random static for a second causes agent x to yell, and the experience of a blue piano causes agent x to say “I smell coffee”. Surely this represents the vast majority of possible worlds, so it’s still fortunate that we’re not in these worlds, so the original argument can still work.
Tally of imaginable wacky worlds means nothing in terms of fortune without weighting their likelihoods to arise. Imaginability does not entail even shots at accessibility.