Propositional attitudes have an attitude type (belief, desire, etc.), and a content. A popular idea in the literature on intentionality is that attitude type is determined by functional role and content in some other way. One can find this view, for example, in Fodor (1987, 17), Dretske (1995, 83), or Loewer (2017, 716). I don't see how it could be correct.
Many experiences have phenomenal properties: there is something it is like to have them. A puzzling fact about these properties is that we appear to know about them in a special, direct fashion: we are "acquainted" with the phenomenal properties of our experiences. Another, related puzzle is that we appear to know about these properties with absolute certainty: if you have an experience as of looking at a red wall, you can conclusively rule out the possibility that you have an experience as of looking at a green wall.
In Schwarz (2018), I put forward a tentative explanation of these facts. I argued that it would be useful for an agent in a world like ours to have a credence function defined over a space that includes special "imaginary" propositions that are causally tied to stimulations of their sense organs in such a way that any given stimulation makes the agent certain of a corresponding imaginary proposition. What we conceptualise as propositions about phenomenal properties (of our experience), I argued, might be such imaginary propositions.
I've decided to write somewhat regular short pieces on interesting papers I've recently read. This one is about Smithies, Lennon, and Samuels (2022).
Smithies, Lennon, and Samuels (henceforth, SLS) criticise the view that there are a priori connections between having a belief with a certain content and other states that would be rational given this belief. A simple example of the target view says that believing P is being disposed to act in a way that would bring one closer to satisfying one's desires if P were true. A more complicated example of the target view, on which SLS focus, is Lewis's. According to Lewis, for a mental state to be a belief state with such-and-such content, the state must, under normal conditions, be connected in a certain way to behaviour, perceptual experiences, and other propositional attitudes. SLS deny this.
I wrote this short piece for a special issue of the Journal of Consciousness Studies on Chalmers's "The Meta-Problem of Consciousness" (2018). Much of my paper rehashes ideas from section 5 of my "Imaginary Foundations" paper, but here I try to present these ideas more simply and directly, without the Bayesian background.
Earlier this year, I read Tyler Burge's Origins of Objectivity. It's a very long book. Here is an abridged version. A few comments below.
Origins of Objectivity
Representation is a basic explanatory kind in psychology that should be distinguished from mere information-carrying. The most fundamental type of representational state is perception. In perception, an organism attributes properties to objects in its environment. To do this, the organism does not need linguistic capacities, nor does it need to know (or otherwise represent) necessary and sufficient conditions for being the relevant object. Instead, the science of perception reveals that it is sufficient that the organism stands in a suitable causal relation to the object and that its perceptual state involves certain constancies (for shape or colour or distance or whatever) which characterize the object "objectively", abstracting away from contingencies of the present stimulus.
I like the starting point — to think of intentional states as explanatory scientific kinds. Burge doesn't say what exactly he means by this. I would put it as a kind of functionalism: intentional states are characterized (at least in part) by their functional inter-connections and their relationship to environmental causes, behaviour and other psychologically relevant facts.
There's an exciting new theory in cognitive science. The theory began as an account of message-passing in the visual cortex, but it quickly expanded into a unified explanation of perception, action, attention, learning, homeostasis, and the very possibility of life. In its most general and ambitious form, the theory was mainly developed by Karl Friston -- see e.g. Friston 2006, Friston and Stephan 2007, Friston 2009, Friston 2010, or the Wikipedia page on the free-energy principle.
Stalnaker holds a combination of views that seem independent to me, but closely connected to him. One is a kind of reductive naturalism about intentionality. On this view, the point of attributing beliefs and desires is to give a high-level characterisation of the subject's behavioural dispositions, their functional architecture and their causal relations to the environment. Another of Stalnaker's views is externalism about mental content. This says that intentional characterisations are relational: even when two subjects are perfect intrinsic and functional duplicates, they may still differ in their beliefs and desires, depending on what objects and properties they are causally related to.
I gave a talk about the Canberra Plan on Tuesday (slides) in which I mentioned that I disagree with Lewis and Kim about the semantics of "pain": they say "pain" denotes whatever occupies the pain role in the species under consideration (or whatever is the relevant kind); I think "pain" rather denotes the property of being in a state that realises the pain role. One of the reasons I gave for my preference is that "pain" would be rather exceptional if it worked as Lewis and Kim believe.
Does the conceivability of zombies threaten type-A materialism, the claim that all mental truths are a priori entailed by the physical truths?
We can imagine beings exactly like us in all physical respects, but lacking consciousness. But this doesn't threaten type-A materialism (as I mentioned here). After all, it isn't a priori that materialism is true. It could have turned out that ectoplasmic states, rather than brain states, occupy the causal roles that, by analytic necessity, belong to mental states. Suppose it turned out that way. Then duplicating only our physical constitution would result in a being that is physically just like us, but lacking consciousness. So by type-A materialist lights, it is conceivable that things are such that there are beings physically just like us without consciousness.
Colours are physical properties of external objects. One such colour is Pure Green: the shade of green that looks not at all yellowish or blueish. However, if people are asked to identify the shade of green that looks not at all yellowish or blueish, they come up with (slighly) different shades: what looks pure green to me looks slighly blueish to you; what looks pure green to you looks slightly yellowish to me. What shall we make of this?
We could claim that one of the groups is simply right about Pure Green and the other wrong, even though there is no way to find out which is which. That is incredible.
In the latest issue of PNAS, there's an article on blindsight in ordinary people: The researchers induced local and temporary blindness by magnetically de-activating certain parts of brain area V1. When forced to choose, the subjects then often guessed correctly the direction or colour of a patch which they didn't consciously see.
(As usual, "consciously" is here functionally defined: what the subjects were missing is not some kind of non-functional, phenomenal consciousness, but a state with a certain functional role, leading in particular to utterances like "I saw a yellow patch". A case of truly phenomenal blindsight would be somebody who behaves in every way as if she consciously sees the patches, but who nevertheless doesn't see them consciously.)
Some philosophers seem to believe that narrow content must be defined without resort to external objects, leaving only bizarre options like phenomenalism, conceptual role semantics and global descriptivism. But that's wrong. Narrow content can and should be defined by external causal relations just like wide content.
By narrow content, I mean a kind of mental content that doesn't much depend on the subject's environment. Completely narrow content is altogether intrinsic to the subject. But hardly anyone believes in completely narrow content. The question is whether there is an interesting kind of content shared between my intentional states and the states of my twin on twin earth -- and those of swampman and those of a brain in a vat.
One can think of perception as a relation between states (or acts) and objects, the objects that are perceived. Alternatively, one can think of it a relation between a state and a content, the information acquired or represented in the perception.
Content is something that excludes possibilities. Suppose I have a perception of an elephant standing in front of me. What possibilities are thereby excluded? There are at least two reasonable answers: 1) the exluded possibilities are possibilities where there is no elephant in front of me; 2) the excluded possibilities are possibilities where I do not have that experience. Regarded as sets of possible situations, on the first account, the content of my perception is a set of situations in which there is an elephant in front of me. On the second, it is a set of situations where I have the phenomenal experience I actually have, even if it is caused by evil scientists. (Strictly, "I" need not be me, but can be whatever is in the center of the relevant situation.)
Panpsychism is the view that all physical things have, besides their physical properties, also psychological or phenomenal properties. The psychological properties are commonly assumed to be intrinsic. The idea is that physics only tells us about the structural and relational properties of things, but remains silent on what it is -- intrinsically -- that has all these dispositions and stands in all these relations to other things. So if we want to attach fundamental psychological properties to electrons (for example), we may well say that they are those physically unknown intrinsic properties: electrons ultimately are pain (say). But that's not essential to what I mean by "panpsychism". If you say that all physical entities have fundamental and irreducible, but extrinsic psychological properties, that's also panpsychism.
Stalnaker's "Lewis on Intentionality" (AJP 82, 2004) is a very odd paper. The aim of the paper is to show that Lewis's account of intentional content as developed in "Putnam's Paradox" -- global discriptivism with naturalness constraints -- faces various problems and conficts with what Lewis says elsewhere.
The first thing that's odd about this is that in "Putnam's Paradox", Lewis doesn't develop an account of intentional content. Rather, he discusses Putnam's model-theoretic argument and suggests that if one holds something like global descriptivism about linguistic content, adding external naturalness constraints on the interpretation of predicates would be an attractive way to block Putnam's argument for underdetermination.
A zombie world is a world physically just like our world but in which there is no consciousness. Must a type-A materialist deny the conceivability of zombie worlds? No, not quite.
Compare the rather uncontroversial hypothesis that "the HI virus" denotes the (type of) virus responsible for most AIDS infections. Is it conceivable that a world could be biologically just like ours but not contain the HI virus? Yes, for it might turn out that scientists have been wrong all the time and no virus is involved in most AIDS infections. If it turned out this way, our own world would be a world biologically just like ours but not containing the HI virus.
"Dynamical basis of intentions and expectations in a simple neuronal network" (PNAS subscription required, there's a free abstract):
[R]ecent indirect evidence suggests that intentions and expectations may arise in behavior-generating networks themselves even in primates [...]. In that case, interestingly, the intentions and expectations inferred from behavioral observations are not always identical to the intentions and expectations that are consciously accessible [...]. In this study we have demonstrated how such intentions and expectations arise automatically in the feeding network of Aplysia.
The "intentions and expectations" found are basically this: if you repeatedly present an A-stimulus to one of Aplysia's central pattern generators, and then switch to a B-stimulus, the pattern generator will respond as if it received another A-stimulus. Only after several B-stimuli will it switch to responses adequate for B. In this sense the animal expects to receive further A-stimuli, and intends to produce further A-behaviour. In a similar, slightly strechted, sense one could say that the animal believes to be in an A-environment (which is an environment containing seaweed). This belief is a certain state of the synapse linking Aplysia's neurons B20 and B8.
If the individuation of mental states depends at least partly on their causal roles, then it depends on the laws of nature (including possibly psychophysical laws). For if the laws differ between world 1 and world 2, a state with a given intrinsic nature can have causal role R in world 1 but lack R in world 2.
Assume world 1 is our world and world 2 is a world that contains a perfect spatiotemporal duplicate of our galaxy but lots of weird things elsewhere that contradict our laws. So the laws of world 2 are not the laws of our world. Then our duplicates in world 2 could have quite different mental states than we do.
But that sounds strange. I would have thought that my mental states do not depend upon what goes on outside the milkyway. We might also get the externalist problem about self-knowledge: If whether I believe P or Q depends on far away events, how can I know I believe P rather than Q if I don't know about these far away events?
Why not simply use a notion of content on which belief isn't closed under strict implication? Then it will be much easier to say that reasoning always delivers new content.
There is no shortage of fine-grained notions of content. We could use English sentences, or classes of intensionally isomorphic sentences, or bundles of tuples of objects and properties ('singular propositions') together with modes of presentation, or whatever. The tricky part is to say what determines whether a subject has a belief or desire with such a content.
As Robbie Williams remarked in the comments, perhaps what we do when we reason is putting parts of our fragmented belief space together. However, I doubt that this will do as a general solution.
First, at least in the context of an interpretationist account of content, it doesn't suffice for fragmentation that the relevant beliefs are somehow stored in different parts of the brain. Rather, if my beliefs are fragmented, say, into a compartment in which I believe P and one in which I don't, this must show up in my behaviour, more or less as follows: 1) In some contexts, the best explanation of some of my actions involves the assumption that I take the world to be P; but also 2) in some contexts, the best explanation of some of my actions involves the assumption that I don't take the world to be P; Moreover, 3) the discrepancy can't be explained as a change of belief.
(This is a follow-up to the previous post.) I think I've found a better way to provide for things like population-dependence in a Lewisian semantic framework. The trick is to regard it as a kind of index-dependence without explicitly introducing population-coordinates into the indices.
Recall, we want "pain*" to denote whatever state occupies the pain-role in the relevant population. Unfortunately, the relevant population isn't just the most salient population in the context of utterance, for we want to say things like
This is going to get a bit weird and technical. I wonder how a Lewisian semantics (along the lines of "Index, Context and Content" and "General Semantics") for terms like "pain" can make true everything Lewis says about such terms.
Assume that
1) Necessarily, for all x, x is in pain* iff x is in a state that plays the pain-role in normal members of the kind to which x belongs.
By "the pain-role" I mean the causal role attributed to pain by folk psychology. By "pain*" I mean whatever satisfies the condition expressed by (1). So (1) is more like a definition than an assumption. Lewis believes that our ordinary concept of pain roughly satisfies (1), but for what follows this doesn't matter. I think it's clear that we could have concepts for which something like (1) holds. Lewis's example of having a certain number stored in memory, as denoting a state of pocket calculators, sounds plausible to me (with the pain-role replaced by the role attributed to the state of having a certain number stored in memory by folk pocket calculator theory).
According to the Lewis-Nemirow ability hypothesis, knowing what it's like to see red is having a certain cluster of abilities. According to almost everybody who writes about the ability hypothesis, the hypothesis also claims that knowing what it's like neither is nor involves any kind of knowledge-that. This is indeed suggested by some of Lewis' remarks, in particular by this one on p.288 of "What Experience Teaches" (in Papers):
The Ability Hypothesis says that knowing what an experience is like just is the possession of [...] abilities to remember, imagine, and recognize. It isn't the possession of any kind of information, ordinary or peculiar.
One has to read the rest of the paper to find out that by "information", Lewis here most probably means exclusion of possible worlds. At any rate, it is clear from the rest of the paper that Lewis doesn't claim that all Mary learns are abilities.
I agree that it sounds fairly plausible to say that phenomenal states have a kind of representational content built into them. But I don't find that plausible anymore if it's combined with the assumption that being of phenomenal type Q is an intrinsic and essential property of phenomenal states. Here's an intuition pump.
Consider a world just like ours except that flying-pigs qualia have traded places with crooked-image qualia. That is, in this world, people have the kind of phenomenal experience we have when we look at flying pigs when they look at crooked images, and vice versa. But our duplicates at this world are not halluzinating flying pigs when looking at crooked images. No, they are not at all mislead by their experiences. For instance, they are not at all inclined to say that there are flying pigs, or that they are seeing flying pigs in these cases. Nor do they draw any of the inferences we would draw if we had the impression of seeing flying pigs. Instead, they typically infer that they are looking at a crooked image. And they would judge their experience to be veridical just in case there really is an image hanging crooked before them.
As promised here are some remarks on the content of phenomenal states. Or rather, on Horgan and Tienson's remarks on the content of phenomenal states in their paper "The Intentionality of Phenomenology and the Phenomenology of Intentionality". Dave Chalmers pointed me at that last week. I should say that this post is not meant to be a response to Dave's comment. I can't respond to it yet because I haven't understood it yet. I'll have to read the other papers he mentions.
Non-reductive (a posteriori, type-B) materialists say that even though phenomenal terms denote physical states or properties, the phenomenal way things are is not a priori entailed by the physical way things are. This means that no amount of physical information can tell us what our phenomenal terms denote. That is, non-reductive materialism implies that the projects of naturalising linguistic and intentional content are doomed. I would say that contraposititvely, since there are good reasons to believe in the project of naturalising linguistic and intentional content, non-reductive materialism is doomed.
I guess I should to clarify my argument. The position I want to argue against consists of the following two claims:
1) "pain" denotes a physical entity, say CFF.
2) For no P that only contains physical terms is "PI am in pain" a priori.
"" is the material conditional. I've chosen the pain example only for brevity: if you think it matters, feel free to replace "pain" by something like "the phenomenal quality of my current red-experience".
Brandt Van der Gaast points out that Michael McDermott proposes something like the semantics I sketched on behalf of type-B materialism in his "The Narrow Semantics of Proper Names" (Mind 1988). That's true. But I think McDermott is almost silent on the matter crucial to type-B materialism, and there is no acceptable way to fill the silence without spoiling type-B materialism.
First, on behalf of type-B materialism a reply to yesterday's post. (Thanks to Sven Rosenkranz for pointing out something like this to me.)
What makes it the case that the red-quale is the referent of "the distinctive quality of my current red-experience"? Not causal or counterfactual relations. Not demonstrative baptising. Not other kinds of verbal and non-verbal behaviour. Right. But these attempts to naturalize semantic properties are doomed anyway. They presuppose that semantic properties are generally independent of how things appear to us, which they are not. In fact, how things appear to us is an essential component in many "modes of presentation" determining the reference of terms. E.g., the referent of "red" is what appears to us under normal conditions in the way red things appear to us. In a same manner, the referent of "the distinctive quality of my current red-experience" is what appears to me in this distinctive way. Which is the red-quale, which in turn is a property of brain states. But no amount of physical information will tell you how this property of brain states appears to me. Phenomenalism as a semantic doctrine may have been too extreme, but it was not entirely wrong.
Some strings of symbols and noises and brain states have semantic properties. What makes it the case that a particular string or noise or state has the semantic properties that it has? One possible answer is that nothing does: semantic properties are fundamental, primitive and inexplicable. That's incredible. Not only because semantic properties just don't feel fundamental, as Fodor pointed out long ago. Much worse, it would make little sense of our use of semantic vocabulary: we believe that our expression "the moon" denotes that heavenly body up there; but how do we know? Maybe it really denotes the largest red thing in Alaska, or Gottlob Frege, or nothing at all. If reference is fundamental, facts about how "the moon" is used, how it got introduced, what conceptual role it plays, in what kinds of causal or counterfactual relations it stands to other things, how people verify statements containing it, what degree of naturalness various candidate referents have, etc. must all be considered irrelevant. On this view, there is a possible world where all these facts about use etc. obtain, but where "the moon" denotes Gottlob Frege. Obviously that's silly. Semantic properties are clearly determined by facts about use, baptising, causal chains, naturalness, etc. (In so far as they are determined at all, that is. No doubt sometimes the facts about use etc. are insufficient to settle whether a string or noise or state has semantic property A or B. But then it really is indeterminate which of the properties it has.)
There is a curious problem about rejecting both premise 2 and 3 in this familiar argument:
I believe that premise 3 is almost certainly false: why can't 'pain' denote CFF at our world and D-fiber firing at other worlds? Or, even better, CFF in humans at our world and other states in other beings here and elsewhere? Some claim that 'pain' must rigidly denote a kind of diagonal state that all beings who are in pain share. But I've never seen a convincing argument why this should be so. Crispin Wright argues (in "The Conceivability of Naturalism") that a) the reference-fixing description for 'pain' is something like 'state of feeling painful', which is itself rigid, and b) necessarily, pain satisfies this description. But it is not at all obvious to me that the reference-fixing description for 'pain' is 'state of feeling painful', rather than, for example, the non-rigid 'state that feels painful' or something physicalistically more acceptable.
Question: What exactly is wrong with something like this as a (physical-cum-indexical) conceptual analysis of "pain" (in my idiolect)?
the state I am in now
One obvious problem is that it's too unspecific: pain is not the only state I am currently in. But that's not the only problem. What else?
Is it a priori that I feel pain now? Or does my knowledge that I feel pain depend on empirical information? Could it turn out that I don't feel pain? Could it have turned out?
The world's first brain prothesis is interesting for several reasons:
Firstly, of course, it illustrates that when philosophers disagree about what would happen in a particular thought experiment, it is of very little help to carry out the experiment in reality: Will these rats become zombies?
Secondly, they are creating a hippocampus prosthesis. I guess they will also try what happens if that prosthesis is stimulated from the outside. There is a slight chance that this will have considerable effects on learning. I don't expect that we might one day learn just by stimulating the prosthesis. But we might learn much more easily by doing so.
Since narrow content is not determined by external factors, it depends much more on other propositional states than wide content. For example, if you believe that Aristotle was human whereas I believe he was a poached egg, the narrow content of all our beliefs about Aristotle will differ. When I believe that Aristotle was Alexander's teacher, you can't have a belief with exactly the same narrow content unless you also come to believe that Aristotle was a poached egg. Likewise for imaginings: When we both imagine Aristotle teaching Alexander, our imaginings cannot have the same narrow content.
Similarly, I think, if Ted believes that for any atoms there is a fusion, whereas Cian disbelieves this, they cannot share any imagining about atoms.