I’m delighted to be guest blogging here at Brains. My plan for my series of posts here is to introduce the set of questions at the center of my current project; shamelessly exploit the brilliant readers of this blog for my own purposes invite others to weigh in on these questions; and briefly present and motivate my own approach to answering them.
The set of questions at issue concern the mind-world boundary.
- Is there a profound, principled distinction between what is “within” the mind (mentality) and what lies outside minds (the non-mental)?
- If so:
What is the basis for this distinction? What characteristics distinguish mental phenomena from non-mental phenomena?
- If not—if our ordinary ways of distinguishing mentality from the non-mental are unstable, or don’t reflect a significant distinction between two kinds of phenomena:
What explains the intuition, widespread in philosophy, that there is profound, principled boundary between the mental and the non-mental?
A brief comment on how these questions are framed. One might think that what qualifies a state as within the mind differs from what qualifies it as mental. For example, perhaps the mind is the brain, but some states that occur within the brain are not mental states; perhaps a single mental state or phenomenon could occur in isolation, and not as part of a mind; etc. I’m not going to worry too much about the relation between “within the mind” and “mental”, but of course I’d be happy to discuss it.
I was spurred to reflect on the mind-world boundary by puzzlement about an influential philosophical thesis, advanced by Tyler Burge. Burge argues that we must sometimes construe mental contents widely (individuate them partly by factors external to the mind) in order to capture the thinker’s “epistemic perspective”: “how things seem to him, or in an informal sense, how they are represented to him” (Burge 1979, 25). What I found puzzling about this thesis is the implication that the boundary of the mind is fixed independently of the epistemic perspective. To put it loosely: if we don’t draw the “mind” boundary by reference to the epistemic perspective (or the metaphysical determinants thereof), then how do we draw it? If we deny that how things seem to me is purely a matter of factors within my mind, what possibilities remain for understanding “within the mind”?
(In a 2012 paper, I argue that there is no way of understanding the mind-world boundary that does justice to the leading positions in the debate over content externalism. I conclude that, since these positions are defined in terms of that boundary, there is no univocal thesis of content externalism or content internalism.)
The question what distinguishes mentality from the non-mental realm also bears on the dispute over whether the mind extends beyond the body. To defend a position on the extended mind thesis, one must invoke some reasonably clear conception of “within the mind”—or, at least, a particular sufficient (or necessary) condition for qualifying as within the mind. Moreover, the extended mind thesis is inconsequential unless the mind-world boundary matters.
A guiding assumption of this project is that there seems, to many philosophers (and folk), to be a profound, principled contrast between what is within the mind and what lies outside it. I do not assume, however, that there is such a contrast.
To understand why philosophers (and/or the folk) accord significance to the mind-world boundary, it will be illuminating to survey the principal characteristics thought to be unique to mentality. A list of mentality’s allegedly distinctive characteristics will also help us in approaching the questions above. For example, if we determined that none of the characteristics on the list were uniquely possessed by mentality, then—absent some additional distinctive characteristics—we would have some reason to doubt that the boundary of the mind was profound. And if we determined that mentality’s unique characteristics were only contingently unique to the mental—that is, that these characteristics could in principle occur outside the mental realm—then we would have some reason to doubt that the boundary of the mind was principled.
The task, then, is to assemble a list of characteristics that are regarded (by a substantial segment of philosophers, philosophical tradition, and/or folk) as (i) unique to mentality, and (ii) significant enough to ground a non-trivial contrast between what is within the mind and what lies outside it.
I’ve come up with a list of five such characteristics. Again, I’m neutral (for now) on whether any of these are unique to mentality—or, for that matter, as to whether they are possessed by mentality. I’m also neutral on the relations between them.
Mentality appears profoundly distinctive because of one or more of the following.
- Phenomenality. Only mental states are such that there is “something it’s like” to be in them.
- Privacy. Only mental states can be known via a method uniquely available to their bearers.
- Intentionality. Only mental states are intrinsically intentional.
- Rationality. Only mental states can distinguish genuine reasoning from “brute” processing.
- Agency. Only mental states can distinguish intentional action from mere behavior.
I’d be grateful to know whether I’ve missed any candidates here. What kinds of characteristics have been thought to be distinctively mental? (I should note that “being nonphysical” is not included in this list, because I think that characteristic is always derived from one or more of those on the list.)
One of the classics is that the mind is not spatially extended: consciousness is in time, but not in space.
Maybe this is covered by a combination of phenomenality and privacy, but besides the method being uniquely available, it is also a quite different method than external sense perception. So, taking a cue from Brentano, mental phenomena can only be perceived by internal perception.
Thanks, Carlo — these are both interesting suggestions. About the idea that the mind isn’t spatially extended: that is a classic view, as you say, but I’m looking for characteristics that might explain why philosophers have thought there was a significant contrast between the mental and the non-mental. The idea that the mind is non-spatial seems, to me at least, standardly an inference from the premise that the mind has certain special features. (Ditto for “the mind is nonphysical”.)
And yes, I regard the various versions of the “special method” claim–whether this is internal perception, direct acquaintance, etc– as falling under the combination of phenomenality and privacy.
Thanks for your thoughts on this!
Hi Brie, what a great question! I have a concern about your characteristic (2), Privacy, namely that it also seems possible to know in a uniquely first-personal way things about oneself that aren’t mental, e.g. that one’s body (or a certain part of one’s body) is moving or at rest, or that one’s heart is beating, etc.
When these special ways of knowing concern one’s *intentional* activities — what one is intentionally doing — I am happy to embrace the conclusion that the matters in question are mental after all: an intentional action is a mental process whose components extent beyond the brain (and sometimes the body). But it seems hard to take this position for, say, the beating of one’s heart or the reflex movement of one’s leg, even though one has a uniquely first-personal way of knowing “from the inside” that these things are happening.
If this is right, then there are also some non-mental states that “can be known via a method uniquely available to their bearers”. What do you think?
Hi John — thanks for your comment. (And for your work managing this blog!)
I can see that I should have been clearer about what I was intending with this list. To clarify: I’m not saying that any of these 5 characteristics are unique to mentality; in fact, at this point I’m not even saying that mentality possesses any of them. (I’m skeptical about some of them: e.g. the “rationality” claim. But put that aside.)
The exercise is, in effect, sociological: to identify features that some (reasonably significant) thread of the philosophical tradition has regarded as uniquely mental, so as to understand what kinds of factors motivate the view that there is a profound divide between the mental and the non-mental.
Having said that: your examples raise an interesting challenge to the “privacy” claim. My initial thought is that these examples show what a more detailed version of the claim needs to include. (Obviously, all 5 of these claims need to be spelled out more thoroughly.) The claim would need to include an “in principle” clause. It seems to me that someone else could, in principle, know that your heart is beating by use of a similar method–e.g., if their phenomenal states were properly linked up with the conditions of your body. After all, it’s the (broadly causal) link between your heart’s beating and your sensations that enable you to know that your heart is beating by use of a method others can’t use to know this (viz., how it feels).
… So although I’m not committed to the privacy claim, I do think that claim — if suitably modified — could be defended against those counter-examples. What do you think?
Hi Brie, thanks for your clarification, and the kind words. I would think that even in the science-fictional state you describe, the other person’s knowledge of the state of my body would still be different from my own, as she would have to identify whose body she was aware of in this instance, whereas the ordinary awareness of one’s own body “from within” can ground knowledge of one’s body that is identification-free. (Of course these concepts of identification-freedom, immunity to error through misidentification, etc., are famously tricky, but it’s difficult to spell them out in a way that rules in the self-ascription of uncontroversially “mental” characteristics while clearly ruling out things like the state of one’s own body.)
Thanks, John. I think the fictional case can be spelled out in a way that avoids this asymmetry. Basically, I think that your knowledge of your heartbeat is “identification-free” in only a weak sense. You don’t need to go through a process of identification, but that’s only because you’re psychologically predisposed to take that sensation to indicate or track something within your own body. What you’re naturally predisposed to take sensation S to track, and what sensation S actually tracks, could both be altered in a fictional scenario, so that you could be predisposed to take sensation S to track (and sensation S could track) another creature’s heartbeat.
My view on this issue is nonstandard, since I think that the only candidates for genuine immunity to error through misidentification are states known through direct acquaintance; although for other states we don’t go through a process of identification, per se, that’s only because we implicitly rely on a default assumption that (roughly) a particular phenomenal state tracks those states. (Obviously, I’m not claiming that we actually entertain this assumption. This point is purely about the epistemology.)
Thanks, Brie, I thought your position might be something like that. I myself am doubtful that the arguments used to argue that bodily self-awareness isn’t IEM in some strict sense don’t also lead to the conclusion that the same holds for the awareness of many (perhaps all) of even our conscious mental states. Probably this is too complex a question to be worth pursuing here, but Daniel Morgan has a very nice paper (I think it’s this one: https://www.academia.edu/8467045/Thinking_about_the_body_as_subject) that argues for this sort of parity between mental and bodily self-ascription: basically, his argument for any state S where we’re not perfectly infallible about whether we’re in S, it’s possible that our apparent S-states really arise from the S-states of someone else due to fictional rewiring, etc. So unless direct acquaintance rules out even this kind of non-luminosity, then by parity of reasoning we should conclude that just as in your view it’s only due to a default assumption that we take our seeming bodily sensations to track the true states of our own bodies as opposed to someone else’s, so it’s only due to such an assumption that we take our apparent mental states to track the true states of our own mind.
Thanks for that reference — I look forward to reading it. And yes, I think this discussion could easily take us very far afield. But I can’t resist adding one point. The IEM claim I’m making isn’t that we’re infallible about whether we’re in a given state–pain, say. Nor is it that pain is luminous. It’s just this. I cannot have knowledge by acquaintance that someone is in pain, while erring in thinking that it’s me who is in pain. (Note that this is compatible with my being fallible in thinking that I’m in pain, and with pain’s being non-luminous.) This follows pretty straightforwardly from the metaphysics of acquaintance. So I think the place to press here is on the claim that we have any such knowledge by acquaintance.
Thanks for this fascinating post. Regarding phenomenality: I tend to think that the standard definition in terms of “what it is like” to be in a mental state prejudices questions about the mind-world boundary in favor of traditional internalist conceptions. If we think of consciousness as relational, such that mind-indpendent objects and their properties can be constitutive of mental states, then the picture looks much different than when we define conscious states in ways that don’t depend on what they provide awareness of. I have in mind as the alternative here a naive realist conception of consciousness that has been defended by people like John Campbell, Alva Noe, Mike Martin, and others.
This issue seems tied to the question of epistemic perspective you raise as well: the difficulty with the internalism/externalism debate in epistemology, it seems to me, is a shared assumption on both sides that conscious awareness can only provide a subject with “internal” reasons, that is, reasons that have their justificatory force independently of their connection to the truth. If we think of consciousness as relational, however, then it is not clear that demand for awareness of reasons is incompatible with the demand that how things are must bear on our epistemic standing.
Thanks, James, for this contrasting perspective. The views you mention are precisely the kind of views that led me to consider how we conceptualize “within the mind”. For example, the view that consciousness is relational immediately prompts the question: relational with respect to what? That is: how should we conceptualize the thing that consciousness is not intrinsic to? If that thing is just the brain, then the conclusion would be that the mind is not the brain. But I take it that, on at least some of those views, consciousness is said to be relational relative to the mind. This raises the question: how are we fixing the “mind” boundary, given that it’s independent of the “consciousness” boundary (or, at least, of the “factors that fix consciousness” boundary)?
A brief note on the “what it’s like” construal of phenomenality. I don’t quite see how this is itself internalist. Maybe this will help: this construal doesn’t exclude the possibility that phenomenal states have intrinsic intentional content. My goal, with this initial list, is to identify leading philosophical motivations for taking mentality to differ profoundly from the non-mental realm. I take it that one central motivation for this position is the intuition that only (though perhaps not all) mental states have a “what it’s like” quality.
Hi Brie, I enjoyed your post and am looking forward to the rest of the series! In the sociological spirit of the endeavor, I wonder whether there might be a 6th possibility. I’m having trouble stating it precisely but it would be something along the lines of: Only mental states contribute to my being the self that I am. (Or perhaps: having the identity that I do.) This would be a view that comes out of work on the psychological approach to personal identity.
Hi Amy! I’m not sure I have a good response to this — that is, either a good way of formulating a candidate for the identity-conferring feature of mentality, or a good reason to think that we needn’t add such a candidate to the list.
But my inclination is towards the latter. I’m inclined to say that mentality contributes to my being the self that I am, in whatever way it does, only because it is mental–that is, only because it has some other distinguishing feature of mentality. E.g., a bundle theorist might single out phenomenality and intentionality. One who identified the self with character, shaped through practical deliberation and intentional decisions, might single out rationality and agency. An advocate of the memory criterion might think that memories are intentional (& maybe phenomenal) states that are appropriately linked to previous experiences (=intentional-phenomenal states). etc.
One view that might resist this strategy is the hacceity view. So I’ll have to consider whether a motivation for taking the mental to differ profoundly from the non-mental is that only mental states — namely, mental hacceities — are identity-constituting. (“Mental hacceity” may be inappropriate, since hacceities are non-qualitative…)
Another term that seems worth mentioning in this context is ‘subjectivity’, sometimes specified as ‘ontological subjectivity’ in opposition to ‘epistemic subjectivity’ (cf. Searle), though it is certainly not always clear what is meant by this. In one reading, it just is another term for what you call ‘privacy’. In another, it just emphasizes that any mental state requires a subject, is somebody’s state – though this is hardly distinctive of mentality. Moreover, it may be thought of as highlighting that subjects of mental states experience and understand themselves in relation to objects. In this reading, it is close to your feature ‘intentionality’.
These features of mentality as belonging to a subject and relating us to objects are also very relevant to the issue of externalism, because we experience ourselves as spatially limited creatures and think of our mental states as states of that organism of those creatures and as of relating us to objects without these objects thereby becoming part of the mind. So, however we draw the boundary between mind and the rest of that organism, this is why at least from a common-sense point of view the idea that mind extends beyond that organism is so disturbing.
Finally, since I have now specified the subject of mental states as a ‘creature’ (or ‘animal’), let me mention another feature that many views implicitly or explicitly ascribe to minds: it is biological.
Thanks, Michael. The question whether to add “subjectivity” to the list is an excellent one — I’ve puzzled over this. In the end, I’m not sure how to cash it out in a way that doesn’t reduce to one of the other characteristics. The closest I’ve come is: point of view. (Some philosophers, e.g. Kriegel, think that conscious states have a built-in “for-me-ness” — but I confess I can’t find this property when I introspect.)
On some views, some mental states are ontologically subjective. But I take it that this feature is posited on the basis of some other feature–presumably, some feature already on the list. (It is for this reason that I don’t include “nonphysical” on the list, although of course some regard that as a unique feature of mentality.) And being biological don’t fit on the list, since it’s not regarded as unique to mentality.
Thanks for your response, Brie. I agree that it is not so clear how “subjectivity” should be understood and was gesturing towards something like what you call “point of view”. What seems to me to be important, particularly in discussions of externalism, is the thought that (a subject’s) mind might be essentially intentionally related to entities external to it (to its objects), while still being clearly distinct from them. This is not exactly captured by your feature intentionality because some internalists will think that intentionality does not require any relations to external objects and some externalists that these objects are somehow part of mind. So I was trying to recruit the word “subjectivity” to express this thought, because in many traditions subjects are thought to be essentially related to objects.
And the other feature the term suggests is belonging to a subject as a larger unit. Neither this feature as such, nor, as you point out, the idea that this subject (and its states) are biological is indeed unique to the mind. I only mentioned it because it seems to me to be important that even apart from the question of how the mental can be distinguished from the non-mental states of the subject, the very idea of mental states being states of a biological, spatially limited being, already excludes some forms of externalism.
Hi Brie,
Interesting stuff!
One sociological factor (which is perhaps what you intend under “privacy”) is issues about external world skepticism. Most philosophers find Cartesian doubt at least prima facie worrisome and it quickly motivates a line between that which can be doubted and that which seems immune from that sort of doubt. (Rorty on this question argued that incorrigibility was the mark of the mental, and then argued nothing is incorrigible, so there’s no mental.)
More broadly, there seems to be something about the appearance/reality distinction. To explain illusory perception, we come to realize there’s a way things appear to us that is at times differs from the way things are. So, (maybe) appearances = mentality; reality = everything else. Of course, this is contentious (to say the least), but it may motivate the urge to draw a line, both folk-psychologically and philosophically.
Hi Josh. Yes, I agree that the factors you mention have been centrally important in driving this distinction. I’m inclined to include both of them under the “privacy” label (as you mention). But I do see that the skepticism-immunity point is slightly different. “Our knowledge of X is especially secure (immune from skepticism)” differs from “We have a special method of achieving knowledge of X.” Another possible moral is that I need a more wide-ranging epistemic category. Arguably, these two epistemic claims are closely related: at least traditionally, it’s because we have a special method of knowing our own mental states that this knowledge is especially secure. And the special security of self-knowledge is traditionally evidence for the “special method” claim.
The “appearance / reality” divide does seem to fall out of these epistemic issues. On a veil of perception view, this divide matches up with both the “special method” claim (direct vs. indirect) and the “especially secure” claim (knowledge of appearances is more secure because more direct).
Hi Brie,
Fascinating post, what comes to my ‘mind’
Shareability-Outside of a blood transfusion, it is the one thing we most easily share with other agents that we ascribe mind to. Not just humans but pets and even children ascribe mind to their playthings or for that matter teenagers name their cars ‘Brad’. Phantom limb may assign error but really co-opts ability of mind to empathize and other social abilities.
Relationality-Once again socially minds have awareness of given status or ascribe status to other agents. Mathematics may be another co-opting of mind’s natural relationality.
The inside-outside of the mental is a natural characteristic of the mind’s natural social function.
Thanks, Vic — some provocative thoughts! I may be missing something here, but both of these factors strike me as reasons we might deny a sharp division between mind and world–since in that division “world” includes other minds.
As you inquired; what is the motivating factor for the perception of a mental / non- mental divide? It may be the experience factor that we can break the perceived divide with members in our own group(s) and maintain the divide with non-members.
Very thought provoking post, thanks!
What do your think of treating the concept of mind as akin to a natural kind concept? Your list (and perhaps many additional, intuitively less central features) would be the “surface features” that need to be explained by the kind’s (currently unknown) underlying nature. The clustering of the features themselves would justify the defeasible assumption that we’re dealing with a kind, and therefore a mental/non-mental difference, with no need to have some definite criterion in advance.
On this model, one would be hostage to empirical fortune when theorizing about mental features and hypothesizing about a feature’s internal or external status. But the degree of hazard would vary based on our current empirical confidence about the mind’s nature. Presumably one would be pretty safe in saying that historical things external to the body are also external to the mind, and reasonably safe in saying that some aspects of the activity in visual cortex strongly correlated with conscious experience are internal to the mind. As far as content goes, it would depend on what the particular theory of content in question talked about. History? visual cortex?
Given our current state of knowledge, claims going beyond our own kind of mind would be on very shaky ground indeed. And any claims about internal/external that greatly outstrip our current empirical understanding would have to be relativized to some theory about the mind’s nature. (The extended mind thesis could be an aspect of some such theories.)
Wondering what you think of this approach – the goal would be to avoid your apparent demand for a clear criterion in advance of theorizing, while still motivating the view that there’s a mental/non-mental divide. Does this fit in your framework somewhere? Or is it playing in a different field? Or is it maybe doomed to failure?
Thanks, Dan. This is a great question, and really gets to the nub of my project. I have two responses.
The first is to your suggestion that we might treat “mind” as a natural kind. I do think that some aspects of mentality can profitably be treated this way. But I’m hesitant to think of “mind” (or “mentality” generally) as a natural kind, since doing so would commit us to saying that the presence of surface features are neither necessary nor sufficient for mentality. To put the worry in a nutshell: if phenomenality (etc) are merely surface features, then it’s conceptually possible that there be phenomenal experiences that are not mental. (That would be “fool’s mentality”, the equivalent of “fool’s gold”: something shiny and yellow, etc., but which wasn’t gold.) And I’m inclined to think that that isn’t conceptually possible. So treating “mentality” as a natural kind would change the topic, from mentality to schmentality.
… Having said that, I’m not claiming that we can generate, from the armchair, a list of necessary and sufficient conditions for mentality. Maybe we can get only a partial list. And maybe the concept of “mind” doesn’t lend itself to a tidy list of that sort: e.g., maybe there are paradigm cases of mental states, and then other states that qualify as mental in virtue of similarity to those paradigm cases.
This brings me to my second point, which concerns your final paragraph. I’m not looking for a genuine criterion for mentality, and I’m not trying to motivate the view that there’s a mental/non-mental divide — I’m neutral (for the sake of this project) on the question whether there is one. Rather, I’m trying to identify what has driven some to think that there is this divide, so that I can make a proposal about the kinds of states that drive that impression. That proposal will appear in the next post, which I’m working on now.
Cool – I look forward to the next post. (Maybe the natural kind approach could be added to the list of reasons why “some” think there’s a divide? I would put myself in that camp – not sure who else belongs there. Probably Millikan, maybe Lycan, maybe Brentano [if Uriah Kriegel is right]. I’d be curious to know who else you think might belong there.)
I think fool’s mentality and twin-mentality make sense, in the end. (Non-mental phenomenality seems easy – proto-phenomenality in certain versions of type-F monism, or John Heil’s qualities would seem to qualify. But I wouldn’t call those fool’s mentality, since too many other features of the mind stereotype are missing.) I think we’re misled by the fact that phenomenality, intentionality, rationality etc. are pretty deep for “surface” features of a stereotype. But compare “bronze”: arguably the stereotype is not composed of surface features, but comparatively deep ones: copper, tin, and zinc. But you could get twin-bronze via twin-copper, twin-tin, and twin-zinc. Similarly, I think you can get twin-mentality via twin-phenomenality, twin-intentionality, and twin-rationality. On the other hand, maybe this would just be a very different kind of mind – though I’m not sure even that would undermine the natural kind strategy. (But that is probably a long discussion – feel free to ignore!)
Hi Dan — I see the attraction of the natural kind approach, but it’s not so clear to me that that would be a reason for thinking that there’s a deep mental/non-mental divide, as opposed to a view about one side of the divide. The kind of divide I have in mind doesn’t follow simply from a division between one natural kind and things not belonging to that kind.
We may simply disagree about fool’s mentality. (Proto-phenomenality is proto for a reason.) Part of our disagreement may stem from the fact that I think of “mentality” as belonging, most basically, to states rather than to (entire) minds; whereas you may be thinking of mentality as “belonging to a (larger) mind”. In any case, my point about phenomenality being not merely a surface feature is, essentially, just the Kripke point: pain is not known by surface features, but rather by essential features. But of course there are many who reject this Kripke argument.
I agree entirely that a “deep” divide (a somewhat subjective matter) certainly doesn’t follow from the mental being a natural or (perhaps better) real kind, but the empirically discovered facts could certainly yield that result. For example, I’d say life vs. non-life is a deep divide, even though fuzzy at the edges. Would that count as “deep” for you? The real kind “mind” could even turn out to be non-physical, which would presumably be deep enough!
Perhaps more importantly, I think you’re right that I’m taking whole minds to be primary, whereas you’re starting from mental states. Probably a symptom of our differing approaches from the get-go: my starting place is an empirical bet, whereas yours seems to be a conceptual point that I’d reject (especially insofar as it relies on the Kripke argument). This also means I’ll be getting off the boat very early in your next series of posts. With reference to the “A Hypothesis” post, I wouldn’t accept your crucial second starting assumption. Yes, mentality may be special, but likely not “in virtue of” one of the 5 characteristics (as opposed to the underlying explanation for their clustering).
The living/non-living contrast is interesting for comparison. 150 years ago someone might have argued themselves into believing that individual “vital states” were special, and this is what made them living states. But of course it turns out that individual living states are special only by being states of living things. My empirical bet is that the mental/non-mental contrast is exactly like this.
Hi Dan, I’m not sure I have an intuition about the living / non-living divide — whether it’s “deep” in the way that the mental / non-mental divide is often taken to be deep. I certainly agree that “mentality is a natural kind” is compatible with “there is a deep divide between the mental and the non-mental”. Your description of our disagreement seems right to me. About the “vital states” issue: I do think our conception of phenomenal properties differs crucially from the conception of “vital” properties, in a way that blocks this comparison. But this once again registers our difference about the Kripke point.
Thanks, Brie! looks like it all comes down to our difference about Kripke.
Hi Brie, it looks like I have to move our discussion to a new thread, so here it is. You write:
I am inclined to agree with this, but don’t see why the same doesn’t hold for proprioception. You claim that even in a rewiring case where my pain-like experiences result from the pains of someone else, these experiences can’t give me knowledge by acquaintance that someone is in pain, while leaving room for error that it’s me. But one who defends the parity of mental and bodily self-ascription can make a similar move with respect to proprioception: in a rewiring case where my experiences as of my body from the inside result from the bodily states of someone else, these experiences can’t give me knowledge by proprioception of someone’s bodily states, while leaving room for error that it’s me.
Put another way — what reason is there to deny that proprioception is just as much a way of being “acquainted” with one’s bodily states as introspection is a way of being acquainted with one’s mental states, at least if we haven’t set up the concept of acquaintance in a way that assumes a certain mind/world dichotomy?
Hi John. You say:
You claim that even in a rewiring case where my pain-like experiences result from the pains of someone else…
I didn’t intend to make that claim. When I know my own pains by acquaintance, I’m directly acquainted with the pains themselves. So I don’t know my pains by an experience that merely results from those pains. The contrast with proprioception is that, in knowledge by acquaintance, I’m not relying on a causal-nomological link between how I know a certain state (the feeling of heartbeat or crossed legs) and the state known (the heartbeat or crossed legs).
Perhaps what you’re suggesting is this. Proprioception is, by definition, knowledge of one’s own body. So if I can know that a heart is beating, while being in error in thinking that it’s my heart that’s beating (because I’m rewired to someone else), then this isn’t proprioceptive knowledge. Is that the idea?
Hi Brie,
Yes, that last point is part of what I was driving at — that the possibility of rewiring cases doesn’t undermine the immunity to error through misidentification (or identification-freedom, as I prefer to think of it) of proprioceptive knowledge, because what you have there isn’t proprioception at all.
This point could lend support for the more radical-seeming idea I was driving at toward the end of my comment: that proprioception might be a form of acquaintance with one’s bodily states, albeit one that’s sometimes mistaken for other forms of awareness (or seeming awareness) that aren’t such forms of acquaintance at all (say, because they’re hallucinatory, or because they derive from some kind of rewiring). Once we allow, as I take it you do, that knowledge by acquaintance isn’t always infallible even where it concerns (what you would call) our mental states, I can’t tell what would rule this possibility out. And if it’s coherent, then we needn’t say that proprioceptive knowledge relies on a causal-nomological link between proprioceptive awareness and the states of one’s body (though of course a certain causal-nomological background is required for proprioception to be possible) — rather, on this view proprioception acquaints a person directly with the state of her body, no less than introspection acquaints her directly in touch with the state of her mind.
Right. If you define proprioception that way, then it is IEM, though only trivially. The difference with acquaintance (as I see it) is that, when one is acquainted with a pain, one has evidence that “pain is occurring” that could not qualify as knowledge unless it’s knowledge of one’s own pain. By contrast, one could know that a heart is beating by relying on proprioceptive evidence (or if “proprioception” is defined first-personally, then evidence indistinguishable from proprioceptive evidence), but be mistaken in thinking that it’s one’s own heart that is beating.
The difference, as I see it, is that in some restricted circumstances, we can have evidence about pain that entails that we’re in pain–viz., by attending to the painful feeling. But I don’t see how one can have evidence that one’s heart is beating that entails this. And it seems to me that evidence one has that one’s heart is beating may stem from another’s heart, in a way that allows one to know “there is a heart beating” but be in error in thinking “my heart is beating”.
This asymmetry is just what I am trying to resist. (What follows channels that argument of Daniel Morgan’s, at least as I recall it.) If we accept that we’re not infallible in recognizing our pains as such, then we can imagine a rewiring situation where it’s arranged so that something that a person is liable to misrecognize as a pain — a certain sort of tickle, say — is reliably produced by the pains of someone else. When the person on the receiving end in this situation is acquainted with what seems to her to be a pain, this *is* evidence that pain is occurring that isn’t necessarily knowledge that she herself is in pain.
Now, for my part I want to resist the conclusion that the self-ascription of pain isn’t immune to error through misidentification, but I don’t see why the case of bodily self-ascription should be any different: in each instance we have a case where in ordinary circumstances (i.e., ones where we’re not rewired or worried about rewiring) a person can self-ascribe properties in a manner that’s doesn’t require identifying herself with the person who possesses the property in question, or making any assumptions about the causal processes that hook her up to the property she is (or seems to be) aware of.
Hi John,
We seem to be getting into increasingly finer-grained descriptions of our disagreement. I’d deny that the scenario you present in your first paragraph is one in which S knows by acquaintance that someone is in pain, but does not know that she herself is in pain. In that scenario she can’t know that she’s in pain by acquaintance–for she’s acquainted only with the tickle. The difference here isn’t just stipulative, re: the metaphysical relation of acquaintance. There’s an epistemic difference as well. In the scenario you describe, her evidence (the presence of a tickle) differs from the evidence she’d have if she were acquainted with pain.
This may help: in rejecting the claim of infallibility, I’m denying that there is any kind of mental state (e.g., pain) about which subjects are generally infallible. But there is a kind of evidence that is present only when one IS in pain, namely, the pain itself. Suppose that I carefully attend to my pain, and recognize it AS pain. In that case (and skipping over some details here), I know by acquaintance that I’m in pain. Now suppose that I attend to my pain, and mistake it for a tickle. In that case, I don’t know that I’m in pain. But this isn’t error through misidentification, since I don’t know that something is in pain. (And of course EM involves knowledge that something is F, yet mistakenly thinking that it is I that am F. )
Hi Brie, sorry for my slow reply. I agree that we are pretty deep in the weeds here, so maybe we should continue this thread in the discussion over e-mail. I’ll give it one more try here, though.
The point of the rewiring case I am borrowing from Daniel Morgan is to fit the description you say at the end is impossible to fill. If I sometimes mistake certain sorts of tickles for pains, and if things are set up such that the only time I feel a pain-like tickle is when someone else is in pain (and, perhaps, I know about this set-up), then — by certain criteria, at least — the presence of a painful-or-pain-like sensation *does* give me knowledge that someone is in pain without settling the question who it is, just as in your view the possibility of being rewired to someone else’s body means that the presence of bodily sensations can give me knowledge that someone’s body is in a certain state without settling the question whose.
Now, I absolutely *don’t* draw from this possibility the conclusion that the self-ascription of pains isn’t IEM, or that we don’t have a distinctively first-personal way of knowing when we’re in pain. The point is rather to raise a “bad company” objection to a common line of thought concerning bodily sensations, namely that the possibility of rewiring cases shows that bodily awareness isn’t distinctively first-personal, and that the knowledge of one’s body “from within” depends on knowing which causal-nomological facts govern the relation between one’s bodily sensations and the body whose state they derive from. I *don’t* think this is true of bodily awareness: rather, in my view the most a rewiring case shows is that it’s *possible* for bodily self-knowledge not to be IEM, while *ordinarily* our bodily self-knowledge doesn’t depend on knowledge of causal-nomological facts about proprioception (any more than our mathematical knowledge depends on our knowledge of causal-nomological facts about memory, as Burge has argued). But Morgan’s example is supposed to show that if you reject this position with respect to bodily awareness you need to do the same with respect to introspection, holding that the self-ascription of pains, etc. depends on the knowledge of such causal-nomological facts as that you’re not given pain-like tickle sensations whenever a certain other person feels a pain, etc.
I hope this clarifies what I’m after here. I’m happy to let you have the last word for now, and hope we can discuss it more down the line.
John, I know you were suggesting to move the discussion elsewhere, but just a quick thought: might it be that the pain case is too close to the case of bodily awareness to draw out the distinction between the mental and the non-mental that Brie and others are after? After all, pain can be thought of as tracking certain bodily states, just as proprioception does. Yet, the kind of rewiring case described by Daniel presumably wouldn’t work for thoughts/beliefs, or would it? So, arguably, one could defend a distinction between self-ascriptions that enjoy de-facto immunity (such as self-ascriptions of bodily states and perhaps also pain) and those that enjoy logical immunity (such as self-ascriptions of thoughts). I’m not sure whether this distinction matters as much as some people have thought (I tend to think it doesn’t – on my view, so-called de facto IEM is all we need for most intents and purposes), but it might nonetheless point to a possible distinction between the mental and the non-mental.
This is interesting. I think it might come down to how you understand beliefs, and how you think we ordinarily come to have knowledge of them. If it’s a matter of tracking down conscious state — what Brie calls occurrent beliefs — via introspection, then there will be the possibility of error (I think Williamson discusses mistaking a belief for a hope), and so the possibility of a suitable rewiring story. And certainly if you want to preserve — as I don’t — a distinction between occurrent and dispositional beliefs, then with regard to the latter there will be the possibility of rewiring stories, where one person’s dispositional beliefs are made to give another person belief-like phenomenology. It might be, though, that a story like Moran’s where we know our beliefs through an exercise of rational agency can rule out these possibilities — but I doubt that it wouldn’t rise again in another form.
Regarding the point about pains tracking bodily states, I take Brie’s view to be that while a person’s bodily states themselves don’t meet her criterion of Privacy, her sensations of them are supposed to. It’s the tenability of this distinction, or at least a certain way of motivating it, that I’m trying to put pressure on here.
Very interesting stuff! I wonder if the contemplative approach you take will provide you the traction you need, however. Another way of looking at the mental/nonmental divide is to ask what kinds of systems are solving what kinds of problems. So when you ask, “What is the basis for this distinction?” the obvious answer is that (at least) two different types of systems are solving two different kinds of problems. In the case of the ‘mental,’ the cognitive systems involved have no access to the causal structure of the systems solved (brains). Mental posits, you could say, allow problem-solving absent high-dimensional causal information. Mental posits, in this sense, are a family of heuristic devices, ways to deal with complex systems absent access to that complexity.
Is it simply a coincidence that at least four (I would argue all five) of the characteristics you list involve deep causal puzzles?
Would you be willing to include ‘causal discontinuity’ on your list? Isn’t this discontinuity, the causal inscrutability of the mental, what motivates your entire project in the first place? If the mental could be seamlessly integrated into our neurobiological understanding, for instance, it would simply be another domain among others.
Following up on Scott’s comments, I wonder if any of your concerns are answered by the approach taken in “Space, self, and the theater of Consciousness” which you can read on my website :
https://people.umass.edu/trehub
Thanks, Arnold, for this reference — I look forward to reading your paper.