Path: utzoo!utgpu!jarvis.csri.toronto.edu!mailrus!ukma!rutgers!elbereth.rutgers.edu!harnad From: harnad@elbereth.rutgers.edu (Stevan Harnad) Newsgroups: comp.ai Subject: Re: Question on Chinese Room Argument Summary: The Chinese Room Argument is not about linguistic matters Message-ID: Date: 24 Feb 89 03:02:55 GMT References: <3312@uhccux.uhcc.hawaii.edu> Organization: Rutgers Univ., New Brunswick, N.J. Lines: 106 lee@uhccux.uhcc.hawaii.edu (Greg Lee) of University of Hawaii wrote: " No, there aren't "two senses of "understand," a subjective and an " objective one," [otherwise we couldn't say] 'He understands, and I do too' As I've suggested already, this is simply not a linguistic matter. The distinction I'm after is already there with "pain" (although we don't have two senses of pain as we do of understanding -- the reason for this will become clearer as we go on). Consider "I'm in pain and he is too." Apart from the obvious fact that I don't mean he's in MY pain (which is already a difference, and not a "linguistic" one but an experiential and conceptual one), it makes sense to say "He SEEMS to be in pain (but may not really be in pain)," but surely not that "I SEEM to be in pain (but may not really be in pain)." (Please don't reply about tissue damage, because that's not what's at issue here [I didn't say "I seem to have tissue damage"] -- or about lobotomy, which may very well change the experiential meaning of pain for me.) The difference (for me) between my pain and his pain is that mine is directly experienced (by me) and his is only inferred (by me) from his behavior. Now the case of "understanding" is quite similar, except that the behavioral criteria for the inference are much more exacting -- so much so that there I CAN say "I SEEM to understand (but may not understand)." The reason I can say this is apparent upon a little reflection, and provides further evidence that there is both an objective and a subjective sense of understanding. Follow carefully: When I say "I only SEEM to understand," I mean objective understanding, not subjective, i.e., "I do feel a (subjective) sense of understanding but I can't provide the behavioral evidence of objective understanding, so I don't relly understand in the objective sense." Subjective understanding, on the other hand, is as certain as subjective pain (which happens to be the only kind of pain -- the objective side of pain is the tissue-damage story, and we rightly don't call that "pain"). You can't say "I only SEEM to feel a (subjective) sense of understanding (but I don't realy understand in the subjective sense)" any more than you can say "I only SEEM to feel pain (but not really)." Another point: I said that subjective understanding was PRIMARY for the issues about mind-modeling under discussion here. In the human case, the subjective sense of understanding and the evidence for objective understanding tend to swing together in the vast majority of instances. The correlation between them is not perfect, as the problem cases already discussed -- S without O and O without S in us -- indicate, but this is not relevant because these occasional dissociations all occur in US, in whom the primary S is not in doubt. Symbol crunchers can't be granted minds on the strength of our occasional mental lapses! Perhaps if people (or objects) habitually went around emitting coherent glossolalic discourse in foreign languages ("speaking in tongues") that they claimed (in English) not to understand, or if they emitted nothing but jargonaphasia that they kept feeling fervently to be full of meaning, things might look a little different, but that's not the way it is; S and O are quite tightly coupled, and S is clearly primary. In fact, in a world without S, what would it even MEAN to ask whether or not an event or a performance by a device "really" involved O ("objective understanding")? It seems to me all you'd have would be events and performances that could be "interpreted" by people with S as being instances of O. But why bother? And if there were no people with S at all, the whole problem of O seems to vanish altogether, leaving only a world of objects, events and performances. (To a methodological epiphenomenalist like me, it's a profound puzzle why the world ISN'T in fact like this -- why there should be any S at all.) The foregoing, let me repeat, was not a "linguistic" analysis. I simply tried to remind everyone about what we all mean by pain and understanding, and on what experiences this is based. I have not had to be hypothetical or paradoxical here. Everyone knows the difference between the subjective sense of understanding in ourselves and the objective evidence of it in ourselves and others; everyone knows the difference between understanding English and not understanding Chinese. But watch the torrent of strained sci-fi that is again going to well up by way of quarreling with the obvious in subsequent postings... " [No, it's not true that we can be] "WRONG (very wrong) about (2) " [objective] but not about " (1) [subjective] [understanding]. " If this were so... 'I thought I understood, but I was wrong' would be " self-contradictory. As I said above, there are two senses of understanding, subjective and objective. The above statement could be paraphrased: "I thought I understood it in the objective sense; it turns out I only understood it in the subjective sense," i.e., it only FELT AS IF I understood it. But in Searle's room the ISSUE is whether there's any mind there feeling understanding at all (or feeling anything, for that matter) rather than just a body that's ACTING AS IF it understood (i.e., that can be interpreted -- or misinterpreted -- as understanding by people who do have understanding). Ceterum sentio: This is not a linguistic matter. " Pardon me if I implied that only philosophers do second-rate linguistics. I won't make the obvious repartee, but will just repeat that these are not linguistic matters... [In a later posting Lee mixes up the syntax of the (putative) symbolic "language of thought" -- whose existence and nature is what is at issue here -- and the syntax of natural languages: Not the same issue, Greg.] -- Stevan Harnad INTERNET: harnad@confidence.princeton.edu harnad@princeton.edu srh@flash.bellcore.com harnad@elbereth.rutgers.edu harnad@princeton.uucp BITNET: harnad@pucc.bitnet CSNET: harnad%princeton.edu@relay.cs.net (609)-921-7771