From owner-cse584-sp07-list@LISTSERV.BUFFALO.EDU Wed Apr 18 11:43:05 2007 Received: from ares.cse.buffalo.edu (ares.cse.Buffalo.EDU [128.205.32.79]) by castor.cse.Buffalo.EDU (8.13.6/8.12.10) with ESMTP id l3IFh4tV021005 for ; Wed, 18 Apr 2007 11:43:04 -0400 (EDT) Received: from front1.acsu.buffalo.edu (coldfront.acsu.buffalo.edu [128.205.6.89]) by ares.cse.buffalo.edu (8.13.6/8.13.6) with SMTP id l3IFh0pe022870 for ; Wed, 18 Apr 2007 11:43:00 -0400 (EDT) Received: (qmail 18332 invoked from network); 18 Apr 2007 15:43:00 -0000 Received: from mailscan1.acsu.buffalo.edu (128.205.6.133) by front1.acsu.buffalo.edu with SMTP; 18 Apr 2007 15:43:00 -0000 Received: (qmail 18254 invoked from network); 18 Apr 2007 15:42:59 -0000 Received: from deliverance.acsu.buffalo.edu (128.205.7.57) by front3.acsu.buffalo.edu with SMTP; 18 Apr 2007 15:42:59 -0000 Received: (qmail 23753 invoked from network); 18 Apr 2007 15:42:49 -0000 Received: from listserv.buffalo.edu (128.205.7.35) by deliverance.acsu.buffalo.edu with SMTP; 18 Apr 2007 15:42:49 -0000 Received: by LISTSERV.BUFFALO.EDU (LISTSERV-TCP/IP release 14.5) with spool id 4824149 for CSE584-SP07-LIST@LISTSERV.BUFFALO.EDU; Wed, 18 Apr 2007 11:42:49 -0400 Delivered-To: CSE584-SP07-LIST@LISTSERV.BUFFALO.EDU Received: (qmail 5039 invoked from network); 18 Apr 2007 15:42:49 -0000 Received: from mailscan7.acsu.buffalo.edu (128.205.6.158) by listserv.buffalo.edu with SMTP; 18 Apr 2007 15:42:49 -0000 Received: (qmail 4068 invoked from network); 18 Apr 2007 15:42:49 -0000 Received: from pollux.cse.buffalo.edu (128.205.35.2) by smtp5.acsu.buffalo.edu with SMTP; 18 Apr 2007 15:42:49 -0000 Received: from pollux.cse.buffalo.edu (ag33@localhost [127.0.0.1]) by pollux.cse.buffalo.edu (8.12.10+Sun/8.12.9) with ESMTP id l3IFgmKF005518 for ; Wed, 18 Apr 2007 11:42:48 -0400 (EDT) Received: (from ag33@localhost) by pollux.cse.buffalo.edu (8.12.10+Sun/8.12.8/Submit) id l3IFgmtM005517; Wed, 18 Apr 2007 11:42:48 -0400 (EDT) MIME-Version: 1.0 Content-Type: TEXT/PLAIN; charset=US-ASCII X-UB-Relay: (pollux.cse.buffalo.edu) X-PM-EL-Spam-Prob: : 7% Message-ID: Date: Wed, 18 Apr 2007 11:42:48 -0400 Reply-To: Albert Goldfain Sender: "Philosophy of Computer Science, Spring 2007" From: Albert Goldfain Subject: comments on syntactic semantics To: CSE584-SP07-LIST@LISTSERV.BUFFALO.EDU Precedence: list List-Help: , List-Unsubscribe: List-Subscribe: List-Owner: List-Archive: X-UB-Relay: (pollux.cse.buffalo.edu) X-DCC-Buffalo.EDU-Metrics: castor.cse.Buffalo.EDU 1029; Body=0 Fuz1=0 Fuz2=0 X-Spam-Status: No, score=-2.5 required=5.0 tests=AWL,BAYES_00 autolearn=unavailable version=3.1.7 X-Spam-Checker-Version: SpamAssassin 3.1.7 (2006-10-05) on ares.cse.buffalo.edu X-Virus-Scanned: ClamAV 0.88.6/3120/Wed Apr 18 09:00:32 2007 on ares.cse.buffalo.edu X-Virus-Status: Clean Status: R Content-Length: 3664 Just a few things I found interesting and jotted down during the syntactic semantics slideshow in class: (1) On internalization as a union of markers and meanings In syntactic semantics, a syntactic domain of markers S and a semantic domain of meanings S' is unioned to form a set S'' (= S U S') of *markers*. The presupposition is that members of S' (i.e., meanings) can serve as markers in S'' (I actually believe that, once internalized, they can). But this does not follow from just a set theoretic union. Consider: Let R be the set of characters in the roman alphabet {a,b,...,z} Let G be the set of characters in the greek alphabet {alpha,beta,...omega} It is not the case that R U G is a set of characters in the roman alphabet...the kinds of things in the new set are not the same. I believe the fancy mathy way of saying this is that the roman alphabet is not closed under union with arbitrary alphabets. Thus, I believe a hidden premise of the union approach to internalization is that internalized meanings *are* markers (i.e., suitable to be the sort of thing semantics forms relations between). This is an interesting issue in AI researc (google "symbol grounding problem" for some approaches to internalization). (2) On delayed solipsism and latent understanding A cognitive agent may have a sensory input that yields an association between the term "tree" and the internalized representation of the tree (height, color, etc)... let's just label this association ("tree", internalized(tree)) Importantly, both elements in this ordered pair are *markers*. Now suppose some agent A has made this association but then loses its sensory input entirely. Well it still has the association ("tree", internalized(tree))...so it still has a syntactic semantics for "tree". But what is to distinguish this from another agent B that was always solipsistic and has a hard-coded association ("tree",fake-internalized(tree)) ...both are just relations between markers? Does agent B also have a syntactic semantics for "tree"? (yes, I do note the resemblance between agent A and Hellen Keller :)). (3) On multiple realizability of cognition Just something interesting to point out. Since multiple realizability relies on two steps: (a) abstracting A from I, and (b) re-implementing I' from A, there are two "points-of-failure" for realizing cognition in a computational agent: (a) the abstraction of cognition from human cognition was not a good one (and that is for cognitive scientists to decide). and (b) the abstraction of cognition from human cognition was okay but the re-implementation of cognition in a computational agent (and that is for AI researchers to decide). It had not occured to me that this might fail because of failed (or weak) abstractions (i.e., someone else to blame other than AI researchers :-P). best abstract of cognition / \ / \ / \ / \ / \ stronger abstract \ / \ \ / \ \ weak abstract of cognition \ / \ \ \ / \ \ \ human cognition weak Impl stronger Impl best Impl (4) Methodological solipsism and contex-independent sense If we take the viewpoint of methodological solipsism can we even keep talking about a context-independent sense? What does a methodological solipsist think an ontologist studies? From owner-cse584-sp07-list@LISTSERV.BUFFALO.EDU Wed Apr 18 11:49:44 2007 Received: from ares.cse.buffalo.edu (ares.cse.Buffalo.EDU [128.205.32.79]) by castor.cse.Buffalo.EDU (8.13.6/8.12.10) with ESMTP id l3IFni6B021420 for ; Wed, 18 Apr 2007 11:49:44 -0400 (EDT) Received: from front3.acsu.buffalo.edu (upfront.acsu.buffalo.edu [128.205.4.140]) by ares.cse.buffalo.edu (8.13.6/8.13.6) with SMTP id l3IFncL3023362 for ; Wed, 18 Apr 2007 11:49:38 -0400 (EDT) Received: (qmail 24283 invoked from network); 18 Apr 2007 15:49:38 -0000 Received: from mailscan6.acsu.buffalo.edu (128.205.7.95) by front3.acsu.buffalo.edu with SMTP; 18 Apr 2007 15:49:38 -0000 Received: (qmail 24044 invoked from network); 18 Apr 2007 15:49:37 -0000 Received: from deliverance.acsu.buffalo.edu (128.205.7.57) by front3.acsu.buffalo.edu with SMTP; 18 Apr 2007 15:49:36 -0000 Received: (qmail 1112 invoked from network); 18 Apr 2007 15:49:28 -0000 Received: from listserv.buffalo.edu (128.205.7.35) by deliverance.acsu.buffalo.edu with SMTP; 18 Apr 2007 15:49:28 -0000 Received: by LISTSERV.BUFFALO.EDU (LISTSERV-TCP/IP release 14.5) with spool id 4824273 for CSE584-SP07-LIST@LISTSERV.BUFFALO.EDU; Wed, 18 Apr 2007 11:49:28 -0400 Delivered-To: CSE584-SP07-LIST@LISTSERV.BUFFALO.EDU Received: (qmail 8121 invoked from network); 18 Apr 2007 15:49:28 -0000 Received: from mailscan6.acsu.buffalo.edu (128.205.7.95) by listserv.buffalo.edu with SMTP; 18 Apr 2007 15:49:28 -0000 Received: (qmail 27578 invoked from network); 18 Apr 2007 15:49:27 -0000 Received: from pollux.cse.buffalo.edu (128.205.35.2) by smtp1.acsu.buffalo.edu with SMTP; 18 Apr 2007 15:49:27 -0000 Received: from pollux.cse.buffalo.edu (ag33@localhost [127.0.0.1]) by pollux.cse.buffalo.edu (8.12.10+Sun/8.12.9) with ESMTP id l3IFnRKF005602 for ; Wed, 18 Apr 2007 11:49:27 -0400 (EDT) Received: (from ag33@localhost) by pollux.cse.buffalo.edu (8.12.10+Sun/8.12.8/Submit) id l3IFnQjP005601; Wed, 18 Apr 2007 11:49:26 -0400 (EDT) References: MIME-Version: 1.0 Content-Type: TEXT/PLAIN; charset=US-ASCII X-UB-Relay: (pollux.cse.buffalo.edu) X-PM-EL-Spam-Prob: : 7% Message-ID: Date: Wed, 18 Apr 2007 11:49:26 -0400 Reply-To: Albert Goldfain Sender: "Philosophy of Computer Science, Spring 2007" From: Albert Goldfain Subject: Re: comments on syntactic semantics To: CSE584-SP07-LIST@LISTSERV.BUFFALO.EDU In-Reply-To: Precedence: list List-Help: , List-Unsubscribe: List-Subscribe: List-Owner: List-Archive: X-UB-Relay: (pollux.cse.buffalo.edu) X-DCC-Buffalo.EDU-Metrics: castor.cse.Buffalo.EDU 1029; Body=0 Fuz1=0 Fuz2=0 X-Spam-Status: No, score=-2.5 required=5.0 tests=AWL,BAYES_00 autolearn=unavailable version=3.1.7 X-Spam-Checker-Version: SpamAssassin 3.1.7 (2006-10-05) on ares.cse.buffalo.edu X-Virus-Scanned: ClamAV 0.88.6/3120/Wed Apr 18 09:00:32 2007 on ares.cse.buffalo.edu X-Virus-Status: Clean Status: R Content-Length: 3931 Ok...I found some typos in the original...read new post below: On Wed, 18 Apr 2007, Albert Goldfain wrote: > Just a few things I found interesting and jotted down during the syntactic > semantics slideshow in class: > > (1) On internalization as a union of markers and meanings > > In syntactic semantics, a syntactic domain of markers S and a semantic > domain of meanings S' is unioned to form a set S'' (= S U S') of > *markers*. > > The presupposition is that members of S' (i.e., meanings) can serve as > markers in S'' (I actually believe that, once internalized, they can). > But this does not follow from just a set theoretic union. Consider: > > Let R be the set of characters in the roman alphabet {a,b,...,z} > Let G be the set of characters in the greek alphabet {alpha,beta,...omega} > > It is not the case that R U G is a set of characters in the roman > alphabet...the kinds of things in the new set are not the same. I believe > the fancy mathy way of saying this is that the roman alphabet is not > closed under union with arbitrary alphabets. > > Thus, I believe a hidden premise of the union approach to internalization > is that internalized meanings *are* markers (i.e., suitable to be the sort > of thing syntax forms relations between). This is an interesting issue > in AI researc (google "symbol grounding problem" for some approaches to > internalization). > > (2) On delayed solipsism and latent understanding > A cognitive agent may have a sensory input that yields an association > between the term "tree" and the internalized representation of the tree > (height, color, etc)... let's just label this association > > ("tree", internalized(tree)) > > Importantly, both elements in this ordered pair are *markers*. > Now suppose some agent A has made this association but then loses its > sensory input entirely. Well it still has the association ("tree", > internalized(tree))...so it still has a syntactic semantics for "tree". > But what is to distinguish this from another agent B that was always > solipsistic and has a hard-coded association > > ("tree",fake-internalized(tree)) > > ...both are just relations between markers? Does agent B also have a > syntactic semantics for "tree"? > > (yes, I do note the resemblance between agent A and Hellen Keller :)). > > (3) On multiple realizability of cognition > Just something interesting to point out. Since multiple realizability > relies on two steps: (a) abstracting A from I, and (b) re-implementing I' > from A, there are two "points-of-failure" for realizing cognition in a > computational agent: (a) the abstraction of cognition from human cognition > was not a good one (and that is for cognitive scientists to decide). and > (b) the abstraction of cognition from human cognition was okay but the > re-implementation of cognition in a computational agent was not a > good one(and that is for AI researchers to decide). > > It had not occured to me that this might fail because of failed (or weak) > abstractions (i.e., someone else to blame other than AI researchers :-P). > > best abstract of cognition > / \ > / \ > / \ > / \ > / \ > stronger abstract \ > / \ \ > / \ \ > weak abstract of cognition \ > / \ \ \ > / \ \ \ > human cognition weak Impl stronger Impl best Impl > > (4) Methodological solipsism and contex-independent sense > If we take the viewpoint of methodological solipsism can we even keep > talking about a context-independent sense? What does a methodological > solipsist think an ontologist studies? > From owner-cse584-sp07-list@LISTSERV.BUFFALO.EDU Wed Apr 18 20:17:54 2007 Received: from ares.cse.buffalo.edu (ares.cse.Buffalo.EDU [128.205.32.79]) by castor.cse.Buffalo.EDU (8.13.6/8.12.10) with ESMTP id l3J0Hs2l010411 for ; Wed, 18 Apr 2007 20:17:54 -0400 (EDT) Received: from front3.acsu.buffalo.edu (upfront.acsu.buffalo.edu [128.205.4.140]) by ares.cse.buffalo.edu (8.13.6/8.13.6) with SMTP id l3J0HoSh052546 for ; Wed, 18 Apr 2007 20:17:50 -0400 (EDT) Received: (qmail 9802 invoked from network); 19 Apr 2007 00:17:50 -0000 Received: from mailscan7.acsu.buffalo.edu (128.205.6.158) by front3.acsu.buffalo.edu with SMTP; 19 Apr 2007 00:17:50 -0000 Received: (qmail 9788 invoked from network); 19 Apr 2007 00:17:49 -0000 Received: from deliverance.acsu.buffalo.edu (128.205.7.57) by front3.acsu.buffalo.edu with SMTP; 19 Apr 2007 00:17:49 -0000 Received: (qmail 12454 invoked from network); 19 Apr 2007 00:17:39 -0000 Received: from listserv.buffalo.edu (128.205.7.35) by deliverance.acsu.buffalo.edu with SMTP; 19 Apr 2007 00:17:39 -0000 Received: by LISTSERV.BUFFALO.EDU (LISTSERV-TCP/IP release 14.5) with spool id 4840242 for CSE584-SP07-LIST@LISTSERV.BUFFALO.EDU; Wed, 18 Apr 2007 20:17:39 -0400 Delivered-To: CSE584-SP07-LIST@LISTSERV.BUFFALO.EDU Received: (qmail 21557 invoked from network); 19 Apr 2007 00:17:36 -0000 Received: from mailscan4.acsu.buffalo.edu (128.205.6.136) by listserv.buffalo.edu with SMTP; 19 Apr 2007 00:17:36 -0000 Received: (qmail 9124 invoked from network); 19 Apr 2007 00:17:35 -0000 Received: from castor.cse.buffalo.edu (128.205.32.14) by smtp5.acsu.buffalo.edu with SMTP; 19 Apr 2007 00:17:35 -0000 Received: from castor.cse.Buffalo.EDU (rapaport@localhost [127.0.0.1]) by castor.cse.Buffalo.EDU (8.13.6/8.12.10) with ESMTP id l3J0HZYh010390 for ; Wed, 18 Apr 2007 20:17:35 -0400 (EDT) Received: (from rapaport@localhost) by castor.cse.Buffalo.EDU (8.13.6/8.12.9/Submit) id l3J0HZjp010389 for CSE584-SP07-LIST@LISTSERV.BUFFALO.EDU; Wed, 18 Apr 2007 20:17:35 -0400 (EDT) X-UB-Relay: (castor.cse.buffalo.edu) X-PM-EL-Spam-Prob: : 7% Message-ID: <200704190017.l3J0HZjp010389@castor.cse.Buffalo.EDU> Date: Wed, 18 Apr 2007 20:17:35 -0400 Reply-To: "William J. Rapaport" Sender: "Philosophy of Computer Science, Spring 2007" From: "William J. Rapaport" Subject: Re: comments on syntactic semantics To: CSE584-SP07-LIST@LISTSERV.BUFFALO.EDU Precedence: list List-Help: , List-Unsubscribe: List-Subscribe: List-Owner: List-Archive: X-UB-Relay: (castor.cse.buffalo.edu) X-DCC-Buffalo.EDU-Metrics: castor.cse.Buffalo.EDU 1029; Body=0 Fuz1=0 Fuz2=0 X-Spam-Status: No, score=-2.5 required=5.0 tests=AWL,BAYES_00 autolearn=unavailable version=3.1.7 X-Spam-Checker-Version: SpamAssassin 3.1.7 (2006-10-05) on ares.cse.buffalo.edu X-Virus-Scanned: ClamAV 0.88.6/3125/Wed Apr 18 16:44:38 2007 on ares.cse.buffalo.edu X-Virus-Status: Clean Status: R Content-Length: 1680 Albert wrote: | From: Albert Goldfain | Subject: Re: comments on syntactic semantics | To: CSE584-SP07-LIST@LISTSERV.BUFFALO.EDU | ... | > | > Thus, I believe a hidden premise of the union approach to internalization | > is that internalized meanings *are* markers (i.e., suitable to be the sort | > of thing syntax forms relations between). Yes; I discuss this in: Rapaport, William J. (2006), "How Helen Keller Used Syntactic Semantics to Escape from a Chinese Room", Minds and Machines 16(4): 381-436, Sect. "Thesis 1" http://www.cse.buffalo.edu/~rapaport/575/F01/turingtest.chrmarg.html#rapaport06-hk | > ...both are just relations between markers? Does agent B also have a | > syntactic semantics for "tree"? Yes. See below. | > (yes, I do note the resemblance between agent A and Hellen Keller :)). See above :-) | > It had not occured to me that this might fail because of failed (or weak) | > abstractions (i.e., someone else to blame other than AI researchers :-P). Interesting! | > (4) Methodological solipsism and contex-independent sense | > If we take the viewpoint of methodological solipsism can we even keep | > talking about a context-independent sense? I'm not sure what you mean by a "context-independent sense"; sense of what? And independent of what context? | > What does a methodological | > solipsist think an ontologist studies? Interesting question. Perhaps an ontologist studies Kantian phenomena? The classic source on this is: Fodor, Jerry A. (1980), "Methodological Solipsism Considered as a Research Strategy in Cognitive Psychology", Behavioral and Brain Sciences 3: 63-109. Unfortunately, not on line.