From owner-cse575-fa07-list@LISTSERV.BUFFALO.EDU Fri Sep 28 11:22:37 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 l8SFMbge020579 for ; Fri, 28 Sep 2007 11:22:37 -0400 (EDT) Received: from front2.acsu.buffalo.edu (upfront.acsu.buffalo.edu [128.205.4.140]) by ares.cse.buffalo.edu (8.13.8/8.13.6) with SMTP id l8SFMQsN070013 for ; Fri, 28 Sep 2007 11:22:26 -0400 (EDT) Received: (qmail 24915 invoked from network); 28 Sep 2007 15:22:21 -0000 Received: from mailscan6.acsu.buffalo.edu (128.205.7.95) by front2.acsu.buffalo.edu with SMTP; 28 Sep 2007 15:22:21 -0000 Received: (qmail 18845 invoked from network); 28 Sep 2007 15:22:02 -0000 Received: from deliverance.acsu.buffalo.edu (128.205.7.57) by front2.acsu.buffalo.edu with SMTP; 28 Sep 2007 15:22:02 -0000 Received: (qmail 9676 invoked from network); 28 Sep 2007 15:21:58 -0000 Received: from listserv.buffalo.edu (128.205.7.35) by deliverance.acsu.buffalo.edu with SMTP; 28 Sep 2007 15:21:58 -0000 Received: by LISTSERV.BUFFALO.EDU (LISTSERV-TCP/IP release 14.5) with spool id 2788404 for CSE575-FA07-LIST@LISTSERV.BUFFALO.EDU; Fri, 28 Sep 2007 11:21:57 -0400 Delivered-To: CSE575-FA07-LIST@LISTSERV.BUFFALO.EDU Received: (qmail 25010 invoked from network); 28 Sep 2007 15:21:49 -0000 Received: from mailscan6.acsu.buffalo.edu (128.205.7.95) by listserv.buffalo.edu with SMTP; 28 Sep 2007 15:21:49 -0000 Received: (qmail 18196 invoked from network); 28 Sep 2007 15:21:40 -0000 Received: from nickelback.cse.buffalo.edu (HELO ?128.205.35.24?) (128.205.35.24) by smtp2.acsu.buffalo.edu with SMTP; 28 Sep 2007 15:21:40 -0000 User-Agent: Thunderbird 1.5.0.10 (X11/20070301) MIME-Version: 1.0 Content-Type: text/plain; charset=ISO-8859-1; format=flowed Content-Transfer-Encoding: 7bit X-UB-Relay: (nickelback.cse.buffalo.edu) X-PM-EL-Spam-Prob: : 7% Message-ID: <46FD1C04.2060209@buffalo.edu> Date: Fri, 28 Sep 2007 11:21:40 -0400 Reply-To: Scott Settembre Sender: Introduction to Cognitive Science From: Scott Settembre Subject: Connectionism vs Symboli :: Smolensky vs Fodor To: CSE575-FA07-LIST@LISTSERV.BUFFALO.EDU Precedence: list List-Help: , List-Unsubscribe: List-Subscribe: List-Owner: List-Archive: X-UB-Relay: (nickelback.cse.buffalo.edu) X-PM-EL-Spam-Prob: : 7% X-DCC-Buffalo.EDU-Metrics: castor.cse.Buffalo.EDU 1336; Body=0 Fuz1=0 Fuz2=0 X-Spam-Status: No, score=-2.2 required=5.0 tests=AWL,BAYES_00 autolearn=ham version=3.1.8 X-Spam-Checker-Version: SpamAssassin 3.1.8 (2007-02-13) on ares.cse.buffalo.edu X-Virus-Scanned: ClamAV 0.90.2/4419/Fri Sep 28 03:36:28 2007 on ares.cse.buffalo.edu X-Virus-Status: Clean Status: R Content-Length: 2849 I have read both Fodor's paper (though I will probably have to read it a few more times before I truly understand all he said) and Smolensky in the Cummings book. For anyone that has not read these, probably best to read the Smolensky one first, then the Fodor, since the Fodor (chapter 16) paper is in response to Smolensky (chapter 17). I am having a hard time figuring out how to understand why Fodor discounts connectionism, especially because it seems like we actually (all) have a working connectionist machine in our head. I believe his argument centers around his belief that SINCE some portions of thought are optimally symbolic-based (for example, reasoning) that ALL of thought is not be connectionist-based (not even visual or auditory?). Or I may be wrong, maybe what I am reading is that Smolensky claims that ALL thought is connectionist based by his model and that Fodor is disagreeing with that? It seems today that we have a good understanding of the strengths and weaknesses of currently implemented ANNs. Although the limitations of current ANNs and how closely they match up with the neural network in our heads (see Churchland in Cummings page 211), may imply there is functionality there that we have still not yet discovered. We now understand why an ANN works (for a non-mathematical understanding, an ANN is like finding the highest peak in a multi-dimensional landscape) from a mathematical perspective (though that may not be how our actual brain works, it may be unnecessary to know in actuality how our brain works, if the brain is just an implementation of the processing of cognition). For some cognition tasks, it seems that a neural network would be a more natural choice. Smolensky did an excellent job of showing in his coffee example, how coffee can be represented by a series of features that are present (or not present) and that it depends on context. Depending on CONTEXT is the key idea here. This contextual deluge is handled very nicely by an ANN, but would make most knowledge-engineers quake at their keyboard if they were needed to code all the different possible contexts in a recognition task in symbolic logic. In addition, ANN's seem to take on the "poverty of the stimulus" quite easily (given enough samples, but not all samples) being able to generalize and categorize. For other tasks, like reasoning, although this can probably be implemented in an ANN, it seems very unnecessary. Our reasoning tasks seem to have neatly evolved over time to quite "nearly" match that of first order logic (FOL). Can FOL be implemented in an ANN? Probably, but it seems like a connectionist approach here (for reasoning) would be purely an implementation, and not address the essense of such cognition. So I guess my confusion comes down to, "why not both?"