harnad@mind.UUCP (Stevan Harnad) (06/26/87)
In article 208 of comp.ai.digest: JMC@SAIL.STANFORD.EDU (John McCarthy) asks: > I imagine that the alleged point at issue and a few of the positions > taken could be summarized for the benefit of those of us whose > subjective probability that there is a real point at issue is too > low to motivate studying the entire discussion but high enough to > motivate reading a summary. The point at issue concerns how symbols in a symbol-manipulative approach to the modeling of mind can be grounded in something other than more symbols so that their meanings and their connections to objects can be independent of people's interpretations of them. One of the positions taken was that connecting a purely symbolic module to peripheral (transducer/effector) modules in the right way should be all you need to ground the symbols. I suggested that all this is likely to yield is more of the toy models that symbolic AI has produced until now. To get human-scale (Total Turing Test) performance capacity, a bottom-up hybrid nonsymbolic/symbolic system may be needed, one in which the elementary symbols are the names of sensory categories picked out by inductive (possibly connectionist) feature-filters (categorical representations) and invertible analogs of sensory projections (iconic representations). This model is described in "Categorical Perception: The Groundwork of Cognition" (Cambridge University Press 1987, S. Harnad, ed., ISBN 0-521-26758-7). Other alternatives that have been mentioned by others in the discussion included: (1) symbol-symbol "grounding" is already enough and (2) connectionist nets already generate grounded "symbols." If you want the entire file, I've saved it all... -- Stevan Harnad (609) - 921 7771 {bellcore, psuvax1, seismo, rutgers, packard} !princeton!mind!harnad harnad%mind@princeton.csnet harnad@mind.Princeton.EDU
marty1@houdi.UUCP (M.BRILLIANT) (06/26/87)
Will the proponents of the various views described below, and those whose revelant views have not been described below, please stand up? In article <915@mind.UUCP>, harnad@mind.UUCP (Stevan Harnad) writes: > In article 208 of comp.ai.digest: JMC@SAIL.STANFORD.EDU (John McCarthy) > asks: > > > I imagine that the alleged point at issue and a few of the positions > > taken could be summarized ..... > > The point at issue concerns how symbols in a symbol-manipulative > approach to the modeling of mind can be grounded in something other > than more symbols so that their meanings and their connections to > objects can be independent of people's interpretations of them. > ..... One of > the positions taken was that connecting a purely symbolic module to > peripheral (transducer/effector) modules IN THE RIGHT WAY should be > all you need to ground the symbols. Caps mine. Position 1 is that the peripherals and the symbolic module have to be connected in the right way. Harnad's position is that > .... a bottom-up hybrid nonsymbolic/symbolic system may be > needed, one in which the elementary symbols are the names of sensory > categories picked out by inductive (possibly connectionist) feature-filters > (categorical representations) and invertible analogs of sensory projections > (iconic representations)..... This looks like a way to connect periperals to a symbolic module. To the extent that I understand it, I like it, except for the invertibility condition. If it's the right way, it's a special case of position 1. Harnad has called the "right way" of position 1 "top-down," "hard-wired," and other names, to distance himself from it. I'm not sure there are any real proponents of position 1 in such a narrow sense. I support position 1 in the wide sense, and I think Harnad does. > ..... Other alternatives that have been > mentioned by others in the discussion included: (1) symbol-symbol "grounding" > is already enough .... They don't care about the problem, so either they or we can go away. They (and I) want this discussion to go to another newsgroup. > ..... and (2) connectionist nets already generate grounded "symbols." Is that a variant of Harnad's position, i.e., "(possibly connectionist)"? I think the real subject of discussion is the definition of some of the technical terms in Harnad's position, and the identification of which elements are critical and which might be optional? Might some of the disagreement disappear if the definitions were more concrete? M. B. Brilliant Marty AT&T-BL HO 3D-520 (201)-949-1858 Holmdel, NJ 07733 ihnp4!houdi!marty1
harnad@mind.UUCP (Stevan Harnad) (06/27/87)
marty1@houdi.UUCP (M.BRILLIANT) of AT&T Bell Laboratories, Holmdel writes: > But a "physically analog" sensory process (as distinct from a digital > one) can be approximately modeled (to within the noise) by a continuous > transformation. The continuous approximation allows us to regard the > analog transformation as image-forming (iconic). But only the > continuous approximation is invertible. I have no quarrel with this, in fact I make much the same point -- that iconic representations are approximate too -- in the chapter describing the three kinds of representation. Is there any reason for expecting I would object? > the "hybrid" three-layer system... does not have a "symbol-cruncher > hardwired to peripheral modules" because there is a feature extractor > (and classifier) in between. The main point is the presence or > absence of the feature extractor... The symbol-grounding problem > arises because the symbols are discrete, and therefore have to be > associated with discrete objects or classes. Without the feature > extractor, there would be no way to derive discrete objects from the > sensory inputs. The feature extractor obviates the symbol-grounding > problem. The problem certainly is not just that of discrete symbols needing to pick out discrete objects. You are vastly underestimating the problem of sensory categorization, sensory learning, and the relation between lower and higher-order categories. Nor is it obvious that symbol manipulation can still be regarded as just symbol manipulation when the atomic symbols are constrained to be the labels of sensory categories. That's a bottom-up constraint, and symbolic AI normally expects to float down onto its sensors top-down. Imagine if your "setq" statements were constrained by what your elementary symbols were connected to, and their respective causal interrelations with other nonsymbolic sensory representations and their associated labels. > Why does Harnad say "invertibility is a necessary condition > for iconic representations..., NOT for grounding" Because the original statement of mine that you quote was a reply to a query about whether ALL representations had to be invertible for grounding. (It was accompanied by alleged counterexamples -- grounded but noninvertible percepts.) My reply indicated that only iconic ones had to be invertible, but that both iconic and categorical (noninvertible) ones were needed to ground symbols. > Position 1 [on the symbol grounding problem] is that the peripherals > and the symbolic module have to be connected in the right way. Harnad's > position is... a special case of position 1. I'm afraid not. I don't think there will be independent peripheral modules and symbolic modules suitably interconnected in the hybrid device that passes the Total Turing Test. I think a lot of what we consider cognition will be going on in the nonsymbolic iconic and categorical systems (discrimination, categorization, sensory learning and generalization) and that symbol manipulation will be constrained in ways that don't leave it in any way analogous to the notion of an independent functional module, operating on its own terms (as in standard AI), but connected at some critical point with the nonsymbolic modules. When I spoke earlier of the "connections" of the atomic symbols I had in mind something much more complexly interdigitated and interdependent than can be captured by anything that remotely resembles position 1. Position 1 is simply AI's pious hope that a pure "top-down" approach can expect to meet up with a bottom-up one somewhere in between. Mine is not a special case of this; it's a rival. > "...and (2) connectionist nets already generate grounded "symbols." Is > that a variant of Harnad's position, i.e., "(possibly connectionist)"? No. In my model connectionistic processes are just one possible candidate for the mechanism that finds the features that will reliably pick out a learned category. They would just be a component in the categorical representational system. But there are much more ambitious connectionistic views than that, for example, that connectionism can usurp the role of symbolic representations altogether or (worse) that they ARE symbolic (in some yet to be established sense). As far as I'm concerned, the latter would entail a double grounding problem for connectionism, the first to ground its interpretation of its states as symbolic states, and then to ground the interpretations of the symbolic states themselves (which is the standard symbol grounding problem). -- Stevan Harnad (609) - 921 7771 {bellcore, psuvax1, seismo, rutgers, packard} !princeton!mind!harnad harnad%mind@princeton.csnet harnad@mind.Princeton.EDU