Re: RT list: Is it right to think about RT as a materialising theoretical attempt?

Dear All,
Just to say that as a paid-up member of the RT list hoi polloi, I very much look forward to being able to follow the discussion on this very interesting post by Jose Luis. Please share ! (rather than just posting to him.) many thanks

Robin Setton

Sent from my iPad

On 24 sept. 2012, at 09:57, Jose Luis Guijarro Morales <joseluis.guijarro@uca.es> wrote:

> Hi there!
>
> I am having a debate on the "materialistic" tendency of the theory that Sperber and Wilson put forward in "Relevance. Communication and Cognition". There is a guy, Malfet, who seems to be very conversant in computational theory, who thinks that S&W failed in their attempt. I have been arguing with him for a long time now, but, as I am not, in any sense, aware of computational theory, I seem to run short of arguments and he keeps beating me up in his theoretical frame, since I have no possibility to respond in THAT frame. To show you an example, let me paste one of his last postings:
>
> ______________________________________________
>
> "MALFET WROTE: If for Relevance Theory a Turing machine is simply a way to describe something materially without needing to bother with futzy underlying physical details, the notion is being grossly misapplied to disasterous effect for the integrity of the theory.
>
> In a limited sense, it is indeed true that understanding the underlying materiality of a computational device is neither necessary nor sufficient for understanding its functional process. But, that's not because these features have been subsumed by a less demanding descriptive logic…rather they have been displaced into another ontological organization that is equally demanding (though in different ways).
>
> To understand this, we need to make a distinction between formal systems and physical systems. I described this earlier in the thread, but I'll do it again here because the relevant emphases are now somewhat different.
>
> Chess is a quintessentially formal system. The universe of a chess game is composed of exactly 238 bits worth of information: 64 squares, each bearing either one of twelve classes of piece or no piece, plus a single boolean value to declare whether it is currently black's turn or white's turn. We can perfectly describe a chess game in these terms because it, as a system, is self-reducing. Whether the white queen is 3mm left of center on its square or 1mm right simply doesn't matter because the information is architectured into discrete categories. There is a space called e6 and a space called e7, but there is no e6.3.
>
> The power of a formal system is that it can be implemented on a Turing machine. If we were to download an online chess program and play a game together, we wouldn't say that we were "simulating" chess. It would in fact be chess, because as a formal system its materiality exists purely as an expression of formality. It's an abstraction to the core, with no other "real" basis to describe it. I could build a chess game set out of moldy bits of cheese in the back of my fridge and it would still be chess (though perhaps smellier than desired). The true chess savants can supposedly play entire games over the phone, with no need for physical representations of the games at all.
>
> This is in sharp contrast to a physical system, such as the movement of the planets. As with chess, I think there is no doubt that these orbits are the result of a fully determinist process. The movement of the planets is governed entirely by the laws of physics, which are fixed and knowable. However, the system is not self-reducing. To *perfectly* describe the course of the planets, we would need to know the position, quality, and velocity of every atom in the universe. Of course, that's not likely necessary for practical purposes. A far less precise set of measurements and parameters would no doubt produce a description that is as useful as we need it to be.
>
> A physical system can be simulated on a Turing machine, but it cannot be implemented. Even if I were to write an incredibly sophisticated algorithm describing the movement of the planets, it would be absurd to argue that I had actually created planetary motion.
>
> So which are S&W claiming? Is pragmatic function a formal system to be implemented or a physical system to be described? I have no idea, and I've asked this question a few times in this thread. The fact that I still don't know after having read Sperber and Wilson's book and several of their articles is deeply perplexing to me.
>
> As S&W articulate their relationship to the rest of the field, they seem to be marching under the colors of formalism. The achievement of Turing was to imagine a way for formal systems to manifest in the real world, and likewise Dan Sperber's indebtedness to him would seem to suggest that he sees himself moving in this direction. And, if this is the route they intend to take, they'll be in good company. Chomsky is certainly claiming that syntax is a formal system. Fodor in his early works especially was claiming that much of cognition works through formal processes. Any talk of modules in the literature has traditionally been part of formal cognitivism, and anyone using the term computational (just about everyone these days) is unambiguously flagging themselves as interested in formal systems.
>
> But, as you say [he means, ME, of course], it would be much easier to treat all of this as a purely descriptive account of a physical system, and perhaps that is what Sperber and Wilson intended all along. If that's the case, I would suggest that their language is deeply misleading. All of Professor Sperber's talk about Turing is wholly irrelevant, and certainly there is nothing cognitivist about such an approach. Nevertheless, it's a viable position, to be sure. However, a purely descriptive formalism like this one is in no sense materialist. The only materiality of a physicalist system would be the synapses of the human brain (or, as D.S. points out, something reducible to them). Anything else would be an abstraction, an idealized system that can be useful descriptively but that cannot be said to exist in any material sense (much like, I would argue, Grice's maxims).
>
> This is the rock/hardplace in which RT currently finds itself. If RT stakes out pragmatics as a formal system, it needs to commit itself to computationally accessible ontologies. If RT instead argues that pragmatics is a physical system, the theory loses its claim on materialism. That's a tough position to be in.
>
> So which do S&W actually claim? Indeed, I would love to know! As it stands, it seems they want to reap the benefits of both worlds while paying the price of neither. "
> _____________________________________________
> (The underlining is mine)
>
>
> The whole debate may be followed here, although there are many off topic postings and a lot of repetitions, for we are a few participants in it. Maybe it's too long for people not fully interested in the topic, but I think it's fair to paste its link here just in case someone would like to be aware of it (and even participate!).
>
> http://www.lingforum.com/forum/viewtopic.php?t=5462
>
> Any hints that would help me in the debate? Have I grossly misunderstood RT real achievement as a scientific step forward?
>
>
> José Luis Guijarro
> Facultad de Filosofía y Letras
> Universidad de Cádiz
> 11002 Cádiz, España (Spain)
> tlf: (34) 956-011.613
> fax: (34) 956-015.505
Received on Mon Sep 24 09:59:54 2012