In article <5mdc6o$l27$1@news.fas.harvard.edu>, kubo@abel (Tal Kubo) writes: >Manfred Jaeger <jaeger@flamingo.stanford.edu> wrote: >>Matthew P Wiener wrote:
>>> [...] I appeal to the EXTREMELY WELL-KNOWN ubiquity of 0-1 laws >>> to claim that whatever the truth is regarding TM mind evolution, >>> I'm already covered.
>>Ubiquity of 0-1 laws? There are just as many cases where no 0-1 law >>holds as there are 0-1 laws. Add a function symbol to your language, >>and you don't get a 0-1 law any more. Make the probability of >>an edge in your random graph a function of the cardinality >>of the domain, and the validity of a 0-1 law depends on the exact >>form of this function. [...]
So you apply it to a form where it does hold. The mental model isn't forced into some logician's favorite straightjacket, doomed forever to defy probabilistic analysis because of bad notation. Sheesh.
>The problem is much more basic than that. Even granting the most >generous hypothetical 0-1 laws, and sidestepping some of the >technical issues you mention, the facts of life about 0-1 laws >**prevent Matthew's argument from making any sense**.
Sure it makes sense.
>e.g.: a syntactic 0-1 law is out of the question;
Eh? Why not?
> semantic 0-1 laws >won't help;
I don't use them. In fact, I've repeatedly denied that there are relevant semantics due to the lack of physical correlates.
> with probability 1 evolution can't be adequately quantified;
I've only pointed out repeatedly that this is why I don't bother spelling out more detail. Whatever particular model I consider is wrong.
>with probability 1 the limit theory is not recursively enumerable;
So what?
>Goedel's theorem can't possibly be applied. And so on.
I apply it to the base theory, not the limit theory.
>In other words, it's just wall-to-wall bogosity wrapped in jargon.
Your comments obviously are. -- -Matthew P Wiener (weemba@sagi.wistar.upenn.edu)
