Re: Inconsistency, etc.

[Tim Gwinn <***>]

Howard,
I think there exists a happy medium between not mentioning von Neumann at
all and what I consider to be an excessive and repetitive focus on von
Neumann from you.
Tim

> -----Original Message-----
> From: ROSEN Forum [mailto:*** Behalf Of Howard
> Pattee
> Sent: Friday, June 04, 2004 4:34 PM
> To: ***
> Subject: Re: Inconsistency, etc.
>
>
> Howard wrote: I am in full concordance with Rosen, although I may
> prefer slightly different expressions. I also treat von Neumann?s
> and Chaitin?s concepts of complexity as complementary to this one.
>
> Tim replied: Those two are indeed concepts of complexity entirely
> different from, and
> inequivalent to, the concept of Rosennean complexity. This discussion list
> is about Rosennean complexity.
>
> HP: In my opinion, it was also Rosen?s view that to fully
> understand complexity requires more than just discussing his
> model. In any case, I have found that to fully grasp Rosen?s
> concept of complexity one should also understand how it relates
> to other complementary concepts. I think you might find that Von
> Neumann?s views are not so entirely different that they are
> uninstructive. It is important to understand that formal
> inequivalence does not mean conceptual irrelevance. Quite the
> contrary, complementary models must be conceptually relatable,
> otherwise they are useless.
>
> For example, Rosen describes von Neumann?s concept of complexity
> (AS, p. 331) to contrast it with his own, and I find such
> contrasts illuminating. Rosen then uses complexity to define his
> concept of error:
>
> Rosen: ?However, we have also seen that such non-reducible,
> non-conjugate encodings must therefore bifurcate from one
> another. It is exactly the appearance of such bifurcations which
> we showed in the preceding chapter to be the essence of error.?
>
> It was instructive to me to find that von Neumann, for whom error
> was a basic issue, defined one type of error consistent with
> Rosen, as a property of differing descriptions of systems and not
> of systems themselves. In my opinion, his other concept of error
> also relates fundamentally to Rosen?s formalisms, but if you
> prefer, I will not discuss it on this list.
>
> Howard