Re: Maximally constrained

[Tim Gwinn <***>]

After re-reading the paper, I think I agree with Judith here. (I admit I did not clearly see this before re-reading.) He carefully writes around any assertion that an organism actually is a maximally constrained system. We do not know that they are in fact maximally constrained system. Rather, he argues that we do not know what organisms are, but only what they are like. I would liken it to a reverse-engineering problem where we do not know about how it was put together in order that it became an organism,  we can instead only make conjectures about its construction and organization by way of comparing its known properties and behaviors with properties and behaviors of other known physical or formal systems such that we will hopefully gain some insight into what an organism is; without drawing the further conclusion that if an organism can be like some other system in some ways, then an organism is in fact an instance of such a class of system. Such a conclusion would require extensive and conclusive evidence. In this paper, his focus is on how organisms can be likened to machines, and what are the ramifications of such a comparison.

 

The paper begins:

"No one knows or can express what an organism is. The best we can do is to try to do say what an organism is like. For at least the past four hundred years, the pervasive metaphor in biology has ben that of the machine. However, no one can say what a machine is either; or rather, many people have said many different things over the years; things which are inequivalent or even mutually contradictory....Machines themselves have been variously considered as physical or material systems, as the executors of specific functions, and as mathematical abstractions. The remarks which follow are intended to clarify the sense (if any) in which the organism is like a machine; in the process, we may hope to learn some new things about both; about the extent to which machines can be considered lifelike, and to which organisms are machine-like." [p.107, ital orig]

 

I think that the quote Howard provides...:

"The results of our analysis above suggest that, whatever else may be true of relatively simple but nevertheless highly evolved contemporary cells, they must at least behave physically like maximally (nonholonomically) constrained, programmable mechanical systems." [p.122, ital org]

...which is immediately followed by...:

"Therefore, it appears that a sensible strategy along these lines is to attempt to fabricate such a  system artificially. This would at least give us a conceptually well-defined end-point to aim at." [p.122]

...demonstrates the caution Rosen is using. If behaviorally organisms act like maximally constrained systems, then a logical next step in the comparison process would be to move from a mathematical argument to an empirical study: to try to construct such a system. Then we would know more about the requirements for constructing such system, how such systems work physically and how well these physical realizations compare (or don't compare) to organisms.

 

I think the mathematical argument that organisms are in fact instances of maximally constrained systems is compelling. But it remains unproven empirically to this day; therefore, conclusions drawn from comparisons between the mathematical conclusions and organisms can be equally compelling, but do not serve as proof.  I think that Rosen is making a similar - if subtle - distinction between compelling and unproven.

 

Regards,

Tim

 

> -----Original Message-----
> From: ROSEN Forum [mailto:***]On Behalf Of Judith
> Rosen
> Sent: Tuesday, November 02, 2004 4:56 PM
> To: ***
> Subject: Re: Maximally constrained
>
>
>  He was referring to a machine, as Tim pointed out in his first
> post on this
> whole discussion:
>
>  > Tim Gwinn wrote: Rosen later notes that this is the only kind of
> mechanical system which can accomplish the experimental result
> that Morowitz
> pointed out years earlier: that a bacterial cell could be carefully frozen
> to absolute zero (where all dynamics (all momenta) are removed) and then
> re-thawed (with no real control over the specific imparted
> momenta) and the
> cell could continue to grow.>
>
> My father was not talking about the bacterial cell being a "maximally
> etc....." but rather the only kind of machine that could appear as if...
>
> It is the machine that he referred to as having the "maximal number of
> non-holonomic constraints".
>
> Judith
>
> > Howard Pattee wrote: On p. 414 of Anticipatory Systems, Rosen points out
> that informational constraints are not a part of Newtonian mechanics but
> "can be regarded as arising out of [Newtonian mechanics] by the imposition
> of a sufficient number of non-holonomic constraints." The only problem I
> have is why "a sufficient number" has become "maximal number.">