[Humanist] 31.339 different from the sum of its parts

Humanist Discussion Group willard.mccarty at mccarty.org.uk
Fri Oct 6 08:16:30 CEST 2017


                 Humanist Discussion Group, Vol. 31, No. 339.
            Department of Digital Humanities, King's College London
                       www.digitalhumanities.org/humanist
                Submit to: humanist at lists.digitalhumanities.org



        Date: Thu, 5 Oct 2017 13:24:15 +0200
        From: Tim Smithers <tim.smithers at cantab.net>
        Subject: Re:  31.335 different from the sum of its parts?
        In-Reply-To: <20171005051657.8F2CC7EC9 at s16382816.onlinehome-server.info>

Dear Willard,

This may well be a distraction from what you really ask.

Machines, including computers, are, I think, best understood
as level hierarchies, and not as ecosystems.  Complicated
machines, like computers, are designed and built this way.
Here ...

 "... A level consists of a medium that is to be processed,
  components that provide primitive processing, laws of
  composition that permit components to be assembled into
  systems, and laws of behavior that determine how system
  behavior depends on the component behavior and the structure
  of the system. ..." (p 95)

These words are taken from

 The Knowledge Level by Allen Newell, 1982
 PDF available here:
 <http://cs.uns.edu.ar/~grs/InteligenciaArtificial/Allen%20Newell%20-%20The%20knowledge%20level.pdf>

This paper is, I think, an example of good writing about how
computers work as machines.  It is a foundational paper in AI,
though it's never been very widely known, let alone well
understood in AI. 

Using his level analysis of the computer as a machine, Newell
introduces a new concept of knowledge: as a capacity for
rational action.  This formed the starting point for work on
Knowledge Modelling and Knowledge Management that came out of
AI in the 1980s and 1990s.  As a concept of knowledge it
differs strongly from the classical notion of knowledge as
"verifiable true belief," used by others who developed
(competing) approaches to Knowledge Management.  A similar
level-based approach is used to design and building database
systems, and David Marr used a level hierarchy closely related
to Newell's, as a basis for understanding vision.  (A no
longer current theory of vision today.)

There's no talk of ecosystems here, nor of "the whole is more
than the sum of its parts," or even "the whole is different
from the sum of its parts."  Levels do hide the details and
complicatedness of some of the machine functioning, thus
allowing us to understand and treat what the machine does, and
can be made to do, in ways completely abstracted from these
details and complications.  This, I would say, is what makes
something a machine, and it necessarily involves combining
components, components that themselves can properly be
considered as machines.  Machines are hierarchically composed
of machines to make different machines.  This is essentially
the same story W Brian Athur presents in his book "The Nature
of Technology what it is and how it evolves" (2009).

An important difference between machines and ecosystems, I
think, is that in machines (composed of machines) functional
dependency is unidirectional, whereas in ecosystems,
functional dependency is often bidirectional.  In machines,
such as computers, the proper working of a machine (at some
level) depends upon the proper working of the machines that
compose it, but the working of these component machines do not
depended upon the proper working of the machine they compose.
Not usually, at least.  And if they do, it usually means the
machines is badly designed, or that something has gone wrong.

The interesting thing about ecosystems is that they are not
composed this way.  Component parts do inter-depend to work,
sometimes as pairs, but also in networks.  This is, of course,
what makes ecosystems harder to study and understand, and what
gives rise to the kinds of complex behaviour sometimes
described using phrase like "the whole is more than the sum of
its parts."

One last thing.  The thoroughly confusing effect of this
phrase, on our understanding of things like ecosystems, is not
"the whole," it's "the sum."  The idea that the combining of
parts involves some sort of summation notion, even only by
analogy, goes against how we see the way the world is put
together, when we really look, and come to understand it.
Summation is, I would say, completely the wrong notion here.
So, it's hardly surprising that we find complexity
complicated.  Ecosystems display compositions of
inter-dependent nonlinear functionality, not summation, which
is a linear unidirectional affair.

Now.  If we network computers together, what do we get?
Ecologies?  Perhaps?  But not ecologies much like we might be
ecologies, I'd suggest.  Networking computers may give us
ecologies of machines.  We, like other living things, appear
to be ecologies made of ecologies, not ecologies made of
machines.  It's only if we insist on seeing natural things as
machines that ecologies might seem to be made of machines, or,
worse, to think that brains are computers.  

Best regards,

Tim

> On 05 Oct 2017, at 07:16, Humanist Discussion Group <willard.mccarty at mccarty.org.uk> wrote:
> 
>                 Humanist Discussion Group, Vol. 31, No. 335.
>            Department of Digital Humanities, King's College London
>                       www.digitalhumanities.org/humanist
>                Submit to: humanist at lists.digitalhumanities.org
> 
> 
> 
>        Date: Wed, 4 Oct 2017 09:20:20 +0100
>        From: Willard McCarty <willard.mccarty at mccarty.org.uk>
>        Subject: the sum of the parts
> 
> 
> Robert Jervis, in System Effects: Complexity in Political and Social 
> Life (Princeton, 1997) and Frank Golley, in The History of the Ecosystem 
> Concept: More than the Sum of the Parts (Yale, 1993), argue, as Jervis 
> writes, that "If we are dealing with a system, the whole is different 
> from, not greater than, the sum of the parts." He cities work across 
> several disciplines, pointing to biology as the starting point of many 
> scholars who have reached this conclusion. What about the disciplines of 
> computing?
> 
> Our machine, comprised of numerous 'black boxes' whose inner dynamic 
> workings are in principle unknowable, would seem to me an example worth 
> our consideration. It was explicitly designed as such an 'ecosystem', 
> though in the language of human neurophysiology (von Neumann 1945). Its 
> user-interface is obviously not more than but different from all the 
> operations which sum to it. (Indeed, from an engineering perspective, 
> those operations are clearly more than what they sum to.) Golley writes 
> that "There was an exact moment of birth" for the concept of an 
> ecosystem, "when the English ecologist Arthur Tansley created the word 
> and presented it in a technical paper" in 1935. Here's the passage:
> 
>> THE ECOSYSTEM
>> [...]
>> Clements' earlier term "biome" for the whole complex of organisms
>> inhabiting a given region is unobjectionable, and for some purposes
>> convenient. But the more fundamental conception is, as it seems to
>> me, the whole system (in the sense of physics), including not only
>> the organism-complex, but also the whole complex of physical factors
>> forming what we call the environment of the biome - the habitat factors
>> in the widest sense. Though the organisms may claim our primary
>> interest, when we are trying to think fundamentally we cannot
>> separate them from their special environment, with which they form
>> one physical system.
>> 
>> It is the systems so formed which, from the point of view of the
>> ecologist, are the basic units of nature on the face of the earth....
>> These ecosystems, as we may call them, are of the most various kinds
>> and sizes.
> (Tansley, "The Use and Abuse of Vegetational Concepts and Terms", 
> Ecology 16.3, p. 299)
> 
> There is the historical question when we extend Robin Gandy's 
> "confluence of ideas in 1936", i.e. "the almost simultaneous appearance 
> in 1936 of several independent characterizations of the notion of 
> effective calculability" (in The Universal Turing Machine, ed. Herken), 
> among which was Turing's, to the systems-thinking of the above. But 
> there's also the question of computer systems as ecologies 
> that include and perhaps now come close to defining us.
> 
> Who has written along such lines?
> 
> Yours,
> WM
> -- 
> Willard McCarty (www.mccarty.org.uk/), Professor emeritus, Department of
> Digital Humanities, King's College London; Adjunct Professor, Western
> Sydney University and North Carolina State University; Editor,
> Interdisciplinary Science Reviews (www.tandfonline.com/loi/yisr20)




More information about the Humanist mailing list