paper3

Best algorithm for LZW ..C language

that the resulting answers are indistinguishable from those generated by a
native Chinese speaker, does the subject ``understand'' the stories?
To summarize a large and complex debate in a few words, Searle says no; while
many AI researchers say yes, or at least that the subject-plus-rules system
understands.
.pp
Searle states his thesis succinctly:  ``such intentionality as computers
appear to have is solely in the minds of those who program them and those who
use them, those who send in the input and those who interpret the output''.
And the antithesis could be caricatured as
``maybe, but does it \fImatter?\fR''.
Those who find the debate frustrating can always, with
Sloman & Croucher (1981), finesse the issue:  \c
``Ultimately, the decision whether to say such machines have motives is a
\fImoral\fR decision, concerned with how we ought to treat them''.
.[
Sloman Croucher 1981 robots emotions
.]
.sh "Autopoiesis \(em natural and artificial"
.pp
Autonomy is a striking feature of biological systems.
Not surprisingly, some biologists have made strenuous attempts to articulate
what it means to them; to pin it down, formalize and study it in a
system-theoretic context.
However, this work is obscure and difficult to assess in terms of its
predictive power (which must be the fundamental test of any theory).
Even as a descriptive theory its use is surrounded by controversy.
Consequently this section attempts to give the flavor of the endeavor, relying
heavily on quotations from the major participants in the research, and goes on
to describe some practical computer systems which appear to satisfy the
criteria biologists have identified for autonomy.
.rh "Homeostasis."
People have long expressed wonder at how a living organism maintains its
identity in the face of continuous change.
.sp
.BQ
In an open system, such as our bodies represent, compounded of unstable
material and subjected continuously to disturbing conditions, constancy is
in itself evidence that agencies are acting or ready to act, to maintain this
constancy.
.FQ "Cannon, 1932"
.sp
.[
Cannon 1932 wisdom of the body
.]
Following Cannon, Ashby (1960) developed the idea of ``homeostasis'' to
account for this remarkable ability to preserve stability under conditions of
change.
.[
Ashby 1960 design for a brain
.]
The word has now found its way into North American dictionaries, eg Webster's
.sp
.BQ
Homeostasis is the tendency to maintain, or the maintenance of, normal,
internal stability in an organism by coordinated responses of the organ
systems that automatically compensate for environmental changes.
.FQ
.sp
The basis for homeostasis was adaptation by the organism.
When change occurred, the organism adapted to it and thus preserved its
constancy.
.sp
.BQ
A form of behavior is \fIadaptive\fR if it maintains the essential variables
within physiological limits.
.FQ "Ashby, 1960, p. 58"
.sp
The ``essential variables'' are closely related to survival and linked
together dynamically so that marked changes in any one soon lead to changes in
the others.
Examples are pulse rate, blood pressure, body temperature, number of
bacteria in the tissue, etc.
Ashby went so far as to construct an artifact, the ``Homeostat'', which
exhibits this kind of ultrastable equilibrium.
.pp
Homeostasis emphasizes the stability of biological systems under external
change.
Recently, a concept called ``autopoiesis'' has been identified, which
captures the essence of biological autonomy in the sense of stability or
preservation of identity under \fIinternal\fR change
(Maturana, 1975; Maturana & Varela, 1980; Varela, 1979; Zeleny, 1981).
.[
Maturana 1975 organization of the living
.]
.[
Maturana Varela 1980 autopoiesis
.]
.[
Varela 1979 biological autonomy
.]
.[
Zeleny 1981 Editor Autopoiesis  a theory of living organization
.]
This has aroused considerable interest, and controversy, in the system
theoretic research community.
.rh "Autopoiesis."
The neologism ``autopoiesis'' means literally ``self-production'', and a
striking example occurs in living cells.
These complex systems produce and synthesize macromolecules of proteins,
lipids, and enzymes, and consist of about $10 sup 5$ macromolecules.
The entire population of a given cell is renewed about $10 sup 4$ times
during its lifetime (Zeleny, 1981a).
.[
%A Zeleny, M.
%D 1981a
%T What is autopoiesis?
%E M.Zeleny
%B Autopoiesis:  a theory of living organization
%I North Holland
%C New York
%P 4-17
.]
Despite this turnover of matter, the cell retains its distinctiveness and
cohesiveness \(em in short, its \fIautonomy\fR.
This maintenance of unity and identity of the whole, despite the fact that
all the while components are being created and destroyed, is called
``autopoiesis''.
A concise definition is
.sp
.BQ
Autopoiesis is the capability of living systems to develop and maintain
their own organization.
The organization that is developed and maintained is identical to that
performing the development and maintenance.
.FQ "Andrew, 1981, p. 156"
.sp
.[
Andrew 1981
.]
Other authors (eg Maturana & Varela, 1980; Zeleny, 1981a) add a corollary:
.sp
.BQ
a topological boundary emerges as a result of the processes [of development
and maintenance].
.FQ "Zeleny, 1981a, p. 6"
.sp
This emphasizes the train of thought ``from self-production to identity''
that seems to underly much of the autopoietic literature.
.pp
Operating as a system which produces or renews its own components, an
autopoietic system continuously regenerates its own organization.
It does this in an endless turnover of components and despite inevitable
perturbations.
Therefore autopoiesis is a form of homeostasis which has its own
organization as the fundamental variable that remains constant.
The principal fascination of the concept lies in the self-reference it
implies,
This has stimulated a theoretical formulation of the notion of circularity or
self-reference in Varela's (1975) extension of Brown's
``calculus of distinctions'' (Brown, 1969).
.[
%A Varela, F.J.
%D 1975
%K *
%T A calculus for self-reference
%J Int J General Systems
%V 2
%N 1
%P 5-24
.]
.[
Brown 1969 Laws of Form
.]
Along with other work on self-reference (eg Hofstadter, 1979), this
has an esoteric and obscure, almost mystical, quality.
.[
Hofstadter 1979 Godel Escher Bach
.]
While it may yet form the basis of a profound paradigm shift in systems
science, it is currently surrounded by controversy and its potential
contribution is quite unclear (Gaines, 1981).
.[
Gaines 1981 Autopoiesis some questions
.]
Indeed, it has been noted that an
``unusual degree of parochialism, defensiveness, and quasi-theological
dogmatism has arisen around autopoiesis'' (Jantsch, 1981).
.[
Jantsch 1981 autopoiesis
.]
.pp
There has been considerable discussion of the relation between autopoiesis and
concepts such as purpose and information.
Varela (1979) claims that
``notions [of teleology and information] are unnecessary for the
\fIdefinition\fR of the living organization, and that they belong to a
descriptive domain distinct from and independent of the domain in which the
living system's \fIoperations\fR are described'' (p.\ 63/64).
In other words, nature is not about goals and information; we observers invent
such concepts to help classify what we see.
Maturana (1975) is more outspoken:  \c
``descriptions in terms of information transfer, coding and computations of
adequate states are fallacious because they only reflect the observer's domain
of purposeful design and not the dynamics of the system as a state-determined
system'';
.[
Maturana 1975 organization of the living
.]
presumably goals are included too in the list of proscribed terms.
Some have protested strongly against this hard-line view \(em which is
particularly provocative because of its use of the word ``fallacious'' \(em
and attempted to reconcile it with ``the fact that the behavior of people and
animals is very readily and satisfactorily described in terms of goals and
attempts to achieve them'' (Andrew, 1981, p. 158).
In his more recent work Varela (1981) diverged further from the hard-line
view, explaining that he had intended to criticize only ``the \fInaive\fR use
of information and purpose as notions that can enter into the definition of
a system on the same basis as material interactions'' [his emphasis].
.[
Varela 1981 describing the logic of the living
.]
He concluded that ``autopoiesis, as an operational explanation, is not quite
sufficient for a full understanding of the phenomenology of the living,
and that it needs a carefully constructed complementary symbolic
explanation''.
For Varela, a symbolic explanation is one that is based on the notions of
information and purpose.
It is clear, though, that while some allow that autopoiesis can \fIcoexist\fR
with purposive interpretations, it will not \fIcontribute\fR to them.
.pp
Is autopoiesis restricted to \fIliving\fR systems?
Some authors find it attractive to extend the notion to the level of society
and socio-political evolution (eg Beer, 1980; Zeleny, 1977).
.[
Beer 1980
.]
.[
Zeleny 1977
.]
Others (eg Varela, 1981) stress the renewal of components through material
self-production and restrict autopoiesis to chemical processes.
Without self-production in a material sense, the support for the corollary
above becomes unclear, and consequently the whole relevance of autopoiesis
to identity and autonomy comes under question.
.rh "Artificial autopoiesis."
Although one can point to computer simulations of very simple autopoietic
systems (eg Varela \fIet al\fR, 1974; Zeleny, 1978; Uribe, 1981), there seems
to have been little study of artificially autopoietic systems in their own
right.
.[
Varela Maturana Uribe 1974 autopoiesis characterization and model
.]
.[
Zeleny 1978 experiments in self-organization of complexity
.]
However there are examples of computer systems which are autopoietic and
which have arisen ``naturally'', that is to say, were developed for other
purposes and not as illustrations of autopoiesis.
It is probably true that in each case the developers were entirely unaware
of the concept of autopoiesis and the interest surrounding it in system
theory circles.
.pp
.ul
Worm programs
were an experiment in distributed computation (Shoch & Hupp, 1982).
.[
Shoch Hupp 1982
.]
The problem they addressed was to utilize idle time on a network of
interconnected personal computers without any impact on normal use.
It was necessary to be able to redeploy or unplug any machine at any time
without warning.
Moreover, in order to make the system robust to any kind of failure,
power-down or ``I am dying'' messages were not employed in the protocol.
A ``worm'' comprises multiple ``segments'', each running on a different
machine.
Segments of the worm have the ability to replicate themselves in idle
machines.
All segments remain in communication with each other, thus preserving the
worm's identity and distinguishing it from a collection of independent
processes; however, all segments are peers and none is in overall control.
To prevent uncontrolled reproduction, a certain number of segments is
pre-specified as the target size of the worm.
When a segment is corrupted or killed, its peers notice the fact because it
fails to make its periodical ``I am alive'' report.