Upload
magali
View
217
Download
3
Embed Size (px)
Citation preview
BOOK REVIEW
Arturo Carsetti: Epistemic Complexity and KnowledgeConstruction
Theory and Decision Library A, Springer, Dordrecht, 2013,vii+151, $129, ISBN 978-94-007-6012-7
Magali Fernandez-Salazar
� Springer Science+Business Media Dordrecht 2014
This book constitutes a major contribution to our understanding of the mechanisms
of knowledge construction and a rare attempt to bridge the gap between biological
and connectionist models, on the one hand, and cognitive models on the other. The
volume documents a revolution now occurring in the cognitive sciences and in the
field of epistemic complexity, a revolution that permits the approach to the problem
of knowledge construction from the standpoint of both theoretical models and
simulation.
The first chapter of the book presents an up-to-date account of a number of recent
trends in the field of self-organization theory. In particular, entropy, algorithmic
complexity, self-referentiality and cellular automata are widely discussed. The
second chapter focuses on some advances obtained, from a modelistic point of view,
in biological computing. The alternative splicing and the interface between ruler and
coder constitute the essential backbone of the chapter. This chapter also introduces
alongside the classical concept of reflexive model, the new and fruitful concept of
self-organizing model. The third chapter is dedicated to the link between semantic
information and algorithmic complexity. The possible design of a biological
computer is widely discussed with respect to the introduction of non-standard
models and limitation procedures. The fourth chapter concerns the genesis of the
cognitive code with respect to the development of meaning considered both from
the point of view of government and use. Finally, the last chapter investigates the
functional nature of morphogenesis as it unfolds at the level of the emergence of
semantic forms and, in particular, of eigenforms.
M. Fernandez-Salazar
Neuropathic Pain Research, SND-CNRS Sorbonne University, Maison de la Recherche 28,
rue Serpente, 75006 Paris, France
M. Fernandez-Salazar (&)
Institute of Diagnostic Neuroradiology, University-Hospital of Greifswald, Walther Rathenau Str.,
46, 17489 Greifswald, Germany
e-mail: [email protected]
123
Minds & Machines
DOI 10.1007/s11023-014-9339-5
According to Carsetti, natural selection should be considered, at the co-
evolutionary level, as the effective coder in action (as also advocated, for instance,
by J. Maynard Smith). As such, it necessarily appears as linked to a continuous
process of emergence of meaning. Actually, at the biological level,
‘‘… what is innate is the result of an evolutionary process and is programmed
by natural selection. Natural selection is the coder (once linked to the
emergence of meaning): at the same time at the biological level this
emergence process is indissolubly correlated to the continuous construction of
new formats in accordance with the unfolding of ever new mathematics, a
mathematics that necessarily moulds the coder’s activity. Hence the necessity
of articulating and inventing a mathematics capable of engraving itself in an
evolutionary landscape in accordance with the opening up of meaning. In this
sense, for instance, the realms of non standard-models and non-standard
analysis represent, today, a fruitful perspective in order to point out, in
mathematical terms, some of the basic concepts concerning the articulation of
an adequate intentional information theory. This individuation, on the other
hand, presents itself not only as an important theoretical achievement but also
as one of the essential bases of our very evolution as intelligent organism’’ (p.
112).
Hence the importance, in Carsetti’s opinion, of articulating and inventing
mathematics capable of engraving itself in an evolutionary and self-organizing
landscape. At the level of a cognitive system sensibility (pace Kant) is not
‘‘… a simple interface between absolute chance and an invariant intellectual
order. On the contrary, the reference procedures, if successful, are able to
modulate canalization and create the basis for the appearance of ever-new
frames of incompressibility through morphogenesis. This is not a question of
discovering and directly exploring (according, for instance, to Putnam’s
conception) new territories, but of offering ourselves as the matrix and arch
through which they can spring autonomously in accordance with ever
increasing levels of complexity. There is no casual autonomous process
already in existence, and no possible selection and synthesis activity via a
possible remnant through reference procedures considered as a form of simple
regimentation. These procedures are, in actual fact, functional to the
construction and irruption of new incompressibility: meaning, as Forma
formans, offers the possibility of creating a holistic anchorage, and is exactly
what permits the categorial apparatus to emerge and act according to a
coherent arborization’’ (p. 111).
The volume aims to investigate, first of all, in which way the tools offered, for
example by non-standard analysis, can be of help to outline new approaches with
respect to the inner structure of an information theory also capable of taking into
account the teleonomical aspects proper to the cognitive actions expressed by living
beings. According to Carsetti, it is not possible to explain the whole complexity of
M. Fernandez-Salazar
123
the self-organizing and living processes within a general Markovian frame even if
expanded by taking into consideration the role of natural selection and the
differentiation processes. In addition, we have to make essential reference to a
logical level of explanation able to take into account the role played by meaning at
the level of the self-organization processes as well as the dialectics between surface
information and depth information. Certainly natural selection rewards the
flexibility and the supply of variability; why, however, does the evolution appear
to reward the supply not of a purely stochastic variability, but of a varied and
articulated complexity and consequently, of a constrained complexity?
Taking up an old thesis introduced by H. Atlan, the author remarks that in a
natural self-organizing system (a biological one) the goal has not been set from the
outside: what is self-organizing is the function itself with its meaning. The origin of
meaning in the organization of the system is an emergent property. Moreover, the
origin of meaning is strictly connected to precise operations of observation and self-
observation. This last remark constitutes one of the key points of the book: if we
want to capture the secret flavor of intentional information, we need measures
capable of taking into consideration the growth processes at stake, the statistical
fluctuations living at the microscopic level, etc. The Shannonian measure concerns
essentially stationary processes articulating in a one-dimensional landscape; on the
contrary a true measure of information for life and hereditary structures should
concern semantic information at work as implemented by specific, self-organizing
coupled systems.
As is well known, true cognition appears constrained by mathematical limitations
imposed by a number of specific analytical tools: computability and the Turing
universe, incompressibility and the oracles in action, self-organizing nets, deter-
ministic chaos, non-linear mathematics, second-order structures, and so on. It is, in
particular, with respect to this particular framework, that, in the author’s words, the
simulation activity, the construction, for instance, of an adequate semantics for
natural language, presents itself as a form of interactive knowledge of the complex
chain of biological realizations through which Nature reveals itself to our brains in a
consistent way (by means, for example, of the intelligent design of specific
experiments at the level of an extended Turing universe). In certain respects, this
can be considered as the true turning point of the book: Carsetti claims that the
simulation work, in effect, offers the semantic net real instruments in order to
perform a self-description process and to outline specific procedures of control as
well as a possible map of an entire series of paths of creative imagination. In turn,
the progressive (and selective) exploration of these paths will allow external
information to form in an emergent way, thus exploiting new and even more
complex patterns of interactive expression and action. When we pass from a world
of objects to a world of constructions we are no longer exclusively faced, for
instance, with boolean algebras, first-order structures and simple observational acts;
we are really faced with a dynamic and functional universe characterized by inner
circularity, by self-organization and by the presence of specific categorization
processes as well as of precise evolutionary differentiation patterns. It is exactly the
correct framing of this particular kind of laboratory of possible emergence that will
assure the successive revelation of ever new levels of depth information.
Epistemic Complexity and Knowledge Construction
123
In its technical part the book investigates the complicated plot existing between
depth information and surface information from the point of view of computability
theory. In particular, Carsetti remarks that the attempts of Church and Kleene as
well as Post’s version of Godel’s theorem not only show, in terms of recursive
function theory, that formal axiomatic systems are incomplete but they also give, in
certain respects, some hints in order to outline an information-theoretic version of
Godel’s theorem, a version that will be given later by Chaitin. In this version we
find precise suggestions about the possibility of introducing effective measures of
the information power of formal axiomatic systems. Carsetti claims that insofar as
we manage to realize, by self-reflection on our own reasoning, that our logical and
mathematical inferences can be formalized by a given formal system, we also
realize that self-reflection is itself part of our mathematical reasoning; actually, it is
at the basis of the effective construction of the undecidable. Thus, we can better
realize in which sense depth information can be defined but, at the same time, it
cannot be effectively computed.
In accordance with some original intuitions by Hintikka, the author argues that
depth information can be thought of as surface information at infinite depth. In
certain respects, we can simply affirm that it can be calculated by an infinite process
during which one can never know how close one is to the final value. In this sense,
we need more adequate measures of meaningful complexity, also capable, for
example, of taking into account the dynamic and interactive aspects of depth
information. We have, in particular, to outline new models for the interface existing
between the observer and the observed system. At the level of this kind of model,
emergence (in a co-evolutionary landscape) and truth (in an intensional setting), in
many aspects, will necessarily coincide. Moreover, a coupled system in this
theoretical perspective must be considered as a multiplicative unit: a source-forge
for other coupled systems. Here we can precisely identify the role played by self-
organizing models: the reflexive models are limited because they give an account
only for the relationships between perception and existence at the standard level,
between creativity and meaning as limited to the realm of invariance. The self-
organizing models, to the contrary, aim to examine the work done by the ruler (as
well as by the connected ‘editor’) with respect to that particular transition point
provided by the link existing between the realm of invariance and the paths of
morphogenesis. In this sense, it is necessary to refer to new models of the real-
number line; these models, however, should be objectively identified (as in the case
of Harthong-Reeb theoretical construction) also with reference to the relationship
existing between seeing and thinking as well as to the ongoing processes of
construction of our neural circuitry. Hence the necessity of a continuous
exploration, in evolutionary terms, of those particular modules intersecting
mathematical investigations and epigenetic growth (at the neural level) that really
identify the secret architectures of the ‘knowing I’. We really have the possibility to
enter the territories of epistemic complexity even if to continue the line of thought
taken in this book it appears necessary to develop, in mathematical terms, a more
incisive and accurate analysis of the role played, at the functional level, by the self-
organizing models.
M. Fernandez-Salazar
123
This is a fine work, a fascinating set of discussions concerning an extremely
interesting area. It contributes to the current and lively debate about the nature of
cognition and about the role played by symbolic dynamics and epistemic
complexity in the development of biological and cognitive activities. Anyone
who is interested in functional and neural models concerning knowledge construc-
tion will be informed by the wealth of material presented in this volume.
Epistemic Complexity and Knowledge Construction
123