Go to Semiotic Review of Books Home Page
Go to SRB Highlights
Go to SRB Archives
This review appeared in Volume 1 (1) of The Semiotic Review of Books.
Structural semiotics has reached a crossroads. In the 60's and 70's, notably in Europe, it produced a methodological model with theoretical and scientific ambitions, which, in the course of the 80's underwent a process of critical reassessment of its hypotheses and the philosophical tradition upon which they rested. Three directions are currently being explored: (1) The poststructuralists claim that structuralism failed to deliver on its promises and they assign this failure to the impossibility for the "human sciences" to be truly scientific. Their deconstructionist and antitheoretical stands lead to the restoration of poetic exercises and literary essayism; (2) The methodologists consider semiotics to be a mere methodology without any particular object; (3) finally there is a scientific trend, both theoretical and rationalist, whose proponents contend that structural semiotics has as its object natural phenomena of a certain kind and that it is now in a position to bridge the gap with the natural sciences without submitting its object to any form of reductionism. The latter direction forms the conceptual framework within which the research program in semio-cognitive morphodynamics developed over the last few years.
A similar evolution can be observed in psychology, notably in the study of categorization and (proto)typicality (Lakoff 1987). The emergence of discontinuities is, at long last, recognized as dynamic "catastrophe" processes of bifurcations of stable states (e.g. the categorical perception of phonetic phenomena). The same trend can be observed in qualitative physics, information processing theory and, generally, in artificial intelligence research, domains in which some aspects of morphodynamic models are now used.
It is in this context that a research program attempting to link together organically semiotics, cognitive science and some of the natural sciences was initiated. This program comprises three sub-projects, each one being a considerable endeavour of its own.
Morphodynamic models make it possible to solve one of the most important problems created by the modern conception of science. Since the triumph of physicalist mechanism in the XVIIth and XVIIIth centuries, the mechanics of forces has overshadowed the dynamics of forms. The claim that it is impossible to link physical objectivity with a qualitative ontology of the perceived world is at the origin of the gap that developed between modern science and Aristotelianism. The issue concerns the structuring of the world into things, forms, states of affairs, events, processes which are both perceptible and describable; if indeed, all these phenomena are reduced to formal descriptions of meaning structures correlated to the subject's cognitive acts, then phenomenalization cannot be conceived as a natural process of matter but only as the result of a subjective activity. This approach, which is taken for granted since the age of classicism, (e.g. Hobbes, Leibniz) must now be considered to be a fallacy. Morphodynamic theories -- e.g. the theory of singularities and their unfolding, the theory of bifurcation of non-linear dynamic systems, the theory of critical physical phenomena -- have shown that there exists a level of reality which is qualitative, morphological, structural and organizational and which emerges from physical reality itself. This level can be called "phenophysical", a neologism coined by Per Aage Brandt. Fundamental physics (viewed as a kind of "geno-physics") expresses itself phenological as a "pheno-physics" which can be described and explained through morphodynamic models.
Once a pheno-physics is elaborated, it is possible to undertake the founding -- in a non reductive manner -- of the phenomenological structures of meaning upon objective phenomenology. Thus, a qualitative ontology can be built upon a physical basis since emerging qualitative structures are viewed as intrinsically meaningful. Founding semiotic naturalism upon pheno-physics makes it possible to account for perception and language in terms of ontological realism. The term "semiophysics" (semiophysique) was coined by Rene Thom (1988) and suggests a scientific program that could unite, under the heading of "physics of meaning" various domains of contemporary research such as: (1) Gestalt theory and Phenomenology in as much as they attempt to apprehend protogeometrical entities as morphological basic units. (2) Ecology of visual perception, from J.J. Gibson (1979) to David Marr (1982) and their followers, to the extent that they contend that there exist some qualitative and morphological structures in the environment which are objective without being merely physical (i.e., which are pheno-physical) and which are made explicit through perception. Marr's ecologism may be expressed in terms of morphodynamic models. (3) Greimas' semiotics of the natural world (1987: 17-47), because it deals with the universe as a set of sensory qualities endowed with some measure of organization. According to this view, morphological "surface structures" (gestalten) emerge from physical "deep structures", thus generating a "figurative language" that is common to perception and language. (4) Peircean Aristotelianism given its concern for substantial forms and entelechies. Like many other philosophers and scientists (e.g. Leibniz, Kant, Goethe, Maxwell, Brentano, Driesch, Stumpf, Husserl, D'Arcy Thompson, Valery), C.S. Peirce was indeed fascinated by the morphology of the natural world. He understood that its diversity and complexity could not be explained merely in terms of physical, mechanical or thermodynamic forces, but that it was necessary to take into account the basic notions relating to Aristotelian "hylemorphism", in particular the concepts of substantial forms and entelechies. For Peirce an entelechy belongs to thirdness in so far as some matter (firstness) becomes determined by a form. The entelechy realizes itself through matter in a programmed manner, the form implementing the program (plan, cause, law) as inner finality. Entelechies are signs, whose objective physical nature itself is the interpretant. The phrase "physics of meaning", referring to the mathematical expression of the morphodynamic concept of Aristotelian entelechy seems to be congruent with the Peircean semiotic tradition. (See Peirce's Collected Papers: 1.22 and 6.355, and The Writings of C.S. Peirce, 1982: 330).
Cognitive theories, bearing upon the perceptual grasping, intellection, and construction of forms by subjects, call for a cognitive morphodynamic theory that could also account for the dynamic models of performance, thus relating cognition to the physics of meaning.
In view of the shortcomings of classical cognitivism which have been emphatically pointed out in resent years by Putnam (1988), Searle (1984), Dreyfus and Dreyfus (1986) and others, an examination of the contributions made by the morphodynamic models of semiocognitive processes is now in order. It concerns in particular the difficulties encountered by classical cognitivism when dealing with the relationship between symbolic mental representations and the physical world as a set of meaningful structures. In this respect, one of the thorniest issues is the question of intentionality, i.e. the question of meaning not as mere denotative reference but as relation to a qualitatively and phenomenologically structured universe.
As was mentioned earlier, sub-symbolic connections is an interesting alternative to standard cognitivism for conceptualizing computational processes.3 According to connectionism, entities endowed with semantic content are described as attractors of dynamic systems; semantics can thus be equated with topology, or, in other words, with attractors' inner structures. Semantics is therefore conceived as a topologico-dynamic rather than logicocombinatory system.
In a well known article, Fodor and Pylyshyn (1988) harshly criticized connectionism on the ground that a consequent mental connectionist must necessarily impute internal syntactic structures to mental representations in order to account for their productivity, generalivity, systematicity, compositionality and inferential consistency, but that the semantic associationism of connectionist models makes combinatory syntax impossible and prevents one from developing any theory of actantial relationship in discourse. The morphodynamic approach has the marked advantage of providing an answer to their criticism by showing that the point of view of dynamic connectionism makes it possible to elaborate a structural actantial syntax. More on this later.
Gibson (1979), Marr (1982) and others have developed the hypothesis that perception makes explicit various levels of reality which are not merely physical, but nevertheless objectively exist, in particular the level of morphological reality. This presupposes that there exist topological mental representations which are not logicocombinatory.
Research in mental imagery, e.g. Kosslyn (1980), Sheppard and Cooper (1982), tends to concur with this. In topologico-geometric mental representations, geometric properties are opposed to algebraic ones, as are analogical and depletive ones to propositional and descriptive ones respectively. Objects are represented by their form and spatial relations, not by formal relations among symbols.
The morphodynamic approach allows the modelling of perception processes in a way which perfects those proposed by Marr, and which validates in part the ecologism hypothesis. There are objective qualitative discontinuities that are conveyed by optical and acoustic signals, detected by transductors, and used as a basis for the cognitive construction of forms, events, and processes in space-time. This mixture of phenophysical properties and perception processes is the ground for the "qualitative ontology" of the projected environment (viz. Jackendoff), i.e. the semiotics of the natural world.
In view of such a qualitative ontology, it is possible to develop a structural actantial syntax, whose basic idea is the morphodynamic schematization of the localist hypothesis. It is assumed that, on a deep cognitive level, syntactic representations of a logico-combinatory nature, and perceptual representations of a topologico-dynamic nature become compatible. But how? Suppose a visual scene that is to be described linguistically: the essential information is provided by its case infrastructure (i.e. the set of semantic actantial roles comprising the scene within a spatio-temporal framework). In order to linguistically represent this scene, the actants must be reduced to their loci, i.e. to space-time balls, incidentally a point about which Thom's, Marr's and Langacker's theories concur. This means that the space of the scene is categorized into domains whose centres are the balls. Such a diffusion of contours is effectuated by means of a potential f. If the situation is simplified through reducing the balls to points, a potential f w varying temporally is obtained, thus defining an "actantial graph". In this manner, it is possible to schematize both topologically and temporally as places and paths the spatio-temporal interactions (events and processes) of the actants. The localist hypothesis claims that these interaction schemata are sorts of archetypes for actantial relations in general. This results in the construction of an iconic schematization of deep actantial relations through which case grammars and narrative structures based on actantial models can be appropriately mathematized. Note that this approach anticipated, and concur with current research by cognitive linguists such as for instance Ray Jackendoff; Leonard Talmy and Ronald Langacker (op. cit)..
Topological and spatio-temporal schematism is also well suited to account for aspectuality. From a semiotic point of view, aspectuality can be considered to be an overdetermination of temporality through which the functions of narrative statements (énoncés) are transformed into processes categorized as inchoative, terminative, durative, perfective and imperfective, etc. According to the formalist (logicocombinatory) perspective on narrative actantial syntax, aspectuality can only be extrinsic. In Greimas' theory for instance, (Greimas and Courtes 1982: 18-19) aspectuality merely pertains to the discursive, exteroceptive level of the semiotics of the natural world and results from the transformation of logical into spatiotemporal structures. By contrast, the morphodynamic approach construes aspectuality as an intrinsic, constitutive feature of actantial syntax. Aspectuality thus encompasses all the grammatical determinations of becoming, moving, processing, overcoming, etc., i.e. of the action of time in substances (in the sense of qualitative ontology). Examples include, stable states reversibly occupying an open temporal interval; beginnings and endings of processes (inchoative, terminative), events, corresponding to changes of qualitative states, i.e. qualitative discontinuities through which the operation of (external) time upon internal qualitative space becomes manifest; actantial interactions described by verbs.
The morphodynamic approach to actantiality leads to results which are interestingly akin to those of some cognitive linguists (e.g. L. Talmy) as far as actantial modalities are concerned (Brandt 1988). Morphodynamic schematism is not only topological and dynamic in the external space-time, but also in internal spaces, within which there exist potentials which account for actantial configurations and whose bifurcations generate actantial interactions. Suppose indeed that actants were provided with a kind of autonomous inner energy and that external dynamics were introduced within external spaces (distinct from space-time), then it becomes relatively easy to deduce from this some morphodynamic schemata applicable to modalities. This is Talmy's central idea, namely that modalities form a closed class (syncategorematic) specifying grammatically a dynamic set of forces pertaining to the concepts of obstacle, resistance, blockage, overcoming, co-operation, competition, interaction, etc. External dynamics are modal dynamics corresponding in narrative semiotics to addressers' activities.
The morphodynamic approach also casts a new light on the fundamental issue of intentionality. The problem is actually twofold: (1) cognitive systems, for instance in humans, behave in an intentional manner, i.e. they appear as being directed, governed, ruled by beliefs, knowledge, purposes as if they were determined by internal representations which act as efficient and final causes; (2) on the other hand, cognitive systems are oriented toward their environment, notably through perception, directionality and referentiality. As Searle (1984), after Husserl, asserts, intentionality is an intrinsic property of certain mental states which are characterized by the fact that they are oriented toward objects and states of the world. The problem of problems (as Husserl says) is perceptual intentionality: how to escape mentalist idealism and methodological solipsism?
Classical cognitivism is indeed "blind". It lacks semantics in its very principle since semantics is restricted to the denotative dimension. It is devoid of intrinsic and internal intentionality and it has no other possibility but to resort to circular reasoning when dealing with the fundamental notions of semantics. As Pierre Ouellet (1983) showed, this vicious circle can be found in Wittgenstein's Tractatus. A proposition describing a state of affairs possesses a local form (logische Form) which is its syntactic articulation and a form of meaning (Form der Sinn) which is of a gestaltico-semantic nature. Thus structured, the proposition is a logical image of the state of affairs; it is the form of the reproduction (Form der Abildung). The proposition presents itself in its form as representation; it is the form of representation (Form der Darstellung). By unfolding itself (exposing itself), it unfolds the way in which the state of affairs unfolds and presents itself. Logical forms thus become forms of reality (Formen der Wirklichkeit) by repeating themselves. This is the solipsist circle of logicist idealism. The relationship "proposition state of affairs" is projective, i.e. the logical form and the form of meaning project themselves and structure the state of affairs in conformity with themselves. The mimesis between the logical form of meaning and the form of reality is actually a self-representation of the logical form of meaning. Consequently, it is impossible to elaborate a satisfactory theory of intentionality based upon classical mentalism.4
In conclusion, it should be obvious that the program of research for a semio-cognitive morphodynamics constitutes a "morphological turn" for standard structural semiotics which can thus join the natural sciences and the cognitive sciences while remaining within the purview of its intellectual heritage. Naturally, a physics of meaning will require mathematical, epistemological and experimental means which might appear to be excessively demanding in the eyes of those semioticians who are used to content themselves with mere conceptual descriptions. But, failing to meet this challenge would lead semiotics to dissolve in the literary vagueness of post-structuralist essayism. This would be all the more regrettable since structuralism has now succeeded in acquiring the mathematics and empirical methodology required by its original ambitions, concepts and projects.
1. This text is based on the position paper I gave at the beginning of the panel on Semiotics and Cognitive Science which I organized as part of the 4th congress of the International Association for &miotic Studies (Perpignan, April 4, 1989.) it was translated into English by Paul Bouissac with the collaboration of David Savan.
2. Contemporary cognitivism is dominated by the paradigm of classical cognitivism which is mentalist, representationist, symbolic, computational and functionalist, e.g. Fodor (1980) and Pylyshyn (1984). It assumes that the environment emits physical information (wavelength, intensities, etc.) which is not significant as such for the cognitive subject. Therefore, this information must be translated by peripheral transductors (retina, cochlea, etc.) into neuronal information that can be processed by the central nervous system through several levels of symbolic mental representations. Mental representation as a psychological reality is taken for granted (note, however, the debate between mental realism on one side and, on the other side both radical physicalism, e.g. Quine (1969), Churchland (1984), and non radical physicalism, e.g. Dennett (1987) who accepts the concept of mental representation as a descriptive concept, not as an objective reality). For classical cognitivists mental representations are considered to be symbolic (in the sense of symbolic logic) and to be expressions of an internal formal language (e.g. Fodor s language of thought). It is hypothesized that there exists a calculus through which these expressions are manipulated by rules of inference, transformations etc. This calculus is implemented physically, hence causally but causality is restricted to the syntactic structure of expressions. However, functionalism distinguishes the implementation (neuronal hardware) from the symbolic calculus itself. After such a computational treatment of the information input, a projection process takes place resulting in the cognitive construction of a projected world. Phenomenological consciousness is considered as the correlate of this projected world. According to this view it is possible to investigate the relationship between consciousness and the computational mind. Jackendoff, for instance, casts in this theoretical framework the whole program of research associated with Gestalt theory and Husserl s phenomenology. Moreover, classical cognitivism holds that the computational mind comprises two types of systems. First, it assumes the existence of modular peripheral systems whose function is to transform the information provided by the transductors Into representations endowed with propositional properties suited to mental calculus. These transductors operate in a strictly bottom-up manner like data-driven computational reflexes; they are specific, computational and informationally partitioned, i.e. insulated from the subject s beliefs and expectations. They formulate hypotheses and make non-demonstrative inferences, thus transforming proximal sensory stimuli into representations of distal objects. Secondly, classical cognitivism makes the assumption that there exist central cognitive systems which are non modular, non-specifial operating in a top-down manner, non partitioned, and interpretative. Since there is no nomological control of their functioning according to some rules, it is not possible to deal with them scientifically. This raises the epistemological problem of semantic holism. For Fodor, central systems are isotrope i.e. all beliefs and knowledge are potentially relevant for the treatment, interpretation and resolution of the output from the modules, and they are Ouinian, i.e. the set of beliefs and knowledge conditions each treatment. Hence Fodor s (e.g. 1981, 1983) radial criticism of Minsky, Schank, Winograd, Newel and of the notion of expert systems, since they consider central systems as if they were modular and specific.
Semantic holism implies that what is significant in the environment for the cognitive subject (i.e. the interaction subject/environment) cannot be derived from the laws of nature, and therefore cannot be integrated in a scientific psychology because science proper can only be nomological. A descriptive discipline, even if it is empirically confirmed is not nomological and cannot be considered a science in the full sense of the term. This is the source of the thesis of methodological solipsism according to which no constitutive reference to the structures of the external world can be included in a scientific psychology, a thesis which leads to the categorical rejection of the point of view expressed in this article in as much as it includes ecology and qualitative ontology in its purview. The approach exemplified by Fodor considers that the only objective reality is the physical one, in the physicalist (geno-physics) sense of the term. This reality acts causally and nomologically upon the computational automatisms of the transductors and modules. On the level of central systems, only the syntactic form of representation acts causally, and therefore signification cannot become an object of scientific inquiry. The counter thesis defended in this article is that there exists a natural semiotics in the environment (a qualitative ontology) which is not encompassed by semantic holism and is not denotative either.
3. Let us consider the evolution of instantaneous global states of non-centred networks S of interconnected elementary units. An instantaneous state is identical with a vector of degrees of activity of various units.In other words, we consider a spatial network of phases N and a global enomorhism T:N->N expressing the local laws of the states of transition of the states of the units with respect to the received information. T is iterated and its asymptotic stable and attractive states are considered.
4. Daniel Dennett's point of view is equally questionable as it considers intentional conceptuality to be a predictive strategy i.e. a heuristics which allows the predicting of the way in which certain systems will behave. Based upon the competence/performance opposition Dennett s thesis contends that cognitive systems such as the brain are intentional (i.e. they are semantic machines) on the level of cinematic competence (the formal and abstract description theory of the functioning) but that they are actually syntactic machines physiologically i.e. on the dynamic level of performance. In so far as syntax does not determine semantics one may wonder how such systems can produce semantics from syntax. Dennett claims that the brain mimics the behaviour of a semantic machine by relying on correspondences between (1) regularities of its internal organization and of its external environment and (2) semantic types. But such a thesis is tenable only if the prime problem of intentionality and directionality has been solved. The thesis may hold for propositional attitudes (intentionality in the trivial sense of folk psychology) but not for perceptive intentionality.
Brandt, P.A. 1988 La charpente modale du sens. Pour une sémiolinguistique morphogenetique et dynamique. Thèse de Doctorat d Etat-Université de Paris lll.
Brentano F. (1973) Psychology from an Empirical Standpoint. Trans. by A.C. Pancurello D.B. Terrell and LL. McAlister. New York: Humanities Press (originally published in 1874).
Churchland P.M. (1984) Matter and Consciousness: A contemporary Introduction to the Philosophy of Mind. Cambridge MA: M.l.T. Press.
Dennett D. (1987) Intentional Stance. Cambridge MA: M.l.T. Press.
Desclés J.P. (1985) Représentation des connaissances: archètypes cognitifs, schèmes grammaticaux. Actes Sémiotiques - Documenta. Paris: C.N.R.S.
Dreyfus H.L. and Dreyfus S.E. (1986) Mind over Machine: the Power of Human Intuition and Expertise in the Era of the Computer. New York: The Free Press.
Fodor J.A. (1981) Representations, Philosophisal Essays on the Foundations of Cognitive Science. Cambridge MA: M.I.T. Press.
Fodor J.A. (1983) The Modularity of Mind. Cambridge MA: M.l.T.
Fodor J.A. and Pylyshyn Z.W. (1988) "Connectionism and Cognitive Architecture: A Critical Analysis ". Cognition 28 1-2:3-71.
Gibson J.J. (1979) The Ecological Approach to Visual Perception. Boston: Houghton Miffin.
Greimas A.J. and Courtés J. (1982) Semiotics and Language: An analytical Dictionary. Trans. by L Crist D. Patte et. al. . . Bloomington IN: Indiana University Press.
Greimas, A.J. (1987) On Meaning. Selected Writings in Semiotic Theory. Trans. by P.J. Perron and F.H. Collins. Minneapolis: University of Minnesota Press.
Hjelmslev, L. (1943) Prolegomena to a Theory of Language. Trans. by F.J. Whitfield. Madison: University of Wisconsin.
Hjelmslev, L. (1959) Essais linguistiques. Copenhagen Nordisk Sprogog Kulturforlag.
Jackendoff, R. (1983) Semantics and Cognition. Cambridge MA: M.l.T. Press.
Kosslyn, S.M. (1980) Image and Mind. Cambridge MA: Harvard University Press.
Lakoff, G. (1987) Women, Fire, and Dangerous Things. What Categories Reveal about the Mind. Chicago: University of Chicago Press.
Langacker, R. (1986) Foundation of Cognitive Grammar. vol.1 Stanford: Stanford University Press.
Lévi-Strauss, G (1988) De près et de loin. Paris: Odile Jacob.
Marr, D. (1982) Vision. A Computational Investigation into the Human Representation and Processing of Visual information. New York: W.H. Freeman and Company.
Ouellet, P. (1983) Le sens de forme du sens. Doctoral thesis. Paris Vll.
Peirce, C.S. (1931-1958) Collected Papers. 8 vol. P. Weiss and C. Hartshorne (eds.). Cambridge MA: Harvard University Press.
---. (1982)The Writings of C.S.Peirse. vol. I.M. Fishet.al. (eds.) Bloomington IN: Indiana University Press.
Petitot, J. (1985) Les catastrophes de la parole.Paris: Maloine.
Pottier, B. (1987) Théorie et analyse en linguistique. Paris: Hachette.
Putnam, H. (1988) Representation and Reality. Cambridge MA: M.l.T. Press.
Pylyshyn, Z.W. (1986) Computation and Cognition. Cambridge MA: M.l.T. Press.
Quine, W.V.O. (1969)Ontological Relativity and Other Essays. New York: Columbia University Press.
Searle, J. (1984) Mind, Brains and Science. Cambridge MA: Harvard University Press.
Shepard, R.N. and Cooper, L.A. (1982) Mental Images and their Transformations. Cambridge MA: M.l.T. Press.
Smolensky, P. (1988) "On the proper treatment of Connectionism." The Behavioral and Brain Sciences. 11, 1-74.
Talmy, L. (1983) "How Language Structures Space." In Spatial Orientation: Theory Research, and Application. H. Pick and L. Acredolo (eds.). New York: Plenum: 225-282.
Thom R (1969) "Topological models in biology." Topology 8: 313-335
---. (1975) Structural Stability and Morphogenesis: An outline of a General Theory of Models. Reading: Benjamin.
---. (1983) Mathematical Models of Morphogenesis. Chichester: Ellis Horwood.
---. (1988) Esquisse d'une sémiophystique. Paris: InterEditions.
Thompson; Darcy W. (1917) On Growth and Form. Cambridge: Cambridge University Press.
Wildgen, W. (1982) Catasprophe Theoretic Semantics. Amsterdam/ Philadelphia: Benjamin.Jean Petitot is Directeur d'études at L'Ecole des Hautes Etudes en Sciences Sociales (Paris). He is the author of Morphogenèse du Sens. I. Paris: Presses Universitaires de France (1iEi) Les catastrophes de la parole. De Roman Jakobson à René Thom. Paris: Maloine (1985) and of numerous articles. He has edited Logos et théorie des catastrophes. A partir de l'oeuvre de René Thom. Paris: Patino (1989). He has recently initiated the Research Progiram in Semio- cognitive Morphodynamics (Maison des Sciences de l'Homme 54 Bd. Raspail 75008 Paris France).