Learning As Self-organization

Karl H. Pribram, Joseph S. King

• Publisher: Psychology Press
• Year: 2013
• Print ISBN: 9780805825862
• eBook ISBN: 9781134997084
• Topic: Psychology
• Subtopic: Cognitive Psychology & Cognition
• Index: Psychology

A year before his death, B.F. Skinner wrote that "There are two unavoidable gaps in any behavioral account: one between the stimulating action of the environment and the response of the organism and one between consequences and the resulting change in behavior. Only brain science can fill those gaps. In doing so, it completes the account; it does not give a different account of the same thing." This declaration ended the epoch of radical behaviorism to the extent that it was based on the doctrine of the "empty organism," the doctrine that a behavioral science must be constructed purely on its own level of investigation. However, Skinner was not completely correct in his assessment. Brain science on its own can no more fill the gaps than can single level behavioral science. It is the relation between data and formulations developed in the brain and the behavioral sciences that is needed.

This volume is the result of The Fourth Appalachian Conference on Behavioral Neurodynamics, the first three of which were aimed at filling Skinner's first gap. Taking the series in a new direction, the aim of the fourth and subsequent conferences is to explore the second of the gaps in the behavioral account noted by Skinner. The aim of this conference was to explore the aphorism: The motivation for learning is self organization. In keeping with this aim and in the spirit of previous events, this conference's mission was to acquaint scientists working in one discipline with the work going on in other disciplines that is relevant to both. As a result, it brought together those who are making advances on the behavioral level -- mainly working in the tradition of operant conditioning -- and those working with brains -- mainly amygdala, hippocampus, and far frontal cortex.

The Neural Systems Level: Process

12.

Morphogenesis and Mental Process

Jason W. Brown

Morphogenesis and mental process

JASON W. BROWN

New York University Medical Center, Department of Neurology

Abstract

Parcellation and heterochrony (neoteny) reflect the pattern and rate of a growth mechanism in morphogenesis. Structure (morphology) and function (behavior) are staged realizations of morphogenetic process. This process continues into adult cognition in the actualization of the mind/brain state. Parcellation obtains in the pruning of cells and connections in early growth, whereas inhibition obtains in a relatively stable morphology with constraints on context: item transforms in microgeny. Selective retardation in process (neoteny) leads to growth at earlier (juvenile) phases. This accounts for the specification of the language areas and elaboration at preliminary phases in mind–for example, dominance, introspection, and creativity.

… we create distinctions, then

Deem that our puny boundaries are things

Which we perceive, and not which we have made.

Wordsworth

Prelude II:221

In the past, efforts were made to show that developmental abnormalities could be interpreted as an attenuation or deviance of normal maturation (e.g., Werner, 1948). These accounts were based for the most part on a temporal or diachronic perspective that gradually fell out of favor with the advent of cognitive science. This was partly a result of work that was critical of simple correlative accounts, but there was also a strong bias against this mode of explanation that resulted from the synchronic perspective typical of cognitive psychology. The specification of systems through growth was incompatible with the concept of modules for various functions conceived as genetically specified organs. In this view, pathology, whether developmental or acquired, represents the perturbation of a system (i.e., a program) that is prewired and no more than fine-tuned by the environment. The development of the brain is treated as an unimportant detail that mediates between the innately delivered program and the mature function. This way of thinking is clearly linked to artificial intelligence models of the mind, where the pattern and direction of growth in the brain are considered no less irrelevant to the final organization of cognition than the mode of construction of a computer to the software.

In recent years, however, there has been a revival of interest in the relationship between normal and pathological development. This field, termed developmental psychopathology (e.g., Cicchetti, 1984, 1993), has led to a rethinking of normal and aberrant development in terms of common underlying mechanisms. The emphasis on timing as a key to the understanding of deviant behavior and the life-span approach to growth and pathology are central to this account. These principles have also been an important feature of the microgenetic theory of adult pathology and its relation to evolutionary and developmental growth trends (Brown, 1988; Brown & Jaffe, 1975). This theory holds that an understanding of the nature of the symptom or error rather than the deficit is of critical importance if a common basis of growth and pathology is to be uncovered. But what precisely are the mechanisms that growth and pathology have in common? This paper is an attempt to specify some of the morphogenetic patterns that may account for behavior in the normal state and in cases of developmental or acquired pathology.

Morphogenesis

Ontogeny is a bridge from the genetic code to the phenotype, whereas phylogeny is the sum, or average, of the ontogenies of a given line. Ontogeny covers a period from conception to adult structure with no clear boundary except death at the end of a developmental sequence. At some relatively stable point in this sequence, an organ is judged to express a phenotype. The phenotype is usually assumed to be directly generated by the genome. For example, in brain organization, there is assumed to be a correlation between the genetic code and the specialized adaptations of the mature organism. Ontogeny is the process through which this correlation is achieved.

The extent to which morphology is specified in the genome is uncertain. Presumably, the genetic code contains a set of instructions for the processes–or events that lead to the processes—that will realize the code in structure. Such processes as are specified by the genetic code are still a long way from phenotypic structure. Neurobiologists recognize this problem and focus, therefore, on the epigenetic functional relations or algorithms that translate code to structure, not the manner in which genes produce nerve cells or circuits (Stent, 1981). However, our knowledge of relations underlying the morphogenetic process, the transition from the genetic code to the developed organism, is still very incomplete.

From the standpoint of mental development, the onset of sensation and learning at birth creates a natural testing ground for the study of genetic specificity. The newborn is prepared for many complex behaviors through a genetic endowment that is relatively uncontaminated by learning. Yet the innate capacities of the mind/brain of the newborn are difficult to specify and continue to be a source of lively debate (see Carey & Gelman, 1991). Speculations range from innate rules, ideas, and/or mental processes to the more conservative notion of constraints on action and perception.

The problem with studies of innateness that begin with the newborn is the focus on behavior as an interaction of a delivered morphology with experience. Circuits in the brain are taken to be the outcome of a prenatal process of growth when, in fact, growth continues on into late life. Similarly, brain function is assumed to be determined by the pattern of electrical activity that maturing circuits generate (i.e., the output of the structure) when the pattern of activity, that is behavior, is an expression of sustained growth. The cleavage between growth and function, or structure and process, results in a neglect of the formative history of the mind/brain, in evolution or in utero. The goal of development is a machine that can be instructed or that can realize a functional program. In this view, the fetal segment of ontogeny is less informative than the confrontation of the innate and the acquired in the earliest period of postnatal life.

In contrast, suppose that development is morphology (i.e., that morphology is an artificial slice through development with behavior its four-dimensional structure) (Brown, 1991; cf. Striedter & Northcutt, 1991). From this standpoint, the newborn is not a starting point to study the innate determinants of language and behavior but, rather, is a phase in a life-span process. Onto- (morpho)-genetic process leading to the mind of the newborn also lays down function after birth. The dichotomy between the innate and the acquired is orthogonal to the nature of mental activity, and this activity is independent of birth as a pivotal event.

With this perspective, an early phase in development deposits morphology, and a late stage deposits function. A common process elaborates both morphology and function, with function being the iteration of growth through the morphology. There have been prior speculations along these lines (e.g., the ideas of Loeb [1907] and Goldscheider [1906], that “… configurations experienced in perception might derive from excitation in the brain resembling the ‘force lines’ that determine form during embryogenesis” (cited in Pribram, 1991, p. 25). A relationship has been suggested between “ontogenetic sculpting” and mechanisms of learning (Thatcher, 1992b) and information representation in the brain (Malsburg & Singer, 1988). Tucker (1992; see also Brown, 1990) commented that physical growth in the brain is psychological growth. The problem is to specify the growth process and determine whether or not this process is related to processes underlying cognition in the adult.

There have been attempts to define the morphogenetic process with greater precision. Goodwin (1982) argued that developing organisms have “an extensive range of morphological potential, describable in terms of probabalistic fields which collapse … into specific morphologies reflecting the particular conditions, internal and external, which act upon them” (p. 52). According to Goodwin, the generative principles that account for the progression from whole to part recur so that a taxonomy of biological form can be achieved through a hierarchic ordering of the transformations. Katz (1983) proposed that “ontogenetic buffer mechanisms” mediate the transition to functional brain architecture. These mechanisms include exuberant growth with specificity through parcellation or pruning of connections, possibly by competitive interaction (Edelman, 1987), and heterochrony, a variation in the timing of developmental process.

It is a thesis of this paper that processes in the development of the brain recur as processes in cognition, that development and cognition–ontogeny and microgeny–are different ways of looking at the same process. Two lines of evidence for this thesis are explored: the relation of parcellation in development to specification in cognitive processing, and heterochrony as a theory of developmental abnormality in relation to pathological symptoms in mature organisms.

Parcellation

Morphology

Parcellation is the pruning of exuberant connections in the growth of the brain as a way of achieving specificity in mature brain structure. According to parcellation theory, the connectivity of the brain is accomplished, at least in part, through a loss of connections. Indeed, most structures in the vertebrate brain have a larger number of neurons during development than in adulthood. The decrease in synapses is even more striking. In macaque neocortex there is a loss of over 2 trillion synapses by the 5th year of life (Rakic, 1989, 1992).

Ebbeson (1984) has written that most, if not all, systems go through phases of diffuse projections that later become more restricted, presumably by the degeneration of selected axonal branches or the loss of selected neurons. The finding of initial proliferation and later elimination in the progression from the general to the specific has been described mainly in the study of sensory systems, where exuberant growth in juveniles with loss of cortical connections (neurons and synapses) in adults is a characteristic feature. There is also evidence for parcellation in the growth of callosal fibers (Innocenti, 1984); connections are initially diffuse and abundant and then become specified through elimination. In studies of cerebral dominance, hemispheric asymmetries are related to callosal thickness and may reflect pruning rather than accentuated growth (Witelson, 1990). Even cytoarchitectonic specificity has been attributed to the gradual connectivity of initially homogeneous neocortex (Creutzfeldt, 1977). Innately driven process determines the connectivity but so does experience. Early visual deprivation in animals can prevent foveal specificity and lead to a more ancestral state of diffuse or ambient perception (Ebbeson, 1984). Experience enhances the specification through constraints on emerging form (Brown, 1988).

At some point in morphogenesis, presumably after most anatomical connections are established, the parcellation effects that produced the structure of the brain give way to parcellation-like effects that characterize processing within this structure. The transition is from the elimination of connections in development to the inhibition of connections in maturity. The question is, Is specification in neuronal development only analogous to later processes of differentiation through inhibition, or are development and behav...

Table of contents

• Cover
• Halftitle
• Title
• Copyright
• Table of Contents
• Foreword
• Keynote
• I. The Behavioral Level: Learning
• II. The Network Level: Self-Organization
• III. The Neural Systems Level: Process
• IV. The Social Level: The Organization of Self in Society
• V. The Transcendental Level: Self-Organization on a Grand Scale
• Afterword
• List of Authors Cited