Copyright (c) 1997 George F. Michel and Erlbaum
 
 
 
 

The author's research described in the current article was supported by grants from the National Institute of Mental Health (1R0 MH 35528), the National Institute of Child Health and Human Development (1R01 HD 16107 & HD 22399) and the DePaul University Research Council.  Requests for reprints should be sent to G.F. Michel, Psychology, DePaul University, 2219 N. Kenmore Ave., Chicago, IL 60614-3504.  E-mail: gmichel@condor.depaul.edu

     The relationship between the developing psychological abilities and the developing brain of a growing child has been an object of research and theory for more than 100 years (e.g., Baldwin, 1895; Case, 1992; Gesell, 1946; Gibson & Peterson, 1991; Johnson, 1993; Lenneberg, 1967; McGraw, 1946; Prechtl, 1982).  Throughout this period, information about the postnatal developmental changes in the nervous system has been employed to account for presumably universal sequences in the development of various psychological abilities.  The modern discipline of developmental neuropsychology, since it examines the relation between the developing nervous system and psychological functioning, would appear to be ideally suited to provide an empirical and conceptual framework for characterizing the interface of neural and psychological development.
     However, developmental neuropsychology derives from the discipline of neuropsychology which traditionally combined two sets of information, neither of which incorporated a developmental orientation.  One set from neurology derived from the examination of the functional consequences (including social, cognitive, emotional, sensory, or motor disturbances) of various neural disorders produced by infections, toxic conditions, anoxia, focal damage due to stroke or injury, and such.  One objective of such examination is to provide a precise description of symptoms associated with damage to specific brain areas so that a patient may be assigned to a syndrome category on the basis of the symptoms.  Obviously, these "symptoms" overlap the second set of information derived from psychology.
     Psychological functioning is a broad and complicated phenomenon.  To understand it, it is necessary to segment it into smaller chunks for examination and analysis.  Historically, human psychological phenomena were divided into the three categories of thought, feeling, and action which subsequently multiplied into a large number of faculties.  A major concern in psychology is whether the categories of psychological functioning are meaningful units or merely arbitrary divisions.   Therefore, the pattern of functional deficits associated with specific neurological disorders has been considered to provide a "natural cleavage" for the identification of fundamental categories of psychological phenomena.   For example, case studies of brain damaged individuals support such divisions as:   certain aspects of speech production can occur in the absence of certain aspects of speech comprehension and vice versa; specific visuo-spatial tasks can be accomplished despite severe linguistic dysfunction; there is a lexical-nonlexical process of reading aloud and a distinction between surface and deep dyslexia; and implicit and explicit memory processes are distinct.  Thus, neuropsychological data seem to provide insight into the identification of "natural" categories of psychological phenomena.  Interestingly, a recent review by Rovee-Collier (1996) demonstrates that the categorizations of memory processes will have to be altered, if information from infant development is considered.
     Most neurologists assume an executive control model of brain-behavior relations in which the brain directs and psychological processes follow.  When applied to developmental phenomena, the model assumes that brain development precedes and potentiates changes in psychological functions and not the reverse.  Consistent with the assumption of executive control, the basic procedure in neurology is to correlate a functional deficit to a particular area of damage.  However, it can be argued that such correlations do not identify an executive control mechanism.  Rather, they identify what the damaged brain does to solve the problem of the missing area.  The missing area may have typically supported the function that has been disturbed or that the disturbance may be the consequence of the functional reorganization after damage.  This may be a small semantic difference but it represents a major theoretical difference.  The correlations do not indicate how brain areas enable specific functions; that is, they do not identify the neural means by which functions are achieved.  For example, the corpus callosum supports the transmission of certain forms of information from one hemisphere to another that are essential for accomplishing several high level functions, as demonstrated by the functional deficits associated with damage to the callosum.   However, nothing is known about the properties of the code by which information is transmitted or how the code affects the forms of information that can be transferred.  If such information were known, perhaps rehabilitation techniques that used alternative pathways could be discovered that would provide the interhemispheric transfer needed to support the achievement of the functions disrupted by callosal damage.
     To aid the neurologist in the assessment of the correlation of functional deficits with neural disorders, the neuropsychologist often administers a battery of standardized tests of general and specific psychological functioning to patients.  When the patients are children, standardized tests with specified "age-norms" are often employed so as to chart a patient's relationship to a population of similarly aged individuals.  Thus, these techniques, when unencumbered by conceptual confusions and methodological difficulties, permit neuropsychological research and practice to take the form of  "normal science" (Kuhn, 1962).  That is, well-understood "standard" techniques (e.g., standard psychological testing instruments) are brought to bear on well-defined problems (individuals suffering various sorts of brain damage) to provide conventionally acceptable interpretations of brain-psychology relations.
     However, when neuropsychological procedures are applied to individuals who are not suffering from conventional neural disorders, or when neuropsychologists attempt to assess the significance of "soft" neurological signs, or when the correlational procedures are used to provide explanatory theories for brain-behavior relations, then neuropsychology opens itself to issues beyond the "normal science".  These issues require examination of the conceptual framework within which the normal activities of the discipline are conducted.  Such examination usually includes the critical investigation of conventionally accepted and familiar techniques, concepts, and assumptions in order to identify the conditions on which their justification and usefulness depend.  Consequently, the meaning of standardized test results, the assumption of executive control in brain-behavior relations, and the value of case studies may be questioned.  This disrupts the normal routines of the discipline and too often leads some to dismiss these endeavors, and the debates that ensue, as mere semantics.
     Of course, neuropsychologists have always vigorously debated the proper methods that should be used in constructing theories of brain-behavior relations.  Although they have acknowledged that the dynamics of the processes generating brain-behavior relations are not observable, the observed correlation between brain damage and alterations in psychological functioning is used often as evidence in support of one sort of hypothesis about brain-behavior relations and for disconfirmation of other hypotheses.  Thus, case studies are accorded the status of experimental manipulations (Shallice, 1988).  These supported hypotheses are then used to explain normal brain-behavior relations or are used in the design of research on individuals not suffering from brain disorders.   Although this approach in neuropsychology has been criticized frequently (e.g., Efron, 1990; Ellis, 1987; Marshall, 1980), it is especially misleading in developmental neuropsychological research.
     Behavior develops through the interplay of various factors that are internal and external to the boundaries of the individual.  Although it is quite fashionable to acknowledge that behavior develops from the interaction of internal and external factors, often that acknowledgment is relatively sterile because the concept of interaction used connotes little more than an additive event.  The more interesting notion of interaction concerns the continual dynamic interplay among factors in which feedback, reciprocal influence, and circular causation occur.  In this dynamic sense of interaction, neuropsychological development involves non-linear causality; hence, analyses of the end-product would not allow accurate reconstruction of the generative processes.  To understand development, it is necessary not only to identify which factors play a role, but also to identify the processes by which they exert their influence.  Standard techniques and conventional concepts, suspect in adult neuropsychology, are even more questionable in developmental neuropsychology.  New principles, concepts, and techniques must be employed.
     Some of the theoretical and empirical work concerning the developmental dynamics of the biology-psychology interface has taken place within the tradition of developmental psychobiology.  This work may be unfamiliar to many psychologists and biologists because too often psychobiology is defined primarily by research activities that combine biological techniques with psychological topics.  Hence, any activity in which a conventional psychological topic (e.g., learning, motivation, memory, perception, attention, aggression, sleep, psychopathology) is examined in relation to a conventional biological technique or procedure (e.g., measures of neural functioning or anatomy) becomes psychobiology.  However, the meaning of psychobiology in developmental psychobiology implies integration of biology and psychology at the conceptual level to form a new discipline (Michel & Moore, 1995).  Also, developmental psychobiology was derived in part from a natural history approach to the study of animal behavior.  That approach requires adoption of a multi-leveled approach to an organism that has an adaptive relation to its environment throughout its lifespan.  This results in a somewhat different view of development (see Gottlieb, 1992).
     As a result of their different ancestry, developmental psychobiology and developmental neuropsychology characterize the biology-psychology interface somewhat differently.  Therefore, they will differ both in the manner by which research is designed and conducted and how results are interpreted.  Moreover, since developmental neuropsychology is, in part, motivated by clinical issues and developmental psychobiology is, in part, motivated by the acquisition of knowledge for its own sake, they will differ in the type of results that are considered pertinent.  Nevertheless, it is my contention that some of the ways by which developmental psychobiology approaches the study of brain-behavior relations might be incorporated into the framework of developmental neuropsychology.  I will describe three principles of developmental psychobiology concerning the dynamics of development that relate to the biology-psychology interface.  Then, I will briefly examine four issues in developmental neuropsychology (handedness, sex differences in behavior, critical periods, and modularity of brain structure/function) from a developmental psychobiological perspective and in relation to those three principles.  Finally, I will identify four contributions that developmental psychobiology can make to the conceptual framework of developmental neuropsychology.  My hope is that the particular twist on these notions from developmental psychobiology will be useful as developmental neuropsychologists approach their task.
Developmental Principles
     Developmental explanations involve a multiplicity of causal factors.  One principle to be derived from the notion of dynamic interaction is that the identification of one factor as contributing to the development of some characteristic does not preclude conclusions about the importance of other factors.  For example, discovery of the contribution of a genetic factor should not preclude discovery of the importance of non-genetic factors, including experience.  Conversely, identification of an experiential contribution to development of some psychological function should not preclude discovery of a genetic contribution.  Moreover, both genetic and experiential constructs require "unpacking" by developmental investigations that identify intrinsic and extrinsic genetic influences and, in mammals, maternal influences and self-generated social and environmental experiences (Atchley & Hall, 1991; Michel & Moore, 1995).   In unraveling the dynamic interaction in development it becomes apparent that all factors that have been discovered to contribute to the development of some function are equally important and that additional factors may yet be discovered.  Consequently, if one type of experiential factor is demonstrated as not influencing development, this does not imply that other experiential factors are equally unimportant.  The developmental psychobiological study of handedness and sex differences in reproductive behavior provide good examples of the value of this principle.
     Developmental explanations require specification of how development is accomplished.  A second principle derivable from a dynamic interaction is that the identification of the factors affecting development does not reveal how they accomplish their effect.  A study which identifies a difference in some factor (e.g., size of some brain structure or type of experience) as being related to a difference in developmental outcome does not reveal the process by which such a difference produces the change.  Developmental psychobiological research must describe exactly "how" that factor affects development and often such investigation reveals important, but not obvious, contributions of individual experience.
     Developmental explanations require the study of development.  A third principle is that the discovery that some ability or function is present early in life (e.g., at birth) and is similar to that of the adult or that some ability appears to develop independently of typical variation in social or physical environmental conditions should provoke additional attempts to comprehend the development of the ability.  Too often, such discoveries satisfy researchers and bring their investigations to a close.  However, this satisfaction does not identify the origin of the ability (development to) or the developmental consequences of the ability (development from).  Identifying aspects of the individual that do not respond to typical variations in environmental conditions only specifies a problem for developmental investigation, it does not provide an answer about how the individual's ability develops.  Also, the elicitation of adult-like patterns from immature individuals need not indicate presence of adult neurobehavioral organization.  Although elements of the neurobehavioral system that typically enter into the organization of the adult pattern may be present earlier, these early occurring elements may depend upon the presence of environmental support conditions that are unnecessary for the adult.  Explanations, based upon empirical evidence, that account for differences between early and adult sensitivity to such contextual support should become part of the developmental account of the adult pattern.
Four Issues in Developmental Neuropsychology.
     Handedness.  Hemispheric specialization of function is one of the most intriguing and popular issues in neuropsychology.  Various kinds of research on clinical and nonclinical populations indicates that the organization and control of speech and the comprehension of both written and auditory language is unevenly supported (i.e., lateralized) by neural mechanisms in the right and left hemispheres of the brain.  Certain spatial abilities are also unevenly supported by the two hemispheres but in apparent opposition to the lateralization of language abilities.  It has been proposed that the neural mechanisms of the left hemisphere are better able to support manifestation of language-like abilities whereas the neural mechanisms in the right hemisphere are better able to support manifestation of certain spatial abilities.
     Interestingly, most researchers choose right-handed males, without a family history of sinistrality, to investigate hemispheric specialization of function (Bryden & Steenhuis, 1991).  Conventional wisdom, supported by the results of some studies, is that the mechanisms supporting left hemisphere language abilities are present in nearly all right-handed individuals but may be less so in left-handed individuals.  Also, males are considered to be more strongly lateralized than females and those with familial sinistrality are considered to be genetically predisposed to sinistrality and hence the character of their hemispheric specialization is more problematic.  Of course, these notions reveal the importance of understanding handedness for understanding lateralization and hemispheric specialization of function.  However, they also raise the question of whether hemispheric specialization is of much adaptive significance for humans, since right-handed males without a family history of sinistrality represent only a minority of the population.
     Since handedness is both an aspect of hemispheric specialization and an influence on the lateralization of other functions, it is a fundamental issue in neuropsychology (Bryden & Steenhuis, 1991).  Yet, its development has been relatively ignored.  Perhaps, it is assumed that handedness reflects the development of neural mechanisms underlying lateralization that are under genetic control.  Indeed, the pattern of expression of hand-use preferences in a population can be predicted by various genetic models (Annett, 1995; McManus & Bryden, 1992; Corballis, 1995).  However, genetic models must be unpacked by developmental investigation to reveal how the characteristic is acquired (Principle 1).  My own work on the development of handedness (Michel, 1987, 1998) demonstrates that when unpacked by developmental analyses many intrauterine and postnatal experiences are shown to play important roles in development of an individual's handedness.
     There is a relation between the specific direction of neonatal supine head orientation preference, as comprised by a general postural asymmetry, and subsequent hand-use preferences in infants (Michel, 1981).  Using systematic longitudinal observations and simple manipulations, we were able to deduce that the relation is derived from self-generated experiences (Principle 2).  That is, head orientation preference, as a component of lateral asymmetry of neonatal posture, results in lateral asymmetry of visual regard of the hands and both hand and arm actions (Michel & Harkins, 1986).  The hand and arm actions are not components of the postural asymmetry but rather are induced by the direction of head orientation.  Similar asymmetries of visual regard and limb action in animal models can result in asymmetries in the functional architecture of the various structures of the nervous system associated with the use of the limbs (Spinelli & Jensen, 1982).  We proposed that the means by which neonatal postural asymmetries can contribute to the development of hand-use preferences (Principle 2) was by influencing the neural support of hand and arm actions resulting in lateralized asymmetries in the ability to coordinate visually guided object prehension and manipulation.  It is likely that such changes in asymmetrical architecture can be accomplished by alterations in denditric growth in various cortical and subcortical structures as a consequence of coincidental activity of neurons (Quartz & Sejnowski, 1998).
     Although the neonatal postural asymmetries have an impact on neonatal left and right hand fisting patterns and arm movements, as well as visual regard of the two hands, neither the neonate's posture nor the hand asymmetries were simply early expressions of the same neural mechanisms responsible for handedness (Principle 3).  Rather, systematic investigation of hand-use at different phases of development revealed differences in the range of conditions under which a hand-use preference may be manifested thereby identifying differences in its organization (Michel, 1998).  The reaching preference of a six month old infant is not the same as that of a 12-month-old, much less the same as the handedness of a six year old child.
     Our research revealed, also, that the specific strength of an infant's preference is affected by the handedness of the mother.  The mother's effect is apparent in patterns of play with the infant (Michel, 1992).  Thus, maternal handedness can influence the direction of an infant's hand-use preference only for those infant's with initial very weak preferences.  Otherwise, individual handedness seems to develop from a neonatal manifestation of a head orientation preference prompted by a postural asymmetry that may reflect intrauterine influences (Previc, 1991).  Thus, the majority of infants exhibit a rightward head orientation preference which promotes laterally asymmetric sensorimotor experiences that contribute to an early right hand-use preference.  A leftward neonatal head orientation preference is associated with an early left hand-use preference.  An initial preference in prehension facilitates the exploration of objects and the elaboration of manipulatory skills that yields a later manifestation of unimanual manipulation preference.  Manual differences in manipulatory skill yield subsequent differences between hands in bimanual manipulatory actions that result in "bimanual" hand-use preferences.  Subsequent hand-use continues to build upon these early preferences resulting in the expanded domain of preference associated with adult handedness.  Thus, a developmental psychobiological approach to the study of handedness can provide a different perspective from which to approach the development of hemispheric specialization of function.
     Sex Differences in Behavior.  The conventional notion is that males and females differ in many aspects of brain-behavior relations.  There is extensive documentation of sex differences in psychological functioning.  Sometimes, these differences are considered to reflect differences in child-rearing and socialization.  However, many sex differences in psychological functions do not vary across cultures nor do they appear to be deliberately taught or fostered in child-rearing.  When research reported the neuroendocrine mechanisms underlying sex differences in the reproductive behaviors in rats, similar mechanisms were proposed to operate in humans and the rat became a model for theoretical explanations of "species-typical" human sex differences and even of sex partner preferences (LeVay, 1991).
     Although most of the sex differences in the behavior of rats is dependent upon specific hormones associated with puberty, many of the differences are not reversed by switching the presence of these sex typical hormones.  The nervous system seems to be sex differentiated in its sensitivity to sex typical hormones.  The sex differentiated component of the neural sensitivity to pubertal hormones appears to be dependent upon prenatal and neonatal exposure to certain hormones typically associated with the male genotype (Yahr, 1988).  Alteration of the rat pup's early hormonal condition could reverse sex-typical sensitivity to pubertal hormones and reverse sex differences in adult behavior.  In consequence, it is generally assumed that the mammalian nervous system differentiates as female unless it is exposed early in development to certain hormones (e.g., testosterone) associated with the male genotype.  Early exposure to testosterone leads to the development of neural architecture that supports the manifestation of male-typical behaviors.
     However, the systematic research of Celia L. Moore (reviewed in 1992) has challenged this model and provided a dynamical interactional model of the development of sex differences in behavior for rats that appears to be even more relevant for understanding human sex differences.  Moore observed that each mother rat treats each of her male offspring differently than each of her female offspring in only one specific way.  The different treatment is primarily in the time the mother spends licking the anogenital area of her pups.  Mother rats must engage in anogenital licking in order to enable their offspring to urinate and defecate, otherwise the young would die before they become competent at these activities.  During anogentital licking, the mother also recycles many of the nutrients and most of the water lost through nursing the young.  Although all pups receive anogenital licking, males receive far more of it than do females.  Indeed, male pups exhibit greater sensitivity of maternally elicited "reflexes" that facilitate licking by the mother.
     Moore found that the perineal odor, associated with the male pup's urine, attracted the extra attention from the mother.  The perineal odor is a consequence of the early secretion of testosterone in the male and may involve its impact on the development of the preputial gland as well as the production of metabolites in the urine.  If  Moore made mother rats insensitive to odor of the urine, then they could not discriminate males from females and these male offspring were then reproductively and behaviorally deficient as adults.  Moreover, if female pups are provided with additional anogenital stimulation either artificially or by making that area attractive to mothers, as adults, these females will exhibit more male-like behavior.  This is especially significant because these females were not exposed to early testosterone; yet their nervous system, like that of a typical male, is more sensitive to male-typical pubertal hormones.  Moreover, although the male rats that were not treated as males by their mothers had the same exposure to neonatal testosterone levels, as adults, they were deficient in their sensitivity to male-typical hormones, as are normal females.
     Subsequent research by Moore (Moore, Dou, & Juraska, 1992) showed that anogenital licking has an impact on the size and architecture of the sex differentiated neural structures involved in the manifestation of sex differentiated behavior (Principle 2).  Thus, the development of sex differences in behavior of rats represents a dynamical interaction among neonatal and adult hormones, neonatal reflexes and odors, and maternal behavior; neither one of which is more important than the other (Principle 1).  The anatomical and neural architectural structures supporting sex differentiated behaviors are a consequence of a complex pattern of hormonal and experiential influences extended over time.  Although some of these differences in experience are prompted by sex differences in certain behavioral "reflexes" of neonatal pups that exhibit a similarity to adult behavioral differences, Moore found that neonatal behavior differed from that of the adult in both pattern and contextual support (Principle 3).
     Generalizing from Moore's research, we might expect that subtle but powerful experiential influences contribute to the development of "species-typical" sex differences in humans.  Rather than being satisfied with a demonstration that some differences in psychological functioning between males and females occur cross-culturally and/or relate to sex differences in gross neuroanatomy, developmental neuropsychologists should be seeking to identify the developmental origins of such differences.  Even when there is no obvious attempt to promote sex differences during child rearing or when sex differences occur early in the life of the child, it is possible that subtle experiential influences are contributing to their development.
     Critical Periods.  The concept of critical period is used to describe the notion that at certain points in development the individual is more receptive or vulnerable to environmental influences than at others.  One text in developmental neuropsychology uses the criteria specified in Nash (1978) for identifying a critical period (Spreen, Risser, & Edgell, 1995, p. 140) in which the sensitivity of the individual to an external stimulus must be triggered by some internal (maturational) factor with specific onset and offset times.
     The concept of critical period derives from 19th century embryology, when it was discovered that developmental trajectories of certain anatomical structures could be more easily altered by environmental manipulations at certain stages than at others.  When employed in the study of teratogenic agents in embryology, the effect of the teratogen (an external stimulus, usually chemical) on subsequent development was maximal only at certain stages and the critical stage differed among different structures.  However, when the concept was incorporated into natural history studies of animal behavior and subjected to rigorous developmental psychobiological examination, the concept lost much of its explanatory power.  Consequently, the concept was changed to "sensitive period".
     Sensitive period refers to the observation that the development of some specific behavioral characteristic of the animal is especially sensitive to the influence of certain stimuli at a particular stage of its development.  However, that same behavioral characteristic can be influenced by other, non-typical, stimuli at other stages of development (Principle 1).  Moreover, both the onset and offset of the sensitive period is influenced not simply by internal processes but by previous experiences (Principle 2), particularly self-generated experiences (Michel & Moore, 1995, pp. 38-43 & 407-409).  Thus, as the development of the behavior becomes better understood, the notion that the "period" is critical becomes less useful (Bateson, 1979).
     Indeed, too often definitions of critical periods are not distinct from the definition of development.  For example, Colombo (1982) defined critical period as "the time between the emergence anatomically or functionally of a given biobehavioral system and its maturation.  The system may be affected in this emergent but immature state (for better or for worse) by exogenous stimuli and this effect can be permanent should the system "harden" to maturity" (p. 263).  However, the development of any ability usually refers to the time between the initial emergence of the ability and the achievement of its "adult" or constant form.  Hence, according to Colombo, all of development is a critical period - a poor  definition, at best.  Although the concept of critical period may provide some descriptive service in the investigation of unusual conditions that disrupt normal development, it too misleading for use in understanding normal development.   From a developmental psychobiological perspective, the concept of critical period need not be incorporated into the conceptual framework of developmental neuropsychology.
     Modularity of Brain Structure/Function.  Until recently, most models of cognition assumed a centralized, linearly structured hierarchical organization with an executive control process.  Coordination of perception and action were presumed to be accomplished by executive processes or routines.  A similar organization was presumed for brain structure-function relations.  Researchers sought specific areas of the brain that presumably created and controlled the programs for coordinated action, complex perceptual abilities, comprehension and production of speech, reading and writing, etc.  However, focus upon the evidence that normal individuals can perform simultaneously two complex demanding tasks without detriment to either and that neurological patients can exhibit syndromes in which they have deficits in specific grammatical rules but not others, etc. has lead to the adoption of a modular notion of brain-behavior relations.  That is, it is presumed that cognition is composed of many semi-independent and separate, but interconnected, processors and the brain's structure represents this modular architecture.  The latter is consistent with modern neurobiological accounts of cortical organization in which certain functions appear to be related to specific anatomical columns (DeYeo, Felleman, Van Essen, & McClendon, 1994; Merigan, 1993).
     The modularization of cognitive functions automatically lends itself to a diagrammatic representation.  The functions can be arranged within a diagram that both separates them and identifies their interconnections.  Although this is not necessary, these diagrams often presume a modular architecture that permits some components to perform normally when others may have been totally eliminated.  Thus, the consequences of damage to a module or a connection can then be determined and predictions made about the deficits in cognitive functioning.  Modern neuropsychology employs a model of brain-behavior relations that assumes individual components function in the same manner whether they are joined together or disconnected.  Indeed, current neuropsychological theory depends upon the presumption of distinct cognitive modules that are also spatially separated in the brain (Bradshaw & Mattingley, 1995; Shallice, 1988).
     There can be little argument with the fact of modularity, only about its nature and extent.  The contents of any module must be unpacked and the details about how it operates must be determined.  Finally, for developmental neuropsychologists, the developmental history of the module must be identified.  For example, only a tiny minority of humans have ever achieved the ability to read and write with alphabetic script.  However, literacy appears to establish cognitive modules that can be selectively impaired by brain injury to produce a wide variety of different forms of reading and writing disorders.  Although some may argue that literacy builds upon "hard-wired" language modules, or certain combinations of language modules and visuo-spatial modules, there are reports of individuals with brain damage who show impairment of reading and writing skill without apparent impairment of other cognitive functions.  Of course, from the perspective of developmental psychobiology, the modules must be constructed during development and the notion of "hard-wired" modules is misleading.
     There is evidence from developmental neurobiological research that the apparent structural modularity of the brain arises as an epiphenomenon of simple developmental growth processes (Purves, Riddle, & LaMantia, 1992,1993).  Moreover, experiential processes profoundly affect neocortical structure-function relations, even in adults (Darian-Smith & Gilbert, 1994; Merzenich, Recanzone, Jenkins, & Nudo, 1990).  Developmental psychobiological (e.g., Bekoff, 1988) research demonstrates that functions that are modular at one level of description (e.g., behavioral) are intermingled at another level (neural).  Therefore, the observation that an individual can perform simultaneously two demanding tasks does not imply that they are subserved by distinct and separate sets of neural modules.
     The characteristics of dynamic systems theory (Michel, 1991; Thelen & Smith, 1994) appear to allow separate functions to be created from a multiplicity of interdependent processors.  That is, within the context of ongoing cycles of neural activity, signal enhancing processes emerge in relation to task, biomechanical, and contextual constraints.  There is no executive control hierarchy but rather a coalitional heterarchy in which information and control activities are decentralized and interactive.  Within the constraints of task, biomechanics and context, the ongoing neural activity either achieves state of dynamic equilibrium or not.  If not, some level of equilibrium must emerge eventually; otherwise, the system disintegrates.  When a perturbation (alteration in task or context) challenges a state of equilibrium by exceeding the buffering capacities of that state, a series of fluctuations begin.  From within the context of this systemic disorder and disequilibrium, a new dynamic equilibrium is afforded.  This resultant equilibrium is not a "set point" but a dynamic state.  Although it remains to be determined whether dynamic systems theory can provide the basis for comprehending neuropsychological phenomena without the postulation of "hard-wired" modules of structure-function, it is consistent with a developmental psychobiological perspective.
Contributions.
 There are four contributions that a developmental psychobiological approach can make to the field of developmental neuropsychology.
     1.  During development, new functional achievements derive from abilities present at earlier phases.  However, developmental psychobiological investigations have revealed that the derivation may not be constrained by any functional categorization of the abilities or by our perception of such categorization (Fentress, 1991).  That is, abilities that are apparently functionally distinct and separate may be developmentally related.  This developmental relationship is identifiable only by empirical investigation.  For example, infant supine kicking activities may be related to both neonatal "reflexive stepping" patterns and subsequent initial walking patterns although both intuitively and perceptually, kicking appears to be a functionally distinct pattern (Thelen & Cooke, 1987).   It is likely, therefore, that supine kicking activity may play an important role in the development of bipedal locomotion.  Similar developmental psychobiological investigations of prehatching spontaneous leg movements of chickens revealed their developmental relationship to postnatal locomotion abilities despite apparent differences in pattern (Bekoff, 1988).   Thus, developmental causality may not be identified necessarily by either an immediate temporal relationship or by an obvious functional similarity among phenomena.
     For the developmental neuropsychologist, the development of initial reading ability may not require practice with, or even exposure to, books or other reading materials; rather it may depend upon the development of certain forms of neural processing derived from motor skills acquired some time before reading begins.  Some difficulties in the acquisition of reading abilities may reflect differences in the timing of cognitive processes that, in turn, may have derived form mechanisms involved with motor control (Wolff, Michel, Ovrut, & Drake, 1990).  If so, then reading acquisition may be facilitated in dyslexia by creating alternative procedures for timing the cognitive processes involved in reading.  Perhaps, the success associated with individual tutoring programs for dyslexia resides in the inadvertent discovery of such alternative procedures.
     Typically, the developmental neuropsychologist will start a developmental inquiry by focusing upon the "finished" adult product and then searching for earlier elements of this product.  Since development likely builds upon the available raw materials, there is no logical reason that recycling of earlier achievements should be restricted to functionally similar categories.  Hence, developmental psychobiological research typically begins with the abilities of the "younger" individual and that individual's commerce with the sociocultural and ecological context to identify where these abilities lead (Michel & Moore, 1995).  In this way, precursors of adult abilities can be found which initially may not have been intuitively obvious.  This does not presume that the individual begins as a tabula rasa.  Rather, it presumes that the structure of the initial system may not be inherently the same as the adult structure.
     2.  The development of the nervous system depends on its context.  That context can include hormonal conditions (that, in turn, may be dependent upon sensory stimulated neural events), biomechanically created patterns of sensory feedback prompted by spontaneous activity of motor processes, behaviorally created alterations in the social and physical environment that affect sensory systems that, in turn, alter hormonal condition and neural state.  Hence, brain and behavioral development are in reciprocal interdependence.  For each behavioral ability whose development is demonstrated to depend upon the developmental achievement of a specific neural state, the development of that neural state is likely to have depended upon the achievement of a specific social and/or physical environmental condition prompted by a previous behavioral event.
     This means that neuropsychological development will exhibit a coactional non-linear causality that will require investigations to begin with the specification of abilities and states at one phase and then to try to identify their consequences for subsequent phases of development (development from).  Also, this means that the individual's development will be influenced by self-generated experiences.  That is, the social and physical environmental consequences of the individual's own actions will generate sensory experiences that will contribute toward the individual's subsequent neural development.  Thus, neuropsychologists will have to expand their expertise to include hormonal, immunological, biomechanical, and other non-neural factors when seeking the causes of behavior.
     3.  The individual is both a structured system and a part of a structured system (an Umwelt of perceivable social and physical environmental events).  This means that different combinations of internal and external conditions and events can form similar achievements and that quite different achievements can emerge from rather slight variations in these combinations.  Hence, developmental "timing" (e.g., "critical periods", "readiness" phenomena, "sleeper" effects,) depends upon combinations of internal and external events, not on internal "clocks".  Although certain neural structures exhibit clock-like characteristics and these may be incorporated into simple rhythmic events, developmental phenomena are not likely to be fully explained by such structures.  The timing and sequential organization of developmental phenomena are not the consequence of internal control systems.  This means that "vulnerability" will fluctuate because the effects of the same perturbation will be minimal or maximal depending upon the "state of coherence" of the individual-context system at the time of the perturbation.  Thus, to understand how a perturbation can have an effect will require understanding the state of coherence of the individual in context.
     4.  Developmental neuropsychology has greatly benefited from incorporation of research conducted on animal models of human phenomena.  However, adoption of a developmental psychobiological perspective will allow greater incorporation of natural history oriented animal research.  A natural history orientation to the study of animal behavior means that the research is conducted in order to comprehend the animal's behavior in its natural setting (Lehrman, 1971).  The emphasis is upon understanding the animal for its own sake and not as a model of human phenomena.  Humans are extensively symbolic in so many aspects of functioning, that animal models can match the realms of human neuropsychology only imperfectly.  Animals can only be weak models of human neuropsychological functioning, although they can reveal basic processes by which neural architecture may be organized and reorganized by function.
     The investigation of the biopsychology of various animal species can provide a conceptual context within which the conceptual framework by which we examine human biopsychology and neuropsychology can be challenged .  That is, in attempting to examine ourselves, we must be able to generate a frame of reference that is not completely dependent upon the original object of examination (i.e., ourselves).  Therefore, if we are to avoid simplistic confirmation of conventional wisdom and potential self-delusion, we must explore the development of other species so as to challenge and break our anthropocentric frame of reference.  Of course, it is not possible to completely eliminate the anthropocentric bias in the study of other animals.  However, when we approach that study with the intent to discover the animal's "world" rather than our own, we can achieve a perspective on both the animal's world and our own that is less based upon our socially and culturally derived self-reflective intuition (Detheir, 1969).  That is, the animal will challenge our attempts to cast it as a simplified form of human.  From this comparative perspective, we can come to better understand ourselves.
Conclusions.
     One impediment to understanding brain-behavior relations is that it incorporates an issue that affects theory choice in science in general - the issue of parsimony.  The assumption of parsimony means that we will prefer the simplest conceptualizations of brain-behavior relations.  Our preference for intuitively appealing explanations of any complex phenomenon might lead them to appear to be simpler and hence, intuitively appealing explanations for brain-behavior relations will appear simpler and more parsimonious.  Lay physics is simpler for most people than scientific physics.  Folk psychology is simpler than scientific psychology.  Are explanations that appear to be simpler always to be preferred over those that appear to be more complex?  Currently, developmental psychobiological investigation reveals that development is complex and often non-intuitive.  Given the assumption of parsimony, then such investigation may appear unnecessarily complicated.  Indeed, translation of the developmental psychobiological framework into procedures for practical application would require extensive alteration of nearly all social institutions (Michel & Moore, 1995).  It remains to be determined whether the extra effort needed to conduct developmental psychobiological research is warranted in terms of its eventual practical application.
     Current application of developmental neuropsychology is constrained not just by historical precedent, but also by societal values and procedures for reimbursement.  Approaches to neuropsychological assessment that begin to incorporate some of the framework characteristic of developmental psychobiology are time consuming and costly (e.g., Kaplan, 1991).  For example, Edith Kaplan has proposed that assessment adopt a process approach that would provide information about how the individual performed the tasks of the assessment rather than information about which aspects of the tasks were and were not completed within the time limits of the test.  Theoretically, the information obtained from a process approach specifies exactly the circumstances under which a patient's symptoms disappear and what contextual supports must be provided to enable effective functioning by the brain damaged individual.  This information could be used to establish new, nonintuitive, procedures for rehabilitation.  That her proposal has not been universally accepted attests to the difficulties facing the emergence of a developmentally rich neuropsychology.  That some have adopted her methods attests to the possibility that developmental neuropsychology may be on the verge of achieving a new level of science.  Now it is time for developmental psychobiologists to begin specifying the practical applications of their work.  Further collaboration between neuropsychologists and psychobiologists should foster such growth.