Vid Discussion
Developmental Psychology 1992, \W. 28, No. 6,1018-1029
Copyright 1992 by the American Psychological Association, Inc. 0012-1649/92/13.00
Developmental Psychology in the Context of Other Behavioral Sciences
Robert A. Hinde St John's College, Cambridge
and Medical Research Council Group on the Development and Integration of Behaviour Madingley, Cambridge, United Kingdom
Certain characteristics of psychology that have been instrumental to its success, such as emphases on an experimental approach, on group means, on theory-driven research, and on analysis but not synthesis, can be detrimental to progress if taken too far. In addition, psychology's success has led to its fragmentation into subdisciplines, with too little cross-fertilization. Future progress may depend on a focus not only on individuals, but also on individuals in a network of social relation- ships whose course is influenced by social norms and values. In this connection, it is helpful to distinguish a number of levels of social complexity and to come to terms with the dialectical relations between them. We must pay more attention to description as a first stage in the analysis of process, recognizing that description can never be perfect and that it must embrace the several levels of social complexity. We must also come to terms with the relations between the several levels of complexity, and thus between the several subdisciplines appropriate to them. This multidisci- plinary approach can be based on a study of relatively stable human behavioral characteristics and must include the relations among individuals, relationships, and culture. Some pointers can be found in a judiciously used evolutionary approach.
A centenary is a time to look back, to see how much has been achieved in the preceding decades. But the enormous progress made by developmental psychology hardly needs underlining, and self-congratulatory backslapping would be out of place in a journal to be read by developmental psychologists. A centen- ary is also a time to look forward, to assess the present con- straints on further progress and ask how they can be overcome. Let us then start with some gross generalizations, not because I believe for a moment that they are without exceptions, but be- cause they help focus attention on some current problems in developmental psychology.
In its early days, psychology needed to establish itself as a distinct discipline and to achieve recognition as a respectable branch of science. It achieved distinctiveness from biology/phy- siology by focusing on the psyche, and from philosophy pri- marily by adopting an experimental approach. It achieved re- spectability by attempting to ape physics—again by the use of the experimental method—and also by attempting a hypothe- tico-deductive approach, by an emphasis on objectivity, and also by the use of statistical tools.
Each of these, taken too far, has brought problems. A focus on the psyche came into conflict with pressures to study behav- ior objectively, which in turn led to a neglect of process. An overemphasis on an experimental approach led to an underem- phasis on people in the real world and to single-variable studies.
I am grateful to Patrick Bateson (1991) for editing a series of essays that brought together many of the issues discussed in this article, and to him, John Fentress, and Joan Stevenson-Hinde for comments on an earlier draft of the article.
Correspondence concerning this article should be addressed to Rob- ert A. Hinde, The Master's Lodge, St. John's College, Cambridge CB2 1TP, United Kingdom.
Attempts to ape physics led to an underestimation of the im- portance of description: much of classical physics dealt with everyday events for whose analysis description was unneces- sary, but the complexity of human behavior demands an initial descriptive phase.
The undervaluing of description led also to a belief that re- search should always be theory-driven. This is fair enough if not overstated, but it can lead to an unwillingness to allow the data to suggest problems. During the forties and fifties, a hypothe- tico-deductive approach led to a particularly narrow, theory- driven approach, with a focus on limited experimental situa- tions. Although theory-driven research is often a first priority, important advances can also follow if novel phenomena are seized and studied with the best tools available. Examples are provided by Bowlby's (e.g., 1969) following up of the finding that disturbed adolescents had had major and repeated separa- tion experiences in childhood, Andrew's (1991) discovery of the effects of testosterone on the persistence of motor patterns and the duration for which events are held in the working mem- ory, or Horn's (1985, 1991) discovery of asymmetries in brain function in chicks. Finally, an overemphasis on statistics can lead to a focus on group means with a neglect of individual differences.
These caveats are not intended to play down in any way the extraordinary progress made by psychology in general or by developmental psychology in particular. But a retrospective view and a recognition of past constraints can warn usof future dangers. And there is one further issue that is a direct result of psychology's success: its fragmentation into subdisciplines. De- velopmental psychology has, properly, become a field in its own right but, as a result, has become partially cut off from clinical, personality, physiological, and social psychology and from biol- ogy. Developmental psychology has focused largely on changes
1018
APA CENTENNIAL: INTEGRATING BEHAVIORAL SCIENCE 1019
with age and on group averages, but we need also to understand individuals, the primary concern of much clinical and personal- ity psychology. Physiological analysis leads to functions of parts, and although individuals function as wholes, that func- tioning depends on, but cannot be entirely explained by, pro- cesses within parts. And physiological analysis can aid behav- ioral understanding. A recent example is the manner in which Rosenblatt's (1991) survey of the physiology of parturition poses new questions about the onset of maternal responsive- ness. Social psychology is concerned largely with group phe- nomena, and developmentalists—knowing that children grow up in groups, that relationships are crucial to their develop- ment, and that values, expectations, and hopes held by the child and others shape development—need social psychologi- cal expertise. And ethology, as we shall see, can contribute prin- ciples and perspectives of importance to developmental psy- chologists.
We have here demands that could conflict: science proceeds by analysis, but one needs synthesis and the study of wholes as well; one needs to specialize, but that means studying only parts of a whole; one needs to describe phenomena, but one also needs to understand process; one needs concepts to cope with intangibles, but one must not lose discipline. It would be folly to suggest that these problems can be readily solved, but the following sections address a series of relevant and interre- lated issues.
There are three main themes. The first concerns the need to focus not only on individuals, but also on individuals in a net- work of relationships, and this in turn requires us both to distin- guish levels of social complexity and to come to terms with the dialectical relations between them. The second is that descrip- tion is a necessary first step, but can never be perfect. Our categories and concepts are essential heuristically but are never absolute, because we have at present no entirely satisfactory way of coping with entities that are both isolatable and intercon- nected and mutually influence each other. The third is the need to integrate developmental psychology with other disciplines. Here, because of my own biases, I refer especially to ethology, though I am aware that some of the ideas I ascribe to ethology also had other roots.
Levels of Complexity
Children grow up in a network of relationships and usually within families, which form parts of larger groups. It is thus necessary to come to terms with a series of levels of social complexity: physiological and psychological systems, individ- uals, short-term interactions between individuals, relationships involving a succession of interactions between two individuals known to each other, groups, and societies (Figure 1). Each of these levels has properties not relevant to lower levels, and at each level new descriptive and explanatory concepts are needed. For instance, we may describe the behavior of two individuals in an interaction as "meshing" well, but meshing is a concept irrelevant to the behavior of an individual in isola- tion. Furthermore, each level affects and is affected by other levels. Thus the course of an interaction depends both on the natures of the participating individuals and on the relationship of which it forms a part, and the nature of a relationship is
influenced both by the component interactions and by the group in which it is embedded. Furthermore, each of these levels influences, and is influenced by, the physical environ- ment and by the sociocultural structure of ideas, values, myths, beliefs, institutions with their constituent roles, and so on, more or less shared by the individuals in the relationship, group, or society in question.
Recognition of these levels is in no way an argument for unidi- rectional reductionism, because the dialectical relations be- tween levels are crucial (cf. Fentress, 1991). Each level, includ- ing that of the individual, must be thought of not as an entity but rather in terms of processes continually influenced by the dialectical relations between levels (Hinde, 1987,199 la). It will be apparent that such an approach always demands liaison be- tween a variety of disciplines.
Each level, as well as the sociocultural structure, has both objective and subjective aspects. For example, relationships have objective aspects that are apparent to an outside observer and subjective aspects that are specific to each participant, known in their entirety only to him or her, and shared only partially. Similarly, the objective aspects of the sociocultural structure may be partially codified in laws and customs, but the subjective aspects may be subtly different for each indi- vidual.
The position taken here is not so extreme as that of some who espouse dialectical determinism (review Hopkins & Butter- worth, 1990). It is of course basic that development must be studied at several levels simultaneously and that stability is, if not always momentary, at least dynamic. But though emphasiz- ing process, 1 would argue that the view that organism and environment are inseparable is not helpful. Although the diffi- culties of boundary definition matter and must be borne in mind, the essential thing is to come to terms with the continu- ous interplay (e.g, Markova, 1990; Mead, 1934). And while re- jecting linear causal chains, I would stop short of saying that development can never be adequately predicted on the basis of individual elements, although that perhaps reflects an aspira- tion rather than an achievable goal.
Description
How can one cope with multiple levels of analysis simulta- neously? How can one nail down entities constituted by continu- ous dynamic processes? Description and categorization are clearly necessary as a preliminary to—or as a part of (Carey, 1990)—analysis, but in describing such phenomena one inevita- bly simplifies the complexity of real life. A delicate balance must be struck between using categories and concepts that one can handle and distorting nature. And that one is compromis- ing must not be forgotten.
This is a lesson that is being learned slowly by ethologists. For example, the early concept of the Fixed Action Pattern (FAP), used to refer to a species characteristic movement pattern, seemed clearcut. Gradually, it became apparent that all FAPs were variable, and the concept became replaced with that of the Modal Action Pattern (Barlow, 1977). Similar issues arise in child development. In studies of preschoolers, "aggressive be- havior" seemed a clearcut category, but experience soon showed that the category boundaries are hard to define and
1020 ROBERT A. HINDE
SOCIETY
SOCIO-CULTURAL STRUCTURE
INTERACTION
INDIVIDUAL BEHAVIOUR
PHYSIOLOGICAL FACTORS
Figure 1. The dialectical relations between successive levels of social complexity.
PHYSICAL ENVIRONMENT
that the category is itself heterogeneous. But subdivision met similar problems. The subcategories into which it is usually divided (e.g., instrumental aggression and teasing aggression [Feshbach, 1970]) themselves have shady boundaries. It is nearer the truth (though not necessarily facilitatory of research) to recognize that aggressive acts involve other behavioral sys- tems—for instance, tendencies to acquire objects (flcquisitive- ness) or status (assertivenesty—and that the nature of the aggres- sion shown depends on interactions between these systems (Fig-
Aggressiveness
Acquisitiveness
Assertiveness
Figure 2. Model of the relations among three propensities and ag- gressive behavior. (Aggression would be shown if the current state were represented by a point above the striped surface.)
ure 2). As another example, Stevenson-Hinde (1991) has argued that although fear behavior and attachment behavior are to be thought of as discrete behavioral systems, "the postulation of discrete behavioural systems should not obscure relations be- tween them. Activation of a fear behaviour system may lead to activation of an attachment behaviour system" (pp. 325-326) and activation of the attachment behavior system may inhibit the fear.
This tendency for systems to change their state or even their properties according to the broader context within which they are operating has been repeatedly stressed by Fentress (e.g., 1991). For instance, at a lower level of analysis, Getting and Dekin (1985) have shown that the neural networks operative in the swimming of the mollusc Tritonia are reconfigured into different functional circuits according to the behavioral state of the animal and that the neurones involved cannot be clearly categorized as motor neurones, central pattern generators, and so forth. And at higher levels of social complexity, relationships or families may change their characteristics with the context. Fentress sees the difficulties in understanding behavior and development as stemming in large part from the difficulty of comprehending that, at all levels of complexity, systems must be both self-organizing and interactive with other systems. "Varying forms of behavioural taxonomy clarify certain proper- ties of expression while potentially obscuring others. Unitary 'boxes connected by arrows' taxonomies often do not work, in part because they too easily draw our attention away from the properties of the arrows that in turn may affect the properties of the boxes" (Fentress, 1991, p. 98).
From this perspective, it is not surprising that individuals behave differently in different social contexts. Stevenson- Hinde (1986) has pointed out that so-called "child characteris- tics" refer to characteristics that lie on a continuum from indi-
APA CENTENNIAL: INTEGRATING BEHAVIORAL SCIENCE 1021
vidual characteristics to relationship or situation characteris- tics, with height and weight but few if any psychological charac- teristics at the individual end, temperament dimensions close to but at varying distances from it, and attachment categories near the relationship end.
It is possible that we should view some age changes in the same way. We are accustomed to the concept of "age-appro- priate behavior," ascribing underlying similarities across ages to "heterotypic continuity" (Kagan, 1971) and the changes to changes in the system concerned, but they could also be caused by changes in relations with other systems. For example, digit span increases between infancy and adulthood, but the evi- dence indicates that memory span remains constant after about age 4, the changes being caused by a domain-specific increase in knowledge about the materials (Carey, 1990).
Because of this lability in the elements and in the relations between elements, every generalization should be accompa- nied by a statement of its limitations—a requirement that makes description of both behavior and context even more nec- essary. As an example of the importance of this, Radke-Yarrow, Richters, and Wilson (1988) found that higher rates of initial child compliance were related to more positive mother-child relationships only in families categorized as "stable," and ma- ternal use of harsh enforcement was associated with more nega- tive mother-child relationships only in "chaotic" families. Again, Stevenson-Hinde and Shouldice (1990) found that mothers of securely attached children tended to overestimate their children's shyness, whereas mothers of insecurely attached children tended to underestimate.
Therefore, we must recognize that description and classifica- tion nearly always involve trying to push nature into pigeon holes when the fit is by no means perfect and that, for psycholo- gists, description must embrace the several levels of social com- plexity.
Developmental psychology, concerned with what children do or can do at different ages, has not neglected description, but it is worth emphasizing two issues. First, at the behavioral level, there are two routes to description: (a) one that refers ultimately to patterns of muscular contraction, and (b) one that refers to the consequences of action or the meanings behind action. Each has its uses and advantages (Hinde, 1966). Taking a lead from studies of lower species, such as fish, some researchers (e.g., Blurton Jones, 1972) have attempted to describe children's behavior by focusing solely on the former route. However, chil- dren are not fish, and such attempts have proved on the whole sterile. Better ways for describing children's behavior, which do not assume that behavior is all we are interested in and take account of the meanings behind actions, are available (e.g., Caldwell, 1969; Lytton, 1973).
Second, description is necessary at each level of social com- plexity, and the more complex the phenomenon, the more se- lective description must be. A special problem arises in the description of relationships (and higher order phenomena). De- velopmental psychologists normally study interactions, for in- stance, studying mother-child play across a number of dyads and making generalizations across dyads. Relationships involve a number of types of interaction and cannot be described from generalizations across dyads about interactions, because the different interactions within each relationship affect each
Interactions approach
Dyad
A - B
C - D
E - F
A - B
C - D
E - F
Relationships
Dyad
Interaction type
X S > Generalization
J
Y > > Generalization
J
approach
Interaction type
A - B X-"|
A - B Y \ > Generalization
A - B zJ
C - D
C - D
C - D
Y > > General izat ion
Figure 3. The contrast between achieving generalizations about interactions and generalizations about relationships.
other. Rather, each relationship must be described, and only then can generalizations be made across dyads (Figure 3).
Attachment theory involves a procedure for categorizing some aspects of mother-child relationships (Ainsworth, Blehar, Waters, & Wall, 1978; Cassidy & Marvin, 1989). A means for classifying other characteristics of relationships is given by Hinde (1979,1991a).
Relations Between Levels
Even though analysis tends to move from more complex lev- els to less, the importance of crossing and recrossing in both directions cannot be overestimated. The relations between lev- els of social complexity are well established in studies of physiol- ogy and behavior (e.g., Andrew, 1991; Horn, 1991; Hutchison, 1991), but they are equally important in developmental studies. It is not only that similar principles of organization may be repeated at different levels, it is also necessary to trace causal relations between them.
Both the experiences a child has in interactions with others and the effects of those experiences on the child himself or herself depend on his or her nature. Those interactions will affect and be affected by the relationships of which they form part, and those relationships are similarly related to the family or group. Each of these levels may also be affected by the socio- cultural structure, by the myths and values current in the fam- ily, group, or society. Thus we need to come to terms with the dialectical relations between levels.
Consider, as an example, the genesis of a fear of snakes. Chil- dren brought up in an institution who have never seen a snake
1022 ROBERT A. HINDE
show little fear if they first encounter one at 30 months, but they avoid a snake crawling on the ground from about 3 years (Prechtl, 1950). Children also show spontaneous fears of other objects or situations that might have posed a real threat in hu- man's environment of evolutionary adaptedness, such as spiders, heights, darkness, and being alone. Humans are much less prone to develop spontaneous fears of other situations that are genuinely lethal in modern society but that were not present earlier in evolutionary history, such as cars or bombs (Marks, 1987). It is thus not unreasonable to suppose that a propensity to fear, or to learn to fear, snakes is part of the human biological heritage.
Anecdotal evidence suggests that the extent of the fear is much influenced by social referencing. The child looks at others, and especially at a trusted other, and imitates their re- sponse (Emde, 1980; Klinnert, Campos, Sorce, Emde, & Svedja, 1983). Comparative evidence provides strong support for this view thus: (a) Wild-reared rhesus monkeys tested in the laboratory nearly always show fear of snakes; (b) laboratory- reared monkeys do not show fear of snakes; (c) laboratory- reared monkeys shown a videotape of a wild-reared monkey showing fear of a snake become afraid of snakes thereafter; and (d) laboratory-reared monkeys shown a "doctored" videotape of a wild-reared monkey apparently showing fear of a flower do not become afraid of either flowers or snakes (Mineka, 1987). There is thus clear evidence that rhesus monkeys have a pro- pensity to fear snakes that depends for its full realization on the experience of seeing others respond fearfully to snakes. This in turn increases the plausibility of a similar explanation of snake fears in humans.
Some individuals develop snake phobias, showing a fear of snakes out of all proportion to the threat they present, a fear that is irrational and is beyond voluntary control. It is reason- able to suggest that the role of snakes as a symbol in our culture is related to these issues. Snakes play an important part, and have played an even more important part, in our mythology. In the myth of the garden of Eden, in the Rubens paintings of snakes gnawing at the genitals of those cast down into Hell, snakes symbolize evil. Therefore, if we are really to understand fear of snakes and the symbolic role of snakes, we must come to terms with a series of dialectical relations among the propensity
Fear of Snakes
Snake Myths Behaviour of caregiver
Propensity to fear snakes
.Snake fear & phobias
Figure 4. The genesis of fear of snakes.
to fear snakes, social referencing within relationships, and snake myths within the sociocultural structure (see Figure 4 and further discussion in Hinde, 1991a).
Let us consider a very different example that also suggests complex links between the levels of social complexity. In a study of families in a research apartment in Bethesda, Mary- land, Radke-Yarrow et al. (1988) found the following: (a) There was a high level of concordance in negative affect between the members of mother-child dyads, indicating strong interdepen- dence within the dyad, (b) In families in which the mother- younger child were concordant in negativity, both mother- older child relationships and sibling relationships tended also to be concordant, (c) Mothers showed more negative affect in families of low socioeconomic (SES) status. When the instabil- ity and unpredictability of life circumstances in the families were examined, it appeared that the link between low SES and maternal negative affect was primarily due to the corrosive hardship of unpredictability and disorganization, (d) As men- tioned earlier, the relations between indices of maternal control interactions and the nature of the mother-child relationship varied with family stability. Thus higher rates of child compli- ance were related to more positive mother-child relationships only in stable families, and maternal use of harsh enforcement was related to more negative mother-child relationships only in more negative ones, (e) The relation between child characteris- tics and the mother-child relationship differed according to the sex of the child. Thus shy girls had more positive relation- ships with their mothers than nonshy girls, whereas shy boys had worse relationships than nonshy boys. Radke-Yarrow et al. ascribed this difference to other child characteristics associated with shyness. A very similar finding in Britain by Simpson and Stevenson-Hinde (1985) was ascribed to maternal values: Mac- coby and Sants (personal communication) showed that Califor- nian mothers like little girls to be shy and little boys not to be. Although these data were cross-sectional, they strongly suggest influences among individual characteristics, relationships, fam- ily characteristics, and the sociocultural structure of beliefs and values. Dunn's (1991) important studies of sibling relationships within the family led to a similar conclusion.
Not clearly demonstrated by these data, but important in the long run, are influences up the levels of complexity. The nature of the family depends on those of the family members and on their relationships, and the values and beliefs of a society stem ultimately from processes in individuals.
Nature-Nurture: Relatively Stable Characters, Constraints on Learning
For logistical reasons, every study in developmental psychol- ogy has limits. One cannot trace all the dialectical relations shown in Figure 1 in every investigation. A starting point is therefore needed. Can one identify simple items or properties of behavior that can fill this role? The previous discussion of the difficulties of describing behavior indicates that one must be content with approximations, with categories heuristically useful but shady at the edges.
A false start involved the view that behavior or propensities could be divided into those that are innate and those that are learned or otherwise acquired. Although this error has long
APA CENTENNIAL: INTEGRATING BEHAVIORAL SCIENCE 1023
been recognized (Bateson, 199 la, 1991b; Oyama, 1985), it still persists. Development involves an interplay between the indi- vidual and the environment. The current state of the individual influences which genes are expressed, and individuals influ- ence and change the world they encounter. At the present time twin and adoption studies are providing new insights into the interactions between genetic and environmental factors in de- velopment (e.g., Plomin & de Fries, 1983; Scarr & Kidd, 1983).
Although the dichotomy of innate versus learned behavior is false, it is possible to arrange characters along a continuum from those that are relatively stable with respect to environmen- tal influence to those that are relatively labile (Barlow, 1989; Hinde, 1966,1991a). Thus there are some characters that ap- pear in virtually the whole range of environments in which life is possible ("stable" characters): either the processes involved in their development are so regulated that they appear over a wide range of experiential influences, or the factors relevant to their development are ubiquitous. By contrast, characters at the la- bile end of the continuum appear only over a narrow range of conditions. It will be noted that this formulation differs from the innate-learned dichotomy in that (a) it involves a contin- uum and (b) a characteristic may be influenced by experience but yet is stable because the relevant influences lack specificity or are ubiquitous. However, the level of analysis at which the character is denned may be crucial. Thus the broad details of the motor pattern of smiling form a stable human characteris- tic, yet its fine details and the circumstances in which it is given are labile. Furthermore, development may be stable up to a certain point and labile thereafter, or labile first and stable later.
For some (but not all) problems such relatively stable charac- ters can provide us with starting points, provided, however, that we remember that they will be subject to variation. It is imprac- ticable to make a list of such characters, partly because a list of mundane characters would be tedious, and partly because the cross-cultural data are not adequate to prove cross-cultural sta- bility for any characters. However, they might include aspects of perception, motor patterns, stimulus responsiveness, motiva- tion, cognitive processes, predispositions to learn (including the capacity for language), and so on (see Hinde, 199la).
Of course, each such "relatively stable" character itself poses a developmental problem. Because the degree to which genes are expressed may depend on the environment, and because suscep- tibility to the environment may depend on the genetic constitu- tion, the constraints on their variability themselves involve an interplay between genetic and environmental influences. The same is true for subsequent development.
The importance of constraints on learning and predisposi- tions to learn must be emphasized here (Hinde & Stevenson- Hinde, 1973; Seligman & Hager, 1972). The earlier work on this subject concerned animals, where cross-species comparisons threw genetic constraints on development into relief. For in- stance, the chaffinch (a small bird) has to learn its song, but it will learn only songs with a note structure similar to the spe- cies-characteristic song. The bullfinch learns preferentially the song its father (biological or adoptive) sang (Thorpe, 1961). Even the capacity to acquire individual distinctiveness in sing- ing behavior, essentially creative in nature, is to be seen in this light (Marler, 1991).
Humans as a species presumably also have similar con- straints, though we recognize them only in the observation that some tasks or experiences are difficult to learn. Indeed it can be agreed that efficient learning requires inbuilt constraints (John- son-Laird, 1990). Within the human species, similar con- straints probably operate in autism. Autism is known to have genetic "bases" and involves specific deficits in understanding of emotion caused by beliefs (Baron-Cohen, 1991). It has also been suggested that male and female humans differ in their predispositions to learn (Hinde, 1987).
Constraints or predispositions may equally well be environ- mental in origin in both animals (Bateson, 1987; Gottlieb, 1991) and humans (e.g., Butterworth & Bryant, 1990; Sameroff & Chandler, 1975). Insofar as an individual is what he or she is as a consequence of prior experience, and future development depends on current state, all development is channeled by expe- rience.
Relationships and Individuals
The critical question for the developmental psychologist is how individual characteristics are affected by the relationships experienced. Strong associations between parenting practices and child characteristics, involving social behavior (e.g., Baumrind, 1971; Bretherton, 1985; Maccoby & Martin, 1983), affective behavior (e.g., Easterbrooks & Emde, 1988; Radke- Yarrow et al., 1988), and cognitive dimensions (e.g., Goswami & Bryant, 1990) have been demonstrated, and although it must be assumed that influences operate both ways, in at least some cases there is an effect of parenting practices on the child.
But the issues are not simple. First, some individual charac- teristics may be influenced by relationships more than others, and the extent to which any one characteristic is affected may change during development. Thus the propensity to show fear may be relatively independent of relationships from 0 to 6 months, subsequently modified by relationships and reinforce- ment, and later still become relatively fixed (Stevenson-Hinde, 1988).
Second, in the case of relatives, and especially parents, it is by no means easy to distinguish genetic from experiential influ- ences. First, similarities in genetic constitution may predispose the child to respond to environmental events similarly to, for example, the parents. Second, similarities in genetic constitu- tion may cause the child to select or create an environment similar to that to which the parents preferentially respond. Third, parents may be predisposed genetically to provide their children with an environment conducive to the development of particular characteristics. For example, shy parents may both pass on genes associated with a predisposition to develop a behavioral style that might be labeled as shy and create an envi- ronment in which their children saw few strangers. Finally, par- ents and others may react differently to children of different genotypes (e.g., Jaspers & Leeuw, 1980; Plomin, 1986; Plomin & de Fries, 1983; Scarr & McCartney, 1983).
With regard to the processes whereby interactions within re- lationships have long-term effects on child behavior, until re- cently most work focused on reinforcement and modeling. Current interest centers on possible cognitive intermediaries between attachment relationships and subsequent interactions.
1024 ROBERT A. HINDE
The quality of the child's attachment relationship with the mother predicts the character of later peer interactions (e.g., Sroufe, 1983; Sroufe & Fleeson, 1986; Turner, 1991), suggesting that it affects some aspects of the child. Bowlby (1969), taking a lead from Craik (1943), postulated that the child forms internal working models of self, of others, and of their relationships. During the last decade, this idea has achieved increasing promi- nence. Initially, although heuristically useful, it was too ill-de- fined to serve as a scientific concept (cf. MacCorquodale & Meehl, 1954). For instance Main, Kaplan, and Cassidy (1985, p. 68) described it as "a mental representation of an aspect of the world, others, self, or relationships to others that is of special relevance to the individual," and elsewhere as "a set of conscious and/or unconscious rules for the organization of in- formation." In this and other articles, additional properties were ascribed to the concept, many of which were isomorphic with the phenomena they were seeking to explain (Hinde, 1989a). Furthermore, there were both methodological and con- ceptual differences in the way in which the concept was used by different workers (Crittenden, 1990). Now, however, the con- cept of internal working model is in an exciting stage of develop- ment, involving inputs from work on cognition by both cogni- tive psychologists (e.g., Johnson-Laird, 1983,1990) and develop- mental psychologists/psychiatrists (e.g., Stern, 1985, 1991). Bretherton (1990) conceptualized internal working models as systems of hierarchically organized schemata, with the models of self, others, and the world interlinked and mutually influenc- ing each other (cf. Fentress, 1991). Used in this way, the concept is becoming more than a useful metaphor and is able to inte- grate data on psychopathology, the transmission of patterns of parenting across generations, the relations between communi- cation within attachment relationships and communication about such relationships to third parties.
Effects of Relationships on Relationships
A child grows up in a network of relationships, and the dif- ferent relationships may affect each other (Hinde & Stevenson- Hinde, 1988a, 1988b). The influence of relationships on rela- tionships has been of interest to three groups of workers. Pri- matologists have come to recognize that relationships affect relationships within primate groups and, with observational and experimental evidence, that the mother-infant relationship is crucially affected by others (Hinde, 1972,1983). Child devel- opmentalists have demonstrated that the quality of a particular relationship in the family may be related to that of another. For instance, the marital relationship may be related to the mother- child relationship (Christensen & Margolin, 1988; Easter- brooks & Emde, 1988; Engfer, 1988; Meyer, 1988), the mother- child relationship may be related to the sibling relationship (Dunn, 1988a, 1988b), divorce may have long-term sequelae for the children (Hetherington, 1988; see also Rutter, 1988), and effects of inadequate parenting may be transmitted across gen- erations (Belsky & Pensky, 1988; Caspi & Elder, 1988; Gross- man, Fremmer-Bombik, Rudolph, & Grossman, 1988; Patter- son & Dishion, 1988). Some of the mechanisms involved are summarized by Hinde and Stevenson-Hinde (1988b).
A third group concerned with the effects of relationships on relationships has been the family systems theorists. Although
there are many points of contact between them and develop- mental psychologists working on similar problems (P. Minu- chin, 1985), the family systems theorists (themselves diverse) have developed a rather distinctive orientation and vocabulary. They emphasize the family as an open system, with organized patterns of interaction that are circular in form. The family system has homeostatic features that maintain the stability of the patterns within it, but may periodically undergo perturba- tions requiring a reorganization of patterns. The individuals constituting the family are seen as interdependent and distrib- uted across subsystems which have their own integrity and whose interactions are governed by implicit rules and bound- aries (e.g., P. Minuchin, 1988).
It will be apparent that many of the properties emphasized by family systems theorists are compatible with the interdepen- dent yet self-organizing systems whose importance is empha- sized by Fentress (1991) at lower levels of analysis (see earlier discussion). It is important to recognize that the family as a whole can have properties with some degree of independence from the behavior of its component units—analogous (only) to Hoyle's (1964) finding at a quite different level that the regular stepping movements of insects are not accompanied by fixed patterns of electrical activity in the motor neurones. Further- more, the concepts of family systems theorists are clearly po- tentially compatible with the dialectical relations between lev- els shown in Figure 1.
My own view is that the level of dyadic relationships merits special attention for developmental psychologists, in that it is by interactions within relationships that development is af- fected. From there it is possible to assess how interactions and relationships are affected by third parties (e.g., Clarke-Stewart, 1978; Corter, Abramovitch, & Pepler, 1983; Barrett & Hinde, 1988), other relationships, and the sociocultural structure. To foster further links between family systems theorists and devel- opmental psychologists, it would be desirable to clear up a few conceptual issues (Hinde, 1989b). For instance:
1. Family systems theorists emphasize the family as an "orga- nized whole" and ascribe to it homeostatic properties. But the maintenance of the family as a functioning unit depends on the behavior of individuals within their relationships. Family orga- nization may derive either from personal goal seeking—in which individuals attempt to create a family that suits their own personal needs, and the resulting pattern of relationships is a consequence but not a goal of the behavior of individuals—as well as from interpersonal goal seeking, involving efforts to make the constituent relationships conform to an ideal or de- sired pattern. In both cases, the goals may be unconscious or loosely defined. In any case, the processes that contribute to stability are diverse and may reside in one or more individuals or relationships. And it may involve attempts to approach an equilibrium or goal state whatever the current state, to ap- proach it only so long as the current state remains within cer- tain limits, or to avoid an undesirable state.
2. The same individual may belong to more than one subsys- tem. Thus the mother is part of both the spouse and the mother-child subsystem. An advantage of this subsystem ap- proach is that it permits description of separate patterns for different subsystems composed of the same people (e.g., spouses are also parents). An advantage of the relationships
APA CENTENNIAL: INTEGRATING BEHAVIORAL SCIENCE 1025
approach advocated here is that it calls attention to the effects of interactions on interactions within a relationship and be- tween relationships, so that a mother's marriage may affect (or be affected by) the mother-child relationship. But there is a clear need for an unambiguous definition of the subsystem con- cept. For instance, does the concept have a subjective reality for the participants? If it does, how does it differ from relation- ships? If it does not, as would appear to be the case with the "three generational subsystem" postulated by P. Minuchin (1985), is its reality confined to the mind of the therapist?
3. Some family therapists downplay the role of the individ- ual so far as to hold that attempts "to quantify the relative input of members of a system" do not make sense in a systems frame- work (P. Minuchin, 1985, p. 300). Developmental psychologists may disagree, especially if the questions asked are concerned with changes in or differences between relationships or families and are carefully phrased (Hinde, 1979).
4. Earlier, the importance of distinguishing between the ob- jective and the subjective reality of aspects of the sociocultural structure was noted. This distinction could be important to family systems theorists. Are family tasks, family myths, and family style descriptive concepts useful to the therapist, or are they (consciously or unconsciously) part of the perceptions of the participants, and thereby influencing process? And the fam- ily world view, concerned with the family's self-perceptions, may be shared by family members, but there may also be marked differences between family members in the way they perceive the family.
These somewhat academic points are perhaps tangential to the clinical achievements of the family systems approach and are intended only to supplement the important efforts made by P. Minuchin (1985) and others to bridge the gap between clini- cians and developmentalists. In my own view, the means by which relationships affect relationships pose crucial problems for developmental psychology.
Individual Differences
Developmental psychology's successes would have been im- possible without the use of statistical techniques. Nevertheless, their widespread and proper use has resulted in a focus on group means and a neglect of individual differences, although it is often differences that help us understand processes (Dunn, 1991; Rutter, 1991). Furthermore, a neglect of individual differ- ences can lessen the value of the data to the clinician who is attempting to deal with individual cases. The advantages of a case study approach have recently been described forcefully by Radke-Yarrow (1991): for example, bringing balance to re- search dominated by group or variable-oriented research and greater understanding of the relations between behavioral sys- tems; light thrown on children at the extremes and on children who show resilience in adverse circumstances or who fail in favorable ones; and the sharpening up of data on turning points in development usually obscured by group variance, thus per- mitting process to be studied more directly. To the biologist, individual differences raise the further question as to whether they merely represent noise in the system or whether they are adaptive: some examples of the application of this approach to children are given in a later section.
As Radke-Yarrow (1991) pointed out, a revival of interest in individual case studies is a recent phenomenon, and statistical techniques for dealing with multiple levels of data on few indi- viduals are at present poorly developed (but see D. H. Barlow & Hersen, 1984; Kazdin, 1982). Although this certainly does not mean that case studies should not be pursued, there is another approach that could take us some of the way. Statistical tech- niques that rely on linear correlational procedures can be mis- leading, and for many purposes it is preferable to attempt to categorize children (Hinde & Dennis, 1986). Examination of those children who appear to be exceptions to the initial catego- rization can lead one to new generalizations. Iteration of such a procedure can approach the individual.
An example that goes some of the way toward this goal is shown in Figure 5, which plots an index of maternal warmth against maternal strong control in 4-year-olds. The children are categorized according to whether they were in the top third, middle third, or bottom third on aggression in preschools. In three replications, aggression was found to be lower when con- trol and warmth were more or less in balance, that is, in the central area, termed authoritative after Baumrind (1971), than in the authoritarian, permissive, or indulgent areas. This of course does not necessarily mean that the dimensions plotted were the crucial ones: in fact, the mother-child relationships differed between areas in many dimensions in addition to warmth and control. But there were a few high aggressives in the authoritative area and a few low aggressives in the others. These exceptions were found to differ from other individuals in the same area on some of these other aspects of the mother- child relationship. If the sample size were adequate this proce- dure could be iterated to approach the individual level.
Links With Ethology
Many developmental psychologists imagine that any input biology/ethology might have concerns parallels between ani- mal and human behavior. Of course, parallels can be found, especially in relatively simple patterns—for instance, in rooting behavior and the Moro reflex. Some human expressive move- ments can be traced back to prehuman forms (e.g., Eibl-Eibes- feldt, 1975; van Hooff, 1972). But anthropomorphism is danger- ous, and parallels can be misleading. In some cases they are revealing only if one finds the right level of analysis. For exam- ple, behavioral development is disturbed by separating infant from mother for a week or two in both rhesus monkeys and humans. However, the evidence indicates that human children are more disturbed if they are away from home in a strange place during the separation period, whereas rhesus monkeys are more upset if they stay in the familiar group environment and the mother goes away. The difference seems to be that, in rhesus monkeys, the mother-infant relationship is more dis- turbed under the latter conditions, because the mother has to reestablish her relationships with her group companions when she returns as well as to cope with her demanding infant. What is common between monkeys and humans is that the more the mother-infant relationship is disrupted, the more the infant is disturbed (Hinde & McGinnis, 1977; cf. Rutter, 1991).
As this last example shows, rather than simple parallels, one should look for principles abstracted from animal data whose
1026 ROBERT A. HINDE
o o
O
Authoritarian
I
Indulgent
-1.5 -1 -.5 0 .5 1
Maternal warmth
1.5
Figure 5. Relations between maternal warmth and maternal control at home, and the aggression shown by 4-year-olds in preschool. (The children are categorized according to whether they were in the top third [circled crosses], middle third [open circles], or bottom third [filled circles] on aggression.)
applicability to the human species can be tested. A classic ex- ample is Bowlby's (1969) use of Harlow's data on rhesus mon- keys (Harlow & Zimmerman, 1959) to show that contact com- fort, and not just food reinforcement as had previously been supposed, was crucial in the mother-child relationship.
Ethologists, unlike developmental psychologists, have em- phasized that full understanding of a structure or behavior de- mands answers to four distinct questions. Thus the question "Why does the thumb move in a different way from the fingers?" could be answered developmentally (the growth of digit rudiments and nerve fibres), causally (the structure of bones, muscles, and nerves), functionally (the thumb's role in grasping, etc.), or in terms of evolution (the human species' monkeylike ancestors presumably had similar thumbs). The importance of the last two questions has been overemphasized by some sociobiologists and neglected by most developmental psychologists. However, such issues can make a not inconsider- able contribution to understanding child development.
First, they have implications for practice. Thus the finding that, across mammals, the frequency of suckling is inversely related to the concentration of the milk and that humans have relatively dilute milk was a strong argument against schedule feeding (Blurton-Jones, 1972). The studies of Klaus and Ken- nell (1976; Kennell, 1986) showing that allowing mothers to have immediate postpartum contact with their infants has at least short-term (but not necessarily long-term [Fleming & Corter, 1988]) beneficial effects were influenced by compara- tive functional considerations. And, at a more theoretical level, Bowlby (1969) cast new light on the so-called "irrational fears of childhood" (fears of darkness, falling, being left alone, etc.) by arguing that they would have been functional in our environ- ment of evolutionary adaptedness.
Second, diverse facts about human behavior, which appear initially to be isolated and independent, can be integrated from an evolutionary perspective. Thus various aspects of the
mother-infant relationship are seen to form a functional whole when seen against the probable sociosexual arrangements in our environment of evolutionary adaptedness (Hinde, 1984).
Third, the links between situations and outcomes can some- times be understood in functional terms. For example, infanti- cide and voluntary abortion are more common when the infant is not the putative parents' own, the infant has poor reproduc- tive potential, or circumstances are adverse and reproductive effort might be wasted. If the incidence of infanticide or abor- tion is taken as an indicator of parental motivation, the data are in harmony with the view that motivation is low when further parental investment in the current offspring might decrease the mother's long-term reproductive success (Daly & Wilson, 1984).
It has even been suggested that behavior that seems maladap- tive in our society may have been functional in others. Thus De Vries (1984) found that children with a "difficult" temperament were more likely to survive famine, perhaps because they were more demanding. Similarly, Main and Weston (1982) suggested that the behavior of infants whose relationships with their mothers were avoidant permitted the maintenance of organiza- tion, control, and flexibility with mothers who do not welcome physical contact and who are restricted in emotional expression (see also Egeland & Farber, 1984). It has also been suggested that the relations between early family relationships and subse- quent personality or behavioral characteristics are adaptive (Belsky, Steinberg, & Draper, 1991; Hinde, 1986, 1991b), al- though the evidence is far from secure.
A fourth possible payoff from an evolutionary-functional approach is that our changing adjustment to our changing cul- ture could be greatly facilitated by an understanding of where we started. This does not imply that there are human character- istics that are independent of culture, but rather that it is helpful to distinguish biological desiderata, resulting from natural se- lection in our environment of evolutionary adaptedness, from
APA CENTENNIAL: INTEGRATING BEHAVIORAL SCIENCE 1027
the desiderata of our particular culture, and to consider the relations of each to the psychological desideratum of mental health (Hinde & Stevenson-Hinde, 1990).
Conclusion
My plea, therefore, is for a truly multidisciplinary approach that focuses on the dialectical relations between levels of social complexity and, most particularly, on those relations among the individual, interactions, and relationships. Such an ap- proach requires not only a descriptive base but also recognition that description can never be precise, and that both descriptive and explanatory concepts are concerned with entities that are interconnected and mutually influence each other.
References
Ainsworth, M. D. S, Blehar, M. C, Waters, E., & Wall, S. (1978). Pat- terns of attachment. Hillsdale, NJ: Erlbaum.
Andrew, R. J. (1991). Testosterone, attention and memory. In P. Bate- son (Ed.), The development and integration of behaviour (pp. 171- 190). Cambridge, England: Cambridge University Press.
Barlow, D. H., & Hersen, M. (1984). Single-case experiential designs. New York: Pergamon Press.
Barlow, G. W (1977). Modal action patterns. In T. A. Sebeok (Ed.), How animals communicate (pp. 98-136). Bloomington: Indiana Univer- sity Press.
Barlow, G. W (1989). Has sociobiology killed ethology or revitalized it? In P. P. G. Bateson & P. H. Klopfer (Eds.), Perspectives in ethology, Vol 8: Whither ethology? (pp. 1-46). New York: Plenum Press.
Baron-Cohen, S. (1991). Do people with autism understand what causes emotion? Child Development, 62, 385-395.
Barrett, J., & Hinde, R. A. (1988). Triadic interactions: Mother-first- born-secondborn. In R. A. Hinde & J. Stevenson-Hinde (Eds.), Re- lationships within families: Mutual influences (pp. 181-192). New \brk: Oxford University Press.
Bateson, P. (1987). Biological approaches to the study of behavioral development. International Journal of Behavioral Development, 10, 1-22.
Bateson, P. (1991 a). Are there principles of behavioral development? In P. Bateson (Ed.), The development and integration of behaviour. Cam- bridge, England: Cambridge University Press.
Bateson, P. (Ed.). (199 lb). The development and integration of behaviour. Cambridge, England: Cambridge University Press.
Baumrind, D. (1971). Current patterns of parental authority. Develop- mental Psychology Monographs, 4 (1, Pt. 2).
Belsky, B., & Pensky, E. (1988). Developmental history, personality and family relationships: Toward an emergent family system. In R. A. Hinde & J. Stevenson-Hinde (Eds.), Relationships within families: Mutual influences (pp. 193-217). New York: Oxford University Press.
Belsky, J., Steinberg, L., & Draper, P. (1991). Childhood experience, interpersonal development and reproductive strategy: An evolution- ary theory of socialization. Child Development, 62, 647-670.
Blurton-Jones, N. G. (1972). (Ed.). Ethological studies of child be- haviour. Cambridge, England: Cambridge University Press.
Bowlby, J. (1969). Attachment and loss. 1. Attachment. London: Ho- garth.
Bretherton, I. (1985). Attachment theory: Retrospect and prospect. In I. Bretherton & E. Waters (Eds.), Growing points of attachment theory and research. Monographs of the Society for Research in Child Development, 50,1-2.
Bretherton, I. (1990). Communication patterns, internal working mod- els and the intergenerational transmission of attachment relation- ships. Infant Mental Health Journal, 11, 237-252.
Butterworth, G., & Bryant, P. (Eds.). (1990). Causes of development. New York: Harvester Wheatsheaf.
Caldwell, B. M. (1969). A new "approach" to behavioral ecology. In I. P. Hill(Ed),MinnesotaSymposiaonChildPsychology, Vol. 2. Minneap- olis: University of Minnesota Press.
Carey, S. (1990). On the relations between the description and the expla- nation of developmental change. In G. Butterworth & P. Bryant (Eds.), Causes of development (pp. 135-160). New York: Harvester Wheatsheaf.
Caspi, A., & Elder, G. H., Jr. (1988). Emergent family patterns: The intergenerational construction of problem behavior and relation- ships. In R. A. Hinde & J. Stevenson-Hinde (Eds.), Relationships within families: Mutual influences (pp. 218-240). New York: Oxford University Press.
Cassidy, X, & Marvin, R. S. (1989). Attachment organization in pre- school children: Coding guidelines. Seattle, WA: MacArthur Working Group on Attachments.
Christensen, A., & Margolin, G. (1988). Conflict and alliance in dis- tressed and non-distressed families. In R. A. Hinde & J. Stevenson- Hinde (Eds.), Relationships within families: Mutual influences (pp. 263-282). New York: Oxford University Press.
Clarke-Stewart, K. A. (1978). And Daddy makes three: The father's impact on mother and young child. Child Development, 49, 466- 478.
Corter, C, Abramovitch, R., & Pepler, D. (1983). The role of the mother in sibling interactions. Child Development, 54,1599-1605.
Craik, K. (1943). The nature of explanation. Cambridge, England: Cambridge University Press.
Crittenden, P. (1990). Internal representational models of attachment relationships. Infant Mental Health Journal, 11, 259-277.
Daly, M., & Wilson, M. (1984). A sociobiological analysis of human infanticide. In G. Hausfater & S. B. Hrdy (Eds.), Infanticide: Compara- tive & evolutionary perspectives. New York: Aldine.
De Vries, M. W (1984). Temperament and infant mortality among the Masai of East Africa. American Journal of Psychiatry, 141, 1189- 1194.
Dunn, J. (1988a). Connections between relationships: Implications of research on mothers and siblings. In R. A. Hinde & J. Stevenson- Hinde (Eds.), Relationships within families: Mutual influences (pp. 168-180). New York: Oxford University Press.
Dunn, J. (1988b). The beginnings of social understanding. Oxford, En- gland: Blackwell.
Dunn, J. (1991). Relationships and behaviour: The significance of Rob- ert Hinde's work for developmental psychology. In P. Bateson (Ed.), The development and integration of behaviour (pp. 375-388). Cam- bridge, England: Cambridge University Press.
Easterbrooks, M. A., & Emde, R. N. (1988). Marital and parent-child relationships: The role of affect in the family system. In R. A. Hinde & J. Stevenson-Hinde (Eds.), Relationships within families: Mutual influences (pp. 83-103). New York: Oxford University Press.
Egeland, B., & Farber, E. (1984). Infant-mother attachment: Factors related to its development and changes over time. Child Develop- ment, 55, 753-771.
Eibl-Eibesfeldt, I. (1975). Ethology. New York: Holt, Rinehart & Win- ston.
Emde, R. (1980). Levels of meaning for infant emotions. In W A. Col- lins (Ed.), Development of cognition, affect and social relations. Hills- dale, NJ: Erlbaum.
Engfer, A. (1988). The interrelatedness of marriage and the mother- child relationship. In R. A. Hinde & J. Stevenson-Hinde (Eds.), Rela-
1028 ROBERT A. HINDE
tionships within families: Mutual influences (pp. 104-118). New York: Oxford University Press.
Fentress, J. C. (1991). Analytical ethology and synthetic neuroscience. In P. Bateson (Ed.), The development and integration of behaviour (pp. 77-120). Cambridge, England: Cambridge University Press.
Feshbach, S. (1970). Aggression. In R H. Mussen (Ed.), Carmichael's manual of child psychology, Vol II. New York: Wiley.
Fleming, A. S., & Corter, C. (1988). Factors influencing maternal re- sponsiveness in humans: Usefulness of an animal model. Psychoneu- roendocrinology, 13,189-212.
Getting, P. A., & Dekin, M. S. (1985). Tritonia swimming: A model system for integration within rhythmic motor systems. In A. I. Sel- verston (Ed.), Model neural networks and behavior (pp. 3-20). New York: Plenum Press.
Goswami, U, & Bryant, P. (1990). Phonological skills and learning to read. Hillsdale, NJ: Erlbaum.
Gottlieb, G. (1991). Experiential canalization of behavioral develop- ment: Theory. Developmental Psychology, 27, 4-13.
Grossman, K., Fremmer-Bombik, E., Rudolph, X, & Grossman, K. E. (1988). Maternal attachment representations as related to patterns of infant-mother attachment and maternal care during the first year. In R. A. Hinde & J. Stevenson-Hinde (Eds.), Relationships within families: Mutual influences (pp. 241 -261). New \brk: Oxford Univer- sity Press.
Harlow, H. E, & Zimmerman, R. R. (1959). Affectionate responses in the infant monkey. Science, 130, 421-432.
Hetherington, E. M. (1988). Parents, children, and siblings: Six years after divorce. In R. A. Hinde & J. Stevenson-Hinde (Eds.), Relation- ships within families: Mutual influences (pp. 311-331). New York: Oxford University Press.
Hinde, R. A. (1966). Animal behavior. New York: McGraw Hill. Hinde, R. A. (1972). Social behavior and its development in subhuman
primates. Eugene, OR: Oregon State System of Higher Education. Hinde, R. A. (1979). Towardsunderstandingrelationships. London: Aca-
demic Press. Hinde, R. A. (Ed.). (1983). Primate social relationships. Oxford, En-
gland: Blackwell. Hinde, R. A. (1984). Biological basesof the mother-child relationship.
In J. D. Call, E. Galenson, & R. L. Tyson (Eds.), Frontiers of infant psychiatry (pp. 284-294). New York: Basic Books.
Hinde, R. A. (1986). Some implications of evolutionary theory and comparative data for the study of human prosocial and aggressive behavior. In D. Olweus, J. Block, & M. Radke-Yarrow(Eds.), Develop- ment of anti-social and prosocial behavior. San Diego, CA: Academic Press.
Hinde, R. A. (1987). Individuals, relationships & culture. Cambridge, England: Cambridge University Press.
Hinde, R. A. (1989a). Continuities and discontinuities: Conceptual is- sues and methodological considerations. In M. Rutter (Ed.), Studies of psychosocial risk (pp. 367-384). Cambridge, England: Cambridge University Press.
Hinde, R. A. (1989b). Reconciling the family systems and the relation- ships approaches to child development. In K. Kreppner & R. M. Lerner (Eds.), Family systems and life span development (pp. 149- 164). Hillsdale, NJ: Erlbaum.
Hinde, R. A. (199 la). A biologist looks at anthropology. Man, 26, 583- 608.
Hinde, R. A. (1991 b). When is an evolutionary approach useful? Child Development, 62, 671-675.
Hinde, R. A. (1992). Auf dem wege zu einer Wissenscheft zwischer- menschlicher Beziehungen [Towards a science of relationships]. In A. E. Aulagen & R. Von Salisch (Eds.), Zwischenmenschliche Bezie- hungen. GOttingen, Germany: Hogrefe.
Hinde, R. A., & Dennis, A. (1986). Categorizing individuals: An alter-
native to linear analysis. InternationalJournal of Behavioral Develop- ment, 9,105-119.
Hinde, R. A., & McGinnis, L. (1977). Some factors influencing the effects of temporary mother-infant separation—some experiments with rhesus monkeys. Psychological Medicine, 7,197-222.
Hinde, R. A., & Stevenson-Hinde, J. (Eds.). (1973). Constraints on learn- ing: Limitations and predispositions. London: Academic Press.
Hinde, R. A., & Stevenson-Hinde, J. (1988a). Epilogue. In R. A. Hinde & J. Stevenson-Hinde (Eds.), Relationships within families: Mutual influences (pp. 365-385). New York: Oxford University Press.
Hinde, R. A., & Stevenson-Hinde, J. (Eds.). (1988b). Relationships within families: Mutual influences. New York: Oxford University Press.
Hinde, R. A., & Stevenson-Hinde, J. (1990). Attachment: Biological, cultural and individual desiderata. Human Development, 33,62-72.
Hopkins, B., & Butterworth, G. (1990). Concepts of causality in expla- nations of development. In G. Butterworth & R Bryant (Eds.), Causes of development (pp. 3-32). New York: Harvester Wheatsheaf.
Horn, G. (1985). Memory, imprinting and the brain. Oxford, England: Clarendon Press.
Horn, G. (1991). Cerebral function and behaviour investigated through a study of filial imprinting. In P. Bateson (Ed.), The development and integration of behaviour (pp. 121-148). Cambridge, England: Cam- bridge University Press.
Hoyle, G. (1964). Exploration of neuronal mechanisms underlying be- havior in insects. In R. F. Reiss (Ed.), Neural theory and modeling. Palo Alto, CA: Stanford University Press.
Hutchison, J. (1991). How does the environment influence the be- havioural action of hormones? In P. Bateson (Ed.), The development and integration of behaviour. Cambridge, England: Cambridge Uni- versity Press.
Jaspers, J. M. F, & Leeuw, J. A. de (1980). Genetic-environment covar- iation in human behaviour genetics. In L. J. T. van der Kamp (Ed.), Psychometrics for educational debate. Chichester, England: Wiley.
Johnson-Laird, P. N. (1983). Mental models. Cambridge, MA: Harvard University Press.
Johnson-Laird, P. N. (1990). The development of reasoning ability. In G. Butterworth & P. Bryant (Eds.), Causes of development. New York: Harvester Wheatsheaf.
Kagan, J. (1971). Change and continuity in infancy. New York: Wiley. Kazdin, A. E. (1982). Single-case research design. New York: Oxford
University Press. Kennell, J. (1986). John Lind Memorial Lecture. World Congress of
Infant Psychiatry, Stockholm, Sweden. Klaus, M. H., & Kennell, J. H. (1976). Maternal infant bonding. St.
Louis, MO: Mosby. Klinnert, M. D, Campos, J. J., Sorce, J. F, Emde, R., & Svedja, M.
(1983). Emotions as behavior regulations: Social referencing in in- fancy. In R. Plutchik & H. Kellerman (Eds.), The emotions, Vol 2. San Diego, CA: Academic Press.
Lytton, H. (1973). Three approaches to the study of parent-child inter- action: Ethological, interview and experiment. Journal of Child Psy- chology and Psychiatry, 14,l-\l.
Maccoby, E. E., & Martin, J. A. (1983). Socialization in the context of the family: Parent-child interaction. In M. Hetherington (Ed.), Mus- sen: Handbook of child psychology (Vol. 4, pp. 1-103). New York: Wiley.
MacCorquodale, K., & Meehl, P. E. (1954). Edward C. Tolman. In W K. Estes, S. Koch, K. MacCorquodale, P. E. Meehl, C. G. Mueller, W N. Schoenfeld, & W S. Verplanck (Eds.), Modem learning theory (pp. 177-266). New York: Appleton-Century-Crofts.
Main, M., Kaplan, N, & Cassidy, J. (1985). Security in infancy, child- hood and adulthood. In I. Bretherton & E. Waters (Eds.), Growing
APA CENTENNIAL: INTEGRATING BEHAVIORAL SCIENCE 1029
points of attachment theory and research. Monographs of the Society for Research in Child Development, J0(Serial No. 1-2), 66-104.
Main, M., & Weston, D. R. (1982). Avoidance of the attachment figure in infancy. InC. M. Parkes& J. Stevenson-Hinde (Eds.), The place of attachment in human behaviour. London: Tavistock.
Markova, I. (1990). Causes and reasons in social development. In G. Butterworth & P. Bryant (Eds.), Causes of development (pp. 186- 214). New York: Harvester Wheatsheaf.
Marks, I. M. (1987). Fears, phobias and rituals. New York: Oxford Uni- versity Press.
Marler, P. (1991). Differences in behavioural development in closely related species: Birdsong. In P. Bateson (Ed.), The development and integration of behaviour (pp. 41-70). Cambridge, England: Cam- bridge University Press.
Mead, G. H. (1934). Mind, self and society. Chicago: University of Chi- cago Press.
Meyer, H.-J. (1988). Marital and mother-child relationships: Develop- mental history, parent personality, and child difficultness. In R. A. Hinde & J. Stevenson-Hinde (Eds.), Relationships within families: Mutual influences (pp. 119-141). New York: Oxford University Press.
Mineka, S. (1987). A primate model of phobic fears. In H. Eysenck & I. Martin (Eds.), Theoretical foundations of behavior therapy. New York: Plenum Press.
Minuchin, P. (1985). Families and individual development: Provoca- tions from the field of family therapy. Child Development, 56, 289- 302.
Minuchin, P. (1988). Relationships within the family: A systems per- spective on development. In R. A. Hinde & J. Stevenson-Hinde (Eds.), Relationships within families: Mutual influences (pp. 7-26). New York: Oxford University Press.
Oyama, S. (1985). The ontogeny of information. Cambridge, England: Cambridge University Press.
Patterson, G. R., & Dishion, T. J. (1988). Multilevel family process models: Traits, interactions and relationships. In R. A. Hinde & J. Stevenson-Hinde (Eds.), Relationships within families: Mutual influ- ences. New York: Oxford University Press.
Plomin, R. (1986). Development, genetics and psychology. Hillsdale, NJ: Erlbaum.
Plomin, R., & de Fries, B. C. (1983). The Colorado adoption project. Child Development, 54, 276-289.
Prechtl, H. F. R. (1950). Das Verhalten von Kleinkindern gegenuber Schlangen [The behavior of children towards snakes]. Weiner Zeitz. f. Philosophic Psychologie und Paedagogie, 2, 68-70.
Radke-Yarrow, M. (1991). The individual and the environment in hu- man behavioural development. In P. Bateson (Ed.), The development and integration of behaviour (pp. 389-410). Cambridge, England: Cambridge University Press.
Radke-Yarrow, M., Richters, J., & Wilson, W E. (1988). Child develop- ment in a network of relationships. In R. A. Hinde & J. Stevenson- Hinde (Eds.), Relationships within families: Mutual influences (pp. 48-67). New York: Oxford University Press.
Rosenblatt, J. (1991). A psychobiological approach to maternal be- haviour among the primates. In P. Bateson (Ed.), The development and integration of behaviour (pp. 191-222). Cambridge, England: Cambridge University Press.
Rutter, M. (1988). Functions and consequences of relationships: Some psychopathological considerations. In R. A. Hinde & J. Stevenson- Hinde (Eds.), Relationships within families: Mutual influences (pp. 332-353). New York: Oxford University Press.
Rutter, M. (1991). A fresh look at "maternal deprivation." In P. Bateson (Ed.), The development and integration of behaviour (pp. 331-374). Cambridge, England: Cambridge University Press.
Sameroff, A. F, & Chandler, M. (1975). Reproductive risk and the continuum of caretaker casualty. In F. D. Horowitz (Ed.), Review of child development research. Vol. 4. Chicago: University of Chicago Press.
Scarr, S., & Kidd, K.. K. (1983). Developmental behaviour genetics. In J. J. Campos & M. M. Haith (Eds.), Mussen, Handbook of child psy- chology (Vol. 2, pp. 345-434). New York: Wiley.
Scarr, S., & McCartney, K. (1983). How people make their own environ- ments. Child Development, 54, 424-435.
Seligman, M. E. P., & Hager, J. L. (Eds.). (1972). Biological boundaries of learning. New York: Appleton Century Crofts.
Simpson, E. E., & Stevenson-Hinde, J. (1985). Temperamental charac- teristics of three- to four-year-old boys and girls and child-family interactions. Journal of Child Psychology and Psychiatry, 26, 43-53.
Sroufe, L. A. (1983). Infant-caregiver attachment and patterns of adap- tation in the preschool: The roots of competence and maladaption. In M. Perlmutter (Ed.), Minnesota Symposia in Child Psychology (Vol. 16, pp. 41-83). Hillsdale, NJ: Erlbaum.
Sroufe, L. A., & Fleeson, J. (1986). Attachment and the construction of relationships. In W Hartup & K. Rubin (Eds.), Relationships and development (pp. 51-72). Hillsdale, NJ: Erlbaum.
Stern, D. (1985). The interpersonal world of the infant. New York: Basic Books.
Stern, D. (1991). Diary of a baby. London: Fontana. Stevenson-Hinde, J. (1986). Towards a more open construct. In D.
Kohnstamm (Ed.), Temperament discussed. Lisse, Netherlands: Swets & Zeitlinger.
Stevenson-Hinde, J. (1988). Individuals in relationships. In R. A. Hinde & J. Stevenson-Hinde (Eds.), Relationships within families: Mutual influences (pp. 68-81). New \brk: Oxford University Press.
Stevenson-Hinde, J. (1991). Temperament and attachment: An eclectic approach. In P Bateson (Ed.), The development and integration of behaviour (pp. 315-330). Cambridge, England: Cambridge Univer- sity Press.
Stevenson-Hinde, J., & Shouldice, A. (1990). Fear and attachment in 2.5 year olds. British Journal of Developmental Psychology, 8, 319- 333.
Thorpe, W H. (1961). Birdsong. Cambridge, England: Cambridge Uni- versity Press.
Turner, P. (1991). Relations between attachment, gender and behavior with peers in school. Child Development, 62,1475-1488.
van Hooff, J. A. R. A. M. (1972). A comparative approach to the phyto- geny of laughter and smiling. In R. A. Hinde (Ed.), Non-verbal com- munication (pp. 209-237). Cambridge, England: Cambridge Univer- sity Press.
Received October 4,1991 Accepted June 16,1992 •