Sunteți pe pagina 1din 25

Childhood Experience, Interpersonal Development, and Reproductive Strategy: An

Evolutionary Theory of Socialization

Author(s): Jay Belsky, Laurence Steinberg and Patricia Draper
Source: Child Development, Vol. 62, No. 4 (Aug., 1991), pp. 647-670
Published by: Wiley on behalf of the Society for Research in Child Development
Stable URL:
Accessed: 08-04-2016 21:17 UTC

Linked references are available on JSTOR for this article:
You may need to log in to JSTOR to access the linked references.

Your use of the JSTOR archive indicates your acceptance of the Terms & Conditions of Use, available at

JSTOR is a not-for-profit service that helps scholars, researchers, and students discover, use, and build upon a wide range of content in a trusted
digital archive. We use information technology and tools to increase productivity and facilitate new forms of scholarship. For more information about
JSTOR, please contact

Wiley, Society for Research in Child Development are collaborating with JSTOR to digitize,
preserve and extend access to Child Development

This content downloaded from on Fri, 08 Apr 2016 21:17:06 UTC
All use subject to
Theoretical Paper

Childhood Experience, Interpersonal

Development, and Reproductive Strategy:
An Evolutionary Theory of Socialization

Jay Belsky
The Pennsylvania State University

Laurence Steinberg
Temple University

Patricia Draper
The Pennsylvania State University

BELSKY, JAY; STEINBERG, LAURENCE; and DRAPER, PATRICIA. Childhood Experience, Interpersonal
Development, and Reproductive Strategy: An Evolutionary Theory of Socialization. CHILD DE-
VELOPMENT, 1991, 62, 647-670. The concept of "reproductive strategy" drawn from the field of
behavioral ecology is applied to the study of childhood experience and interpersonal develop-
ment in order to develop an evolutionary theory of socialization. The theory is presented in
terms of 2 divergent development pathways considered to promote reproductive success in the
contexts in which they have arisen. One is characterized, in childhood, by a stressful rearing
environment and the development of insecure attachments to parents and subsequent behavior
problems; in adolescence by early pubertal development and precocious sexuality; and, in adult-
hood, by unstable pair bonds and limited investment in child rearing, whereas the other is
characterized by the opposite. The relation between this theory and prevailing theories of social-
ization, specifically, attachment, social-learning, and discrete-emotions theory, is considered and
research consistent with our evolutionary theory is reviewed. Finally, directions for future re-
search are discussed.

In 1982 Draper and Harpending offered pair bonding in a manner that makes biologi-
a novel interpretation of the early cross- cal sense. Drawing on concepts from mod-
cultural and subsequent psychological liter- ern evolutionary theory, and particularly
ature on father absence, which suggested Trivers's (1974) parental investment theory,
that boys from families in which parents di- Draper and Harpending (1982) argued that
vorced frequently engage in exaggeratedly early experience "sets" the reproductive
and stereotypically masculine behavior dur- strategy that individuals will follow in later
ing childhood, and that girls from such life. Whereas children growing up in father-
homes tend to be sexually "promiscuous" absent homes stemming from divorce de-
during adolescence (e.g., Biller, 1981; Heth- velop behavior profiles consistent with an
erington, 1972). In particular, these anthro- expectation that paternal investment in child
pologists proposed that early family experi- rearing will not be forthcoming and that pair
ences shape children's orientations toward bonds will not be enduring, those from

Work on this paper was supported by a grant from the National Institute of' Child Health
and Human Development (R01HD15496) and by a NIMH Research Scientist Development
Award to the first author (K02MH00486). The authors would like to thank the following individu-
als for their helpful feedback on an earlier draft of this manuscript, as well as five anonymous
reviewers: Urie Bronfenbrenner, Robert Burgess, Jim Chisholm, E. Mark Cummings, Kathryn
Hood, Kevin MacDonald, Nora Newcombe, Michael Rutter, Steve Suomi, Michelle Surbey, and
Marsha Weinraub. First author's address: College of Health and Human Development, The
Pennsylvania State University, University Park, PA 16802.

[Child Development, 1991, 62, 647-670. ? 1991 by the Society for Research in Child Development, Inc.
All rights reserved. 0009-3920/91/6204-0010$01.00]

This content downloaded from on Fri, 08 Apr 2016 21:17:06 UTC
All use subject to
648 Child Development

father-present households anticipate the op- behavior development we advance draws

posite and thus defer sexual activity once heavily upon concepts basic to the fields of
they reach biological maturity and seek to behavioral ecology (e.g., Krebs & Davies,
establish and maintain enduring, close, het- 1981) and sociobiology (Hamilton, 1964).
erosexual relationships. From sociobiology we take the maxim that
natural selection tends to favor behavior that
Central to Draper and Harpending's
increases fitness, that is, the representation
(1982) application of modern evolutionary
of an individual's genes (relative to unre-
theory to the study of father absence is the
lated individuals) in future generations.
proposition that these distinctly different
From behavioral ecologists we take the
"reproductive strategies" maximize the re-
maxim that behavioral strategies that con-
productive prospects of children in the dis-
tribute to reproductive success are faculta-
tinct environments in which they grow up.
tive, that is, contextually conditional. In
In essence, they theorized that humans have
other words, optimal strategies depend upon
evolved to be sensitive and responsive to the
the options available to individuals given
context of early rearing, and as a conse-
the physical, economic, and social ecology
quence they develop certain behavioral pat-
(Crawford & Anderson, 1989). A given pat-
terns and psychological orientations that
tern of behavior is only optimal in the con-
subsequently guide their reproductive func-
text of a specific environment. From the per-
tioning. Fundamentally, evolution has
spective of modern evolutionary theory,
primed humans to learn particular lessons
complex, highly social organisms such as hu-
during the first 5-7 years of life that will mans evolved to modulate social behaviors
shape their subsequent pair-bonding and
like mating and parenting in response to par-
child-rearing behavior.
ticular environmental cues (Hinde, 1982).
We believe that Draper and Harpend- Thus, the lifespan theory of socialization and
ing's (1982) insightful analysis of the father- interpersonal development we advance does
absence literature lays the seeds of a far not presume biological determinism in any
more general theory of early experience, so- narrow sense. Instead, and consistent with
cialization, and psychological development contemporary behavioral ecology, it as-
than even they realized-one that enables sumes an evolved flexibility of organismic
students of human development to bring response to variations in context, in the ser-
several contemporary theories of psycholog- vice of biological goals (i.e., reproductive
ical and behavioral development into line fitness).
with several currents within modern evolu- Our theory underscores the need to dis-
tionary thinking. Whereas Draper and Har- tinguish among ultimate, distal, and proxi-
pending restricted their focus to father ab- mate causes of behavioral development
sence, we contend that by thinking more (Tinbergen, 1963). A proximate cause is
generally about early affective experience in closely spaced in time to a given phenome-
the family in light of contemporary theories non. A distal cause is further away, yet intri-
of psychological and behavioral develop- cately and causally related to the proximate
ment, it is possible to build a far more gen- mechanism. An ultimate cause concerns the
eral and evolutionary-based theory of social- evolutionary or biological function of a phe-
ization. The primary purpose of this article nomenon, with "why" a process or phenom-
is to articulate this evolutionary theory of so- enon occurs, rather than "how" it occurs. In
cialization and lifespan interpersonal devel- this article we contend that contextual stress,
opment. In so doing, we attempt to integrate a distal cause, shapes parental behavior, a
several diverse areas of inquiry in the field more proximal influence, which affects in-
of developmental psychology, including re- terpersonal and behavioral development
search on contextual stress and its effects on and, thereby, somatic development, and that
parents, parenting behavior and child devel- the ultimate function of these processes is
opment, the antecedents and consequences to enhance (at least in the environment of
of variations in security of attachment, the evolutionary adaptedness) reproductive fit-
genesis of behavior problems, the anteced- ness. Ultimate, distal, and proximate expla-
ents and consequences of variations in pu- nations, then, are by no means mutually ex-
bertal timing, and influences on sexual be- clusive, nor should they be regarded as
havior. providing alternative explanations. Rather,
they are linked together in a system of cau-
Consistent with the arguments of sation.
Draper and Harpending (1982), the theory
of childhood experience and psychological Consistent with Draper and Harpend-

This content downloaded from on Fri, 08 Apr 2016 21:17:06 UTC
All use subject to
Belsky, Steinberg, and Draper 649

ing's (1982) father-absence analysis, we as- describe-one geared toward opportunistic

sume that evolution has designed humans to advantage taking, the other toward mutual
vary their mating and child-rearing behavior commitment and reciprocal benefit-are of-
in accordance with the contextual conditions ten considered in terms of immoral versus
in which they develop, so as to maximize moral or improper versus correct behavior,
their reproductive success. This notion that we regard them as being distinctively, yet
rearing context shapes life history, which is equally, well suited for the particular eco-
itself systematically related to patterns of logical niches in which they develop. While
pair bonding and parenting in a manner de- Western majority culture may value one
signed to maximize the dispersion of an or- more than another, the evolutionary pro-
ganism's genes in subsequent generations, cesses which led to their emergence existed
has a long history in biobehavioral research apart from contemporary moral evaluations.
on many species of animals (Konishi, Emlen, Additionally, although the term "strategy"
Rickets, & Wingfield, 1989; Lack, 1947), implies a conscious plan to many, we pre-
however surprising it may appear to tradi- sume no such awareness and use the term
tionally trained social scientists. There is in the same sense that behavioral ecologists
good reason to believe that humans, too, do when studying the behavior of many spe-
have evolved to modify their reproductive cies of animals.
behavior (i.e., mating and parenting) in the
service of fitness considerations and in ac- In the second major section of this arti-
cle, we review evidence consistent with the
cord with social and ecological variables.
theory we advance. We must note at the out-
Although we will argue that there is set, following Hinde and Stevenson-Hinde
much about contemporary human develop- (1990, p. 67), that "there can be no proof"
ment that can be illuminated by an analysis that the developmental phenomena which
of reproductive strategies, the fact that the our evolutionary theory seeks to explain "are
current ecology of human development con- the products of natural selection operating
trasts markedly with the environment of evo- in our environment of evolutionary adapted-
lutionary adaptation (i.e., the context of early ness. All that can be said is that, in such
humans in which behavior evolved) means cases, a wide range of otherwise apparently
that the predictions our theory generates are independent facts are integrated better by
necessarily constrained. Even though we do the theory of natural selection than by any
not expect that early social context, rearing, other." Finally, in a third section, we outline
and psychological and behavioral orienta- limits of the theory, particularly in terms of
tion will necessarily forecast the number of alternative explanations, and propose two
offspring borne, given widespread access research designs capable of disconfirm-
to both contraception and abortion, we do ing critical predictions derived from the
anticipate that these factors, particularly theory-predictions that are important be-
in combination, predict other markers of re- cause they cannot be derived from related
productive strategy. These markers include theories of socialization.
pubertal timing, sexual activity, and pair
bonding. An Overview of the Theory
This article is divided into three major Evolutionary ecologists theorize that in
sections. In the first we outline the evolu- order for any organism to reproduce, effort
tionary theory of socialization and lifespan must be apportioned among three funda-
interpersonal development and propositions mental tasks-(1) growth and development,
derived from it. Specifically, we distinguish (2) mating, and (3) parenting (including ga-
two prototypic developmental trajectories mete production). Yet species differ dramati-
that lead to two distinct reproductive strate- cally in how they apportion effort across
gies, and we describe the contextual condi- these tasks. Relative to other species, hu-
tions and developmental processes hypothe- mans emphasize growth and development,
sized to give rise to them. It remains unclear as seen in the prolonged period of juvenile
whether the phenomena and processes we dependence and delayed sexual maturation.
detail are equally characteristic of male and And in contrast to most other mammals, we
female development. Certain aspects of the are unusual for the importance we attach to
theory are more strongly supported in stud- pair bonds and for the high levels ofbiparen-
ies of males, while others are more strongly tal care required to rear children to maturity.
supported in studies of females. But despite this generalization about the
species, there is substantial diversity in the
Although the two social orientations we ways in which different populations of hu-

This content downloaded from on Fri, 08 Apr 2016 21:17:06 UTC
All use subject to
650 Child Development

mans manage (1) growth and development, grating such a theory with the more strongly
(2) mating, and (3) parenting. environmental theory we advance is via the
notion of differential susceptibility to envi-
A central tenet of the theory we advance
ronmental experience: Whereas some indi-
is that a principal evolutionary function of
viduals may be genetically predisposed to
early experience-the first 5-7 years of
life-is to induce in the child an understand- respond to contextual stress and insensitive
rearing by maturing early and engaging in
ing of the availability and predictability of
relatively indiscriminate sexual behavior,
resources (broadly defined) in the environ-
others may be genetically predisposed to re-
ment, of the trustworthiness of others, and
spond to sensitive rearing by deferring sex-
of the enduringness of close interpersonal
ual behavior and establishing enduring pair
relationships, all of which will affect how
bonds in adulthood. Thus, while nurture
the developing person apportions reproduc-
may determine the direction development
tive effort. Individuals whose experiences in
takes, nature may determine the likelihood
and around their families of origin lead them
that, and the extent to which, an individual
to perceive others as untrustworthy, rela-
will be influenced by a particular set of envi-
tionships as opportunistic and self-serving, ronmental conditions.
and resources as scarce and/or unpredict-
able will develop behavior patterns that Figure 1 outlines the interrelation of the
function to reduce the age of biological mat- major domains of the theory (context, rear-
uration (within their range of plasticity) (see ing, psychological development, somatic de-
Barkow, 1984, footnote 7, for a similar pre- velopment, reproductive strategy) and con-
diction), accelerate sexual activity, and ori- trasts the two developmental pathways
ent them toward short-term, as opposed to mentioned in the preceding section on re-
long-term, pair bonds. In other words, a dis- productive strategy. It is important to reiter-
proportionate amount of reproductive effort ate that even though these developmental
will be allocated toward the individual's pathways and related reproductive strate-
growth and development and mating rather gies are discussed as distinct types, it may
than toward parenting. Individuals, in con- be best to regard them as prototypes that
trast, whose experiences lead them to per- characterize contrasting ends of a continuum
ceive others as trustworthy, relationships as of ecologically sensitive behavioral develop-
enduring and mutually rewarding, and re- ment rather than as the only two expressions
sources as more or less constantly available of interpersonal development that are pos-
from the same key persons will behave in sible. The choice, then, is between what
ways that inhibit (relative to the first type) Stearns (1982), following Bradshaw (1965),
age of maturation, will defer sexual activ- refers to as discrete versus continuous phe-
ity, and will be motivated to establish-and notypic plasticity. In discrete plasticity there
be skilled in maintaining-enduring pair are few if any intermediate phenotypes, and
bonds, all of which will serve to enhance all individuals show one or the other of only
investment in child rearing. This second a small number of discrete phenotypes that
type of person, then, disproportionately in- are triggered by the environment, whereas
vests reproductive energies in parenting ef- in continuous plasticity the phenotypic re-
fort rather than individual growth and devel- sponse is matched or scaled to the environ-
opment or mating. In essence, we argue that ment (Chisholm, 1988).
early experiences and the psychological and Although the figure and the discussion
biological functioning they induce lead indi- that follows are consistent with a path
viduals to engage in either a "quantity" or a model, whereby context affects rearing and,
"quality" pattern of mating and rearing.
in turn, psychological and behavioral devel-
In arguing that experience shapes de- opment, somatic development, and repro-
velopment, we do not mean that all or even ductive behavior, it may not be best to
most of the variance on human behavior is conceptualize the theory in terms of such
contextually controlled. Behavior-genetic a linear flow. An alternative, though not
studies strongly indicate that this is by no necessarily mutually exclusive, approach in-
means the case (Plomin, Loehlin, & De volves more of a cumulative-conditional-
Fries, 1985). In fact, an obvious alternative probability conceptualization. This alter-
to the theory we advance is one of genetic native, which we favor, states that when
polymorphism which asserts that individu- antecedent conditions A, B, and C obtain,
als may be genetically predisposed to one the probability of D is greater than when
developmental pathway or another (see only two of these antecedent conditions ob-
Rushton, 1985). One possible way of inte- tain, and that the probability of D is even

This content downloaded from on Fri, 08 Apr 2016 21:17:06 UTC
All use subject to
Belsky, Steinberg, and Draper 651

Marital discord Spousal harmony

High stress A. FAMILY CONTEXT Adequate $ resources
Inadequate $ resources

Harsh, rejecting B. CHILDREARING Sensitive, supportive,

insensitive Infancy / Early Childhood responsive
Inconsistent Positively affectionate

Insecure attachment Secure attachment

Mistrustful internal C. PSYCHOLOGICAL / Trusting internal working
working model BEHAVIORAL model
Opportunistic interpersonal DEVELOPMENT Reciprocally-rewarding
orientation interpersonal orientation

Aggressive Anxious
Noncompliant Depressed

Early maturation / puberty D. SOMATIC Later maturation / puberty


Earlier sexual activity Later sexual activity

Short-term, unstable E. REPRODUCTIVE Long-term, enduring
pair bonds STRATEGY pair bonds
Limited parental investment Greater parental investment

FIG. 1.--Developmental pathways of divergent reproductive strategies

less when just a single such condition ob- success, and especially the chances of estab-
tains. Thus, whereas a path-oriented theory lishing an enduring pair bond in which a
predicts that an effect will obtain only when man and a woman reciprocally exchange re-
the immediately preceding influential con- sources for purposes of facilitating the rear-
dition exists, a conditional theory presumes ing of their progeny (Draper & Belsky,
multiple paths to an outcome and greater 1990). Thus, in our view and that of others
and lesser probabilities of an outcome ensu- (e.g., Burgess & Draper, 1989), contextual
ing given varying antecedent conditions. Al- stressors, among them, marital discord, sin-
though we are not inclined to make such gle parenthood, and unstable employment,
specifications at this time, variations on a will foster more insensitive, harsh, rejecting,
conditional theory might differentially inconsistent, and/or unpredictable parent-
weight the importance of proximal versus ing behavior. This, in turn, will induce in
distal antecedent conditions in forecasting the child what students of attachment theory
certain behavioral developments. refer to as an insecure or mistrustful internal
working model (particularly of the insecure-
Theoretical Propositions and Predictions
avoidant variety) of the self, others, and rela-
The first feature of Figure 1 to be con-
sidered is the relation between context and tionships (Ainsworth, Blehar, Waters, &
child rearing. Consistent with traditional
Wall, 1978; Bowlby, 1969; Sroufe, 1979).
That is, when contextual stress and corre-
cultural anthropology (Whiting & Whiting,
lated rearing patterns are extant, children
1975), we presume that patterns of child
will be relatively more likely to come to
rearing reflect and are derivative of the gen- view themselves as unlovable and as unwor-
eral ecology in which families reside and
thy of love; others as undependable and
that, implicitly (if not explicitly), rearing
uncaring; and, as a consequence, close inter-
strategies represent attempts by parents to
personal bonds as ephemeral and unde-
prepare their children for the world that they
"expect" their offspring to encounter. Where
we differ from the traditional view is in theo- In making these observations, particu-
rizing that a critical feature of this (often un- larly about the influence of the ecology of
conscious) interpretation of the future in- rearing, it is important to point out that sub-
volves expectations regarding reproductive jective perceptions of stress are just as im-

This content downloaded from on Fri, 08 Apr 2016 21:17:06 UTC
All use subject to
652 Child Development

portant as, if not more important than, objec- For many social scientists today, such
tive conditions of stress. Because parents' psychological and behavioral patterns are re-
child-rearing behavior serves a mediational garded as problems because they appear
function between context and development, dysfunctional within mainstream society.
a parent's subjective experience is critically We speculate, however, that these behavior
important for understanding the context- patterns are functional when viewed from
rearing linkage. This is not to say that only the standpoint of evolutionary biology (see
parental behavior mediates the influence of Chisholm, 1988). This is because they may
context. It is just as reasonable to presume, serve to mediate and, indeed, contribute to
especially after infancy, that the child's di- the influence that early experience exerts on
rect experience with the context-via watch- somatic development. In particular, we the-
ing parents argue or being witness to a sin- orize that externalizing and internalizing be-
gle parent's relationships with a series of haviors may have evolved as proximal biobe-
partners, for example-will result in the havioral mechanisms that accelerate the
child developing particular understandings timing of puberty-within the individual's
of what the future will hold with respect to range of plasticity. As we will make clear in
relationships. a moment, such early maturation is consid-
ered to be an important component of an op-
In contrast to the "scenario" of insensi- portunistic reproductive strategy, as it not
tive rearing and insecurity just depicted, the only affects postpubertal sexual behavior but
child growing up in an environment in also makes earlier procreation possible.
which resources are (or at least are per- Sex differences in externalizing and
ceived to be) relatively abundant and/or
internalizing behaviors are commonly re-
predictable-including care and nurtur- ported and typically attributed to cultural
ance-develops a distinctly different work-
conditioning (e.g., Campbell, in press; Rut-
ing model of the interpersonal world. When
ter, 1970). Our theory presumes, in addition,
parents trust and count on others and rear an evolved basis of the difference in how
their children to engage in reciprocally re-
so-called behavior problems typically are
warding and enduring interpersonal bonds
manifested by boys and by girls, especially
and to expect paternal investment, children
prior to puberty. In particular, we theorize
develop predispositions to selectively at- that males and females are predisposed to
tend to, encode, and anticipate experiences
express their "behavior problems" in differ-
that are strikingly different from those of
ent ways because different biological pro-
children who experience insensitive, re-
cesses must be energized in order to influ-
jecting, or neglectful care. Whereas rejected
ence biological maturation in the two
or neglected children will home in on
slights, offenses, and mistreatments-
because they have learned to be vigilant In the case of females, we suspect that
about being taken advantage of-secure "internalizing" problems serve to lower me-
children focus on empathic and considerate tabolism, store fat, and thereby stimulate
social encounters and, in so doing, foster menarche. Such experientially and behav-
more of them. For all children, then, subse- iorally induced maturation sets the stage for
quent experience builds on past experience sexual activity and, thereby, procreation.
and, importantly, expectations and under- Early maturation, then, is part of a constella-
standings set earlier in life are reinforced via tion of biological and behavioral processes
self-constructing interpersonal processes. that facilitate a quantity-oriented reproduc-
tive strategy. So too is accelerated sexual ac-
This linkage between context, rearing, tivity with multiple mates. We thus predict
and psychological orientation to the inter- that early maturing girls, particularly those
personal world fosters, as we have implied, who come from stressed families and have
correlated patterns of interpersonal behavior developed insecure outlooks on life in gen-
during childhood. Under conditions of fam- eral (and on close relationships in particular)
ily stress, insensitive rearing, and insecure and who have developed internalizing
feelings, behavior problems are more likely symptoms before puberty, will be more sex-
to develop-high levels of aggression, im- ually active and have less enduring hetero-
pulsivity, and/or noncompliance with adults sexual relationships across their life course.
and socialization norms (externalizing symp-
toms) or high levels of sadness, depression, Early sexual activity among females
and/or social withdrawal (internalizing may foster reproduction in still other ways.
symptoms). In the human female there is an extended

This content downloaded from on Fri, 08 Apr 2016 21:17:06 UTC
All use subject to
Belsky, Steinberg, and Draper 653

period between menarche and actual fertil- in their life course, will bear fewer progeny,
ity (Lancaster, 1986; Short, 1976). Moreover, and both mother and father will invest time
there is some suggestive evidence that both and resources more heavily in them.
sexual behavior and associating with nonre-
lated males shortens and regulates the men- To many readers it will no doubt seem
strual cycle (e.g., Cutler, Garcia, & Krieger, counterintuitive to assert, as we and others
1979; Trevathan & Burleson, 1989). A do (MacDonald, 1988), that a developmental
shorter and more regular cycle affords, over trajectory characterized by attachment inse-
a lifetime, more fertile periods and, thus, curity, behavior problems, early maturation,
more opportunity to conceive children (Me- precocious sexuality, unstable pair bonds,
tral, 1981; Trevathan & Burleson, 1989). limited parental investment, and high total
Thus, we speculate that early puberty and fertility makes "biological sense." In mod-
accelerated sexual activity might actually ern Western society, this reproductive strat-
function to shorten the period of postpuber- egy appears clearly dysfunctional and disad-
tal subfertility and, in so doing, organize the vantageous. Why, then, would it have
menstrual cycle to permit more conceptions evolved? To answer this question it is im-
across the life span. portant to keep in mind that past environ-
ments of evolutionary adaptation have been
The prevalence of externalizing symp-
highly variable. In some contexts, such as
tomology among males and its association
those affording rapid expansion into favor-
with family conflict and divorce (see below
able and previously unexploited niches,
for a review of literature) lead to the sugges-
there may have been advantage to rapid re-
tion that it is aggressive, disobedient, and
production and no penalty (in terms of off-
oppositional behavior that, via some unspec-
spring survival) to transient bonds between
ified biological mechanism, stimulates ear-
sexual partners and loose attachment be-
lier (than would otherwise be expected) mat-
tween mothers and offspring-provided
uration among boys. Conceivably, processes
only that other group members could be re-
involving androgenic activity might be in-
lied upon to provide food and protection to
volved. In any event, aggressive, noncompli-
juveniles. This scenario assumes an environ-
ant behavior, which seems to be inherently
opportunistic vis-a-vis others, should also mentally favorable context (Draper & Har-
pending, 1982).
foster indiscriminate and opportunistic sexu-
ality and, in concert with earlier puberty, in-
crease the likelihood of such males becom-
What evolutionary mechanism, though,
would underlie an accelerated reproductive
ing fathers before other men. It seems likely
schedule in the context of unfavorable envi-
that earlier maturing, opportunistic males
ronments such as poverty or social and fa-
would be the least likely to use contracep-
milial instability, as we propose? We point
tion. Moreover, if such males come from
again to the maxim that organisms have been
maritally discordant or single-parent homes,
selected to reproduce themselves and to at-
it seems unlikely that they will anticipate
tend to important environmental cues in the
and seek enduring pair bonds and thus antic-
process. In the absence of indications that
ipate investing time, energy, and resources
delayed maturation and reproduction can
in their offspring (i.e., they will evidence
have benefits, early sexual activity and high
low paternal investment).
fertility have much to recommend them.
As can be seen in Figure 1, a dramati- This strategy may be associated with higher
cally contrasting pattern of development and offspring mortality, but from the point of
behavior is anticipated for children whose view of fitness, individuals living in such ad-
developmental experiences have fostered verse circumstances who delay reproducing
a secure rather than insecure outlook on may well be selected against (i.e., leave few
life and relationships. These children, we or no offspring). In such an environment, a
theorize, will be especially skilled at estab- man who invests disproportionately in one
lishing and maintaining close friendships woman and in children (who may not be his
during middle childhood (rather than own) will leave relatively few of his own off-
developing behavior problems), will not be spring behind. Likewise, a young woman
early maturers, will delay sexual intercourse who waits for the right man to help rear her
(rather than engage in indiscriminate mat- children may lose valuable reproductive op-
ing), and will possess the interpersonal skills portunities at a time when her health and
and the motivation to establish and maintain physical capability are at their peak and
enduring, heterosexual pair bonds. As par- when her mother and senior female kin are
ents they will bear offspring at a later point young enough to be effective surrogates. In

This content downloaded from on Fri, 08 Apr 2016 21:17:06 UTC
All use subject to
654 Child Development

such circumstances, nonbonded and rela- opmental research on the determinants and
tively indiscriminate sexuality, as well as consequences of variations in socialization
high fertility, can be positively selected. It practices.
is in this sense, then, that we assert that both
In a second subsection, we turn atten-
of the reproductive strategies that we de-
tion to the most novel predictions made from
tailed make "biological sense," in that they
the theory, namely, that pubertal timing is
are optimal given the contexts in which they
sensitive to rearing context, rearing styles,
develop-and for which they were selected.
and/or antecedent psychological and behav-
ioral functioning (i.e., linkages of A, B, or C
There is much about the evolutionary with D in Fig. 1). In a third subsection, we
theory we have outlined that is consistent review research linking pubertal timing
with other theories of socialization and psy- with postpubertal sexual activity (i.e., D and
chological development, especially social E in Fig. 1). Then, in a final section, we
learning (e.g., Bandura, 1977), attachment briefly summarize research emanating from
(Bowlby, 1969), and discrete emotions the- sociology and psychology pertaining to the
ory (Malatesta & Wilson, 1988). Indeed, the determinants of pair bonding. In particular,
general proposition that experience shapes we consider evidence that rearing experi-
psychological orientation and behavioral ences in childhood, such as exposure to di-
functioning, which feeds back to affect ex- vorce, can influence pair bonding, and we
perience and maintain earlier-established reinterpret research on the adult personality
developmental trajectories, is embraced by correlates of successful and unsuccessful
virtually all modern theories of, and marriage in terms, respectively, of mutually
perspectives on, human development (e.g., beneficial and opportunistic orientations to-
Bronfenbrenner, 1977; Elder, 1981; Epstein ward relationships. With regard to Figure 1,
& Erskine, 1983; Wachtel, 1973). What is this is research pertaining to linkages be-
unique about our theory, however, is that it tween B and E and between C and E.
integrates rather diverse developmental
phenomena-including contextual stress, Family Context, Rearing Patterns,
rearing patterns, attachment styles, behavior and Development
problems, pubertal timing, sexual activity, Basic to our theory are several proposi-
and pair bonding processes-in a manner tions linking family context, patterns of child
that extends, rather than violates, these other rearing, and children's psychological and
perspectives. Most important, though, our behavioral development. In particular, the
theory places all of these phenomena in evo- theory asserts, consistent with much current
lutionary perspective, using the notion of re- thinking regarding the determinants and
productive strategy as a guiding principle, consequences of socialization practices (Bel-
and generates a critical prediction unique to sky, 1984; Elder, Nguyen, & Caspi, 1985;
it-namely, that developmental experiences McLoyd, 1990; Patterson, 1986), that under
and behavior patterns derivative of them conditions of real or imagined stress parental
serve to regulate the timing of puberty and, behavior becomes less affectively positive
thereby, sexual behavior and pair bonding. and more insensitive, perhaps to the point
of psychological or physical abuse, and that
Research Evidence exposure to such rearing practices under-
mines the psychological well-being of the
There is no shortage of research consis- child. In contrast, a positive, prosocial inter-
tent with many of the propositions advanced personal disposition is presumed to derive
above, and in this section we seek to summa- from warm, sensitive, and what have been
rize it. Our goal is to underscore the wealth labeled "authoritative" rearing practices
of data that are consistent with our perspec-
(Baumrind, 1971), and these are further pre-
tive, while acknowledging evidence that is sumed to be promoted by social supports
not. We begin by examining research on the and economic resources (e.g., Lempers,
contextual and rearing antecedents of pat- Clark-Lempers, & Simons, 1989). In es-
terns of behavior that might be conceptual- sence, when resources become limited, pa-
ized as "opportunistic." Accordingly, the rental patience for and tolerance of young
first subsection addresses the interrelation
children's behavior declines, and, as a re-
of context, rearing, and psychological/be- sult, adults become more prone to anger and
havioral development, that is, linkages be- inconsistent disciplinary practices (Belsky,
tween sections A and B and between B and
C in Figure 1. In many respects, this first
subsection reviews rather traditional devel- Context and rearing.-Consideration of

This content downloaded from on Fri, 08 Apr 2016 21:17:06 UTC
All use subject to
Belsky, Steinberg, and Draper 655

research on child abuse (Belsky, 1980; Cic- tal sphere to influence parental behavior, it
chetti & Carlson, 1989), economic depriva- must affect the psychological functioning of
tion (Burgess & Draper, 1989; McLoyd, the individual adult in the parental role. Be-
1990), occupational stress (Bronfenbrenner cause stress is a subjective experience, we
& Crouter, 1982), marital discord (Belsky, would expect to find evidence that psycho-
1981; Emery, 1988), and psychological dis- logical distress relates directly to parental
tress (McLoyd, 1990) reveals a consistent re- functioning. Much recent research indicates
lation between contextual stress and "dys- that anxiety, hostility, and depression are as-
functional" parenting. Child maltreatment is sociated with the very same patterns of pa-
more likely to occur under conditions of eco- rental behavior that have been associated
nomic deprivation, an association that can- with income loss, under- and unemploy-
not be attributed simply to increased abuse ment, poverty, and marital discord (e.g.,
reporting in lower-class communities (Pel- Conger, McCarty, Yang, Lahey, & Kropp,
ton, 1978; Steinberg, Catalano, & Dooley, 1984; Kochanska, Kuczynski, Radke-Yarrow,
1981). Elder's research on the Great Depres- & Welsh, 1987).
sion (Elder et al., 1985) and recent work on
contemporary families that suffered serious The fact that contextual stress adversely
income loss as a result of recession in the influences adult psychological well-being
1980s (Lempers et al., 1989) show, consis- and thereby parental functioning provides
tent with McLoyd's (1990) review of the lit- the basis of our assertion that parents, via
erature on parenting practices among impov- their rearing, are "preparing" their children
erished parents, that limited financial for the world they expect them to encounter
resources are associated with the use of as adults. Whereas affectively supportive
power-assertive child-rearing techniques in behavior and harmonious parent-child ex-
disciplinary encounters. "Lower class par- changes convey-inadvertently or inten-
ents are more likely to issue commands tionally-a life-course message that the
without explanation, less likely to consult world is a caring place and that relationships
the child about his or her wishes, and less can be counted upon, rearing that is af-
likely to verbally reward the child for behav- fectively negative and fosters conflict and
ing in desirable ways. Poverty also has been coercion conveys to the child just the op-
associated with diminished expression of
affection and less responsiveness to the so- Rearing and development.-The analy-
cioemotional needs explicitly expressed by sis just offered presumes that children de-
the child" (McLoyd, 1990, p. 322). In light velop psychological and behavioral orien-
of these findings, it is noteworthy that when tations consistent with the life-course
rhesus monkeys are experimentally de- experiences for which parents "prepare"
prived of resources and thus must spend them. Some of the most compelling-yet
more time and energy foraging for food, least surprising--evidence that insensitive,
they, too, become less solicitous of their off- rejecting, and/or inconsistent rearing leads
spring's needs (Rosenblum & Paully, 1984). children to view the world as uncaring and
It is not just limited income that can relationships as untrustworthy and to be-
compromise parental functioning, but also have in opportunistic ways comes from the
certain characteristics of the job (Crouter, study of the development of abused and ne-
in press) as well as marital processes (Bel- glected children (for a review, see Young-
sky, 1981). Work conditions that generate blade & Belsky, 1989, in press). Not only are
feelings of distress, including lack of job infants and toddlers subjected to mal-
protection, dirty work, excessively close treatment at very high risk of developing in-
supervision, overload, boredom, and secure attachments to their maltreating par-
underutilization, have all been linked to ent (e.g., Cicchetti & Braunwald, 1984;
more irritable or less involved patterns of Lyons-Ruth, Connell, Zoll, & Stahl, 1987),
parenting (e.g., Bronfenbrenner & Crouter, but in their relations with age-mates both
1982; Moen, 1982). Marital conflict and low during the preschool and school-age years
levels of marital satisfaction are associated they are more aggressive, less cooperative,
with less responsive, more intrusive, and less empathic, and less trusting than other
more affectively negative patterns of moth- children (e.g., Herrenkohl & Herrenkohl,
ering and fathering (Jouriles, Pfiffner, & 1981; Howes & Eldredge, 1985).
O'Leary, 1988; Olweus, 1980).
While the role of the child in precipi-
In order for contextual stress originating tating child maltreatment cannot be dis-
in the economic, occupational, and/or mari- counted, it is noteworthy that research on

This content downloaded from on Fri, 08 Apr 2016 21:17:06 UTC
All use subject to
656 Child Development

the antecedents and sequelae of patterns of been characterized by shared control, sensi-
infant-mother attachment in nonmaltreating tivity, and responsiveness (Holden & West,
populations is generally consistent with evi- 1989; Kuczynski, Kochanska, Radke-Yarrow,
dence just summarized. American and Euro- & Girniss-Brown, 1987; Parpal & Maccoby,
pean research indicates that insensitive, in- 1985). In contrast, disobedience, noncompli-
trusive, and/or unresponsive care during the ance, and aggression appear to be fostered
first year of life is associated with insecure by coercive, negatively demanding, and
attachment relationships (Ainsworth et al., physically punishing patterns of parenting
1978; Belsky, Rovine, & Taylor, 1984; Gross- (Martin, 1981; Patterson, 1986). This appears
mann, Grossmann, Spangler, Suess, & Unz- to be true not only in contempor-
ner, 1985). Infants and toddlers insecurely ary Western society but around the world
attached to their parents are less able to tol- (Rohner, 1975). Thus, meeting children's so-
erate frustration as 2-year-olds (Matas, Ar- cial and emotional needs in a supportive, re-
end, & Sroufe, 1978), are more prone to so- sponsive manner fosters a social orientation
cial withdrawal in the preschool peer group that values mutually beneficial interactions
(Waters, Wippman, & Sroufe, 1979), evince and relationships, whereas patterns of rear-
less sympathy for the distress of preschool ing that are negative, inconsiderate, and co-
age-mates (Waters, Wippman, & Sroufe, ercive lead children to behave in ways that
1979), are less ready to interact with are self-centered. As our earlier analysis in-
friendly, unfamiliar adults as are 1-year-olds dicated, these latter outcomes are more
(Main & Weston, 1981) and 3-year-olds (Lilt- likely to arise under conditions of contextual
kenhaus, Grossmann, & Grossmann, 1985), stress, whereas the former processes are
are responded to less positively by other more likely to characterize well-resourced
preschoolers with secure attachment histor- families.
ies (Jacobson & Wille, 1986), are liked less
by preschool classmates (LaFreniere & Childhood Experience, Psychosocial
Sroufe, 1985), and during the early elemen- Development, and Pubertal Timing
tary school years are more likely to be A distinguishing proposition of our the-
judged by parents and/or teachers to have ory concerns the experiential, psychologi-
serious behavior problems, including ag- cal, and behavioral antecedents of pubertal
gression and noncompliance (Erickson, timing. We have proposed that individu-
Sroufe, & Egeland, 1985; Lewis, Feiring, als whose early family experiences are high
McGuffog, & Jaskir, 1984; Renken, Egeland, in stress and who evince problematic behav-
Marvinney, Mangelsdorf, & Sroufe, 1989). ior during childhood should be more likely
Research on the developmental corre- to undergo pubertal maturation earlier than
lates of patterns of child rearing that is not children whose childhood experiences are
focused on attachment security generates more pacific. Because the age of pubertal on-
findings that are strikingly consistent with set is a significant predictor of the age of
the research just summarized (for a review of onset of sexual activity (see below), early
the last decade's research, see Belsky, 1990). maturers have additional reproductive
Consider, for example, evidence from two prospects and may be seen as potentially
recent longitudinal studies, one indicating more reproductively opportunistic. Differ-
that high levels of observed maternal posi- ential pubertal timing therefore provides a
tive involvement (affectionate contact and critical somatic link between early social ex-
verbal stimulation) during the first 2 years of perience and later reproductive behavior.
life predicted low levels of mother-reported More important, because traditional social-
behavior problems at age 4, even after con- ization theories do not yield obvious predic-
trolling for observed maternal behavior at tions about somatic outcomes of variation in
age 4 (Pettit & Bates, 1989), and the other early social experience and behavioral de-
showing that high levels of maternal respon- velopment, findings linking family relations
sivity and acceptance during the infant and or children's psychosocial functioning with
toddler years forecast high levels of consid- later physical development distinguish our
erateness at age 10, again after controlling theory from its less biologically oriented
for the same dimensions of parental behav- predecessors.
ior measured at age 10 (Bradley, Caldwell, Because the timing of puberty is
& Rock, 1988).
strongly influenced by genetic (Plomin &
In sum, there is clear evidence that co- Fulker, 1987) and nutritional factors (Mar-
operation and compliance develop among shall, 1978), developmentalists have paid lit-
children whose parental relationship has tle attention to the possibility that early so-

This content downloaded from on Fri, 08 Apr 2016 21:17:06 UTC
All use subject to
Belsky, Steinberg, and Draper 657

cial ecology and behavioral development leads to earlier pubertal maturation. Using
may influence pubertal maturation. Typi- data from an ongoing, prospective study of
cally, timing of puberty is treated as an inde- New Zealand girls followed from birth, Mof-
pendent variable, and differences in the per- fitt, Caspi, and Belsky (1990) did not find
sonality and social development of early and that the presence of either internalizing or
late maturers are viewed as "outcomes" of externalizing problems at age 7 was pre-
maturational processes (see Brooks-Gunn & dictive of earlier maturation, however. More
Reiter, in press). The fact that much of this research on this question is clearly needed
research is cross-sectional or retrospective in before we can fully evaluate this aspect of
design or, where longitudinal, disregards the theory.
the possibility that the "outcomes" of dif-
ferential pubertal timing may derive from Social antecedents of pubertal tim-
prepubertal experiences encourages us to ing.-However unusual it may seem to stu-
reconsider extant interpretations of the evi- dents of human development that timing of
dence. Following this, we consider research puberty is affected by prior and contempora-
on the social antecedents of pubertal timing. neous social conditions, this notion is widely
accepted among scientists who study repro-
Personality characteristics of early and ductive development among nonhuman pri-
late maturers.-We would anticipate that in mates and other mammals. In this section,
adolescence early maturers would be more we examine several lines of evidence de-
likely than late maturers to evince aggres- rived from both human and animal research
sion, extraversion, and psychological and be- that support the hypothesis that the timing
havioral "dysfunction," including both in- of maturation is contextually plastic, socially
ternalizing problems such as depression and mediated, and accelerated by intrafamilial
anxiety (especially among girls) and exter- stress.
nalizing problems such as impulsivity and
conduct disorders (especially among boys). The plasticity of human reproductive
This is in fact the case. Among boys, early functioning in response to social stimuli is
maturers are more popular (consistent with well established. For example, studies have
the expected higher level of extraversion) documented the synchronization of men-
(Petersen, 1985); more likely to be involved strual cycling among dormitory mates
in problem behavior (Duncan, Ritter, Dorn- (McClintock, 1980), that sexual activity may
busch, Gross, & Carlsmith, 1985); and, de- induce or quicken ovulation (Trevathan &
spite their external appearance of adult ma- Burleson, 1989), and that cross-sex contact
turity, more likely to experience "frequent induces menstrual regularity and shortens
temper tantrums" (Livson & Peskin, 1980, p. the menstrual cycle (Trevathan & Burleson,
73). Among girls, early maturers are more 1989). Although these studies do not exam-
popular (Simmons, Blyth, & McKinney, ine pubertal timing specifically, they pro-
1983), yet have more self-image and emo- vide clear evidence of the link between
tional difficulties (including anxiety and de- social experience and human endocrinolog-
pression) (Aro & Taipale, 1987; Simmons & ical functioning. Rather than viewing the
Blyth, 1987), and are more involved in prob- hormone-behavior relation as unidirec-
lem behavior (Aro & Taipale, 1987; Magnus- tional, recent developments in socioendocri-
son, Statin, & Allen, 1986). nology indicate that the link between social
experience and hormonal activity is recipro-
Because longitudinal studies indicate cal. As Worthman (in press) notes, there are
that many of these behavioral patterns are reasons to expect that this reciprocal relation
stable between childhood and adolescence
should be especially potent in the hormonal
(Caspi & Bem, 1990; Digman, 1989), it is regulation of sexual behavior and reproduc-
quite possible that observed personality and tion, including the regulation of pubertal
behavioral differences between early and maturation.
late maturers antedate and affect puberty,
rather than the reverse. Needless to say, an Most of the direct evidence of social in-
adequate test of this proposition requires the fluences on pubertal maturation comes from
assessment of preadolescent psychological research on family relations in nonhuman
functioning, pubertal timing, and behavior primates. Among most nonhuman primates
during adolescence. Although there are nu- living in the wild, pubertal maturation is as-
merous data sets in which this question can sociated with increased distance in the
be examined empirically, we know of only parent-child relationship, either in terms of
one study that directly tested the hypothesis increased physical distance or heightened
that behavioral dysfunction in childhood aggression and, ultimately, forced migration

This content downloaded from on Fri, 08 Apr 2016 21:17:06 UTC
All use subject to
658 Child Development

(Caine, 1986; Steinberg, 1989). When volun- with their mother) matured faster physically
tary or forced migration is prevented among over the next 12 months than did age-mates
animals in captivity, however, reproductive who began the year at an equivalent stage of
maturation is inhibited, especially in the puberty but reported initially closer family
case of females. This effect has been docu- ties.
mented in a variety of mammals, including
several species of monkeys as well as ham- The constrained time period of the
sters, wolves, and wild dogs (Evans & Steinberg study leaves open the possibility
Hodges, 1984; Levin & Johnston, 1986; Tar- that there were pre-assessment differences
diff, 1984). Similarly, the sexual develop- in youngsters' rate of maturation (although
ment of males housed in captivity in large not in their actual pubertal status at the ini-
groups is inhibited by the presence of domi- tial assessment) that caused the observed
nant males and stimulated (or disinhibited) differences in parent-child relations. For
by the removal of more dominant members this reason, the aforementioned study of
from the groups (Goy, 1986, personal com- Moffitt et al. (1990) is especially interesting.
munication). When separation of juvenile In their sample of New Zealand girls, family
from parent occurs after an inhibitory pe- conflict at age 7 was significantly, albeit
riod, however, reproductive maturity is rap- modestly, predictive of earlier menarche,
idly attained (Evans & Hodges, 1984; Tar- even after the effects of parental divorce
diff, 1984). Taken together, these studies (which may also accelerate puberty-see be-
suggest that reproductive maturation spe- low) and weight (which is known to be asso-
cifically may be inhibited by physical close- ciated with earlier maturation; Frisch &
ness to parents and accelerated by distance McArthur, 1974) were taken into account in
from them. a path analysis.

Although fewer studies of pubertal tim- Further evidence that preadolescent

ing and family relations exist on humans, the contextual stress is associated with acceler-
available evidence parallels findings from ated maturation among girls is found in re-
the animal research. Several studies, em- search on the developmental consequences
ploying a diverse set of methodologies, have of father absence. Jones, Leeton, McLeod,
shown that conflict and distance between and Wood (1972), Moffitt et al. (1990), and
parents and children increases at puberty; Surbey (1990) each report that girls reared
this effect is more characteristic of females in father-absent households attain menarche
than males, and is more commonly observed at an earlier age than their counterparts in
in mother-child than father-child relations, intact homes; moreover, early father absence
although findings in this general direction (Jones et al., 1972; Moffitt et al., 1990) and
have been reported across studies of all four years of father absence (Surbey, 1990) are
parent-child dyads (e.g., Hill, Holmbeck, especially predictive of early puberty in
Marlow, Green, & Lynch, 1985a, 1985b; girls. Surbey also reports that higher levels
Papini & Sabby, 1987; Steinberg, 1987, of stress are associated with earlier menar-
1988; Susman et al., 1987). Sociobiological che in families in which fathers were pres-
explanations of this phenomenon make good ent. This finding, along with those of Moffitt
sense, since postpubertal distance in the et al. (1990) concerning family conflict, sug-
parent-child relationship would minimize gests that it is not the absence of a father that
inbreeding and increase reproductive fitness speeds pubertal maturation, but the stress
(Steinberg, 1989). associated with divorce and with growing up
in a single-parent household (McLanahan &
Of course, correlational studies of pu- Booth, 1989; McLoyd, 1990).
berty and parent-child distance do not in-
dicate the causal direction of the relation Thus, despite the powerful role that ge-
between these two variables. Especially netic factors play in influencing individual
interesting, therefore, are two studies of hu- differences in pubertal and sexual matura-
man adolescents that suggest that parent- tion, a number of studies suggest that girls
child distance may precede pubertal onset, who have been exposed to stressful condi-
at least in homes of girls, and that female tions in the family both before and around
adolescents who have more strained family the time of puberty may attain menarche ear-
relations mature earlier than their peers lier and mature at a faster rate. We suggest
whose family relations are closer. In the first that early stress may affect pubertal timing
of these investigations, Steinberg (1988) by making the individual more biologically
found that girls who reported more strained reactive to social conditions as she ap-
relations with their parents (and especially proaches the age of pubertal onset. Whether

This content downloaded from on Fri, 08 Apr 2016 21:17:06 UTC
All use subject to
Belsky, Steinberg, and Draper 659

similar processes are operative among males which our theory draws attention and puber-
is not known, since most of the relevant ani- tal timing might best be conceptualized as
mal and human research has focused on fe- curvilinear. Whereas moderate and even rel-
males. Obviously, the greater precision with atively high levels of stress may stimulate
which the timing of menarche can be mea- earlier maturation, extreme stress may shut
sured compared with parallel indices of pu- down physical maturation. The biological
bertal maturation in males contributes to our system may have evolved to accelerate sex-
limited knowledge about influences on the ual development when reproductive pros-
timing of male puberty. pects are perceived as limited and a com-
petitive edge is needed-thus earlier
Although the research we have consid- maturation-but to defer sexual develop-
ered provides intriguing support for our ment when the very survival, not simply the
propositions concerning contextual stress reproductive prospects, of.the organism are
and pubertal timing, we would be remiss if threatened (Surbey, 1987). In essence, un-
we did not draw attention to evidence that
der conditions of moderate to high stress, ca-
would seem to be inconsistent with our the-
loric resources may be marshaled in an effort
ory. Perhaps the most noteworthy is that re- to enhance reproductive possibility, but
garding the secular trend in physical devel- when stress becomes too extreme, the sys-
opment, as it provides the best evidence that tem functions to conserve resources by in-
pubertal timing is contextually plastic: For hibiting sexual development.
the past 200 years, age of menarche has de-
creased substantially in industrialized socie- It is beyond the scope and purpose of
ties. Although this change is routinely pre- this article to discuss the possible neuroen-
sumed to be the (not necessarily exclusive) docrinological processes by which early and
result of improvements in nutrition and gen- proximate social experiences may be related
eral living conditions (Eveleth & Tanner, to the timing of puberty. Suffice it to say that
1976; but see Adams, 1981, for an alternative it is widely accepted that the timing of pu-
explanation), we must acknowledge the pos- berty is regulated by a neuroendocrine sub-
sibility that industrialized society may in system that is intertwined with other endo-
certain respects be more stressful than its crine systems, and that this system is both
preindustrialized counterpart. Consider in dynamic and responsive to environmental
this regard not only that people live in closer stimuli (Brooks-Gunn & Reiter, in press). It
proximity to nonrelatives than they ever did is also important to note that the role of
before, but also that the intuitive sense that stress in the acceleration of puberty may be
parent-child relations and family life have organizational rather than activational. In
become more harmonious in recent centu-
other words, there may be important indi-
ries remains unconfirmed. vidual differences in neuroendocrine func-
Moreover, even if we attribute the secu- tioning that do not emerge somatically until
lar trend to improvements in nutrition and adolescence, but which originate in child-
health and regard it as inconsistent with our hood experience (e.g., Phoenix, Gpy, Gerall,
prediction that stress accelerates pubertal & Young, 1959). The notion that early
timing, the possibility must be entertained events may affect pubertal development is
that nutrition and health affect reproductive consistent with other research on the organi-
development in a manner different from zational effects of hormones on pubertal
other sources of influence. Consider in this maturation (e.g., Coe, Hayashi, & Levine,
regard animal research indicating that early 1988) as well as with data from the Standard
life stress accelerates physical development Cross-Cultural Sample (Murdock and others,
(Denenberg, Garbanati, Sherman, Yutzey, & 1971), indicating that average adult stature
Kaplan, 1980) and human research indicat- among males increases as a function of the
ing that anovulation and amenorrhea can re- degree to which the individual's society is
sult from starvation, excessive exercise, or characterized by high levels of stress and fa-
exceedingly high levels of stress (Frisch & ther absence during infancy (Gray, personal
McArthur, 1974; McClintock, 1980; Surbey, communication, no date). Such evidence,
1987). and our theorizing, together suggest that one
profitable strategy for enhancing under-
In point of fact, evidence that matura- standing of endocrinological processes in-
tional delay in human females results from volved in sexual maturation might involve
extreme physical exercise and starvation comparing the endocrinological systems of
(i.e., anorexia nervosa) raises the prospect prepubertal school-age children whom our
that the relation between some stressors to theory predicts will mature at different rates

This content downloaded from on Fri, 08 Apr 2016 21:17:06 UTC
All use subject to
660 Child Development

(i.e., those from high-risk environments pulsivity, independence striving, and un-
evincing behavior problems and nondisor- conventionality accounts for the clustering
dered age-mates from low-stress families). of problem behaviors in adolescence, in-
cluding substance abuse, delinquency, and
Pubertal timing and sexual activ- precocious sexual activity. According to our
ity.-We have argued that processes that in-
theory, these behaviors should cluster to-
duce early puberty are part of a reproductive
gether, but not because they are problem-
strategy that has evolved to foster early mat-
atic. Rather, we predict that these behaviors
ing and limited parental investment. More-
covary because they all manifest the same
over, we have theorized that in the envi-
opportunistic, risk-tasking reproductive
ronment of evolutionary adaptation, early
maturation fostered the early onset of sexual
activity in the service of procreation. Al- The links between family structure and
though one would no longer predict, given the onset of sexual activity are predicted
current cultural conditions, contraceptive from most theories of adolescent develop-
accessibility, and the widespread availabil- ment, but the mediating mechanisms pre-
ity of abortion, that early maturation neces-
sumed important in these theories are not
sarily leads to early childbearing, there is
operative. Specifically, the higher rate of
good reason to expect that it is related to sexual activity among adolescents from
earlier sexual activity.
single-parent homes is not attributable to
The onset of sexual activity in adoles- less vigilant parental monitoring (despite
cence is strongly linked to biological devel- the finding that single parents are less vigi-
opment: Boys and girls who mature early lant; Dornbusch et al., 1985), nor is it appar-
initiate intercourse at a younger age than ently a transient reaction to divorce-related
their peers (Aro & Taipale, 1987; Magnusson stress, since earlier sexual activity is ob-
et al., 1986; Smith, Udry, & Morris, 1985). served among girls from single-parent
The hormonal changes of puberty-and es- homes regardless of when or whether their
pecially the surge in testosterone-increase parents divorced (Newcomer & Udry, 1984,
the adolescent's sex drive, interest in sex, 1987). We contend that an interpretation of
and level of arousal when exposed to sexual this evidence in terms of reproductive strat-
stimuli (Smith et al., 1985; Udry, 1987; Udry, egy is illuminating.
Talbert, & Morris, 1986).
Childhood Context and Adult Personality
Because not all adolescents of equiva-
Determinants of Pair Bonding
lent pubertal status are equally likely to en- On the basis of our theorizing that link-
gage in sexual activity, studies of nonbiolog-
ages among childhood experience, psycho-
ical factors that appear to augment the direct
logical functioning, pubertal timing, and
impact of puberty on sexual behavior are es- sexual behavior have evolved to serve repro-
pecially interesting. Consistent with our ductive functions, we would expect child-
perspective, adolescents who come from
hood rearing experiences or environments
single-parent households or who evince
to be systematically related to variation in
signs of delinquency and substance abuse adult pair-bonding processes. Because such
are more likely than their same-age, compa-
linkages are assumed to be mediated, at least
rably developed peers to be sexually active in part, by the psychological orientation of
(Jessor, Costa, Jessor, & Donovan, 1983; the individual, adult personality should also
Newcomer & Udry, 1987; Thorton & Cam-
be related to patterns of pair bonding. We
burn, 1987). now consider evidence consistent with such
At first glance, the link between sexual notions.
activity and various "problem" behaviors
and environments is easily explained within Childhood family context and adult
conventional theories of socialization, but pair bonding.-Social scientists interested
upon further reflection one is forced to won- in the intergenerational transmission of di-
der about the underlying mechanism. In a vorce and the consequences of growing up
society that no longer labels nonmarital sex- in a single-parent family have conducted nu-
ual activity as "deviant," one must ask why merous studies linking childhood family
this behavior covaries with such censured context and adult pair bonding, though
behaviors as delinquency and substance rarely have these inquiries been discussed
abuse. Jessor (Jessor & Jessor, 1977; Jessor in such terms. Nevertheless, research on dat-
et al., 1983) has suggested that an underlying ing behavior, sexual activity, timing of mar-
personality constellation composed of im- riage, and marital instability are all, in es-

This content downloaded from on Fri, 08 Apr 2016 21:17:06 UTC
All use subject to
Belsky, Steinberg, and Draper 661

sence, investigations of pair bonding and marital quality, perhaps more than stability,
relationship processes. to be systematically related to personality or
psychological characteristics of adults; and,
With regard to dating, there is consistent
specifically, that marriages are more likely to
evidence that divorce and family conflict are
be satisfactory and stable when individuals
related to increased heterosexual activity
have personality characteristics that promote
among offspring (Demo & Acock, 1988; Kin-
mutually beneficial social exchange than
nard & Gerrard, 1986). Perhaps the first to
when they have personalities more geared
discern such a link was Hetherington (1972),
toward opportunism and advantage taking.
who observed that adolescent girls from di-
vorced families, but not girls whose fathers Terman (1938; Terman & Buttenwei-
had died, were more heterosexually assert- ser, 1935) was among the first to document
ive than girls from intact families. This evi- empirical associations between personality
dence is consistent with Booth, Brinkerhoff, characteristics of spouses and marital
and White's (1984) finding that, compared to functioning-personality characteristics, in
college students with intact families, those fact, that can easily be interpreted in terms
whose parents were divorced or perma- of the distinction we have been emphasizing
nently separated dated more often, and this between a secure, mutual-benefit orienta-
activity increased further if parental or tion and an insecure, opportunistic, and
parent-child conflict persisted during and self-centered one. Consider, in this regard,
after divorce (see also Kinnard & Gerrard, his summary of the self-descriptions of the
1986). Also in accord with such findings is most happy and least happy couples whom
consistent evidence that males and females he studied. Happily married women, he
not living with both biological parents initi- noted "are characterized by kindly attitudes
ate coitus earlier than their counterparts in toward others and by the expectation of
intact families (Hogan & Kitagawa, 1985; kindly attitudes in return. ... They are co-
Kinnard & Gerrard, 1986; Newcomer & operative. . . . They enjoy activities that
Udry, 1987). As we noted earlier, this does bring... pleasurable opportunities to oth-
not appear to be due to less vigilant monitor- ers . . . their expressed attitudes imply a
ing by single parents (Newcomer & Udry, quiet self-assurance and a decidedly opti-
1987). mistic outlook on life.... Unhappily married
women, on the other hand, are characterized
In light of all these results pertaining to by emotional tenseness .... They give evi-
dating and sexual behavior, it should not be dence of deep-seated inferiority feelings to
surprising to discover that children of di- which they react with aggressive atti-
vorce marry earlier than other individuals tudes. . . . They are inclined to be irrita-
(Carlson, 1979; Mueller & Pope, 1977), even ble. .... They are egocentric" (Terman, 1938,
after socioeconomic factors, are controlled pp. 145-146, emphasis added). Not surpris-
(Keith & Finlay, 1988; McLanahan & Bum- ingly, findings with regard to men were
pass, 1988). Moreover, in view of the fact much the same.
that age at marriage is itself related to marital
stability (e.g., Bumpass & Sweet, 1972; In the time since Terman's (1938) clas-
Glenn & Supancic, 1984), there is also little sic study, a large quantity of data have been
reason to be surprised that children of di- collected that are generally consistent with
vorce are more likely to experience, in adult- it (Dean, 1966; Eysenck, 1980; Pickford, Sig-
hood, the dissolution of their own marriages nori, & Rempel, 1966). One criticism that
(e.g., Glenn & Kramer, 1987; Keith & Finlay, can be wielded against all of this work is that
1988). This, too, cannot be attributed en- measures of personality and marriage were
tirely to socioeconomic factors (McLanahan obtained at roughly the same point in time.
& Bumpass, 1988). Fortunately, two recent longitudinal studies
demonstrate that personality actually fore-
Adult personality and marital qual- casts future marital quality. In one investiga-
ity.-The fact that dating, sexual activity, tion of 300 couples followed from their en-
and marital stability are each related to expo- gagements in the 1930s through 1980, Kelley
sure to divorce in childhood provides some and Conley (1987) found that higher levels
support for the contention that childhood of neuroticism (among men and women) and
family experiences are related to adult pair- lower levels of agreeableness (among men
bonding processes. In light of the emphasis only) predicted both greater marital instabil-
we place on the role of experience in foster- ity and lower marital satisfaction. In the sec-
ing opportunistic orientations toward others ond longitudinal study, Skolnick (1981) re-
and toward relationships, we would expect ported that a self-confident and nurturant

This content downloaded from on Fri, 08 Apr 2016 21:17:06 UTC
All use subject to
662 Child Development

(versus hostile) orientation during adoles- possibility that parents, for example, who are
cence positively predicted marital satisfac- genetically predisposed to rear their off-
tion at midlife. Although none of the find- spring in an insensitive, unresponsive, re-
ings reviewed prove to be particularly jecting, neglecting, and/or inconsistent man-
surprising, they are consistent with our pre- ner may bear children who are genetically
diction that psychological profiles which in- predisposed to be aggressive, develop be-
cline one to relate to others in an opportunis- havior problems, mature earlier than others,
tic manner are associated with less stable and become sexually active at earlier ages.
and less satisfying pair bonds in adulthood. Certainly consistent with this line of rea-
soning is evidence that onset of puberty
(measured by peak height velocity, menar-
Strong Inference Tests of the Theory che, or ratings of secondary sex characteris-
In summary, then, there is a rather ex- tics [Fischbein, 1977a, 1977b]) and age of
tensive literature that is in accord with pre- first sexual intercourse (Martin, Eaves, &
dictions derived from our evolutionary Eysenck, 1977) show substantial genetic in-
theory of socialization and interpersonal de- fluence (Plomin & Fulkner, 1987). However,
velopment. The evidence reviewed indi- because recent research indicates that a di-
cates (a) that under conditions of contextual mension of personality labeled "agreeable-
stress parental care is less sensitive, more ness," which defines a nice, friendly, and
inconsistent, and more affectively negative; trusting personal style of relating to others,
(b) that infants and young children exposed shows minimal genetic influence (Bergeman
to such parenting practices are more likely et al., in press), and that aggression might be
than others to develop insecure attachments one of the least heritable aspects of behavior
to their parents and to grow up evincing (for review, see Plomin, Nitz, & Rowe, in
"problematic" behavior and opportunistic press), we do not believe it is tenable to
styles of relating to others; (c) that rearing assume that genetics account for the entire
conditions which promote such patterns of relation between context, rearing, psycho-
psychological and behavioral development logical and somatic development, and
predict, in the few studies available, earlier reproductive strategy.
pubertal maturation; (d) that early puberty
Because the postulated effects of con-
is related to opportunistic styles of relating
text, experience, and psychological and be-
to others, including "problem" behavior, as
havioral development on puberty represent
well as (e) early onset of sexual activity; (f)
the most original-and uncanny-predic-
that early rearing environments character-
tion of our theory, it is essential to design
ized by marital dissolution and family con-
strong inference research (Platt, 1964) capa-
flict are predictive in adolescence and adult-
ble of disconfirming the prediction that un-
hood of frequent dating, early marriage, and
der conditions of contextual stress, "prob-
divorce; and, finally, (g) that marital dissolu-
lematical" rearing, and/or childhood
tion as well as poor marital quality are more
behavior problems, children will mature
likely when spouses are emotionally unsta-
earlier than their peers. Two research de-
ble, insecure, and self-centered.
signs seem most appropriate, one experi-
Despite the abundance of evidence con- mental, the other correlational. The correla-
sistent with our perspective, none of these tional design would involve identical and
data actually confirm the major precepts of fraternal twins reared apart and/or adopted
the theory. Indeed, the most fundamental and biological children. The central goal of
criticism that can be directed against the re- the research would be to test the effect of
search summarized is the same one that has the hypothesized antecedent conditions
been directed at much socialization research (contextual stress, problematic parenting,
(Plomin et al., 1985; Scarr & McCartney, and child behavior problems), after control-
1983). It involves the fact that although most ling for the heritability of pubertal timing
socialization theories, ours included, pre- and, ideally, of problematic parenting and
sume that childhood experiences causally behavior problems as well. If the hypothe-
influence psychological and behavioral de- sized antecedents of early puberty did not
velopment and, in our case, even somatic de- emerge once behavior genetic variance was
velopment, virtually all evidence amassed in statistically controlled, then a critical theo-
an attempt to verify them confounds genetic retical proposition would have been discon-
and environmental processes. firmed.

When this line of argument is applied to From a more experimental standpoint,

the theory we have advanced, it raises the we propose randomly assigning children

This content downloaded from on Fri, 08 Apr 2016 21:17:06 UTC
All use subject to
Belsky, Steinberg, and Draper 663
from troubled families and at high risk for theory-namely, that individuals are se-
the development of behavior problems to a lected on the basis of behaviors that max-
treatment condition known to be effective in imize reproductive fitness; and second, that
preventing behavior problems or to a non- students of Bowlby have presumed that a se-
treatment control group. On the basis of the cure, trusting attachment was the norm
theory outlined in this article, two predic- within the environment of evolutionary
tions would be tested. The first is that chil- adaptedness (e.g., Ainsworth et al., 1978;
dren assigned to the control treatment would Sroufe, 1979). Although Lamb et al. were
mature earlier than those in the experimen- correct on both counts, we believe they
tal group; the second is that within the ex- missed an opportunity to reinterpret Bowlby
perimental group, those for whom the treat- in light of a contemporary understanding of
ment proved most effective in preventing behavioral evolution (Hinde & Stevenson-
externalizing and/or internalizing behavior Hinde, 1990; Main, 1990). In essence, we
problems would mature later than would have argued that the variations in attach-
those within the experimental group for ment security that Bowlby's theory so
whom the treatment proved less effective. clearly anticipated evolved to serve repro-
Needless to say, if both research designs just ductive fitness goals in an ecologically sensi-
outlined were implemented and effects of tive manner.
antecedent conditions on puberty were dis- We also believe that our view extends in
cerned, this would provide very strong evi-
new and interesting directions several other
dence in favor of the theory.
popular perspectives on development. Nei-
ther Elder's (Elder et al., 1985) sociologi-
cally informed analysis of the life-course
By placing the study of socialization in consequences of economic deprivation nor
evolutionary perspective, we have sought to Patterson's (1986) social-learning formula-
integrate a variety of areas of inquiry that tion of family processes that give rise to
for the most part have remained relatively antisocial behavior-to cite but two
separate in the developmental literature. Al- examples-is informed by evolutionary
though a number of the topics our theory thinking in general or by ideas pertaining to
links together have been recognized as reproductive fitness more specifically. Nev-
related-as demonstrated in our summaries ertheless, both of these views can be easily
of research on contextual stress and harsh assimilated into the framework advanced in
parenting, attachment insecurity and behav- this article in a manner that simultaneously
ior problems, early puberty and sexual activ- preserves and enhances their contribution to
ity, and experiences in the family or origin the study of socialization and interpersonal
and subsequent pair bonding-no current development. While Elder and Patterson
theory of socialization provides a basis for each seeks to explain how particular life ex-
linking these literatures within an overarch- periences foster particular patterns of behav-
ing conceptual framework. We believe that ioral development (i.e., proximate causa-
an evolutionary perspective that emphasizes tion), neither addresses why human
individual differences in reproductive strat- development works in the way in which
egy offers a promising foundation. they describe it (i.e., ultimate causation). By
focusing on questions of ultimate causation,
One important way in which we have our theory seeks to extend and incorporate
recast traditional ideas about development theirs.
concerns our attempt to extend contempo-
rary attachment theory in a manner consis- The feature of our theory that distin-
tent with the modern view of evolution. In guishes it from others is the prediction that
his original formulation of attachment the- prepubertal rearing experiences and be-
ory, and "in the context of the evolutionary havioral developments influence the timing
theory of the time" (Main, 1990), Bowlby of puberty and that these developmental
(1969) asserted that attachment behaviors events and processes collectively affect ado-
evolved because actions on the part of the lescent sexual behavior and adult pair bond-
infant such as crying, following, and clinging ing. Central to our theory is the notion
served to increase the likelihood of species drawn from modern evolutionary biology
survival. Lamb, Thompson, Gardner, Char- that humans, like many other animals, often
nov, and Estes (1984) critiqued this feature adjust their life histories in response to con-
of the theory on two grounds: first, that it textual conditions in a manner that will en-
failed to incorporate understandings derived hance reproductive fitness-or at least
from the "modern synthesis" of evolutionary would have in the environment of evolution-

This content downloaded from on Fri, 08 Apr 2016 21:17:06 UTC
All use subject to
664 Child Development

ary adaptation. In this regard it is important on social and psychobiological processes

to note that the influence which context, connected to pubertal maturation, we be-
rearing, and behavioral development may lieve the theory highlights the need for new
have on pubertal timing today may be less research on learning in infancy and early
than it once was. In fact, in light of the secu- childhood. Cosmides and Tooby (1987) as
lar trend in pubertal timing, the age of repro- well as Symons (1989) have recently argued
ductive maturation may be less plastic and that humans have evolved to process infor-
less responsive to the developmental experi- mation in particular ways or to learn some
ences to which our theory draws attention things more easily than others because they
than it was when evolutionary pressures first are especially important with regard to re-
shaped the processes of human develop- productive fitness. When such ideas are ap-
ment. We continue to expect prepubertal ex- plied to the developmental theory advanced
periences to influence pubertal timing, but in this article, the prospect is raised that
acknowledge that the magnitude of effect there might be certain features of experience
may be more limited than once might have and relationships that young children are
been the case. keen at discerning because these help set
the youngster on a developmental trajectory
In order to test this prediction, it will toward a reproductive strategy consistent
be necessary not only to discount behavior with the demands of his or her perceived
genetic explanations but to consider seri- environment (Draper & Harpending, 1982;
ously the conditional version of our theory. Weinrich, 1977). Thus, children should be
This strongest version of our theory asserts especially attuned-through their own emo-
that puberty will be earlier among children tional experience-to whether their own
(1) who grow up in contexts of stress, (2) and needs are being served via the care they are
experience parent-child relationships that being given. Associations between attach-
are rejecting or aversive, (3) and evince pre- ment security and responsive maternal care
pubertal behavior problems. In other words, can be reinterpreted in this manner.
adequate tests of the theory will need to take
into consideration rearing context, parent- Kurt Lewin once wrote that there is
child relations, and prepubertal behavioral nothing so practical as a good theory. More
development, at least when it comes to try- specifically, we would argue that a good the-
ing to predict pubertal timing. ory should do three things: (1) help to reor-
ganize old findings in a manner that gener-
In only a limited manner have we ad- ates new understanding; (2) generate new
dressed issues concerning the possible and testable hypotheses; and, as a result, (3)
neuroendocrinological or endocrinological lead to new inquiry and new discoveries. As
mechanisms-either in adolescence or to the first of these aims, we believe that our
earlier-that may link social experiences perspective both enhances and extends ex-
and somatic development. Certainly the isting models of socialization in a manner
findings of Moffitt et al. (1990), Steinberg that is consistent with recent developments
(1988), and Surbey (1990) underscore a need in the fields of evolutionary biology and be-
for more research on the physiological and havioral ecology. As to the second of these
hormonal processes that mediate the associ- aims, we have attempted to set forth a num-
ation between family relations and pubertal ber of specific and testable hypotheses about
timing. More generally, however, more re- the links between context, early experience,
search attention needs to be paid to the rele- social behavior, somatic development, sexu-
vant developmental processes that are set in ality, and family formation. Whether the the-
motion well before puberty but that influ- ory we have advanced actually leads to the
ence its timing and tempo. It is time that development of new knowledge awaits em-
developmentalists interested in adolescence pirical inquiry.
begin to systematically examine pubertal
timing as an outcome of social experience,
rather than presuming the reverse causality. References
At the very least, we hope that our theorizing
stimulates research that challenges the "con- Adams, J. (1981). Earlier menarche, greater height
ventional wisdom" that treats pubertal tim- and weight: A stimulation-stress factor hy-
ing as the independent variable and psycho- pothesis. Genetic Psychology Monographs,
logical and behavioral development as 104, 3-22.
dependent variables. Ainsworth, M., Blehar, M., Waters, E., & Wall, S.
(1978). Patterns of attachment. Hillsdale, NJ:
In addition to stimulating new research Erlbaum.

This content downloaded from on Fri, 08 Apr 2016 21:17:06 UTC
All use subject to
Belsky, Steinberg, and Draper 665
Aro, H., & Taipale, V. (1987). The impact of tim- erman & C. Hayes (Eds.), Families that work
ing of puberty on psychosomatic symptoms (pp. 39-83). Washington, DC: National Acad-
among fourteen- to sixteen-year-old Finnish emy Press.
girls. Child Development, 58, 261-268. Brooks-Gunn, J., & Reiter, E. (in press). Pubertal
Bandura, A. (1977). Social learning theory. Engle- development. In S. Feldman & G. Elliott
wood Cliffs, NJ: Prentice-Hall. (Eds.), At the threshold: The developing ado-
Barkow, J. (1984). The distance between genes lescent. Cambridge, MA: Harvard University
and culture. Journal of Anthropological Re- Press.

search, 40, 9-14. Bumpass, L., & Sweet, J. (1972). Differentials in

Baumrind, D. (1971). Current patterns of parental marital instability: 1970. American Sociologi-
authority. Developmental Psychology Mono- cal Review, 37, 754-766.
graphs, 4, 1-103. Burgess, R., & Draper, P. (1989). The explanation
Belsky, J. (1980). Child maltreatment: An ecologi- of family violence: The role of biological be-
cal integration. American Psychologist, 35, havioral and cultural selection. In L. Ohlin &
320-335. M. Tonry (Eds.), Family violence (pp.
Belsky, J. (1981). Early human experience: A fam- 59-116). Chicago: University of Chicago
ily perspective. Developmental Psychology, Press.
17, 3-23. Caine, N. (1986). Behavior during puberty and ad-
Belsky, J. (1984). The determinants of parenting: olescence. In G. Mitchell & J. Erwin (Eds.),
A process model. Child Development, 55, Comparative primate biology: Vol. 2A. Be-
83-96. havior, conservation, and ecology (pp. 327-
Belsky, J. (1990). Parental and nonparental care 361). New York: Liss.
and children's socioemotional development: Campbell, S. (in press). Developmental perspec-
A decade in review. Journal of Marriage and tives on child psychopathology. In M. Hersen
the Family, 52, 885-903. & T. Ollendick (Eds.), Handbook of child
Belsky, J., Rovine, M., & Taylor, D. G. (1984). The psychopathology (2d ed.). New York:
Pennsylvania Infant and Family Develop- Plenum.
ment Project: III. The origins of individual Carlson, E. (1979). Family background, school,
differences in infant-mother attachment. Ma- and early marriage. Journal of Marriage and
ternal and infant contributions. Child Devel- the Family, 41, 341-353.
opment, 55, 718-728. Caspi, A., & Bem, D. (1990). Personality continu-
Bergeman, C., Chipuer, H., Plomin, R., Pedersen, ity and change across the life course. In L. A.
N., McClearn, G., Nesselroade, J., & Costa, P. Pervin (Ed.), Handbook of personality theory
(in press). Openness to experience, agree- and research (pp. 549-575). New York:
ableness, and conscientiousness assessed Guilford.
separately by age and gender: An adoption/ Chisholm, J. (1988). Toward a developmental evo-
twin study. Journal of Personality and Social lutionary ecology. In K. MacDonald (Ed.),
Psychology. Sociobiological perspectives on human devel-
Biller, H. (1981). Father absence, divorce, and opment (pp. 78-102). New York: Springer
personality development. In M. Lamb (Ed.), Verlag.
The role of the father in child development Cicchetti, D., & Braunwald, K. (1984). An organi-
(pp. 319-358). New York: Wiley. zational approach to the study of emotional
Booth, A., Brinkerhoff, D., & White, C. (1984). The development in maltreated infants. Infant
impact of parental divorce on courtship. Jour- Mental Health Journal, 5, 172-183.
nal of Marriage and the Family, 46, 85-94. Cicchetti, D., & Carlson, V. (Eds.). (1989). Child
Bowlby, J. (1969). Attachment and loss: Vol. 1. maltreatment: Theory and research on the
Attachment. New York: Basic. causes and consequences of child abuse ne-
Bradley, R. H., Caldwell, B. M., & Rock, S. L. glect. New York: Cambridge.
(1988). Home environment and school perfor- Coe, C., Hayashi, K., & Levine, S. (1988). Hor-
mance: A ten-year follow-up and examination mones and behavior at puberty: Activation or
of three models of environmental action. concatenation. In M. Gunnar & W. A. Collins
Child Development, 59, 852-867. (Eds.), The Minnesota symposia on child psy-
Bradshaw, A. (1965). Evolutionary significance of chology (Vol. 21, pp. 17-41). Hillsdale, NJ:
phenotypic plasticity in plants. Advances in Erlbaum.
Genetics, 13, 115-155. Conger, R., McCarty, J., Yang, R., Lahey, B., &
Bronfenbrenner, U. (1977). Toward an experimen- Kropp, J. (1984). Perception of child, child-
tal ecology of human development. American rearing values, and emotional distress as me-
Psychologist, 32, 513-531. diating links between environmental stress-
Bronfenbrenner, U., & Crouter, A. (1982). Work ors and observed maternal behavior. Child
and family through time and space. In S. Kam- Development, 54, 2234-2247.

This content downloaded from on Fri, 08 Apr 2016 21:17:06 UTC
All use subject to
666 Child Development
Cosmides, L., & Tooby, J. (1987). From evolution interactional perspective. In M. Magnussen
to behavior: Evolutionary psychology as the & V. Allen (Eds.), Human development: An
missing link. In J. Dupre (Ed.), The latest and interactional perspective (pp. 113-147). New
best: Essays on evolution and optimality (pp. York: Academic Press.
277-306). Cambridge, MA: MIT Press. Erickson, M. F., Sroufe, L. A., & Egeland, B.
Crawford, C., & Anderson, J. (1989). Sociobiology: (1985). The relationship between quality of
An environmentalist discipline? American attachment and behavior problems in pre-
Psychologist, 44, 1449-1460. school in a high-risk sample. In I. Bretherton
Crouter, A. C. (in press). Processes linking fami- & E. Waters (Eds.), Growing points of attach-
lies and work: Implications for development ment theory and research (pp. 147-193).
in both settings. In R. Parke & S. Kellam Monographs of the Society for Research
(Eds.), Advances in family research, Vol. 2. in Child Development, 50(1-2, Serial No.
New York: Erlbaum. 209).
Cutler, W. B., Garcia, C. R., & Krieger, A. B. Evans, S., & Hodges, J. (1984). Reproductive sta-
(1979). Sexual behavior frequency and men- tus of adult daughters in family groups of com-
strual cycle length in mature premenopausal mon marmosets (Callithrix jacchus jacchus).
women. Psychoneuroendocrinology, 4, 297- Folia Primatologica, 42, 127-133.
309. Eveleth, P., & Tanner, J. (1976). Worldwide varia-
Dean, D. (1966). Emotional maturity and marital tion in human growth. New York: Cambridge
adjustment. Journal of Marriage and the University Press.
Family, 28, 454-457. Eysenck, H. (1980). Personality, marital satisfac-
Demo, D., & Acock, A. (1988). The impact of di- tion, and divorce. Psychological Reports, 47,
vorce on children. Journal of Marriage and 1235-1238.
the Family, 50, 619-648. Fischbein, S. (1977a). Onset of puberty in MZ and
Denenberg, V., Garbanati, J., Sherman, G., Yut- DZ twins. Acta Geneticae Medicne et Gemel-
zey, D., & Kaplan, R. (1980). Infant stimula- lologiae, 26, 151-158.
tion induces brain lateralization in rats. Sci- Fischbein, S. (1977b). Intra-pair similarity in
ence, 201, 1150-1152. physical growth of monozygotic and dizygotic
Digman, J. (1989). Five robust traits and dimen- twins during puberty. Annals of Human Biol-
sions: Development, stability, and utility. ogy, 4, 417-430.
Journal of Personality, 57, 195-214. Frisch, R., & McArthur, J. (1974). Menstrual cy-
Dornbusch, S., Carlsmith, J., Bushwall, S., Ritter, cles: Fatness as a determinant of minimum
P., Leiderman, H., Hastorf, A., & Gross, R. weight necessary for their maintenance or on-
(1985). Single parents, extended households, set. Science, 185, 949-951.
and the control of adolescents. Child Devel- Glenn, N., & Kramer, K. (1987). The marriages
opment, 56, 326-341. and divorces of the children of divorce. Jour-
Draper, P., & Belsky, J. (1990). The relevance of nal of Marriage and the Family, 49, 811-825.
evolutionary thinking for issues in personality Glenn, N., & Supanic, M. (1984). The social and
development. Journal of Personality, 58, demographic correlates of divorce and separa-
141-162. tion in the United States: An update and re-
Draper, P., & Harpending, H. (1982). Father ab- consideration. Journal of Marriage and the
sence and reproductive strategy: An evolu- Family, 46, 563-576.
tionary perspective. Journal of Anthropologi- Grossmann, K., Grossmann, K. E., Spangler, G.,
cal Research, 38, 255-273. Suess, G., & Unzner, L. (1985). Maternal sen-
Duncan, P., Ritter, P., Dornbusch, S., Gross, R., & sitivity and newborn orientation responses as
Carlsmith, J. (1985). The effects of pubertal related to quality of attachment in Northern
timing on body image, school behavior, and Germany. In I. Bretherton & E. Waters (Eds.),
deviance. Journal of Youth and Adolescence, Growing points of attachment theory and re-
14, 227-236. search (pp. 233-256). Monographs of the So-
Elder, G. (1981). History and the life course. In ciety for Research in Child Development,
D. Bertaux (Ed.), Biography and society: The 50(1-2, Serial No. 209).
life history approach in the social sciences Hamilton, W. (1964). The genetical theory of so-
(pp. 77-115). Beverly Hills, CA: Sage. cial behavior. Journal of Theoretical Biology,
Elder, G., Nguyen, T., & Caspi, A. (1985). Linking 7, 1-52.
family hardship to children's lives. Child De- Herrenkohl, E. C., & Herrenkohl, R. C. (1981).
velopment, 56, 361-375. Some antecedents and developmental conse-
Emery, R. (1988). Marriage, divorce, and chil- quences of child maltreatment. New Direc-
dren's adjustment. Beverly Hills, CA: Sage. tions for Child Development, 11, 57-76.
Epstein, E., & Erskine, N. (1983). The develop- Hetherington, E. M. (1972). Effects of paternal ab-
ment of personal theories of reality from an sence on personality development in adoles-

This content downloaded from on Fri, 08 Apr 2016 21:17:06 UTC
All use subject to
Belsky, Steinberg, and Draper 667
cent daughters. Developmental Psychology, Kelley, E., & Conley, J. (1987). Personality and
7, 313-326. compatibility: A prospective analysis of mari-
Hill, J., Holmbeck, G., Marlow, L., Green, T., & tal stability and marital satisfaction. Journal
Lynch, M. (1985a). Pubertal status and of Personality and Social Psychology, 52, 27-
parent-child relations in families of seventh- 40.

grade boys. Journal of Early Adolescence, 5, Kinnard, K., & Gerrard, M. (1986). Premarital sex-
31-44. ual behavior and attitudes toward marriage
Hill, J., Holmbeck, G., Marlow, L., Green, T., & among young women as a function of their
Lynch, M. (1985b). Menarcheal status of mothers' marital status. Journal of Marriage
parent-child relations in families of seventh- and the Family, 48, 757-765.
grade girls. Journal of Youth and Adoles- Kochanska, G., Kuczynski, L., Radke-Yarrow, M.,
cence, 14, 301-316. & Welsh, J. (1987). Resolutions of control epi-
Hinde, R. (1982). Attachment: Some conceptual sode between well and affectively ill mothers
and biological issues. In C. M. Parkes & J. and their young children. Journal of Abnor-
Stevenson-Hinde (Eds.), The place of attach- mal Child Psychology, 15, 441-446.
ment in human development. London: Tavi- Konishi, M., Emlen, S., Ricklets, R., & Wingfield,
stock. J. (1989). Contributions of bird studies to biol-
Hinde, R., & Stevenson-Hinde, J. (1990). Attach- ogy. Science, 46, 465-472.
ment: Biological, cultural, and individual de- Krebs, J., & Davies, N. (1981). An introduction
siderata. Human Development, 33, 48-61. to behavioral ecology. Sunderland, MA:
Hogan, D., & Kitagawa, E. (1985). The impact of Sinauer.
social status, family structure, and neighbor- Kuczynski, L., Kochanska, G., Radke-Yarrow, M.,
hood on the fertility of black adolescents. & Girniss-Brown, D. (1987). A develop-
American Journal of Sociology, 90, 825-855. mental interpretation of young children's
Holden, G. W., & West, M. J. (1989). Proximate noncompliance. Developmental Psychology,
regulation by mothers: A demonstration of 23, 1-8.
how differing styles affect young children's Lack, D. (1947). Darwin's finches. New York:
behavior. Child Development, 60, 64-69. Cambridge University Press.
Howes, C., & Eldredge, R. (1985). Responses of LaFreniere, P. J., & Sroufe, L. A. (1985). Profiles
abused, neglected, and non-maltreated chil- of peer competence in the preschool: Interre-
dren to the behaviors of their peers. Journal lations between measures, influence of social
of Applied Developmental Psychology, 6, ecology, and relation to attachment history.
261-270. Developmental Psychology, 21, 56-69.
Jacobson, J. L., & Wille, D. E. (1986). The influ- Lamb, M., Thompson, R., Gardner, W., Charnov,
ence of attachment pattern on developmental E., & Estes, D. (1984). Security of infantile
changes in peer interaction from the toddler attachment as assessed in the Strange Situa-
to the preschool period. Child Development, tion: Its study and biological interpretation.
57, 338-339. Behavioral and Brain Sciences, 7, 127-147.
Jessor, R., Costa, F., Jessor, L., & Donovan, J. Lancaster, J. (1986). Human adolescence and re-
(1983). Time of first intercourse: A prospec- production. In J. Lancaster & B. Hamburg
tive study. Journal of Personality and Social (Eds.), School-age pregnancy and parenthood
Psychology, 44, 608-626. (pp. 17-38). New York: Aldine.
Jessor, R., & Jessor, S. (1977). Problem behavior Lempers, J., Clark-Lempers, D., & Simons, R.
and psychosocial development: A longitudi- (1989). Economic hardship, parenting, and
nal study of youth. New York: Academic distress in adolescence. Child Development,
Press. 60, 25-49.
Jones, B., Leeton, J., McLeod, I., & Wood, C. Levin, R., & Johnston, R. (1986). Social mediation
(1972). Factors influencing the age of menar- of puberty: An adaptive female strategy? Be-
che in a lower socio-economic group in Mel- havioral and Neural Biology, 46, 308-324.
bourne. Medical Journal of Australia, 21, Lewis, M., Feiring, C., McGuffog, C., & Jaskir, J.
533-535. (1984). Predicting psychopathology in six-
Jouriles, E., Pfiffner, L., & O'Leary, S. (1988). year-olds from early social relations. Child
Marital conflict, parenting, and toddler con- Development, 55, 123-136.
duct problems. Journal of Abnormal Child Livson, N., & Peskin, H. (1980). Perspectives on
Psychology, 16, 197-206. adolescence from longitudinal research. In J.
Keith, V., & Finlay, B. (1988). The impact of pa- Adelson (Ed.), Handbook of adolescent psy-
rental divorce on children's educational at- chology (pp. 148-187). New York: Wiley.
tainment, marital timing, and likelihood of di- Liutkenhaus, P., Grossmann, K. E., & Gross-
vorce. Journal of Marriage and the Family, mann, K. (1985). Infant-mother attachment at
50, 797-809. twelve months and style of interaction with a

This content downloaded from on Fri, 08 Apr 2016 21:17:06 UTC
All use subject to
668 Child Development
stranger at the age of three years. Child De- McLanahan, S., & Bumpass, L. (1988). Intergener-
velopment, 56, 1538-1542. ational transmission of family disruption.
Lyons-Ruth, K., Connell, D. B., Zoll, D., & Stahl, American Journal of Sociology, 94, 130-152.
J. (1987). Infants at social risk: Relations McLoyd, V. (1990). The declining fortunes of
among infant maltreatment, maternal behav- black children: Psychological distress, par-
ior, and infant attachment behavior. Develop- enting, and socioemotional development in
mental Psychology, 2(2), 223-232. the context of economic hardship. Child De-
MacDonald, K. (Ed.). (1988). Sociobiological per- velopment, 61, 311-346.
spectives on human development. New York: Metral, G. (1981). The action of natural selection
Springer-Verlag. on the human menstrual cycle: A simulation
Magnusson, D., Statin, H., & Allen, V. (1986). Dif- study. Journal of Biosocial Science, 13,
ferential maturation among girls and its rela- 337-343.
tion to social adjustment in a longitudinal per- Moen, P. (1982). The two-provider family: Prob-
spective. In P. Baltes, D. Featherman, & R. lems and potentials. In M. Lamb (Ed.), Non-
Lerner (Eds.), Life span development and traditional families: Parenting and child de-
behavior (Vol. 7, pp. 74-101). Hillsdale, NJ: velopment. Hillsdale, NJ: Erlbaum.
Erlbaum. Moffitt, T., Caspi, A., & Belsky, J. (1990, March).
Main, M. (1990). Cross-cultural studies of attach- Family context, girls' behavior, and the onset
ment organization: Recent studies, changing of puberty: A test of a sociobiological model.
methodologies, and the concept of condi- Paper presented at the biennial meetings of
tional strategies. Human Development, 33, the Society for Research in Adolescence, At-
48-61. lanta, GA.
Main, M., & Weston, D. R. (1981). The quality Mueller, C., & Pope, H. (1977). Marital instability:
of the toddler's relationship to mother and to A study of its transmission between genera-
father. Related to conflict behavior and the tions.Journal of Marriage and the Family, 39,
readiness to establish new relationships. 83-93.
Child Development, 52, 932-940. Murdock, G., & others. (1971). Outline of cultural
Malatesta, C., & Wilson, A. (1988). Emotion/cog- materials (4th ed.). New Haven, CT: Human
nition interaction in personality develop- Relations Area Files.
ment: A discrete emotions, functionalist anal- Newcomer, S., & Udry, J. (1984). Mothers' influ-
ysis. BritishJournal of Social Psychology, 27, ence on the sexual behavior of their teenage
91-112. children. Journal of Marriage and the Family,
Marshall, W. (1978). Puberty. In F. Falkner & J. May, 477-485.
Tanner (Eds.), Human growth (Vol. 2, pp. Newcomer, S., & Udry, J. (1987). Parental marital
212-244). New York: Plenum. status effects on adolescent sexual behavior.
Martin, J. A. (1981). A longitudinal study of the Journal of Marriage and the Family, 49,
consequences of early mother-infant interac- 235-240.

tion: A microanalytic approach. Monographs Olweus, D. (1980). Familial and temperamental

of the Society for Research in Child Develop- determinants of aggressive behavior in ado-
ment, 46(3, Serial No. 190). lescent boys: A causal analysis. Develop-
Martin, N. G., Eaves, L. J., & Eysenck, H. J. mental Psychology, 16, 644-660.
(1977). Genetical, environmental, and person- Papini, D., & Sabby, R. (1987). Adolescent puber-
ality factors influencing the age of first sexual tal status and affective family relationships: A
intercourse in twins. Journal of Biosocial Sci- multivariate assessment. Journal of Youth
ence, 9, 91-97. and Adolescence, 16, 1-15.
Matas, L., Arend, R., & Sroufe, L. (1978). Continu- Parpal, M., & Maccoby, E. E. (1985). Maternal re-
ity in adaptation in the second year: The rela- sponsiveness and subsequent child compli-
tionship between quality of attachment and ance. Child Development, 56, 1326-1334.
later competence. Child Development, 49, Patterson, G. R. (1986). Performance models for
547-556. antisocial boys. American Psychologist, 41,
McClintock, M. K. (1980). Major gaps in menstrual 432-444.

cycle research: Behavioral and physiological Pelton, L. (1978). The myth of classlessness in
controls in a biological context. In P. Komen- child abuse cases. American Journal of Or-
ich, M. McSweeney, J. Novack, & N. Elder thopsychiatry, 48, 569-579.
(Eds.), The menstrual cycle (Vol. 7, pp. 7-23). Petersen, A. (1985). Pubertal development as a
New York: Springer. cause of disturbance: Myths, realities, and
McLanahan, S., & Booth, K. (1989). Mother-only unanswered questions. Genetic, Social, and
families: Problems, prospects, and politics. General Psychology Monographs, 111,
Journal of Marriage and the Family, 51, 205-232.
557-580. Pettit, G. S., & Bates, J. E. (1989). Family interac-

This content downloaded from on Fri, 08 Apr 2016 21:17:06 UTC
All use subject to
Belsky, Steinberg, and Draper 669
tion patterns and children's behavior prob- The social and psychological effects of pu-
lems from infancy to four years. Develop- berty on white females. In J. Brooks-Gunn &
mental Psychology, 25, 413-420. A. Petersen (Eds.), Girls at puberty (pp. 111-
Phoenix, C., Gpy, R., Gerall, A., & Young, W. 140). New York: Plenum.
(1959). Organizing action of prenatally admin- Skolnick, A. (1981). Married lives: Longitudinal
istered testosterone propionate on the tissues perspectives on marriage. In D. Eichorn, J.
mediating mating behavior in the female Clausen, N. Haan, M. Honzika, & P. Musser
guinea pig. Endocrinology, 65, 369-382. (Eds.), Present and past in middle life (pp.
Pickford, J., Signori, E., & Rempel, H. (1966). In- 243-265). New York: Academic Press.
tensity of personality traits in relation to mari- Smith, E., Udry, J., & Morris, N. (1985). Pubertal
tal happiness. Journal of Marriage and the development and friends: A biosocial expla-
Family, 28, 458-459. nation of adolescent sexual behavior. Journal
Platt, J. (1964). Strong inference. Science, 146, of Health and Social Behavior, 26, 183-192.
347-353. Sroufe, L. A. (1979). The coherence of individual
Plomin, R., & Fulker, D. (1987). Behavioral genet- development. American Psychologist, 34,
ics and development in early adolescence. In 834-841.
R. Lerner & T. Foch (Eds.), Biosocial- Stearns, S. (1982). The role of development in the
psychosocial interactions in early adoles- evolution of life histories. In J. T. Bonner
cence (pp. 63-94). Hillsdale, NJ: Erlbaum. (Ed.), Evolution and development (pp.
Plomin, R., Loehlin, J. C., & DeFries, J. C. (1985). 47-73). New York: Springer-Verlag.
Genetic and environmental components of Steinberg, L. (1987). The impact of puberty on
environmental influences. Developmental family relations: Effects of pubertal status
Psychology, 21, 391-401. and timing. Developmental Psychology, 23,
Plomin, R., Nitz, K., & Rowe, D. (in press). Behav- 451-460.
ioral genetics and aggressive behavior in Steinberg, L. (1988). Reciprocal relation between
childhood. In M. Lewis & S. M. Miller (Eds.), parent-child distance and pubertal matura-
Handbook of developmental psychopathol- tion. Developmental Psychology, 24,
ogy. New York: Plenum. 122-128.
Renken, B., Egeland, B., Marvinney, D., Steinberg, L. (1989). Pubertal maturation and
Mangelsdorf, S., & Sroufe, L. A. (1989). Early parent-adolescent distance: An evolutionary
childhood antecedents of aggressive and pas- perspective. In G. Adams, R. Montemayor, &
sive withdrawal in early elementary school. T. Gullotta (Eds.), Advances in adolescent de-
Journal of Personality, 57, 257-282. velopment (Vol. 1, pp. 82-114). Beverly Hills,
Rohner, R. (1975). Parental acceptance-rejection CA: Sage.
and personality: A universalistic approach to Steinberg, L., Catalano, R., & Dooley, D. (1981).
behavioral science. In R. Brislin et al. (Eds.), Economic antecedents of child abuse and ne-
Cross-cultural perspectives on learning (pp. glect. Child Development, 52, 975-985.
66-90). New York: Halsted. Surbey, M., (1987). Anorexia nervosa, amenor-
Rosenblum, L., & Paully, G. (1984). The effects of rhea, and adaptation. Ethology and Sociobiol-
varying environmental demands on maternal ogy, 8, 475-615.
and infant behavior. Child Development, 55, Surbey, M. (1990). Family composition, stress,
305-314. and human menarche. In F. Bercovitch & T.
Rushton, J. (1985). Differential K theory: The so- Zeigler (Eds.), The socioendocrinology of pri-
ciobiology of individual and group differ- mate reproduction (pp. 71-97). New York:
ences. Personality and Individual Differ- Liss.
ences, 6, 441-452. Susman, E., Inhoff-Germain, G., Nottlemann, E.,
Rutter, M. (1970). Sex differences in children's re- Loriaux, D., Cutler, G., & Chrousos, G.
sponse to family stress. In E. J. Anthony & C. (1987). Hormones, emotional dispositions,
Koupernik (Eds.), The child in his family (pp. and aggressive attributes in young adoles-
27-54). New York: Wiley. cents. Child Development, 58, 1114-1134.
Scarr, S. & McCartney, K. (1983). How people Symons, D. (1989). A critique of Darwinian an-
make environments: A theory of genotype en- thropology. Ethology and Sociobiology, 10,
vironment effects. Child Development, 54, 131-144.
424-435. Tardif, S. (1984). Social influences on sexual matu-
Short, R. V. (1976). The evolution of human repro- ration of female Saquinas oedipus oedipus.
duction. Proceedings of the Royal Society of American Journal of Primatology, 6, 199-
London, 195, 3-24. 209.
Simmons, R., & Blyth, D. (1987). Moving into ado- Terman, L. (1938). Psychological factors in mari-
lescence. New York: Aldine de Gruyter. tal happiness. New York: McGraw-Hill.
Simmons, R., Blyth, D., & McKinney, K. (1983). Terman, L., & Buttenweiser, P. (1935). Personal-

This content downloaded from on Fri, 08 Apr 2016 21:17:06 UTC
All use subject to
670 Child Development

ity factors in marital compatibility. Journal of Waters, E., Wippman, J., & Sroufe, L. A. (1979).
Social Psychology, 6, 143-171. Attachment, positive affect, and competence
Thorton, A., & Camburn, E. (1987). The influence in the peer group: Two studies in construct
of the family on premarital sexual attitudes validation. Child Development, 50, 821-829.
and behavior. Demography, 24, 323-340. Weinrich, J. D. (1977). Human sociobiology:
Tinbergen, N. (1963). On the aims of methods of Pair-bonding and resource predictability. Be-
ethology. Zeitschfrift fur Tierpsycologie, 20, havioral Ecology and Sociobiology, 2,
410-433. 91-118.
Trevathan, W., & Burleson, M. (1989, August). Whiting, B., & Whiting, J. (1975). Children of six
Human pair bonding and non-ovulatory sex- cultures. Cambridge, MA: Harvard Univer-
ual activity in females: New hypotheses. Pa- sity Press.
per presented at the first annual meeting of Worthman, C. (in press). Socioendocrinology: Key
the Human Behavior and Evolution Society, to a fundamental synergy. In T. Ziegler & F.
Northwestern University, Evanston, IL. Bercovitch (Eds.), The socioendocrinology of
Trivers, R. (1974). Parent-offspring conflict. Amer- primate reproduction. New York: Liss.
ican Zoologist, 14, 249-264. Youngblade, L., & Belsky, J. (1989). Child mal-
Udry, J. (1987). Hormonal and social determinants treatment, infant-peer attachment security,
of adolescent sexual initiation. In J. Bancroft and dysfunctional peer relationships in tod-
(Ed.), Adolescence and puberty (pp. 147- dlerhood. Topics in Early Childhood Educa-
163). New York: Oxford University Press. tion, 9, 1-15.
Udry, J., Talbert, L., & Morris, N. (1986). Biosocial Youngblade, L., & Belsky, J. (in press). The social
foundations for adolescent female sexuality. and emotional consequences of child mal-
Demography, 23, 217-230. treatment. In R. Ammerman & M. Herson
Wachtel, P. (1973). Psychodynamics, behavior (Eds.), Children at risk: An evaluation of fac-
therapy, and the implacable experiments. tors contributing to child abuse and neglect.
Journal of Abnormal Psychology, 82, 324- New York: Plenum.

This content downloaded from on Fri, 08 Apr 2016 21:17:06 UTC
All use subject to