1. Art and Diseño

DevelopmentalPsychology
1996, Vol. 32, No. 3,425-434
Copyright 1996 by the American PsychologicalAssociation, Inc.
0012-1649/96/$3.00
The Development of Behavior Before Birth
W i l l i a m P. S m o t h e r m a n a n d S c o t t R . R o b i n s o n
State University of New York at Binghamton
Study of the fetus in vivo provides a simple system for experimental study of early neurobehavioral
development. This review summarizes research on the development of behavior before birth including studies where fetuses are exposed to stimuli that mimic features of the neonatal environment,
such as milk and an artificial nipple. These stimuli reliably evoke responses from fetal subjects,
including species-typical behavior such as the stretch response and oral grasping of the artificial
nipple. Contingent presentations of the nipple and milk can result in classical conditioning, including activation of the endogenous opioid system. Quantitative analysis of fetal motor behavior coupled
with the use of ecologically relevant sensory manipulations provide a means for assessing integrated
output of the developing nervous system.
A general article with the title "The Development of Behavior
Before Birth" implies a number of topics that might be discussed. One may anticipate that such an article would provide
an historical overview of the field of behavioral embryology or
a survey of the diverse approaches used to measure behavior
in utero. Recent reviews have appeared that have adequately
described historical context (Oppenheim, 1982, 1992) and research methods that currently are permitting study of prenatal
behavioral development (Nijhuis, 1992; Smotherman & Robinson, 1988a), so these issues will not be addressed in the present report. Alternatively, one might expect such an article to
pose speculative scenarios for prenatal learning or cognition, or
to suggest how parents can interact with the fetus within the
womb. The subject of prenatal enrichment has received widespread attention in the popular press, but dubious claims by
authors and marketeers actually may have impeded empirical
investigation of fetal behavioral capacities. Rather than belaboring issues that have been adequately addressed elsewhere, or
which do not merit scientific discussion, this article will focus
on the findings and implications that have emerged from experimental study of fetal behavior, primarily in nonhuman species.
It is our thesis that the mammalian fetus is interesting in its
own right because it exhibits a rich and sophisticated behavioral
repertoire (Robinson & Smotherman, 1992a; Smotherman &
Robinson, 1990), and that study of the mechanisms that regulate behavior in the nonhuman fetus can provide unique insights into the general rules that direct neurobehavioral development and the research methods needed to identify them
(Robinson & Smotherman, 1988; Smotherman & Robinson,
1988b, 1994a).
Behavioral Study From Points of Origin
Abundant evidence now exists that newborn mammals are
behaviorally competent, albeit different from adults, soon after
birth (Alberts & Cramer, 1988). Neonatal rodents, which are
born in a very immature (altricial) condition, are capable of
recognizing and orienting toward the lactating mother (Blass,
1990), attaching to the nipple, engaging in organized suckling
behavior to extract milk, soliciting caregiving from the mother
as an essential element of the homeostatic regulation of water,
nutrients, and temperature, expressing cyclic periods of motor
activity and wakefulness, and learning important information
about littermates, contingencies within the nest environment,
and sensory cues that will serve to direct adult feeding, social,
and reproductive behavior. Even more extensive lists of behavioral competence can be produced for the newborn offspring
of other species, including humans. But the mere existence of
sensory and motor competence in the newborn raises an important question about the origins of behavior because it implies
that behavioral organization has its roots in the prenatal period.
Development emerges from interactions between systems
(Hofer, 1981; Kuo, 1967; Oyama, 1985). The essence of the
epigenetic view of development is that the organism and the various systems that it comprises exhibit a dynamic interplay with
the local environment. Changes in the environment, or in the
quality of interactions, can lead to unanticipated developmental
outcomes that are evident in the whole organism. These broad
principles of behavioral development are well recognized in the
postnatal period, during which one can readily visualize interactions between the infant and its caregivers and inanimate
environment. However, one of the uninformed reactions that
nearly all fetal researchers have experienced is the viewpoint,
implicitly held or explicitly stated, that events antecedent to
birth are driven by maturational processes, and that truly interactive behavioral development does not commence until after
birth. This viewpoint is well-illustrated by teratologieal studies,
which have identified chemical agents that are transmitted to
the fetus and that are often presumed to exert an impact at the
cellular or subeellular level to produce behavioral effects that
are evident after birth. If the roots of behavior extend into the
William P. Smotherman and Scott R. Robinson, Laboratory of Perinatal Neuroethology, Center for Developmental Psychobiology, Department of Psychology,State University of New York at Binghamton.
Research reported in this article was supported by National Institutes
of Health Grants HD 2823 l, HD 28014, and HD 16102.
Correspondence concerning this article should be addressed to William P. Smotherman, Laboratory of Perinatal Neuroethology, Center
for Developmental Psychobiology, Department of Psychology,EO. Box
6000, Binghamton University, Binghamton, New York 13902-6000.
425
426
SMOTHERMAN AND ROBINSON
prenatal period, then researchers must extend a broader, interactive perspective of behavioral development to encompass the
fetus and its intrauterine environment.
An example of how a broader behavioral perspective can contribute to understanding prenatal development is provided by
studies of the effects of fetal exposure to alcohol. Fetal alcohol
syndrome (FAS) and less severe fetal alcohol effects (FAE) are
well known consequences of maternal alcohol abuse during gestation (Abel, 1984). Consequences of fetal alcohol exposure include anatomical dysmorphologies (e.g., facial abnormalities
and lung hypoplasia) and behavioral disorders evident before
(altered behavioral state organization) and after birth
(hyperactivity, learning disabilities). The mechanisms of prenatal alcohol effects generally are assumed to involve basic embryological processes, such as gene expression, cellular physiology or intercellular communication. But the case can be made
that alcohol also alters the expression of behavior by the fetus,
and in so doing changes the interplay of factors that serve to
regulate conditions in utero and promote further behavioral development. For example, alcohol exposure is known to suppress
fetal motor activity for several hours (McLeod et al., 1983;
Smotherman et al., 1986). In turn, chronic reductions in movements of the mouth and limbs can produce effects on fetal morphological development. Fetal akinesia deformation sequence,
which results from experimental inhibition of fetal movement,
can result in microstomia and micrognathia, retarded lung development, skin and facial abnormalities, joint contractures
and altered bone growth, short umbilical cord, and long-term
movement disabilities (Moessinger, 1988). These effects occur
because the expression of movement stimulates the growth and
shaping of the oral cavity, lungs, skin, bones, and umbilical
cord. Alterations of fetal behavior in utero thus can serve as
the proximal mechanism for morphological and behavioral abnormalities. Viewed in this fashion, the ability of agents such as
alcohol to alter fetal behavior, and thereby produce lasting
effects on morphological and behavioral development, provides a new dimension to the term behavioral teratogen
(Smotherman & Robinson, 1987a).
Improvements in technology and procedures that provide direct access to the fetus in vivo (e.g., Smotherman & Robinson,
1991 ) are generating the impetus for prenatal developmental
research to move beyond the simple documentation that manipulations of pregnant animals can produce effects that are
evident in their offspring after birth. Rather, we see a growing
need for developmental researchers to focus attention on how
prenatal events affect the fetus, its behavior, and its relationship
with environmental conditions in utero. Investigation of behavioral potentials in the fetus will promote understanding of the
mechanisms of normal and abnormal development that lead
to predictable behavioral outcomes after birth. In other words,
behavioral study of the fetus will be necessary to understand
the origins of motor and sensory capabilities of infants and the
mechanisms of altered developmental outcomes.
The Fetus in Its Environment
Behavior is jointly determined by factors internal to the organism, such as the central nervous system (CNS) and effector
systems, and by physical conditions and sensory stimuli imping-
ing on the organism from the external environment. One of the
principal ways that the environment can influence the course of
development is by contributing to the regulation of behavior
expressed at different ages by the developing organism. This fact
is well recognized in postnatal animals; abundant evidence exists that early sensory or motor experiences Can have profound
impacts on adult behavior. However, the role of the environment in directing behavioral development is less well recognized
for the fetus.
Although one might visualize the fetus as a passive passenger
isolated from the vagaries of the outside world, the fetus does
not develop in a vacuum. It is surrounded by and interacts with
a complex environment. The intrauterine environment reflects
some aspects of the postnatal world while providing unique
physical constraints and sources of sensory stimulation for the
fetus (Smotherman & Robinson, 1988c). Because many fetal
researchers are realizing that development emerges from the interplay between the fetus and its intrauterine niche, a number
of laboratories have begun to characterize the stimuli that may
be available to the fetus in utero. Characteristics of the fetal
niche in different sensory modalities have been summarized in
several recent reviews (acoustic stimuli: Abrams, Gerhardt &
Peters, 1995; Fifer & Moon, 1988; Leeanuet & Granier-Deferre,
1995; mechanical stimuli: Ronca, Lamkin, & Alberts, 1993;
chemical stimuli: Robinson & Smotherman, 1991a; Schaal, Orgeur, & Rognon, 1995).
The most important elements of the intrauterine environment consist of the wall of the uterus, the extraembryonic membranes (amnion and chorion) that envelop the fetus, and amniotic fluid. The myometrium of the uterus is composed of
smooth muscle that provides an elastic restraint around the fetus that can suppress some aspects of motor activity while facilitating other forms of coordinated movement (Robinson &
Smotherman, 1987; Smotherman & Robinson, 1986). Rhythmic contractions of the uterine muscles before the onset of labor
also provide rhythmic pressure stimulation that appears to promote neurobehavioral development (Sadowsky, et al., 1992). In
addition to contributing to the physical restraint of the fetus,
the principal role of the extraembryonic membranes is maintenance of the fluid pool that surrounds the fetus. The amniotic
fluid provides free space necessary for limb and body movements; motor activity is suppressed under conditions ofreduced
amniotic fluid volume (oligohydramnios) or increased fluid
viscosity. Amniotic fluid also is regularly swallowed and inspired, contributing to the normal development of the oropharyngeal cavity, lungs, and digestive system (Moessinger, 1988 ).
Interactions between the fetus and amniotic fluid are bidirectional. One of the principal means by which the volume and
composition of amniotic fluid is regulated is by active fetal behavior (ingestion, breathing, micturition). Mechanical and
acoustic stimuli can be transmitted through the maternal abdomen with some attenuation (Abrams, Gerhardt, & Peters, 1995;
Ronca, Lamkin, & Alberts, 1993 ), providing the fetus with potential information about maternal behavior and voice (Fifer
& Moon, 1988). Chemical stimuli produced by the mother or
present in maternal diet can be passed to the fetus by transport
across the placenta, which is attached to the vascularized endometrial lining of the uterus, and thereby introduced into the
fetal circulatory system. Olfactory and gustatory stimuli found
SPECIAL SECTION: PRENATAL BEHAVIOR
in amniotic fluid that gain access to fetal sensory receptors directly by diffusion from capillaries in the olfactory epithelia,
can provide abundant information about maternal physiology,
individual chemical signatures, and diet selection before birth
(Hopper, 1988; Maruniak, Silver, & Moulton, 1983).
The environment contributes to the regulation of behavior in
two principal ways: (a) it is a source of sensory stimuli, some of
which can elicit specific behavioral responses, and (b) it provides a physical context in which behavior occurs. Both aspect s
of environmental influence have been demonstrated in the expression of behavior by the fetus. For example, rodent fetuses
exhibit stereotypical motor responses following infusion of
novel chemosensory fluids into the mouth. A small volume of a
solution prepared from lemon odor extract reliably evokes facial wiping in the rat fetus on the last 2 days of gestation (E20E21 ). This behavior involves a series of 5-10 forelimb strokes
in which the paw slides from ear to nose along the sides of the
face (Robinson & Smotherman, 1991 b). Overhead forelimb
strokes are expressed by juvenile and adult rodents in response
to aversive taste-odor stimulation and during spontaneous
grooming behavior (Berridge & Fentress, 1986; Johanson &
Shapiro, 1986). The observation that the wiping response is expressed by fetal rats on E20 of gestation, but not on El9
(Smotherman & Robinson, 1989), suggests the sudden prenatal
emergence of this organized pattern of behavior. However, experimental manipulation of conditions at the time of behavioral
assessment has revealed that younger fetuses are capable of expressing the facial wiping response as well. In a typical testing
situation, the fetal subject is exteriorized from the uterus and
amniotic sac (ex utero) and is freely suspended within a supportive fluid environment (warm, buffered saline bath). Because it remains connected through the umbilical cord to the
placenta, which is attached to the uterus, the fetal subject remains healthy in this exteriorized position, providing clear visual access for the experimenter to assess fetal behavior
(Smotherman & Robinson, 1991 ). E20 fetuses tested ex utero
consistently exhibit the wiping response to lemon, but El9 fetuses do not. If fetal subjects are allowed to remain within the
intact amniotic membranes (in amnion), facial wiping is expressed at the earlier age (Robinson & Smotherman, 1991b).
Detailed analysis of movements associated with fetal responses
to lemon in amnion and ex utero has suggested that the amniotic membranes provide structural support for the head and
forelimbs that reduces head activity and facilitates paw-face
contact. The amnion thus serves as a kind of external scaffolding that promotes the expression of organized behavior. The influence of the amniotic sac on the expression of facial wiping
behavior in the fetal rat provides a clear example of the dual
roles of the environment in the control of behavior during the
prenatal period. Appreciation of the role of the intrauterine environment as a codeterminant of fetal behavior can provide a
source of novel hypotheses concerning the origin, expression
and regulation of behavior during the prenatal period.
E x a m i n i n g Age-Related Change in the Absence
o f Experience
Postnatal studies of behavioral development are replete with
examples of age-related differences. Typically, the same experi-
427
mental treatment administered at different ages is reported to
produce behavioral effects in younger subjects that are less consistent, more poorly organized, evoked by a narrower (or
broader) array of stimuli, or otherwise less well-developed than
the effects produced in older subjects. The discovery of age-related differences sometimes is taken as evidence of behavioral
maturation, sometimes as indicative of the influence of experience on emerging behavior. Deprivation or enrichment experiments can extend these findings by specifically implying a role
for sensory experience in behavioral development. However, the
complexity of the postnatal environment in general, and the dependence of infant mammals on maternal care in particular,
make it very difficult to disentangle accruing experience from
other age-related changes during development.
In contrast to the postnatal period, the needs of the fetus are
met through maternal physiology, not maternal behavior, and
the fetus is relatively buffered from the moment-to-moment
changes that typify life outside the womb (Smotherman & Robinson, 1994a). As a result, it is possible to present discrete
forms of sensory stimulation to the fetus with certainty that the
fetus has never before experienced similar stimuli. By applying
these sensory manipulations to different subjects across a range
of ages, it is possible to estimate changes in the expression of
behavior that occur in the absence of explicit experience. An
example of this research approach is provided by fetal responses to stimuli typically associated with the postnatal suckling context. Fetal rats exhibit distinctive oral responses upon
exposure to a soft artificial nipple shaped to approximate the
dimensions of the nipples of a lactating rat (Robinson et al.,
1992). When the artificial nipple is gently held in contact with
the mouth, fetuses exhibit mouth activity, licking, and forelimb
treadling (repeated forelimb extensions that do not contact the
face or nipple). Younger fetuses (E 18) show lateral head movements coupled with rhythmic mouthing that eventually results
in passive oral capture of the nipple, followed by cessation of
head movements (Figure l ). Older fetuses (E 19-E21 ) show an
active oral grasping response in which the tip of the nipple is
seized and force (both sucking and biting pressure) is exerted
on the nipple (Figure 2 ). It is important to note that fetuses at
all ages exhibit responses upon their first experience with the
artificial nipple. Although experience is held constant, evidence
suggests that the ability to seize the nipple during the oral grasp
response improves with age: The percentage of mouthing responses that are followed by successful grasping of the nipple
increases across ages E 19-E2 I. Organized behavioral responses
to the artificial nipple are a clear indication of the prenatal development of behavior patterns that are important to the neonate. Moreover, the fact that newborn rats are not responsive to
surrogate nipples when tested soon after birth (PO) underscores the use of fetal study in understanding the origins of
postnatal ingestion and identifying rapid changes in the control
mechanisms that subserve suckling behavior.
Fetal responses to other biologically relevant stimuli, such as
milk, provide an additional example of age-related changes in
responsiveness in the absence of experience. During normal
suckling behavior at the lactating nipple, infant rats exhibit a
predictable sequence of responses associated with the ingestion
of milk (Brake, Shair, & Hofer, 1988 ). After initial orientation
to the maternal ventrum and attachment to the nipple, the pup
428
SMOTHERMAN AND ROBINSON
• Grasping
171 Oral Capture
[] No Response
100"
80"
tO
tl)
60
CO
40
20
0
E17
E18
E19
E20
E21
PO
Figure 1. Development of oral grasping of an artificial nipple in the rat fetus and neonate. On E 18, oral
contact with a soft vinyl nipple elicits lateral head movements and mouthing that results in passive oral
capture of the nipple (left). Over the next 3 days, fetuses exhibit an active oral grasping response to the
nipple. The responsiveness of fetal rats to the artificial nipple stands in contrast to the lack of response
evident in newborns (PO) before suckling experience.
enters a quiet or active sleep state and engages in rhythmic
mouthing activity (sucking) that stimulates oxytocin release in
the lactating mother, triggering milk letdown. Once milk becomes available in a reservoir behind the nipple (associated
with nipple distension), the pup exhibits a stereotypic stretch
response that involves elongation of the body trunk and extension of the rearlimbs (Drewett, Statham, & Wakedey, 1974).
The stretch response is thought to facilitate extraction of milk
from the reservoir, after which the pup typically awakens and
disengages from the nipple. After a short period of elevated motor activity (behavioral activation), the pup reattaches to a nipple, often shifting to a new location on the maternal ventrum,
and reinitiates the suckling sequence (Hall, 1990).
Many of the elements of neonatal responses to milk in the
suckling context are expressed by fetal rats following intraoral
infusion of milk (Robinson & Smotherman, 1992b). Infusion
elicits an initial bout of mouthing activity that is followed over
the course of several minutes by elevated rearlimb activity.
These changes in motor behavior are accompanied by alterations in fetal responsiveness to other forms of sensory stimulation, such as application o f a stiffbristle (perioral probe) to the
vibrissal area near the mouth (Figure 3 ). Unmanipulated fetal
subjects express a unilateral facial wiping response to the perioral probe; after milk, however, the probe is ineffective in evoking the wiping response (Smotherman & Robinson, 1992a). Ultimately, behavioral changes initiated by milk culminate in the
expression of a fetal stretch response that closely resembles the
behavior of newborns after milk letdown. This sequence of behavioral events that follows infusion of milk is not expressed to
other chemosensory fluids, some of which include constituents
of milk. Solutions of simple sugars (lactose or sucrose) and fats
(corn oil) have been reported to produce milk-like behavioral
effects in infant rats (Smotherman & Robinson, 1992c), yet do
not trigger comparable responses in the fetus, which lacks experience with the constitutents of milk and other stimuli associated with suckling. Infant formulas (milk or soy-based) also are
ineffective in promoting rearlimb activity, reduced responsiveness to a perioral probe, and the stretch response in the rat fetus
(Smotherman & Robinson, 1992b). Although milk is ingested
by the fetus after infusion, experimental blockade of the esophagus, which prevents swallowing, has confirmed that the prosensory qualities of milk are sufficient to elicit mouthing and
rearlimb activity, reduced responsiveness, and the stretch
(Robinson & Smotherman, 1994). The unique configuration
of milk's orosensory qualities are recognized by the fetus upon
its first exposure to this biologically important fluid.
The complete sequence of behavioral effects is elicited by
milk infusion in the E21 fetus. Tracing the different response
elements of this sequence earlier in gestation has revealed that
milk evokes fewer and less pronounced behavioral components
at younger ages. For example, the E 19 fetus shows little mouthing activity, only modest increases in rearlimb activity, and no
evidence of the stretch response (Andersen, Robinson, &
Smotherman, 1993). Developmental study of fetal responses to
the artificial nipple and milk has suggested that the behavioral
sequence associated with suckling in the newborn is composed
of simpler response elements that are governed by different control mechanisms and follow independent developmental trajectories. The view emerging from this and related studies of other
patterns of fetal behavior (Robinson & Smotherman, 199 lb) is
that organized patterns of behavior are assembled from simple
precursors, and that the rules governing this assembly may be
quite different than the basic developmental processes that give
rise to the elements themselves.
Cryptopsychobiology: Revealing Behavioral Potentials
The foregoing examples of fetal behavior would seem to imply a steady improvement in behavioral performance with advancing age. The notion of uniform progress and increasing
complexity is a common preconception of the process of behavioral development that may be imposed on the data by the observer as often as accurately reflecting patterns of age-related
change (Smotherman & Robinson, 1995). Not all aspects of
behavior exhibit a continual increase in organization with age.
In fact, a growing list ofexamplesfrom both the nonhuman and
human literatures suggests that some patterns of behavior may
be expressed only for a transient period during early life (so-
SPECIAL SECTION: PRENATAL BEHAVIOR
429
t-- 100
80
O
60
~,
40
Cj)
20
=!
ol
i
O
zi
!
o
NT
SAL SAL NALCTOP BNI
Figure 3. Percentage of E20 fetal subjects that exhibit a facial wiping
response in a bioassay of perioral cutaneous responsiveness administered 1 min after intraoral infusion of milk (black bars). The bioassay
consists of application of a stiff von Frey bristle (the perioral probe) to
the lateral vibrissal pad. Most fetuses that receive no infusion or drug
treatment (NT) or a control infusion and ip injection of saline (SAL)
exhibit facial wiping to this stimulus. However, few SAL-injected Subjects exposed to milk express the wiping response. Administration of a
nonselective opioid antagonist such as naloxone (NAL) or the kappa
opioid antagonist nor-binaltorphimine (BNI) block the effect of milk
and reinstate facial wiping to the probe, whereas the selective mu antagonist (CTOP) has no influence on the milk effect. These data indicate
that the fetus's first experience with milk results in activation of the
endogenous kappa opioid system, which reduces fetal responsiveness in
the bioassay.
Figure 2. Photograph of a rat fetus on E20 of gestation exhibiting oral
grasping of an artificial nipple. To gain access to fetal subjects for experimental study, the pregnant rat is prepared with a spinal anesthetic,
permitting exteriorization of the uterus and fetuses into a buffered saline bath maintained at body temperature (37.5 *C). After delivery
from the uterus and amniotic sac, with the fetus submerged within the
bath (preserving the umbilical connection to the placenta, which remains attached inside the uterus), controlled presentation of various
sensory stimuli, such as the soft vinyl nipple, can be used to measure
fetal responsiveness and expression of organized behavior before birth.
called ontogenetic adaptations), whereas other patterns exhibit
a discontinuous developmental trace, disappearing for an intermediate period and reappearing at a later age. The transient
disappearance and later reappearance of behaviors such as stepping, reaching, and grasping in the human infant have provoked
a variety of explanations, ranging from refutation of any relationship between neonatal reflexes and later behavior to dynamical models that emphasize underlying continuity in form
and function (Tbelen, 1984, 1989 ). A biomechanical explanation that may apply t o many examples of retrogressive changes
in development is the constraining or permissive influence of
the environment that exerts differential impacts on behavioral
expression across ages.
The role of environmental conditions in facilitating or suppressing the expression of behavior is illustrated by the postnatal development of facial wiping responses to aversive taste-odor
stimuli in rat pups (Smotherman & Robinson, 1989). As described above, the fetus is capable of expressing facial wiping on
the last 2-3 days of gestation. Infant rats that are placed on the
floor of a testing chamber and exposed to the same lemon stimulus only a few hours after birth can express facial wiping, but
do so only after a brief period of behavioral activation that resuits in displacement from a prone posture. The wiping response disappears over the next few days of postnatal life and
is not expressed consistently until 11-13 days after birth. The
performance of facial wiping during the postnatal period thus
follows a presence-absence-presence pattern that is contrary to
conventional views of progressive improvement during development. The early disappearance of the wiping response corresponds to the emergence of contact righting in the rat, which is
reflected in the infant turning to a prone posture when in contact with a solid substrate (PeUis, Pellis, & Teitelbaum, 1991 ).
Maintenance of a prone posture in infant rats--the goal of
righting--necessitates that all four legs remain in contact with
the ground to provide postural support. But if the forelimbs are
freed from this support function, they become available for
other motor responses, including facial wiping. This behavioral
conflict hypothesis has been confirmed by immersing infant rats
in a buoyant fluid medium to neck depth, thereby eliminating
paw contact with the ground. In this suspended posture, 1- to 3day-old rat pups consistently express facial wiping responses to
a lemon infusion (Smotherman & Robinson, 1989 ). The pup's
responsiveness to contact with a substrate is the proximal mechanism responsible for the interaction between righting and fa-
430
SMOTHERMAN AND ROBINSON
cial wiping behavior. Evidence from the rat fetus suggests that
substrate responses emerge during the prenatal period. Placement of the forepaws of an E21 rat fetus on an angled substrate,
while remaining submerged within the fluid environment of the
saline bath, reduces the performance of facial wiping in response to lemon infusion. However, testing the fetus on a submerged substrate 1 day earlier (E20) has no influence on facial
wiping behavior. These findings from fetal and neonatal rodent
subjects confirm that patterns of developmental change are dependent upon interactions between the developing organism
and its immediate environment. Assessment of behavioral performance at any given age must take into account the physical
and sensory conditions present at the time of testing.
Presentation of stimuli typical of other periods of development (e.g., artificial nipple or milk delivery to the fetus) or manipulation of environmental context (e.g., immersion of the neonate in a buoyant fluid medium) are two methods that can
reveal behavioral abilities that may be hidden at various times
during development (for other examples see: Bekoff & Kauer,
1984; Hall & Williams, 1983; Pfister, Cramer, & Blass, 1986;
Smotherman & Robinson, 1989). In a functional sense, environmental manipulations that promote the expression of behavior function in a manner analogous to the amniotic membranes in the study of facial wiping in the El9 fetus, described
above: They provide supportive scaffolding that permits the expression of organized action patterns. The function of scaffolding is to provide a temporary replacement for missing elements
in a structure before the complete architecture is in place. In
terms of behavioral development, this may be accomplished by
exposing animals to stimuli normally absent at a particular age,
or by releasing the animal from environmental constraints, or
by mimicking the effects of stimuli on underlying neural systems. The latter approach recently has been applied to investigate the ontogeny of fetal responses to milk.
The cascade of behavioral effects that follows milk infusion is
dependent, in large part, on milk-induced activity in the endogenous opioid system of the fetus (Smotherman & Robinson,
1992c). Delivery of milk to the E21 fetus triggers a brief period
of activity at the kappa class of opioid receptors, which in turn
promotes changes in rearlimb activity, sensory responsiveness,
and the stretch response (Figure 3). Administration of antagonist drugs that block opioid activity at all receptors (naloxone)
or selectively within the kappa system (nor-binaltorphimine)
eliminates opioid activity and blocks the effects of milk on motor and sensory behavior (Smotherman & Robinson, 1992a,
1992d). Conversely, administration of agonist drugs that
mimic activity at kappa opioid receptors (such as U50, 488)
can promote many of the effects of milk, resulting in elevated
readimb activity and reduced responsiveness to perioral stimuli. Kappa opioid manipulations also alter the responsiveness
of fetal rats to chemosensory stimuli other than milk, which
ordinarily are ineffective in eliciting suckling behavior. Fetuses
treated with a kappa agonist that receive an intraoral infusion
of lemon extract (which ordinarily evokes the facial wiping
aversion response) promote expression of the stretch response.
The finding that kappa opioid activity is a natural concomitant
of exposure to milk suggests that the inability of younger fetuses
(E 19) to express suckling responses to milk may be due to the
inability of milk to evoke activity in the kappa system. This
hypothesis has been examined by administering the kappa agonist and infusing nonmilk fluids to fetuses across a range ofgestational ages. On E 19, kappa opioid activity promotes some behavioral elements associated with the sequence of events leading
to the stretch (e.g., rearlimb activity and postural extension of
the body trunk), but has no effect on other components (e.g.,
mouthing and expression of the stretch, Andersen, Robinson, &
Smotherman, 1993). Data such as these are contributing to the
view that even simple action patterns, such as the fetal stretch
response, are the product of multiple behavioral systems that
are regulated by independent mechanisms, which are coupled
over developmental time. In broader terms, experiments such
as these imply that the behavioral performance of a subject provides an underestimate of its behavioral potential at any given
age. Given the appropriate eliciting stimuli and supportive environment, young animals can express behaviors that ordinarily
are performed only during restricted periods earlier or later in
development.
Plasticity in Early Development: R a p i d Sculpting
o f Responses
Because it is possible to measure the response of a fetus upon
its first exposure to a particular sensory stimulus, the experimenter also can measure the role of experience in shaping responses to subsequent exposures. Various learning paradigms
have been used to demonstrate that fetuses are capable of acquiring and expressing habituation and associative learning in
utero (Robinson & Smotherman, 1995; Smotherman & Robinson, 1987b). For example, pairing a novel taste-odor stimulus
(mint) with an aversive consequence (e.g., chemically induced
illness produced by ip injection of lithium chloride) results in
a conditioned aversion upon subsequent exposure to the mint
stimulus. Conditioned changes in fetal behavior have been demonstrated in fetal rats as early as E 17 of gestation (Smotherman
& Robinson, 1985 ). Studies such as these confirm that fetuses
can detect stimuli that are introduced into the intrauterine environment, form associations based on the contingencies between CS and US, and retain these associations into the postnatal period. Comparable studies of associative learning have not
been conducted with human subjects, although abundant evidence suggests that human fetal exposure to patterned sound
stimuli, such as maternal voice, can affect acoustic preferences
of the neonate. These findings suggest that learning has the potential for contributing to behavioral development during the
prenatal and neonatal periods.
Virtually all of the experiences of the newborn occur in some
relation to milk obtained at the nipple. Newborn rats typically
experience milk during suckling 1-2 times per hour. Because
some of the behavioral effects of milk are protracted, persisting
for periods of 30-60 rain (Robinson & Smotherman, 1992c),
behavior is continually shaped by the pup having just been exposed to milk or in the process of obtaining milk. These facts
suggest that milk may serve as a focal event that mediates sensory experiences in the suckling and nest environments
(Smotherman & Robinson, 1994b). This working model of
learning in the newborn is supported by experiments involving
contingent presentations of milk and other sensory stimuli to
the rat fetus (Arnold, Robinson, Spear, & Smotherman, 1993;
SPECIAL SECTION: PRENATAL BEHAVIOR
100"
Bioassay
T1
T2
T3
() . . . .
5 . . . . 10 . . . .
- "
CS = nipple
,,
x
US=milk
~' CS US
T3:. " ...................
0
15
1~5. . . .
Response in B i o a s s a y
m 80"=-
CS
~
2"0m'in
O
~
,\\\
60"
40"
x
x
~io 4"5 gos
431
o~ 20"
0
ti
~
o
o
¢.o
09
SAL N A L C T O P BNI
rlt~
Figure 4. Experimental timeline for single-session classical conditioning in the rat fetus (left). The exper-
imental session consists of four phases: (a) a series of conditioning trials (T I-T3) in which the artificial
nipple is presented as the conditioned stimulus (CS) and intraoral infusion of milk is delivered as the
unconditioned stimulus (US), (b) a 9.5-rain delay, (c) reexposure to the nipple CS, and (d) measurement
of the conditioned response of the fetus in a behavioral bioassay of perioral cutaneous responsiveness.
Reexposure to the nipple CS in subjects receiving paired presentations ofCS + US (black bars) promotes
endogenous opioid activity that is measured as reduced cutaneous responsiveness in the bioassay (right).
Treatment with naloxone (NAL) or the mu antagonist CTOP blocks the expression of conditioned changes
in responsiveness, whereas control injection of saline (SAL) or the kappa antagonist nor-binaltorphimine
(BNI) does not interfere with conditioned reduction of facial wiping in the bioassay.
Robinson, Arnold, Spear, & Smotherman, 1993). Fetuses that
receive a series of three paired presentations o f a CS (an artificial nipple) followed by an US (intraoral infusion of milk) exhibit altered behavioral responses when the CS is later presented
alone. Specifically, fetuses are less responsive to a perioral cutaneous stimulus 1 min after reexposure to the nipple CS (i.e.,
they show reduced facial wiping in a bioassay where a perioral
probe is applied to the lateral vibrissal pad of the fetus, Figure
4). Reduced responsiveness is evident only in fetuses exposed
to paired presentations of the CS and US, and is not observed
in control subjects presented with the CS alone, US alone, or
US and CS separated by intervals of several minutes during
conditioning trials. This pattern of results indicates that contingent exposure to milk and the artificial nipple results in classical
conditioning of sensory responsiveness in the fetal rat.
Further experiments have demonstrated that the conditioned
change in responsiveness following reexposure to the nipple CS
is dependent upon endogenous opioid activity. Blockade of opioid receptors with naloxone (NAL) after the series of conditioning trials, but before reexposure to the nipple CS, eliminates the conditioned change in response: NAL-treated subjects
show high levels of facial wiping that do not differ from controls.
Administration of selective antagonists of opioid receptors has
produced a paradoxical finding. Milk is known to promote activity in the kappa opioid system when presented to the fetus,
and the artificial nipple has no effect on opioid activity apart
from being paired with milk. But reexposure to the nipple CS
following pairings with milk results in conditioned activity in
the mu opioid system, with no evidence of kappa activity.
Therefore, contingent presentations of the artificial nipple and
milk result in activation of a different neurochemical system
than is activated by either the US or CS before conditioning.
Because control fetuses that have equal exposure to both the
nipple and milk fail to exhibit conditioned changes in opioid
activity, the temporal relationship between stimuli during the
fetus's first experiences with milk and the nipple dictates subsequent behavioral and neurochemical consequences. Additional
experiments recently have extended these findings by confirming that conditioned opioid activity can be triggered by the nipple CS after only a single pairing with milk.
In light of other studies of learning during the perinatal period, it may be unremarkable that fetal rats can quickly learn
an association between a stimulus that mimics an important
feature of the postnatal suckling environment (the nipple) and
another stimulus that has important consequences for neonatal
biology (milk). An unexpected finding to emerge from research
on fetal learning is that classical conditioning also affects fetal
behavioral and neurochemical responses to the US, milk
( Robinson & Smotherman, 1994). Fetal subjects that receive a
series of three nipple-milk pairings continue to exhibit increased opioid activity upon a fourth infusion of milk. However,
it is necessary to block activity in both the kappa and mu opioid
systems to counter the behavioral effects of milk after conditioning and restore fetal responsiveness to perioral stimulation.
This finding indicates that milk, after its contingent presentation with the artificial nipple, comes to evoke activity in both
the mu and kappa opioid systems. Control fetuses that are exposed to unpaired presentations of milk and the artificial nipple
continue to show activity only in the kappa opioid system, much
like initial fetal responses evoked by milk. The implication of
conditioned changes in opioid responses evoked by milk and
the artificial nipple is that the neurochemical consequences of
432
SMOTHERMAN AND ROBINSON
experience with milk and other stimuli encountered in the suckling situation are likely to be altered permanently after the first
exposure to milk. At both a behavioral and more mechanistic
level of analysis, initial fetal responses to milk are governed by
identifiable rules. But these rules undergo rapid reorganization
when the fetus experiences milk in a particular configuration
with other stimuli associated with suckling. Behavioral potentials that are present at the end of gestation are rapidly reconfigured after birth through the newborn's interactions with the
environment, providing an experimental example of the oftenmentioned plasticity of behavior and the CNS during perinatal
development.
T h e P r e m a t u r e Infant: N o t a Fetus, N o t a N e w b o r n
Study of fetal behavior sometimes is justified by reference to
health concerns encountered by clinicians in obstetrical, perinatal, or pediatric practice (e.g., Nijhuis, 1992). For example,
preterm infants represent a large population of infants at risk
who require medical intervention for their survival and healthy
development, usually in the setting of the neonatal intensive
care unit (NICU). Technological and procedural advances in
perinatology now are promoting the survival of infants born
before 24-26 weeks gestational age and weighing only 500-750
g. After delivery, these infants are placed in a highly artificial
environment within the NICU, where they are exposed to abnormal light-dark cycles, noise, intermittent and unpredictable
noxious stimulation, artifcial modes of feeding (e.g., gavage
and parenteral), and separation from typical patterns of caregiving (Field, 1990). Experiences in the NICU are likely to have
far-reaching consequences for the development of the infant.
Even when age is calculated from conception, allowing for
differences in maturity at birth, preterm infants typically exhibit behavioral difficulties that include disrupted behavioral
state organization, altered responsiveness to patterned stimulation, and problems associated with feeding. The clinician is
faced with the dual tasks of ensuring the preterm infant's survival while simultaneously promoting growth, development,
and eventual adaptation to the environment outside the NICU.
Thus, the goal of the clinician is to merge the atypical conditions
faced by the preterm infant with the age-typical developmental
trajectory of a full-term infant. Clinicians differ in their opinions of whether these preterm infants should be viewed as a fetus removed from its age-typical intrauterine environment or a
newborn only partially equipped to cope with the demands of
extrauterine life.
How preterm infants are viewed has ramifications for the
treatments they receive within the NICU. Some current strategies for managing the health and well-being of preterm infants
are based on alternative views that the preterm infant is best
treated like a full-term newborn or like a fetus. Intervention
treatments that are designed to promote normal neonatelike
responding are in widespread practice, such as intensive human
handling (Schanberg & Field, 1987) or nonspecific sensory
stimulation. Other approaches involve attempts to imitate features of the prenatal environment, such as floating infants on
temperature-regulated, undulating water mattresses (Korner,
1980), or to provide patterned sensory stimulation that is appropriate for the neurobehavioral abilities of the infant (Als et
al., 1986). The latter strategies have become practical only with
advances in our understanding of the behavioral capacities and
developmental requirements of the fetus in utero. Recognition
of fetal behavioral potentials can promote recognition of the
special needs and abilities of preterm and other high-risk neonates and serve as a heuristic, suggesting novel approaches to
preterm care.
Because the environment of the preterm infant differs in almost all respects from the full-term fetus or newborn, and because behavioral development reflects the dynamic interplay between organism and environment before birth as well as after
birth, we should not expect the preterm infant to be the same
as a fetus or a full-term newborn. Although fetal research has
suggested new ways of caring for preterm infants, many of the
practices in the NICU ignore the fundamentally interactive nature of behavioral development. One of the basic implications
of fetal research conducted with animal subjects is that perinatal development emerges from the interactions of an organism
with unique behavioral potentials with its age-typical environment. Because the preterm infant differs in neurobehavioral potential from the full-term neonate, and yet resides in an environment strikingly different from that of the fetus, it may be
most appropriate to consider the preterm infant as neither a
fetus nor a neonate. Adoption of this viewpoint may suggest alternative, nonintuitive approaches to the management of preterm infants that are based on understanding the mechanisms
underlying developmental change. Because developmental trajectories are relative and are contingent on conditions of measuring performance, it may be inappropriate to evaluate the
progress of such infants by comparison to developmental milestones based on either fetal or neonatal standards of physical
growth or behavioral performance, as is routinely done in pediatric practice. Ideally, treatment strategies should be tailored to
the developmental capacities of the preterm infant while retaining critical interactions that promote adaptation to the postnatal environment.
Perinatal responses to milk and other suckling stimuli suggest
an example of the potential consequences of age-atypical experiences within the NICU (Smotherman & Robinson, 1992c).
Experiments with fetal rats point to early feeding interactions
as a proximal mechanism for promoting behavioral and neural
development in the newborn. Exposure to milk appears to be a
critically important event for the newborn, and the context in
which this fluid is first experienced may alter trajectories for
development of the nervous system and behavior (Robinson &
Smotherman, 1994). The importance of milk and early feeding
interactions may be most evident in premature infants who
must be sustained on artificial nutritional regimens. Gavage
feeding or intravenous infusion bypass the sensory systems normally engaged during nursing and prevent expression of neural
and behavioral responses to milk. The findings obtained from
animal fetuses suggest that the sensory contexts in which milk
is first experienced, or the neurochemical systems that are activated during early feeding, may result in different patterns of
behavioral and neural development (Smotherman & Robinson,
1994b). It is unknown whether gavage or parenteral feeding engage orosensory systems normally active during feeding, and if
not, whether they result in different patterns of response when
more normal feeding is initiated. Providing more naturalistic
SPECIAL SECTION: PRENATAL BEHAVIOR
experiences during feeding situations in the NICU may help
to diminish the long-term consequences of preterm birth on
ingestive behavior when age-typical patterns of food presentation are initiated after release from the NICU.
The u n i q u e circumstances of premature birth highlight the
importance of an integrated approach to the study of behavioral
development. Premature birth represents a kind of natural experiment that can augment study of normal development in the
fetus and full-term neonate. At each age, the developing organism exhibits behavioral potentials that are both the means and
the product of interactions with the immediate environment.
Experimental study of behavioral development, which is best
exemplified by research on n o n h u m a n animals, uses artificial
manipulations o f internal states, neural functions, or sensory
events to challenge the young organism. Study of behavior in
the fetus can contribute to a broader understanding of the developmental process not only by focusing on a unique period in
the life of all mammals, but also by exposing the fetal subject to
unusual environments and behavioral tasks, thereby yielding
information about the lability of early development. The challenge for future research in our laboratory will be to use focal
investigation of the fetal rodent to extract general rules that direct and give shape to behavioral development, and to apply
these principles to broader developmental questions concerned
with other periods of the lifespan or other species, including the
human.
References
Abel, E. L. (1984). Prenatal effects of alcohol. Drug andAlcohol Dependence, 14, 1-10.
Abrams, R. M., Gerhardt, K. J., & Peters, A. J. M. (1995). Transmission of sound and vibration to the fetus. In J. P. Lecanuet, N. A.
Krasnegor, W. P. Fifer, & W. P. Smotherman, (Eds.), Fetal development: A psychobiological perspective (pp. 315-330). Hillsdale, NJ:
Erlbaum.
Alberts, J. R., & Cramer, C. P. (1988) Ecology and experience: sources
of means and meaning of developmental change. In E. M. Blass (Ed),
Handbook of behavioral neurobiology Vol. 9 (pp. 1-39). New York:
Plenum.
Als, H., Lawhon, G., Brown, E., Gibes, R., Duffy, E H., McAnulty, G.,
& Biickman, J. G. ( 1986 ). Individualized behavioral and environmental care for the very low birth weight preterm infant at high risk
for bronchopulmonary dysplasia: Neonatal intensive care unit and
developmental outcome. Pediatrics, 78, 1123-1132.
Andersen, S. L., Robinson, S. R., & Smotherman, W. P. (1993), Ontogcny of the stretch response in the rat fetus: kappa opioid involvement.
Behavioral Neuroscience, 107, 370-376.
Arnold, H. M., Robinson, S. R., Spear, N. E., & Smotherman, W. P.
(1993). Conditioned opioid activity in the rat fetus. Behavioral Neuroscience, 107, 963-969.
Bekoff, A., & Kauer, J. A. ( 1984): Neural control of hatching: fate of
the pattern generator for the leg movements of hatching in posthatching chicks. Journal of Neuroscience, 11, 2659-2666.
Berridge, K. C., & Fentress, J. C. (1986). Contextual control oftrigeminal sensorimotor function, Journal ofNeuroscience, 6, 325-330.
Blass, E. M. (1990). Suckling: Determinants, changes, mechanisms,
and lasting impressions. Developmental Psychology, 26, 520-533.
Brake S. C., Shair, H., & Hofer, M. A. (1988). Exploiting the nursing
niche: The infant's sucking and feeding in the context of the motherinfant interaction. In E. M. Blass (Ed.), Handbook of behavioral neurobiology: vol. 9. Developmental psychobiologyand behavioral ecology (pp. 347-388 ). New York: Plenum Press.
433
DrewetL R. E, Statham, C., & Wakerley, J. B. (1974). A quantitative
analysis of the feeding behaviour of suckling rats. Animal Behaviour,
22, 907-913.
Field, T. M. (1990). Neonatal stress and coping in intensive care. Infant
Mental Health Journal. 11, 57-65.
Fifer, W. P., & Moon, C. (1988). Auditory experience in the fetus. In
W. P. Smotherman & S. R. Robinson (Eds.), Behavior of the fetus,
(pp. 175-188 ). Caldwell, N J: Telford Press.
Hall, W. G. (1990). The ontogeny of ingestive behavior. In E. M.
Stricker, (Ed.), Handbook of behavioral neurobiology: vol. 10. Neurobiology of food and fluid intake (pp. 77-123 ). New York: Plenum
Press,
Hall, W. G., & Williams, C. L. ( 1983 ). Suckling isn't feeding, or is it? A
search for developmental continuities. In J. S. Rosenblatt, R. A.
Hinde, & C. Beer, & M.-C. Busnell, (Eds.), Advances in the study of
behavior vol. 13 (pp. 219-254). New York: Academic Press.
Hopper, P. G. ( 1988 ). Adaptive fetal learning: prenatal exposure to garlic affects postnatal preferences. Animal Behaviour, 36, 935-936.
Hofer, M. A. ( 1981 ). The roots of human behavior New York: W. H.
Freeman.
Johanson, I. B., & Shapiro, E. G. (1986). Intake and behavioral responsiveness to taste stimuli in infant rats from 1 to 15 days of age. Developmental Psychobiology, 19, 593-606.
Korner, A. E (1980). Maternal deprivation: compensatory stimulation
for the prematurely born infant. In R. W. Bell & W. P. Smotherman
(Eds.), Maternal influences and early behavior (pp. 337-352). New
York: Spectrum Press.
Kuo, Z.-Y. (1967). The dynamics of behavior development: An epigenetic view, New York: Random House.
Lecanuet, J.-P., & Granier-Deferre, C. (1995). Human fetal auditory
perception. In J. P. Lecanuet, N. A. Krasnegor, W. P. Fifer, & W. P.
Smotherman (Eds.), Fetal development: A psychobiological perspective (pp. 239-262). New York: Erlbaum.
Maruniak, J. A., Silver, W. L., & Moulton, D. G. (1983) Olfactory receptors respond to blood-borne odorants. Brain Research, 265, 312316.
McLeod, W., Brien, J., Loomis, C., Carmichael, L., Probert, C., & Patrick, J. (1983). Effect of maternal ethanol ingestion on fetal breathing
movements, gross body movements, and heart rate at 37 to 40 weeks'
gestational age. American Journal of Obstetrics and Gynecology, 145,
251-257.
Moessinger, A. C. (1988). Morphological consequences of depressed or
impaired fetal activity. In W. P. Smotherman & S. R. Robinson
(Eds.), Behavior of the fetus (pp. 163-173). Caldwell, NJ: Telford
Press.
Nijhuis, J. G. (Ed.). ( 1992 ). Fetal behavior: Developmental and perinatal aspects. Oxford, UK: Oxford University Press.
Oppenheim, R. W. (1982). The neuroembryological study of behavior:
progress, problems, perspectives. Current topics in developmental biology, 17, 257-309.
Oppenheim, R. W. (1992). Pathways in the emergence of developmental neuroethology: antecedents to current views of neurobehavioral
ontogeny. Journal of Neurobiology, 23, 1370-1403.
Oyama, S. ( 1985 ). The ontogeny of information. Cambridge, UK. Cambridge University Press,
Pellis, V. C., Pellis, S. M., & Teitelbaum, P. ( 1991 ). A descriptive analysis of the postnatal development of contact-righting in rats (Rattus
norvegicus ). Developmental Psychobiology, 24, 237-263.
Pfister, J. E, Cramer, C. E, & Blass, E. M. (1986). Suckling in rats extended by continuous living with dams and their preweanling litters.
Animal Behavior, 34, 415-420.
Robinson, S. R., Arnold, H. M., Spear, N. E., & Smotherman, W. E
(1993). Experience with milk and an artificial nipple promotes con-
434
SMOTHERMAN AND ROBINSON
ditioned opioid activity in the rat fetus. DevelopmentalPsychobiology,
26, 375-388.
Robinson, S. R., Hoeltzel, T. C. M., Cooke, K. M., Umphress, S. M.,
Murrish, D. E., & Smotherman, W. P. (1992). Oral capture and
grasping of an artificial nipple by rat fetuses. Developmental Psychobiology, 25, 543-555.
Robinson, S. R., & Smotherman, W. P. ( 1987 ) Environmental determinants of behaviour in the rat fetus II: The emergence of synchronous
movement. Animal Behaviour, 35, 1652-1662.
Robinson, S. R., & Smotherman, W. P. (1988). Chance and chunks in
the ontogeny of fetal behavior. In W. P. Smotherman & S. R. Robinson (Eds.), Behavior of the fetus (pp. 95-115 ). Caldwell, N J: Telford
Press.
Robinson, S. R., & Smotherman, W. P. (1991a). Fetal learning: Implications for the development of kin recognition. In P. G. Hopper
(Ed.), Kin recognition (pp. 308-334). Cambridge, UK: Cambridge
University Press.
Robinson, S. R., & Smotherman, W. P. (1991b). The amniotic sac as
scaffolding: Prenatal ontogeny of an action pattern: Developmental
Psychobiology, 24, 463-485.
Robinson, S. R., & Smotherman, W. P. (1992a). Fundamental motor
patterns of the mammalian fetus. Journal of Neurobiolog~, 23, 15741600.
Robinson, S. R., & Smotherman, W. P. (1992b). Organization of the
stretch response to milk in the rat fetus. Developmental Psychobiology, 25, 33-49.
Robinson, S. R., & Smotherman, W. P. (1992c). The emergence of behavioral regulation during fetal development. In G. Turkewitz (Ed.),
Annals of the New YorkAcademy of Sciences: Vol.662. Developmental
Psychobiology (pp. 53-83). New York: New York Academy of
Sciences.
Robinson, S. R., & Smotherman, W. P. (1994). Behavioral effects of
milk in the rat fetus. Behavioral Neuroscience, 108, 1139-1149.
Robinson, S. R., & Smotherman, W. P. ( 1995 ). Habituation and classical conditioning in the rat fetus: Opioid involvements. In J. P. Lecanuet, N. A. Krasnegor, W. P. Fifer, & W. P. Smotherman (Eds.), Fetal
development: A psychobiological perspective. New York: Erlbaum.
Ronca, A. E., Lamkin, C. A., & Alberts, J. R. (1993). Maternal contributions to sensory experience in the fetal and newborn rat (Rattus
norvegicus). Journal of Comparative Psychology, 107, 61-74.
Sadowsky, D. W., Martel, J. K., Jenkins, S. L., Poore, M. G., Cabalum,
T., & Nathanielsz, P. W. (1992). Pulsatile oxytocin administered to
ewes at 120 to 140 days gestational age increases the rate of maturation of the fetal electrocorticogram and nuchal activity. Journal of
Developmental Physiology, 17, 175-181.
Schaal, B., Orgeur, P., & Rognon, C. (1995). Odor sensing in the human
fetus: Anatomical, functional and chemo-ecological bases. In J. P.
Lecanuet, N. A. Krasnegor, W. P. Fifer, & W. P. Smotherman (Eds.),
Fetal development: A psychobiological perspective (pp. 205-237).
New York: Erlbaum.
Schanberg, S. M., & Field, T. M. ( 1987 ). Sensory deprivation stress and
supplemental stimulation in the rat pup and preterm human neonate. Child Development, 58, 1431-1447.
Smotherman, W. P., & Robinson, S. R. (1985). The rat fetus in its environment: Behavioral adjustments to novel, familiar, aversive and
conditioned stimuli presented in utero. Behavioral Neuroscience, 99,
521-530.
Smotherman, W. P., & Robinson, S. R. (1986). Environmental determinants of behaviour in the rat fetus. Animal Behaviour, 34, 18591873.
Smotherman, W. P., & Robinson, S. R. (1987a). Prenatal influences on
development: behavior is not a trivial aspect of prenatal life. Journal
of Developmental and Behavioral Pediatrics, 8, 171-176.
Smotherman, W. P., & Robinson, S. R. (1987b). Psychobiology of fetal
experience in the rat. In N. A. Krasnegor, E. M. Blass, M. A. Hofer, &
W. E Smotherman (Eds.), Perinatal development: A psychobiological
perspective (pp. 39-60 ). Orlando, FL: Academic Press.
Smotherman, W. P., & Robinson, S. R. (Eds.). (1988a). Behavior of the
fetus. Caldwell, NJ: Telford Press.
Smotherman, W. E, & Robinson, S. R. (1988b). Dimensions of fetal
investigation. In Smotherman W. E & Robinson S. R. (Eds.), Behavior of thefetus (pp. 19-34). Caldwell, NJ.: Telford Press.
Smotherman, W. P., & Robinson, S. R. (1988c). The uterus as environment: The ecology of fetal experience. In E. M. Blass (Ed.), Hand-
book of behavioral neurobiology: vol. 9. Developmentalpsychobiology
and behavioral ecology (pp. 149-196 ). New York: Plenum Press.
Smotherman, W. P., & Robinson, S. R. (1989). Cryptopsychobiology:
The appearance, disappearance and reappearance of a species-typical
action pattern during early development. Behavioral Neuroscience,
103, 153-160.
Smotherman, W. P., & Robinson, S. R. (1990). The prenatal origins of
behavioral organization. Psychological Science, 1, 97-106.
Smotherman, W. E, & Robinson, S. R. ( 1991 ). Accessibility of the rat
fetus for psychobiological investigation. In H. Shair, G. A. Barr, &
M. A. Hofer (Eds.), Developmental psychobiology: New methods and
changing concepts (pp. 148-166). New York: Oxford University
Press.
Smotherman, W. E, & Robinson, S. R. (1992a). Kappa opioid mediation of fetal responses to milk. Behavioral Neuroscience, 106, 396407.
Smotherman, W. E, & Robinson, S. R. (1992b). Dimethyl disulfide
mimics the effects of milk on fetal behavior and responsiveness to
cutaneous stimuli. Physiology & Behavior, 52, 761-765.
Smotherman, W. P., & Robinson, S. R. (1992c). Prenatal experience
with milk: Fetal behavior and endogenous opioid systems. Neuroscience and Biobehavioral Reviews, 16, 351-364.
Smotherman, W. P., & Robinson, S. R. (1992d). Opioid control of the
fetal stretch response: Implications for the first suckling episode. Behavioral Neuroscience, 106, 866-873.
Smotherman, W. P., & Robinson, S. R. (1994a). Caveats in the study of
perinatal behavioral development. Neuroscience and Biobehavioral
Reviews, 18, 347-354.
Smotherrnan, W. P., & Robinson, S. R. (1994b). Milk as the proximal
mechanism for behavioral change in the newborn. Acta Paediatrica,
(Supplement), 397, 64-70.
Smotherman, W. P., & Robinson, S. R. ( 1995 ). Tracing developmental
trajectories into the prenatal period. In J. P. Lecanuet, N. A. Krasnegor, W. P. Filer, & W. P. Smotherman (Eds.), Fetal development: A
psychobiological perspective (pp. 295-314). New York: Erlbaum.
Smotherman, W. P., Woodruff, K. S., Robinson, S. R., del Real, C.,
Barron, S., & Riley, E. P. (1986). Spontaneous fetal behavior after
maternal exposure to ethanol. Pharmacology Biochemistry and Behavior, 24, 165-170.
Thelen, E. (1984). Learning to walk: Ecological demands and phylogenetic constraints. In L. P. Lipsitt (Ed.), Advances in infancy research
vol. 3 (pp. 213-250). Norwood, N J: Ablex.
Thelen, E. (1989). Self-organization in developmental processes: can
systems approaches work? In M. R. Gunnar & E. Thelen, (Eds.),
Systems and development, Minnesota Symposium on Child Psychology Iiol. 22, (pp. 77-117). New York: Erlbaum.
Received D e c e m b e r 10, 1993
Revision received August 3, 1995
AceeptedAugust 18, 1995 •