Lateral Bias in Infant Chimpanzees (Pan troglodytes)

Copyright 1990 by the American Psychological Association, Inc. 0735-7036/90/S00.75

Journal of Comparative Psychology 1990, Vol. 104, No. 4, 309-321

Lateral Bias in Infant Chimpanzees (Pan troglodytes) William D. Hopkins

Kim A. Bard

Language Research Center, Georgia State University, and Yerkes Regional Primate Research Center, Emory University

Division of Reproductive Biology, Yerkes Regional Primate Research Center, Emory University

Carolyn L. Fort Yerkes Regional Primate Research Center, Emory University This study documents the presence, strength, and direction of lateralization in chimpanzees (Pan troglodytes) over the first 3 months of life. Nursery-reared chimpanzees (7 males and 5 females) were repeatedly assessed on a behavioral scale. Lateral bias was measured for 4 behaviors: handto-mouth, hand-to-hand, defensive grasp, and first step. Hand-to-mouth was significantly lateralized for the sample. Eight of the 10 chimpanzees that showed hand-to-mouth used the right hand. Lateral bias for defensive grasp was positively related to lateral bias both of first step and of hand-to-mouth. Lateral bias in hand-to-mouth was inversely related to lateral bias in handto-hand. Strength of lateralization increased as chimpanzees matured. These laterality effects in infant chimpanzees were expressed under conditions of emotional arousal. Moreover, degree of laterality may be a predictor of responsivity to stress.

Recently, there has been a resurgence of interest in the nature and distribution of lateralized behavior among nonhuman primates in relation to that for humans (see MacNeilage, Studdert-Kennedy, & Lindblom, 1987). There is considerable debate over the factors that affect the distribution of lateralized behavior in both human and nonhuman primates. Therefore, it is important to discern those lateralization traits that human and nonhuman primates may share. Approximately 90% of humans are considered predominantly right-handed (Annett, 1985), although the notion of clear dichotomies of right- and left-handers is considered to be outdated (Healey, Liederman, & Geschwind, 1986). InThis study was supported by National Institutes of Health (NIH) Grant RR-00165 to die Yerkes Center, NIH Grant RR-03591 to R. B. Swenson at the Yerkes Center, and by National Institute of Child Health and Human Development (NICHD) Intramural Research Program Funds through the Laboratory of Comparative Ethology, NIH. Additional support was provided by NICHD Grant 06016 to the Yerkes Regional Primate Research Center and National Research Service Award Research Training Fellowship HD-07105 to Kim A. Bard. We gratefully acknowledge the assistance provided by the Divisions of Reproductive Biology and Veterinary Medicine, the animal care staff in the Great Ape Nursery, and numerous student assistants at the Yerkes Research Center. Grateful appreciation is extended to two anonymous reviewers and Jeannette Ward for their very constructive critiques. We are indebted to Stephen J. Suomi for his support and encouragement, Josh A. Schneider for photographic services, and Kathleen A. Platzman, a Brazelton-certified examiner, for training and for comments on the manuscript. Yerkes Regional Primate Research Center is fully accredited by the American Association for Accreditation of Laboratory Animal Care. Correspondence concerning this article should be addressed to Kim A. Bard, Division of Reproductive Biology, Yerkes Regional Primate Research Center, Emory University, Atlanta, Georgia 30322.

309

stead, the distribution of hand preference is considered to form a continuum from strongly right-handed to strongly lefthanded persons. Additionally, hand preference seems to be influenced by factors such as heredity (Annett, 1985; Levy & Nagylaki, 1972), neurological insult (Geschwind & Galaburda, 1985), development (Young, Segalowitz, Corter, & Trehub, 1983), and task demands (Fagot & Vauclair, 1988; Healey et al., 1986; Hopkins, Washburn, & Rumbaugh, 1989). For example, a greater incidence of left-handedness has been linked to mental retardation and reading disabilities as well as other developmental disorders (Pipe, 1988; Satz, Soper, & Orsini, 1988). Given the influence of these factors, many argue that biological components or dispositions may influence both hand preference and other lateralized processes in humans (Witelson, 1987). This argument is bolstered by the fact that anatomical asymmetries exist in the brain of both human adults and neonates (Geschwind & Levitsky, 1968; Wada, Clarke, & Hamm, 1975; Witelson, 1977; Witelson & Pallie, 1973). Further support is found in the observation that other asymmetries are lateralized very early in life, such as turning responses (see Turkewitz, 1977) and processing of speech and nonspeech sounds (Molfese & Molfese, 1979). A growing literature suggests that nonhuman primates evidence anatomical asymmetries similar to those observed in humans (Cain & Wada, 1979; Falk, 1978; Falk, Cheverud, Vannier, & Conroy, 1986; Holloway & de la Costelareymondie, 1982; LeMay, 1985; LeMay & Geschwind, 1975; YeniKomshian & Benson, 1976). Moreover, some behavioral studies have revealed lateralized processing for species-specific vocalizations (Heflher & Heffner, 1984; Petersen, Beecher, Zoloth, Moody, & Stebbins, 1978), tones (Dewson, 1977; Pohl, 1983), facial discrimination (Hamilton & Vermeire, 1988; Morris, Hopkins, Bolser-Gilmore, & Washburn, in press), line orientation (Hamilton, 1983; Hopkins & Morris,

310

K. BARD, W. HOPKINS, AND C. FORT

1989), dot localization (Jason, Cowey, & Weiskrantz, 1984), and form recognition (Hopkins, Washburn, & Rumbaugh, 1990), despite the rather negative findings for hand preference. It is unclear, however, whether asymmetries are present early in life in nonhuman primates. Only a few studies have examined the ontogeny of lateralized behaviors such as hand preference or head turning. MacNeilage et al. (1987) argued in their recent review that a significant number of studies on hand preference in nonhuman primates have used young subjects, thus skewing the findings. Their assumption is that ontogeny does influence both strength and direction of hand preference in nonhuman primates. However, this assumption is based retrospectively on only a handful of observations. Very little empirical data that directly examine lateralized behavior in infant nonhuman primates have been collected. A number of studies have reported lateralized behaviors in adults of various primate species (see MacNeilage et al., 1987; Ward & Hopkins, in press; Warren, 1977, 1980, for reviews).

Figure 1. The stepping reflex and first step of a young chimpanzee (11 days old).

Figure 2. A cloth is placed lightly over the face to elicit the defensive grasp.

Lehman (1980) reported that older infant macaques show stronger hand preferences than younger infants, although the distribution of right- and left-handed subjects was equal. Vauclair and Fagot (1987) reported similar findings for a group of baboons. More recently, Fagot (in press) reported that four baboons evidenced a right-hand preference for goaldirected reaching behaviors as early as the first 2 weeks of life. Among great apes one infant chimpanzee and one infant orangutan were found to have right-hand preferences on bimanual and unimanual tasks, and stronger preferences were found in the orangutan (Bresard & Bresson, 1983). Righthand preference was displayed in food and nonfood reaching, a right leading limb was exhibited for hand-limb locomotion, and left-hand preference was displayed for self-touching by an orangutan between 12-70 weeks of age (Cunningham, Forsythe, & Ward, 1989). Chorazyna (1976) reported that a single chimpanzee developed hand preferences that shifted from primarily left-hand preference to the right after 37 weeks of age. Behavioral assessments performed on very young human and nonhuman primates provide the opportunity to discover early asymmetries. The Neonatal Behavioral Assessment Scale (NBAS; Brazelton, 1973, 1984) has been used for humans (see Francis, Self, & Horowitz, 1987, for a review) and great apes (Bard, 1990; Bard & Platzman, 1989; Hallock, 1982; Hallock, Worobey, & Self, 1989; Redshaw, 1989), and a substantially modified version has been used for rhesus macaques (e.g., Levin, Schneider, Ferguson, Schantz, & Bowman, 1988; Schneider, 1987; Suomi, 1987) and cebus monkeys (Visalberghi & Suomi, 1988). The NBAS is also useful for measuring the neurobehavioral integrity of neonatal chimpanzees (Bard, Platzman, Suomi, & Schneider, 1989). Moreover, chimpanzees' performance on the NBAS is in many respects indistinguishable from humans' (Bard & Platzman, 1989; Bard, Platzman, & Suomi, 1987).

LATERALIZED BEHAVIOR IN CHIMPANZEE INFANTS

The similarities between human and chimpanzee neonates are more remarkable than the differences (Bard, 1990; Bard, Platzman, Lester, & Suomi, 1990; Hallock et al., 1989). Neonatal chimpanzees and humans attend to the same social and nonsocial stimuli with the same quality of engagement. Overall motor performances are essentially identical. Furthermore, the stability of the autonomic nervous system is comparable for both species. The areas in which chimpanzee and human neonates differ relate to behavioral state, specifically in the clusters of behavior involved in state regulation and range of state (Lester, 1984). Chimpanzee neonates exhibit a higher degree of alertness and a smaller range of state than human neonates (Bard, 1990). As a part of the NBAS, self-calming or self-quieting behaviors are recorded. Self-calming behaviors can take the form of sucking, either on the wrist, fist, or thumb; focusing attention on either a visual or auditory stimulus; or maintaining a preferred body posture, including hand-to-hand clasping (Gianino & Tronick, 1988). These behaviors are used by infants when they are fussy or crying to return to a quiet, alert behavioral state (Brazelton, 1973, 1984; Sammons, 1989). Chimpanzee and human neonates do not differ in their ability to self-calm nor in their ability to bring their hand to their mouth for sucking (Bard, 1990; Hallock et al., 1989; Redshaw, 1989). To date, there are no published reports of laterality in the self-quieting behaviors of either human or chimpanzee neonates. However, asymmetries both in the expression of positive emotional states and in self-regulatory behavior (e.g., self-clasping) were suggested by Trevarthen (1986) for slightly older human babies. In this study we report observations of the ontogeny of lateralized behaviors in chimpanzee infants from birth

311

Figure 4. The self-calming behavior of hand-to-mouth can take the form of sucking on a clenched fist.

through the first 3 months of age. We have three specific aims: (a) to describe the presence, strength, and direction of lateralized behaviors; (b) to identify the contexts in which lateralized behaviors occur; and (c) to describe changes in the strength or the direction of lateralized behaviors over the first 3 months of life. Method

Subjects Twelve chimpanzees (Pan troglodytes), 1 males and 5 females, served as subjects. The subjects had been placed in a nursery because of inadequate maternal care within 24 hr after birth (n = 9), within 1 week of birth (n = 1), or within the first 3 weeks of life (n = 2). All subjects were kept in incubators at the Yerkes Regional Primate Research Center nursery for the 1st month of life and subsequently given standard nursery care.

Procedure

Figure 3. Hand-to-hand grasps are one self-calming behavior used by young chimpanzees.

The NBAS was administered every other day from 2 days after birth through 6 weeks and then once a week through 12 weeks of life. The mean number of tests conducted in the 1st month of life was 12.25 (range, 5-16); in the 2nd month of life, 6.33 (range, 3-9); and in the 3rd month, 2.75 (range, 0-4). The NBAS is a test of neurobehavioral integrity and consists of 28 behavioral items and 18 reflexes. The entire NBAS assessment takes between 25-35 min and was administered at the same time each day, midway between feedings, in a dimly lit and quiet room. The primary examiner, Kim A. Bard, was trained to 90% reliability by a Brazelton-certified examiner. The NBAS is a test designed for neonatal humans. We decided, however, to extend the testing for chimpanzees beyond the neonatal period for several reasons. First, we wanted to describe the developmental performance of chimpanzees. Second, the NBAS provided enriching stimulation for the infants and a structure in which to interact. Third, we found that many items continued to provide valuable information on the neurobehavioral functioning of chimpanzees past the neonatal period.

312

K. BARD, W. HOPKINS, AND C. FORT

Figure 5. The self-calming behavior of hand-to-mouth usually takes the form of sucking on the right thumb.

Lateral bias may be displayed in 2 of the 28 behavioral items of the NBAS (i.e., defensive grasps and hand-to-mouth behavior) and in IS of the 18 reflexes measured during the NBAS (no asymmetries can occur for glabella, nystagmus, or sucking). The procedures involved with assessing lateral bias for behaviors and reflexes are presented in the following sections. For the purposes of this study, more focus was placed on the lateral bias of reflexes and behaviors than is typical for NBAS testing. Behaviors. The lateral bias of four behaviors of the NBAS was measured: first step, defensive grasps, hand-to-hand grasp, and handto-mouth behavior. The walking reflex (illustrated in Figure 1) was elicited by gently leaning the infant forward from a standing position. The foot that moved forward first was noted. Defensive grasps were recorded after a cloth was placed to cover the face (illustrated in Figure 2). Upper quadrant swipes, swipes to the midline, and actual

grasps of the cloth were recorded. Hand-to-hand grasps (Figure 3) and hand-to-mouth behaviors (Figures 4 and 5) were two of the selfcalming behaviors used by chimpanzee infants. When infants became fussy or cried, they were allowed a 15-s period in which to quiet themselves. Only hand-to-mouth behaviors and hand-to-hand grasps that occurred after a period of fussiness or crying were scored as these self-calming behaviors. Both the absolute frequency and the lateral bias of these behaviors were noted. The hand that grasped was recorded in hand-to-hand. If the fingers of both hands intertwined, a hand-to-hand behavior was recorded, but no lateral bias was recorded. Reflexes. The reflexes of the NBAS are scored on a 4-point scale (0 = absent, 1 = weak response, 2 = normal response, and 3 = obligatory hyperactive response) that is designed to discriminate between normal and abnormal responses. An asymmetry is recorded for a reflex only when a different score is recorded for the right and left elicitation of a reflex. This scoring system does not distinguish between degrees of a normal response. The following list briefly describes the 15 reflexes in which asymmetries were noted if they occurred: (a) plantar (foot) grasp; (b) hand grasp; (c) ankle clonus (a tremor of the foot when steady pressure is applied); (d) Babinski (flexion of the foot and toes when the side of the foot is stroked); (e) stand (extension of the legs with brief support of weight); (f) walk (stepping action with hip and knee flexion); (g) place (flexion of the foot when back of foot is stroked); (h) Gallant (incurvation of the trunk when side of spine is stroked); (i) crawl (movement of hips, legs, and feet and freeing of face); (j) spin (tonic deviation of the eyes and head when rotated); (k) tonic neck reflex (TNR; asymmetric reflex elicited by turning the head to the side when supine); (1) Moro (extension of the arms and hips and abduction of the shoulders with loss of support to the head); (m) rooting (head turn with mouth open on stimulation of the side of the mouth); (n) passive movements of the arms (resistance to extension and recoil when arms are moved); and (o) passive movement of the legs (AndreThomas, Chesni, & Saint-Anne Dargassies, 1960; Brazelton, 1973, 1984;Prechtl, 1977).

Data Analysis A z score was calculated for each behavior and each reflex in order to determine the lateral bias for each subject. Such a calculation was

Table 1 Z Scores for Each Subject and Lateralized Behavior Hand-to-mouth

Hand-to-hand

Defense

Step

z z z Subject n n n n z Males -2.24* \ Scott 3.00" 5 -1.07 14 9 -1.00 Donald 4.90** 0.00 10 1.09 21 12 35 -1.73 Lamar 1.73 -1.00 4 2.12* 4 3 8 2.00* Fritz NA NA 1.73 12 -0.82 6 Chip -1.90 0.45 5 -1.15 10 19 1.50 16 1 -0.77 Barney -1.00 15 -0.83 13 -1.81 15 NA 1.41 Jarred 8 1.00 -1.41 9 8 Females 3.74** 1 Katrina 0.58 14 14 3 0.00 -1.00 3.71** 2.71** Carole 2 11 21 -1.41 0.28 13 3.71** NA Lillie 21 1.41 0.71 8 8 Sheena 3.31** 0.50 -0.24 17 33 16 -1.41 18 37 Sierra 4.43** 27 -2.79** 1.89 18 23 1.88 Note. Positive values reflect a right lateral bias, and negative values reflect left lateral bias. NA indicates that no behavior was observed. *p o •* o o fn

§

n o

c

91 ° 71 ~ 71

O O 0 OO

S

-< o o o U

3

— o o o o o— 1

— O0 O 0

u _>

O O O tS fN — —

— O mO—

s>

N

8888588 OOOOTTT

a 88E83 c — o o o — ca

R

O m O t~- O O ro

(N OO fl Tf O

N

O S oo «^i O

r^ O Q oo —« O O >n O O O —0 O O

— * fN ro O O • 6S-c-93S 3 « a i 5 b P M •