Invited symposium
Chair: Carolyn Rovee-Collier
Discussant: Carolyn Rovee-Collier
Publication of the classic volume Early Behavior: Comparative andDevelopmental Approaches (Hess, Stevenson, & Rheingold, Eds.; Wiley, 1967)marked the formal debut of experimental research with infants as we know it.Each chapter in the volume, authored by one of the participants, has sincegiven rise to an entire field of infancy research--learning (habituation,conditioning, discrimination), visual recognition memory, long-term memory,preferential looking, sucking behavior, inhibition of distress, socialinteraction, the development of visual attention and perception, andcomparative development. In honor of their seminal contributions to ourscience, ICIS has reunited these esteemed researchers and asked them to reflecton the early programmatic work that they reported in this volume, whatmotivated their particular research questions, the problems that they faced inperforming it, and its reception by the scientific community at the time.During the roundtable discussion that will follow the formal presentations, theparticipants will comment on where their early work has led, the remaining gapsthey perceive in current research, and the problems facing current researchersin their respective fields. The titles of the presentations are the titles ofthe original chapters.
Details of individual items:
paper
This book was edited by myself, Eckhard Hess, and Harriet Rheingoldand was published by Wiley in 1967. It was the product of two researchconferences, in 1963 and 1965, that were organized by the United StatesSocial Science Research Council on the topic of early development. Theinvited participants represented differing theoretical and empiricalapproaches to the study of the development of the young of differentspecies. The first version of their reports was presented at theinitial meeting, and their final papers were discussed at the secondone. These conferences marked the first instance in which a broadspectrum of thought was brought to bear on the study of the basicproblem of early development. The content of the chapters covered notonly a range of different species but also a range of behavior processesfrom the very simple to the very complex. The viewpoints of theparticipants varied from those in which genetic and innate organizationsof perception and behavior were deemed to be most critical to those inwhich conditioning and learning were thought to be sufficient forunderstanding behavioral development. The research reported in thesechapters and the diverse points of view of the authors havesignificantly influenced the concepts and designs of subsequentgenerations of researchers as well as the public welfare, health, andeducational policies of many nations.
paper
The belief that the young of animals and humans learn new behavioreasier and retain it longer than adults originated in folklore and wasthe foundation for the idea that infancy is the period during whichadulthood is structured as well as numerous social movements andeducational practices over the last century and a half. As of themid-1960s, however, developmental studies bearing on this belief weregrossly deficient. A major source of divergence among them was thefailure of early investigators to control and equate motivational andreinforcing conditions in animals of different ages. This chapterfocused on the development of learning abilities and how their basicdimensions and parameters change with age. When we equated motivationallevels, we found that simple learning ability did not change with age,but younger animals forgot more rapidly. Their rapid forgetting raises the question of how early experiences canbe important for later behavior. In fact, early memories might persistinto adulthood if developing organisms encounter occasional reminders,or reinstatements, of the original event. We found that preweanlingscontinued to exhibit a high degree of conditioned fear 4 weeks later ifthey received one shock a week between training and testing. Theperiodic shocks alone were insufficient to produce learned fear. It istempting to speculate that reinstatement may be one basis for thewidespread belief that early experiences are important in determiningadult behavior. It is probably adaptive that early memory is weak and extends over onlya short duration. If all early fears persisted into adulthood, theywould be an enormous hindrance. Similarly, many instrumental responses,if not forgotten or otherwise eliminated from the repertoire, wouldimpede development.
paper
Clinging, grooming, and play are stereotyped activities of youngchimpanzees and account for most of their social behavior. Playincludes slapping, wrestling, and pushing, accompanied by half-closedeyes, a broad smile, and a panting laugh. Grooming consists of closevisual inspection of the spot being groomed, with coordinated probingand picking movements of the fingers and lips. Grooming is apreoccupation of adults but occurs in incomplete form in young animals.Clinging maintains infants’ contact with their mothers; it occurs oftenbetween young chimpanzees in late childhood but is rarer in adolescents. This chapter focused on the role of these activities in the socialdevelopment of the chimpanzee. The rewarding effectiveness of theseactivities, their selection in preference tests, and their occurrence insocial interactions vary with motivational factors. Although play isthe preferred activity of young chimpanzees in familiar situations, forexample, manipulations that increased emotional arousal decreased playand increased clinging; as clinging subsequently decreased, playincreased. Highly aroused young prefer to cling to the most familiarobject; moderately aroused young prefer to play with more novelobjects. Presumably, the ability of these activities to maintainarousal within an optimum range contributes to their rewarding effects. Social interactions can also be interpreted within an arousalframework. The infant’s relation to its mother is characterized byarousal-reducing activities such as clinging; infants not permitted tocling become actively distressed. With age, clinging decreases and playincreases. Also with age, stimuli that were aversive at younger agesincreasingly evoke play. This perspective can account for individualdifferences in social responsiveness during infancy and early childhood.
paper
We sought a handle on the world of the newborn by providing detaileddescriptions of two organized patterns of behavior that are present atbirth--sucking and looking. We found that sucking consistently loweredinfants’ movement before and after feeding relative to light foreheadstroking. This effect existed before the first postnatal feeding butwas greater for good suckers. A hydraencephalous child with only abrain stem responded normally, suggesting that the congenital relationbetween sucking and movement is organized subcortically. We alsomonitored the two components of sucking with a nipple that regulatedmilk delivery. Infants typically produced one expression (nipplelapping) movement for each suction (negative pressure) movement. Whenmilk was given for each expression, some suction dropped out, and theratio shifted in favor of expression; when milk was given for eachsuction, however, the ratio was unaffected. Although expression did notdrop out, it was flexible. When different threshold pressures ofexpression were required to obtain milk, infants adjusted their level ofexpression accordingly. Visual preferences were determined by recording the corneal reflectionsof infrared marker lights on two panels placed perpendicular to infants’line of sight. Infants looked more at panels of intermediate brightnessand random shapes with an intermediate number of turns, but they lookedlonger at a simple than at more complex checkerboards. Ocularorientation was determined by a frame-by-frame analysis of the distancebetween the center of the pupil and the corneal reflection of infraredlights marking the position of a black triangle on a white field. Forcontrol subjects, infrared lights formed a hypothetical triangle.Relative to controls, ocular orientations of experimental subjects wereless variable and more concentrated on the vertices but not the linearcomponents of the triangle.
paper
This chapter addressed the origin of knowledge about the environment.Beginning with the perceptual capacities and processes present at birth,I explored the role of learning through perceptual experience per se asdistinct from learning through actions or consequences of action. Toanswer this, I recorded responses of newborns from different species tovisual stimuli. Previously, newly hatched chicks were thought to peck randomly atobjects, subsequently pecking at food items only through differentialreinforcement. However, chicks’ initial pecking at forms was nonrandomand adaptive relative to natural food sources. Although feedinginfluenced subsequent pecking preferences, so did exposure to particularforms and nonreinforced pecking in the presence of other chicks. Thelatter effects are not interpretable within a learning framework.Newborn monkeys preferred patterned stimulation, but dark rearingreduced this preference to chance. Human newborns younger than 1 day also exhibited nonrandom preferencesfor visual patterns. Infants <1 week fixated schematic faces (scrambledor nonscrambled) significantly longer than ovals with two black dots,which they fixated longer than gray or white ovals. The preference forfacial characteristics remained stable through 6 months and mayfacilitate the development of social responsiveness. At 2 months,infants’ preference shifted from stripes to bull’s-eyes; over repeatedexposures, infants fixated variable patterns longer than constant onesbut conserved total fixation time. These studies revealed that acquisition of knowledge about theenvironment begins at the first look. During the early weeks,information acquired from a single look at a certain target may beretained only long enough to cause attention to shift to anothertarget. Even at this minimal level, the knowledge acquired givesdifferential exposure to parts of the environment, allowing immatureorganisms to become familiar with the environment--a basic form oflearning.
paper
This chapter focused on the stimulus determinants of behavioral changein human infants. Neonates habituated to repeated presentations of anolfactory stimulus and renewed responding to a novel odor. Infantshabituated to a mixture renewed responding less to the component thatwas more similar to the mixture (adult judgments). This result impliesthat olfactory habituation is a higher nervous-system function. Over alternating exposures, neonates sucked more on a standard nipplenipple than nipple-only controls (positive contrast) and less on arubber tube than tube-only controls (negative contrast). When rewardedwith 5%-dextrose, tube-sucking increased over baseline and decreasedduring extinction; tube-sucking by controls given dextrose betweentrials progressively decreased. Thus, dextrose made a poor elicitor ofsucking more effective. Dextrose reinforcement also establishedconditioned head-turning in newborns. For the experimental group, abuzzer sounded, the left cheek was stroked, and left turns wereimmediately followed by dextrose; left turns increased duringconditioning and decreased during extinction. Controls given dextrosenoncontingently 8-10 s after stroking exhibited no response change. Ina second study, stroking one cheek was paired with a tone (S+) anddextrose, whereas stroking the other was paired with a buzzer (S-) andno dextrose. Ipsilateral responding to S+ increased during acquisitionand decreased during extinction. In a third study, right turnsfollowing right-sided stimulation were reinforced with dextrose in thepresence of S+ but not S-; after initial training, the contingency wasreversed. Responding increased to S+ and decreased when it became S-;by the end of reversal training, infants responded more to the new S+.Similar procedures were effective in shorter sessions with olderinfants. The finding that environmental events can selectively alter responseswithin only a few days of birth raises optimism about infants’ learningpotential and its experimental study
paper
This chapter describes a longitudinal study of the development ofconditioned head-turning, a motor component of feeding behavior.Infants were newly born, 3 months, or 5 months old when the experimentbegan. All roomed with their mothers at the Institute for Care ofMother and Child in Prague, Czechoslovakia, and concurrentlyparticipated in other conditioning studies (eye-blinking, orientationalhead-turning) as well. The conditioning stimuli were sounds of a bell or buzzer; theunconditioned reinforcement was milk presented from one side or theother through a nipple connected to a thermos bottle. Stabilimetermovements, respiration, and head rotations in a padded head-cradle wererecorded polygraphically. Testing occurred daily during a normalfeeding period after the morning nap. Initially, left turns *30o to the bell were rewarded with milk to astringent conditioning criterion. Next, the bell was presented withoutmilk to an extinction criterion (the same), followed by reconditioning.Next, infants learned a discrimination--left turns to the bell and rightturns to the buzzer, which was reversed twice. Four phases of learningwere observed: (1) indifference to the CS, (2) inhibition of generalactivity, (3) unstable or isolated CRs, (4) stable conditioning. Theduration of phases, trials to criterion, and CR latency decreased overage groups, but speed of extinction did not--possibly because newborns,who had taken longer to condition, were older when undergoingextinction. The youngest group also reconditioned significantly faster,perhaps for the same reason; the fastest conditioners met thediscrimination criterion by 21/2 months.