The Behavioral and Social Sciences: Achievements and Opportunities (1988)

Below is the uncorrected machine-read text of this chapter, intended to provide our own search engines and external engines with highly rich, chapter-representative searchable text of each book. Because it is UNCORRECTED material, please consider the following text as a useful but insufficient proxy for the authoritative book pages.

Motivational anc3 Social Contexts of Behavior

Motivational and Social Contexts of Behavior On any list of plights of the human condition, one would find prejudice, addiction, violence, and suicide. They represent the extreme and morally most troublesome phenomena concerning the origins and regulation of appetites, purposes, and sentiments. Questions about these phenomena often lie close to the surface of everyday life, and they have been addressed since the earliest times at that level, frequently in texts of religious, ethical, legal, and philo- sophical reflection. The questions take such practical forms as: Why do strong emotions arise? Why does intoxication hold such overwhelming appeal to some people but not to others? How is criminality bred and how can it be contained? Long-established traditions of reflection have yielded many compelling an- swers to such questions, forming the core of commonsense wisdom about them. But together these answers are often found to be contradictory. (Common sense, schooled by life, is greatly tolerant of logical contradiction.) Occasionally, one answer achieves predominance over its competitors, usually by acquiring the cloak of religious, ideological, or political authority. Research in these areas thus has peculiar burdens and special responsibilities. One burden derives from the power of common sense to insulate us from surprise. Many scientific theories about familiar matters, when proven by em- pirical research, are deemed obvious, fully expected, and not having needed scientific verification. Yet common sense could also have deemed directly con- trary findings to be obvious and fully expected. The special responsibility of research is to remain skeptical, to be unwilling to accept too readily any result 49

50 / The Behavioral and Social Sciences that accords with common sense, to insist on scientific proof. A more significant burden is that research Endings may not support a currently dominant idea. The responsibility of research is to follow theoretical leads and empirical results wherever they lead. In this chapter we discuss research on affective states and processes; the linkages between health, behavior, and social contexts; the causes and control of violent crime; and the nature of social interaction. Among the matters now under intensive study are the social and motivational conditions that affect vulnerability to depression, cardiovascular disease, and risk of addictive be- havior; the competing environmental cues and psychobiological processes that affect eating behavior and body weight; the ways that parental practices in managing children and criminal justice practices in sentencing offenders affect the course of"criminal careers," and the manner in which the social origins, size, and divisions of responsibility in a task group can affect what it produces. The research methods involved in these areas of investigation range from bio- physics to cultural analysis. Research on motivational and social contexts of behavior, perhaps more than any other area in this report, accents the promising opportunities and the sharp challenges of multidisciplinary science. AFFECT AND MOTIVATION How and why do drives and desires grow and decline? Are certain human emotions primary and universal? By what processes do children reach emo- tional maturity, and why do some fail to do so? How do people and animals regulate their appetites under changing conditions of scarcity or abundance? Why does exposure to stable environmental circumstances not stimulate iden- tical or consistent behavior? These and similar questions animate research on motivational systems and affective states and processes. At the physiological level, there is research, primarily using animals, on the motivations associated with hunger, thirst, and sexual behavior. By studying animal behavior in conjunction with brain events and structures, significant progress has been made in uncovering the neural features responsible for these motivational systems. There have also been major advances in understanding homeostatic systems that operate to control the intake and expenditure of energy and, in humans, the neuromuscular codes that make the face a prime vehicle for emotional expression. At least as important as these advances at the physiological level is better recognition of the complexity of motives and emotion. These phenomena can- not be reduced to the workings of single variables. The relations between cognition and emotion are complex, and emotion can be understood only in context, which almost always involves the cultural meaning system in which behavior is embedded. Moving from primarily individual phenomena, such as

Motivational and Social Contexts of Behavior / 51 pain and hunger, to more interactive phenomena, such as indignation and envy, the cultural components become more and more compelling. The un- derstanding of affect and motivation thus draws on talents and skills ranging from those of chemists, physicists, and computer experts staffing brain-imaging facilities, through biomedical and behavioral specialties traditionally concerned with personality and mental illness, to sociologists and anthropologists study- ing the contexts, occasions, and meanings of emotional expression. Emotional Expression, Perception, and Maturation Identification of the facial, vocal, and postural correlates of different emotions and development of increasingly sensitive measures of the activity of peripheral and central nervous systems hold important promise for illuminating the course of normal emotional development. They also provide major new opportunities to investigate the natural history, diagnosis, treatment, and prevention of af- fective illnesses, such as depression. One new approach that has already proven useful involves measuring the relative tension of various facial muscles that affect the facial expression of emotions. This muscle tension can be monitored electronically by surface elec- trodes, and the corresponding facial expressions can be monitored by television cameras and evaluated by anatomically based coding systems. Techniques of electronic recording, display, and analysis of the action of all the muscle groups that contribute to expression in the face now enable precise quantitative de- scriptions and realistic computer simulations of emotional expression. These and other advances have permitted detailed study of the development, percep- tion, and decoding of emotional displays and explorations of deceptive com- munication of emotion. Such information can provide, among other applica- tions, a new empirical basis for testing and monitoring the utility of pharmacological and other strategies used in the treatment of affective illnesses. Recent research indicates that a small number (S-10) of facial expressions seem to have the same meaning cross-culturally. The role of such communi- cations in regulating social order, as well as more elaborate learned expressions, is now under study. Communications of affect- "nonverbal communica- tions" are an important part of socialization since this is a powerful way for peers and elders to convey their attitudes about objects and behaviors of im- portance in their culture. Researchers are beginning to use technical advances in measuring facial expression to aid in studying the development of children's competence at deciphering expressions and to see how this form of commu- nication regulates behavior and stimulates emotional growth. For example, a new "social referencing" paradigm is being used to study infant social-emotional sensitivity. An experimental demonstration of this par- adigm in uncertain or potentially fearful situations involves the response of 1- year-olds to a "visual cliff," an illusion of discontinuity created by shifting grid

A 2 / The Behavioral and Social Sciences 5:'~ :: ::: ~:2:~ ~::~: : ,, ~ I:: ~ ~ ~ ~~ ~ .':.: ~~ ~ .:: t~,.,.'~,~'~.'~'..:'', .~ ~ .~ :2:.'.; :.:.:.:,:::::.: :~ : :s as,. ~2~,~',~,~:~'-'~.~'~,~,~,~ ,~:~,~'~',:~'~:~: - , ~:~"'',":,0 a,:.. ."': it: :: in' ''I :':::: ~ ''""'','' ':" ~~.~.~ :.~: F:::.'~ ''a' 'I ~ I: :: ::,.,:.,. I::: pi,..,,., ,~.:,',Y,,,~,~ ~2" ~'2' ':: bib'. ::::.:.:.:''.:.:.: ;,':':' i: .::: I:: ,': .': ~~ ':- V2:~: ~ ..'.,,,,,,.,' ':: ::: :

Motivational and Social Contexts of Behavior / 53 patterns laid over a flat surface. When the depth of the visual cliff is intermediate and the infants might or might not cross it, they look to their mothers. When mothers pose a "fear" expression, none of the infants cross; when mothers pose a "happy" expression, nearly all cross. In another experiment, when infants' mothers respond to a stranger in an unfriendly rather than neutral or friendly manner, the infants respond to the stranger with cardiac acceleration (char- acteristic of fear), less smiling, and more distress. This kind of evidence not only reflects infant sensitivity to emotional expression but also the use of that sensitivity in regulating behavior. Recent studies indicate that young children up to 3 years distinguish pri- marily a positive and a negative emotional category with little differentiation, for example, among excited, happy, or proud, and among sad, angry, or afraid. Over the next several years children begin to differentiate within the positive and negative categories in ways that are in some respects universal and in others culture specific. One intriguing finding is that children only gradually- between 3 and 13 years of age come to realize that it is possible to simulta- neously experience different emotions, such as being happy that one's lost dog has come home but sad that the dog has been injured. At first, only the positive or negative emotion is acknowledged; later, children will typically say that the two emotions can be experienced at the same time. Research is proceeding on how learning of this sort in childhood may contribute to the emotional capacity to cope with the stresses, complexities, and responsibilities of life. A developing mosaic of findings is beginning to yield a detailed picture of the link between a child's interaction with its parents, the child's emotional FACIAL EXPRESSION Are facial expressions of emotion entirely cul- ture-specific, or are there some universal expressions across the human family? If there are universal expressions, what are the common muscular gestures of the face that comprise them and what emotions do they express? Researchers interested in these questions have traveled far, in this case to mountain hamlets in New Guinea whose inhabitants have had minimal contact with the outside world the people have never worked for West- erners or traveled to commercial centers; neither speak nor understand Eng- lish or pidgin; and have never seen photographs, magazines, or motion pictures that might convey outsiders' modes of expression. T .1 · 1 r 1 · .1 · r 1 In Ine video Irames snown In this ngure, the Instructions were to pose emotions for: "your friend has come and you are happy"; "your child has died"; "you are angry and about to fight"; and "you see a dead pig that has been lying there for a long time." The facial displays corresponding to these emotions—happiness, sadness, anger, disgust are easily recognized by and the same as those of Westerners. A variety of experiments similar to this one have been done in other cultures. In fact, the four emotions shown here and a few others (fear, surprise) are easily and consistently produced and rec- ognized across preliterate and literate cultures.

54 / The Behavioral and Social Sciences perceptiveness, social competence, and peer group status. For example, a num- ber of researchers have documented the principal way that fathers and mothers differ in their style of child play, mothers being more verbal and didactic while fathers are more physical and arousing. In turn, these stylistic differences are reflected in children's social development, with the children of highly verbal mothers and physically very playful fathers being rated more socially competent with their peers. Other recent work indicates links between a child's ability to accurately recognize emotional expressions and the child's social status among peers; children who more correctly identify facial expressions are more popular. Some evidence further suggests that early peer status is predictive of later social and emotional adjustment. For example, there are findings that different types of low peer status vary in their stability and degree of association with later adjustment problems; specifically, the status of rejected children is more stable over time than that of either neglected (by peers) or popular children. Another line of work indicates that unpopular children do not necessarily differ in their knowledge about what to do in various social situations, but they do differ in the way they construct goals: popular children have more friendly, assertive goals, while unpopular ones have more task-oriented goals. Moreover, children who attribute social rejection to their own personal inadequacies are less likely to cope effectively by changing their styles than those who do not. Pilot intervention studies focusing on these and related dimensions of social competence are proving effective in promoting change in the social status of some socially rejected children. There is obvious promise as well as challenge in trying to map out the antecedents of emotional expression and discover how it operates on a fine-grained level to regulate social behavior and contribute to social competence. Emotive Circuitry and Metabolism in the Brain Researchers are beginning to gain very detailed knowledge about the nature of the different brain circuits involved in such motivational and affective pro- cesses as aggression, eating, drinking, and sexual behavior. The classical studies in the 1950s that proved that animals will work for electrical stimulation of certain brain areas to the exclusion of all else. They had an enormous impact. Animal studies descending from this pioneering work have shown that there are separate motivational systems for defensive fighting (between members of the same species), predatory fighting (usually between members of larger and smaller species), and feeding, but that the neural mechanisms of predatory attack are more closely related to the mechanisms of feeding than to those of defensive fighting. Motivational systems passing through the hypothalamus appear to be acti- vated at different stages by food and water, as well as by opiates and stimulants. The latter substances appear to activate the systems more strongly than the

Motivational and Social Contexts of Behavior / S5 more biologically useful rewards; this finding is important for understanding biological support for addictive behavior. In other studies, researchers have determined that, for most right-handed people, the right cerebral hemisphere is more specialized for both expression and reception of emotions, in compar- ison with the specialization of the left hemisphere for language. The exact significance of this asymmetry is still unknown; it is as yet an intriguing clue as to how the brain organizes emotion. Over the last few decades, a new discipline of neurochemistry has emerged. A wide variety of neurotransmitters, chemicals that transmit impulses from one nerve cell to another, have been identified more than 50, with estimates as high as 250. Different chemicals seem to operate in brain systems to carry out different functions. For example, one category of neurotransmitters, the en- dorphins, are especially involved in the modulation of pain; another, the cate- cholamines, operate in the control of activation and mood and play a critical role in episodes of depression. A variety of innovative technologies are making the anatomical and func- tional features of the brain more visible and quantifiable. Using new imaging methods, researchers can now measure neuroanatomical features, conduct studies of the volume and density of specific brain structures, and study the metabolism of living brains with unprecedented margins of safety and precision. Such projects involve interdisciplinary collaboration on a broad front. For example, high-quality PET studies of brain structure use precise cyclotron targetry, math- ematically advanced techniques of data reconstruction, high-purity radiophar- maceuticals, advanced diagnostic techniques, ingenious neuropharmacological strategies, and sophisticated behavioral methods and assessments. These meth- ods make possible the discovery and analysis of detailed brain circuits that underlie normal and abnormal affective processes, the localization of receptors that are affected by psychoactive drugs, and an ever-expanding horizon of related studies in humans and animals. Biobehavioral Rhythms Behavior that is rhythmic ranges from the relatively exotic, like the mating of marine organisms that takes place for only one day, at precisely the the same time every year, to the familiar sleep-awake cycle of most mammals. Rhyth- micity is a process that affects almost every realm of psychological function, from diurnal changes in thresholds for discriminating simple auditory and visual signals, to monthly changes in human mood (correlated partially with the female menstrual cycle), to annual variation in breeding and hibernation. Behavioral research on biological rhythmicity is advancing along two main fronts. First, there is a rapidly developing theory of biological cyclicity, espe- cially daily (circadian) rhythms. This is a quantitative, formal set of models that holds for a wide range of organisms, from insects through mammals,

56 / The Behavioral and Social Sciences including humans. Based on mathematical oscillator theory, this model has proven useful to understanding behavioral patterns as seemingly disparate as activity rhythms in laboratory mammals and cyclic seasonal depression in hu- mans. A second front is the analysis of specific psychologically relevant behav- ior. Disturbances in sleep, depression, and other forms of psychopathology all have cyclic components. Perhaps more important, some syndromes may be caused directly by jarring alterations in normal cyclicity, such as can arise from long-distance air travel. Patterns of Food Consumption How do animals and people regulate their weight? If adult animals are ex- perimentally starved or fattened and then given free access to food, they gen- erally return to their original weight. One explanation for this result is that the brain senses the body's weight (perhaps monitoring the size of fat stores), just as it senses blood pressure or carbon dioxide, and is set biologically to "defend'? a certain weight through shifts in food consumption or metabolism, similar to the way a thermostat is set to maintain a particular temperature. This set-point theory of weight regulation has been a fruitful point of departure for research. The theory suggests that most overweight people may happen to have a high set-point, so that maintaining lower weight is a constant battle between set- point regulation and psychological or social pressures to be thinner, leading to the bioregulatory pathologies of anorexia and bulimia and the frustrations of tens of millions of dieters. But researchers are also focusing on the ways in which this view of eating behavior is incomplete. One line of work with animals and people indicates that eating often occurs not in response to a present weight deficit but in anticipation of one, implying an important role for learning. In the last decade, there has been a remarkable coming together of three disparate traditions of studying how food-seeking behavior adjusts to environmental opportunities: operant conditioning (learning for rewards), the economic theory of resource utility maximization, and functional approaches in ethology. In all three do- mains, the immediate issue is how animals and humans settle on more or less optimally rewarding behavior. In operant conditioning, the issue is expressed in terms of how a pigeon allocates key pecks or a rat pushes on a number of different keys that distribute reinforcements (usually, access to food) according to different schedules (that is, temporal patterns of an irregular, chancy nature). In economics, the issue is generalized to how inputs pegged at various prices combine to yield the most satisfactory overall return or value. In ethology, the issue appears as the problem of understanding patterns of foraging in the wild for a variety of possible food sources. In an operant experiment, an animal is left for some time in an environment

Motivational and Social Contexts of Behavior / 57 where it can engage in a number of its usual activities, such as grooming and drinking, but in order to eat it must work by pushing on certain keys. The experimenter differentially rewards some patterns of behavior, and that inter- mittent schedule of rewards is found to be highly controlling of what the animal chooses to do with its time. The relationships between schedule changes and behavior changes can be expressed as mathematical regularities. When com- bined with information about an animal's nutritional needs and characteristics of food availability in its natural environment, these regularities have been shown under certain conditions to yield accurate predictions about foraging behavior. Ethology and operant conditioning, two research fields that were diametrically opposed in the past, have thus begun sharing ideas and methods. The laboratory operant box has become a tool of research on natural adaptation. The laboratory situations studied by animal behaviorists seem to be, in some ways, more complex and more natural (in terms of simulating the conditions of free-ranging animal behavior) than the situations usually arranged for human research subjects. However, when roughly comparable naturalistic situations of activity allocation are posed to people, their behavior turns out to be well described by the same mathematical generalizations. Investigators of these dif- ferent topics have recognized their common analytical interests and are now working together to develop predictions of behavior based on various possible fundamental principles. One very important issue is the empirical assessment of the economic principle of cost-benefit maximization in comparison with an alternative principle that derives from the animal studies and some of the human research: that attention is accorded to respective activities so that each yields the same average rate of reinforcement rather than the same marginal rate. Eating must also be studied in the context of other behaviors; rather than experiments in which caged animals are given access to a food cup but little else to do, the present challenge is to design theories and experiments com- bining food availability with numerous other available activities (sexual behav- ior, caring for the young, other forms of interaction, and so on) as competing alternatives. The study of foraging illustrates an important trend in the attempt to un- derstand motivation: the merging of functional-adaptive approaches with the study of mechanisms. Claims about the adaptive value of certain behaviors— for example, the greater selectivity in choosing mates by females because of their greater investment of the offspring- have prompted research to test such sociobiological claims and to search for the mechanisms that mediate the be- havior. The senses of smell and taste are also important in regulation of eating. Particularly for elderly people, loss of appetite can often be explained by a failure of the sense of smell. And since the smell system is much affected by Alzheimer's disease, subtle alterations in smell sensitivity may be key elements

58 / The Behavioral and Social Sciences in developing early diagnostic techniques. Research is also directed toward understanding genetically determined individual differences in taste. For ex- ample, two-thirds of people tested taste the compound PTC (phenylthicarban- ide) as distinctly bitter; the other third do not. These two groups respond very differently to sweeteners, one finding them much more potent than the other. It should be noted that scientific and clinical interests in developing a com- prehensive understanding of taste and smell are joined by practical interests in improving the prediction of sensory effects of food ingredients based on their chemical composition. On a different track, researchers are working to understand the functions that eating serves aside from being a source of nutrition, most notably as a vehicle for social symbolism and communality. For example, in most of south Asia which accounts for close to 20 percent of the world's population food is a major medium of communication that reinforces the hierarchial nature of the social system. Rules about who eats what food handled by whom play a powerful role in supporting and defining the caste structure. To fully under- stand food selection, the many and complex functions of food need to be taken account of in behavioral theories. BEHAVIOR AND HEALTH Although it is still common to think about health as though it were strictly a matter of human biology the research province of medical scientists, phys- iologists, pharmaceutical chemists—such thinking has become outdated. Until about 1940, the leading causes of death and disability in the United States were acute infectious diseases, such as pneumonia and tuberculosis; today, such diseases are not usually fatal, and the leading causes of death are illnesses that are significantly affected by behavioral and social factors, such as cigarette smoking, excessive drinking, illicit drug use, bad dietary habits, violence, stress, and refusal or inability to maintain recommended medical regimens. Of the 2 million deaths in the United States every year, about 1.5 million are due to heart disease, cancer, stroke, or fatal injuries and overdoses; for all of them, behavioral and social causes are known to be major risk factors. Studies of life expectancy and mortality among cigarette smokers, former smokers, and nonsmokers yield estimates of more than 300,000 excess deaths a year due to smoking. Epidemiological studies in metropolitan areas indicate that one of every five adults suffers from a diagnosable form of motivational or emotional disorder during a given year, most commonly alcoholism or depression. Fundamental research concerning the behavioral and social contexts of health and illness has increasing applications to prevention and treatment. Research taking place in laboratory experiments, accompanied by rigorous theoretical modeling, and in field sites such as schools, workplaces, and clinics addresses

Motivational and Social Contexts of Behavior / 59 such questions as: How do social influences and motivational dispositions affect the onset, maintenance, and cessation of drug, alcohol, and tobacco use? How pervasively do environmental stressors affect the body? What are the social and behavioral factors and processes that increase or decrease vulnerability to dis- ease? How do variations in the incentives and costs of health care affect how and whether People seek treatment, provision of care. and the outcomes of treatment? 1 ~ Prevention of Drug, Alcohol, and Tobacco Abuse "Peer pressure" and "youth culture" are bywords today of influences on adolescent alcohol, drug, and tobacco using behavior. This was not the case 20 years ago, when these ideas, borrowed from social science, first began to revise the popular and scientific thinking that linked youthful smoking to the straightforward desire to emulate adults, teenage drinking to juvenile rebellion, and illicit drug initiation to predatory pushers. Since then, a phalanx of careful studies, including ethnographic investigations among drug users in and out of school, panels of adolescents followed at yearly intervals through young adult- hood, and repeated national surveys of successive classes of students, have revealed a far more detailed and accurate picture, suggesting bases for selecting among and developing new approaches to treatment and, especially, prevention of problems induced or aggravated by alcohol, tobacco, or drug abuse. Contrary to earlier impressions that pushers cut some innocent children out of the pack and introduce them into a lonely world of drugs, drug use among school-age children has proven to be an intensely social activity. Researchers studying friendship patterns across time among secondary-school students in New York, for example, have found that the most closely correlated character- istic of close friends, aside from their age and sex, is their level of marijuana use. About half of the time, this pattern results from acquiring new friends with similar established patterns of drug use or nonuse; the other half of the time, existing friendships lead to subsequent matching of patterns of consumption or abstinence. Experimentation with different substances seems to be related mainly to environmental opportunities and broad social factors. Student drug use in junior and senior high school is mostly a friendly or party activity, which varies depending on the demographic characteristics of the students in the school and the overall peer perception of the harmfulness of use. But the evidence impressively supports one long-suspected factor: young people reared with poor self-images are especially attracted to intoxicants, finding in them a com- mon bond with other troubled youngsters, which may lead them together into increasing alienation from noninvolved peers and activities. Young people's use of intoxicating substances is known to follow a remark- ably standard sequence, going from alcohol to cigarettes, to marijuana, to pills

60 / The Behavioral and Social Sciences or powders, to intravenous drugs. But moving any distance in this sequence is neither inevitable nor typical. In recent years, more than 90 percent of high school seniors have tried alcohol, 70 percent cigarettes, SO percent marijuana, 17 percent cocaine, and 1 percent heroin. The proportions who use any of these daily is much lower: about 20 percent for cigarettes, S percent for alcohol, S percent for marijuana, and less than 1 percent for other substances. A strong curvilinear association has been discovered between age and the prevalence of narcotic (opiate) abuse, with use beginning between the 6th and 12th grades, cohort use peaking by age 2S, and among the continuing steady, heavy users, a rapid falling off after age 40. Adolescent friendship networks and general societal prevalence influence the onset of use. Formation of spe- cialized supply-focused groups is important to its continuance. Later associative transitions, especially marriage, tilt strongly toward the termination or sharp reduction of use in early adulthood. In the older years, a combination of de- clining health, more serious judicial jeopardy, and the prospect of permanent loss of social supports, such as employment, careers, and family relations, where these exist, lead to fewer users. Evaluating the Effects of Interventions Research on the effectiveness of preventive intervention against early alcohol, tobacco, and drug use has advanced considerably in the past decade, yielding more detailed understandings of developmental sequences and their condi- tioning at crucial transitions in the contemporary life course, while offering possible ways to reduce the problem. Net positive results have been demon- strated in some research-based health education experiments that capitalize on the willingness of students to behave with appropriate self-regard in the pres- ence of supportive peers. Such programs avoid the zealous overkill of many past programs and teach how to reject offers to share drugs, including alcohol and cigarettes, in socially graceful ways. These lessons have proven very cred- ible when a distinction is made between the risks of frequent heavy use and those posed by trial or occasional use, especially when this information is received from respected and properly trained adults and older youths. Effec- tiveness is clearly increased when these lessons are reinforced through a variety of channels, including mass media and voluntary group efforts as well as school curricula. Some of these experiments have now been expanded and modified into major national campaigns and commercially published packages. How- ever, the expansion and modification of these intervention techniques preceded expansions of evaluation research to determine if and how well the positive results hold up across different social environments and with other"teachers" than the original research-based intervention teams. There has been a substantial gain in understanding how education and media can affect the use of cigarettes. An important statistical finding is that cigarette sales per capita decreased most during the period (1967-1970) when nation-

Motivational and Social Contexts of Behavior / 61 wide "equal time" anti-cigarette-smoking television advertisements were in place. This decrease coincided with several other factors: rising cigarette excise taxes, more general information concerning the medical risks inherent in reg- ular cigarette smoking, and a shift in the normative acceptance of smoking and of the appropriateness of social intervention by physicians, family, and even strangers, based on increasing concern about health and fitness in general. Subsequent studies conducted on smaller geographic scales in rigorous, ex- perimentally controlled programs with special news features and direct ed- ucational efforts in small groups have confirmed the positive effects of"health promotion" advertising aimed at preventing smoking among adolescents as well as adults. The key questions now concern the transferability of such efforts, the persistence in their efforts, and the design of more efficient ways to achieve these objectives. Further experimental and longitudinal research is needed to answer these questions. Addiction and the Marketplace The character of addictive substances as market commodities is a relatively new approach that has proved very useful. Recent studies of alcohol, tobacco, and even heroin use have clearly shown that laws of the market operate for each of these substances: when prices rise there is an overall decrease in the amount of substance consumed, although the total amount of money spent on the substance increases. Data are available from the 1960s and 1970s, when real alcohol and cigarette prices shifted from year to year because of the seesaw effects of higher excise taxes and overall price inflation that cheapened these taxes. These data suggest a substantial negative relationship between price and consumption (elasticity figures of about - 0.5, that is, a decrease of one-half percent in consumption for each one percent increase in unit price) for alco- holic beverages and cigarettes. These elasticity results have important impli- cations for public health policy, because the level of consumption is closely related to rates of illness and death due to lung cancer, cirrhosis of the liver, drunk driving crashes, overdoses, and the like. New data show how different age and sex groups respond differently to price increases (such as occur when excise taxes are raised). For cigarettes, the effect of price increases appears to be strongest among 15- to 24-year-olds, and it is stronger among women than men. In general, the effect seems to be through changes in "participation" (that is in smoking or not smoking) rather than in "amount" (that is, the number of cigarettes smoked by those who do). The effect of alcohol price changes, however, appears strongest among males aged 20 to 50 and involves amount as well as participation. Confirmation of these findings and more details about how price changes affect participation and amount, and how different social groups respond, could be very useful in forming public policies.

62 / The Behavioral and Social Sciences Prevention of AIDS It is likely that between 1 million and 1.S million Americans have already been infected with the AIDS (acquired immune deficiency syndrome) virus. At present in the United States, AIDS is concentrated among homosexual and bisexual males and users of illicit intravenous drugs, and there is also relatively high incidence of infection in children born to infected mothers, recipients of blood transfusions, and sexual partners of those in these high-risk groups. The epidemic is spreading much faster in the black and Hispanic population than in the white population. However, present evidence indicates a likelihood that the virus is capable of spreading to a very large proportion of sexually active individuals, well beyond the groups now identified as high risk. The consensus among biomedical experts is that fully effective treatments and vaccines may not be available for at least 5 years and maybe considerably longer. In the interim, practical efforts to cope with the immense social, psy- chological, and economic problems of the AIDS epidemic will require contri- butions from the social and behavioral sciences in a number of areas, including ones that have been relatively neglected in recent years. One of those neglected areas is basic research on human sexual behavior. There is a renewed appreciation of the need for vigorous programs of research on such behavior, which has always been a sensitive undertaking in American society and which has lacked extensive public investment. But understanding the variety of sexual behaviors and how people integrate those behaviors into their lives will be crucially important in persuading individuals to modify their sexual behavior in order to diminish the risk of spreading AIDS. Lack of knowl- edge about sexual behavior also seriously handicaps current efforts to predict the future course of the AIDS epidemic in the United States. For example, there is a remarkable lack of reliable data on the rate and character of sexual contacts for the national population and for groups with elevated risks of infection. Even simple estimates of the size of the current population of homosexual men rely on data collected by Kinsey in the 1940s, which were considered suspect then due to the biased samples in Kinsey's surveys. There is an immediate, crucial need for better knowledge of human sexual behavior because of the AIDS crisis, and such research will have practical applications in many other areas, including fertility and population studies, teenage pregnancy, and treatment of sexual dysfunction. Better understanding of human sexual behavior will also substantially contribute to the study of other sensitive and powerful motivational and social factors that affect personal and health behavior. A second research area that is given added urgency by AIDS is the com- munication of health information. The goal of public education about AIDS is to increase public knowledge and understanding of the epidemic and to en- courage people to avoid behaviors that have high risks of spreading the infec-

Motivational and Social Contexts of Behavior / 63 tion. The most difficult challenges include producing effective and publicly acceptable education programs for youth; diffusing information to relatively hard-to-reach populations, such as cultural minorities intravenous drug users, and prostitutes; and providing information that fully informs without panicking the public, so that realistic public health policies can be debated and formed. Massive and decentralized public education programs will have numerous unique elements and may involve a slow reaming process with considerable trial and error. Rigorous evaluations of the effects of educational programs will be important in order to learn from experience and thereby improve the pro- grams over time. The technology for conducting evaluation studies does not need to be created, but it does need to be implemented correctly. What is learned will not only be relevant to AIDS, but also to numerous kinds of educational programs and to knowledge about social influences on behavior more generally. Three other lines of research acquire great practical importance due to the AIDS epidemic: (1) systematic historical study of earlier societal reactions to epidemic diseases, such as leprosy, tuberculosis, polio, and syphilis; (2) careful and continuously updated estimation of the costs and benefits of alternative intervention, treatment, and care strategies; and (3) microcultural studies of whether and how educational information, screening, intervention programs, and changing availability of devices, such as sterile needles or barrier contra- ceptives, may shift behavior to reduce the risk of transmission. A particular need under the last category is to determine how and under what contingencies individuals in different circumstances respond to reaming the results of serum testing for AIDS infection. Stress, Risk of Illness, and Behavior How are changes in physiological functioning (for example, the endocrine and immune systems) affected by social and behavioral factors? Psychological stress, which has a variety of negative effects on health, may be mediated through the emotional system, and recent work suggests that stress can ad- versely influence the functioning of the immune system. Other research has found that social support mechanisms can, to a remarkable degree, reduce a person's sensitivity to certain stressors and increase compliance with medical regimens. Psychological stress can markedly alter social roles and behavior patterns. For example, there is a direct association between bereavement and mortality in men, but not in women. Among the elderly, living alone is associated with poorer nutritional intake and poorer health for older men, but not for older women. Behavioral and social scientists have begun to struggle with the com- plex issue of how differing social support networks mitigate or contribute to a variety of disorders such as depression, schizophrenia, or cardiac disease. One

64 / The Behavioral and Social Sciences line of research examines how the effects of social networks on health outcomes for the patient are mediated by psychological mechanisms such as the percep- tion of personal control. This kind of factor appears far more significant than has been recognized by health care providers. A rapidly developing area of research is the role played by risk, vulnerability, and protective factors in the etiology, maintenance, and recovery from such diverse mental conditions as schizophrenia, affective disorders, and (in chil- dren) attention-deficit. disorder. Risk factors are those elements in the individ- ual, the family, and the environment that are associated with higher incidence rates of such disorders in the population. Vulnerability refers to a condition of the individual, which is sometimes reflected by palpable, specific markers and sometimes conceptualized as a generalized or specific predisposition to path- ological outcomes. The study of risk and vulnerability factors offers insights into the potential power of different predictive models of etiology of disease while providing, in turn, a rational basis for norms of preventive intervention with those individual who are prone to the development of mental and physical illness. Help-seeking channels and mutual-help groups are related contextual influences on health behavior that have come under increasing research scru- tiny. There are many studies on the deleterious effects of job loss, such as studies finding an increase in serum cholesterol in laid-off workers following plant closings. The most dramatic research traced the existence in some individuals of psychological depression for up to half a century after the economic de- pression of the 1930s and its job disruption. Most recently there has been increasing interest in studying the neuroendocrine correlates of stress, in an effort to understand the physiological mechanisms by which environmental events affect physical health. For example, recent biobehavioral research sug- gests that psychophysiological responsiveness (reactivity) to emotional stress may be a marker of processes involved in the development of cardiovascular disorder. Recent investigations have also examined sex differences in neuroen- docrine responses to stress, as well as the effects of occupational stressors on adrenocortical and adrenomedullary hormones. A number of studies have be- gun to consider the regulatory effect of the central nervous system (CNS) on immunity. Under consideration are endocrine, neurotransmitter, and opioid influences as modulating processes, as well as a direct neuronal link between the CNS and immunocompetent tissues. In addition to regulatory mechanisms, attention has also been directed to feedback processes between the immune system and the CNS. These areas of research have been greatly facilitated by the availability of new bioassay and measurement techniques. Behavior and Health Care Delivery Systems Research on health care financing has taken on major importance as health care costs have risen to a level exceeding the national defense b fidget. A sig-

Motivational and Social Contexts of Behavior / 6S nificant step in that research was a large-scale social experiment initiated in the 1970s, involving 7,700 individuals in six areas of the country, that yielded important information on the specific consequences of alternative forms of health care organization and services. In one aspect of its design, the study compared the styles of medical practice (especially, hospitalization practices) of a health maintenance organization (HMO) versus a traditional fee-for-service system. Many observers view HMO medicine as cheaper, but skeptics have charged that this is an artifact of selection bias- namely, of enrolling generally healthier clientele from the outset. To test the issue, individual families were assigned randomly to each system, in both cases paid for entirely by third parties. Analysis of the results showed that the families in HMOs still had 40 percent fewer hospital admissions over a 5-year period than the fee-for-service group; this difference translated into a financial savings of 28 percent. The health-status consequences of this organizational difference and the relative satisfaction of the clients are now being analyzed. On another issue, analysis of medical follow-up data obtained during the experiment showed that members of families who were assigned to health insurance plans that paid all of their medical costs made substantially increased use of services about 40 percent greater than those in plans with a sub- stantial deductible (as high as $1,000 depending on family income). For the free-care group, there was measurable improvement in health outcomes for low-income people with certain problems, such as bad eyesight, decayed teeth, or, most seriously, high blood pressure. But for participants not afflicted with these particular problems, no overall improvement in health was evident de- spite the increased use of medical services. Much of the increased use, such as treating relatively minor injuries in the emergency room rather than at home (for example, cuts not needing stitches), was of minimal benefit. On balance, the data strongly suggested no single experiment or study can be said to prove that it would be cost-effective to improve health care by further sub- sidizing medical treatment for people afflicted with certain health deficiencies, but the complete removal of economic considerations in access to medical services would not be a cost-effective way to improve the health of the popu- lation. CRIME AND VIOLENCE The problem of crime has been in the forefront of national attention for several decades. Surveys have consistently shown that Americans rank it among the nation's most serious domestic problems. This has been especially true for certain serious crimes involving violence or invasion of the home murder, forcible rape, aggravated assault, robbery, and burglary. Few demands on the behavioral and social sciences are more insistent than the call for knowledge on how to reduce the threat of these dangerous acts.

66 / The Behavioral and Social Sciences At the same time, criminal and violent behavior is a challenge and an op- portunity for scientific research. Why do people help each other or hurt each other? Why do they voluntarily conform to some, but not other social norms? To what degree and under what conditions is behavior controlled by material or biological needs, character traits, internalized values, or calculations of moral, economic, or physical consequences? Competing theories of behavior face stern tests in having to account for the "deviant cases" of predatory crime in largely civil, law-abiding societies. Because of the inherently stealthy nature of crime, direct systematic obser- vation of it is rarely possible, and the systems for reporting crime have generally been weak and nonuniform. In recent years, however, the main source of data—the police-generated FBI Uniform Crime Reports- has been supple- mented with two new sources: self-report surveys of individuals who, on a confidential basis, describe their own criminal activity; and victimization sur- veys, which report the experience of crime by the general population. Although all three data sources have limitations, they are now available for mutual val- idation and calibration, and it has been found that together they yield infor- mation that is consistent or can be readily reconciled. Criminal Careers and the Effects of the Criminal Justice System One of several promising approaches developed on the basis of interpreting these results in recent years has been to focus on criminal careers, that is, the development of criminal and noncriminal behavior as a life-long process. A1- though sporadic participation in crime is most common among adolescent males, nearly half of all urban-dwelling American males can expect to be ar- rested for some kind of nontraffic offense at some time during their lives. Yet all three data sources show that a relatively small number of"career criminals" or "violent predators" are responsible for a grossly disproportionate percentage of the crimes committed. A few people begin criminal activity quite young some as early as 8 years of age. The highest rate of criminal activity in the general population occurs in the mid to late teens. But there is a heavy drop- out rate from adolescent criminality, so that by age 30 or so those still com- mitting violent crimes are largely the very serious, resistant, high-rate offenders who are unlikely to abandon their criminal activities much before age 40, if then. Research has not yet provided information about the dynamics and rea- sons for termination. There is a strong nexus between drugs and crime. Regular users have much higher rates of criminal activity than nonusers or those who have discontinued use, but the directions of causality between drug use, other criminal behavior, and other age-related processes are still ambiguous. There are also important questions about the way in which the criminal justice system affects the course of criminal careers. During the twentieth cen-

Motivational and Social Contexts of Behavior / 67 fury, the principal strategies to prevent crime have been rehabilitation, deter- rence, and incapacitation. Studies of prison-based rehabilitation have largely reported failure whether because there are barriers to the use of theoretically sound techniques or because there are flaws in the techniques themselves is not always clear. And while deterrence is effective, research on the structure and operations of the criminal justice system shows that it is very difficult to increase deterrence much beyond present levels. Simply stiffening the penalties on the books does not translate into a tougher, more deterrent crime system at the operating level. More policing, including recruitment of citizen-watch groups, and additions to jail capacity, is helpful, but limited in extent. Incapacitation is at present attracting substantial scholarly interest, and con- siderable research has tried to measure quantitatively the cost-effectiveness of imprisonment. While a jailed burglar cannot break into homes, jail is an ex- pensive proposition. If a 22-year-old burglar whose criminal career is likely to terminate anyway in S years is imprisoned for 10 years, the last S years of jail occupancy may be "wasted." Considerations other than cost-effectiveness clearly matter a great deal in setting correctional policies, but in taking these kinds of calculations into account, it is very important to have precise data on criminal activity. Knowledge of the "normal" course of criminal careers may provide a useful tool for measuring with greater accuracy the social benefit of incapaci- tation. Since criminal careers occur over time, longitudinal panel techniques are now likely to yield more powerful results than additional cross-sectional stud- ies. Some of the main understanding of criminal career patterns came about from early panel studies that began in the 1930s. Additional studies are now needed to fill out in much greater detail the main points of the criminal life cycle: early childhood development in the family or institutional context; the onset of offending, which may be as early as middle childhood; the age of peak participation or prevalence in the mid to late teens; and the settling into and dropping out of criminal careers at later points in the life cycle. Ideally, such studies might follow a single cohort panel from birth to age 2S or older. The problem with this approach, however, is that it takes decades to complete to get the full payoff in knowledge. An alternative approach is to select several cohort panels of different ages, following each of them for a period of several years to reach overlapping ages, and to ascertain the relative effects of criminal justice interventions. The complex design of such studies, based on interviews, observations, and records, now is being undertaken. Design includes the development of specialized institutional arrangements to collect, maintain, and govern access to the data and, when possible, to link such longitudinal designs to randomized field experiments involving shifts in pol- icing or judicial operations, family-based or school-based interventions, or other changes that may affect criminal behavior (see "Experimental Design," Chapter 5~. Additionally, comparative studies of crime and criminal careers in

68 / The Behavioral and Social Sciences other countries and during other periods of history can provide a richer context and help overcome some interpretive limitations on cohort studies, enabling researchers to pinpoint which factors are specific to this particular period and which are more general to understanding the origins of criminal behavior. Antisocial and Prosocial Dispositions An area that needs to be examined much more systematically concerns the effects or criminal activity of such key life events as marriage, getting a job, Joining the military, moving to a new community, and the like. A negative association has been shown between having a legitimate job and committing violent crimes (for adults), and there is evidence on the relationship between certain kinds of family background and a career in crime. The influence of family structure and dynamics during childhood on criminal behavior seems very important. In correlational studies based on regional statistics, commu- nities with high divorce rates also have high rates of criminal activity. But these data cannot reveal whether it is actually the children of divorced parents in those communities who are responsible for the crimes. Children with only one parent because the other has died are less likely to be involved in offenses than children whose parents are divorced, which suggests that the existence of marital conflict surrounding divorce, rather than the fact of losing a parent, may matter most in this connection. Children of criminals are more likely to commit offenses than children of noncriminals. In recent years, there has been substantial research interest in the problem of family violence, but there is as yet insufficient knowledge of the way in which a history of child abuse or other forms of family violence may be connected to later criminal activity. Researchers now want to use more sophisticated techniques to identify the exact relation- ship between family conflict and later criminal activities, as well as the effect of strategies of intervention. One likely factor in the development of antisocial behavior, by at least some adolescents, is how their parents typically try to manage their children's be- havior. In one study of adolescent males, measures of parental behavior were categorized as monitoring the young man's whereabouts, disciplining him for infractions of various sorts, teaching him problem-solving skills, and reinforc- ing desired behavior. The adolescent's behavior was measured in the areas of delinquency (as indicated by police and self reports), academic performance, and social relations with peers. The study found that higher rates of the (in- tercorrelated) parental behaviors of monitoring and disciplining were associ- ated with decreased delinquency; academic performance and social relations were strongly associated with parental teaching and reinforcement; and the other pairings showed little or no association. That these are probably causal connections was established by training the parents of chronic delinquents to

Motivational and Social Contexts of Behavior / 69 do what other parents typically do about monitoring and discipline; the result was a significant long-term reduction in the delinquent behavior. Another line of relevant research concerns the linkages between empathy, prosocial motivation, and aggressive behavior. Empathy is a class of emotional responses to someone else's situation rather than to one's own. Research has demonstrated that most people respond with empathic distress to someone in danger, discomfort, or pain. There is also evidence indicating that empathic distress is largely involuntary: it is difficult to avoid empathizing with someone in pain or distress without actively engaging in certain evasive perceptual or cognitive strategies, such as looking away from the victim or trying hard to think about other things. It has been clearly demonstrated that empathic distress contributes to pro- social, helping behavior in most people. Such helping behavior is usually mo- tivated internally rather than by the desire for reward or approval (although these motives may also be present); moreover, the occurrence and rapidity of helping behavior are related positively to the intensity of empathic distress. When people give help the intensity of their empathic affect is diminished. An as-yet small body of research suggests that empathy may also reduce unpro- voked aggression in children and that empathy training can reduce the ag- gressive behavior of adolescent boys with delinquent tendencies. The effects of empathy are, of course, limited. For example, the relationship between empathy and helping behavior is often mediated by whether the per- son in need is viewed as similar to the observer (empathic bias). But there is evidence that empathic bias may be reduced by educating people about what they have in common with members of other cultures or groups, as well as with other individuals in their own group. The potential fields of application of empathy research include childrearing practices, the school setting in and out of the classroom, interactions with peers, and the content of television and other media. The challenge is to learn enough about how empathy develops and functions to permit advice to parents, media, and training programs on how to increase the empathic potential of delinquent or criminally prone youth (and others), whose sensitivity to other people may have been blunted by precious life experiences. Apart from this value, a better grasp on the conditioning of a human response tendency like empathy may illuminate the major cohesive, as well as divisive, forces in society. ATTRIBUTIONS AND EXPECTANCIES IN SOCIAL INTERACTION The concepts of self-esteem, social networks, and gender-based differences have been vital to research advances on emotional maturation, health risk and protection, and criminal behavior, discussed above. They have also been used

70 / The Behavioral and Social Sciences to guide research on topics such as educational attainment, language devel- opment, and mental frameworks, a range of topics taken up in Chapter 1, as well as research on bargaining, employment relations, and fertility behavior, discussed in Chapters 3 and 4. These theoretical concepts about social-psychological mechanisms have a familiar ring to them. The terms themselves seem part of today's universal vocabulary, and it is easy to assume that they are commonsense ideas that have been handed down from generations. But the terms, and the nuanced ideas they represent, were actually invented by behavioral and social scientists not long ago, tested and polished in classic theoretical and empirical work, and then became diffused (as are other scientific ideas) through the broader culture through education and the mass media. The concepts have since been adopted so fully, applied so widely, and their very novelty forgotten so rapidly as to seem truly everyday knowledge. In the past, experimental research on interpersonal relations tended to focus on isolated components of what is an interactive process. In the interest of sufficient experimental control, participants were often strangers who were prevented even from directly seeing each other. More recently, however, the study of social interaction has shifted from the identification of particular factors implicated in certain products of social interaction a happy or unhappy mar- riage, a fast or slow work group, a successful or unsuccessful negotiation to an understanding of the process. The emerging paradigm is the simultaneous study of many elements, freely varying in naturalistic settings, assessed over long durations, involving participants who have both a joint history and an anticipated future together, and using multiple methods such as self-reports, observer ratings, behavioral codings, and psychophysiological measurements. This change in the study of social interaction has been possible because of developments that include technological advances, such as high-fidelity video- tape, psychophysiological recording, and powerful microcomputers; advances in statistics, including new techniques for time-series analysis, model estima- tion, and refinements in multiple regression and analysis of variance; and ad- vances in theory formulation, including computer simulation as a medium for writing more formal, quantitative theories. As a result, researchers are now beginning to make substantial new inroads on such questions as when erro- neous expectancies (for example, stereotypes about race or gender) are most likely to be perpetuated, which interactive features improve the efficiency of task groups, or what principles govern the growth of initial encounters to close relationships. Expectancies, Self-Concepts, and Motives In an important experiment in the 1960s, some teachers were led to expect that certain of their lst- to 6th-grade students would "bloom" academically.

Motivational and Social Contexts of Behavior / 71 Although the designated students had been randomly chosen, subsequent test- ing following a period of instruction showed they in fact did better than stu- dents who had not been designated in that way. Although this study aroused considerable controversy and was criticized on a variety of methodological grounds, subsequent research has generally supported its conclusions. The findings were important not merely in showing that teachers were able to influence students, but also because the teachers, operating with an incorrect hypothesis, behaved in such a way as to bring about the confirmation of the hypothesis. And the teachers were not just misled by the false expectancy to misjudge the performance of the designated students; they created a situation in which the students indeed did perform better by independent, objective measures, and the teachers were unaware that they had done anything special to create this state of affairs. In its general form this concept is known as a "self- fulfilling prophecy." Research on self-fulfilling prophecies shows the manner in which features of personality, at least in terms of how a person appears to outside observers, develop and change over time through social interaction. Rather than being preprogrammed and static, like a phonograph record, personality is, to a con- siderable degree, constructed and reconstructed through relations with others. This phenomenon has been clearly shown in studies of aggressive boys. This work identifies a process in which these boys expect aggression from their peers and accordingly interpret ambiguous provocations as reflecting definitely hostile intentions rather than seeing them as neutral probes or as possibly accidental. This leads the boys to retaliate aggressively, a reaction that confirms their reputations for aggressiveness and elicits peer counterreactions that fur- ther strengthen the boys' beliefs about the hostility of those around them. Unfortunately, however, this work has not yet discovered how such cycles begin. An important advance along these lines would be longitudinal research that permits the separation of intrapersonal and interpersonal factors. The subject of motives and how they are internalized is also an active area of current research. It has been shown that providing auxiliary rewards for performance of an initially attractive activity seems to rob the activity of some of its intrinsic interest. In a classic controlled experiment, for example, children showed a decline in their interest in playing with "colored markers" in a free- play period that took place some time after they had been rewarded for playing with them. Subsecuent experiments have shown that this decline in intrinsic motivation does not occur if the reward is given as a sign of competence or excellent performance; in fact, intrinsic motivation can be enhanced under these circumstances. The effects of reward structure on intrinsic motivation are also mediated by differences among individuals in their motivations to be competent. A particularly important feature of human social interaction is that people intuitively understand many of the principles governing social processes and,

72 / The Behavioral and Social Sciences as a consequence, are capable of manipulating them for their personal purposes. Research of the last two decades has examined the skills that people use in presenting their personalities to other people. For example, for some people whose self-image is competent, if that self-image is threatened, then they will try to protect it by handicapping themselves in ways made available by the environment, so that they have an excuse for poor performance if it occurs and gain credit from good performance if that occurs. In the case of strategic self- presentation, in which the personality presented to others is not in line with the person's self-concept, the attempt at social deception can be unmasked by subtle features of facial expression. A major line of research in social interaction focuses on the microprocesses through which information is communicated nonverbally, and sometimes unconsciously, about personality. Development of Close Relationships After years of studies restricted to individuals in isolation or to relationships between (usually) two people who had never seen each other before and prob- ably never will again, research on relationships is moving in new directions. Changes in the internal and interpersonal processes that accompany the emerg- ence of an enduring relationship are now being seen in laboratory studies of interaction between people who expect that their initial encounters will be followed by more lengthy ones. The earlier research had shown that, in situ- ations where experimental interactions between subjects were entirely casual and without any long-term consequences, people tend to rely on stereotypes and first impressions. More recent research has shown that even the possibility of additional future interaction has dramatic effects on individual behavior, both before the first contact takes place and during its initial stages. The anticipation of future interactions appears not only to heighten attention to the other person, but it also to contribute to the development of a more individuated impression of that person. For example, one recent study shows that a person anticipating future interaction with someone else pays more attention to information inconsistent with his or her preconceptions about the other person and tends to process that information carefully for its implications about the other's traits and attitudes. In contrast to tendencies to discount such information, as in the intrapersonal processes supporting self-fulfilling proph- ecies, the person who anticipates a future relationship frequently relies on unexpected information to revise his or her initial impressions. In laboratory studies, the possibility of a future with another person changes the format of interaction regarded as most appropriate from one of "social exchange," in which a benefit is given in response to receipt of a benefit, to one of"social responsiveness," in which the benefits given take into account the perceived needs of the other person. The details of interactional differences under these two circumstances are being explored; people sharply distinguish

Motivational and Social Contexts of Behavior / 73 the two kinds of relationships and have characteristic ways of expressing their different levels of responsiveness. The shifts in motivation as relationships develop are captured in recent theories that analyze systematically the initial patterns of incentives and the formal transformations that may be performed on them. Some measures of the changing nature and quality of social interactions over extended periods of time are provided by research on shifts in levels of satis- faction over the course of an enduring marriage or career. For example, marital satisfaction is typically high early in marriage, decreases with the birth of the first child, reaches its nadir as the children enter adolescence, and increases as children leave the household. Work satisfaction increases until approximately age 40, levels off through the mid-50s, and rises again thereafter. There are a number of competing explanations for these trends. The investigation of spousal, parent-child, coworker, and worker-supervisor relationships at different phases in the life cycle is the next step in understanding social interaction processes as they evolve over long periods of time. Small Groups and Behavior Frequently, large organizations such as governments, corporations, or uni- versities create small, temporary groups to carry out specific tasks. The capacity to form, activate, and dissolve such ad hoc committees or working parties is a major adaptive resource of organizations and societies and has become a fron- tier area for research on decision making and other fundamental group pro- cesses. Some of the research conducted on ad hoc groups is designed to examine the scope and variation of basic propositions concerning group influences on behavior. One important instance is the following effect: when an individual is asked in experiments to perform simple but physically demanding tasks, he or she typically exerts substantially less effort when led to think that a group or people is doing the tasks at the same time than when led to think he or she is performing them alone. While the total work done increases steadily as the actual number of people in the group increases, the average effort exerted by each individual decreases steadily with group size. This type of research pro- vides a baseline for studying the conditions under which various kinds of incentives induce greater or lesser individual productivity: How much does the effect obtain when the labor involved is mental rather than physical? When the group is comprised of friends rather than strangers? When each individual's task is unique? And what is the trade-off between greater individual effort and greater need for coordinating differentiated tasks? Research on these questions is obviously important for understanding productive efficiency in many set- tings. In a different vein, a series of experiments brought together strangers who

74 / The Behavioral and Social Sciences 12 2 8 - o ID ~ 4 At O l o Am_ · Flrst dinero. murder O Second dogr" murder /` · Manslaughter O Not guilty _~ Am' _ _ - 1 1 1 1 1 60 Of 120 Deliberation Rime (minutes) 180 240 JURY DECISION MAKING How do trial juries reach decisions? What is the effect of different agendas for making a decision, of changing jury sizes, of varying decision rules for example, unanimity versus majority? Among the new experimental methods for studying collective decision making is the simulated jury trial, in which a group of people are selected from actual jury panels to watch a videotaped trial proceeding and then, in an actual jury room, deliberate and reach a verdict while their discussions are recorded. This figure illustrates the waxing and waning of groups of jury members speaking or voting for particular verdicts across the four-hour deliberation of one such mock jury deciding a murder trial. The pattern displayed in this 12-member, unanimity-rule jury deliberation is associated with the most common of several observed decision agendas: "presumed innocent." The jury first addressed basic evidentiary issues: whether the defendant was guilty of any of the charges. During this period, the members were evenly split in a typical pattern among five possibilities— not guilty, manslaughter, second-degree murder, first-degree murder, and undecided. During this "evidence-driven" phase, few shifts of opinion occur concerning the proper verdict. In a series of votes after about two hours, the jury decided that "not guilty" was untenable, and the debate shifted directly to the verdict. Faction sizes then began to shift dramatically. The final ver- dict, second-degree murder, was the modal one for all mock juries viewing this trial. It was also the one favored by most judges and attorneys, and it was the verdict of the original jury at trial. The pattern shown in this figure, and the probability of reaching a par- ticular verdict, would differ if the size of the jury were smaller, the decision required a certain majority but not unanimity, or the jury adopted a different discussion agenda, such as "murder (first or second degree) versus non- murder (manslaughter or not guilty)."

Motivational and Social Contexts of Behavior / 75 differed noticeably with respect to race, age, sex, social class, and language use attributes differentially valued in the society. The people were presented a series of perceptual problems to be solved (for example, does a rectangle enclosing smaller black and white rectangles contain more black area or white area?) for which their status characteristics were objectively irrelevant. In the absence of prior information about one another's task-relevant skills and knowledge, persons with high-status characteristics were often attributed more of the required skills than were low-status persons. Consequently, they were allowed to initiate more interactions, and their solutions were more readily accepted by the group. In short, the status ordering produced within the group reproduced the external social order rather than responding to the problem at hand. An important and revealing group process occurs in trial juries, which vary from one jurisdiction to another in size and in the proportion of assent required to assert a verdict. How do differences in size and rules- say, a unanimity rule in a jury of 12 compared with a 6-person majority in a jury of Affect the decision process and outcome? One can, of course, speculate, and jurists have, but there has been little empirical knowledge to support these speculations. What goes on in a jury room has not been subject to close examination except by later reconstruction of the proceedings from the memories of jurors, which are known to be fallible at the level of detail needed. However, carefully de- veloped studies of experimental juries have now yielded powerful new knowl- edge. In an example of such experiments, several panels of experimental jurors, selected from actual jury pools, watched a videotaped enactment of a real trial, and then met in regular jury rooms to decide on a verdict. Among other things, the group size and decision rule employed were varied from panel to panel. The proceedings were recorded on videotape (with the jurors' knowledge) and then subjected to intensive computer-aided analysis. One major finding con- cems different decision rules. Researchers found that when unanimity is re- quired, the deliberations are far more thorough than when it is not required. A large fraction of the additional discussion, uncovering and eliminating serious errors of fact and law, occurs after a decision would have already been reached under a plurality rule. Moreover, nonunanimous juries more often reach ex- treme verdicts (for example, first-degree murder rather than second-degree murder) than unanimous ones do. With a nonunanimous rule, members of small factions also contribute less to discussion, and larger factions attract new members more quickly. This finding is particularly significant because in other contexts it has been shown that the outcome of a group discussion is deter- mined by those members who frequently shift their opinions, rather than by . . . . . . . those Wit :1 more extreme 1nltla . positions. These findings focus attention on the dynamic interactions among group members rather than on dependencies between pairs of variables measured at only two points in time. They have led to the development of more complete

Z6 / The Behaviom1 and Social Sconces cheoredcaI accounts of the dynamics of interaction in groups. They have also been ched by legal policy makers concerned with constitutional guarantees of due process and the selection of representative juries. Such research on group decision making exemplifies the theoretical and p~cdcal conthhutions Tom research on social interaction. The Social Coustruchon of Gender It is becoming increasingly clear that human sexual differences are actuary composhes of Enable elements There are chromosomal differences Tar tamale and far male mammals), anatomical differences, physiological differences, psy- chological differences, behavioral differences, and sociocultural differences. The latter three categories are generaDy studied under the rubric of gender research, With several components: primp gender idend~ (~ or Embed partner choices (heterosexual, homosexuals gender-relevant behavior styles (~male Tom boyishness, nary"), and " male"e~mi sexuaHy dimorphic nonsexual capacities Jar example, geomethc ability). Recognition that gender can be heated as a beha~om1 and social construct quite distinct (though not necessarLy divorced) Tom biological differences has led to important advances in a number of gelds. New Guinea has proven to be something of an anthropological laboratory on gender differences. For decades, this island was know to house cultures supply some of We eddy mom ~odc g=~r band and peachy: dmab iced homosexuals, elaborate notions of menstrual pollution, ceremonies of sex-role reversal and the most extreme know doctrines of male supremacy. Analysts relying on earDer theories Were unable to explain or interpret these unusual beheL and pmchces. More recently, however, researchers working with social-construction the- vies have begun to unravel a consistent native logic underlying these exotic dam, detailing the may in which New Guinea peoples use Mod, sexual ac~vides, and dtuaHy concocted substances to think about and manipulate kinship sod gender identities, health and disease, lid and death. In some areas of New Cuinea, far example, boys are druaLy "ground into men by Ceding them male- gro~n or male-hunted Mods: in other areas, by putting semen on or in their bodies: and in yet other areas, by ritually bleeding them to dd them of "~male blood." In even case, a logic comes to the surface concerning social identities that can be manipulated because they are bound up with substances that can be manipulated. ~itchcrak bebeL and notions of kinship have also been shown to link consistently with this logic. Furthermore, and more Musingly, re-

Motivational and Social Contexts of Behavior / 77 gional variations for example, between an emphasis on adding male sub- stances and an emphasis on deleting female ones have been shown to covary systematically with variations in political organization between different tribal groups. Researchers trying to understand social change in Europe have also begun to adopt a social-constructionist approach to gender systems. Western history is not without sexual exoticisms the nineteenth century has attained a status among social historians similar to New Guinea in anthropology. The Victorian evidence shows the extent to which notions of gender may be reconstructed in response both to received ideas and to emerging social, economic, and political patterns. For example, outbursts of peculiar sexual theorizing (say, that masturbation causes insanity or that sexual restraint is analogous to capital accumulation) have been shown to be related to stressful changes in family structure and class relations. The interpretive methods leading to these breakthroughs in understanding gender systems in geographically and historically diverse locales are just be- ginning to be fumed toward understanding gender systems closer to home. A growing focus is the complex interrelationship between what goes on in the public place of work and the private place of the family. One of the most challenging questions is to explain the motivational force and historical resi- lience of gender beliefs: for example, that "women's place" is in the home or working as a nurse or elementary teacher, which draws upon their"natural" capabilities, while men are "better suited" for work that requires highly rational thought or the exercise of authority. To the degree that earlier research on gender behavior was heavily influenced by such traditional models of male and female behavior, it tended to leave unstudied a large range of behavior that did not fit the stereotypes, such as the diverse activities of women in the public domain and that of men in the domestic or private domain. Over the last decade, under the impetus of new theory and its implications, researchers have compiled a much improved record of the public and political activities of women: biographies of hitherto little-known activists; data on women's labor organizations and political action in the workplace; data on women's rights movements at different points in American history; and descriptions of reform movements directed at broad social issues temperance, slavery, and peace. Research on men's domestic activities lags substantially behind this work, but when more data are uncovered, one can anticipate a thorough sorting out of competing theories and stereotypes in light of a more complete historical re- cord. OPPORTUNITIES AND NEEDS The anatomy and control of motivational processes; the codification of emo- tional expression; and the understanding of eating, sexual, and other behavior in animals and humans have all been illuminated in recent years by continu-

7~3 / The Behavioral and Social Sciences ously improved technology, better controlled research designs, and more so- phisticated theoretical ideas. Research has given insights on the role of peer pressure, mass media, and market forces in changing societal levels of alcohol, tobacco, and drug abuse. Controlled policy experiments have revealed effective strategies for designing medical cost-control measures without detrimental ef- fects on health. The career dimensions of criminal activity have been opened to study, and important lessons have been developed for crime prevention and control. The study of small group processes has been enriched by innovative experimental approaches such as simulated jury trials facilitated by video re- cording equipment. The subtlety and deep-seated effects of gender differences and divisions have been clarified by sustained cross-cultural and historical investigation. Research in these areas often involves complex interactions with large, well-established professional groups and organizations: for studies of biobehavioral aspects of health, health care providers, physicians, and hospi- tals; for studies of criminals, the police, the judiciary, and the penal system; for realistic studies of juries and the courts. In each case, complex questions of access, confidentiality, legality, and skepticism must be answered before any study can even begin. In the case of longitudinal studies, these issues must be repeatedly confronted, often in the context of changing institutions and tech- nologies. We therefore propose new expenditures of approximately $56 million annually to make possible further developments in this research. One of the outstanding needs is to renew the technological bases for labo- ratory and field research, for which we estimate about $10 million per year is needed. The tools of neuroscience, especially in the realm of surgical and physiological equipment and neuroimaging devices, have undergone revolu- tionary changes in precision, skill requirements, and expense. New generations of audio and video recording and synthesizing equipment permit more elegant research designs and aural and visual displays than were possible when ex- perimental stimuli and theoretical models had to be constructed, presented, and controlled by hand or with mechanical or photographic media. Recent improvements in physiological measurement technology using biochemical and endocrinological assay techniques now make it possible to simultaneously examine behavioral and biological states and responses and to do so in social settings such as homes or workplace. Recent statistical innovations permit investigators of interaction and communication to examine reciprocal causation over time more readily and precisely than in the past. At the same time, ad- vances in video technology permit more naturalistic social interactions to be recorded cheaply and preserved indefinitely so that the qualities of interper- sonal relationships can be sampled over long periods, which is increasingly important as more investigators move from the study of first impressions to the study of longer-term close relationships. Finally, the revolution in micro- electronics has made mainframe computing capacity available in desktop or mini machines, affecting practically every kind of research.

Motivational and Social Contexts of Behavior / 79 While new equipment for behavioral and physiological measurement and analysis is much more powerful and flexible than that available only a few years ago, budget limitations on equipment acquisition and methodological training have left many laboratories and field stations in a state of actual or potential obsolescence. The laboratory facilities in most university departments were developed and outfitted before the advent of modern video, computer, and other technologies. The present stock and potential demand for modernization of laboratory equipment needs to be systematically canvassed, especially in multiuser laboratories, to determine an appropriate set of goals and schedules for upgrading. Provision for these facilities should also include technical sup- port, training in the methods and analytical techniques appropriate to new technologies, and graduate and postdoctoral research opportunities, discussed below. Of the $10 million more annually that we estimate is needed for tech- nological upgrading, we estimate that $5 million should be added to laboratory equipment expenditures, $2 million more specifically for neuroimaging tech- nology acquisition and access, $2 million for computer hardware and for soft- ware development, and $1 million for upgrading of research animal care. The need for intensive technical training in the operating skills, underlying principles, and new theoretical possibilities linked to the technological ad- vances requires an infusion of support for research fellowships, traineeships, and institutes. Overall, we believe that $8 million annually should be added for this training. At the graduate level, the need is not to increase the overall number of graduate students, but to reduce their heavy reliance on teaching and outside income during these years, so that as new doctoral scientists they will have research experience well beyond the boundaries of their own disser- tation topics. At the postdoctoral level, the need for additional training oppor- tunities should be more closely integrated with the importance of building and sustaining collaborative affiliations and countering the fragmenting effects on science of rapid growth in knowledge and shifts in technique. A sharp increase in fellowship support is needed to achieve the kinds of technical skills required in some of these areas, and leads to a recommendation of an increment of $5 million, with the majority ($3 million) allocated to postdoctoral support and the minority ($2 million) to predoctoral support. We also recommend that $3 million be committed to advanced training institutes. The value of longitudinal studies—including those with experimental com- ponents has been thoroughly proven by outstanding work in each of the research areas discussed in this chapter. Prospective longitudinal studies are usually the most effective way to uncover and generate certainty about causal relationship with extended temporal or developmental structures. But the time between their design and the reaping of their results can be as much as from 20 to 30 years, during which time there are inevitable improvements in tech- nique and changes in the scientifically and practically significant questions. To some degree an ongoing study can accommodate changes, but some shifts in

80 / The Behavioral and Social Sciences knowledge and interest will be of sufficient magnitude to warrant beginning new studies on a topic before older ones are completed. The past decade has seen a decline in the rate of new longitudinal studies, cutbacks in the frequency and scope of data collection in several mature studies, and high rates of inactivation in both new and mature projects. It is time to revitalize this critical aspect of the research system. Commitments are needed for longitudinal studies of normal affective and motivational patterns, emo- tional maturation, and the incidence and effectiveness of treatment across the lifespan for affective illness; risk factors and preventive interventions (including marketplace variables) with respect to alcohol, tobacco, and drug use and sexual and dietary patterns; the role of social networks in affecting health, particularly later in life; the development of criminal behavior in the critical period from middle childhood to young adulthood, especially in the context of family behavior and criminal justice policies; the development of close per- sonal relationships; and the development of changes in gender roles and re- lationships. For this substantial agenda we recommend a new commitment of $15 million annually, which will permit approximately 10 large-scale longi- tudinal studies to be carried out. It is essential that the multidisciplinary character of research on motivation, behavior, and social contexts be recognized and appropriately supported by research agencies. To do so will require some reconsideration of relevant review processes, broadening of administrative protocols to encourage collaborative research efforts, and review of staffing for extramural research. In particular, one of the major routes for advancing collaborative research is the creation of various types of research workshops. They can include brief, recurrent meetings of a group of active collaborators and extended workshops (4-6 weeks) at which research is reviewed, planned, and carried out. For these purposes, we recommend an annual increment of $1 million. The staffing requirements of support programs that try to encourage, properly review, and adequately mon- itor collaborative and multidisciplinary work are more demanding than single- investigator review and funding operations. We strongly encourage modest increases in the numbers of program staff in funding agencies so as to permit greater attention to solving the problems of integrating research portfolios, encouraging grantee interchange and collaboration where appropriate, and cultivating a broader knowledge base about the scientific and administrative opportunities, benefits, as well as pitfalls of collaborative or at least more re- ciprocally informed and mutually responsive research. Of particular concern is the segmentation of research on health matters according to disease-specific missions. This segmentation tends to inhibit the basic study of common biobehavioral and sociobehavioral processes and, in- deed, of health-related behavioral and social processes on the whole. There is a need across the health research institutes in the National Institutes of Health (NIH) and the Alcohol, Drug Abuse, and Mental Health Administration

Motivational and Social Contexts of Behavior / 31 (ADAM HA) for increased coordinated research on common biobehavioral processes relevant to health. Such projects need to be reviewed by interdisci- plinary research specialists; the skills and perspective of such specialists cannot be simulated simply by organizing review panels with researchers from different disciplines. In addition, at the National Science Foundation (NSF), review practices discouraging the support of studies with clinical samples should be reconsidered. While NIH devotes substantial efforts to support of fundamental biological research that involves clinical subjects, this is not true of behavioral and social sciences research support from NIH. It would not be duplicative for NSF to increase its investment in such research. There is a growing sense among researchers on affect and motivation that the establishment of national multidisciplinary research centers, in which teams of investigators can combine diverse approaches and methods, is one of the best means of advancing research on these processes. Such centers could bring together researchers who do not now communicate much, particularly those concerned with normal affective and motivational processes (principally psy- chologists) and those concerned with the causes and treatment of affective and motivational disorders (principally psychiatrists and neurologists). An annual expenditure rate for such centers of as much as $9 million is recommended. Finally and in spite of the emphasis given above to collaborative research and centers we recognize that much of the work in all of these areas is advanced through individual investigator grants. Even with refurbished labo- ratories, better trained researchers, and high-quality data bases, there is still a need for grant support that is tailored to the specific work and interests of qualified researchers, and such support must be adjusted to take account of the more complex requirements and longer time horizons of present research frontiers. We recommend a steady increase for individual grants over the next several years to appropriately $13 million more than present funding levels. Most of the increase should go to increasing the size of individual grants to more realistic levels for the work; the balance should be used to fund additional investigators, especially young ones who are currently being denied funding even though their proposals receive high ratings in the review process.