You are on page 1of 31





ABSTRACT. Individual illness belief systems form a cognitive structure which lies beneath the cultural and social aspect of health care in a community. Popular belief systems are different from, yet linked to, expert belief systems. Popular illness terms often help support a stable cultural milieu by linking concepts of causes and significance of types of illness problems with a set of health care seeking choices; as well as linking typical physical and psychological symptoms with associated social problems. This study presents an example of how illness beliefs perform these functions in urban, mainstream America. One hundred and seventeen people with biomedically defined hypertension were interviewed following the Explanatory Models format. The belief held by 72% of this sample was that they had 'Hyper-Tension,' a physical illness characterized by excessive nervousness caused by untoward social stress. The data are used to derive a composite diagram of the cognitive domain of 'Hyper-Tension' in America which demonstrates the various options people have for interpreting their experiences and choosing appropriate therapeutic actions. They use this illness belief system to justify otherwise unwarranted social behavior and to assume various aspects of the sick role. This popular model is compared and contrasted with the professional model of the disease 'hypertension' and with other less frequent models which were observed in this sample.

INTRODUCTION It is generally accepted that all cultures have some sort o f healing tradition. How these differing healing systems function to maintain a stable cultural milieu is a matter o f much wider debate. Frequently healing systems are divided into categories such as: local (primitive, folk) medical system; regional (Ayurvedic, Unani, Chinese) medical systems and the cosmopolitan (modern, Western, scientific) medical system. This type o f classification relies on differences in geographic spread, degree o f professionalization, and underlying theories o f health, disease and healing. Since this type o f analysis tends to stress differences rather than commonalities, functional equivalents which occur across medical systems are more difficult to determine (Press 1980). The type o f problem that develops is illustrated by Press who defines a folk medical system as being (1) any health system at variance with Western, scientific medicine; (2) any health system at variance with a codified, formal and literate medical tradition (Western, scientific, Ayurvedic, classical Chinese, etc.); (3) any system o f health practice at variance with the official health practice o f the c o m m u n i t y or nation (Press 1978). This type o f theoretical muddle arises because researchers allow themselves to be distracted b y differences in healing theory (e.g., biomedical vs. Ayurvedic) or in ritual practices (naturalistic vs. spiritualistic) or in apparent complexity (cosmopolitan vs. local). Horton has documented that, in fact, these

Culture, Medicine and Psychiatry 4 (1980) 197-227. 0165-005X/80/0043-0197 $03.10. Copyright 1980 by D. Reidel Publishing Co., Dordrecht, Holland, and Boston, U.S.A.



differences may be more apparent than real, even when African traditional thought and Western science are contrasted (1967). Perhaps a more fruitful way to approach a theoretical understanding of health care systems is to look at the distribution of healing knowledge in a culture, and then examine how this knowledge is used to shape both social and personal realities. One of the characteristics of nearly all health care systems is that there is a division of knowledge into practitioner (or professional) and popular realms. While this is shown in complex, urban environments, it probably also applies to most small scale societies as well. Thus, despite Lewis' claim that the Gnau "do not have deafly recognized medical experts" (1975:196), elsewhere he admits that "men who have acquired this knowledge are thus the experts in healing and diagnosis" (p. 170). It is also important to recognize that in almost every culture there are a variety of healers among whom a lay person can choose. The individual interpretations of illness that each practitioner makes are sufficiently distinct in most cases to offer a real choice of healing technique. This does not merely apply in those settings where a variety of healing traditions intersect, such as Taiwan (Kleinman 1980), but also occurs in the urban West where most healers may bear the name "physician." The concept that healing styles differ markedly even within a tightly controlled, highly professionalized system like Biomedicine 2 is also not well accepted in the literature, despite Hehnan's (1978) and Freidson's (1970) work. Press, for example, states that in Bogota "no two curers are alike ... each is a distinct and stylistic specialist. It is easy, in comparison, to view the physician [sic] monolithically" (1969). 3 One has only to listen to the violent arguments that occur, for example, between nephrologists and cardiologists as they discuss the care of a critically ill person to realize that such a "monolithic" view is untenable. This occurs even though both are considered to be members of the same speciality (internal medicine). The important thing to note here is that as we attempt to work through the interaction of individual belief systems and expert belief systems, we cannot assume that either is a uniform structure, unaffected by the individuals themselves. Instead we must look at how a particular layman interacts with a particular expert, and from these interactions draw conclusions about the larger system. INDIVIDUAL BELIEF SYSTEMS Sickness is a ubiquitous human experience. By its very nature, sickness is a profound threat to the social and personal existence of the individual. Individuals must therefore have some way of interpreting the nature of the threat that illness poses as well as avenues which may be taken to mitigate these effects. In



other words, a belief system must imbue illness experiences with meaning and provide alternative courses of action. Kleinman has recently described a method to elicit such beliefs in his 'Explanatory Models of Illness' (1980). This methodology appears to function cross culturally. Apparently all people need answers to questions such as "What caused this illness?"; "Why did it begin at a particular time?"; "What is happening to me?"; What will be the outcome of this illness?" and "What should be done about it?". It is important to note that while we are primarily concerned with a layman's interpretation of his illness at this point, expert interpretations of diseases are frequently oriented towards answering the same questions. Indeed, the observation that the cognitive and practical direction of both lay and expert illness explanatory models in most social systems can be summed up into as few categories as Kleinman postulates has immense theoretical implications for the cross cultural study of sickness. Apparently, despite the enormous diversity that is observed in the ways that illness is experienced and dealt with, there are a relatively small number of core functions that are performed by nearly all healing systems. An individual's 'explanatory model' should not be viewed as something which is necessarily unchanging. Instead, it is a concise statment of the individual's illness beliefs which are deemed relevant to a particular aspect of that person's experiences at a particular time. As the aspect of interest is changed, as the experiences change, the explanatory model will change, sometimes with amazing rapidity. Thus, for example, if one inquires about the physical causes of an illness, an explanatory model may be given which will be radically different from the explanatory model given by the same individual if one then asks about spiritual or social causes of the same illness. Since explanatory models usually contain treatment options, the presence of different explanatory models for different aspects of the same illness means that an individual may be involved in a variety of treatment options, which to the outside observer appear to be mutually contradictory, without feeling a cognitive strain (Kunstadter 1976). This is repeatedly observed in settings where there are competing healing systems: in Taiwan an individual may use the remedies of a Western physician, a traditional Chinese doctor and a religious shaman at the same time (Kleinman 1980). Questions about the compatibility of these treatments are usually only raised by the ethnographer. Although it is important to avoid reifying explanatory models, it is also important to recognize that these seem to provide the justification for therapeutic action, and that therefore, while ephemeral, are still a substantive link between the individual's belief systems and the actions that are undertaken as a result of these belief systems. When we speak of individual belief systems, particularly as these are applied



to laymen, there is a danger that we will be seduced by the term "system." Systematicity, coherence and interdependence are aspects of professional belief systems. It is an unusual individual (or, indeed, practitioner) who has worked out all the interconnections of his belief system so that it has become an integrated whole (Berger and Luckmann 1967). It is more likely that a person will have a set of beliefs about illness which are only loosely interconnected, if at all. The connections between the isolated beliefs may be supplied to meet the need to explain a particular situation. In addition, in a particular situation, both the interconnections and the items of individual beliefs appear to be continuously reworked (in part through the construction of a series of explanatory models) to provide a framework for dealing with a particular illness (AmaraSingham 1980). It is only in this sense that we can speak of an individual illness belief system - it appears in fact to be a set of beliefs, a potential system which can be operationalized to cope with a particular experience. Because time and experience are required to work the isolated beliefs into a functioning system, it may not be useful to question people about their explanatory models for hypothetical situations. Unless there has been some reason to work through their beliefs, what will be elicited in that setting is only the first, most tentative, most generalized shaping of the belief data. Having outlined the nature of the relationship between an individual's illness beliefs and the resulting explanatory models, we need to explore the origin of items in these belief systems and explanatory models. There are three major sources: idiosyncratic, popular and expert. The first of these is derived from that person's own experiences, obsevations and reworking of his basic data unrelated to what others say or do. Clearly, the extent to which this is a substantial source of a person's beliefs will vary from person to person - the eccentric will have a greater proportion of idiosyncratic beliefs than a person who toes the party line. Culturally, they are very important since they may serve as a source of new ideas - concepts which may be added to the cultural domain of illness beliefs making it more adaptive, more flexible, and therefore more viable in a changing environment (Buckley 1967). Most of these idiosyncratic beliefs, of course, will remain unstated, or will be rejected the instant an innovator has the temerity to propose such preposterous ideas. From a research standpoint, however, these idiosyncratic beliefs are important because they serve as a source of variability in our data which is impossible to explain - or to ignore. To use another jargon, they are "noise" which must be carefully evaluated before either including or rejecting them from our data base. A second, and probably the most important source of illness beliefs is the lay health system (Chrisman 1977). This is commonly known under a variety of names including "popular health system" and most uses of "folk health system." Other people's illnesses and ideas about illness may be widely discussed in



ordinary conversation, with alternative explanations presented and rejected. This information thereby serves as a major repository of knowledge which can be accumulated into a set of individual illness beliefs. When an illness occurs, in addition to the individual reworking the data in his own set, the social network that is affected around the sick person becomes a, if not the, most important source of information about the available interpretations and practical alternatives. The process of negotiating meanings and potential actions among the idiosyncratic models of illness and the models expressed by the social network is dearly an extraordinarily complex process which has been described elsewhere (Chrisman 1977; Freidson 1970). The capacity of these first two sources of beliefs and models is illustrated by the repeated observation that somewhere between 70 and 90% of all illness episodes are adequately managed without requiring recourse to expert knowledge. The third input to individual illness belief systems and thus to their explanatory models comes from information derived from expert explanatory models. As the term is used here, "experts" are not limited to practitioners of the great healing traditions such as Ayurveda or Biomedicine but may be any individual who is socially recognized as having more extensive knowledge of illness and techniques for dealing with illness. As one who has both more extensive and more systematic knowledge of illness, an expert can supplement and offer correction to individual explanatory models. (In taking a cognitive approach to illness I am setting aside the technical aspects of the expert's practice, which in my opinion, often primarily serve to validate the interpretive aspects.) The expert offers an explanatory model which has whatever prestige the individual expert and his healing tradition may carry (Stone 1979). It is important to remember that the expert as well as the lay health system can only offer explanatory models - belief systems cannot be directly transmitted. As such the explanatory models can then be accepted by the individual as new items in his set of illness beliefs, to be integrated with other beliefs and experiences and included or rejected in the process of producing new explanatory models. CULTURAL ILLNESS BELIEF SYSTEMS Having established the production and function of individual belief systems, let us turn to what can be loosely termed the cultural belief systems. These include the various popular and the expert belief systems. Since I do not wish to postulate a "group mind," what these consist of are the belief systems which seem to be a necessary basis for the explanatory models which are frequently expressed by the members of a particular group. In the case of expert belief systems, these are often codified in textbooks, although the codified elements that a particular expert will use in the explanatory models he offers clients are



almost as problematic as the relationship between the elements of the popular belief system and the explanatory model of a particular layman. Because of these factors, popular illness belief systems usually remain tacit until their implications are worked out by ethnographers or unusual individuals like Muchona, Tumer's (1967) theoretically inclined Ndembu informant. This follows from the nature of lay health systems, which consist of the beliefs which are commonly held and communicated among laymen of a particular group. As a layman becomes able to articulate the popular concepts of illness as a coherent system, he usually ceases to be a mere layman and is recognized as an expert. Despite these caveats, lay systems do appear to exist. Over long periods of time various solutions to illness problems are worked out and passed on. Solutions which better meet the social and psychological needs which arise as the result of illness replace those which are less functional (Lindenbaum 1979). 4 As has been pointed out, the individual is not obligated to comply with the culturally accepted modes of interpreting and healing sickness, but as an ever widening social network becomes involved with the sickness, there usually is increasing pressure to conform. The various health belief systems of any particular group are often revealed by the way the social structures for dealing with health and illness operate. Again, despite apparent diversity, there seem to be a set of core functions which operate cross culturally. As Kleinman (1980: 71) had outlined them, these are: 1. The cultural construction of illness as a psychosocial experience. 2. The establishment of general criteria to guide the health care seeking process and to evaluate treatment approaches that exist prior to and independent of individual episodes of sickness. 3. The management of particular illness episodes through communicative operations such as labeling and explaining. 4. Healing activities per se, which include all types of therapeutic interventions, from drugs and surgery to psychotherapy, supportive care, and healing rituals. 5. The management of therapeutic outcomes, including cure, treatment failure, recurrence, chronic illness, impairment, and death Dunn (1976) would add to this list: 6. A set of activities designed to maintain health and reduce the incidence of illness. The derivation and implication of these functions has been well established in the medical anthropology literature, so that extensive discussion here is not warranted. It should be noted that there are again two major functions which are performed: a cognitive function and an action function. In other words, two questions are answered: "What is happening to this person?"; and "What should be done about it?" One of the most important aspects of the cognitive functions of the lay health system is that they are lined to everyday reality and everyday or commonsense



knowledge. This is in contradistinction to expert health systems which draw their cognitive links to esoteric knowledge systems. This is best demonstrated by Good (1977) who shows that in Iran, despite partial acceptance of the theories of the three high traditions of medicine (Galenic-Islamic, Sacred-Qur'anic and Biomedicine), the lay or popular health system has its own view of clinical reality which is grounded in the exigencies and experiences of everyday life. As such the meanings ascribed to illness terms in popular health belief systems are drawn from the definitions of everyday language rather than professional jargon (Berger and Luckmann 1967). Too often we tend to look at the expert definition since it is most readily available - and then assume that laymen (if they use the term at all) use the same definition. Since the same term may be grounded in expert theories of the causation of disease on the one hand, and in commonsense perceptions on the other, we must be cautious to avoid this danger. Illness terms, or labels, function in a variety of ways. In the first place, the label itself implies a narrowed range of potential explanatory models. Good's (1977) description of heart distress, previously referred to, is an excellent example of the type of explanatory models that an illness term can imply, and how these are interconnected. The label thus serves to link a set of beliefs about the causes and significance of particular types of illness problems with a set of health care seeking choices of available treatment options for these types of illness problems (Kleinman 1980:108). However, because the primary effect of an illness is on the person and on that individual's social role, many popular illness terms tend to be linked with psychological and social implications which are associated with the illness. Thus the illness term will tend to link what are seen as typical symptoms and psychological processes with the typical social problems which are associated with any particular type of illness problem (ibid.). Through this process of definition, both the behavior and the social role of an ill person will be culturally directed. Note however: this does not imply that the behavior and social role will necessarily be culturally determined; there is normally room for individual interpretation of the cultural pattern. These are two problems that arise from this formulation of the social construction of illness. The first of these concerns how people actually use the popular and expert system of knowledge in conjunction with their own idiosyncratic interpretations and their personal experiences to generate explanatory models. In Good's work, for example, we learn that "heart disease" first is associated with "sadness, worry, anxiety." Is that simply a subconscious free association on a societal scale, or are specific mechanisms postulated which would indicate potential interventions? Further exploration is needed to clarify this issue. The second issue is whether this socio-psychological framework has any validity in the urbanized West whose primary healers use materialistic, mechanistic,



stochastic explanations for the occurrence of disease. Theoretically we would predict a severe tension existing between lay and expert explanatory models, to the point that expert intervention should be rejected as being irrelevant to the felt needs of the populace. Indeed, this criticism is widely leveled at the medical profession. But what do the data show? Why do people persist in seeing their doctors? STUDY DESIGN AND METHODOLOGICAL CONSIDERATIONS To address some of these issues, particularly as they arise in a complex, urban setting, I designed a study exploring some of the illness beliefs of mainstream Americans. One hundred and seventeen people who were outpatients in a hypertension clinic were interviewed using a semi-structured questionnaire which followed the 'Explanatory Models' format (Kleinman 1980:106). The research was carded out over a twelve month period beginning September 1977. There are a number of methodological issues which affect the generalizability of the results and must be discussed. Since I was primarily interested in illness beliefs in an urban setting, traditional anthropological techniques to identify a study population (the isolated village) could not be used. Population sampling was beyond our f'mancial and time constraints. As a result, I elected to study a population which was defined biomedically and administratively. These were individuals who were attending a particular clinic at the Seattle Veterans Administration Medical Center. The only other formal criterion was that they had been attending the clinic for at least six months. This time restriction was set to allow the patient's illness beliefs to come into equilibrium with a new clinic environment. These restrictions would be expected to have a number of effects on the data. Selecting people who have elected to seek and remain in medical care will select those whose explanatory models are less likely to be in conflict with the professional's model. In other words, any popular health beliefs found in this population would suggest that these beliefs are much more widely held among the population at large. Administratively, only those individuals who have served in the United States Armed Forces are eligible for health care through the Veterans Administration. This results in a preponderance of males: thus, only 2 of the 103 individuals finally included in the study were women. I cannot predict the effect of this skewed sex ratio on the data. In addition, minority groups were underrepresented. Of the 103, 90 were white (including both women), 9 were black, 2 American Indian and 2 Filipino. The effects of minority group models of illness will be discussed later in the paper. Since the primary interest is in the health beliefs of mainstream Americans, this sample serves our purposes.



Other sociodemographic characteristics of this group are as follows: mean age was 55.4 years with a range of 22 to 79. Mean educational achievement was 11.9 years, with a range from 3 years to a Ph.D. candidate. Mean occupational status measured by the seven point Hollingshead scale was 4.55 with a range of 1 (highest) to 7 (Bonjean et al. 1967). To summarize these data, the population consisted of predominantly white middle aged men with a high school education who were employed in a middle class occupation. The reason for selecting a particular clinic was to ensure that all individuals had a similar physiological experience, and that they were all given the same professional explanatory model. Hypertension was selected as the index medical condition partly for administrative reasons: i.e., there was a clinic in operation whose staff was amenable to our research. However, a second important reason is that uncomplicated hypertension can exist as a chronic condition without significant numbers of other coexisting medical problems. This is not as true for heart disease, diabetes, or chronic lung disease, for example. Individual health beliefs are not neatly segregated into disease specific packages, and we felt that our task would be much simpler if we could limit the discussion to the experience of a single illness. A 22 item semi-structured questionnaire based on Kleinman's Explanatory Models questions (1980:106) was developed and pretested on six individuals who were not part of the study. The resulting revised questionnaire was then given to 117 patients in the clinic by the chief investigator, who was introduced as a graduate student in anthropology and therefore someone who was administratively unrelated to the clinic. The interviews lasted ten to thirty minutes. All interviews were tape recorded and transcribed. Fourteen of the recordings were technically inadequate for transcription (usually due to electrical interference) and were eliminated, having 103 people in the study. All data here stem from these 103 interviews. RESULTS Each interview was analyzed to determine the individual's explanatory model of hypertension. Using the person's own words as much as possible, the models were diagrammed into a set of 'nodes' and 'arrows' where the nodes are important factors in explanations of etiology, pathophysiology and outcome, and the 'arrows' represent the causal relationships which were said to exist between the nodes (Figure 1). The individual models contained a mean of thirteen nodes with a range of four to twenty-six. It can be seen from the example that we are dealing with relatively complex models. As would be expected, the models were not necessarily internally consistent. On occasion individuals gave two or more parallel, nonintegrated models at different points in the interview. When












STO. . . .







inconsistencies were pointed out, a typical response was: "1 never thought of it that way," suggesting that the inconsistencies were not problematic to the person interviewed. The models also had points of merger, where several strands jointed in a single outcome, and also had branchings, where a single factor could result in a variety of outcomes. A manual sorting and grouping technique was devised whereby nodes which were in structurally similar locations in the various individuals' models and which had similar semantic content were grouped for the entire sample of 103 individual explanatory models. This technique allowed the 1300 individual nodes to be collapsed into a set of 59 categories. The categories which occurred in more than 20% of the individual models were selected as being indicative of a portion of a cultural model. These were combined with their first order arrows to give a summary diagram which I call the Cognitive Domain of Hypertension (Figure 2). In this figure the width of the arrows and the size of each node are proportional to the number of people who gave that item in their individual models. The nodes in the cognitive domain model can be further classified as causes, intermediate mechanisms and outcomes of hypertension. The number of people who included a particular node in their individual models is shown in Tables 1-3. The nodes of the summary diagram account for 90% of all the nodes given in the individual models. However, 56% of the people interviewed had at least one concept which is not presented here, so that only 44% of the individual models can be entirely mapped onto the nodes of the cognitive domain. The arrows linking the nodes are more difficult to interpret. To be included they had to be given by at least 5% of the sample, and the largest was only included by 39%. During the interviews I gained the impression that whereas the





concepts expressed here as nodes are fairly stable elements o f the individual's belief system, the links drawn between them are not, and were amenable to being shifted around in response to a different line o f questioning. Theoretically, this is what we would expect. Nonetheless, the wide range o f individual variation coupled with the inherent ambiguities o f these links make it impossible to estimate the "completeness" o f this aspect o f the cognitive domain.

TABLE 1 Causes N Acute Stress Chronic External Stress Chronic Internal Stress Smoking Alcohol Personality Trait Heredity Food Salt Water 57 50 14 15 16 17 25 33 56 43 % 56 49 14 15 16 17 25 32 55 42


DAN BLUMHAGEN TABLE 2 Intermediate Mechanism N Hypertension Vessels Constrict Heart Beats Faster Heart Works Harder Altered Circulation Narrowed Blood Vessels Circulation Cut Off Overweight High Pressure Heart Strain Burst Blood Vessel 63 12 23 34 31 43 16 48 48 33 29 % 62 12 23 33 30 42 16 47 47 32 28

TABLE 3 Outcome N Tiredness Dizziness Headache Flushing Emotional Symptoms Stroke Physical Damage Heart Attack Heart Failure Paralysis Death 18 28 21 31 29 68 29 46 33 17 26 % 18 27 21 30 28 67 28 45 32 17 25

CONTENT OF THE NODES Having described the origin o f the diagram, it is very important to present the contents o f some o f the nodes. The remainder are either self-explanatory and have very little internal variation, or are relatively unimportant to the overall schema; therefore neither will be extensively discussed. There are two major sets o f causes o f this illness. These are roughly divided into psycho-social causes on the left o f the diagram, which result in a state called Hyper-Tension, and a set o f physical-hereditary factors shown at the b o t t o m which affect the pressure in the body. Many o f the people interviewed gave b o t h systems. For clarity in the presentation, the two major systems will be discussed separately.



The most important of the psycho-social causes of the illness 'Hyper.Tension' fall into a category called "Chronic External Stress." Nearly half o f the people interviewed said that this was a significant factor in the production o f their illness. As can be seen from Table 4 this category was used to group three types of social and environmental stresses: First was a single episode o f extreme stress, such as being at Pearl Harbor during the Japanese raid in 1941, or in the 1964 Alaskan TABLE 4 Chronic External Stress N = 50, 49% of Total N
I. Single Episode of Extreme Stress 5

Category %

II. Job Stresses Work Too Hard Specific Occupation Job Stress Unemployment III. Buildup of Normal Stresses Life's Stress and Strain Tension Worry/Anxiety/Nervousness Pressure Insufficient Rest

7 5 13 2 9 11 9 4 4

14 10 26 4 18 22 18 8 8

Totals are greater than 100% because some respondents listed multiple items. earthquake. Secondly, there were a set of stressors related to employment: A particularly demanding occupation, such as being a submariner. Disagreeable colleagues or just too much hard work, possibly as a manual laborer, were also considered to be significant causes. Finally, an accumulation of stress of normal living is also often given as a reason for developing hypertension. The important difference between this category and the next, which I call chronic internal stress, is that here the stressors seem to originate outside the individual and his immediate family network. For a variety o f reasons the two categories seemed to segregate differently and therefore are kept apart in the diagram. Chronic internal stress (Table 5) relates to either psychiatric problems - such as chronic schizophrenia - or long-standing intrafamilial problems. A nasty mother-in-law or an unfortunate marriage with a messy divorce are examples o f this. Interestingly this differed from the category labeled 'Personality Trait' which is an expression of the idea "I've always been a high tension person."


DAN BLUMHAGEN TABLE 5 Chronic Internal Stress N = 14, 14% of Total Category % 50 64

N Psychological Problems Family or Interpersonal Problems 7 9

Totals are greater than 100% because some respondents listed multiple items. Despite the fact that this is an innate characteristic o f the individual, it is not seen as inherited, i.e., derived from a parental trait. People actually said "I've always been a tense person and m y mother was too, but I d o n ' t think it was inherited." Acute Stress (Table 6) is any situation which is seen as temporarily affecting the individual. Examples of this were becoming angry with someone, being stopped by a policeman, or having sex. People viewed themselves as having been made more vulnerable to becoming 'hyper-tense' in such situations. A long series o f such circumstances could result in an altered physical state, but would then be classified under "Chronic External Stress." TABLE 6 Acute Stress N = 57, 56% of Total Category % 14 21 12 9 9 47 19 11

I. Physical Overexertion

8 12 7 5 5 27 11 6

II. Internal Stress Worry/Anxiety Nervous Emotions Anger III. Social Stress Stressful Situation Excitement Social Demands

Totals are greater than 100% because some respondents listed multiple items.

THE MEANING OF HYPER-TENSION All o f these causal factors resulted in a state I call 'Hyper-Tension'. This must be distinguished from the biomedical defmition o f hypertension, which refers to a



sustained elevation of the hemodynamic pressure in the systemic circulation of the body. Since lay terminology is related to everyday English instead of professional jargon the natural translation of this term would be "excessive tenseness." This is what was found when people were asked to describe what they meant by 'Hyper-Tension'. Table 7 gives the descriptive terms that were used. While many of these occurred too rarely to be statistically significant, when they are put together they paint a graphic picture of this illness. There are, however, two components to 'Hyper-Tension'. In the first place it is a state characterized by the experience of nervousness, fear, anxiety, worry, anger, upset, tenseness, overactivity, exhaustion and/or excitement. Secondly it is a condition brought on by the psychosocial factors already discussed which makes the individual more susceptible to becoming 'hyper-tense.' TABLE 7 Hyper-Tension: Descriptive Terms N Nervousness Irritated Nerves Tenseness Fear/Anxiety/Worry Overactive Anger/Upset Exhaustion Frustrated Mind Working Emotional Deterioration Grinding Teeth Impatience High Stress Excited Being up Tight High strung Shook Up 15 2 4 6 4 5 3 1 1 1 1 1 1 1 2 1 1 1

Despite the psychological sound of the experiences associated with 'HyperTension', it is considered to be a physical and specifically not a psychological disorder. It is not an illness which is "all in a person's head," but appears to involve a physical change in the person, although the nature of this change is not specified. Thus, for example, despite the fact that tenseness is a central part of the illness, tranquilizers are n o t considered to be appropriate therapy. The state of 'Hyper-Tension' manifests its existence in various ways. The primary action is on the heart, which may pump faster or harder resulting in an



increase in pressure b o t h on the blood (high blood pressure) and on the body as a whole. Many informants explicitly made this distinction between 'HyperTension' and high blood pressure. As one person said: My interpretation is that the Hyper-Tension moves into high blood pressure and causes high blood pressure. I think that's where it really gets its momentum from, to cause the blood pressure to go up. [Hyper-Tension] is nervousness, a condition, a pressure... I have a fast moving pressure thing. Like on a treadmill, moving all the time, fast, I think . . . and then there's some nervousness about it, some nervousness, some worry and concentration that the load is too heavy for the mind to carry it . . . . it overloads someplace, causes a short circuit. I didn't know it was high blood pressure. [Note the shift in terms.] I got shortness of breath. I'd walk up a short ramp and I was out of breath. Went to the doctor for something else . . . and he discovered that my blood pressure was way up . . . . [The doctor said] 'stabilize your body, you had it stabilized and it should stay there.' It's just like a spring if you had it down now if you release it [by not taking medication] it'll go back up and damage your heart. This distinction has no correlate within the professional definitions o f tb.z disease. HEREDITARY-PHYSICAL CAUSES OF HIGH BLOOD PRESSURE In addition to the people who believed that stress caused their Hyper-Tension there were a substantial number in this study who felt that their illness was caused by physical or hereditary factors, and n o t psychosocial ones. Here the primary model was that a variety o f factors caused a narrowing on the blood vessels. This could be due to hereditary factors, salt which would cause deposits in the blood vessels or excessive weight squeezing the blood vessels. Whatever the cause, this alteration in blood flow was seen as making the heart work harder and thereby increasing the blood pressure. Similarly, too much food was seen as causing a person to become overweight, which would then stress the whole body system. This would occur either by narrowing the blood vessels as has been described, by putting an additional workload on the heart, or simply by increasing the b o d y ' s 'pressure.' F a t t y foods and salt were felt to be particularly dangerous in this regard. Salt reportedly holds on to water, which either makes one overweight, or simply increases the fluid pressure in the 'system.' The concept of 'water' contributing to overweight appears to be widespread, and probably is due to the well-known phenomeno, o f premenstrual weight gain and to the similarly well-known effect of water retention during weight loss. Since hypertension is universally treated with diuretics ('water pills') which are also popularly used for premenstrual edema and weight loss, this professional practice supports a popular cognitive system unrelated to the professional cognitive system.



As Figure 2 indicates, most of the effects of this illness work through the intermediate mechanism of high pressure. Just as there are psychosocial as well as physical causes of high pressure, there are psychosocial and physical results. The psychosocial effects are shown at the top of the diagram and consist of the categories tiredness, dizziness, headache, flushing and emotional symptoms. Most of these categories are self-explanatory. It must be noted that despite intensive clinical research, no consistent correlation can be established between the blood pressure as measured by a physician using a sphygmomanometer (blood pressure cuff) and these symptoms (Kaplan 1978). This includes headaches. These symptoms must therefore be considered to be symptoms of a popular or folk illness and not part of the pathophysiology of (professionally defined) high blood pressure. The only one of these symptoms which requires further discussion here is the category of 'emotional symptoms.' Table 8 lists these. These include altered internal, psychological states as well as a changed social status. If this table is compared with the tables describing Acute Stress and Hyper-Tension (Tables 6 and 7) it is immediately obvious that many of the factors which are said to cause Hyper-Tension are also seen as the result of Hyper-Tension. Some of the individuals recognized this and explicitly called attention to a vicious cycle or positive feedback loop which tends to exacerbate their situation. Thus, for example, the inability to meet social demands which is viewed as a result of their high blood pressure is believed to feed back as an acute stress further exacerbating the person's illness. Similarly, people recognize a set of permanent, physical results from their illness. These consist of primarily stroke, heart attack, and a wide variety of other results which are grouped under the classification of 'physical damage.' Problems with blood vessels were prominent in this category: varicose veins, 'poor circulation' resulting in leg pain, and aneurysms were frequently mentioned. Given the high proportion of social causation, it is interesting that of the two conditions which result from these permanent changes, paralysis and death, the former was more widely feared. Death was considered preferable to a state of permanent invalidism, where people were perceived as being totally dependent on others, with no opportunity to exercise any meaningful social role. Physical death is preferable to social death. THE EXPERT'S MODEL In the introduction a process was described by which individuals combine their own illness experiences and interpretation with the popular and expert models


DAN BLUMHAGEN TABLE 8 Emotional Symptoms N

I. Psychological Symptoms

Nervous Temper Irritable Anxiety/Worry Upset Feel Stress Gets up tight Changed Behavior Tension Hyperactivity Loss of Sleep Easily Distracted II. Changed Social Function Unable to meet social demands Unable to work as well Unable to show restraint Ill at ease Altered behavior Relationships deteriorate

8 3 8 3 1 1 1 1 2 1 1 1 3 3 1 1 1 2

to come up with a set of explanatory models of their ilhaesses. We have demonstrated one way to derive the popular model. How do these link with the expert model? It is important to remember that the physician is the primary folk healer for almost all the individuals in this study. The doctor's expertise is seen as the standard by which popular and personal illness beliefs are judged. The statement "You shouldn't ask me that, I'm not a doctor" was repeatedly encountered in this study. Nonetheless, along the way we have pointed out important areas of difference between professional and popular beliefs. What then is the expert model of hypertension and how does it relate to the popular model? This question, which superficially appears quite easy to answer, is in fact one of the knottiest that faces us. Conventional wisdom would indicate that we should look in the recognized textbooks of medicine, or in the biomedical literature, and use the models expressed there as prototypes of the professional model. This overlooks the nature of the material presented in each of these two places. The biomedical literature is part of the constant reworking of the basic explanatory models and belief systems of medicine. As such, almost any belief will be expressed by some expert who marshaUs data to support his view. Most of these, of course, are rapidly forgotten, except by the archivist. It is assumed by the medical profession that out of the cacophony, truth will somehow emerge.



The authors of textbooks and monographs take the basic data of the biomedical literature and then rework it into a belief system. These are very extensive models. A recent textbook of medicine, for example, spends 18 double columned pages describing hypertension alone (Beeson, McDermott, Wyngaarden 1979). Similarly, large monographs go into ever increasing detail to make all the data fit into some overarching schema (e.g., Kaplan 1978). Frequently, flow models similar to the one developed here are presented. What are discussed in these sources are belief systems, which individual practitioners merge into their previous belief systems and with their clinical experience (Mechanic 1978). This process may be very similar to that described for laymen, but hopefully it occurs on a different level of complexity and coherence (Gaines 1979). As has been described, the expert then uses these beliefs to shape an explanatory model to give to an individual patient. In addition, it is my experience that explanatory models are not usually presented by a practitioner as a whole at one point in time, but rather piecemeal as the need for interpretation of the illness experience arises. Several of the patients in this study had had their illness for thirty years or more and had seen many different practitioners. Not only had they seen different practitioners, but over this period of time the overarching biomedical models of hypertension have changed substantially. Under these circumstances it is impossible to say what professional models have been presented to them (AmaraSingham 1980). Since we are using cross sectional data from an ongoing process, the only alternative that we have is to examine the models which were currently being presented in the clinic. There are two sources of illness explanatory models in this clinic: the two nurse practitioners s who were actually seeing the patients, and a wide variety of pamphlets prepared by the American Heart Association, Searle Pharmaceutical Company, and the National High Blood Pressure Education Program. All of these were important to patient education. The practitioners were asked what they felt was important to teach their patients about hypertension. Responses to each of the five categories of the Explanatory Model were obtained. Both nurse practitioners felt that the cause of High Blood Pressure is unknown. Both stated that stress can exacerbate High Blood Pressure but is not the primary cause. Despite this, both said that they would not try ~too hard to dissuade a person who believed that stressful situations had caused his High Blood Pressure. Heredity and obesity were seen as contributing factors. Both stated that reasons for onset of the disease at a particular time were unknown, and that they therefore did not try to give an explanation for the time of onset. The pathophysiological explanation involved increased pressure across the walls of blood vessels, which resulted in narrowing of the blood vessels. Both believed that patients usually cannot detect symptoms of elevated blood pressure,



but again stated that they did not attempt to change the mind of those who believed they could, except insofar as such beliefs affect medication taking behavior. The primary treatments were dietary, with salt restriction and weight loss being most important. Following these were medications which they expected patients to need for the rest of their lives. Advice for general good health was reported, specifically regular exercise and stopping smoking. The pamphlets, as one might expect, differed little from the opinion of the nurse practitioners, particularly insofar as they were unable to give reasons for the etiology or onset of an illness. One said "In some people, blood pressure is nearly always higher than it should be." This is a good example of how shallow the explanations are that can be given by statistically based medical theories. Each of the pamphlets stressed that emotions were not the primary cause of high blood pressure, and that symptoms were not a reliable sign of how high a person's blood pressure may be. The pathophysiological mechanisms were largely the same as those given by the nurses. One interesting difference in therapies is that most of the pamphlets stressed medications as the first choice of treatment instead of dietary changes. [This is in conflict both with the nurse practioners as well as with medical experts (Kaplan 1978:105).] Several of the pamphlets included closing sections titled something like "what about general health habits?" The selection of topics here is interesting. One says, in part: 1. Get plenty of relaxation. Don't wait until you are exhausted to start to rest. 2. Lose weight (if your doctor believesyou should) then maintain it at the level suggested. 3. Avoid alcohol and tobacco... 4. Settle your problems so they don't worry you and interfere with your rest and peace of mind. 5. Have adequate sleep. To do this you may have to insist on a nap during the day (Curtis 1976). Three of the topics here deal with stress as a cause of illness despite the statement earlier in the pamphlet that "It is a popular misconception that a hypertensive patient is a nervous, compulsive, never resting perfectionist who cannot sit still . . . [some] are the relaxed; well adjusted, quiet, calm type" 0bid). Thus the recommendations for action conflict with the theory of illness which had just been presented. The professional model used in this clinic can therefore be diagrammed as shown in Figure 3. It is obviously very different from the lay model, particularly insofar as it is totally unable to give a satisfying answer to the questions "Why me?"; "Why now?" (Taussig 1980). As if became obvious that there was a wide difference between the popular















and professional models o f the illness, we began asking people if they were aware o f these discrepancies. Of those who said that they read the pamphlets at least occasionally and who believed that their illness was predominantly due to stress (N = 34) 68% stated that they had never found anything that did not agree with their conception o f their illness. I assume that these people are subconsciously filtering out any information that is discrepant with their existing model. More interesting are the 32% who stated that they were aware o f the differences and consciously rejected the information, feeling that while a particular thing m a y be true in general, it didn't apply specifically to them. As one said: Most of the doctors I talk to, some of them that I've talked to are for it and some are against i t . . . I figure in, something of that respect, you are with your opinion and you can have it, but it doesn't necessarily mean that I have to abide by it. I can think and I'm an intelligent being too and I can do research and read and come to my own conclusions about certain things, and just because you disagree with me, doesn't mean that one thing is right and I'm wrong. See? It would be different if I came up against a total amount of learned people who in medicine would say that this is wrong, but you don't." The professional model exists, but the person with the illness retains the right to accept and reject the doctor's ideas. In fact, by far the majority o f the people in this study have developed models which vary widely from that held by their professional advisors. Why is this? Strong social or psychological forces must exist which require a model o f illness based on social stress - otherwise such a 'survival' would rapidly be overcome by the sheer weight o f Biomedical opinion. THE USE OF THE POPULAR MODEL To answer these questions we must go beneath the simple presentation o f the model and examine how the people in this study report using the model to interpret their experiences and to structure behavioral options. Despite the fact that these represent reported and not observed behavior, they were given as examples o f how the person thought he acted under certain circumstances. As



such they demonstrate the conceptual link between the cognitive models and perceived behavior. Much more extensive observation would be necessary to document the links between cognitive models and actual behavior. The majority of the people interviewed (72%) reported believing that 'HyperTension' is a physical reflection of past social and environmental stressors, which is exacerbated by current stressful situations. Vignettes from the interviews illustrate the way these concepts are used to define and redef'me experiences, and to manipulate the social environment to reduce stress. [Mental things] put a strain and stress and tight anxiety and this causes [high blood pressure]. I think that worry can cause - anticipations of something you have to do, I think this raises it, too. Q. Is this what raised it in your case? Well, I know it raised it enough to put me in the hospital at one time . . . . I took a physical aboard an aircraft carrier and it was fine and here come my orders back and I'm going to Operation Deep Freeze, to winter over [in Antarctica] to spend a year on the ice. I went down [for another physical] and everything was normal except my blood pressure and it was way u p . . . So this does have something to do with it. I didn't want to stay on that ice down there for a year. Here even the threat of a potentially disagreeable situation was considered to be enough to give this person his lifelong problem. It is not merely extraordinary events, but even the normal pressures of an occupation can build up to cause illness: Well, I been in t h e . . , car business 30 years and that'll drive you right up a t r e e . . , with my partner we had the largest Buick dealership in the state for ten years and I handled the car desk - all the pressure's on the car desk . . . I had phone calls from [other states], I had people from the office force coming in on the deals, I had seven or eight salesmen coming in on questions, I had sales meetings every morning. And that was six days a week. . . . I think it was the normal things of the car business [that caused my high blood pressure]. Once the cycle of Hyper-Tension and high blood pressure is started, it affects the person's behavior: The slightest emotional problem sends my blood pressure up, and so I've got to constantly work on it by myself and take deep breathing exercises and relaxing exercises... So I just watch anything that distracts me to a point, and I'll just stop it. And even though I might hurt the person's f~elingsI haven't got time to explain to them what I'm doing. Sometimes I just walk off and take care of myself. This person had had a 'nervous breakdown' for which he received psychiatric hospitalization and continues to receive outpatient therapy and medications. Despite the fact that his blood pressure was elevated before the breakdown (by his own report) he feels his Hyper-Tension beame a problem at that time and continues to be his major problem. As such, it permits the behavior he describes which otherwise would be socially unacceptable. A similar response was given



b y a p e r s o n w h o d e m o n s t r a t e s t h e s e m a n t i c linkage o f social a n d p h y s i o l o g i c a l processes: If I get a n g r y . . . I just turn around and go the other way and get away from it. Q. Quite literally you remove yourself from t h e . . . Right, the situation. And then when I get calmed d o w n . . . I go back when I'm not angry and I can sit down and talk about it rationally. Pressures, if I see too many pressures coming up I get busy doing something else and forget about them, or take care of them later, or let them take care of themselves. N o t m e r e l y w o r k s i t u a t i o n s , b u t also f a m i l y r e l a t i o n s h i p s are seen as b o t h a f f e c t i n g a n d b e i n g a f f e c t e d b y this illness. When you speak of Hyper-Tension I think it actually means that a person tends to lose control of his composure. I relate it to that. I'm rather high strung and very sensitive and at times it becomes quite severe and then it doesn't help out a situation . . . . For example, with my wife, we're close naturally . . . . [But] when I do something and she gives me a negative attitude, I go into a point where I say, 'This is the way it's going to be' and I become rather emphatic. And speak of Hyper-Tension, that's where it flares up. Because o f t h e e x t r e m e e f f e c t t h a t H y p e r - T e n s i o n h a s o n h o w t h e i n d i v i d u a l acts, it is u s e d to j u s t i f y o t h e r t y p e s o f illness b e h a v i o r i n c l u d i n g t h e sick role. An insurance salesman reported: I would say when you deal with individuals and don't want to show your innermost feelings you have a tendency to hold back your inner feelings and express your outer feelings. I feel that this puts an extreme pressure on an individual . . . I know in my job it's constantly answering the phone and dealing with different personalities that you can't tell people what you think or it's gonna ruin your career so consequently you hold these feelings back for an unforgiving haft an hour or fifteen or twenty minutes . . . . You gotta tell those S.O.B.'s off and get it over with. I say I can't call people that because if I do that I'm dead . . . . When I restrain myself for a period of time and don't tell people exactly what I think . .. I get weak . . . . At the office if I feel myself not feeling well and I feel sort of drowsy and down I just tell my secretary that I'm going home and don't send any calls home. So I just go home and get in bed and just relax, and I feel much better after two or three hours. It is very o b v i o u s t h a t this p e r s o n is u s i n g his illness as a m e a n s o f t e m p o r a r i l y a t t a i n i n g a sick role status, even t o t h e p o i n t o f b e i n g able t o j u s t i f y going h o m e t o bed. A n o t h e r e x a m p l e o f this c o m e s f r o m t h e d i r e c t o r o f a very large org a n i z a t i o n , d e m o n s t r a t i n g t h a t these p o p u l a r m o d e l s are n o t social s t r a t u m specific. I found that learning a tittle bit about hypertension and having it myself allows me to understand a key member of my staff who's been having a very serious problem in controlling his blood pressure with medical treatment, and I'm very much aware of behavioral pattern changes that rightly or wrongly I tend to attribute to this, and when he gets into one of the more extreme behavior changes, invariably if I force him down to the health center to check him, it's extremely up. I think this kind of understanding can be very helpful in treating this disease.



Again the sick role is used to justify otherwise unacceptable behavior: the choice this executive faced was to either declare his aide physically sick or he would be obligated to fire him. The existence of a popular model of HyperTension enables him to do the former. It should be reiterated that according to the biomedical research into high blood pressure, this type of behavior cannot be attributed to physiological causes (Kaplan 1978). If one takes medication to control the blood pressure, it has an effect on social behavior: Since I've taken these pills... I've never felt so good. Before that why I would be kinda sickly, I'd be scared to go here, scared to go there and [now] I don't mind going anywhere alone. My wife always had to be with me anywhere I went and just being scared I wouldn't make it . . . . Well, now it's all gone. Again, this response cannot be accounted for on a pharmacological basis. The popular model is powerful enough to produce marked behavioral changes which the professional model cannot explain except to label them 'the placebo effect.' OTHER MODELS As was pointed out earlier, the model developed thus far is that held by the majority of the people interviewed. There were some minority groups represented whose models of illness did not correspond to this and need to be mentioned. The most interesting of these models was that given by a Filipino. He felt that eating many kinds of meat f'filed up his stomach, which would cause his blood to rush up and flood his body, making it hot. This is turn would cause him to see flashing lights, and become dizzy and tired. Since the problem was excessive heat, his treatment was to eat the correct foods and take cool baths. This is very clearly a version of the hot-cold or humoral theory of illness which has been widely found in cultures under Spanish influence. (Harwood 1971). This finding is particularly significant in that it indicates that the primary model which was derived is not merely an artifact of the experimental methodology. If the same questions will give a hot-cold as well as a social stress model of illness we must be reaching popular conceptions at some level, and not merely creating another research myth. I was unable to achieve sufficient rapport with the two American Indians in this survey to establish the extent to which native belief systems affected their current illness beliefs. One said that he had completely rejected the old ways. Although there were nine blacks in this study, the models that they gave were by and large indistinguishable from the people in the rest of the study. When one was asked whether he was aware of any traditional family remedies he said: "Theirs' was a regular doctor, and for headache they took aspirin like everybody else."



One small cluster o f beliefs which was not represented in the composite diagram has to do with the concept o f harmful substances in foods, sometimes called "poisons." To quote: The raw meat or red meat, even if it's cooked, from what I understand has a chemical effect on the body that actually leads toward or builds up more toxic poisons in the system . . . . It brings in more toxic poisons faster than any vegetable source. The red meat supposedly has more of a toxic effect on the body than does poultry or fish, for example. Pork is particularly dangerous in this regard. These poisons are normally eliminated by the kidney; but if they build up can cause an excessive amount o f blood. This in turn raises the pressure in the b o d y system and may cause 'HyperTension.' This condition was also referred to as 'High Blood' (Snow 1974, 1977). The connecting link with the predominant cultural belief system was "increased pressure." From there, individuals giving this belief subsystem followed a more typical pattern. While a higher proportion o f blacks (30%) than whites (9%) gave elements o f this belief system, it was present in b o t h groups. This illustrates an important point: while a partictdar belief m a y be more prevalent among one ethnic group, there m a y be significant sharing o f belief systems across ethnic boundaries. A second point that this illustrates is that less prevalent models tie into the composite model at specific linking points. A person who holds a belief that is discrepant at one place may well hold beliefs that are consistent with the majority at other places. These points o f contact, or links, between minority or idiosyncratic belief systems and the majority model are extremely important in the clinical application o f this work. F o r example, the person cited for his beliefs about "poisons" subsequently went into a long diatribe against biomedicine for its reliance on pharmacological rather than nutritional therapeutics. He was asked why he came to the biomedical clinic if he felt that way: O.K. I come to this clinic because, first of all, I've been treated very well here, and [I] actually mean that from a patient standpoint. My blood pressure was detected not because I came here with a complaint of blood pressure, but for other kinds of things, and the blood pressure was detected. I think that nutrition is just one of the treatments or methods that is available to doctors and nurses . . . . I'm not discounting all of the major advances that traditional,medicine has made. I applaud them wholeheatedly, but it's just that I'm saying in conjunction with traditional American techniques, use nutrition also as a major contributing factor of curing or healing or correcting the disease process. So what I'm really saying is that I use both. I use the traditional treatments that I can get here along w i t h . . , the advice of a naturopath. I think they are both very valid. DISCUSSION G o o d (1977:53) presents four tasks that a theory o f illness language must perform. It should provide a framework for understanding:



1. The pathway linking the symbolicwith the affective and the physiologic. 2. The role of language in linking social experience to diseases. 3. The strategic use of illness language. 4. How change in medicallanguage is generated in broader social change. The model that has been presented here performs these tasks, and indeed, clarified some of the issues which were left problematic in Good's own work.

1. The pathways linking the symbolic with the affective and physiological.
People construe their experiences as meaning certain things, as having affected their persons as well as their bodies. In other words, certain experiences become symbolic for the individual. These symbols are used prospectively to structure a set of behaviors which are seen as appropriate in light of the meanings of the symbols - these would include illness behavior, sick role behavior, and post sick role rehabilitation - as well as being used retrospectively in reconstructing their memories to explain why a symbol which became manifest at a single point had deep roots in their past. Particularly powerful symbols, as were observed with some of the individuals in this study, were sufficient to focus all the experiences that person had had, and to overshadow the rest of his life. It is interesting to notice that although the symbol of Hyper-Tension was used to link affective experience with physiological events, the primary link was between affect in the past with affective experiences in the present and future. Thus past social experience operated through physiological mechanisms to affect current and future social stresses. Any mind-body dualism is well concealed here.

2. The role of language in linking social expe~qence to diseases. The terms used to accomplish the linkage of social stress and illness are extremely important. If this disease were called, for example, 'Korotkoff's syndrome' (after the man who first described the sounds used to measure blood pressure) it would be much less amenable to the socio-psychological interpretations that were used by this group. Both medical terms, hypertension and high blood pressure are based on root words that have meanings from affective language as well as from physics. Indeed, the commonsense meaning of hypertension particularly is precisely that given to it in the term I call Hyper-Tension. The same is true with high blood pressure. This was sometimes seen when people would drop out the middle component and refer to their illness as "high pressure problem." This too, is more amenable to psychosocial interpretation. However, the distinction between Hyper-Tension and high blood pressure is also important to understanding the semantics of linking the affective problem with a physical illness. Hyper-Tension is often the physical result of social stress; high blood pressure is one (of many) symptoms of Hyper-Tension. Because of this dichotomy, people could consider themselves "Hyper-Tensive" even though their blood pressure



readings were normal. Since all the people in this study were biomedically defined as having elevated blood pressure, 6 it is impossible to draw any conclusion about the prevalence of people who consider themselves to have Hyper-Tension but do not have biomedically defined hypertension. My anecdotal clinical experience indicates that there are a substantial number of such individuals.

3. The strategic use of illness language. This model demonstrates how people shift the explanatory models they present within a fairly wide range of variation without either affecting their basic illness belief structure or feeling a cognitive strain. Earlier, too, we commented on how easily the links between basic concepts could be shifted around. If the 'nodes' are analogous to the words in a dictionary, and the links to a grammar used to connect words, what we have been describing can be considered a language of illness. Unfortunately, very little is known either about the vocabulary or the underlying logic by which concepts are linked to each other to provide meaning and behavioral options in a specific setting. While no direct data are available, it is easy to see how certain parts of the model would be more likely to be expressed in a clinical setting, whereas others are more likely to be used at home or work. The use of this illness idiom as a means of manipulating the home or work environment has also been well documented. Here there are two matrices which intersect: the cognitive matrix and the social interaction matrix. Elements of each are actively selected to give the individual a variety of options which can be activitated to reach a desired outcome.

4. How change in medical language is generated in broader social change. This task is not as clearly performed by this model; nonetheless, a foundation for understanding changes in illness semantics is laid. The model is, above all, not static but is constantly being reinterpreted, reworked to meet changing circumstances. New elements can be added from idiosyncratic experience or popular or professional explanatory models. Relationships between model elements have similarly been shown to be flexible, readily moulded to meet the explanatory needs of the individual. It should be evident that different illnesses will have differently shaped cognitive domains. In the particular instance of a chronic nonincapacitating illness with ample opportunity for individual exploration of a variety of themes, we would expect a tight cognitive domain with many opportunities for explanatory model development. We would expect less dependence on the professional model and more on popular and idiosyncratic explanations. In an acute, overwhelming, uncommon illness, we would expect a narrower cognitive domain,



with fewer behavioral and treatment options, and greater reliance on professional expertise. The implications of these variations will have to be worked out as we investigate the various meanings that illnesses can have in the American culture. Good's clinical ethnography did not show how one set of illness beliefs were interrelated with the beliefs ab~,ut other illnesses. While we have not discussed this extensively, there are a set of connecting links that should be pointed out. One of the major body systems involved in this illness belief system is the effect on the person's heart. Although the model was developed as an explanation of hypertension, the rudiments of a popular model of "heart disease" can also be seen. Since there are three different arrows leading into the node "heart attack" we can predict that there will be at least three different meanings attributed to this term. A heart attack can mean: a) blood vessel closing off, causing the heart to die; b) the heart being so overworked that it just gives up; or c) a blood vessel bursting causing an "explosion" in the heart. While this needs to be more extensively studied, it provides tantalizing glimpses of the links which tie beliefs about Hyper-Tension to beliefs about other illness. Plain folk say 'Hyper-Tension'; the experts say hypertension, and each thinks the other is taking about the same thing. Perhaps it is this muddying of the waters which allows both to function without cognitive dissonance becoming so great as to cut off interaction. The physician doesn't have to deal with the social stresses that don't fit hismodel, and the layman can reshape the physiological explanation of biomedicine to a more amenable form. Because of the distance that is effected by two disparate def'mitions, the therapeutic alliance is not threatened. But there are occasions when dissonance caused by different models of illness does impede healing. At those times, a full understanding of the illness belief systems which are available to the layman and to the physician, if coupled with a willingness to negotiate a more functional set of explanatory models, may pave the way to a richer, deeper, and above all more satisfying experience to healing. Beyond the actions of healing and beyond even the theoretical understanding of the cognitive aspects of illness in America lie the problematic issues of the relationship between illness beliefs and illness behavior. Data from this study which are not described here will help clarify this. Early analysis indicates that, when this model is cast into a form amenable to statistical analysis, I will be able to explain a greater proportion of the variance in medication taking behavior than has been heretofore possible. Additionally, a greater proportion of the impact of sickness on everyday life can be interpreted using this model than has been possible previously. Forthcoming papers will deal with these findings.

University of Washington




1. Acknowledgments. This study was supported by the Robert Wood Johnson Clinical Scholars Foundation of the University of Washington and the Health Services Research and Development Affiliation Program of the Seattle Veterans Administration Medical Center. Opinions expressed in this paper are those of the author and not necessarily those of the Robert Wood Johnson Clinical Scholars Foundation. An earlier version of this paper was presented to the American Anthropological Association in Cincinnati, Ohio, December 1, 1979. I am greatly indebted to Pamela Amoss, Noel Chrisman, Tom Inui and Arthur Kleinman, all of the University of Washington. Each gave invaluable assistance at all points in the research process and made extensive comments on earlier drafts of this paper. Special mention must be made of Linda LaRocque who performed yeoman duty in transforming often nearly inaudible tape recordings into an accurate transcription. 2. I prefer the term "Biomedicine" for that type of medical practice that elsewhere is referred to as "scientific," "western," "cosmopolitan," or "modern." "Scientific" and "modern" tend to make the competitors sound unscientific and old fashioned. "Western" is too limiting. "Cosmopolitan" has a variety of connotations which are unfortunate. After all, biology is perceived as the source of the theory underpinning this mode of medicine, just as the Vedas are seen as the source for Ayurvedic medicine. In each case, there are regional variations as well as other sources of specific practices which should not distract us. 3. In fairness to Press, the physicians who were accessible to the urban poor he was studying were practicing in a government clinic setting where they were interchangeable, and thus in a sense, monolithic. This generalization would not hold for other social strata in that city, a point he fails to make. 4. This is not to say that solutions which are worked out over generations actually affect the pathophysiological course of the disease except insofar as they muster social support (Cobb 1976). One of the romantic assumptions is that trial and error over a long period of time will lead to a corpus of efficacious medical interventions. An understanding of the difficulty of proving the efficacy of any therapy - despite powerful interventions, powerful statistical techniques and large population samples - will quickly dispel such notions. 5. While Nurse Practitioners were the "experts" here and not physicians, the distinction did not seem to make any difference to the people in the study. These nurses hold an anomalous position, sometimes referred to as "Nurse" and sometimes as "Doctor," indicating their ambiguous status. In one particularly memorable instance, this anomalous status was indicated by someone who referred to the same practitioner first as "nurse" and "she," then as "doctor" and "he"! 6. Interestingly, about 7% of the people seen in this sample did not meet the clinic criteria for the definition of high blood pressure - i.e., did not have the disease hypertension. All of them reported having the illness Hyper-Tension. The reason for this is that while a single elevated blood pressure measurement is not sufficient to diagnose the presence of the disease, it is enough to be referred to the hypertension clinic for treatment. Once a person is in the clinic, it is very difficult to discharge him.
R E F E R E N C E S CITED AmaraSingham, L. R. 1980 Movement among healers in Sri Lanka: A case study of a Sinhalese patient. Culture, Medicine and Psychiatry 4 : 7 1 - 9 2 . Beeson, P., W. McDermott and J. Wyngaarden 1979 Cecil Textbook of Medicine. Philadelphia: W. B. Saunders.



Berger, P. L., and T. Luckmann 1967 The Social Construction of Reality. Garden City; New York: Doubleday & Co. Bonjean, C., R. J. Hill and S. D. McLemore 1967 Sociological Measurement: An Inventory of Scales and Indices. San Francisco: Chancellor Publishing Co. Buckley, W. 1967 Sociology and Modern Systems Theory. Englewood Cliffs, N. J.: Prentice-Hall. Chrisman, N. 1977 The health seeking process: An approach to the natural history of illness. Culture, Medicine and Psychiatry 1: 351-378. Cobb, S. 1976 Social support as a moderator of life stress. Psychosomatic Medicine 3 8 : 3 0 0 314. Curtis, L. 1976 Understanding High Blood Pressure. Pamphlet published by Searle & Co., San Juan, Puerto Rico. Duma, F. 1976 Traditional Asian medicine and cosmopolitan medicine as adaptive systems. In C. Leslie (1976) op. cit. Freidson, E. 1970 Profession of Medicine: A Study of the Sociology of Applied Knowledge. New York: Harper & Row. Gaines, A. 1979 Definitions and diagnoses: Cultural implications of psychiatric help seeking and psychiatrists definitions of the situation in psychiatric emergencies. Culture, Medicine and Psychiatry 3: 381-418. Good, B. 1977 The heart f what's the matter" The semantics f illness in lran" Culture' Medicine and Psychiatry 1:25 - 5 8 . Harwood, A. 1971 The hot-cold theory of disease: Implications for treatment of Puerto Rican patients. Journal of the American Medical Association 216:1153-1158. Helman, C. G. 1978 "Feed a cold, starve a fever" - Folk models of infection in an English suburban community and their relation to medical treatment. Culture, Medicine and Psychiatry 2:107-138. Horton, R. 1967 African traditional thought and Western science. Africa 37:50-71. Kaplan, N. 1978 Clinical Hypertension. Baltimore: Williams & Wilkins. Kleinman, A. 1980 Patients and Healers in the Context of Culture. Berkeley: University of California Press. Kleinman, A. et al. 1976 Medicine in Chinese Cultures: Comparative Studies of Health Care in Chinese and Other Societies. Washington, D. C.: U. S. Government Printing Office for Fogerty International Center, N. I. H., DHEW Publication No. (N.I.H.) 7 5 - 6 5 3 Stock No. 0 1 7 - 0 5 3 - 0 0 0 3 6 . Kuustadter, A. 1975 Do cultural differences make any difference? Choice points in medical systems available in northwest Thailand. In Kleinman et al. (1975) op. cit. Leslie, C. 1976 Asian Medical Systems. Berkeley: University of California Press.



Lewis, G. 1975 Knowledge of Illness in a Seplk Society. London: Athlone Press. Lindenbaum, S. 1979 Kuru Sorcery. Palo Alto: Mayfield Publishing Corp. Mechanic, D. 1978 Medical Sociology. New York: The Free Press. Press, I. 1969 Urban illness: Physicians, curers and dual use in Bogota. Journal of Health and Social Behavior 10:209-218. 1978 Urban folk medicine: A functional overview. American Anthropologist 8 0 : 7 1 84. 1980 Problems in the definition and classification of medical systems. Social Science and Medicine 14B:45-58. Snow, L. 1974 Folk medical beliefs and their implications for cure of patients. Annals of Internal Medicine 81:82-96. 1977 Popular medicine in a black neighborhood. In Spicer, 1977, op. cit. Spicer, E. 1977 Ethnic Medicine in the Southwest. Tucson: University of Arizona Press. Stone, G. 1979 Patient compliance and the rote of the expert. Journal of Social Issues 35:34-59. Taussig, M. 1980 Reification and the consciousness of the patient. Social Science and Medicine 14B:3-14. Turner, V. 1967 The Forest of Symbols. Ithaca: Cornell University.