Linda Garro 1998
Linda Garro 1998
Linda Garro 1998
GARRO
Department of Anthropology
University of California at Los Angeles
With reference to both critiques and empirical studies, the theoretical and
methodological grounding of anthropological research on medical decision
making is examined in this article, giving particular attention to the con-
struction and evaluation ofcognitively oriented decision models. A decision-
modeling study carried out in the Mexican village of Pichdtaro (in con-
junction with James C. Young) frames an exploration of some of the ten-
sions and points of contention about the aims and designs of cognitively
oriented studies of decision modeling. While a decision model can provide a
reasonably good guide to an understanding of treatment actions and the
culturally based rationality that underlies them, such models fall short when
they are oriented primarily around predicting treatment accounts. They
should also attend to the jointly cultural, personal, social, and cognitive
constructive processes through which meaning is conferred upon the oc-
currence of illness, [care seeking, decision models, decision making, Mexico]
Medical Anthropology Quarterly 12(3): 319-340. Copyright © 1998 American Anthropological Association.
319
320 MEDICAL ANTHROPOLOGY QUARTERLY
While not all decision-making studies are cognitively oriented, it is also the
case that not all decision-making studies are overtly concerned with developing
and testing decision models (e.g., Kayser-Jones 1995; Nardi 1983; Sargent 1982,
1989). A common decision-modeling strategy is to build a decision model using
interview data and other information obtained primarily from one sample and then
to validate the model using decisions made by a second, independent sample. In
this article, reference to a decision-making approach or decision-making perspec-
tive refers to an inclusive higher-level category, with a "decision model"—the
more formal representation of the decision-making process that can be evaluated
using actual choices—as a subcategory.
An alternate subtitle for this article is "Pichataro Revisited." Although the ar-
ticle reflects my current thinking, it draws on findings of an earlier decision-mak-
ing study carried out (in conjunction with James C. Young) in Pichataro, a Mexi-
can town where both Purepecha (Tarascan) and Spanish are commonly spoken
(see J. C. Young 1981; Young and Garro 1994).1 This cognitively oriented study
implemented the two-pronged research design described in the preceding para-
graph, separating the data used to construct the decision model from the illness
case histories used to evaluate the model. The Pichataro study provides a base for
exploring the theoretical grounding of decision-making and decision-modeling
studies. The review of relevant literature presented in this article is necessarily se-
lective, but serves to highlight themes and points of debate. An objective is to fur-
ther a discussion of the merits and the limitations of decision-making studies, espe-
cially those that are explicitly cognitive in orientation. To what extent do such
studies open onto an understanding of the rationality involved in making decisions
in a given cultural context?
Some time ago H. Gladwin and Murtaugh observed that "natural decision-
making researchers have various goals" (1984:115). Stating that their primary ob-
jective is to predict behavior to "be useful for policymaking," they note that other
researchers have different goals with some being "interested in finding out the gen-
eral heuristics and procedures widely used in a culture and across cultures to make
decisions" (1984:115). While they regard this objective as "a crucially important
task for cognitive anthropology," it is "somewhat separate" from the concerns that
motivate their work (1984:115). Others see cognition and policy as intertwined,
but they discemibly foreground the policy implications. For example, C. Gladwin
declares that "the goal of decision studies" is to "model how people make real-
world decisions and to identify the specific decision criteria used by most individu-
als in a group in order that policymakers might intervene in the decision-making
process with new policies designed to make things better for the targeted group"
(1989:86). Recent medical anthropological studies that emphasize an applied pol-
icy orientation include Ryan and Martinez (1996) and Bauer and Wright (1996). I
do not review these primarily policy-oriented decision-making studies in this arti-
cle. I focus more specifically on writings that address issues relating to the role of
cognitive theory.
The outline for the remainder of this article is as follows: the first section pre-
sents some broadly based critiques leveled at the theoretical underpinnings of
anthropological studies of care seeking and, at a general level, decision-making ap-
proaches. The second section is a brief overview of theoretical perspectives on de-
cision making, both within and outside anthropology. Both sections highlight
PART 1: DHCISION MODELS OF TREATMENT CHOICE 321
some of the diverse ways scholars think about rationality and decision making,
though it is beyond the scope of this article to delve too deeply into these matters.
Nevertheless, the material covered in these two sections show how the idea of ra-
tionality is "ambiguous," "slippery," and "subject to multiple definitions" (Shore
1996:168-169). A brief sketch of the Pichataro study is presented next, providing a
backdrop for examining some of the tensions and points of contention about the
aims and design of cognitively oriented studies of decision making. Following the
lead of Mathews (1987), the discussion highlights the "framing" of decision situ-
ations as integral to the process of decision making. The companion piece to this
article (Part 2: Divergent Rationalities) explores these issues further in relation to a
study of health care decision making carried out in an Anishinaabe community in
Manitoba, Canada.
own interests and their own satisfactions" and presupposing a "universal Eco-
nomic Man," the role relegated to culture is to provide a "relativized set of prefer-
ences" (Sahlins 1976:102). Consequently, only "the actors (and their interests
taken a priori as theirs) are real" (1976:102); culture "is reduced to an epipheno-
menon of purposeful 'decision-making processes' (as they say)" (1976:86). Al-
though it is never quite clear whether Good's appraisal is intended to encompass
the Pichataro study or not,2 he notes that "utilitarian assumptions often appear in
the common-sense reasoning" in the literature on care seeking (Good 1994:47; see
also 1994:180-181 and Brodwin 1997:82-83). Good goes on to state: "The ana-
lytic conjunction of the utilitarian actor, instrumental beliefs that organize the ra-
tional calculus of care-seeking, and ethnomedical systems as the sum of strategic
actions is uncomfortably consonant with neo-classical economic theories of the
utilitarian actor, the market place, and the economic system as precipitate of value-
maximizing strategies" (1994:47). By suggesting points of correspondence, Good
challenges anthropologists studying care seeking to examine the extent to which
their assumptions and objectives, both implicit and explicit, formulate culture as
"practical reason."
Through their portrayals, other scholars raise objections to the assumed cog-
nitive underpinnings of some anthropological accounts of care seeking and deci-
sion making. Brodwin, for example, notes that his own study "does not isolate
health-seeking as a set of rule-bound observable actions" (1996:191). He situates
his work "away from the utilitarian and cognitive frameworks used by many an-
thropologists to study people's use of medical services in developing countries
(e.g., Sargent 1989; Young and Garro 1982)," frameworks that he says are geared
to portraying "the decision-making calculus that individual care-seekers may fol-
low" (Brodwin 1997:82-83). Writing about "the recent cognitivist shift in the hu-
man sciences," Bibeau depicts the "dominant mechanistic view" as one whereby
"human beings are predominately shaped by 'cognitive blueprints' stored in the
brain in the form of mental maps, scripts, frames, and scenarios that shape actions
of individuals along predetermined lines," which "scientists can leam to 'read' in
order to predict actual behaviors of individuals" (1997:250). As reality is far more
complex than such a view allows, Bibeau urges anthropologists to counteract this
trend by concentrating on the "social grounding of meaning" as revealed by "ac-
tions, interactions, and practices" (1997:250).
Luhrmann observes that the assumption that "people have a coherent clear-
cut set of beliefs" often underlies talk of "rationality" (1989:321; cf. Kirmayer
1992; A. Young 1981, 1982). A corollary assumption is that "people act on their
beliefs, that beliefs are prior to action" (Luhrmann 1989:321; see also A. Young
1981, 1982). Complementing Bibeau's comments about "cognitive blueprints,"
Kirmayer elaborates:
In opposition to this view, Luhrmann and Kirmayer contend that "beliefs are not
fixed or consistent, for they are often presented to justify some action" (Luhrmann
1989:353; see also Kirmayer 1992:330). Although it is unclear whether verbal dis-
course is to be considered a form of action, Bibeau pushes for the "recognition that
behavior precedes belief, that cultural beliefs are revealed by actions, and that the
belief/behavior relation must be inverted" (1997:250).3 Writing from a different
perspective and research agenda, Boster cautions that much of what researchers
identify as influencing choices may instead be "post hoc rationalizations" (1984:387;
for similar admonitions see Good 1986:166-167 and Bloch 1998:25).
In this article, as well as in its companion, I argue for a view of human cogni-
tion that is both flexible and grounded in social and cultural processes. First, with
reference to the preceding discussion about "subjective utilitarianism," and with
the elusive concept of "rationality" again front and center, I establish that some of
the concerns raised above are more closely associated with normative approaches
to decision making than with the more descriptive approach typically adopted by
anthropologists. The general thrust of the position developed here is that while
there is definitely a pragmatic cast to health care decisions, a decision-making ap-
proach does not necessarily entail a formulation of culture as "practical reason"
grounded in a "utilitarian calculus" in the way that the writings of Good (1994) and
Brodwin (1997:73, 82) suggest.
Theoretical Perspectives on Decision Making
Aspiring to construct theories that will account for all types of choice-making
behavior and traversing diverse fields—including economics, psychology, and ad-
ministrative studies—the literature on decision making is voluminous and intri-
cate. At a very general level, a distinction is often drawn between normative and
descriptive models. As Abelson and Levi explain,
decision models can be oriented toward either how people should choose (norma-
tive models) or how they do choose (descriptive models). When a disparity be-
tween the two exists, it raises issues of human rationality. [ 1985:232]
Normative decision models are prescriptive, structuring decision problems in
terms of probabilities and utilities to reach an optimal decision (i.e., one that maxi-
mizes benefits and minimizes losses or costs). The systematic evaluation of alter-
natives for reaching an optimal decision is seen as a rational process; indeed, it is
seen to represent how one ought to reason. Rationality is thus presumed, with de-
faults leading to the question of "why do people not behave rationally?" (Abelson
1976:61). The normative model, postulating a "universal Economic Man" (and an
internal "rational calculus"), comes closest to the "subjective utilitarianism" posi-
tion put forward by Sahlins. Still, explicitly normative models have not garnered
much support from medical anthropologists (though see Fabrega 1973,1974).
Descriptive decision theorists claim that models of the normative type are not
psychologically plausible, as inferences about underlying mental processes, be-
cause they attribute unrealistic information-processing capabilities to the decision
maker (Quinn 1978; Tversky 1972; Tversky and Kahneman 1974). The normative
"rational" decision process is simply an "idealization that isn't there" (Abelson
1976:61). These theorists argue that instead of probability estimations, complex
decisions are made by relying on procedures that simplify the kinds of cognitive
324 MEDICAL ANTHROPOLOGY QUARTERLY
a simplified model of the situation in order to deal with it. He behaves rationally
with respect to this model, and such behavior is not even approximately optimal
with respect to the real world. To predict his behavior we must understand the way
in which this simplified model is constructed, and its construction will certainly be
related to his psychological properties as a perceiving, thinking and learning indi-
vidual. [1957:199]
Although omitted from Simon's formulation and from most other nonanthro-
pological accounts of decision making, it can be added that such a "simplified
model of the situation" is culturally grounded, as cultural understandings or cul-
tural models frame our understanding of how the world works (Quinn and Holland
1987; see Fjellman 1976:88-89).4 Because they emphasize describing or repre-
senting the culturally grounded decision-making process, anthropological ac-
counts are not overtly concerned with academic controversies about "apparently ir-
rational beliefs" (see Sperber 1982, 1985: ch. 2) as evidenced by the sustained
"rationality debate" (e.g., see discussion in Good 1994:10-14 and Shore 1996). If,
for example, affliction attributed to witchcraft is seen to require a specific form of
treatment, this connection can be incorporated in a decision model or other depic-
tion of the decision-making process. Descriptive decision approaches do not deny
"cultural reason" (Sahlins 1976:170).
Another avenue of inquiry among descriptive decision theorists concerns
how the formulation or framing of a decision situation affects choices made. Psy-
chological experiments demonstrate that different ways of framing the same situ-
ation lead to different decisions (e.g., Tversky and Kahneman 1981). The general
conclusion of these experiments is that "even the most elementary normative prin-
ciples cannot be taken as descriptively valid" (McNeil et al. 1988:567). Although
debates about the nature of human rationality are not absent (see Abelson and Levi
1985:232-235), the general aim of descriptive decision theory is to understand the
reasoning process and not to assess whether reasoning occurs in the right way (see
Stein 1996:17-18).
The oppositional pairing of rationality against irrationality and of reason
against emotion have long structured discussions of human cognition and behav-
ior. Nevertheless, a growing contingent of scholars view emotion as integral to
cognitive processes. Damasio, a neurologist interested in the biological underpinnings
of reason and decision making, maintains that "certain aspects of the process of
emotion and feeling are indispensable for rationality" (1994:xiii). He reviews a di-
verse set of findings (including the psychological experiments on framing men-
tioned in the preceding paragraph) to support his assertion that "emotions and feel-
ings may not be intruders in the bastion of reason at all: they may be enmeshed in
its networks for worse and for better" (1994:xii). Within cognitive anthropology
the interdependence of cognition, emotion, and motivation is acknowledged in a
PART 1: DECISION MODELS OF TREATMENT CHOICE 325
These illnesses were diagnosed relatively infrequently. Less than 10 percent of the
case histories we collected were considered curable only by remedios caseros. In
addition, such a diagnosis was usually not the first diagnosis made.
The study of treatment decision making in Pichataro consists of two main
parts. The first part centers on understanding what people do when confronting ill-
ness in order to construct the decision model, the second on evaluating the decision
model using illness histories recorded during a series of visits to a sample of families.
As the basis for building the decision model, we attempted to understand the
nature of the knowledge that individuals bring to the occurrence of illness, how this
knowledge is used to make sense of illness, and the process of making treatment
decisions. The fieldwork yielded opportunities to observe how illness episodes un-
fold, serving to enrich, corroborate, and challenge our evolving understandings
and hypotheses about what people do when faced with illness. Informal talks and
interviews were also important sources of information.
In addition, a number of structured interview approaches were employed. We
designed several interview formats to explore cultural knowledge about illness and
its treatment. These included a "term-frame" interview, which systematically
paired illnesses and propositions, and ranking tasks, which explored assessments
of faith in different treatment alternatives for a set of illnesses and judgments of se-
verity for separate sets of illnesses and symptoms. Other interviews were designed
to learn about patterns of care seeking, such as through contrastive questioning
about treatment alternatives, the posing of hypothetical illness situations, and re-
cording family-based case histories of past illnesses.
To minimize the potential for post hoc rationalization of choices, Boster ad-
vises ethnographers to "elicit informants' evaluations of alternatives indepen-
dently of a particular decision situation" (1984:347). The comments and qualita-
tive judgments put forward in the structured interviews (and obtained apart from
actual illness cases) represent a source of information consistent with Boster's ad-
vice. There was considerable convergence among the multiple sources of informa-
tion used in constructing the decision model. While verbal statements about illness
and treatment actions are the underpinning of the model, the resulting repre-
sentation is not simply a distillation of what we were explicitly told. No one told us
"this is how we decide what to do when someone is ill." Rather, the two strategies
underlying the decision model are theoretical inferences, and the decision model it-
self is a formal representation of these strategies which also incorporates con-
straints on choice.
Across a variety of settings, four criteria consistently came up as important
considerations in the choice of treatment: (1) gravity of the illness; (2) whether an
appropriate home remedy is known for the illness; (3)/e (faith or confidence) in the
effectiveness of remedios caseros or remedios medicos for a given illness; and (4)
expense of treatment and the availability of resources.
The decision-making process in Pichataro can be understood with reference
to two basic strategies or general principles which relate these four criteria to ex-
pected orderings in the use of the treatment alternatives. When an illness is not seri-
ous, the general pattern is "cost-ordered," starting with alternatives that are less
costly, and turning to the most expensive alternative, usually the physician, only as
a last resort after the less costly options have been exhausted. The pattern is to try
self-treatment first, unless an appropriate home remedy is not known. If no home
328 MEDICAL ANTHROPOLOGY QUARTERLY
pragmatically oriented, but human action is understood "as mediated by the cul-
tural design, which gives order at once to practical experience, customary practice,
and the relationship between the two" (Sahlins 1976:55).
of a home remedy and "faith" in treatment for the illness at a household level. Bias
potentially introduced through reliance on these procedures would be in the direc-
tion of indicating the model as failing to account for a given choice, rather than in
the direction of overstating its success rate.
The decision model successfully predicted a high proportion of both initial
and subsequent treatment choices, approximately 90 percent of all treatment ac-
tions. Even when initial home treatments, which represent "a largely routine initial
response to treatment," were excluded from the calculations, the prediction rate is
still more than 80 percent. It was assumed that what remained after excluding in-
itial home treatment involved "active decision making" (Young and Garro
1994:166). These prediction results were interpreted as providing "strong evidence
for the basic validity of the model, and for the contention that the considerations
and assumptions embodied in it represent important aspects of how people actually
make these choices" (Young and Garro 1994:166).9 We also underscored the need for
additional research on "the cognitive bases of the treatment decision-making process,"
in particular on "the dynamics of the faith concept" (Young and Garro 1994:180).
guides to actions, guides that may be modified by new experiences. Concepts and
experience, knowledge and action, are complexly interdependent (Luhrmann
1989:353).
In addition, as perhaps first discussed in Frake's (1961) study on the diagnosis
of "skin diseases" among the Subanun, the evaluation of illness may become a "so-
cial activity" that involves negotiating the relevance of culturally shared knowl-
edge about illness and treatment to particular cases (cf. Chrisman 1977). In
Pichataro, speculating about diagnosis, possible cause, and appropriate treatment,
was common. Cultural knowledge serves as a resource, but there are often multiple
explanatory frameworks that may be culturally and/or personally applicable to a
given situation. Further, while theories of decision making are based on the ideal of
an individual decision maker, actual decision making may be grounded in a social
unit(cf.Mathewsl987).
Even though the outcomes of these social and constructive processes are cen-
tral in evaluating the decision model for Pichataro, these processes are themselves
accorded scant attention in the representation of the decision-making process. In-
stead, the focus is on discovering strategies hypothesized to underlie decision mak-
ing and on relating these strategies to treatment actions. As described in the follow-
ing section, some cognitive anthropologists suggest that decision-making studies
rely too much on predicting outcomes and too little on understanding the underly-
ing cognitive process.
She analyzes transcripts of two conversations in which family members reach de-
cisions about appointing auxiliary personnel for religious festivals in Oaxaca,
Mexico. Mathews's descriptive account points to the interrelationships among
various goals and how they are embedded in scenarios and appraised in reaching a
decision. Who the family members choose as auxiliary personnel "depends ulti-
mately on what they hope to accomplish by having a successful festival which, in
turn, depends on the goals determining their participation in religious service in the
first place" (1987:56). Families differ in assigning priority to one goal over another
in reaching a decision, although they may still try to partially satisfy the goal ac-
corded less priority. Mathews concludes,
Only by investigating how informants represent decisions and structure their solu-
tions can we begin to work toward the formulation of a general theory of human
understanding that will enable us to specify the ways in which individuals use cul-
tural knowledge and personal experience to generate, organize, and select among
goals for the purposes of making a decision. [1987:58].
I do not wish to argue at a general level against Mathews's position; her key
points are cogently presented. However, even though goals per se are not stipulated
in the explanation of treatment choice developed for Pichataro, I submit that they
are encompassed by the two general proposed orderings. Additionally, the goals
and their prioritization are implicit in the statements made by Pichatarefios from
which these orderings derive. Embedded in the two orderings are three goals: alle-
viating sickness, ensuring that a cure is achieved, and minimizing expenditures.
While treatment actions are motivated by the overarching goal of alleviating sick-
ness, the two remaining goals may come into conflict because treatment by a phy-
sician is, in most instances, the alternative judged most likely to achieve a cure and
also the most expensive; self-treatment is judged least likely to lead to a cure and
also the least expensive. Whether one or the other of these goals is prioritized over
the other depends on the severity of the illness. For nonserious illness, the cost-or-
dered strategy emphasizes conserving resources while still attempting to alleviate
the illness. When illness is serious, priority is placed on obtaining a cure. While
constraints on the realization of goals are not featured in the examples Mathews
provides, constraints definitely impact on goals for treatment choices in Pichataro.
Even if the objective of ensuring a cure is accorded priority at an ideational level,
constraints may lead to a situation where minimizing cost assumes greater salience
in determining the reported course of action at the phenomenal level.
I have gone into some detail about implicit goals in the explanation of treat-
ment choice for Pichataro because I want to suggest that Mathews's higher-level
insight about the lack of regard in decision-modeling studies for how individuals
"represent" or "frame" decisions is even more important than explicitly attending
to goals. As Tversky and Kahneman explain, it is "often possible to frame a given
decision problem in more than one way" (1981:453). But it is the framed situation,
and not the process of speculating or converging on a particular framing, that deci-
sion models are set up to handle. Decision models of treatment choice (e.g.,
Mathews and Hill 1990; Ryan and Martinez 1996; Young and Garro 1994) are ori-
ented around accounting for a series of treatment actions. But they generally treat
the framing of decisions either as unproblematic or as a given in constructing and
evaluating a decision model.
334 MEDICAL ANTHROPOLOGY QUARTERLY
Concluding Remarks
As a framework for exploring the relationship between the phenomenal and
ideational orders, I have brought forward a series of issues raised about decision-
making and decision-modeling studies in the anthropological literature. While a
number of these issues are complex and difficult to resolve, I have suggested some
ways of thinking about them and have emphasized the value of converging data.
With the Pichataro study as a backdrop, I have argued that a decision model can
provide a reasonably good guide to an understanding of treatment actions and the
culturally based rationality underlying such decisions. At the same time, I have
also stressed the need for greater attention to the jointly cultural, personal, social,
and cognitive constructive processes through which meaning is conferred on the
occurrence of illness. In line with the arguments put forward by Mathews (1987),
the position taken here is that decision models do not adequately explain how deci-
sion makers draw upon cultural and personal knowledge in framing decision prob-
lems and structuring their resolution. This is an area for future research.
One issue that has not been raised in this article is whether a decision model is
a general-purpose approach. In the following companion piece, I discuss my efforts
PART 1: DECISION MODELS OF TREATMENT CHOICE 335
NOTES
treatment choice for Pich£taro (Young and Garro 1994), lead to the predictions of the cir-
cumstances under which a less preferred source of treatment is "selected."
7. Although she does not provide further elaboration for this point, it is a difficult one
to defend. In the case of Gladwin's proposed hierarchical decision structure, whether the in-
dividual-level cognitive process even approximately follows a hierarchical order is an open
question. It is one thing to state that "tree models use more realistic assumptions about indi-
viduals' cognitive capacities than do linear additive decision models" (Gladwin 1989:11),
and quite another to assert that the tree model captures cognitive processes at an individual
level.
8. The usage offolk follows the definition given by Press: " 'Folk medicine' should be
strictly limited to describing systems or practices of medicine based upon paradigms which
differ from those of a dominant medical system of the same community or society
(1980:48).
9. Here it should be noted that the implications of a highly predictive model are a bit
more modest than what Gladwin (1989) claims, especially in regard to representing cogni-
tive process.
10. Statistical modeling approaches have been proposed as alternatives to decision
modeling (e.g., Chibnik 1980). When applied to treatment decision making, however, one
significant limitation of prediction-oriented statistical models is that they do not adequately
capture the processual nature of illness and care seeking (Mechanic 1979; Weller et al.
1997:225-226). With the objective of discovering statistically significant patterns of asso-
ciation between the use of treatment alternatives and the independent variables examined,
such statistical models focus on a decision outcome and not a sequence of outcomes (see
Kroeger 1983; Stoner 1985 for reviews). A recent study by Weller et al. (1997) comparing a
multivariate modeling approach and a decision-modeling approach can be used to illustrate
this point. In describing the research situation, the authors write: "Seeking a cure for illness
on the Pacific Coast of Guatemala is a dynamic process and throughout the process, a series
of sources and treatments may be sought" (1997:231-232). Nonetheless, the analysis using
the two approaches, as well as their comparison, is confined to the prediction of the initial
treatment alternative used. Because subsequent treatment choices are not independent of
preceding actions, initial treatment actions are studied since these alone are amenable to sta-
tistical analysis. But a limited ability to predict an initial treatment action is incomplete as an
account of care seeking; the initial treatment used cannot stand for an illness episode. It is
also perhaps worth noting that except for dichotomous judgments of severity, none of the
variables in the multivariate modeling analysis pertain to how individuals confer meaning
on an occurrence of illness, nor is this a particular focus in the interviews used to build the
decision model. Whether multivariate models are able to deal with "ideational" aspects (e.g.,
"faith," etiological assessments), which have formed part of other decision models and are
often featured in descriptive anthropological accounts of care seeking in medically plural-
istic settings, remains an open question.
11. As schemas can have motivational force (D'Andrade and Strauss 1992), such a
view can accommodate Kirmayer's position that emotions "determine not what is logical to
do to achieve certain ends but what ends are most pressing in a given situation" (1992:330).
12. Nardi's use of the term model comes close to the concept of a cultural schema or
cultural model even though Nardi's paper predates the 1987 volume edited by Holland and
Quinn devoted to developing this concept.
REFERENCES CITED
Abelson, Robert P.
1976 Social Psychology's Rational Man. In Rationality and the Social Sciences. S. I.
Benn and G. W. Mortimore, eds. Pp. 58-89. London: Routledge and Kegan Paul.
PART 1: DECISION MODELS OF TREATMENT CHOICE 337
Mechanic, David
1979 Correlates of Physician Utilization: Why Do Major Multivariate Studies of Physi-
cian Utilization Find Trivial Psychosocial and Organizational Effects. Journal of Health
and Social Behavior 20:387-396.
Nardi, Bonnie A.
1983 Goals in Reproductive Decision Making. American Ethnologist 10:697-714.
Nisbett, Richard, and Lee Ross
1980 Human Inference: Strategies and Shortcomings of Social Judgment. Englewood
Cliffs, NJ: Prentice-Hall.
Press, Irwin
1980 Problems in the Definition and Classification of Medical Systems. Social Science
and Medicine 14B:45-57.
Quinn, Naomi
1978 Do Mfantse Fish Sellers Estimate Probabilities in Their Heads? American Eth-
nologist 5:206-226.
Quinn, Naomi, and Dorothy Holland
1987 Culture and Cognition. In Cultural Models in Language and Thought. D. Holland
and N. Quinn, eds. Pp. 3-40. Cambridge: Cambridge University Press.
Ryan, Gery, and Homero Martinez
1996 Can We Predict What Mothers Do? Modeling Childhood Diarrhea in Rural Mex-
ico. Human Organization 55:47-57.
Sahlins, Marshall
1976 Culture and Practical Reason. Chicago: University of Chicago Press.
Sargent, Carolyn
1982 The Cultural Context of Therapeutic Choice: Obstetrical Care Decisions among
the Bariba of Benin. Dordrecht, Netherlands: D. Reidel.
1989 Maternity, Medicine and Power. Reproductive Decisions in Urban Benin.
Berkeley and Los Angeles: University of California Press.
Shore, Bradd
1996 Culture in Mind: Cognition, Culture, and the Problem of Meaning. New York:
Oxford University Press.
Simon, Herbert A.
1957 Models of Man: Social and Rational. New York: John Wiley.
Sperber, Dan
1982 Apparently Irrational Beliefs. In Rationality and Relativism. M. Hollis and S.
Lukes, eds. Pp. 149-180. Oxford: Basil Blackwell.
1985 On Anthropological Knowledge. Cambridge: Maison des Sciences de l'Homme
and Cambridge University Press.
Spiro, Melford E.
1987 Collective Representations and Mental Representations in Religious Symbol Sys-
tems. In Culture and Human Nature: Theoretical Papers of Melford E. Spiro. B. Kil-
borne and L. L. Langness, eds. Pp. 161-184. Chicago: University of Chicago Press.
Stein, Edward
1996 Without Good Reason: The Rationality Debate in Philosophy and Cognitive Sci-
ence. Oxford: Clarendon Press.
Stoner, Bradley P.
1985 Formal Modeling of Health Care Decisions: Some Applications and Limitations.
Medical Anthropology Quarterly 16:41-46.
Strauss, Claudia
1992 Models and Motives. In Human Motives and Cultural Models. R. D'Andrade and
C. Strauss, eds. Pp. 1-20. Cambridge: Cambridge University Press.
340 MEDICAL ANTHROPOLOGY QUARTERLY
Tversky, Amos
1972 Elimination by Aspects: A Theory of Choice. Psychological Review 79:281-299.
Tversky, Amos, and Daniel Kahneman
1974 Judgment under Uncertainty: Heuristics and Biases. Science 185:1124-1131.
1981 The Framing of Decisions and the Psychology of Choice. Science 211.453-458.
Weller, Susan C , Trenton R. Ruebush II, and Robert E. Klein
1997 Predicting Treatment-Seeking in Guatemala: A Comparison of Health Service
Research and Decision-Theoretic Approaches. Medical Anthropological Quarterly
11:224^245.
Wikan, Unni
1989 Managing the Heart to Brighten Face and Soul: Emotions in Balinese Morality
and Health Care. American Ethnologist 16:294-312.
Young, Allan
1980 The Discourse on Stress and the Reproduction of Conventional Knowledge. So-
cial Science and Medicine 14B: 133-146.
1981 When Rational Men Fall Sick: An Inquiry into Some Assumptions Made by
Medical Anthropologists. Culture, Medicine and Psychiatry 5:317-335.
1982 The Anthropologies of Illness and Sickness. Annual Review of Anthropology
11:257-285.
Young, James C , and Linda C. Garro
1982 Variation in the Choice of Treatment in Two Mexican Communities. Social Sci-
ence and Medicine 16:1453-1465.
1994[1981] Medical Choice in a Mexican Village. New Brunswick, NJ: Rutgers Uni-
versity Press.