Fears, Phobias, and Preparedness: Toward an herbrawt/classes/390/Ohman.pdfFears, Phobias, and...

download Fears, Phobias, and Preparedness: Toward an herbrawt/classes/390/Ohman.pdfFears, Phobias, and Preparedness:

of 40

  • date post

    12-Jul-2018
  • Category

    Documents

  • view

    215
  • download

    1

Embed Size (px)

Transcript of Fears, Phobias, and Preparedness: Toward an herbrawt/classes/390/Ohman.pdfFears, Phobias, and...

  • Psychological Review2001, Vol. 108, No. 3, 483-522

    Copyright 2001 by the American Psychological Association, Inc.0033-295X/01/S5.00 DOI: 10.1037//0033-295X.108.3.483

    Fears, Phobias, and Preparedness: Toward an Evolved Module ofFear and Fear Learning

    Arne OhmanKarolinska Institute!

    Susan MinekaNorthwestern University

    An evolved module for fear elicitation and fear learning with 4 characteristics is proposed, (a) The fearmodule is preferentially activated in aversive contexts by stimuli that are fear relevant in an evolutionaryperspective, (b) Its activation to such stimuli is automatic, (c) It is relatively impenetrable to cognitivecontrol, (d) It originates in a dedicated neural circuitry, centered on the amygdala. Evidence supportingthese propositions is reviewed from conditioning studies, both in humans and in monkeys; illusorycorrelation studies; studies using unreportable stimuli; and studies from animal neuroscience. The fearmodule is assumed to mediate an emotional level of fear learning that is relatively independent anddissociable from cognitive learning of stimulus relationships.

    Mammalian evolution has required the successful developmentof defense systems to cope with dangers that threatened to disruptthe transport of genes between generations. In the early mamma-lian environment of evolutionary adaptiveness (Tooby & Cos-mides, 1990), disaster could strike fast and without warning,primarily through hunting predators but also through aggressiveconspecifics and from physical events such as falling objects,floods, thunder and lightning, and sudden lack of oxygen. Escapeand avoidance were common strategies designed by evolution todeal with such exigencies. At a minimum, they required a percep-tual system to identify threats and a reflexively wired motorsystem to move the organism away from the danger. With moresophisticated nervous systems, the effectiveness could be ex-panded both at the sensory and the motor ends, and the relationbetween stimulus and response could be rendered less stereotypedby inserting a central motive state between the two. In this way, themotive state could be activated from innate danger stimuli andserve to promote escape through the flexible tailoring of responsesto environmental contingencies (e.g., Archer, 1979). For example,depending on circumstances, the animal would freeze, escape, or

    A portion of this work was completed while Susan Mineka was a Fellowat the Center for Advanced Study in the Behavioral Sciences, Stanford,California, during 1997 to 1998. The research cited as well as preparationof this article was made possible by grants from the Swedish Council forResearch in the Humanities and Social Sciences, the Bank of SwedenTercentennial Foundation, the National Science Foundation (BNS-8507340 and BNS-8216141), the University of Wisconsin GraduateSchool, and the John D. and Catherine T. MacArthur Foundation (95-32005-0).

    We thank Mark E. Bouton, Peter Lovibond, and Stefan Wiens for theircomments on drafts of this article.

    Correspondence concerning this article should be addressed to ArneOhman, Department of Clinical Neuroscience, Psychology Section, Karo-linska Institutet, Karolinska Hospital, Z6, S-171 76 Stockholm, Sweden, orto Susan Mineka, Department of Psychology, Northwestern University,Evanston, Illinois 60208. Electronic mail may be sent to arne.ohmancns.ki.se or mineka@northwestem.edu.

    attack (e.g., Blanchard & Blanchard, 1988). It is this central motivestate that we commonly identify as fear (e.g., Mineka, 1979;Ohman, 1993a). An essential characteristic of fear, therefore, isthat it motivates avoidance and escape (Epstein, 1972).

    Potentially disastrous events sometimes do not strike withoutnotice, however, but may be heralded by subtle cues. For example,to the attentive observer, a predator may announce its presence byfaint sounds or odors. By using the contingency between such cuesand the potentially deadly consequence, the central motive state offear could be conditioned to the cue (e.g., Rescorla & Solomon,1967), which would promote further flexibility in the relationshipsbetween stimulus and response. Furthermore, conditioned fearcues could recruit defensive responses in anticipation of the pred-ator's strike, which provides a decisive edge in the arms racebetween predator and prey (see Hollis, 1982). From this perspec-tive, it is likely that survival-relevant relationships between cuesand consequences could be used by natural Selection to promotetheir preferential and selective association in the brains of animals(e.g., Bolles, 1970; Seligman, 1970). The emergence of moreadvanced nervous systems assured that the outcome that evolutionselected for, avoidance of potentially deadly events or situations,could be achieved through more sophisticated and selective mech-anisms, such as inborn defense responses, Pavlovian conditioning,instrumental learning, and eventually cognition and consciousdeliberation (e.g., Razran, 1971).

    Viewed from the evolutionary perspective, fear is central tomammalian evolution. As a product of natural selection, it isshaped and constrained by evolutionary contingencies. It is acentral thesis of this article that this evolutionary history is obviousin the fear and phobias exhibited and readily learned by humans.We are more likely to fear events and situations that providedthreats to the survival of our ancestors, such as potentially deadlypredators, heights, and wide open spaces, than to fear the mostfrequently encountered potentially deadly objects in our contem-porary environment, such as weapons or motorcycles (e.g., Marks,1969; Seligman, 1971).

    483

  • 484 OHMAN AND MINEKA

    Purpose and Overview

    The purpose of this article is to develop a concept of an evolu-tionarily evolved fear module (Tooby & Cosmides, 1992) thathelps to organize and explain important aspects of human fear andfear learning. The article is organized into eight parts. The first onedevelops the concept of the fear module. Briefly, a fear module isproposed to be a relatively independent behavioral, mental, andneural system that is specifically tailored to help solve adaptiveproblems prompted by potentially life-threatening situations in theecology of our distant forefathers. The second and third partsdiscuss a central feature of the fear module: its relative selectivityregarding the input to which it responds and preferably enters intoassociations. In the fourth part, we evaluate two alternatives, oneevolutionary but nonassociative and one quasi-associative but non-evolutionary, to evolutionarily shaped selective associations as abasis for fear learning. In the fifth and sixth parts, we analyze therole of cognition in fear learning, particularly with regard to theautomaticity with which particular stimuli activate the fear moduleand with regard to the encapsulation of the module from cognitiveinput. In the seventh part, we review research that delineates theneural basis for the fear module. In the eighth part, finally, wedevelop a levels-of-learning concept to reconcile data from humanconditioning with the database from animal research on fear and itsneural mechanisms.

    An Evolved Fear Module

    The Status of Evolutionary Explanations

    From the evolutionary perspective, behavior is more likely to beorganized into relatively independent modules than in terms ofgeneral-purpose mechanisms (Tooby & Cosmides, 1992). Just asour body is composed of a number of independent organs servingsurvival and procreation, behavioral and mental systems can bethought of as composed of organs or independent modules (e.g.,Fodor, 1983). As a result of natural selection, such modules weretailored to solve specific adaptive problems that were recurrentlyencountered in the environment of evolutionary adaptiveness(Tooby & Cosmide's, 1990).

    In our view, evolutionarily based explanations of psychologicalphenomena have no special status that set them apart from othertypes of explanations. Although evolutionary hypotheses are notamenable to direct empirical tests in psychological research, basicpostulates that are testable only indirectly, in terms of their con-sequences, are commonplace in science. To take a classic example,Newton's concept of attraction perhaps did not appear to makesense to his contemporaries, but it did make sense of the data onthe paths of planets when used in his axiomatic system. In psy-chology, however, the widespread commitment to "the standardsocial science model" (Tooby & Cosmides, 1992, pp. 24-34), withits emphasis on learned behavior, collectively known as culture,has planted skepticism among many regarding the role of phylo-genetic influences on human behavior (see, e.g., Delprato, 1980,with regard to fear). Partly, the skepticism is nourished by popularmisuse of evolutionary arguments, in which virtually any psycho-logical phenomenon can be declared post hoc to represent evolu-tionarily shaped adaptations. But if hedged by some reasonablecaveats, we think that evolutionarily based theorizing provides afruitful avenue for analyzing psychological phenomena.

    A putative evolutionary explanation must, just as any scientificexplanation, be open to empirical tests; to be more precise, it mustbe integrated into a conceptual network with testable conse-quences. Here, the distinction between theoretical and metatheo-retical statements may be helpful (Johnston & Turvey, 1980). Thepurpose of a scientific theory is to explain a set of empiricalphenomena, whereas a "metatheory is concerned with justifyingthe asking of certain kinds of questions in a particular area ofinquiry" (Johnston & Turvey, 1980, p. 149). For example, thequestions about fear that we pose in this article are different fromthose that would be posed by someone inspired by, for example, asocial constructivist or psychodynamic metatheory of