Face the Beast and Fear the Face: Animal and Social Fears as Prototypes for Evolutionary Analyses of Emotion


  • Presidential Address presented at the 25th Annual Meeting of the Society for Psychophysiological Research. Houston, TX, October 17–20, 1985. The ideas and work presented in the paper have developed over the years in continuous interactions with a number of friends, colleagues, and students. I would especially like to mention Gunilla Bohlin, Ed Cook, Ulf Dimberg, Francisco Esteves, Mats Fredrikson, Bob Hodes, Kenneth Hugdahl, Staffan Hygge, Peter Lang, Isaac Marks, Susan Mineka, Jack Rachman, and Lars-Göran Öst. Needless to say, the rather wild juxtaposition of ideas and data presented here reflects more my own lack of scientific responsibility than theirs.

  • The research has been supported by a long series of grants from the Swedish Council for Research in Humanities and Social Science.

Address requests for reprints to: Arne Öhman, Department of Clinical Psychology, University of Uppsala, P.O. Box 1225, S-751 42 Uppsala, Sweden.


This paper applies a functional-evolutionary perspective to fear in the context of encounters with animals and threatening humans. It is argued that animal fear originates in a predatory defense system whose function is to allow animals to avoid and escape predators. Animal stimuli are postulated to be differentially prepared to become learned elicitors of fear within this system. Social fears are viewed as originating in a dominance/submissiveness system. The function of submissiveness is to avert attacks from dominating conspecifics. Signs of dominance paired with aversive outcomes provide for learning fear to specific individuals. Data which in general are interpreted as supportive of this conceptualization are reviewed. To explain the mechanism behind the causal relationships suggested in the evolutionary analysis, an information-processing model is presented and empirically tested. It is argued that responses to evolutionary fear-relevant stimuli can elicit the physiological concomitants of fear after only a very quick, “unconsciousness,’ or preattentive stimulus analysis. Support for this notion is presented from backward masking studies where it is demonstrated that conditioned autonomic responses to fear-relevant stimuli can be elicited even with masked stimuli.