26 BUREAU OF AMERICAN ETHNOLOGY [Bull. 175 



level of latent meaning, these seemingly irreconcilable explanations of 

 a given disease actually have the same meaning (pt. 1, p. 11 ; pt. 4, pp. 

 117-128). 



Factors facilitating diagnosis. — Although several aboriginal med- 

 ical specialties are today entirely obsolete and forgotten, it is interest- 

 ing to note that one rarely hears of a patient who had a disease that 

 no one was any longer qualified to cure, perhaps because shamans only 

 diagnose illnesses that one or the other of them has the power to treat 

 effectively. 



Diagnostic criteria. — The single most important diagnostic pro- 

 cedure is the study of the patient's dreams, which supposedly reveal 

 the etiology and nature of his condition. This is a far from senseless 

 procedure, since already Aristotle (n. d.) knew that in sleep one may 

 be more sensitive to inner stimuli — indicative of the onset of a physical 

 illness — than one is in a waking state (cf. Bartemeier, 1950). It is 

 psychologically legitimate to suggest that, given the general knowl- 

 edge of the type of dreams one is supposed to have if one has, e. g., 

 hiwey lak, the patient's unconscious may comply with these cultural 

 expectations and produce a suitable type of dream.^ 



In other instances it is the physical symptom that provides clues 

 to the nature of the illness and causes the shaman to ask whether the 

 patient had the kind of dreams that should, in principle, accompany 

 his physical symptom. Given the fact that, even in the absence of 

 leading questions, all dreams are modified — by means of a secondary 

 elaboration — in the act of remembering and retelling them, it is ex- 

 tremely probable that a patient, who is being insistently questioned 

 by a shaman, will "recall" having had the expected type of dreams. 

 Such retroactive falsifications are presumably especially common in 

 cases where the diagnostician suspects witchcraft and therefore insists 

 that the patient cannot be helped unless and until he reveals the name 

 of the person who bewitched him. This supposition is greatly 

 strengthened by the fact that retroactive falsifications of dreams occur 

 both when one seeks to deceive oneself (Freud, 1950) and when one 

 must comply with social expectations (Devereux, 1951 a). 



In brief, given the tendency of primitives to differentiate less 

 sharply than we do between reality and dream (Kroeber, 1952), and 

 given also their suggestibility (Stoll, 1904), it is extremely probable 

 that, when pressured by an inquisitive diagnostician, the primitive 

 patient will "recall" having had the kind of dreams he is supposed 

 to have had, or else will produce the desired dream the very next 

 night.^" In this context it is noteworthy that, according to Toffelmier 

 and Luomala (1936), the Diegueno boy, whose dream does not fit the 



• The cultural patterning of Plains Indian visions tends to confirm this hypothesis 

 (Devereux, 1951 a). 



'" A Plains Indian patient was systematically Induced by me to "dream out" his conflicts, 

 In order to achieve a cure (Devereux, 1951 a). 



