254 The students who were observed to avoid oral English communication were not the students who were making the least progress in written expression. The low group in written expression was composed of males and females. Most of this group scored low in both English and Spanish written measures. The lowest group of oral English achievers was composed of all females. These females appeared to fit Smith's (1980) description of the "marginal man," the immigrant on the edge of the new society. Smith found that Portuguese females were conditioned to maintain traditional social networks that required them to appear less acculturated and less fluent in English. Female survival skills are deeply related to maintenance of both cultures, the new and the old. To become too integrated into the new is to lose contact with the old. The same culturally regulated sexual roles was noted by Garcia (1980), who observed that in spite of 15 years of revolutionary teaching, Cuban male attitudes toward women have not changed. Garcia's ob servations are confirmed by Montaner (1981) who points out that although there has been a great deal of talk about the liberated female in Cuba, few females occupy positions of importance within the revolu tionary government. Male behavior in Spanish was not observed during physical edu cation primarily because of the difficulty in accompanying males during this class. The fact that rank on Spanish was significantly correlated with participation in physical education appears to be related to cafeteria behavior discussed in Hypothesis Two. Males also appear to be carrying on cultural roles. Those who are more proficient in Spanish are also those students who are most active in physical