way the song of a live tutor is likely to evoke orienting responses more often, and the probability of an association between song stimuli present at the same time will be enhanced: the outcome will be an engram for the song elements present during the orientation. Live tutoring is more effective than tape tutoring, and live tutoring is effective whether or not, for example, the tutor is the very small strawberry finch, or a relatively large white-crowned sparrow singing strawberry finch song. Thus, the effective stimulus event seems not to be some specific visual quality of the white-crowned sparrow.
Tests of predictions based on this model are now underway. These tests involve analyses of the behavior of students and tutors to determine the amount of orientation and rate of habituation to songs of different types that are presented in different ways. Such experiments provide a severe test of the proposed behavioral mechanisms involved in song learning. Even if the proposed mechanisms turn out to be incorrect, analysis of the behavior of students and tutors should provide an approach to understanding the large individual variability found in tutoring studies. Although suggestions can often be made of the kinds of factors producing individual differences, it usually is a mystery why different birds treated the same way differ so markedly. Kroodsma and Pickert ( 1984a) suggested that "This variation must represent real individual differences in how the young males responded to the social situation we provided in the laboratory" (p. 393). They concluded that an understanding of the causes of these individual differences will contribute to an understanding of their importance in determining the fitness and general behavior of individuals.
Although the evidence and critical factors spelled out for the proposed model are behavioral, I have considered the possible physiological events that might be involved. Only by framing behavioral theory in terms of probable physiological mechanisms will we be able to generate unified theories that adequately traverse both the physiological and behavioral landscape. In addition, careful attention to the processes at one level can produce insights into the workings at the other, to the enrichment of each of the several disciplines involved.
The original research discussed here was supported by grants from the University of California, the National Institutes of Health (HD-04343 and MH-38782), and the National Science Foundation (BNS-7914126 and BNS-8004540).
Alcock J. ( 1984). Animal behavior: An evolutionary approach, ( 3rd ed.). Sunderland, MA: Sinauer.
Baker M. C., Bottler S. W., & Arnold A. P. ( 1984). "Sexual dimorphism and lack ofseasonal changes in vocal control regions of the white-crowned sparrow brain"