Major directions of our past research.

Identifying elements of a mate recognition system. In 1977, we identified two individual behaviors that appeared to be critical to cowbirds' mate recognition system, male song production and female copulatory responsiveness to song. When coupled together, these two elements seemed to constitute a "safety net" for species and mate recognition appropriate for a brood parasite in that naïve female cowbirds responded to naive male song. We then found that development of male song could be influenced by auditory experience with adult males, as has been found in all songbirds studied to date. But, of more interest we also found male song development was affected by female social stimulation. These findings showed that traits used by females to assess males during courtship could be influenced by females prior to courtship.  These data, in fact, suggested a role for non-imitative social shaping that was critical to the development of song and to mate recognition.

Males use females' visual responses to shape song. To expand our understanding of social interactions between males and females in early spring well before copulatory responsiveness occurs, we videotaped pairs and trios of males and females. A striking response to us was female wing movement while songs were in progress. We also found that wing stroking and other forms of female social input (gaping and aggression toward another female) affected song content and song development from as early as 120 days of age .  Thus, females respond non-vocally to the variable vocal precursors of mature song, similar to human adults responding to infant babbling. For their part, the young males "read" these visual signals and modify their behavior. Reading a visual system does not involve imitation. Most importantly, we have recently documented the influence of such proximate social experience on developmental rate of song ontogeny and song structure, suggesting the presence of heterochronic mechanisms, an important link to evolutionary and speciation processes.

Innate is not enough: Social experience defines competence. Subsequent research showed that phenotypic plasticity in courtship skills was such that captive male cowbirds could incorporate and retain heterospecific song elements and could direct songs to other species. Wild-caught juvenile males that we housed with canaries after fledging courted canaries when female cowbirds were available. We also studied juvenile males housed only with pairs of female cowbirds.  These males showed no signs of knowing when or how to approach or sing to female cowbirds when in a flock in an aviary setting. Nor did they show a defining feature of male-male interactions: countersinging (CS). In CS, two (or more) males face one another at usually less than one foot and alternate songs between the singers or sometimes sing at the same time. They may exchange 2-20 songs.  All of the above males had been captured as juveniles after fledging and had had early experience with their own species. Thus, these data served as an important incentive to understand how social experiences shape developmental systems .

Limitations of female solicitation assay.  These findings of learned social incompetence led us to recognize the limitations of the female solicitation assay, a technique used for many avian species as a perceptual test of song preference.  As we looked at male cowbirds in aviaries that did not seem to know when or how to sing to females, we realized that we, as the playback engineers, were supplying that knowledge to the females by picking the song and the time to play it. But if such song use represented a female-assessed competency of the male, the playback bioassay was not capturing it. Thus, we have turned to new assays of female preferences including social and copulatory responsiveness and fecundity in aviary settings to complement information about song quality obtained by the copulatory bioassay.  

Recognition of incompetent song use, as opposed to adequate song structure, led us to consider more carefully the different parameters of song that were being studied such as song content, song production, a song's social orientation, and a song's use with other vocal signals such as the cowbird's flight whistle. For example, more dominant males both sing and whistle before and during copulation. Thus, the original element, song structure, is part of a larger family of variables encompassing learning and performance.

Female responsiveness to song has many facets, including malleability. In work using captive flocks in our lab, Freeberg found that female mate preferences, and correlated choices of songs, were malleable. In 20+ years, we had little evidence of female malleability in song preferences. The new data suggested to us that the multi-modal nature of the female's flock environment facilitated new behavioral states, an analogous finding to recent work in filial imprinting.

          We have found that female responsiveness to song is a more complex variable than we previously thought: some females are more or less choosy, some are more or less responsive (how often and how long they solicit) and some are more or less modifiable (at least, at the level of juvenile or adult). Thus, as with the male's vocal signal, female sexual response is also a family of variables, some of which show plasticity. Taken as a whole, these data indicate co-variance on the suite of variables comprising the cowbird's mate recognition system.

Socially compounded variables form a complex system. The assumption of male-female coupling of behavioral elements is a tenable idea, but may arise from very different mechanisms than we originally thought. To make an analogy, we had first identified individual behavioral elements, as if they were chemical elements, whose isolated properties could then be assayed (atomic weight, etc). But, when we combined elements in even simple ways, we found that the stability of the elements' properties were not fixed. Thus, female responsiveness to song is in part a function of what she hears (a "best of n" process).  Male responsiveness to females is also a product of prior and present social interactions that renders active or relatively inert other social and vocal variables (e.g., lack of dominance or lack of song effort affects song effectiveness as a courtship signal). 

          Our self-analysis led us to use more complex settings to allow more of the behavioral elements and their possible interactions to enter into the social equation. In addition to the original behavioral elements of song and female sexual response, we have also begun to focus on socially compounded traits, such as directed song or female approach to males, i.e., behaviors combining several elements across organisms.   A directed song is a socially compounded behavior; both producer and recipient must cooperate. So too, female approach requires a male to remain in place. We believe such dyadic behaviors are closer to the nature of the traits assessed when mate choice occurs, and subsequent sexual selection operates. We are investigating this idea more thoroughly in our current work by looking at the role of social environments in shaping variations in such socially compounded behaviors.

    And so to return to our original focus on individual behaviors, the most conspicuous marker of avian communication to humans may be the song, but the circumstances and behavioral parameters that lead a male to produce it at the right time to the right receiver rest on critical postnatal variables that are functionally as important as song structure itself.  Moreover, for the female, the song may be also be a conspicuous marker, but it may also be historical evidence of investment in months of interactions in which females manipulate many males' songs by affecting acoustic content. Thus, selection is working on individual ontogenies, of which the original elements, song and female response, are simply the most obvious.