A Commentary on David A. Hounshell’s Commentary on “Software as Labor Process”

Abstract

We particularly value David Hounshell's commentary for the way in which it situates our work in relation to both the old "new" and new "new" labor historiog-raphy. We endorse his suggestion to draw a connection between the recent profes-sionalization efforts of "software engineers" and the larger histories of the established engineering disciplines. Hounshell challenges us to engage further with the professionalization literature. In retrospect we could have been more explicit, but it had been our intention to pose provocative questions rather than provide definitive analysis. We expect that our readers will have no difficulty recognizing our description of the historical controversy over "what is programming?" and "what makes for an exceptional programmer?" as attempts to establish autonomous control over what sociologists of professions refer to as "socially valuable body of esoteric knowledge." The psychological profiles and programming aptitude tests of the early 1950s, the certification debates of the mid-1960s, and the persistent tension between the "art" of programming and the "science" of software engineering are expressions of an ongoing struggle by computer programmers to establish professional and intellectuallegitimacy. On several occasions Hounshell reminds us of the severe need for further empirical research on software labor markets past and present. On this point we agree with hirn wholeheartedly. The recent debates over the alleged shortage of skilled American information technology workers have exposed how remarkably little we know even about contemporary software workers, much less the laborers of previous decades. A few of Hounshell's criticisms seem misplaced, however-as though we were actually writing the labor history of software, rather than setting a research agenda for writing this history. But even in our agenda-setting task, we were hampered by the lack of secondary literature and accurate empirical data. For example, Hounshell suggests that "dualladder" programs provided substantial career paths for programming professionals. We did not find evidence to suggest that such programs were either widespread or effective, but this is indeed an open and interesting empirical question. Our brief sketch of this history attempts to make creative use of some well-worn traditional sources, but we agree that there is a great deal of research, both quantitative and otherwise, that needs to be accom-U. Hashagen et al. (eds.), History of Computing: Software Issues