IEEE Technology and Society Magazine - Spring 2013 - 13

it extensively. personality profiling that emphasized
antisocial traits also fell into the trap of treating this
as typically male. A field that had initially appeared
quite gender-neutral rapidly became less so.
"Tower of Babel," the chapter on the role of programming languages in the late 1950s and the 1960s,
presents them as means for achieving control of the
programming activity as much as means for controlling computers."Automatic programming" was a term
used to characterize programming languages - not
that programming became automated, but that by
contrast with low-level machine code, higher-level
languages were supposed to let computer users rather
than the "computer boys" write programs. Scientists
could write algebraic expressions in Fortran; business
people could write in Cobol, and "Susie Meyer" in an
IBM advertisement for the pL/I language "could find
happiness handling both commercial and scientific
applications" with it, despite having no programming
experience - and if she could, so could anyone.
programming languages could thus be seen as
deskilling, enabling managers to replace computer boys
with cheaper and less powerful labor, whether male or
female. A language like Cobol, with its English-like
notation, could let managers understand programs.
Of course, these languages were also expected to help
skilled programmers to resolve the "software crisis" -
the widespread difficulty in producing system software
on time or to function properly. It is not surprising that
for all their merits, programming languages fell short
of fulfilling these (inconsistent) expectations.
One of the strengths of Ensmenger's book is its
showing how what we may take as "obvious," such as
the basic character of computer science as an academic
discipline, emerges from multiple claims to define the
field. Indeed, it was not obvious at the start that there
should be such a field - new technologies often don't
give rise to new disciplines. Academics who worked in
computing wanted an intellectual foundation for programming, employers wanted educational standards,
and computing workers wanted their activity to have
professional standards. The major academic computing
organization, the Association for Computing Machinery (ACM), also had non-academic members, but when
it developed the first computer science curriculum standard, it was heavily theoretical and mathematical, and
in the view of many in computing occupations, gave too
little attention to practical data processing.
The book is not primarily about academic computer science, though in its discussion of the ways in
which women became excluded from the programming profession, it might have added something on
university departments. They largely grew out of electrical engineering and mathematics departments, both
fields that were heavily male in the 1960s, and so the
composition of computer science departments shared
IEEE TECHNOLOGY AND SOCIETY MAGAZINE

SprING 2013

that imbalance. The computer science network was an
old-boys network from the start.
An early ACM definition of computer science
was that it was the study of information. This definition could unite several fields, and it gave rise to the
term "informatics" that several European languages
use instead of "computer science" Ensmenger uses
Thomas Kuhn's notion of "normal science" to argue
that computer science became established as the science of algorithms when it acquired its major textbook
in Donald Knuth's Art of Computer Programming,
which made algorithms central, traced their mathematical lineage, and helped to set the agenda for further work to define the profession.(My first reaction,
as an academic computer scientist, was "of course
that's what it's about!," but Ensmenger rightly points
out that this was a matter of construction.)
The following chapters treat the professionalization
of programming. During the 1960s the cost of computing rose, particularly the cost of software. Although the
term "software crisis" that came from the 1968 NATO
conference on software engineering originally referred
to system software, from the standpoint of corporate
profit, the complexity of application software presented
similar problems. Furthermore, the computer boys continued to appear hard to manage - unkempt, unruly,
and unsocial. They seemed to require autonomy in
order to deal with the computer as only they knew how,
while managers sought to define their role more narrowly as technicians.
During the 1960s two primary organizations
competed to define the programming profession, the
Data processing Management Association (DpMA),
and the ACM. The former offered its Certified Data
processor (CDp) exam, while the latter had its computer science curricular guidelines, as well as regular
conferences of academics and others. Although the
DpMA was focused on professional programmers,
interest in its CDp program declined after a few
years of growth. It was not clear that its certification met industrial needs, and as with earlier aptitude
tests, there were charges of fraud. The ACM, while
narrower, came to play a larger role in defining the
computing field through standard degree programs.
Its definition, however, did not directly address the
professional status of programming.
Software engineering is concerned with managing complexity in programs of thousands or millions
of lines of code. Ensmenger discusses several standard approaches, focusing on the multiple senses of
"manage."There were attempts to make software development an industrial discipline, modeled on factory
mass production. There was the "chief programmer
team" (CpT), akin to a surgical team with a master programmer and skilled assistants who divided the work.
This design could conflict with corporate managerial
|

Table of Contents for the Digital Edition of IEEE Technology and Society Magazine - Spring 2013