|
08.10
Up for the Count
By Michael N. Geselowitz, Ph.D., IEEE History Center
In August 2010, we are reaching
the end of the latest United States census
process. Every 10 years, the United States — as
mandated by the constitution — undertakes to
enumerate its population. The current census
will be controversial because of its political
importance (it is used to determine
congressional representation) and because the
size and complexity of the U.S. population
requires statistical techniques that could
potentially be manipulated. One thing,
however, is certain: the U.S. Census Bureau will use the latest
technology to collect and process the data, as
they have in the past.
For the
1950 census, the Bureau supported the
development of UNIVAC, and accepted delivery of
the UNIVAC I on 31 March 1951. It was formally
dedicated and put into service on 14 June of
that year, and many historians of technology
consider that the birth of non-military
computing. However, perhaps the most important
contribution of the technologists who worked for
or were funded by Bureau to enhancing technology
for humanity, occurred exactly 120
years ago this month.
By 1870, the U.S. population had
grown so large that clerks were not able to
process the data by hand. The Bureau began to
experiment with a number of mechanical devices
to automate the process. For the 1880 census,
an engineer named Henry Hollerith had served on
the technical team and felt that he could
improve the process. In 1884, he filed a patent
for an electromechanical device that could
rapidly read information that had been encoded
by punching holes on a paper tape or set of
cards. This and two subsequent patents for
producing and sorting the cards all issued in
1889. Having incorporated in 1887, Hollerith’s
Tabulating Machine Company was chosen through a
competition in 1889 to process the census of
1890. The project was a huge success and, on 16
August 1890, the total population of the United
States was announced to be 62,622,250 (the
number in 2000 was 281,421,906, and we are of
course waiting for the 2010 results).

Hollerith’s
machine from his 1887 patent.
Photo: Courtesy of IEEE History
Center
However, the monopoly of
Hollerith on this new sort of data processing
enabled him to charge a premium for the use of
his equipment. This situation led the Census
Bureau to develop its own punch-card equipment
(based on Hollerith’s ideas) and to a number of
patent battles which Hollerith ultimately lost,
leading in turn to the founding of more
competitors. In response, in 1911 Hollerith
merged his company with several others into a
firm that eventually became International
Business Machines — IBM! IBM continued to lead
the market, experimenting with various
technologies to improve card processing. Its
forays into electronic counting led it to enter
and, for a long time, dominate the new and growing
electronic computer market. For many years,
punched cards were the main method of computer
input.
The impact of punched card
equipment on data processing, as well as its role
in the subsequent development of electronic
computing, have led IEEE to designate this
technology in its STARS program (Significant
Technological Achievement Recognition Selections
— an online compendium of invited, peer-reviewed
articles on the history of major developments in
electrical and computer science and
technology). Readers are invited to view the
full punched card equipment article on the IEEE
Global History Network at
http://www.ieeeghn.org/wiki/index.php/STARS:Punched_Card_Equipment.
So, happy anniversary (one
anniversary among many possible) to the
information age!

Michael N. Geselowitz, Ph.D.,
is staff director at the IEEE History Center at
Rutgers University in New Brunswick, N.J. Visit
the IEEE History Center's Web page at:
www.ieee.org/organizations/history_center.
Comments may be submitted to
todaysengineer@ieee.org.
|