James W. Perry

(October 21, 1907 - December 24, 1971)

Originally trained as a chemist, James Whitney Perry went on to become one of the chief proponents of using computers for indexing, classifying, and searching documents. A man of many interests, he was able to bring together talents from diverse fields and incorporate them for the advancement of information science.

Education

Perry received his Bachelor of Science degree for chemistry in 1927 as well as a Master of Science degree for the same subject in 1928 from North Carolina State College. Additionally, Perry earned a Scientiæ Magister (Master of Science) degree in 1931 from MIT. He also spent some time studying at Technische Hochschule of Stuttgart and Berlin-Charlottenburg from 1932 to 1933 (Skolnik, 1972).

Professional Career

In 1945, after World War II, Perry returned to MIT in order to develop his ideas on improving library literature searching and methodology. His experience dealing with the literature in his field of chemistry demonstrated to him that the efficiency of the library and document retrieval needed to be brought up to speed (Perry, 1977). In the late 1940s he was charged with developing punch card systems of organizing chemical information (Williams & Bowden, 1999). While exploring these information issues in the sciences, Perry became interested in difficulties of information retrieval in many other disciplines, including law, medicine, and metallurgy (Perry, 1977).

Upon joining Western Reserve University in 1955, Perry began work with computers designed to search abstracts using codes that stood for specific natural language terms (Weinberg, 1996). During his time at Western Reserve University, Perry served as the president of the American Documentation Institute (later the American Society for Information Science) for 1957 (Skolnik, 1972). Some of Perry’s most cited work is from this period. His colleague Madeline M. (Berry) Henderson (1999) writes:

"We were so prolific that we inspired Si Newman to compose a limerick. (Simon Newman was a chemist at the U.S. Patent Office, active in early efforts to automate patent searching procedures.) Si wrote:

        Perry, Berry, and Kent

        Re-announce the self-same event.

        Their abstracts in miniature

        Cover the world’s literature

        Recently doubled by Perry, Berry, and Kent!" (p. 268).

Perry moved on to his last academic post in 1960 at the University of Arizona, Tuscon. There, his work consisted of “design of artificial languages for computers and the translation and processing of natural languages” (Perry, 1977, p. 68) along with “pioneering work on communication systems design and theory” (Skolnik, 1972, p. 69).

Contributions to Reference Librarianship

While at the University of Arizona, James W. Perry made his major contribution to reference librarianship probably without intending to do so. Although he contributed much to the field of information retrieval in terms of computer systems, in the 1960s reference librarianship was not a technologically oriented vocation. Perry’s contribution came about from his idea, which received its final published form in 1963, that “queries [...] may be considered at different stages of formulation or expression as follows:

        Q₀ = some ideal “best” query to obtain information to deal with a given problem or situation

        Q₁ = the mental conception developed by some person as to needed information

        Q₃ = the statement of a query by a person without regard to a given IR [information retrieval] system

        Q₄ = the statement of a query by a person with regard to a given IR system” (Perry, 1963, pp. 21-22).

This idea was significant to reference librarianship when Robert S. Taylor came across it and applied it to “the inquirer’s formation of a question” (Richardson, 2002, p. 189). Taylor also lifts his idea of five filters that an inquirer must use when posing a question from this paper by Perry (Mackavoy, 2006). Perry had written that in order to get the best results of an IR system, an inquirer’s “queries will be influenced by the following factors:

        1) The assignment, i.e., the problem or situation with which he is called upon to deal.

        2) The extent of his knowledge that bears on the assignment.

        3) His estimate of what additional information he needs.

        4) His personal preferences as to sources of information, e.g., preference for laboratory experimentation over library searching.

        5) His expectations as to what kind of query can be made to evoke a useful response from available IR systems without incurring undue expenditure of time, effort or money” (Perry, 1963, p. 24).

Other Areas of Interest

Perry’s early career consisted of work as a chemist. He was involved with the development of the first low-cost synthetic detergent. During World War II, Perry worked with composite rocket propellants and high explosives. Also during this time he studied and published numerous papers on scientific Russian (Perry, 1977).

References

Henderson, M. M. (1999). Prolific abstractors [Electronic version]. Proceedings of the 1998 Conference on the History and Heritage of Science Information Systems, 268.

Mackavoy, M. (2006, November 21). DIS 245 "Info Access" encyclopedia | Robert S Taylor. Retrieved November 21, 2006, from http://ucla245.pbwiki.com/Robert%20S%20Taylor

Perry, J. W. (1963). Defining the query spectrum – the basis for developing and evaluating information-retrieval methods. IEEE Transactions on Engineering Writing and Speech, 6(1), 20-27.

Perry, R. (1977). Perry, James Whitney. In Encyclopedia of library and information science. (Vol. 22, pp. 66-68). New York: Marcel Dekker, Inc.

Richardson, J. V. (2002). Research on reference transactions. In F. C. Lynden (Ed.), Advances in librarianship. (Vol. 26., p. 189). San Diego: Academic Press.

Skolnik, H. (Ed.). (1972). James Whitney Perry. Journal of Chemical Documentation, 12(1), 69.

Weinberg, B. H. (1996). Complexity in indexing systems -- abandonment and failure: implications for organizing the Internet. Paper presented at the 1996 ASIS Annual Conference. Retrieved November 21, 2006, from http://www.asis.org/annual-96/ElectronicProceedings/weinberg.html

Williams, R. V. & Bowden, M. E. (1999, August 11). Abstracts, reviews, compilations, and indexes stored and retrieved using mechanical or electromechanical sorters. Retrieved November 21, 2006, from http://www.chemheritage.org/explore/timeline/MACHINE.HTM

Thomas A. Keswick