Jean Sammet

Mathematics and industrial programming

Sammet was born in New York in 1928, earned a mathematics degree at Mount Holyoke in 1948 and a master’s degree at Illinois in 1949. After teaching and other work, she entered industry at Sperry Gyroscope.

She organized a scientific programming group and then worked at Sylvania on software for MOBIDIC. The move from mathematics to software involved managing people and systems as well as expressing calculations in code.

Sources for this section: [1]

A committee language and a symbolic system

Sammet served on the committee that developed COBOL before joining IBM in 1961. At IBM she directed work on FORMAC, a system for manipulating symbolic mathematical expressions.

Her 1981 account credits assistance from Robert Tobey in the early FORMAC work. The aim was to join formal algebraic capability with an existing numerical programming language, rather than treating an unevaluated expression as an error.

COBOL’s committee history and FORMAC’s team history require different accounts of responsibility. Neither is accurately described by naming a sole inventor and omitting everyone else.

Sources for this section: [2] · [3]

Preserving a field’s development

Her 1969 book Programming Languages: History and Fundamentals documented languages and their development. Her later IBM responsibilities included programming-language technology and work associated with Ada.

Historical classification was itself a form of technical work: a language’s notation, intended users and implementation could differ. A useful history needs evidence for those differences rather than a list of names arranged as inevitable progress.

Sources for this section: [1]

Professional leadership and later recognition

Sammet was president of ACM from 1974 to 1976 and became a member of the US National Academy of Engineering in 1977. She received a Computer History Museum Fellowship in 2001 and died on 20 May 2017.

Her career joined language construction with professional organization and the preservation of computing history. FORMAC is especially revealing because it asks the computer to retain mathematical structure, not merely replace letters with numerical values.

Sources for this section: [2]