Harvard Laboratory for Computer Graphics and Spatial Analysis

Howard Fisher, a Chicago architect, founded the Laboratory for Computer Graphics in the Graduate School of Design at Harvard University with a grant obtained from the Ford Foundation in December 1965. Fisher had observed the computer maps produced by Edgar Horwood's group at University of Washington at a training session held at Northwestern University in 1963. Fisher thought he could produce a more flexible cartographic tool, so he set out to design a software package he called “SYMAP.” This prototype that Fisher built at Northwestern University served as the basis to obtain the grant from the Ford Foundation. Once the laboratory was established at Harvard, a more polished version was developed by a team of programmers. It was accompanied by training materials in the form of a correspondence course, which were made widely available to universities around the world and incorporated into advanced university courses.

SYMAP could handle attributes attached to points, lines, and areas, to produce choropleth or contour maps on a line printer. These displays were crude but readily accessible in the computer center era. The interpolation technique was particularly sophisticated for the time. SYMAP was distributed to over 500 institutions and set an early standard for cartographic software distribution, with an inexpensive charge (initially $100) for source code copies. Other packages followed, including SYMVU, for three-dimensional display; CALFORM, for plotter output; and POLYVRT, to convert cartographic databases.

The laboratory became a center for research on spatial analysis. William Warntz was appointed the second director in 1968, under the amended name, “Laboratory for Computer Graphics and Spatial Analysis.” Warntz worked on the theory of surfaces and sparked a number of developments that later became central to modern software systems. Another team at the laboratory, under the direction of Carl Steinitz, developed techniques for environmental planning and experimented with grid analysis software.

This early laboratory employed almost 40 students and staff by 1971, but it declined rapidly as the funding dried up. A second phase of the laboratory built from this low point, under the direction of Allan Schmidt. Starting from a focus on topological data structures, a team of programmers built a prototype software system called “ODYSSEY.” Laboratory researchers also experimented in cartographic visualization, producing the first spatiotemporal hologram, for example. At the same time, the laboratory hosted a series of annual conferences. At the second high point in 1981, the laboratory staff had surpassed 40. An agreement was signed to transfer ODYSSEY for commercial distribution, but Harvard then decided not to pursue this option. The laboratory was reoriented toward research, and the staff dispersed rapidly. The laboratory continued to exist until 1991 with just a handful of students and researchers. In the last period, some innovative digitizing packages were developed for early personal computers.

While the Harvard software is long obsolete, the laboratory sparked innovations in computer mapping, [Page 220]analysis of spatial information, and geographic information systems. The students and staff associated with the laboratory went on to play important roles in academia and industry.