There was no single “first”
The title of this article suggests one first system. The historical record is less tidy. During the 1960s, several groups in different countries worked on putting maps and spatial data into computers, often with different goals and little contact with one another. Which one deserves the label “first” depends on how a geographic information system is defined, and historians generally describe credit as shared.
Here a GIS means software that stores locations along with attributes, such as soil type or land use, and lets a user combine and analyze them. That differs from simply drawing a map by machine, which is a related but separate idea.
The problem: too much land, too many maps
Governments in the mid-20th century faced a practical difficulty. They needed to decide how land should be used, but information about soils, forests, farmland, water and settlement was spread across many paper maps drawn at different scales. Comparing two maps meant overlaying them by hand, and measuring an area meant counting squares or tracing outlines with a device called a planimeter.
Computers offered a different path. If maps could be converted into numbers, then measuring, comparing and combining them could become routine calculation. The computers of the time were still large and expensive, which is why early projects were mostly run by governments and universities.
The idea of layered maps
The central concept behind GIS is layering. Each theme, such as roads, rivers, soils or forest cover, is kept as a separate layer. Layers can then be viewed alone or combined to answer a question, for example which areas are both near a river and on a certain soil type.
Layering existed before computers. Planners and landscape architects in the 1960s used transparent sheets, each showing one theme, and stacked them over a light table. One well-known landscape architect, Ian McHarg, popularized this overlay method in his 1969 book Design with Nature. The computer versions turned the same idea into data that could be stored, searched and recalculated, rather than redrawn each time.
Canada: a national land inventory
The project most often named in this story is in Canada. In the 1960s the federal government carried out a large land inventory to assess how land could be used for agriculture, forestry, recreation and wildlife. Roger Tomlinson is widely credited with leading the design of a computer system to handle the resulting maps, usually called the Canada Geographic Information System. It scanned map sheets, converted them into digital form and measured areas of land by category. Tomlinson is sometimes called the “father of GIS,” a label that reflects his influence but also simplifies a team effort.
The United States: Harvard and the census
In the United States, a number of efforts ran in parallel. A laboratory at Harvard, founded in the mid-1960s, developed programs for computer mapping and spatial analysis. One of its early programs, SYMAP, produced maps printed on ordinary line printers, using overprinted characters to create shading. These maps look rough today, but they allowed people without drafting skills to experiment with spatial data. Several later mapping programs and many researchers trace their roots to this group.
Around the same time, census authorities worked on ways to organize street and address information so that data could be tied to locations. Methods of this kind, developed in the late 1960s and 1970s, helped make it possible to map population statistics. Other countries also had projects. Several European and other national agencies began experimenting with digital map data in the 1960s and 1970s, and researchers in universities explored related ideas. This is one reason a story centered on one nation is incomplete. That is one reason a broader view of global computing history matters for this topic.
The technical hurdles
Early GIS developers faced limits that are hard to imagine now.
- Input. Maps had to be converted into numbers, either by tracing lines with a digitizer or by scanning. Both were slow and prone to error.
- Storage. Memory and disk space were scarce, so data was often kept on magnetic tape and processed in batches.
- Output. Results came out on line printers or plotters, and graphics screens were rare. Part of the later story of computer graphics is the arrival of interactive displays that made this work easier.
- Cost. Mainframe time was expensive, which is why early users were mostly large organizations. The wider setting is described in the account of how mainframe computers dominated the era.
Developers also had to choose how to represent the world. Some systems stored maps as grids of cells, each holding a value. Others stored points, lines and areas as coordinates. These two approaches, now often called raster and vector, both remain in use.
From specialist tool to everyday map
During the 1970s and 1980s, GIS software spread to utilities, local governments, forestry agencies and universities, helped by cheaper hardware. Falling costs and more capable microprocessors gradually moved the work from central computers to workstations and eventually to personal machines. Commercial software companies appeared, and standards for sharing map data slowly developed.
The legacy is visible in daily life. Online maps, delivery routes and location-based services build on the same basic idea of linking places to data, although the modern versions look far removed from line-printer shading.
How to read the claim
If you see a source naming the first GIS, treat it with some caution. The Canadian project is a strong candidate when the definition emphasizes a large operational system for national land data. Other projects have claims under different definitions, such as computer mapping, spatial analysis or census geography. The more defensible summary is that GIS emerged in the 1960s from several overlapping efforts, and that by the end of that decade the core ideas of digital layers and spatial analysis were in place.