The birth of BASIC, and the decade before Microsoft (1964–1975)

By the time Bill Gates and Paul Allen sat down to write Altair BASIC in early 1975, BASIC was already eleven years old. It had been used by tens of thousands of students at Dartmouth and dozens of other universities. It had been sold commercially by General Electric, Hewlett-Packard, Digital Equipment Corporation, Tymshare, and CompuServe. It had been written by hand for at least four major minicomputer architectures. By one of the Dartmouth co-creators' own retrospective estimate, somewhere on the order of five million people in the world already knew how to write BASIC programs before Microsoft existed. That figure should be read as a participant's after-the-fact guess rather than an audited count, but its order of magnitude is corroborated by the documented spread of BASIC across roughly 80 commercial time-sharing systems in the United States and dozens more abroad.

That spread is the editorial spine of this whole series. Microsoft did not bring BASIC to the world. Microsoft scaled an existing user base further, on cheaper hardware, with sharper licensing. The next section, Gates, Allen, and the Microsoft BASIC dynasty, covers what they actually did. This section covers what came first.

Dartmouth, the invention (1954–1964)

BASIC was created by John G. Kemeny and Thomas E. Kurtz at Dartmouth College in 1963 and 1964. BASIC is an acronym they coined: Beginner's All-purpose Symbolic Instruction Code. Kemeny was chairman of the Dartmouth Mathematics Department; Kurtz was the mathematician who ran Dartmouth's computing centre and had been Kemeny's colleague since 1956.

Kemeny's pedigree

Kemeny was Hungarian-born and emigrated from Budapest with his Jewish family in 1940 as anti-Jewish legislation tightened. He went on to a mathematical career that touched two of the twentieth century's most consequential figures:

• Manhattan Project. Drafted out of his Princeton undergraduate studies, sent to Los Alamos as a "human computer" performing the numerical calculations the bomb design required.
• Princeton, 1948 to 1949. At age 22, before he had earned his PhD, became Albert Einstein's mathematical assistant. During the same period he attended a lecture by John von Neumann at Los Alamos in which von Neumann laid out the architecture of what would later be called the stored-program computer. Kemeny, by his own later recollection in Man and the Computer, distinctly remembered thinking "God, I hope I live long enough to see such a thing."
• Dartmouth, 1954 onward. Don Morrison, Dartmouth's dean of the faculty, recruited Kemeny on Al Tucker's recommendation from the Princeton math department. Per Kemeny's daughter Jenny, both Einstein and von Neumann recommended him to Morrison. Kemeny arrived on the Dartmouth campus in 1954, and per a colleague's recollection, the senior faculty noticed immediately: "Things are going to be different around here."

Punch cards by train to MIT

Before Dartmouth had its own computer, Kurtz spent two years acting as the liaison between Dartmouth and MIT's computer centre. His own first-person account of that period is one of the more vivid pre-DTSS details:

"My job was to act as a liaison between Dartmouth and the MIT computer center. And it involved taking punch cards, and everything was punch cards in those days. And put them into a steel box and going down once every two weeks to MIT. This involved getting the 6:20 train out at White River Junction, and I did that every two weeks. ... Well, I figured out that the data transfer rate, we talk about gigahertz and all this kind of stuff, was 1.67 bits per second. That was the data transfer rate." Tom Kurtz, Birth of BASIC (Dartmouth video).

That ratio (bytes-of-program divided by round-trip-wall-clock-time) is the friction that motivated everything that followed. By 1958 Kemeny had decided Dartmouth needed its own machine.

The LGP-30 and the furniture-budget trick (1959)

Dartmouth's first computer was an LGP-30, paid for through an accounting workaround Kurtz delights in recounting:

"How can we get a computer into the new Bradley building? There's no budget for it. Ah, but there's budget for furniture and furnishings. Ah, a computer is a furniture, right? Yeah, okay. So that's how they figured out how to pay for the LGP 30." Tom Kurtz, Birth of BASIC.

The LGP-30 arrived sometime in 1959 and was placed in the basement of College Hall. Kemeny then secured National Science Foundation funding for undergraduate research assistantships because Dartmouth had no graduate students.

The post-Sputnik (October 1957) NSF was, per Kurtz, "funding all kinds of programs to support science instruction in the universities, graduate level and undergraduate level." Kemeny was, in Kurtz's words, "Johnny on the spot," recruiting students from places like the Bronx High School of Science the way other colleges recruited football players.

DART, ALGOL 30, and the realisation that ALGOL was too complex

Two languages preceded BASIC on the LGP-30 and shaped Kemeny and Kurtz's eventual design:

• DART, written by physics student Bob Hargraves. A higher-level interpreter for the LGP-30 that did basic arithmetic and built-in functions. Per Kurtz, "a language not quite as good as Fortran, but a simple enough language that one could do arithmetic, like A equals B plus C divided by seven or have a square root or something like that." Hargraves demonstrated that something simpler than Fortran could run on a small machine.
• ALGOL 30, written by Steve Garland. A port of ALGOL to the LGP-30. Garland's reaction, per Kurtz: "Gee, if Hargraves can do DART, I can do ALGOL." And he did.

Garland's ALGOL 30 worked but was, in Kemeny's view, still too complex for the audience he wanted to reach. He looked at simplifying ALGOL or Fortran for student use and concluded: "if you made them simpler, it was a different language." That's when he decided to write a new language from scratch, and write the compiler himself.

John McCarthy says: do time-sharing

The architectural inspiration for what would become DTSS came directly from John McCarthy, the AI pioneer, who had been at Dartmouth before going to MIT for better computing facilities. Per Kurtz, on a visit to MIT around 1962:

"John McCarthy, famous in artificial intelligence, had been at Dartmouth and went to MIT because they had better computing facilities at the time that he went. And he said, 'You guys should do time sharing.'" Tom Kurtz, Birth of BASIC.

Time-sharing meant running one job for one second, then the next, then the third. A way to give multiple users at multiple terminals the illusion of having the whole machine to themselves. The LGP-30 was too small for it, but the next generation of hardware would be capable. Kurtz brought the idea back to Kemeny and got an immediate yes.

The 1963 vendor selection

Dartmouth issued a competitive proposal to IBM, General Electric, NCR, Bendix, and Burroughs. GE's proposal won. "The best equipment for our purposes, in terms of what it could do architecturally, but it was also the cheapest." The letter of intent went to GE in summer or fall 1963. Hardware: a GE-225 (later GE-235) with a DATANET-30 front-end.

NSF's peer reviewers pushed back. Per Kurtz:

"We were going to develop a time sharing system using undergraduate students as programmers. And the peer review was, 'you can't have undergraduate students writing software for a major computing system.' Fortunately, Kemeny had such good relations with the people at the Science Foundation that in spite of these slightly negative reviews, they funded us." Tom Kurtz, Birth of BASIC.

The skepticism turned out to be unwarranted. Two undergraduates, John McGeachie (DATANET-30) and Michael Busch (GE-225), shipped the operating systems. Kemeny himself wrote the BASIC compiler, all 3,000 lines of it, by hand, beginning in summer 1963. He hired a Tuck school student named Bill Zani to do test runs:

"He'd wake up at three or four AM and work two hours doing the programming. And he would come in with the code and I'd meet him at 8:30, nine o'clock in the morning. He would go over it with me. And it would be handwritten. I would then have to put into punch cards of that code to be read into the GE computer in Linn." Bill Zani, Birth of BASIC.

Zani has a quiet but real claim in BASIC's history: "I can tell you for sure, I was the first man to see BASIC run."

May 1, 1964, proof of concept

On the night of April 30 to May 1, 1964, McGeachie and Busch were working on the operating systems all night. At what they later guessed was around 4 a.m., they got the system to run a simple BASIC program on two separate teletype machines simultaneously. That moment (one program, two terminals, both seeing it work) is the canonical "first run" of BASIC and DTSS together.

Kemeny was at the keyboards with McGeachie and Zani. Kurtz had gone home (he was an early riser), so the language's lead designer missed his own language's first run. The first BASIC compiler test (Zani's summer-1963 single-program runs) had happened months earlier on the GE-235 with single-user access. May 1, 1964 was the first time-sharing demonstration: BASIC running on simultaneous terminals through DTSS.

Kurtz's later assessment of the morning:

"What really happened on May 1st was a clear proof of concept. A clear demonstration that all the work that had gone into the thinking about whether or not one could actually share a machine amongst several people, the thinking about whether this simple language would work. All of that was proved correct." Tom Kurtz, Birth of BASIC.

By autumn 1964, hundreds of Dartmouth freshmen were using the system across 20 teletypes. By the late 1960s, 85% of Dartmouth undergraduates were using DTSS regularly across every major. By the early 1970s DTSS was running roughly 30,000 programs per day.

The vision, computing for everyone

The standard framing ("Kemeny and Kurtz wanted non-STEM students to be able to use computers") undersells what they were actually arguing. Their vision was universal computing access. The student population at Dartmouth was the test case, not the goal. The goal was a world where computing was a tool any educated adult could use.

In Kurtz's own words, opening the Birth of BASIC video (a Dartmouth-produced retrospective, see Source note at the end of this section for institutional-bias caveats):

"Computing was coming into its own. But in all of the other projects that were undertaken by industry and by universities, the target was research and development of computing ideas and so forth. Whereas, here at Dartmouth, we had a crazy idea that our undergraduate students who are not going to be technically employed later on, social science and humanity students, should learn how to use the computer. Complete nutty idea." Tom Kurtz, Birth of BASIC.

And the deeper version of the same argument, near the end of the video, not just students, but the masses:

"It was the first effort in the history of computing to try to bring computing and make it simpler and bring it to the masses, so that the masses could use it." Tom Kurtz, Birth of BASIC.

Kurtz's "first effort" claim deserves an asterisk. MIT's Compatible Time-Sharing System (CTSS) had been operational since 1961, three years before the May 1, 1964 first run of BASIC on DTSS. Project MAC at MIT, launched July 1963 under Fernando Corbató, was the parallel large-scale time-sharing effort. McCarthy himself had been advocating time-sharing as a democratisation tool at MIT before urging Kemeny and Kurtz to pursue it at Dartmouth. The honest version of Dartmouth's claim is narrower but still substantial: MIT had time-sharing first; Dartmouth had the first time-sharing system designed for non-specialists, with a programming language designed for non-programmers. The two efforts were running in parallel with overlapping democratisation framings; Dartmouth's distinct contribution was the audience (undergraduates from any major) and the language (BASIC, deliberately approachable), not the time-sharing concept itself.

That single editorial choice (programming language for non-programmers, computing for everyone) produced everything BASIC is known for: interactive use, plain-English keywords (PRINT, LET, GOTO, IF...THEN), forgiving syntax, helpful error messages, and the time-sharing system underneath that made the machine always available.

A colleague's summary of Kemeny near the end of the video gets at it cleanly:

"Kemeny was amazing. A visionary. He had a view of where we all ought to be computer literate. Before most of us even realized that computers existed."

Both founders are now gone. John Kemeny died in 1992. Thomas Kurtz died on 12 November 2024, eight months before this series began. Their first-person accounts now live only in archived oral histories at Dartmouth Libraries and the Computer History Museum, and in the Dartmouth-produced Birth of BASIC video this section quotes from. That video is, for now, the most accessible primary-source record of what they actually thought they were doing.

The intervening years, eleven years of BASIC before Microsoft (1964–1975)

Between Dartmouth's first run and Microsoft's Altair port, BASIC spread aggressively through the minicomputer and time-sharing world. By the time Gates and Allen got to it, BASIC was already a well-known commercial language with a decade of dialects and ports behind it. The implementations worth knowing about:

General Electric Mark I time-sharing (mid-1960s)

GE built the original Dartmouth hardware (GE-225, then GE-235) and went on to launch its own commercial BASIC time-sharing service. The first commercial BASIC at scale. This is the moment BASIC stopped being purely an educational language and started being a commercial one, and the first signal that the market for "computing accessible to non-specialists" was real.

HP Time-Shared BASIC and the HP 2000 series (1968)

Hewlett-Packard introduced its bundled time-sharing systems starting in 1968 with the HP 2000 series. Sold until June 1978; supported until 1985. Pricing started at $90,500 for the 2000A.

The HP 2000 was an immediate success and propelled HP to become the third-largest minicomputer vendor of the era. It was used heavily for education (statistics, regression, economics) and for business. It was also the source of canonical BASIC programs of the early seventies, including Mike Mayfield's Star Trek (1971), the text-mode game that defined a genre. The People's Computer Company published programs in HP 2000 format as a standard.

DEC BASIC-8 (1969)

In 1969, David H. Ahl at DEC hired a programmer to write BASIC-8 for the PDP-8. It was an immediate best-seller and became one of DEC's biggest-selling software products. DEC sold packaged PDP-8 plus BASIC systems as the "EduSystem" lineup, aimed explicitly at schools. Ahl later founded Creative Computing magazine in 1974 and is one of the most consequential figures in early educational computing, a name worth knowing.

DEC BASIC-PLUS (1971), the dialect Microsoft would pattern its BASIC on

BASIC-PLUS was DEC's extended BASIC dialect for the PDP-11 and the RSTS/E time-sharing operating system, shipped in 1971. The entire RSTS user environment was written in BASIC-PLUS. Login, logout, accounting, system management, all of it. It was DEC's flagship BASIC implementation.

Microsoft BASIC was patterned closely on BASIC-PLUS when Gates and Allen wrote it four years later. This is the load-bearing fact that connects this section to the next. The dialect Microsoft sold was not invented at Microsoft. It was a port of an established commercial dialect refined by DEC over the previous four years on much larger hardware. The interpreter was new code. The language semantics were inherited.

Tymshare SUPER BASIC (1968) and CompuServe DEC-10 BASIC (1969)

Two of the major commercial time-sharing services. Tymshare introduced SUPER BASIC in 1968. CompuServe launched in 1969 with a BASIC running on the DEC-10. Both targeted business customers buying time on remote machines rather than buying their own hardware.

Five million BASIC programmers before Gates

The strongest single piece of evidence for what BASIC's reach already was by 1975 comes from a Birth of BASIC participant's own retrospective math:

"I once estimated that even before Bill Gates got into the action at all, five million people in the world knew how to write programs in BASIC. There were something like 80 time-sharing systems in the United States that offered BASIC as one of their languages. And it was all over the world. I even got a letter from somebody in Siberia. A student in Siberia wrote me a letter once. This is before Gates, BG." Dartmouth participant, Birth of BASIC.

Five million. Eighty time-sharing systems in the US. A student in Siberia. All before Gates and Allen wrote a single line of Altair BASIC.

This is the number that anchors the rest of this series' editorial argument. Microsoft did not bring BASIC to the world. When Microsoft showed up in 1975, BASIC had already done a decade of real work, academically at Dartmouth, commercially at GE, HP, DEC, Tymshare, and CompuServe, internationally to the point of unsolicited fan mail from Siberia. The "computing for the masses" vision Kemeny and Kurtz had articulated as a "complete nutty idea" in 1963 had been substantially realised by 1975, on the back of minicomputer hardware that cost between fifty and ninety thousand dollars per installation.

What would change in 1975 wasn't the language. It was the hardware. The Altair 8800 cost $400 in kit form. The Microsoft contribution to BASIC's eventual mass-market reach was the work of squeezing an existing commercial dialect onto $400 hardware, then building the licensing machinery to make every major microcomputer vendor pay for it.

That's the next section. It opens, as these things do, with two young men reading Popular Electronics.

Source note, on the Dartmouth video

The Birth of BASIC video this section quotes from extensively is a Dartmouth-produced retrospective, made (per its framing) for BASIC's 50th anniversary in 2014. As institutional self-portraiture, it is celebratory by design. Several of its claims (the "first effort in the history of computing to bring computing to the masses," the "five million BASIC programmers before Gates" estimate, the "NSF's skepticism turned out to be unwarranted" framing) are recollections from project participants in their own promotional material, not independently audited or peer-reviewed. The directional facts (dates, names, hardware, the May 1, 1964 first run, BASIC's spread to commercial time-sharing services through the late 1960s) are corroborated across multiple independent sources and are reliable. The colour and the "we were first, we were right, they doubted us" framing should be read as Dartmouth's institutional voice, not as neutral history. Where this section uses direct quotes from the video, attribution is explicit and the institutional context is part of the read.

Sources

• Wikipedia, Dartmouth BASIC
• Wikipedia, Dartmouth Time-Sharing System
• Wikipedia, BASIC
• Wikipedia, HP Time-Shared BASIC
• Wikipedia, BASIC-PLUS
• Wikipedia, BASIC-8
• Dartmouth, BASIC at Dartmouth
• Computer History Museum, In Memoriam: Thomas E. Kurtz, 1928–2024
• YouTube, Birth of BASIC. Dartmouth-produced video; the primary source for direct quotes in this section. Full transcript filed in the series research folder.