A Queer History of Computing: Part Four

(0)

In Part Four of our ongoing genealogy of queer computing (Part One, Part Two, Part Three), we introduce a second generation of queer scholars who made important contributions to the field of computer science, and from whom we may trace a direct connection back to those familiar foundational figures.

On June 20, 2009 at 4pm at The Hampstead Quaker Meeting House in London, a memorial service was held for Professor Peter Landin. In attendance were his family and the friends whose lives he had touched over the last 78 years. It was a collision of worlds, a sudden mixing of two communities that Landin had kept separate his entire life. Landin's friend and colleague Olivier Danvy likened the event to the memorial for the French mathematical logician Jean van Heijenoort, author of From Frege to Gödel (1967).[1] In the early part of his life, van Heijenoort had been the personal secretary and bodyguard of Leon Trotsky, the famous Russian Marxist revolutionary and theorist, and the founder and first leader of the Red Army. Van Heijenoort left service only two months before Trotsky's murder in Mexico City by Stalinist assassins, but was a devout Trotskyist until his death, publishing extensively on his relationship with the revolutionary figure and editing a volume of Trotsky's correspondence before his own death in 1986. In attendance at van Heijenoort's funeral, Danvy recalls, were two disparate groups of people: on one side the logicians, and on the other the Trotskyists, each one incapable of communicating their own sense of importance of the man to the other.

Peter Landin had also led something of a double life. He was a foundational figure in computer science, and a pioneer of programming language design based on mathematical logic and the Lambda calculus. He was responsible for the invention of the first abstract process virtual machine—a kind of software emulator of a real world computer—ever defined. Many modern programming languages—such as JavaScript, the programming language that underpins much of the Web—make use of or fully rely on Landin's work on functional values, and have implementations based on his definition of a "closure," a programming function that “encloses” a set of variables so that it can be used in different contexts. Yet there was more to Peter Landin than this. All his life he had been a political radical, and since coming out in the 1970s he had been an active member of the Gay Liberation Front, protesting and campaigning on behalf of gay rights in the UK and abroad. His home on Rona Road in Camden had been a famous gay commune, and from the dinner parties he hosted many movements and collaborations were born.[2] And so on this Saturday afternoon two worlds met to commemorate his passing, and once again there seemed an impassable divide between these two parts of Peter Landin's life, these two worlds he simultaneously occupied. 

Landin was part of what we might describe as the second generation of influential queer figures in the field of computer science. This lineage is not simply chronological; there is a direct, genealogical connection between early foundational figures such as Alan Turing and Christopher Strachey, and those who lived through the pioneering gay rights movements of the 1960s and 1970s. The clearest queer lineage that begins with Alan Turing leads to Robin Gandy, his longtime friend and associate. Gandy first met Alan in 1939 as a student at Cambridge,[3] but they became particularly close when they were stationed together during the War and in the years following, and remained friends until Turing's death in 1954.[4] Gandy was never very explicit about his sexuality with friends and colleagues, but he and Turing seemed to share a mutual understanding and often discussed men and sex in a coded, joking way both in person and through correspondence. Landin shares a similar lineage with Christopher Strachey, having spent as brief period as Strachey's assistant after meeting in a bizarre set of circumstances that unite a number of key figures. Similarly, Landin spent a brief period as Christopher Strachey's assistant, the two having met in a bizarre set of circumstances that unite a number of key figures.

***

Peter Landin was born in Sheffield, England on June 5, 1930, eleven years after Robin Gandy and eighteen years after Alan Turing. The only child of an accountant father who had been disabled in WWI, he was educated at King Edward's Grammar School. Later, at Clare College Cambridge he completed a mathematics degree in a rushed two years, and then attempted the very difficult Part III course, but came away with only a 3rd class degree. As Landin tells it, he was unsure of what to do with his life after college, which led him to a now-infamous group of early computer science pioneers:

When I ceased to be an undergraduate, and because of being fast-laned through all these things and leaving with a rather ambiguously low grade degree I was very uncertain of what to do with my life and spent the next six months in a Sheffield reference library trying to avoid making a decision about my life. I used to go out to a cafe just around the corner from this reference library … and one day I was having my coffee in Fields cafe, and a voice came booming across the crosswise tables, and this voice said "I say didn't I see you reading Principia Mathematica[5] in the reference library this morning?" And that's how I got to know the legendary Mervyn Pragnell who immediately tried to recruit me to his reading group.[6]

Mervyn Pragnell is the mysterious figure orchestrating many of these early connections.  He is not only responsible for introducing many of these figures to one another, but also for introducing them to the lambda calculus of American mathematician and logician Alonso Church, which was essential to the development of a mathematical theory of computability. Not much is known about Mervyn Pragnell, as he does not appear to have ever held an academic post or published any research paper. Nonetheless, he was fascinated by mathematical logic in general, and Church's lambda calculus in particular. Much as with Landin, he was known for hanging around London bookshops approaching individuals he saw purchasing volumes on mathematical logic and recruiting them for a reading and discussion group. In an interview from 2000, Rod Burstall—one of many important logicians to get his start in Pragnell's groups – recalls that, while looking for a logic text in a London bookshop, he asked a man whether the shop had a copy. "I'm not a shop assistant," the man responded, and "stalked away," only to return to invite him to join the informal seminar where he would meet Peter Landin and, subsequently, Christopher Strachey.[7]

The sessions were held illicitly after-hours at Birkbeck College, University of London, without the knowledge or permission of the college authorities.[8] Pragnell knew a lab technician with a key that would let them in, and it was during these late night sessions that many famous computer scientists cut their theoretical teeth. This also appears to be the place Landin would first meet Strachey, and it marks the beginning of an important intellectual relationship between these two men. It is unclear how open either man was about his sexuality at the time—Landin, who identified as bisexual, would marry his wife Hanne that same year, and their marriage would last until 1973—but the connection is nonetheless meaningful, as it shows how intimately linked the world of computing was at this time, and how powerful these connections would be to research and development within the field.

Thus in 1960, nearly a decade after Strachey's love letter generator and six years after the death of Alan Turing, Peter Landin was taken on as a research assistant to Christopher Strachey, who at the time was an independent computing consultant working out of his home at 9 Bedford Gardens in Kensington. Having left the National Research and Development Corporation (NRDC), Strachey formally started activities as a private consultant on June 1, 1959. By 1960 he was fully occupied with a number of contracts, many of which he had begun while employed by the NRDC. Strachey took Landin on as a full-time employee specifically for a contract with the Ferranti electrical engineering and equipment firm, for whom he had agreed to deliver a scientific autocode – the term for a family of "simplified coding systems" or programming languages devised for early computers – for the company's new Orion computer. Landin set upon the project with great ambition, imagining an innovative compiler that functioned "as an automatic product of the semantics of the autocode, matching its forms to semantic representations of the instructions of the machine, and generating LISP expressions that could be executed."[9]

[9 Bedford Gardens, London]

While Landin was working for Strachey full time, he was not fully occupied by the Ferranti project, and with Strachey's encouragement he spent much of his time on a theoretical study of programming languages. It gave Strachey a certain satisfaction to be able to claim that he was funding "the only work of its sort being carried out anywhere (certainly anywhere in England)."[10] Whether due to this split in the time spent on the Ferranti project, or due to the overly ambitious and theoretical work Landin was attempting with his compiler, the work was never fully finished, and required additional work by Ferranti's own programming department to bring it into workable condition. Still, the research Landin began here with the support of Strachey was foundational to his study of programming languages. It allowed him to clarify his ideas about programming semantics and led to the publication of "The Mechanical Evaluation of Expressions" in 1964,[11] which showed how to translate programs into lambda calculus and defined the SECD machine, a landmark virtual and abstract machine that emulates a hardware environment within which lambda calculus expressions may be evaluated.[12] Landin hoped this work might form the basis of the design of future computers, and in many ways it has.

 

As Landin was conducting research and raising his children, a cultural shift had begun. In the US and Canada a transformation was underway. The former had seen the now infamous Stonewall Riots in Greenwich Village and the beginnings of a social movement for gay and lesbian rights. The Gay Liberation Front, or GLF, was formed by thirty-seven women and men who broke ranks with the conservative homophile establishment and urged a candlelight march in response to the riots. The GLF first took hold in the UK in 1970, growing rapidly over the next three years before splitting into a number of spin-off organizations such as the London Gay and Lesbian Switchboard, many of which still thrive today. In 1971 it issued a manifesto comprising  a list of immediate demands, including the decriminalization of homosexual acts. While the law criminalizing homosexual activity that led to the arrest of Alan Turing had already been overturned by the Sexual Offenses act of 1967, that legislation set out explicit terms by which homosexual acts would be deemed legal, namely mutual consent, a minimum age of 21, and that sex take place in private between no more than two people. Thus, it was far from the end of the struggle to end homosexual persecution, and in many ways it marks the beginning of a long legal battle that is still ongoing.

1970 would be a transformative year for Peter Landin as well. The previous ten years had truly shaped his career, but he was set to undergo a massive change. In 1964 Landin had ceased working for Christopher Strachey and, through contacts provided through their relationship, was "brain drained" to the US and—along with his wife and two small children children—he moved to New York City to work for Univac, then a major computer manufacturer. The family first took up residence in a hotel, but after asking for a home with a garden they were moved to a half-house in Greenwich Village. Landin published several key works during this time period, perhaps most famous among them a short work titled "The Next 700 Programming Languages" (1966), in which he gave a witty account of how all programming languages of the time were just sugared[13] versions of the lambda calculus.[14] By 1966, Landin was tired of the corporate world of New York City, and so moved with his family to Cambridge, MA to take up a teaching position at MIT. Still, he was disheartened by what he saw as a secretive environment that shunned collaboration, along with a group of colleagues with very different ideas about the logic of programming languages. And so in 1967 he was tempted back to London with the chair position at Queen Mary College, where he remained for the rest of his career.

Then suddenly, in 1970, Landin made the abrupt decision to walk out on the discipline of computer science. After serving as the evaluator on a student's PhD committee, he decided that the field had become too theoretical and retired. Having attained the position of full professor, he was given emeritus status and continued on in a reduced capacity at the university for the next forty years, but for Landin something had changed and he was no longer interested in the kind of innovative research that had occupied the previous fifteen years of his life. It was also during this time that Landin's personal life underwent a transformation. In 1973 he separated amicably from his wife, though he remained close to her and his children for the rest of his life. He was also becoming involved with the GLF and other burgeoning gay organizations, and was even arrested during a gay rights protest in London.[15] A regular on the lawn at Hampstead Heath, frequent dinner party host, collaborator, facilitator, and activist, Landin underwent a substantial transformation as he moved from one life into another. 

It is in this period that Peter Landin's life begins to recede from view. The archive fails, and forty years are devoured by the impassable partition that he erected between his personal and professional life. No doubt there exist many people who could share fond memories of Peter's activist years and his role as an organizer and friend, but these stories have not yet come  to light.[16] Instead, it is his professional service and contacts that remain. What little that exists in the way of memorial and biography has been produced by university colleagues, and while it is currently unclear what will happen to Landin's papers, correspondence, and materials, too often in such cases these details are deemed "personal" and are either excluded or reserved until some future date. Toward the end of his life, Landin became "convinced that computing had been a bad idea, giving support to profit-taking corporate interests and a surveillance state, and that he had wasted his energies in promoting it."[17] It is perhaps unsurprising, then, that there exists almost nothing online about the last forty years of Landin's life and that, despite his influential role in the development of the field of computer science, Landin did not own a computer, a television, or a car.[18]

This is, in part, the reparative work that this essay hopes to accomplish. In linking the professional accomplishments of these men with those personal parts of their lives that even they may have deemed inappropriate for public discussion, my hope is to create a queer archive that links foundational developments in the history of computer science to explicitly queer figures and politics. It is, in part, a refusal of the separation of these worlds, and an acknowledgement of the way in which the sexual lives of these men are part of the historical significance of contemporary computational technologies. It's not that these facts have been hidden or are not known, it's that there is often a compulsion for historians to pass over them in silence. As with Landin, many of these figures have only recently passed away, and many others will be gone in the coming years. As a result, preserving these histories is of particular importance, as is producing an archive that reflects the complex divisions and connections that constitute these lives and this history.



[1] Olivier Danvy, "In Memoriam Peter Landin," Vimeo.

[2] His colleague Richard Bornat notes in a commemorative article in the Formal Aspects of Computing journal, that "It was at one of his dinner parties that those who reinvigorated Gay Pride marches in the mid 80s met, just in time for the battle over clause 28" Richard Bornat, "Peter Landin: a computer scientist who inspired a generation, 5th June 1930 - 3rd June 2009," Formal Aspects of Computing (2009) 21: 394.

[3] The two met at a party in which Gandy was arguing in support of the Communist line in the Winter War between the Soviet Union and Finland. For many years Gandy was a member of the Communist Party, yet somehow escaped scrutiny even after the controversy surrounding the Cambridge Five, the group of Soviet spies believed to have been recruited through the Apostles society at Cambridge.

[4] Gandy was a mathematician and logician, but not technically a computer scientist. In this sense he does not neatly fit into this history, but as one of Alan's closest friends he was the strongest link to his life and work until his death in 1995. Moreover, while Gandy's work in mathematical logic was not explicitly in the field of computing, it should be clear by now that the fields share a common history and are very much aligned.

[5] Principia Mathematica, mentioned previously with regards to Turing, is a three-volume set of texts written by Alfred North Whitehead and Bertrand Russell, published in 1910, 1912, and 1913 respectively. It is an attempt to derive all mathematical truths from a well-defined set of axioms and inference rules in symbolic logic. It is widely considered to be one of the most important and seminal works in mathematical logic and philosophy.

[6] Peter Landin, Untitled talk at "Program Verification and Semantics: The Early Work," BCS Computer Conservation Society Seminar, Science Museum, London, UK, June 5, 2001.

[7] Donald MacKenzie, Mechanizing Proof: Computing, Risk, and Trust, (Cambridge: MIT Press, 2004), 273.

[8] Rod Burstall, "Christopher Strachey – Understanding Programming Languages," Higher Order and Symbolic Computation 13 (2000), 51.

[9] Bornat, Ibid., 393.

[10] C. Strachey, Curriculum Vitae (1971), Strachey Papers, A3.

[11] Landin, P. J. 1964. "The mechanical evaluation of expressions." Computer J. 6, 4, 308-320.

[12] Bornat, Ibid., 394.

[13] The phrase "syntactic sugar" was also coined by Landin in 1964 to describe the surface syntax of A Programming Language (APL) which was defined semantically in terms of the applicative expressions of lambda calculus. It has come to refer to any syntax within a programming language that is designed to make things easier to read or to express, that is, it makes things "sweeter" for humans to use, even if they might be expressed more cleanly or succinctly in a number of alternate styles.

[14] The phrase "The Next 700…" has since been adopted as a kind of meme among computer scientists, spawning a number of speculative papers charting the future of a given field.

[15] Landin had previously been arrested while on a demonstration with the Committee of 100, the 1960s anti-war group founded by Bertrand Russell. He was sentenced Pentonville Prison, but only lasted a week before he became so bored that he paid the fine to be released. (via)

[16] In writing this piece I reached out to one of Peter Landin's children, but did not receive a response.

[17] Richard Bornat, Peter Landin Obituary, The Guardian, September 22, 2009.

[18] In fact, Peter Landin never learned how to drive. He was well known for biking everywhere he went, even into his old age.