— The Alan Turing Digital Archive and References to George Bernard Shaw

Saint Mary’s University, Halifax / September 19, 2022

Over the course of the summer, I have been engaged in an archival research project on the influence of fictive imaginaries (literature, myth, religion, etc.) in the work of digital computing and AI (artificial intelligence) pioneer, Alan Turing. Focusing on the field of AI, this project aims to deconstruct the founding mythos of AI and demonstrate that what is commonly thought to be “impartial” or “objective” science is imbricated with ideas that originate in fictive literature, myth, and religion and which continue to inform contemporary AI discourse.

As our research progressed my focus began to narrow down to the influence of one particular author on Alan Turing: Irish playwright, George Bernard Shaw. This is a connection that Andrew Hodges makes much of in his biography of Turing, Alan Turing: The Enigma. Hodges refers to Turing as “an admirer of Bernard Shaw” (308) and consistently elaborates on Turing’s thought through contradistinction to that of Bernard Shaw. Indeed, a cursory glance at Alan Turing: The Enigma would lead one to believe that Alan Turing was heavily influenced by Bernard Shaw’s plays – Back to Methuselah and Pygmalion in particular.

Although Hodges’ implicit and explicit argument that Turing was influenced by Bernard Shaw appears to support the thesis that Turing’s work is directly influenced by fiction, upon closer examination of the evidence, or lack thereof, this connection becomes rather tenuous. To begin with, the only piece of primary source evidence that Hodges cites is a 1933 letter from a twenty-one year old Turing to his mother where Turing writes, while studying at Cambridge, that “[t]here has been a very good play on here by Bernard Shaw called ‘Back to Methuselah’” (72; qtd in Hodges). This single sentence of Turing’s and the fact that he referred to Back to Methuselah as “a very good play” (74) thus becomes the only substantial piece of evidence that Hodges produces regarding Bernard Shaw’s influence on Turing.

What appears to be going on is that Hodges is using Bernard Shaw to inject a narrative into his biography of Turing. It is not so much that Turing was influenced by Bernard Shaw, but that Bernard Shaw and his writing provides a framework by which to interpret Turing. There is nothing inherently wrong with interpretive frameworks – they are a common device within the biographical genre and provide an effective (and affective) form of narrativization to human lives which commonly resist the convenience of linear narrative. The issue is that Hodges’ claims of Bernard Shaw’s influence on Turing have been picked up in further literature and have resulted in a circular, self-referential discursive apparatus that rests on little more than a single sentence in a single letter.

For instance, in an article for The Journal of Bernard Shaw Studies, Lawrence Switzky examines the “ELIZA effect” – the tendency to see responsive computer programs as more intelligent than they really are (50) – through Bernard Shaw’s Pygmalion. In discussion of Turing, Switzky points out that “Shaw shadowed Turing’s founding proposals about how to measure computer intelligence” (51) and that “none has been a greater devotee of Shaw than Alan Turing” (53). The evidence that Switzky cites is, once again, Turing’s aforementioned 1933 letter to his mother and Hodges Alan Turing: The Enigma. The connection between Turing and Bernard Shaw begins and ends with a single letter and the contrived bibliographical narrativization of Hodges. Upon unraveling the knot of Turing and Bernard Shaw, all threads lead back to Hodges.

Much like the founding myths and metaphors of AI, what appears to be at play here is a foundational mythos surrounding the influence of George Bernard Shaw on Alan Turing; a reified metaphor that goes full circle and begins and ends with Andrew Hodges.

Works Cited

Hodges, Andrew. Alan Turing: The Enigma. Centenary ed., New Jersey: Princeton UP, 2012.

Switzky, Lawrence. “ELIZA Effects: Pygmalion and the Early Development of Artificial Intelligence.”

SHAW: The Journal of Bernard Shaw Studies, vol. 40, no. 1, 2020, pp. 50-68.