Comments

• valerieschafer

  February 7 edited February 7

  @Titaÿna, thank you ! I would add to these questions, after having looked at the reimplementations by Montfort and GingerBeardMan: How is reimplementation a reinterpretation ? What does it tell us also about the legacy/ (and reborn-digital) heritage of Strachey?
• valerieschafer

  February 7

  @markcmarino pointed out to the very interesting piece by Link (http://www.alpha60.de/research/there_must_be_an_angel/DavidLink_MustBeAnAngel_2006.pdf), which is really stimulating, but how can we "queer" a bit this narrative and analysis through CCS ?
• Titaÿna

  February 7

  Thank you @valerieschafer for the responses and the suggestions.

  Regarding the reinterpretation point :

  Yes, I think both of them could be considered reinterpretations rather than reimplementations. At least with GingerBeardMan, the words used for randomness are explicitly listed.

  Since the original source code was lost, we can't determine which part of Roget's Thesaurus was used and how. However, I'm open to being proven wrong on this!

  The fact that the vocabulary was largely based on Roget’s Thesaurus lends a very peculiar flavor to the results.(Stratchey 1954 : 27)

  If other sources exist that could clarify how Stratchey selected the data or materials used for generating the letters, it might be possible to reconstruct it. If anyone has pertinent information, I'd be highly motivated to undertake this task. Perhaps it's partially what GingerBeardMan did, but I'm skeptical that the reconstruction effort would meet historian standards for re-implementation. I'll investigate this further.

  Work cited :

  Strachey, C. (1954, octobre). The « Thinking » Machine. Encounter, 25‑31. https://www.unz.com/print/Encounter-1954oct-00025/
• Titaÿna

  February 7

  @markcmarino pointed out to the very interesting piece by Link (http://www.alpha60.de/research/there_must_be_an_angel/DavidLink_MustBeAnAngel_2006.pdf), which is really stimulating, but how can we "queer" a bit this narrative and analysis through CCS ?

  The question of queerness can be approached through the question, "Can computers think?". In his 1954 article, Strachey discusses the Love Letter Generator, using it to critique what we now refer to as artificial intelligence. It's safe to say he would likely be dismayed by the current terminology used in relation to NLP progress and coding in general, considering the exercise to be a trick:

  It is clear that these letters are produced by a rather simple trick and that the computer is not really 'thinking' at all. This holds true for all programs which make the computer appear to think; upon analysis, they are nothing more than rather complicated tricks. However, sometimes these tricks can lead to quite unexpected and interesting results. (Stratchey 1954 : 27)

  At the same time, it's impossible to ignore the fact that Strachey, though he did not live openly, was a gay man. Consequently, the task of generating love letters, a seemingly straightforward task on the surface, is not merely a simple trick or a straightforward endeavor. It's a process that necessitates a more nuanced approach. In my research, I found some potentially beneficial hypotheses in the work of Wardrip-Fruin :

  "As Jeremy Douglass notes (2000), the love letter generator has often been used as an example in discussions of queer identity rather than considered carefully as a literary project. Certainly, there are reasons for this — Turing and Strachey were both gay, and at least Turing was openly so, at a time when homosexuality was illegal in England. It might also seem, from widely-reproduced outputs of the generator (e.g., that found in Hodges), that it was a love letter generator that 'could not speak its name' — the word 'love' being conspicuously absent.” (Wardrip-Fruin 2006 : 82)

  I'm still pondering whether the "trick" in generating love letters is to demonstrate that traditional love letters can be easily imitated, or if more exploration into Wardrip-Fruin is required.

  It might depend on the concept of whether the "trick", so to speak, in the creation of love letters is to effectively illustrate that the classical, heteronormative, tradition of penning romantic epistles can indeed be seamlessly imitated, thus challenging the unique, personal touch previously associated with them.

  Alternatively, I’m still considering that Wardrip-Fruin theory that this is a way to sign love letter without having to sign them is also a good hypothesis. It does go along with my research question “What does it imply that the love letters are signed by M.U.C. as an institution in this context of having fun around heteronormative expressions of love?”

  This remains an open question in my mind for now.

  Wardrip-Fruin, N. (2006). Expressive processing : On process-intensive literature and digital media. Citeseer. https://citeseerx.ist.psu.edu/document?repid=rep1&type=pdf&doi=a0f0bded35a10fb9ea3a5a000ff2b258770f3e15
• valerieschafer

  February 7

  @jarhmo pointed out to this stimulating post: http://jarahmoesch.com/2011/04/04/critical-code-the_ego_page-queerly-reading-the-computer-code/
  where this sentence (among others) is also food for thoughts :
  "This paper focuses on the programmer’s embodiment of the code and how the programmer brings notions of the world into the code itself, thereby creating an invisible layer of embedded normativity".
• JudyMalloy

  February 8

  Thanks @Titaÿna for this excellent topic!

  Riffing on algorithm antecedents in my paper on "'A WAY IS OPEN', Allusion, Identity, Authoring System, and Audience in Early Text-Based Electronic Literature" I wrote:
  __
  "Over the centuries, sometimes purposefully, sometimes with serendipity, in electronic literature and its precursors, narrative devices emerge, submerge, and emerge again, from a tenth-century bishop’s dice-driven gambling for virtues; to allusions to the worldly Chaucerian narratives of pilgrims on their way to Canterbury in the dice-driven Chaunce of the Dyse (Hammond 1925; Mitchell 2009; Sergi 2011); to echoes of Chaunce of the Dyse in the computer-mediated output of an electronic literature-influential triangle of two men and a computer—as Lytton Strachey’s nephew, Bloomsbury-bred computer scientist Christopher Strachey, and Manchester University’s historic mainframe computer, aka the Manchester University Computer (MUC) (probably the Ferranti Mark 1, which was prototyped by the Manchester Mark I), and Alan Turing, the man who designed a hardwired, noise-based random number generator for the MUC—collaborate in a series of groundbreaking computer-generated love letters created with Strachey’s software and Turing’s hardware (Strachey 1954: 25–31)."__

  And then after riffing on Chaunce of the Dice for a while and returning to the code for the Love Letters

  _
  "Sometimes, an influence can be suggested, but there is no definite proof. Oxford-educated Chaucer scholar, Eleanor Prescott Hammond, wrote the classic paper on Chaunce of the Dyse in 1925. Her work would very probably have been known in Bloomsbury circles. And (whether consciously or not) the shifting gender identities and texts of changing ideas of love, randomly assigned in the Chaunce of the Dyse, echo in the 1950s in the process of Christopher Strachey’s MUC Love Letters (Gaboury 2013)—and later occur and reoccur in the lives and generative poetry of the extraordinarily brilliant Fluxus couple, not-couple, couple, Dick Higgins and Alison Knowles."_

  Judy Malloy, "'A WAY IS OPEN', Allusion, Identity, Authoring System, and Audience in Early Text-Based Electronic Literature" in Electronic Literature as Digital Humanities. Dene Grigar and James O'Sullivan, eds. Bloomsbury Press, 2021. available at
  https://www.bloomsburycollections.com/monograph-detail?docid=b-9781501363474&tocid=b-9781501363474-chapter31
• valerieschafer

  February 8

  @JudyMalloy Thank you for sharing these stimulating quotes (and the link to the book, so much to discover!) and for also opening up this long-term perspective with the Chaunce of Dyse.
• Stefano

  February 10

  Yes I know it is too late but... I was thinking that the difference between a classical poem analysis and a code analysis is also the role of the subject which is doing the job. A subject with body, a personal history, specific competencies. Where the standard critical theory probably tends to hide the differences to concentrate on the rules of the analysis, here we could try to do something different. Should the subject be taken explicitly into account? How?
  A simple example: I've stumbled upon Stratchey code after having done the same kind of program within the goal of explaining to children how "dices" (thanks @JudyMalloy ) interplay with grammar and style rules to create infinite advertisments (or wine labels, recipes, soccer play live comment or letters) which show a taste of intelligence.
  So my surprise, and my pleasure, in discovering Love Letters came, how can I say?, from belly and not from brain. I had the impression that his code was a rewriting of mine. I felt a kind of brotherhood with Stratchey: did we have the same initial ideas, did we have the same overall goals? Had he the same kind of happiness the first time he saw the code runnng and producing letters? Did he fell in his own trap of thinking (hoping) that there were someone behind the letter?
  While I'm not saying that is fundamental to be a programmer to read critically a code, I suppose that to put oneself on the page aside the text (and the history behind, the knowledge on that piece of code accumulated since its creation) is very important.
• cschnitz

  February 11 edited February 11

  This excellent conversation reminds me of a bit I wrote in the introduction to my dissertation, copied below:

  Strachey’s M.U.C. Love Letter Generator is widely-cited as the first work of electronic literature (J. W. Rettberg). Electronic literature scholars and historians of computation have made much over Strachey’s choice to generate these saccharine, over-the-top love letters with his program. Strachey, a close colleague and contemporary of Alan Turing, was also a gay man in the United Kingdom at a time when male homosexual acts were criminal and punishable offenses. Strachey’s generator is, noticeably, genderless, using non-gendered descriptive language and centering only the transient reader as “you” and the M. U. C. as inhabiting the first-person “me.” After all, the Manchester University computer, the signatory of all the letters, has no gender and is not concerned with the gender of its program’s readers. The letters generated are also comical and campy, a parody of the genre and of the act of declaring one’s love in the first place. As Jacob Gaboury writes in “A Queer History of Computing,” Strachey’s generator is “a queer critique of normative expressions of love, enacted through a kind of generative, computational performance, through a purposefully deficient simulation” (Gaboury). The generator itself is also a procedural critique—the program is “a parody of process,” as Noah Wardrip-Fruin writes in “Digital Media Archaeology” (306). Telling someone how much you love them is at once ordinary and absurd, a prosaic task that is nearly impossible to get right in words. Putting the M. U. C. to this task highlights the task’s absurdity and impossibility.

  Strachey was not just setting the M. U. C. up for romantic failure with his love letter generator, though. He was using what the M. U. C. could do—randomize, generate, calculate, print—to make a creative critique on what humans so often fail to do. As a society, we so frequently fail to support the expansive and often queer nature of human love and we fail, constantly and flamboyantly, to figure out how to tell those we love how much we love them. Strachey, with his generator, was showing us the role of creative computation in critiquing some of the qualities that make us most human.

  With an emerging awareness of what the computational can do for social critique, Strachey and his love letter generator set a tone and a sensibility for the programmers, poets, and artists who sought to use computational processes creatively in the decades that followed. In starting with love and using the M. U. C. to fervently, breathlessly, ardently lay bare the deeply human limitations of expressing it, Strachey’s generator makes explicit the rhetorical capacity of the computational to comment and reflect on the social and relational.
• Titaÿna

  February 12

  Dear all,

  Thank you very much for the very informative feedbacks. With Valentine’s day this week I really appreciate these discutions.

  @JudyMalloy, thank you for your insights into the algorithm's origins, its ties to literary and cultural traditions, and its importance. I also appreciate you highlighting the connection to the Chaunce of the Dyse. Your phrase "triangle of two men and a computer" stands out to me, as it illustrates how a love letter written by a computer could challenge the heteronormative way of expressing love.

  @Stefano, don't worry, it's never too late. Thank you very much for your insights. Your description of Strachey's algorithm as more bottom-up than top-down, as we sometimes describe in sociology, underscores the idea that the concept of randomness in computational creativity likely doesn't have a single origin.

  While I'm not saying that is fundamental to be a programmer to read critically a code, I suppose that to put oneself on the page aside the text (and the history behind, the knowledge on that piece of code accumulated since its creation) is very important.

  I fully agree. I believe historians can learn from code and, conversely, code can also provide insights to historians. Understanding the "coding" aspect of the code as a resource is crucial. The ClimateGate incident (Marino 2020) serves as a prime example of a non-programmer attempting to interpret code, or perhaps trying to validate a theory using the code.

  @cschnitz, thanks for sharing !

  Strachey’s generator is, noticeably, genderless, using non-gendered descriptive language and centering only the transient reader as “you” and the M. U. C. as inhabiting the first-person “me.” After all, the Manchester University computer, the signatory of all the letters, has no gender and is not concerned with the gender of its program’s readers.

  Your answer sheds new light on the questions I had. Your work enables me to highlight Strachey's playful yet pointed challenge to conventional expressions of love. You also draw attention to the algorithm's significance as a queer critique and its relevance to contemporary discussions about identity and expression.

  I was asked to present this code briefly at our plenary meeting here in Luxembourg, as it coincides with Valentine's Day. Given that the code pertains to love, I was somewhat stressed about not being able to adequately link the source code to queer history. Nevertheless, your work has offered me ample inspiration on how to proceed.

  Marino, M. C. (2020). Critical Code Studies. MIT Press.
• kevinbrock

  February 16

  A related text that you might find interesting, if not useful, is Jim Brown's "The Machine That Therefore I Am" (Philosophy & Rhetoric vol. 47, no. 4, 2014, pp. 494-514; https://doi.org/10.5325/philrhet.47.4.0494), in which he explores the rhetorical action(s) of a nonhuman rhetor/actor when engaging in similar sorts of iterative document compositions, building on Erasmus' De Copia exercises.