Plagiarism, Technology, and the Dead-End of the Humanities

February 24, 2022

Left: Utagawa Hiroshige, Plum Garden in Kameido, 1857. Right: Vincent van Gogh, Flowering Plum Tree (after Hiroshige), 1887

by Justin E. H. Smith

My friend Agnes Callard stirred up some mischief a while back by writing at least somewhat sympathetically of the deeds of plagiarists, including her own past self in elementary school. “Academia,” she observed, “has confused a convention with a moral rule, and this confusion is not unmotivated.”

In order to count as plagiarism, I take it, a piece of written work must be falsely presented as if written by its presenter, when in fact it was written by someone else. So defined, I suppose I am in most circumstances “opposed to plagiarism,” roughly in the same degree and with the same intensity as Agnes. Though much like her as well, rather than simply taking a stand on the issue —as if that’s what an intellectual were supposed to do—, I am much more interested in figuring out the historical and technological circumstances in which this gesture came to be something like the literary and academic equivalent of murder, the absolute unspeakable act that can only eventuate in total exclusion from the ecumene, this even at a time when information-processing technologies are, like it or not, largely obviating the need for a well-rounded, generally competent person to develop the skill of long-form textual composition at all.

This coincidence of technology and taboo cannot but make one wonder, that is, what the motivation of the academics is to maintain their dead-ender fight against plagiarism even as the skill it counterfeits becomes more irrelevant. An analogous puzzle arises for me when considering popular discussions of “book banning”, where Heinrich Heine’s line is piously cited about how book-burnings are always followed with corpse-burnings as if by some iron law, as if the book were a stable transhistorical entity, rather than something with different powers and threats packed into it depending on the total information ecology in which it is found. Ban all copies of the Vulgate Bible from the isle of Britain — as long as I can use a VPN address there to connect to the wonderfully Leibnizian and encyclopedic Bible Gateway website, I can still read the Word of God in English — in a dozen different translations in fact, as well as in Navajo, Aleut, and Yakut (evangelizing efforts in the modern period seem often to have been pursued principally as a pretext for the study of comparative linguistics, a fact still evident today in the web-based resources of Evangelical institutions). Anyone who is still policing the circulation of printed materials in an era when the vastly greater part of information flow is occurring digitally is simply not paying attention to the realities of our new information ecology. I am not of course saying that it no longer matters when provincial dimwits remove copies of Toni Morrison’s Beloved or Art Spiegelman’s Maus from school libraries, but only that we must be honest about the way information circulates today, especially the way it circulates among the school-kids whose libraries are affected, rather than pretending that books have the same function today that they did in the Reformation.

Often it’s the very people who are most publicly indignant about the thinning of school-library shelves who seem to spend much of the day circulating meme images of splendiferous reading rooms at dreamlike libraries in Lisbon or Budapest and writing captions such as: “OMG, bookgasm”. It’s fairly obvious that such expressions are themselves symptoms of the decline of the book, rather than instances of resistance against this decline. No true bibliophile would ever think to indulge in such tacky fetishization, let alone to put it on display for all the world to see. Nor does one have to be a confirmed Freudian to agree that the fetish emerges in compensation for the loss of an object we could once take for granted.

And as with the reading of books, so too with the student composition of the sort of texts that continue to be valued and described as an apprentice’s first stabs at authorship, and thus as propaedeutic to book-writing.

I do not want to say that the valorization of compositional skill, and the condemnation of fraudulent imitation of it, is only a fetishization and nothing more. But it does seem that the redoubled defense of this skill, as if it were a transhistorical necessity, is best understood as a reaction, on the part of those of us who are old enough to be invested in it, against its evident demise.

This may perhaps be more clearly seen if we consider plagiarism within a broader field of historical practices for the cultivation of literacy. There is in particular a species of writing that appears to belong to the same broad genus as plagiarism, yet that does not involve taking someone else’s work and presenting it as your own, but rather taking your own work and presenting it as someone else’s.

There is a range of reasons why someone might do this. There is, first of all, the hoax, in which a forger seeks some sort of gain, usually financial, by claiming to have in his possession a document whose value depends on its supposed authorship by a more distinguished person than himself. Whatever else might be said of it, forgery requires deep familiarity with the object one seeks to imitate, and is thus a form of mastery even when the master uses his skill for blameworthy purposes. As Ken Alder shows in his remarkable work on the nineteenth-century French forger Denis Vrain-Lucas (1818-1882), in that era forgery was much more than just a potentially lucrative criminal enterprise, but something rather like a parallel world for the exercise of imagination and artisanal ingenuity together, and even for the improvement of history itself by retroactive injection of new facts into it.

Denis Vrain-Lucas’s letter from Joan of Arc to the Parisians

I myself have recently been working on an Italian missionary in Ethiopia, Giusto d’Urbino (1814-1865), who may have forged a Ge’ez manuscript attributed to the Ethiopian philosopher Zera Yacub. Our current inability to conceive why or how someone could go to such great lengths to do something like this —learning the classical Ethiopian ecclesiastical language, among other things, and enough about paleography to create a passable version of a text held to be 200 years old at the time of its discovery—, stems not only from the fact that we are more prepared today to entertain the idea that a seventeenth-century Ethiopian might have written philosophy, but also from the fact that we are less prepared today to entertain the idea that a nineteenth-century missionary could have had the talent and determination to think his way into the intellectual world of a seventeenth-century Ethiopian, and to generate a text as if drawn out of that world by magic. We might previously have been overlooking the richness and depth of early modern Ethiopian thought, I mean, but we might also, now, be forgetting the richness and depth of the high-modern culture of forgery.

Forgery exists on a continuum with forms of imitation that are not blameworthy at all, but that indeed occupied a central place in humanistic and literary education up until the last few decades: namely, the sort of training in which one is made to write in the style of some exemplary predecessor. In this light, d’Urbino’s “copy” of Zera Yacub’s text could generously be interpreted as an exercise by which he demonstrates his thorough internalization of the style and conventions of Ethiopian intellectual tradition (even if he ultimately lied about its true nature to Antoine d’Abbadie, the Paris-based collector prepared to pay him for it).

A century ago a Latin exam might have given you a few lines in English and instructed you to translate them into Latin dactylic hexameter in the style of Virgil’s Eclogues. The point was obviously not to make you fraudulently present yourself as Virgil, but only to make you internalize the feeling for the language that made Virgil such a master of it. Art historians and collectors likewise often have trouble determining whether a painting comes from the workshop of a given artist, or from the artist himself, even if no intentional deceit was involved in its original production. The workshop disciple was learning by imitating, and of course to master such imitation is to gain the dark power to deceive if one wishes. But this power itself, like all art, is morally neutral.

Imitating an exemplar’s style is one way of internalizing their work and their ability, though there is another even more direct way of doing this that has been at the core of many traditions of knowledge-transmission in many societies across the millennia: namely, rote learning, straightforward memorization, word for word, of an entire canon, in the form of sutras, for example, or chapters and verses of scripture.

It is, mostly, transformations in information technology, notably the gradual decrease over the centuries in the expense of production and ownership of written materials, that caused a corresponding diminution in the importance of learning bodies of material in this way. There is less need to internalize a system of knowledge, literally to carry it inside you, when you can simply store it externally in the form of a book that is always at hand.

It is a commonplace of the philosophy of education in our world of technological mediation, for which thinkers like John Dewey and Maria Montessori were unconsciously offering tailor-made recommendations, that rote learning is bad, reactionary, an expectation of only the most backward of knuckle-wrapping school-marms. Generally speaking I honestly do not know where I stand politically — I like liberty and equality both, and I have real trouble understanding what system would harmonize these two desiderata in the most effective way. But I know that at least as far as education is concerned, I am a conservative, or, more precisely, as I’ll explain soon enough, some kind of postmodern conservative.

I am conservative in that I think rote learning is not only good, but that it is the gold standard of learning and would ideally precede and support all other forms (don’t worry, present and possible future employers — I am also a realist, and would never impose my vision on any actually existing classroom). I believe the prevailing assumption of the evil of rote learning, as it develops from Rousseau through Dewey and into the managerial truisms of education schools today, is based on the tragically misguided idea that the principal business of education is to draw out the singular voice of each individual, rather than drawing individuals into a community of shared knowledge, otherwise known as a tradition.

It is not that I do not think each of us has a singular voice, for indeed we do, but that does not mean that we will all spontaneously have something to say about King Lear or Old Yeller or whatever that is worth hearing. This is all the more the case when the actual expectation that a text be read before it is spoken of seems further and further from the realities of the college classroom, so that what is turned in as a “reading response” is often something more like an improvised bluff, an imitation of the outer forms of a lost art that the student has never really studied, much like what I would do if I were commanded at gunpoint to mount a stage and to start dancing ballet. A far better exercise, under such circumstances, remains now as centuries ago to ask the student not to write about King Lear, but to perform it — to internalize considerable portions of the text, without regard for whether the meaning itself is internalized. That can come later, perhaps years later, once it is inside of you.

The idea that each of us has a singular voice, and that it places a stamp of authorship on our compositions, is of course corollary to the prohibition on plagiarism. Significantly, there are other intellectual, expressive, and creative traditions in which the proprietary relationship to one’s output is understood differently, and there may be something to learn from these.

As Charles Upton observes in his truly sui-generis 2016 book, What Poets Used to Know, the closest thing to a living bardic tradition we have in the contemporary United States is likely to be found among the fire-and-brimstone preachers, who have typically internalized the entirety of scripture early in their lives (listen, for example, to Son House riffing on the numerology of the Book of Revelation), and then spend their careers freely improvising on the body of knowledge that is both within them and shared among them equally. They develop and cultivate signature styles of delivery, they have favorite passages of scripture, but their authority in delivering sermons is not generally perceived as authorship in any sense that might be meaningfully protected as intellectual property.

It is when a world such as that of the preacher collides with that of the scholar, for example when managerialization of the preacher’s vocation compels him to go and get credentialed with a doctorate, that a particular danger of plagiarism arises. Thus for example Keith D. Miller argues —compellingly in my view— that Martin Luther King Jr.’s embeddedness in a communitarian tradition that valued “voice-merging” mitigates at least in part the gravity of the accusations of plagiarism against him in his Boston University doctoral dissertation on Paul Tillich and Henry Nelson Wieman. Why, indeed, should voices not be merged? And even if a preacher is compelled to pass through a secular credentialing institution, and would probably do better to play by its rules while inside it, on what grounds can we say that the conventions of authorship, and the epistemology of knowledge production this implies, must win out in any absolute sense over an expressive tradition that is vastly older and in many respects more vital?

For quite a long time, before I was sure of myself as a writer, before I was confident in my own voice, I regularly took large chunks of other peoples’ writing and mixed them with my own, sometimes changing a few of the words to fit more closely with the voice I was still struggling to find. I never let any of these experiments slip out of my private custody, but for some years I was periodically seized by the paranoid fear that I had done so, perhaps in notes that I had shared in a private circle, perhaps even in an article I had published. If in fact I had let this happen, it seems to me it would have been excessively harsh to exclude me from the ecumene of scholars or writers on this basis. In fact it seems to me that the habit of voice-merging, as I had come to practice it, was an entirely salutary, natural, and necessary thing for me to do at the time. In part this need, for me, arose because I was working largely outside of any tradition; no one ever really taught me how to write, and I needed exemplars badly. If I had inadvertently plagiarized, and had been caught out for it, I could only have protested that I had no other choice, I was only learning how to write.

The problem is compounded today by the fact that, while most of us have been cut off from any vital tradition of voice-merging with other human beings in our community, we now quite regularly and as a matter of course, in the production of our own compositions, or imitations of compositions, merge our voices with the imitation voices of machines. This happens, in fact, whenever autocorrect transforms a Latin word that I’ve written into its nearest English equivalent. When I write an e-mail using Microsoft tools, AI attempts to predict what I am about to say, and recommends in a slightly lighter shade the words that might correctly finish the sentence. No student is prohibited from using such tools as autocorrect or auto-fill in order to write a paper, even though quite obviously these are tools of the same general kind as an AI program that simply takes key words and generates an entire essay on them. The paraphrasing that I used to do in my copy-pasted fragments of other people’s writing can now be done with AI tools, and universities are now modernizing their anti-plagiarism arsenals to include paraphrase-detection software. But why permit machine authorship at the level of words or sentences, but not paragraphs?

I have become convinced that the best way to respond to the threats posed by new technologies to the conventions of literacy that we continue to strive to maintain is to return to the practices of knowledge transmission that preceded literacy altogether and that characterized oral cultures: namely, education should involve the rote internalization of texts. Make kids memorize epic poetry, or the Bible, or King Lear, or the Upanishads, through recitation, adding a few lines each day over the course of several years. They do not even need to see the texts in written form, but only to hear them and speak them. Eventually one might make them compose epic poetry in their heads, too, like Solzhenitsyn biding his time in the Gulag by reverting to oral literature when his writing implements had been taken away, adding a few lines each day, accreting a world inside of himself.

It is an advantage of rote memorization that you don’t have to comprehend anything about what you’re learning. Just get it in there, as early as possible, and then you’ll have a whole lifetime to unfold the meanings you are already carrying around inside you without knowing it. Nor is it so important what canonical work you select to fill up on either; the wonderful thing about living tradition is that it can make stone soup out of even the most modest ingredients in the initial text. Entire religions that survive across millennia have been cooked up around what are in the end just a handful of run-of-the-mill popular legends.

Nonetheless —and here is where I will sound particularly conservative, so brace yourself— if your education is in English, then you will do well to master bodies of work that were written in English, as well as in French, German, Anglo-Saxon, Norman French, Latin, and Greek. These are the languages that do the most to shape the meanings of the words you are using, and to write in the proper sense is to deploy words that you know inside and out, phylogenetically, cross-linguistically, morpheme by morpheme. And this is why autocorrect and auto-fill are so insidious: in order to really be writing, you have to know the meanings of words, their entire histories, the way they are interwoven with the meanings of other words in related languages, and the way they descend from ancestor words in ancestor languages. Surely, in this light, it would be far better to have mastery over individual words, and to let AI write the paragraphs built up from them, than to let AI choose and correct our words, but to complete the paragraphs they constitute on our own. And yet under our current brain-dead epistémè, the former approach to the production of texts would be considered “plagiarism”, while the latter is just using helpful word-processing tools. This simply makes no sense.

I have said that in terms of my philosophy of education I am some sort of postmodern conservative. I’ve explained the conservative part, so let me now get to postmodernism.

I suppose what I mean first and foremost by this is that I think humanistic inquiry is best pursued not only in the mode of stern truth-seeking, but also with a reasonable admixture of the spirit of play. Recently, some co-authors and I attempted to inaugurate our own modest “ludic turn” in the history of scholarship by producing a work of historiographical metafiction that deployed the conventions of the apparatus criticus, the abstract, even the referee’s report and the offprint, as the elements of a literary genre yet to emerge. So far, our effort seems mostly to have fallen stillborn from the shipping crates in which the physical copies of the book arrived from our presses in far Estonia. But we’re playing the long game and we continue to expect to see some serious critical uptake soon.

For me personally, this project sometimes reminded me of a movie I saw long ago called Окно в Париж/Window to Paris, a late-Soviet comedy in which the inhabitants of a Leningrad kommunalka discover a top-floor window through which they can instantaneously travel to the City of Light, thus generating all sorts of predictable gags of the same species as one might see in Beverly Hillbillies. I don’t remember much of what happened, but there was a scene in which an extremely blasé and cynical Russian émigré violinist in a Parisian café, bored with his own fate and with his absolute maestroship in relationship to his own instrument, had taken to holding the bow fixed between his legs as he played the violin upside down on top of it, as if masturbating blank-faced and joyless. For me, the move to scholarly metafiction as a mode of expression was in large part the result of a feeling that had been surging in me for some years prior, where my own relationship to the production of straightlaced scholarly footnotes came to feel something like that violinist’s relationship to his instrument must have been.

This then is the cynical account of the turn: I was bored. But the more hopeful account, the one that would start to take hold of my imagination and to push out the boredom as I advanced in the project over the course of many months, is that it was only in pursuing this project that I was able to apprehend the deep sense of the line in Matthew Arnold’s poem, “Empedocles on Etna”: “He fables, yet speaks truth.” The deeper we pushed into the creation of verisimilitude for some historically embedded yet fully contrived figure or event, the more it seemed that we were literally producing history, engaging in the poïesis of history, or, if you like, re-synthesizing the sundered legacies of history as “story” and history as “wie es eigentlich gewesen [ist]”.

If this sounds too rapturous, let me just reel it back a bit and say that I’ve learned first-hand how much you have to know, how confident you have to be in your knowledge, in order to insert a fictional character into the non-fictional world of, say, the Gulag or the nineteenth-century Midwestern carnival circuit. And it seems to me that such creative work of historiographical metafiction, or creating plausible hoax versions of seventeenth-century philosophy texts, or creating fictitious documents that seem as if they could only have emerged from particular local contexts or lost traditions of the past, could be a very good way of helping some textual element to survive and thrive in humanistic education in the era of PhotoShop and auto-fill.

Textuality, in short, can still have a place in humanistic education, but the potentials of the art of composition should be expanded, or perhaps reversed, to include imitation that pushes up to the limit of the forgery. Working in this vein, one discovers that imitation is not necessarily stifling and limiting, but indeed can be a significant conduit to both understanding and creativity, to both artisanal skill and intellectual mastery: in sum, it is work that exercises the whole of the mind.

It is moreover entirely normal and predictable that at the end of a technology’s history as a useful and necessary tool, it should mutate into something that is sooner riffed upon in creative ways than faithfully reproduced. Thus new strains of —generally cringe-worthy— “book art” emerge at the historical moment of the book’s demise, and thus too in general the numerous historical instances across various cultures of the transition of various practical technologies to a decorative or ceremonial function once they are no longer strictly speaking needed (e.g., the traditions of heraldry that emerged from an initial urgency of distinguishing your enemy’s shields from those of your comrades on the battlefield).

We’re going to have to get creative, in other words, as humanists and pedagogues. The arms-race against plagiarism is futile. It cannot be won, and I for one am not going to spend the rest of my career running written work through anti-plagiarism software when that written work itself has already been run through anti-plagiarism-detection software, and so on. The conception of authorship on which our current anti-plagiarism conventions were built is a conception that does not reflect our current technological reality, and that does not preserve or cherish any vision of the humanities capable of bearing them, alive, into our uncertain future.

About the Author

Justin E. H. Smith is an author and professor of philosophy in the Department of History and Philosophy of Science at the University of Paris. The Internet Is Not What You Think It Is, will appear in 2022 from Princeton University Press.

Publication Rights

This essay was first published in Justin E. H. Smith’s Hinternet. Subscribe here. Republished with permission. It wasn’t plagiarised, honest.