Translation Theory, Practice, and Transduction

Or: Reflections on Working Between Arabic and Standard Average European Languages

[NB. These comments were delivered at the recent workshop, Towards an Arab Left Reader. Convened and organized by Profs. Mezna Qato and Chana Morgenstern at Newnham College, Cambridge University, one purpose of the workshop was to address issues that came up in the translation of key texts from the Arab Left.]

It’s daunting to speak on the subject of translation to this audience. Everyone of us in this room translates, whether or not we identify primarily as translators. For myself, I have come to prefer the clunky term of translation practitioner, since it suggests that such activity is practice: which means it’s never quite done and it’s never quite perfect. The word practitioner also speaks to knowledge gained through experience and repetition. Which is the most relevant kind of knowledge, I think, that there is on the subject of translation. 

And for this simple reason, I am not sure I have much to add to what you know by discoursing on the subject. Before I am accused of creating a naïve and false distinction between theory and practice, let me acknowledge what we also all know: that there is no such thing as proceeding without theory. All of us in this room have, to different degrees, read and given thought to questions of meaning and interpretation. Which is to say, for us to have gotten as far as we have in this room, our thinking is informed by the philosophy of language, semiotics, literary theory, postcolonial studies and so on. 

My point here in drawing attention to practice is to underscore something I have learned from experience that it is the practice of rendering texts from one language into another that teaches us the most profound lessons about translation in the sense that concerns us at this workshop. 

It is through the painstaking process of finding, or never finding, the right words and expressions, that we begin to understand source and target languages. In the act of comparing senses, seeking equivalent phrases, and acknowledging—in frustration or glee—irreconcilable differences, we sometimes make discoveries about language writ large, or meaning, or what it’s like to be language-bound humans. In comparison, the lessons of translation theory seem to me much more humble with regards to what we are talking about here. Indeed, by now the very term “translation” has taken on so much metaphysical baggage that it rarely describes the actual material activities of people laboring between languages.

In US academic conversations the study of translation is dominated by questions of literature. This is a problem—and, as a side note, let me say that you are fortunate here in the UK to have real disciplinary and methodological diversity in the study of translation. My discipline, Comparative Literature, has long been in the thrall of its own translation studies. Even if much of this scholarship studies the role of translation in colonial and postcolonial contexts, much of it operates within older, unacknowledged Eurocentric and Christian notions of salvation and communion. As Mona Baker and I (and others) have written with regard to our present, violent moment, much translation activity is military in nature and it is not about human understanding, nor is it in the service of political solidarity.[1] Translation theory in the literary disciplines has been largely blind to this history, choosing instead to focus on a marginal slice of translation activity—literary translation—and imbuing it with a spirit of spiritual transcendence that is taken directly from the early Church. 

I don’t want to rehash this point any longer, since it’s not relevant to the project at hand, which really is one of political solidarity and human understanding. Just because most translation in our world is wielded as a weapon of empire, does not mean that all translation must be. And even if translation must be a weapon, we too know how to arm ourselves. 

What I want to say here is not a political pronouncement on translation as part of anti-imperialist, anti-capitalist struggles and solidarities, as important as that might be. Instead, I want to clear away some of the discursive debris that gets in the way of us doing what we are here to do: which is to take texts that were composed in different cultural and linguistic contexts, for audiences situated in different historical circumstances, dreaming different dreams, and to render them—across time, space, language and culture—into something else: in our case, legible Anglophone texts that speak meaningfully to contemporary audiences. 

I would like to make some points, but they are all might be distilled into two. First, there are longstanding theoretical debates about translation. If I were a proper scholar, I would have already invoked the work of Apter, Benjamin, and Spivak. And now I have. We have all seen battles where charges of infidelity or untranslatability were hurled as if their ground were clear. I would like to suggest that, as compelling as these theoretical debates can be, they are often unhelpful when engaging in particular translation activities. They may be helpful for conceptualizing such activities generally or for commenting on them, but it may be that they do not help us engage in them. So, what I want to say to say to anyone of you who finds those debates unhelpful: you may ignore them as you work, as long as you develop your own tools for reflecting on the work you do. Conversely, to those who find them useful, please ignore me and count yourself lucky.

Second, if I had to describe the feeling I most often experience while engaged in translation, it would be one of constraint and feeling stuck. I know that I am not alone: someone once described translation as trying to ride two galloping horses at the same time, without being crushed under hoof. While I’ve never experienced such equestrian excitement while working on written texts, I do like the metaphor. To translate sometimes really does feel like being trapped between very imperfect options. Given this, I have found it necessary in my personal practice, to look for any choices I have at my disposal. So, that’s my second big note: to remember that we have choices when we translate. There is never solely one way to do it.

I had this spirit of choice in mind when I replied to the invitation to suggest a short reading for the group.[2] The 60-year-old essay by Jean-Paul Vinay and Jean Darbelnet I selected does something very important for me: it gets us away from thinking of “translation” in the metaphysical sense—the translation of philosophers—and gets us into the small procedures and strategies we necessarily and routinely employ when rendering text between languages. What Vinay and Darbelnet sketch out for us is a spectrum of choices—nothing more. But this is already a lot. They describe a range of direct procedures, from borrowing and calque to literal translation; and a range of oblique procedures, from transposition to modulation to equivalence to adaptation. Together, they provide a technical vocabulary for most of the techniques employed in the translations of this workshop.

But if you read the essay, you will have noticed that what they describe does not map easily onto Arabic-English translation. For instance, in the center of Vinay and Darbelnet’s spectrum sits literal translation. In many ways, this is the standard by which most translations are judged to be faithful or not. Unfortunately, it is a technique that the Arabic-English translator will rarely, if ever, have the pleasure of using. Why? The reason is simple: since losing most of its inflections, English relies on word order to make meaning. Thus, the most common sentence structure of Arabic—the jumla fa‘liyya—must undergo real transformations when it is rendered into English. There is no literal translation of it. Same with the manṣūbāttamyīzmaf‘ūl matlaqḥālmaf‘ūl fi-hi—or the iḍāfa. Same with the differences between the aspective sense of Arabic verbs and the tense system of English. Rendering the māḍī into the past tense is not always the best choice, nor is the muḍāri‘ fully equivalent to the present tense. 

The point is not that these various differences are untranslatable, as some might say. It is true these language features do not exist in English. Yet, to say “there is no translation for them” is to insist, in a sense, that renditions must be literal. This is like saying that all hands must be left.

More importantly, for those of us working between Arabic and English, it is not an option. Fortunately, there are plenty of techniques other than literal translation for remolding these foreign linguistic shapes with the clay of English. Using Vinay and Darbelnet’s terms, we spend more of our time performing modulation, which pertains to lexical choices, and transposition, which pertains to shifts in grammar and syntax. Sometimes—while translating aphorisms (ḥikim and amthāl) we reach instead for equivalence. Thus, “illī fāt māt” might become “What’s done is done” or “Don’t cry over spilt milk” or something else. We are also heavy borrowers in English, thus terms like “jihadi” and “takfiri” are now effectively English words. 

As I said, in US academia the subject of translation is dominated by questions of literature. To give one example, the endless argument about foreignizing versus domesticating strategies of translation. But there’s really no argument, since there is a broad consensus among translation theorists that to domesticize a source text is to turn it into pabulum. There is the an apparently anti-capitalist version of this stance which states that a domesticized text is a commercial text, a text decontextualized so as to fit the needs of consumer markets made up of lazy, infantilized consumer-readers. And there is an anti-racist or anti-imperialist version: domesticization whitens the source text and is thus accomplice to cultural appropriation. (As an aside: How odd that cultural appropriation is now a crime. It was not that long ago that the Marxist critic Dick Hebdidge developed that term in order to describe the counter-hegemonic practice of bricolage in working-class British subcultures.) 

So, the translation scolds tell us: don’t domesticate, foreignize! Don’t bring the text to the reader; make the reader go to the text. Make the reader work, for in that work they will become alienated—however slightly—from their own linguistic and cultural context. And thus the reader, and the language itself, will grow beyond themselves. Despite my tone, this line of thinking is rich and compelling. It was first articulated 200 years ago by Friedrich Schleiermacher, and then grew—with the German and British romantics—into an aesthetic position that went far beyond the practices of translation. Eventually it became a foundation for the estrangement poetics of literary modernism. 

My point in mentioning estrangement is not to denounce it. But rather to historicize it. And to point out how odd it is that one particular literary technique—that emerged first within one philosophical-literary movements two centuries ago, and then another a century ago—would continue to dominate so much discourse on translation even in our present. Yet, to assert that literature should be modernist in style is like saying that every moment should be 1968.

By the same token, it makes little sense to think that all translation need be literary, or performed under reining literary-critical stricture. Indeed, as I understand it, the project of the Left Arab Reader is not one of literature. The texts we are working on are not necessarily literary in nature (more are not). Our audiences are not necessarily literary audiences, or audiences expecting literary texts. And our purpose, I think, is not necessarily to “move” audiences in aesthetic ways. 

I have been using words like remolding, reworking, changing and transforming as if nothing were lost in the process. Of course something is lost. We can always translate better. We can practice and improve. If we are doing academic translation, we have the luxury of using paratext—we can introduce and comment, we can supply footnotes. (For an excellent example of this, there is Samah Selim’s recent discussion of her experience translating Arwa Salih’s memoir.)[3]

But what about fidelity? What about faithfulness to the original? Those who want to talk about faith are welcome to, but to do so returns us needlessly to a world of Christian metaphysics. Let us talk about works, not faith. Let us look at grammar and sense. Let us investigate context, resonance, and semantic fields. But let us put aside all talk of faith. While we’re at it, let us stop talking about The Word so we can start talking about words.

To conclude, I want to refer to the linguistic anthropologist, Michael Silverstein, whose work has yet to be taken up by translation theorists. In one essay, “Translation, Transduction, Transformation,” Silverstein observes that translation theory has been based primarily on what Whorf called, ‘Standard Average European Languages,’ whose grammatical structures are relatively proximate (and often fully equivalent) and whose cultural contexts are highly overlapping.[4] If the material on which translation is theorized has come from closely-related—or standard average European—languages, which share common cultural rituals and landmarks, then it may not explain the labor that needs to take place when working with more distal languages, such as Arabic. Moreover, Silverstein points out, because such theory is grounded in Saussurian models of denotative language—signifiers that denote signifieds—it fails to describe non-denotative aspects of language. 

Let me give just one example of non-denotative language: the Peircian category of index, which has no place in the Saussurian universe, and hence in literary theories of translation. But it is a common part of language.

To explain, consider the word “bābā,” which appears across a wide number of Arabic dialects. What does it mean? We might translate it as: father, papa, pops, dad, daddy, paterfamilias, and so on—and these capture some, or much of its meaning. We might discuss the differences between it and the MSA forms of “ab” and “abū.” We could discuss register and tone. We could argue whether the word “daddy” can ever be appropriate now that we have heard it on Ivanka’s lips. 

However, these discussions won’t get at critical aspects of the word in usage. Consider how in some dialects “bābā” is the word a child uses to address their biological father, and also, how that father addresses the child, irrespective of their gender. It does not follow, then, that the meaning of the word “bābā” is also “child.” Rather, this usage is pointing—indexically, outside of language—to the relation these people have with one another. By the same token, the word doesn’t mean “parent-child” relationship: rather, it is the sound that people in this relationship use when signaling this relationship to one another. 

The same is true for many other such words: māmā, ‘ammū, giddū, and so on. Moreover, biological kinship does not begin to contain the uses of these words—all of them can be used, affectionately and ironically, by friends, lovers and so on, again, irrespective of gender. To render these usages involves getting into the structures of human relations in the Arab world—and to puzzle with how to express this lived culture in textual terms. Most importantly, to translate compels us to move beyond purely denotative models of signification.

There are other linguistic, but non-denotative, aspects of language that we tangle with all the time in Arabic. For instance, sub-lexical aspects of language that create sense, but not at the level of denotation, or what Peirce called symbol. We all know how, through jinās, the tri-radical Semitic root system can cast shimmering semantic auras across texts. We don’t have jinās in English, though sometimes we use Latin or Greek roots to approximate the effect. And there are also, for example, sub-root sound patterns in Arabic. The pair q-ṭ, for instance, hints at acts of cleaving and paring, which is then elaborated variously in the tri-radical roots q-ṭ-ṭq-ṭ-rq-ṭ-‘q-ṭ-fq-ṭ-m and so on. Similarly the pair gh-m suggests something murky, which is then expanded across a number of tri-radical roots, such as gh-m-m, gh-m-rgh-m-ḍ and gh-m-q, and so on, all of which retain the gesture toward obscurity, albeit with different shades. A deep awareness of these sound patterns can be found in the canons of poetry and prose. These are not cases of onomatopoeic words, though those exist as well. Authors employ such patterns to generate senses that, again, cannot be explained by Saussurian terms. 

These examples of index and symbol are neither extreme nor rare. These are as commonplace in everyday talk as they are in prestige textual traditions. They are present in routine family interactions, public speech, slogans and poetry, religious discourse and profane. 

The point is, that unlike translators from Spanish to French or from French to English, we cannot translate these features without also translating quite a bit of their social and cultural contexts. The term Silverstein uses for this process is transduction, which he describes as a process of conversion. But this is not conversion in the sense of religious faith. Rather, the metaphor is one of energy and dynamics. In processes of transduction, it is energy, not meaning, that is converted. Through transduction, the energy of wind or falling water becomes electricity. Friction occurs, and there is resistance. There are inefficiencies and contingencies. But for all the slippages, there can be sparks and shocks, and sometimes the lights come on. 

I can think of no better metaphor for what we are doing than that. 


[1]   See, for example: Mona Baker, “Interpreters and Translators in the War Zone: Narrated and Narrators,” The Translator 16: 2 (2010), 197-222, and “Narratives of Terrorism and Security: ‘Accurate’ Translations, Suspicious Frames,” Critical Studies on Terrorism, 3: 3 (2010), 347-364; Elliott Colla, “Dragomen and checkpoints,” The Translator 21:2 (2015), 132–153; and Moira Inghilleri, “You Don’t Make War Without Knowing Why: The Decision to Interpret in Iraq,” The Translator 16:2 (2010), 175-96.

[2]   Vinay and Darbelnet, “A Methodology for Translation,” in The Translation Studies Reader, ed. Lawrence Venuti (New York: Taylor and Francis, 2007), 128-137.

[3]   Samah Selim, “Politics and Paratext: On Translating Arwa Salih’s al-Mubtasarun,” Alif: Journal of Comparative Poetics 38 (2018), 180-202.

[4]   Silverstein, “Translation, Transduction, Transformation: Skating ‘Glissando’ on Thin Semiotic Ice,” in Paula G. Rubel and Abraham Rosman, eds., Translating Cultures: Perspectives on Translation and Anthropology (Oxford: Berg, 2003). Thanks to Michael Lucey for introducing me to this essay.