Skip to content
Give readers a window on the world. Click to donate.

Translating the Dictionary

By Janet Hendrickson

In her poetry pamphlet Treasure of the Castilian or Spanish Language, published by New Directions, Janet Hendrickson experimentally translates Sebastián de Covarrubias's four-hundred-year-old Spanish dictionary. In the essay below, Hendrickson describes the process of bringing Covarrubias's idiosyncratic entries into English, both erasing and translating his work to create a new poetic text.

For several years, I translated a dictionary. It was the first monolingual dictionary of a European language, the Treasure of the Castilian or Spanish Language by Sebastián de Covarrubias, originally published in 1611. When I tell this to English speakers, they often seem puzzled or amused. When I tell this to Spanish speakers, they often say brightly, oh, Covarrubias. He’s something of a Samuel Johnson, I sometimes explain to English speakers as a cultural translation, except that Covarrubias came first, which might make Johnson the translation, and to me Covarrubias is a more interesting writer. His Treasure is copious, encyclopedic, filled with citations from authority, personal anecdotes, popular refrains, etymological and theological speculation. It stretches to more than 1,500 pages in the version I worked from; the entry for “elephant” alone is twelve pages. The translation is unconventional in that I also erased most of the book to make a series of prose poems from it, in a volume much shorter than the original: forty-eight pages of text. The English Treasure is small, materially, published as a poetry pamphlet with a stapled binding and a pale pink cover. Its appeal as an object lies in its slender portability; to me, though, it has felt very big. For a time, working through this dictionary was one of the practices that governed my life.

My process was this: I translated as I read the dictionary, choosing only the parts that were interesting to me—sentences, fragments, and sometimes a single phrase extracted from an entry. My rule was to follow the order of the original text.1 Sometimes I translated the Spanish with unnecessarily archaic terms; sometimes I opted for contemporary, overly literal equivalents, approaching the intended meaning at a slant. For example, the tesoro of the original title became “treasure” rather than “treasury” or “thesaurus,” choices more appropriate to the genre. My method was doubly faithful: to Covarrubias, his words and practices of citation and collection, and to my experience of encounter with him. My aim was to contain and be contained by him: to represent him, although not as summary or digest. Rather, I sought to make a new poetic object from the material of his book.

I never was quite sure what to call this project. Some friends, particularly Spanish-speaking ones, suggested I sign my name as author and call it an original work. My editor, in discussions about this, suggested tags that could have clearly signaled my intervention: “transcreated,” “overdrafts,” after Basil Bunting, “translated, selected, and erased.” I had originally pitched and conceptualized the project simply as translation—the term, paired with “selection,” that we ultimately chose. By naming and first thinking of this project as translation, I wanted to comment, through its form, on what translation is. All translation defines, erases, errs, selects, invents; sole authorship is a false prerequisite to place something in the category of art. But these arguments are easy to make and, among translators, trite. There’s also the fact that, while I normally work with contemporary writers with whom I communicate directly, I never can know or talk with Covarrubias. The work I did was impossible without him; it was made from him; it is him and is not. And so I want to talk about this work—this specific work, and translation’s work—through my relationship to words and silence, to faith and faithfulness, to the people I translate.


Words and silence: fidelity to the source text is not singular, but rather involves competing fidelities to the enmeshed parts of that text—to meaning, syntactical logic, rhythm, and sound. Covarrubias, like a translator, is concerned with an elusive original: in his case, not a source text, but the sources of the Spanish language itself. The stated aim of his Treasure was etymological: to trace the origins, classical and vernacular, of contemporary Spanish. More materially, however, the relationship between sound and writing is central, if not to Covarrubias’s investigations, then to the linguistic regulation framing his project. Covarrubias writes in a preface, “No one should be scandalized that the words in this, my book, are written as they sound, without maintaining their proper spelling, since this will be amended immediately in the same discourse: I give as an example Philipo, which should not be sought under Ph, but rather F.” Although Covarrubias was inconsistent in applying this principle, he was not the first or only Spanish writer to advocate for phonetic over etymological spellings.  The author of the first Spanish grammar, Antonio Nebrija, declared in 1492, “[W]e must write as we pronounce, and pronounce as we write, because if things were any other way, the letters would have been found in vain.” To write as we pronounce, words are written as they sound, the letters found: the standardization of spelling implies a shared linguistic identity, a common way of speaking; agency belongs to the speaking public or even the language itself.

When I write that I was faithful to my reading experience, I’m referring less to my reading of the sound of the Spanish alphabet than to the pause I sometimes felt in apprehension. I still let Spanish sound determine the architecture of the book: I titled each lexical poem, that is, each entry, with the Spanish headword and the English following in parentheses, and I retained the alphabetical order of the Spanish. In terms of fidelity to meaning, however, and the ways words mean, I wanted to hold my intermittent experience, as a non-native speaker of Spanish (or any seventeenth-century language), of some words’ strangeness. With deliberate inconsistency, my translation maintains the gap between the reading and comprehension of words or phrases whose luster, as I saw it, might have dulled with an English choice that emphasized transmission of meaning or sought to echo an imagined effect of the original. I translated idiomatic expressions literally; sometimes I turned to early modern bilingual dictionaries to find archaic stand-ins; sometimes I translated words through their contemporary, rather than seventeenth-century, usage. As long as an “equivalent” could be found in a dictionary, or a chain of thesauruses and dictionaries, I considered that English term a fair choice.

The first two pages from the letter A of Covarrubias's dictionary, with the sentences and phrases Hendrickson translated in yellow.

Yet I was also faithful, or inconsistently faithful, not just to words, but to an author who at once charmed and was indifferent to me. There is something unstandard about Covarrubias’s work, something unbounded: however much he remained a stranger to me, he was there. Covarrubias is evident in the text in a way that contemporary lexicographers are not, whose personalities (though not always their biases), as a rule, are sublimated into the standardized prose of their collaboratively authored objects. Covarrubias writes as he thinks: his Treasure is an essay, if an essay shows the mind at work; it is a poem, if one understands poesis as the making it lays bare. Formally, the dictionary reads as a diary, whose entries, by their length, mark the amount of time the author had or allowed himself to give to any subject. Covarrubias apparently worked in alphabetical order and, as he explains in some entries, was ill as he wrote, and he feared he might not live to finish the text. Individual entries and the aggregate space devoted to each letter, as a rule, grow progressively shorter after the longest letter, C.

The variation in prolixity throughout the dictionary, however, also reveals Covarrubias’s associative, accumulative, memory-driven ways of gathering thought. A few excerpts from the entry for candela, candle, demonstrate his method. Covarrubias ends this entry, which extends for a page and a half, with several failed efforts to conclude. Turning from a historical discussion on the use of candles in funerary rites, he states:

And because my rule is not to deal with materials ad longum, but rather only the etymologies of words and that which is necessary to illustrate them, I will not go on any longer about this matter or others that offer me the opportunity to do so at each step, because this work would be long, and my daring would be so great in the desire to address and comprehend in a single volume that which professors in every faculty have written in many, that I neither have talent for it, nor can I guarantee such a long life, that, for better or worse, could put an end to this work.

However, rather than quit the entry as promised, Covarrubias turns away from the thought of his own death and resumes his previous discussion on funereal candles:

I close with the devotion and holy use of burning candles in the burials of the faithful, for whom, lacking men to carry the candles, angels are substituted to carry out this ministry in their place . . .

After briefly citing an example of angels burning lamps in the prison cell of the sixth-century Visigothic martyr Hermengild, Covarrubias moves on to define a verb that shares a root with candle:

Encandilar, dazzle, to perturb the sight of he who comes before us and blind him. From there dazzle has meant to deceive with words, and dazzler, the notoriety of a bad, deceitful female.

This conclusion—or perhaps extension—of the entry demonstrates the tensions between completion and comprehensiveness, between Covarrubias’s ambition and the limitations imposed by his mortality. It also shows a disconnect between words and knowledge: the light that should illuminate blinds, and speech deceives. To write as one pronounces implies a faithfulness only to the material form of language, matching letters to sound, rather than meaning to reality. To pronounce does not promise a fidelity to truth.

Left, the title page of Covarrubias's original dictionary (public domain via Wikimedia Commons). Right, the cover of Janet Hendrickson's poetry pamphlet.

I write as I pronounce; I write as I hear: I usually misremember Covarrubias’s (or through him, Nebrija’s) statement as one of these two first-person singular formulations. Translation can be a false memory: morally right, or so it sometimes seems, even or especially (to those favoring the target language) when factually wrong, asserting an independent life. I first encountered Covarrubias while writing a graduate seminar paper on the relationship between narration and knowledge in an early picaresque novel, Lazarillo de Tormes, whose anonymous authorship has sometimes been attributed to Covarrubias’s father. The professor leading the seminar suggested I consult Covarrubias for historically contemporaneous definitions of the two key verbs of my inquiry, “to know” and “to believe.” I looked up the relevant entries but was distracted by the prose beyond them. The pleasures of reading doubly—for the useful and for the unexpected—brought about this translation project. I think now, however, that my academic inquiry into knowing and believing also anticipated this work with Covarrubias, particularly as the experiment foregrounds the knowledge and concealment occasioned by this and any translation’s competing fidelities.

As I teach my language students every semester, Spanish has two verbs for “to know”: saber and conocer. As these verbs show, knowledge can consist of apprehension of reality independent of the self, or it can be generated relationally. Saber is associated with objective fact. According to Covarrubias, “it means to understand and deal with wisdom. To know what one wants, to have news of something.” Conocer, for its part, connotes familiarity gained through firsthand experience. The virtues associated with this verb in Covarrubias’s definition tie less to the intellect than to the body and its relationship with the soul, to social bonds and moral choice:

To know one’s sin, to confess it. To know a woman carnally. To know about a case, to be a judge in it . . . Desconocido, ignorant, the ingrate who has lost the knowledge and memory of the good received.

This kind of knowledge requires the existence of multiple people: the woman whose body one knows with one’s body; those whose generosity one forgets. Conocer demands that one recognize a social or moral reality and then make something out of it, often with words: a statement, as in a confession, a judgment.

While saber signals objective certainty, conocer is an act of earned trust: trust in others to reveal themselves, in one’s own ability to perceive them correctly. To believe, meanwhile—creer—is “to assent with that we do not understand or feel, the proper act of faith.” Belief excludes sensory perception and, by definition, proof: “In human things, those who place little faith in others have this refrain: ‘Seeing is believing,’ which, strictly speaking, is not believing.” Covarrubias concludes this entry with a list of words built around creer as a root that signify its opposite: “Credulity, incredulous, incredible, disbelieve”—a found poem.

The Cuenca Cathedral, where Covarrubias is buried, and where he served as schoolmaster and choir director during his life. Photo © Janet Hendrickson.

To believe, or rather, belief, like saber and conocer, also has a partner term, faith, which, through its related word “fidelity,” remains coupled to translation theory and practice, despite translators’ insistence on its limited prescriptive and descriptive usefulness. In contemporary English, faith, a noun without a corresponding verb, might seem external to the believer. Faith is something one has, and to possess it can seem an individual choice: the term calls to mind religious conversion, for instance. In Covarrubias’s definition, however, faith implies an interpersonal relationship, like conocer. He writes:

Sometimes faith means promise, as in “I give my faith and my word”; other times fidelity, as in “I have faith in so-and-so,” which means I trust him; sometimes it means belief, as in “I have faith in what so-and-so told me.” Other times conscience: “So-and-so is in good faith” [. . .] Other meanings are hope or credence, as in “With faith that you would take this as good, I did this in your name.”

Faith here largely consists of trust in someone to act or communicate on another’s behalf. Interestingly, in English as well as Spanish, the burden of faith is often communicated through the part of speech the root word takes—noun or adjective—and the preposition that follows it: to have faith in someone means she might betray that faith; to be faithful to someone places the responsibility on you.

My translation of the entry for “faith” consists of a single phrase that follows the excerpt above: “the testimony that one is alive.” In context, this phrase defines a bureaucratic expression. Here is the full sentence in which it appears, followed by the entry’s conclusion: “Faith of life, the testimony that one is alive: an ordinary diligence to charge a periodical pension. And many other phrases that do not come to me at present. Finally faith is one of the three theological virtues, sine qua impossible est placere Deo”—without which it is impossible to please God. After further citation, Covarrubias concludes the entry with an incomplete list: “Infused faith, acquired faith, etc.” This passage demonstrates the style by which Covarrubias earned my trust in his companionship—that is, why I liked him, and why, for a time, I gave my writing to his thinking: a phrase lyrically suggestive when extracted, a first-person acknowledgment of his limitations, spontaneous thought, accumulative logic.

Translation is rarely discussed through the term “faith,” but rather through variants on its adjectival form: faithful, fidelity. Covarrubias defines fiel, faithful, as “he who keeps faith and loyalty, he who deals with the truth and does not deceive the other. With regard to weight, we call faith balance.” As I mentioned earlier, the principal fidelities in my translation experiments with Covarrubias were double: to his text and my reading experience. The tension between these fidelities, however, does not mean equilibrium: placed on opposite ends of a balance, neither my book nor Covarrubias’s could, at least materially, level the other’s weight. A traditional notion of faithfulness to the original text might in fact be faithful to the reader who expects to find it reproduced—in this case, a reader who might be deceived by my translation’s partial truths or truths competing with the original. Covarrubias does not need me to speak on his behalf. But given the connection between faith and trust in others, I might ask, perhaps, what kind of trust I have in Covarrubias when I take most of him out. And what do I know, what and whom does a reader know, conocer, through that unconventional trust?


Translation is a kind of knowing: it is a way of listening, over the course of many drafts, to what another says, and saying that other back in one’s own words, listening to oneself. I write as I hear; I write as I pronounce. The listening to the original text, through the process of translation, takes place as listening must: by silencing one’s own voice. That silence, I want to qualify, is not repressive but generative. One speaks in turn.

Here is another excerpt of a definition from Covarrubias, from tesoro, the keyword of the title: “a lurking corner, a hiding place, where money, gold or silver, pearls and jewels and like things were hidden so long ago that they left no memory or trace, nor knowledge of the person to whom they belonged; it follows that whoever finds the treasure has a right to it.” One reason I did not claim the translation, however original, as original writing is because there is knowledge of the person to whom its source belonged, especially in Spanish; because Covarrubias’s words belong to many; and also because there is not much knowledge of Covarrubias among English readers, and English already dominates more than enough. Translation is a kind of communal writing, unfenced land. The translator resides with the source text not as a settler, but a guest, where others have passed and will pass, and from which, changed, she will move on. One must retain a knowledge and memory of the good received.

1. As a note, my original was not the original, that is, the first published edition of the text, but a 2006 version published by Iberoamericana-Vervuert that integrated the 1611 Treasure with a Supplement, unpublished until 2001, that Covarrubias was working on up to his death. This version of the Treasure also modernized spelling and regularized alphabetization of the entries. 

Related Reading:

"Na Dem Dey Hunt Whales": Onyinye Uwolloh's New Moby-Dick in Nigerian Pidgin

Translating the Ancient Female Voice as Queer

On Sestinas and Literary Translation

Published Mar 15, 2021   Copyright 2021 Janet Hendrickson

Leave Your Comment

comments powered by Disqus
Like what you read? Help WWB bring you the best new writing from around the world.