Classics & English
Brian Earl

Apuleius’ The Golden Ass in Translation and Adaptation

  • Faculty Advisor

    John Schafer

  • Faculty Advisor

    Brian Bouldrey

Published On

May 2015

Originally Published

NURJ 2014-15
Honors Thesis

Cooper Hewitt, Smithsonian Design Museum | PHOTO


This paper explores the linguistic, historical, cultural, and philological problems of translating an ancient Roman novel, Apuleius’ The Golden Ass, for a contemporary American audience. The introductory essay surveys centuries of Western thought on the duties, goals, methods, problems, and dangers of translation, expanding on Dryden’s framework of categorizing a translation as metaphrase, paraphrase, or imitation. The introduction also practically demonstrates how using different translation philosophies leads to significantly different translations, and reflects on which choices are most effective, arguing against Nabokov’s assertion that metaphrase is the only acceptable means of translation. Following the introduction, the paper provides a new translation of the first third of The Golden Ass, targeting an audience of adults who do not have any formal training in the classics. The third section of the paper demonstrates how translations change when targeting different audiences; specifically, I translate brief sections of the novel for three groups: children, academics, and students just beginning to study the classics. The final section of the paper is an adaptation of the same material translated in the second section; instead of translating the material, I retell the story in a contemporary setting, exploring how the novel might have been written if Apuleius had lived in the modern American Midwest.

The content here is an excerpt from the introduction of the paper, examining the duty and methods of translation.

Introduction: On Translating Books 1-4 of The Golden Ass

I. The Duty of the Translator

In my experience, Latin classes often encourage understanding a text in its original language. Translation is a crutch for students who lack fluency, treated almost as a dirty word—we’re not translating Latin, we’re reading it. Nobody cares if I sing of arms and a man, but arma virumque cano—that’s another story. But reading Latin in Latin only succeeds when everyone discussing the text possesses some level of fluency; when we wish to share the text with someone who can’t read the language, we must somehow convey the author’s work in words accessible to the new audience—and for many Latinists, this “somehow” presents a great challenge. We might paraphrase or summarize, offer a stiff and robotic word-by-word recounting, or flounder for any words at all. We might describe the essence of the text or dart around the original meaning, but we almost always fail to give our non-Latinist friends the same experience of reading the original. Often, the best we can do is simply to describe our experience, but hearing that Dido’s lament is moving and powerful cannot compare to the personal experience of being moved by her powerful words.

Even so, sharing an alien text is one of the great joys of learning a foreign language. Just as the astronomy student would share the discovery of a new planet with her friends without using a name like Kepler-20f, or the biologist would explain the fascinating mechanisms by which the body functions in everyday language, so the classicist is eager to communicate the works of Greece and Rome via translation, leaving behind the original Greek and Latin. Translation, in the words of Dante Gabriel Rossetti, “endow[s] a fresh nation, as far as possible, with one more possession of beauty” (65). As Friedrich Schleiermacher states, translation “bring[s] two people together who are . . . totally separated from each other . . . [into the] immediate relationship . . . of author and reader” (39). Translation is about sharing art and beauty—connecting people who are otherwise separated by time and distance.

With this in mind, the translator faces certain duties. How can he best create, present, and preserve this art? Vladimir Nabokov says the translator’s chief duty is to provide as literal a translation as possible, “to reproduce with absolute exactitude the whole text, and nothing but the text” (134). Anything less than this sends Nabokov into “spasms of helpless fury”:

“A schoolboy’s boner is less of a mockery in regard to the ancient masterpiece than its commercial interpretation or poetization. . . . The term “free translation” smacks of knavery and tyranny. It is when the translator sets out to render the “spirit”—not the textual sense—that he begins to traduce his author. The clumsiest literal translation is a thousand times more useful than the prettiest paraphrase. (127)”

I—and more importantly, scholars and translators who are much more knowledgeable and skilled than me—disagree. Octavio Paz, for instance, argues that “literal translation . . . is not translation” (154). But why? Why is Nabokov incorrect? Wouldn’t a translator want to aim for “absolute exactitude”?

No: among other reasons, literal translations will inevitably fail to capture the mood and tone of an author in a way that is accessible to the new audience.

Mood and tone—the experience of reading something—define a text; rendering this “spirit” of a text is a necessity, not a denigration. Readers will remember few, if any, specific words and phrases after completing a novel, but they will recall how they felt, whether they laughed or cried or recoiled in disgust. John Dryden, who prepared a translation of Ovid’s Epistles in 1680, proposes that “the sense of an author, generally speaking, is to be sacred and inviolable,” and this must be preserved in translation above all else (21). It is not the individual words that matter so much as the feelings they convey. While a literal translation is often sufficient for these purposes, the literal translation is not an end in itself, and when it fails the translator must know how to appropriately depart from the original text. Dryden expands on these thoughts:

“Words are not like landmarks, so sacred as never to be removed; customs are changed, and even statutes are silently repealed, when the reason ceases for which they were enacted. As for the other part of the argument, that [the author’s] thoughts will lose their original beauty by the innovation of words; in the first place, not only their beauty, but their being is lost, where they are no longer understood . . . I grant that something must be lost . . . in all translations; but the sense will remain, which would otherwise be lost, or at least maimed, when it is scarce intelligible. (28)”

The translator must modernize the text at the expense of individual words; failure to do so will result in the greater losses of beauty, intelligibility, and experiencing the text in the way it was meant to be experienced. Walter Benjamin agrees with Dryden’s sentiments, arguing that a translation should produce an “echo of the original” (77). Translation is not copying or recreation; it is creation in a new form. An echo is not speech, but something entirely different, sound waves reflected off a distant surface—yet an echo can have the same effect on the ears it touches.

Nearly as important as preserving the author’s sense is preserving the original images, whether in plain narrative or in figurative language. The translator should not add to or embellish the author’s imagery any more than he should subdue or replace it. Why speak of the setting sun when the author describes the sky growing darker and darker? Why say someone walked like a wounded deer when the original says someone tottered unsteadily? In Reading Rilke, William Gass compares fifteen different translations of Rilke’s writing. He scathingly criticizes the translators (including himself) when they fail to preserve Rilke’s images (67). At the same time, he interrogates the English translations for connotations that might subvert the imagery; the word “terrifying,” for instance, has different colors than the German schrecklich. Gass goes to great lengths to preserve as much of the original as he can—sense, mood, tone, image; he deliberates about each and every word in his translation.

In this translation of The Golden Ass I prescribe to all these duties—conveying the text faithfully, preserving the sense and images of the original, presenting the author so that the modern audience appreciates him as much as his ancient audience did. My chief aim, however, is to create an artistic product. “The hallmark of bad translations,” writes Benjamin, is the tendency to impart only information without striving for any new artistic ends (71). Paz, likewise, describes translation as “a literary operation” (157). And so, I want my translation to stand on its own, to provide pleasure to the reader, to be my own small contribution to what the powers-that-be call “art.”

II. The Methods of Translation

No translation can exactly recreate the mood and meaning of a literary work, but it can approximate it. How? It is not just a language we must translate when working with Latin—it is an entire culture. Separated by an ocean and two thousand years, how can we recreate a text so a modern American audience will appreciate it? The translator’s most important skill is his ability to make careful, informed decisions. Lydia Davis succinctly quips, “No choice is simple, even one that seems simple” (62). The translator must be able to recognize the sort of decisions he makes—even the instinctive ones—and articulate and justify each and every choice.

Dryden describes the various strategies for approaching a translation project. “All translation,” he writes, “may be reduced to . . . three heads.” These heads are:

“metaphrase, or turning an author word by word, and line by line, from one language into another . . . paraphrase, or translation with latitude, where the author is kept in view by the translator, so as never to be lost, but his words are not so strictly followed as his sense; and that too is admitted to be amplified, but not altered . . . [and] imitation, where the translator (if now he has not lost that name) assumes the liberty, not only to vary from the words and sense, but to forsake them both as he sees occasion . . . taking only some general hints from the original. (17)”

These are the translator’s tools: metaphrase, paraphrase, and imitation. They are three distinct vehicles that transport the reader from one place to another—like automobiles, ferries, and trains. Transferring from one to another in a single trip can be inconvenient, frustrating, or downright impossible, and so it is better to pick a mode and stick with it. Schleiermacher, who favors metaphrase, writes that imitation should never be combined with the other methods, for doing so turns the entire translation into “mere imitation, or to a still more repulsively conspicuous and confusing mixture of translation and imitation that throws the reader mercilessly back and forth like a ball” (51).

However, sometimes a journey necessitates merging methods. Just as someone driving from Milwaukee to Detroit will either need to take a long and circuitous route through Illinois—with potentially brutal, unforgiving traffic in Chicago—or else take an expensive but relaxing ferry across Lake Michigan, some passages will prevent translators with similar dilemmas. A direct translation (metaphrase) might be clunky, confusing, or arduous to read, when a paraphrase presents a smoother, more elegant route, but at the cost of the original language. In such instances, the translator might choose to mix methods rather than religiously sticking with one, just as the driver might decide the easiest journey involves paying the extra fare for the ferry. Indeed, in many cases metaphrase and paraphrase complement each other quite well.

Like automobiles, ferries, and trains, metaphrase, paraphrase, and imitation each have advantages and disadvantages; depending on the goal of a particular translation, one method may prove more effective than the others. Metaphrase provides the most “literal” translation of a text—every word in the original will have something that corresponds to it in the translation, and there will be no additions. This allows the reader to understand what the original text “says” more than any other method, and the reader will be able to draw their own conclusions about the text free from interpretive actions from the translator. So why not exclusively use metaphrase?

One problem lies in that what is idiomatic in a foreign language might not make sense in the new tongue. Some idioms transfer perfectly—the Latin verb replicare (to turn/fold back/over) can be used with thoughts; in English, we can also say “I turned these thoughts over.” Other idioms, however, do not translate well—Latin speakers would say there was a manus (hand) of robbers, but in English, we would say a “group” or a “band.” In some instances of idiomatic trouble, the translator can integrate any method of translation fairly seamlessly into his work. A metaphrase will reproduce the original text faithfully, though the phrase might strike the readers’ ear as odd or nonsensical (e.g., the Danish use the idiom “to take off the clogs”); a paraphrase would describe or explain the idiom (“to die”); and an imitation would choose an idiom from the target language to replace the original (“to kick the bucket”).

The problems of metaphrase extend beyond figures of speech—a translation based solely on metaphrase will struggle to capture the mood and tone of the original. Roman authors and American authors have different styles of writing. As an inflected language, Latin has much more flexibility regarding where words can fall in sentences. This allows authors to place subjects, objects, and verbs in various positions to add or subtract emphasis. Literary Latin—including that of Apuleius—is marked by long, twisting sentences, using copious nested relative clauses; subjects, verbs, and objects are often separated by a surprising amount of space. On the other hand, American English favors short, clear, concise sentences, which Davis points out in her essay on translating Proust, whose verbosity rivals that of Apuleius (58). As Edward Seidensticker states, “An English sentence hastens to the main point and for the most part lets the qualifications [adjectives, relative clauses, etc.] follow after” (143). Latin does not do this, sometimes beginning with descriptions before revealing the subject, and often delaying the verb until the very end of a clause or sentence. While a modern writer can mimic some of Latin’s emphatic choices, such word order will be unusual in English, making the sentence stand out compared to its counterpart in the original. The translator will inevitably modernize and normalize the word order of sentences; retaining the original word order would be unsustainable for more than a sentence or two, reading like utter gibberish.

When faced with cumbersomely long sentences, the metaphraser has two options: break up the sentences into smaller chunks to adapt to a more modern style; or retain the old sentence length at the risk of leaving the reader lost and confused, needing to reread passages over and over again to track the flow of meaning. The Romans were accustomed to their characteristically long sentences, and so they would not have had much trouble following them, keeping track of subjects, verbs, and other crucial information. So, if the metaphraser chooses to retain original sentence divisions, the reader will have a much more difficult time following the text than the original audience (though to be fair, even the Romans would have had to stop, think about, and reread the most convoluted Latin sentences). Benjamin warns, “A literal rendering of the syntax…is a direct threat to comprehensibility” (79). However, if the metaphraser chooses to break up the long sentences into more digestible chunks, he sacrifices fidelity to the original. In either case, the metaphraser will be able to translate what the text means, but not the exact experience of reading it.

Paraphrase, then, attempts to address these shortcomings of metaphrase. The translator concedes the inevitable sacrifices to the original, and errs on the side of making the text accessible to the new audience. What does a modern American care if the Romans used long sentences? Confusingly long sentences could ruin the comic timing in a text like The Golden Ass. Paraphrase strives to achieve the cadence, flow, and idiom of the target language, capturing the ideas and tone of the original, but often at the cost of not translating particular words. Garry Wills, introducing Christopher Logue’s translation of the Iliad, explains that poets paraphrase or “‘take liberties’ . . . not to get away from” the original, but to move toward it (xi). And translators will find themselves using paraphrase in nearly any translation; as Dryden points out, “every language is so full of its own properties, that what is beautiful in one, is often barbarous, nay sometimes nonsense, in another . . . It is impossible to render all those little ornaments of speech in any two languages” (21, 30). Therefore, Dryden concludes, “it would be unreasonable to limit a translator to the narrow compass of his author’s words: ‘tis enough if he choose out some expression which does not vitiate the sense” (21). Paraphrase fills the gaps between languages that metaphrase creates. The two modes can be used in conjunction, one supplementing the other to create a smooth, accessible, and (mostly) faithful translation.

Imitation wanders far from the original text, and might preserve only themes, mood, tone, and content. There will not be much similarity on the level of language—the imitator will use different diction, different sentences, different paragraphs. Whole scenes may change. Until one examines an imitation and an original broadly on a thematic level, the only similarity between the texts might be the title—if that. Dryden nicely sums up the goal of the imitator: “to write like one who has written before him, on the same subject; that is, not to translate his words, or to be confined to his sense, but only to set him as a pattern, and to write, as he supposes that author would have done, had he lived in our age, and in our country” (19). Thus, adaptation—transporting a text to a new environment or medium, retelling the story in a new setting or form—falls under imitation’s umbrella.

Beyond these three main tools, Schleiermacher introduces two different philosophies a translator can employ. The translator can try to make the translation of the foreign language sound as much like the target tongue as possible, or he can push the limits of his native language, mimicking the original, trying to force his own language into the foreign mold (42). The decision between these two philosophies will affect many subsequent choices the translator makes. For example, a translator trying to completely adapt to the target language will often sacrifice a writer’s particular style if such a style feels unnatural in the new language.

In the case of this translation, I have attempted to make English sound as Apuleius-esque as I can. Apuleius has a distinct loquacity and even makes up his own words—as a result of imitating this style, my translation might feel oddly verbose at times, a tad foreign to the ear, but “transmitting a feeling of foreignness,” as Schleiermacher puts it, is only an advantage (46). Infusing the work with a sort of exoticism will remind the reader they are reading a foreign text, aiding their imaginations in picturing a milieu in another country, characters speaking another language, a culture with customs different from their own. Benjamin also supports this mode, sharing a quote from Rudolf Panwitz: “The basic error of the translator is that he preserves the state in which his own language happens to be instead of allowing his language to be powerfully affected by the foreign tongue” (81).

Incidentally, metaphrase is the tool that most naturally allows the translator to push his words toward the original language. Translating word-by-word rather than idea-by-idea not only ensures everything written remains accounted for, but ensures that it is accounted for in a way similar to that in which it was originally presented. Schleiermacher writes, “The more closely the translation follows the phrases of the original, the more foreign it will strike the reader” (46). And so, metaphrase allows the translator to be most faithful to the original both in literal meaning and in preserving the way in which things are said—a way that might seem unnatural to the new audience.

But that is not to say using metaphrase to push the boundaries of language to something foreign is the “best” way to translate; the purpose and audience will guide the translator just as much as—if not more than—a general philosophy on translation. Metaphrase, paraphrase, and imitation each belong in every translator’s arsenal. The translator must be conscious of his choices and aware of the effect each tool has, permitting him to make appropriate decisions to achieve his particular goal.

I intend my translation to be what I call an “accessible metaphrase.” I strive to translate nearly every word of the original Latin, but to do so in such a way that the result does not sound like a translation. I do not confine myself to Apuleius’ sentence divisions or word order. I frequently break up his sentences and paragraphs into more digestible chunks (and on a few occasions, merge his sentences). When the main subject or verb is delayed for a long period of time, I move this crucial information to the beginning of the sentence; while in some cases it is delayed to provide a surprise or a punch line, in English such a delay would render the sentence incomprehensible. I make a few additions, inserting background information necessary for understanding some of Apuleius’ more obscure comparisons and allusions. Occasionally Apuleius will repeat an idea with synonyms for emphasis, but English cannot always replicate such repetition in a smooth and elegant manner. In cases like these I omit a few Latin words from my translation. When metaphrase fails, I paraphrase, but I try to do so sparingly.

A final tool of translation is to collaborate with other people and other books. Translators would be foolish not to use the many resources available to them. For example, in the introduction to his translation of the Iliad, Christopher Logue recalls that when he began his decades-long project, he was not confident in his Ancient Greek; he relied upon older translations as guides (vii). Similarly, I regularly consulted the translations of Sarah Ruden and P. G. Walsh when working on this project, and I also read Robert Graves’s. Donald Frame supports this practice; he argues that translation is “a cumulative undertaking, and therefore borrowing—or stealing—whenever you see that your own best solution to a problem is clearly inferior to someone else’s” is perfectly acceptable (82). The exception, however, is the “rare cases” where a translator improves on the original (83). Using previous translations can help decipher difficult passages in the original and be a source of inspiration in deciding how to render a certain phrase.

Meanwhile, Davis takes a slightly different approach; she prefers to separate her work from her predecessors, at least when writing her first draft. She refuses to “look at other translations,” to read the whole work in the original language, or to research the author’s life (54). Essentially, she translates blind, but she will then use all these resources when she revises her translation on second and subsequent drafts. This allows her to create something that is original without squandering the work done before her; she writes, “Often another version made me confirm that mine felt right to me, or it induced me to make it better. Sometimes I found a word I hadn’t thought of using: in Moncrieff, say, ‘housetop’ for ‘roof’” (56). Her method ensures the “freshest” take on a work, as free as possible from outside influences.

Finally, translators can collaborate by discussing their translation project with other people. William Weaver makes it a point to consult the original author or people who knew him as part of his translation process (122-3). Even with ancient texts, when such consultations are impossible without a time machine, discussing the translation with others—from those who know the foreign language well to skilled writers in the target tongue—can be invaluable.

Download full thesis

Brian Earl (’14) is the recipient of the Joseph Clyde Murley Prize for Excellence in Honors Thesis Research (2014). He currently works as a technical writer at Epic and enjoys having fun in the outdoors during his free time.