WHAT I LOST WHEN I TRANSLATED JORGE LUIS BORGES
University of Puerto Rico
Some fifteen years ago, after I had been translating for three or four years and thought—a little prematurely it turned out—that I knew what I was doing, I wrote an article in which I made a claim for translation as a certain kind of literary criticism: literary criticism in its text-analysis modality. Basically, I said that in the process of textual analysis that goes into the translation, the first thing and the last thing the translator-cum-literary critic needed to do was look closely at the text, the words on the page, and when that written text had been absorbed, its stylistic peculiarities and tics analyzed and understood, the translator’s job was virtually done, because everything else followed from that. My approach to translation then, and in fact my approach to translation now, is that I attempt to reproduce in the translation the markers of style that I identify in the source text.
In my ignorance and naiveté, it never occurred to me that this approach to literary translation was in any way daring or anti-establishment, certainly not that it ran counter to standard practice. I had come to the work of translation from the literary studies of twenty-five or thirty years ago, and so I knew—heavens, we all knew—that Hemingway’s style was not Faulkner’s or Joyce’s or Virginia Woolf’s or Henry James’s, that de Quincey did not write like Jay McInery or Paul Auster. My reasoning with respect to translation was very clear to me: Why on earth would a person translate these writers, whose voices were so different from one another, without making a good-faith effort to show the reader in that other language, whatever the language might be, why this writer was prized for his or her writing as well as his or her stories? I thought it unfair of me, when translating a literary work, to accommodate the writer’s voice to my own sense of appropriateness or to an English-language readership’s sense of what a “well-written” sentence was. I told myself that that other reader in that other language and culture, reading a translation, needed to be able to detect the writer’s difference or uniqueness, somehow, or I might as well produce a summary of the plot. That was my thinking and frankly, I still like it.
But I discover that I have been theorized into anti-establishmentarianism. In book after book, article after article, anthology introduction after anthology introduction, Lawrence Venuti, for instance, talks about, and bemoans, the dominant Anglo-American translation ideology of what he calls “fluency,” which is a strategy that consists of reducing the hills and valleys and chasms and skyscrapers of the stylistic landscape of the original text into one broad pampas of target-language sameness. In a word, simplification or flattening of the source-language style into “acceptability”, “readability” in the target language. And as I see when I read translations, Venuti may be right, this may be going on. At the yearly conferences of the American Literary Translators Association, translators do talk about making the translation read smoothly, “as though it had been written in the target language”, and that may be a code-phrase for what Venuti calls the homogenization of the source-language style. But while I, too, want my translations to read as though they really had perhaps been written in English, I know that English is a remarkably accommodating language, which has been home to an enormous variety of writers with an enormous variety of styles and approaches to writing. Often, writers don’t write “naturally”, don’t write “smoothly” or “readably”, and that is a conscious decision in virtually every case; therefore, in my view, I as a translator should respect that, and try to teach my readers to respect it, too. I have always, in my literary translations, tried to resist the fluency, defined as “flattening out”, that Venuti has spoken of, and have recently pledged to myself to produce translators’ notes to try to raise the level of recognition of the choices that I and other translators make.
But I don’t want to set myself up as a model to follow, don’t want now, as I seem to have done in my early days, to lay down rules for every translator. All I feel comfortable with saying is that this approach works for me, and it has, I believe, worked for the particular writers I have translated.
From the beginning of my work in translation, I have translated “stylish” writers, writers who have a recognizable voice: Reinaldo Arenas, who is baroque and verbally playful and multi-stylistic, and very musical and rhythmic; Fernando Arrabal, whose style in any given novel will be, if you’ll forgive the apparent redundancy, stylized to an extreme, artificial, not spoken, not natural; Ana Lydia Vega, who uses the vernacular in remarkable ways; Edgardo Rodríguez Juliá, whose Renuncia del Héroe Baltasar uses mock-historical document style, academic lecture style, baroque prose-poetry style, anything but a “natural Spanish” style; and now Borges, who has been called perhaps the greatest Spanish stylist of the twentieth century.
When I went to Arenas with a problem that involved being “faithful” to what was on the page, he would as often as not say to me, “Don’t worry about the words here, catch the meaning in another way, work with the intention behind the meaning, but most of all, catch the rhythm, the word play”. Arenas knew that his brand of the baroque depended more on accumulation and effect, sometimes, than on the precision of a particular word. And his advice certainly kept me from writing “standard English”; I knew I had to respond to, and reenact, what he had done in the Spanish, which was not “standard”. When I was doing La renuncia, I felt that it was important to distinguish between the many styles Edgardo uses: one for the lectures and one for the historical documents and one for the broadsides and one for Bishop Larra and one for Alejandro Juliá Marín. In the historical documents, for example, I spent days with the Oxford English Dictionary, a dictionary of the etymology and history of English, to be sure that none of the words that I wanted to use for the mock-eighteenth-century texts in Spanish had entered English at a later date than that. One of my worst moments was having to invent eighteenth-century circumlocutions for the word “landscape,” which is a concept that Baltazar uses in the documents written by him and quoted by the lecturer in the second lecture but a word that came into English too late for an English Baltazar to have used it. In other words, I felt that I needed to put as much effort into the English, to make it as believable in its own way, as antiquated and even archaic, as Edgardo had done for the Spanish. And I was delighted when the reviewer for the Library Journal said that footnotes would have helped to identify the historical personages and put them into their historical context! She fell for the recreation of the “historical” aspect of the novel, even to the point of failing to remember that it was a novel. I did much the same for Arenas’ El Mundo Alucinante, which is a “rewriting” of the memoirs of a late-eighteenth-century friar; for that, I attempted to recreate the style of Benjamin Franklin in the Autobiography and James Boswell in his Life of Samuel Johnson, to meet Arenas in his own recreation of that dated, breathless, unconsidered, unconsciously hilarious style.
I was very glad to find, as I worked on Borges, that he would have approved of this approach. Because while he is very accepting of every conceivable different style of and approach to translation, in the essay titled “Las Versiones Homéricas” he does offer one cardinal rule: translators should treat those things that are part of the esthetic surface of the text one way—creatively, and with faithfulness to the text’s peculiarities—and those things that belong to the fabric of the language another way—the way one would treat prepositions, for instance. English doesn’t say “depend from” just because Spanish says “depender de”; English doesn’t say “dream with” just because Spanish says “soñar con”; and Spanish doesn’t say “presidir sobre” just because English says “preside over”. Borges’s examples go beyond prepositions, of course, to the tics or “frases hechas” of the language: when Agustín Moreno says “¿Qué hacen todo el santo día?” Borges notes that the holiness of the day comes from the language, not the poet. It would be absurd in English to say anything but “the livelong day” or “all day long” — those are two of the standard idioms. To try to write “holy day” or some other such literally-translated phrase is to import an æsthetic effect or potential thematic issue into the English where none exists in the Spanish. Of course, sometimes those buried or invisible metaphors of the language do need to be taken into account in a translation, one does need to analyze the constellations of metaphors and motifs that are at work in a text to see whether the frases hechas have been (consciously or unconsciously) folded into the mix by the author, but as often as not, for most writers, they are simply part of the machinery that the language has crafted for expressing itself in.
I was gratified, of course, to find that Borges and I shared this what to me seemed eminently common-sense view of translating: let the language be language, in a natural way, and let the writer’s style be the writer’s style.
When I began this project, the collected fictions of Borges, I had known Borges’s work in English, but I had never really read much of him in Spanish, even though I had translated several essays of his in the seventies. (And it’s incredible how ignorance of the larger corpus shaped my decisions back then). When I began to translate him for the collected fictions — or began to read him carefully in Spanish before I started translating — I was struck by certain elements of his style, and reading around in the criticism, the biographies, some writers’ memoirs, I discovered that I was on to something, even if I was fifty years later than the Spanish-language world in discovering it. (Talk about rediscovering the wheel!) But what I saw was interesting from a Translation Studies or reception-studies point of view, because what I saw was that English-language writers and critics always commented with great wonder and admiration on Borges’s themes, the subjects and philosophico-literary treatment of his stories, his playing with genres, whereas Spanish-language writers and critics, especially at the beginning of his or their career, almost invariably commented on another aspect of his work: his style, his prose, his writing itself. Not that the themes and subjects and genre play didn’t startle and waken Spanish-language readers’ imaginations, sometimes even change their lives and art — Carlos Fuentes, for instance, has spoken very movingly about the influence on him of Borges’s subjects and cultural eclecticism1. But to writers and readers in Spanish, the subjects or “stuff” of the fictions was often simply not as shocking, not as disorienting, not as liberating, not as “new”, as the prose itself was. Mario Vargas Llosa, for example, who is no great admirer of Borges the fabulist, has talked about the profound way Borges changed not only writing in Spanish, but the very Spanish language:
Borges’s prose is an anomaly, for in opting for the strictest frugality he deeply disobeys the Spanish language’s natural tendency toward excess. . . . [In] Borges there is always a logical, conceptual level to which all else is subservient. His is a world of clear, pure, and . . . unusual ideas that . . . are expressed in words of great directness and restraint. . . . Borges made a radical innovation in the stylistic tradition of Spanish. By purifying it, by intellectualizing and coloring it in such a personal way, he showed that the language. . . was potentially much richer and more flexible than tradition seemed to indicate . . . (10).
In the light of this kind of appraisal of Borges’s style, this insistence on its newness and, in a way, unconventionality, it was clear to me that if I was to do the kind of translation that I believed in and that Borges himself counseled, I had to be, or become, sensitive to the elements of that style, its deviation from “standard” literary Spanish (which is, as we all know, a pretty decent working definition of the concept “style”). Another thing that struck me as I began to read in and around Borges was the way Borges himself talked about his style. In virtually every one of the prefaces he wrote to his volumes of fictions, he disavows what he calls his earlier, baroque style in favor of a new “plain style”, a notion he borrowed from Kipling. Clearly, Borges himself felt that he was doing something that he, at least, had not done before: he was purifying, streamlining his style, paring it down, trimming away the fat, bringing it out of an earlier complexity into a “plainer” mode. (Not, he said, that it was “simple”, for there was nothing simple about it; it was just not as decorative and/or shocking and relentlessly “avant-garde”). So that became one important rule for me: the prose of the translation was to be as “frugal”, as “direct”, as “restrained”, in Vargas Llosa’s words, or as tight, economical, and efficient, as I saw it, as Borges’s own prose was. As in the Spanish, every word had to carry its own weight. I had to shift gears out of the baroque of the writers I had recently been translating, Reinaldo Arenas and Edgardo Rodríguez Juliá, and into a taut classicism.
But what I needed to guide me in the actual choices I had to make were the details of the style, and there were two things that made me aware of the most remarkable of Borges’s hallmarks: his adjectives. First there was a remarkable sentence by Borges himself in the preface to El Hacedor: “To left and right, absorbed in their waking dream, rows of readers’ momentary profiles in the light of the ‘scholarly lamps’, as a Miltonian displacement of adjectives [hypallage] would have it. I recall having recalled that trope here in the Library once before, and then that other adjective of setting — the Lunario’s ‘arid camel’, and then that hexameter from the Aeneid that employs, and surpasses, the same artifice: Ibant obscuri sola sub nocte per umbran’”. This hypallage, as it is called in English — arid camel, scholarly lamps — was, I realized, everywhere in Borges, for it both opens and closes the fictional corpus. In the first sentence of the first “biography” in A Universal History of Iniquity (1935) we read this: “In 1517, Fray Bartolomé de las Casas, feeling great pity for the Indians who grew worn and lean in the drudging infernos of the Antillean gold mines. . . ”, laboriosos infiernos. And at the end of his career, in one of the last fictions that he wrote, “The Rose of Paracelsus”, in the volume Shakespeare’s Memory, Borges uses this trope twice: fatigado sillón/ “weary chair”, and mano sacrílega/ “sacrilegious hand”. Thus we are presented with a stylistic trait, a fingerprint, that identifies Borges throughout his career. Other clear examples of this technique are una cicatriz rencorosa/ “a vengeful scar”, alcohol pendenciero/belligerent alcohol”, biblioteca ilegible/ “illegible library”, and dentelladas blancas y bruscas/ “brusque, white bites”.
(A brief digression concerning these adjectives and the “transparency” that Lawrence Venuti talks about: I have a “translation stalker”. People talk about the Translation Police (Borges had mentioned the Fact Police with respect to Martín Fierro); this guy is a translation policeman gone bad — he is so furious at translation-malefactors (under his definition of a translation crime) that he has become a bully and a scold. He lives somewhere in Maryland, is clearly well-educated and well-read, and has no patience with “odd” adjectives. He has already sent me three bulging envelopes filled with pages torn from a yellow legal pad; the pages are handwritten, single-spaced, and covered with insulting comments about “my” adjectives: “Whoever heard of ‘belligerent alcohol’— you must have been drinking yourself” or “‘Illegible library’! — It’s the books that are illegible, because of the bad lighting or whatever, not the library! This is yet another example of your translatorese!” Obviously this person believes that the English of a translation should be “fluent”, as Venuti defines that word: perfectly “normal” English, with none of the small shocks and momentos de asombro that the original author may have attempted to achieve in his or her text. Just as obviously, this person didn’t read my Afterword to the Collected Fictions, where I tried to explain this aspect of the prose).
The second part of my awakening to the importance of Borges’s adjectives came in that Vargas Llosa essay that I quoted a second ago. There, Vargas Llosa specifically mentions Borges’s “strikingly original use of adjectives and adverbs”2. That made me realize that I had somehow to deal with the words that Spanish-language readers and commentators had puzzled over for years; I could not simply translate them into invisibility. One of the most famous opening lines in Spanish literature is this: Nadie lo vio desembarcar en la unánime noche: “No one saw him slip from the boat in the unanimous night”. What an odd adjective, “unanimous”. It is so odd, in fact, that one is sorely tempted to put something like “all-encompassing”, so as to make it “comprehensible” to the reader. But it is just as odd in Spanish, as Vargas Llosa has told us, and it clearly responds to Borges’s intention, expressed explicitly in the story “The Immortal” (which I’ll talk about in just a second), to let the Latin root govern the Spanish (and, by extension, English) usage. In “The Dead Man” there is a “splendid” woman: Her red hair glows; indeed, I believe that in Borges, splendid always has either the etymological sense of glowing or the sense only slightly metaphorized from that, of glorious. Somewhere else there are “concave” hands: cupped, of course. And there are many more “odd” examples besides. These two techniques, hypallage and etymologized adjectives and adverbs, are present throughout the entire course of Borges’s career. They are also traits he surely found and recognized in some of the English writers he most admired, and sometimes translated — Emerson and Thoreau and de Quincey and Sir Thomas Browne — all of whom employed words with their etymological force, though none of them were so radically “classical” or “plain” as Borges himself. Indeed, unlike them Borges used the technique of what I’ve called etymologized words as a way of cutting through the baroque, trimming it down, not perpetuating it — as a way of making an efficient writing, packing a great deal of meaning into the story by freighting words with not just dictionary meaning, but their entire historical significance.
Along this line, I began to notice all the other places where Borges talked about the distinguishing characteristics of the particular style of one story or another. Many of Borges’s stories are what I call “found fictions” or “edited fictions”, many are pseudo-translations; they are stories that masquerade as documents discovered by a person who then publishes them or, if the case demands, translates and publishes them. “Brodie’s Report” is one example of this kind of story; perhaps the most famous example, which I’ll talk about in a second, is “The Garden of Forking Paths”. In one particular case, “The Immortal”, the narrator notes that the document, discovered in a volume of Pope’s translation of the Iliad, “teems with Latinisms”, abunda en latinismos. There’s never anything accidental in Borges; if he says the document was found tucked away in a translation, then in my view, he’s saying something about translation, and we know about Pope’s Homer, his Vergilian imitations, how very elevated and eighteenth-century they were. So — Borges wanted this English to “abound in Latinisms”, perhaps as Pope’s had. And I made it do so: my soldiers “magnanimously covet the steel blade”, as they do in the Spanish; my deserts are “vagrant deserts,” to match the Spanish’s difusos desiertos. (I am not a slave of the cognate, however, as one who compares my translation with the original will see; here, I believe that “diffuse deserts” — in English — is too alliterative, that the Spanish does not sound so alliterative as the calqued English would; and therefore, I decided to use a still Latinized but slightly different word than difusos for my deserts).
It hardly seems strange that once I became so enmeshed and engrossed in the words of Borges’s writing, from time to time other lights would come on, some further insights into his stylistic intentions would come to me — i.e., not strange that the text would make a demand on me. For example, often Borges’s stories are purported to be translations from the English, as I noted earlier.
This is the case with “The Garden of Forking Paths”. The part of the story’s brief preface that slapped me awake one day, that gave me pause, reads as follows: “The statement which follows — dictated, reread, and signed by Dr. Yu Tsun, former professor of English in the Hochschule at Tsingtao — throws unexpected light on the case”. This statement was dictated, reread, and signed in England, I realized, during World War I. Thus, it had to have British spelling and use British words for things: “flat”, not apartment; “aeroplane”, not “airplane”, etc. I felt that for the sake of verisimilitude, for the sake of following Borges’s express intentions for the found document, I had to switch from the American English of the rest of the book to a British English that I really wasn’t sure I mastered. We called in a Brit to vet the story, but I have been told, and in the New York Times Sunday book review, no less, that we still missed one Americanism. We did the same thing in other stories where an Englishman or Irishman is talking or writing. What I mean to say by this is that I tried to be attentive not only to the marks of Borges’ style, but also to the hints that he left within the stories about the language that should be used in them.
Another strategy similar to this is related to the issue of Borges’ recognizable vocabulary. I had promised myself that as the single translator of the fictions (this time around, I mean), I would try as hard as I could to give English readers the same sense of words repeated throughout the career as Spanish language readers got. (Heretofore, Borges had had almost twenty separate translators, and they did not consult with one another about which words to use). Those words that Borges used over and over again — laberinto, atroz, tigre, etc. — I attempted to repeat, as well. It was part of my strategy of reenactment. What that meant in practice was that I had to go back and forth, back and forth, through the corpus, jotting down words or remembering them so that I could try to make them ring all down through the years and stories, grateful, too, that I was working in the age of computers. Whenever I had chosen a word for my translation that didn’t work in a later context, I would try to find a mediating word that would work in both, or all three, or all four contexts, through the years. Interestingly, late in the editing stage I ran across a word that I realized I didn’t understand; this was the word profundo as used in an early fiction, “The Dread Redeemer Lazarus Morel”, in A Universal History of Iniquity. The phrase it appeared in was profundo pistolón. I simply could not figure that one out. So I went hunting, and what I discovered was that Borges almost always used the word profundo as an adjective for zaguán, the entrance to the typical Buenos Aires house of the turn of the century. And suddenly I knew what it meant: long and narrow: the long, narrow entryway, the long, narrow pistol used by a Civil War-era scoundrel.
But back to the distinguishing marks of Borges’s style. As I began to edit and revise my translations, I discovered that they seemed choppy to me, that I could never manage to read with any speed, that I kept getting stopped by what were remarkably short sentences, by periods or by the semi-colons that linked otherwise independent clauses together. It has been my experience through the translation of a couple of million words of Spanish or so that Spanish writers do not use many semi-colons; they use commas and conjunctions, or frequently relative pronouns, to link clauses together so they flow. They employ a style filled with compound-complex sentences; they concatenate clauses, pack a sentence with all the baggage it will bear — and then pack in a little more and sit on it. Not Borges. Borges apparently wanted to slow the reader down by using the speed bumps, those policias muertos, of the period and the semicolon. As I began to look more closely, I realized some other things. Borges, of course, as I had known since the beginning, likes parallelism, chiasmus, subtle repetitions-with-variations. He is a very classical writer, in that sense. But what I also realized was that he is a paratactic rather than hypotactic writer, using coordinating conjunctions (and, but, etc.) much more often than subordinating. This is also the style of Whitman, from whom Borges seems to have borrowed those “mismatched catalogs” that he is famous for. And punctuating with the semi-colon correlates with that tendency toward parataxis, for it suppresses or soft-pedals the causal connections that another writer might make with subordinating conjunctions and other sorts of explanations. The suppression frees Borges from having to make explicit how one detail or fact or sentence is related to the other — does a consequently go here? a while? a nonetheless? a because? a despite? — and I believe it adds to the mysteriousness that we sense in some of the statements, the sense that some unexplained or inexplicable thing lies under the surface of this prose. The semi-colon also produces spareness, and a particular, recognizable rhythm. It became clear to me in a way that I had never really analyzed before, but only intuited, that Borges was, in a word, a man who had not just studied but absorbed the rules of classical rhetoric.
(Another parenthetical remark, this time about rhetorical devices: Borges is not a very acoustic writer; he is an intellectual, and not so much sensory, writer. So I was shocked when I thought I discovered a story in which Borges seems to be using onomatopoeia. It is the short fiction “Pedro Salvadores”, in the volume In Praise of Darkness. The story is about a man who for several years hides in the cellar from the dictator Rosa’s vigilante mob or private army; either his eyes become accustomed to the dark or he himself becomes accustomed to his ersatz blindness, and he hears things. See if you hear what I hear. The Spanish runs like this: “Algunos ecos de aquel mundo que le estaba vedado le llegarían desde arriba: los pasos habituales de su mujer, el golpe del brocal y del balde, la pesada lluvia en el patio”. I think I hear, onomatopoetically, the footsteps and the rain above his head. This is the way I rendered it in English: “Now and again, echoes of that world he could not enter would reach him from above: his wife’s footsteps as she went about her routine, the thump of the water pump and the pail, the pelting of rain in the patio”).
This, then, was the experience of translating Borges that I had under my belt when one day I was asked to do a radio interview on the centenary. During the interview, the interviewer, thinking no doubt of that dreadful adage that has been attributed, apocryphally, to Robert Frost, “poetry is what gets lost in translation,” asked me just that: What did my translation lose, what will English-language readers not get from my translation of the stories? Although the impertinence of the question took me a little by surprise, I told him more or less what I’ve told you: English-language readers inevitably lost the shock of the new that Spanish-language readers got the first time they read Borges’s prose. It was not possible to recover that today, that “deep disobedience” of the norms of the Spanish language. But I had, I said, attempted to reproduce the peculiarities of Borges’s Spanish prose in my English. As for the more pointed intention of the question, I declined to answer; the critics can do that. The more interesting question is what I myself lost when I translated Borges.
The three years I spent with the stories were some of the most intense and fascinating years I have ever spent, comparable only to those wonderful years in graduate school when you’re reading all the wonderful literature of the world for the first time. For me, Borges was, as he has been for many people, a veritable textbook on the art of storytelling, the art of spellbinding, and most of all, the art of writing. But my experience of the power of language used at the highest level of consciousness and creativity, the power of language as used by Borges, has had an unforeseen and somewhat sad result: it has made almost every other writer, stylist, pale in comparison. I think it is no insult to say that almost nobody writes as well as Borges did, at the level of language. There are great storytellers, greater probably than Borges, but not many stylists that even come close. I feel honored to have worked with him, and to have had that experience, but the price that must be paid for experience is, as we know, a terrible loss of innocence. Discovering the strategies of Borges’s style, and working to reenact them in English, entailed losing any shred of illusion, any hint of unselfconsciousness about the way the rest of us mortals write. As
Borges himself showed us in so many stories — “The Aleph”, “The Garden of Forking Paths”, “The Gift”, “Blue Tigers”, “Shakespeare’s Memory — a blessing is always a mixed blessing. As Borges noted sadly, he inherited a library, and blindness; we who study Borges inherit great sight, yet the rest of the library somehow fades.
- In a lecture given before the Anglo-Argentine Society, Carlos Fuentes speaks about the influences of the content of Borges’ stories:
Borges . . . enriched our Spanish home with all the imaginable treasure-houses of world literature, East and West, and he permitted us to go forward with a sense of possessing more than we had written, which was all that we had read, from Homer to Milton to Joyce — all of them, probably, and along with Borges, the same blind seer.
Borges attempted a supreme narrative synthesis. In his stories, in order to give us the most complete portrait of what we are — thanks to our memory of what we have been — the literary imagination appropriates unto itself all cultural traditions. [Borges, too, then, was something of a conqueror of other lands!] For example, Spain’s Arab and Jewish heritage, mutilated by royal absolutism, . . . reappeared, marvellously fresh and lively, in Borges’s tales. I certainly would not have had this early, fraternal revelation of my own Arab and Jewish heritage without such stories as “Averroes Search”, “The Zahir”, and “The Approach to al-Mu’tasim” (“The Accidents of Time”, at the Royal Society of Arts, London, November 6, 1990, published in The Borges Tradition, ed. Norman Thomas di Giovanni, London: Constable, 1995, pp. 52-52).
- Mario Vargas Llosa, A Writer’s Reality (Boston, New York: Houghton Mifflin, 1991; originally published by Syracuse University Press, 1991), p. 2; I first met the longer extract quoted below in Borges: A Life, James Woodall (New York: HarperCollins Basic Books, 1996), p. xxvi, though I am quoting slightly different material here.