A word is nothing but a metaphor for an object or, in some cases, for another word.

This aspect of
language is admirably shown in the third part of Gulliver's Travels, where Dean Swift decribes a "project"
at the school of languages in the Academy of Lagado, whereby, in order to do away with the bothersome
intermediary of words, people will avail themselves of the objects in question, carried in a sack and
brought out when the thing otherwise would have been named, thus avoiding the necessity of words
and the dangerous nuances entailed therein. Following this notion of Swift's, then, we can see that a
word in translation is at two removes from the object under description. The word "dog" and the word
perro may conjure up a like image in the mind of the Englishman and the mind of the Spaniard, but other
subliminal images may accompany the two versions and thereby give the two words further differences
beyond sound; the Portuguese word cão is closer to the Latin root that gives us "canine," opening up our
minds to broader connotations.

Following the example above and showing that, like words, no two metaphors are alike regardless of
similarity, we can take the case of the reader's past experience with dogs: one may have had a delightful
pet in childhood and, therefore, is warmed by the word as he comes across it, while another may have
been bitten by a vicious cur at the same period in life and will get a chill or a feeling of fright. Continuing
on doggedly, we must also take cultural differences into account. Among some peoples, Muslims, for ex-
ample, the dog is considered a vile creature, worthy of a swift kick, while others, notably those of
northern Europe, dote on him. So that "dog" can never translate perro in all of its hidden senses. A more
succinct example is the fact that cocks do not crow alike in the ears of different peoples: an American
rooster sings "cock-a- doodle-doo," but carry him to Mexico and he will say qui-qui ri-qui

In light of the above, then, translation is really what we might call transformation. It is a form of
adaptation, making the new metaphor fit the original metaphor, and in a bad translation the results can
be most procrustean. Jorge Luis Borges had a fine sense of how words are used and of their Swiftian
limitations when he told his translator not to write what he said but what he wanted to say. In this case
Borges was taking advantage of a Spanish idiom in order to produce (if we may approach his
terminology) a bifur- cated sense: the expression quiero decir in Spanish literally means "I want to say"
(sometimes carried to "I am trying to say"), but it has come to have the idiomatic sense of I mean. What
the Argentinian author was doing was stressing the inadequacy of words as we strive for some platonical
form of expression that would be more practical than the Lagadian solution. Looking at the English equiv
alent of quiero decir, "I mean," we must remember how often we use just that phrase to correct or
reword what we are "trying to say" ("What I mean is...," etc.).

More deadly even than personal and cultural nuances in hindering an "exact" translation is the very
sound of languages and the words that constitute them. We have already seen how the crow of the cock
differs, so it is quite natural that the names of objects should receive varying sounds. This makes for
extreme difficulty in the translation of poetry, as might be imagined, especially when rhyme is involved.
The rooster has shown us that onomatopocia varies from tongue to tongue, and authors will marshal the
very sounds of their language in order to squeeze out its ultimate effects. Shakespeare says, "A drum! a
drum! Macbeth doth come," but this thunderous announcement would bumble along in French as "Un
tambour! un tambour! Macbeth vient." It is obvious that the trans- lator will have to take liberties with
the text in order to preserve the spirit of what Shakespeare "wants to say." The other side of the coin is
seen (or heard) when Verlaine organizes the sounds peculiar to the French language to imitate the wail
of a violin (it sounds more like a cello to me) as he says, "Les sanglots longs des violons de l'automne."
English simply has no matching nasal sounds in words that would convey the meaning, unless we turn to
trom- bones, and then we have changed instruments, although that may be more in keeping with what
translation is all about.

Words and phrases, then, are not just descriptions of the ob- jects or circumstances entailed, but more
often than not denote the spirit involved. Almost as difficult as poetry to render into another language
are curses and oaths. The meanings can be quite different, but the spirit is universally human enough to
be the same. There fore, when we translate a curse, we must look to the feelings behind it and not the
words that go to make it up. In English, when we insuit someone's maternal descent, we call the person
a "son of a bitch," while in Spanish he is an hijo de puta, "son of a whore." The closest in English to this
latter is the archaic "whoreson," which, even if understood, would not arouse much more than a ripple
of indignation. Portuguese leaves it up to the listener's imagination to deduce the impact by simply
saying filho da mãe, "son of your mother," which is patent and obvious, but leaves the door open to all
manner of vile conjectures regarding one's dam. A most com- mon insult in Spanish is cabrón, "cuckold."
There is no really exact equivalent in English, for the word itself would be about as effec- tive as
"whoreson." Indeed, many people would not know what you were talking about.

It should also be noted that Spanish is completely different from English in the way it arrives at the term
for the hapless hus- band. English goes to the European cuckoo, who lays its eggs in other birds' nests
(the American does no such thing; here the cul- prit is the cowbird), while Spanish resorts to the image
of the billy goat. There is irony behind this epither as the goat has traditionally been the symbol of male
sexuality, as portrayed in the figure of satyrs and other creatures. Thus the sex victim is derisively called
by the name of the one who has wronged him. This is in the same spirit as the word dunce, derived from
the name of Duns Scotus, reputed to be the wisest man of his day, or akin to the way we call the village
dimwit Einstein. This whole concept makes it difficult to render into English the numerous conceits found
in Mediterra- nean literature and folklore having to do with horns. Saint Jerome should have been more
alert to the consequences when he nodded in his translation of the Hebrew of the Old Testament and
led Mi- chelangelo to give his Moses a pair of horns for all to see.

In Genesis, Adam is given the delightful privilege of naming the newly created animals. This was also the
inventive chore of the discoverers and explorers of the New World when they came upon flora and fauna
unknown to their philosophies. The Mexican writer Andrés Iduarte used to say that he wanted to be the
first person on the moon simply to be able to name things. The newcomers had recourse to three
different methods in their nomenclature: they could accept the Indian name, in a version usually colored
by their own tongue; they could assign a name that identified the creature as one approaching a known
animal in the Old World; or they could apply an entirely new and descriptive name to the being. We have
many examples of all three methods. Woodchuck, quetzal, and jaguar are examples of the first, but some
Spaniards, when they spied the last for the first time, baptized it rigre, even though they had never been
to India or seen a tiger. The Portuguese, who had been there, were more accurate, calling it onça
(ounce). The Amer ican robin, sometimes pedantically and more accurately called mi- gratory thrush, a
translation of its Latin scientific name (turdus migratorius), which can be counted on to set schoolboys a-
giggling, is really quite different from the European variety. Ex- amples of descriptive names would be
prairie dog and armadillo.

We are what we are called, to such a degree that ever so many adjectives have been coined from proper
names, for example, Chur- chillian, cervantino, balzacien. Without a name we have no identity.
Sometimes a name is what gives an object existence. Bill Klem, for many years dean of National League
umpires, cloquently described his position as creator through nomenclature when he said, "It ain't
nothin' till I say what it is. It ain't a ball, it ain't a strike, it ain't nothin'." In these cases of linguistic
creation the translator must recreate and he must do so wisely and with extreme care. He must know
that tigre can mean "tiger" in English only when the creature is a denizen of the Old World. When a tigre
turns up in Venezuela, it must perforce be rendered as "jaguar" for the sake of accuracy and at the
expense of all the connotations carried in the original misnaming on the part of the discoverers.
Somehow the Venezue lan dictum, "Donde ronca tigre, no hay burro con reumatismo" (Where the jaguar
growls there are no donkeys with rheumatism) would lose its strength if we substituted jaguar for tigre in
Spanish, yet that is what we must do in English. The result is as flat as substituting "groundhog" for
"woodchuck" in "How much wood could a woodchuck chuck if a woodchuck could chuck wood," or akin
to what happened to the name of the brokerage house when Mr. Bean was replaced by Mr. Smith:
Merrill, Lynch, Pierce, Fenner, and Bean went out of tune on its last note when it be came Merrill, Lynch,
Pierce, Fenner, and Smith. Somebody hit a black key.

There are nuances of meaning that sometimes lurk in differing titles for the same creature or object. In
English, "vulture" and "buzzard" are words that describe the same bird, but when these names are
turned into epithets and applied to human beings, they are quite different: an "old buzzard" is quite
another thing from an "old vulture." The first connotes foolish senility, the second rapac ity. In
Portuguese there is a subtle difference between burrice and asneira, activities ascribed to humans as
they are compared to a burro or an asno, two different ways of saying donkey. In English, there is a slight
difference in tone if we call a person an "ass" or a "jackass" (the jenny seems to have escaped any stigma
in this one). In American usage, with the confusion between "ass" and "arse," the term "asshole" (which
Julio Cortázar loved because it had no real equivalent in Spanish) seems to carry more of the spirit of the
beast of burden rather than the anal orifice. Even the English, who normally keep the two terms apart by
proper pronunciation, often adopt the American way in this instance.

In speech and in writing what we do essentially is choose the word (or metaphor) that we think,
sometimes instinctively, best describes or conveys the meaning of what we want to communicate
(Borges again). The author makes his or her choice and puts it down in writing. Along comes the
translator, who must then make another choice, but in a different language and on a different level.
Sometimes the one possible word in the original is faced by several possible translations in the second
language (I eschew "target lan guage" because when I was in the infantry a target was something to
shoot at and, ideally, kill, which does, indeed, often happen in the matter of translation). What could be
set forth as a classic ex- ample between Spanish and English is the word rama. If translated as "branch,"
it can be applied equally to a tree or a bank (fiscal variety), both of which meanings obtain in both
languages. If we are to translate rama as "limb," however, perfectly legitimate if it belongs to a tree, we
have brought into our English version the nuance of an arm or a leg, something the Spanish word does
not contain, and then there is "bough," which can be applied neither to a bank nor to a human
appendage. This would bring on un- wanted ramifications if the Spanish translator is called upon to
translate the line from the nursery rhyme, "When the bough breaks."

It is this matter of choice that bedevils the translator as he seeks to approach the language he is working
from as closely as possible. In certain cases it comes down to such elemental things as articles. I have
always maintained that translation is essentially the closest reading one can possibly give a text. The
translator cannot ignore "lesser" words, but must consider every jot and tittle. In the case of Latin and
Russian, for example, there are no articles. Bringing them into English and many other languages, the
translator must decide what the author "wanted to say." Is it "the dog" or "a dog?" The important
difference between the two concepts (as far as we who work in English are concerned) was left to the
unspoken imagination of the Romans.

In most works of literature the opening line is normally a very pithy one, giving all kinds of directions and
definitions to the work as a whole. Those of us who did Latin in high school with Miss Whitford had to
commit to memory those first ten lines of the Aeneid that start out with "Arma virumque cano" (without
long marks, a sign of our new exalted status as we left Cicero behind). The standard classroom version,
backed up by any number of pon- ies, was "I sing of arms and the man..." Rolfe Humphries says, "Arms
and the man I sing"; Shaw uses it for the title of a play. Then some years later, along came Allen
Mandelbaum (I sing of arms and of a man") and Robert Fitzgerald ("I sing of warfare and a man at war"),
using the indefinite article instead of the traditional definite one. The meaning is clearly different in
these two differing choices. In the one case, Aeneas would seem to be the man, the one anointed by the
gods for his sacred mission to refound Troy as Rome, while in the other he is a man who happened to be
picked by fate to fulfill that high endeavor. Both versions make sense, bur the discrepancy is obvious. If
we accept one or the other, we are making two quite different choices as to the interpretation of
Acneas's position. Is he a demigod (his mother is Venus) above all other men, or is he more human,
Everyman, chosen by lot for this noble destiny? Unfortunately, unlike Borges, Virgil was not around to tell
his translators what he meant or wanted to say. Neither do we have any explanations from his
contemporaries. We can only be envious of his Russian translators, who can follow right along. without
any article problem and not be faced with the difficulties incumbent upon two different poems.

All of this shows us that the process of translation is one of choice. The skill of the translator lies in the
use of instinct or, better, what Ortega y Gasset calls "vital reason," using Alexander's actions at Gordia as
an example of how it works, a kind of acquired instinct, much the way we put on the brakes when a dog
dashes in front of our car. This necessarily human part of choice is illustrated by the often cited report of
what a computer in Japan did when it attempted to translate the phrase "Out of sight, out of mind,"
coming up with the perfectly logical meaning of "Confined to an insane asylum." As we have said before,
however, the past experience of the individual will affect the translator in the same way that it does the
reader. People have a kind of liking for certain words, either from experience and background or by
cultivated preference, as exemplified by the vocabulary of W. C. Fields. This becomes evi- dent as we find
ourselves having to rewrite what we have done because we have favored some words too much. There
might even be some atmospheric influence, for I have found myself using a word on Monday and then
changing it to something else on Tues- day, only to go back to the original choice on Wednesday. This
might be the influence of academe, as those of us who teach have become accustomed to living by MWF
and TuThS (although the S has been long gone for some time now). Perhaps there are MWF words and
TuThS words, with our minds resting and going blank on Sunday.

These incessant changes are the bugbear of the translator. It is my feeling that a translation is never
finished, that it is open and could go on to infinity, like the figure on the old box of Aunt Jem- ima
pancake flour (Aunt Jemima holds a box that shows Aunt Jem ima holding a box that shows Aunt Jemima
holding a box...). The phenomenon in question is doubtless because the choices made in translation are
never as secure as those made by the author. Since we are not writing our own material, we are still
unsure whether or not the word we have used is the best one, either for meaning or for sound or for
ever so many other reasons. I am always distressed when I receive the usually handsome copy of a book
I have translated. I like the dust jacket most of the time and if it is by Knopf the description at the end of
the type used, because when I start to read the thing, on page one already, I start having second
thoughts about word choice and how it would have been so much better had I said this instead of that. I
rarely read these through, only when I must use one as a textbook, because it is simply too upsetting to
run across so many "should-have-beens." Rationally, however, there is really nothing wrong with the
trans- lation and any number of reviewers have allowed so, but that old matter of the proper choice
remains and I am dissatisfied. It is this feeling possibly that lies beneath the need for new translations of
old books every so often while the original text goes on and on in all its glory.

George Steiner speaks of this in his masterful After Babel and Jorge Luis Borges illustrates the matter in
his ficción entitled "Pierre Menard, Author of the Quijote." The fact is that there is a kind of continental
drift that slowly works on language as words wander away from their original spot in the lexicon and
suffer the accretion of subtle new nuances, which, as the authors mentioned above demonstrate, result
from distortions brought about by time and the events that people it. The choice made by an earlier
translator, then, no longer obtains and we must choose again. Through some instinct wrought of genius,
the author's original choices of word and idiom seem to endure. This is in line with my dissatisfaction
with what I have translated, as I mentioned above. When some- thing of my own finally comes out in
print, I am most often rather pleased with it, and I rarely see any need for emendation. The pas sage of
time has removed me from the sense that this is my work, and it seems to be that of someone else. In
the case of my own stuff, I feel that this fellow has done quite well here, this is just the way I would have
put it, while in the case of translation I wonder why the boob said it this way when it would have been so
much better put another way. Perhaps literary translation should be a con- tinuous process, what the
jargon calls "on-going," a labor of Sisy- phus, as it were.

The author who knows his language inside out can be either the easiest or the hardest to translate. If he
has what might be termed a classical style or use of language, that is, if his sense of words is so pure that
as metaphors they approach the object por- trayed most closely, the translator is on his mettle to find
that same closely approaching word in his language. A writer like Gabriel García Márquez has this gift of
language, and he is so exact in his choice of words, getting ever so close to what he wants to say, that,
indeed, it is difficult to make a botch of a translation of his work as he leads you along to a similar
closeness in English of metaphor (word) and object.

My most amazing experience along these lines, being led by the author, occurred, however, with quite a
different kind of writer, one who was most difficult to translate for the same reason that he knew his
language so well, the Cuban poet and novelist José Le- zama Lima. As I worked away at the arduous but
rewarding task of rendering his novel Paradiso into English, I came upon a few lines from two American
poets that Lezama had translated into Spanish, Walt Whitman and Hart Crane. At the moment I did not
have a text of the originals at hand and in order to facilitate the process of translation I rendered the
lines back into English. When I finally did get hold of the original versions I found that I had missed by
only a couple of words in both cases. I attribute this result to the fact that Lezama Lima had made such a
perfect trans- lation of the poets that I would inexorably and of needs have to arrive at a version quite
close to the original. Would that everything I did could have turned out so close.

In most cases Lezama offered more difficult challenges. Like James Joyce, he had such a grasp of the
language that he could see beyond its confines and, since his mind was broader than the lan- guage,
went about inventing neologisms and restructuring the tongue in quite a logical way so as to express
thoughts and feelings that lay beyond the norms of its expression. The translator is thereby put to the
test to expand his own language in order to match what the original is saying. Sometimes, however, he
will find that his language does have an expression or turn of phrase or con- struction that fills the gap
that the adventurous author is trying to take care of in his language. The problem for the translator here
is that his version will be commonplace and will not show the au- thor's newly coined discovery.
Therefore he must lay his standard translation aside and seek something new that will both give the
meaning of what the author wants to say and make it quite clear that something new has been added to
both languages.

This close knowledge of the language works in an inward fash ion as well, and there, too, it defies the
skills of the translator. I know of an outstanding example and one that I really think impos sible to render
into any other language. It is the epigraph that fol- lows the title of the Brazilian João Guimarães Rosa's
novel Grande Sertão: Veredas (absurdly translated into English as The Devil to Pay in the Backlands,
although I don't know what else could have been done with it). The epigraph states, "O diabo na rua no
meio do redemoinho" (The devil in the street in the middle of the whirl- wind). Rosa has put the devil
not only in the middle of the whirl- wind in the street but also in the very word for whirlwind: re-demo-
inho; one of the words for devil (demon) in Portuguese is demo, and there he is in the middle of the
word as well. Thomas Colchic has received a Guggenheim grant to produce a new and proper version of
this great novel and I do not envy him as he faces this particular problem.

Another aspect of a deep knowledge of one's own language is a thoroughgoing familiarity with local
expression and idiom on the part of an author. In many cases this closeness to regional expres sion
makes translation difficult, sometimes impossible when it comes to preserving the flavor of the original.
An example of this is the title of Juan Rulfo's story in El llano en llamas, "Es que somos muy pobres." A
very simple statement to read that becomes impos sible to translate because of that es que. It precludes
a translation as "We're very poor," and "The fact is, we're very poor" would re- move it from the mouth
of the peasant girl who utters the phrase. There are other cases where customs and manners play a
strong role in the formation of words and expressions. Probably the most difficult aspect of translation is
the necessary but often futile at tempt to preserve or convey a cultural milieu and its concomitants
through words. Even within one language different regions pro- duce different nuances and meanings for
the same words. A Puerto Rican in Buenos Aires who innocently announces that he is going to catch a
bus (Voy a coger la guagua) would be arrested as a child molester. Bicha in Portugal is a line, a queue; in
Brazil it is a drag queen.

When a translator is faced with a work dealing with the pampa and its gauchos, he must be wary of
transporting the locale and its inhabitants to the American West. Despite their similarities, the gaucho
and the cowboy are two completely different creatures, and Martin Fierro must never sound like
Trampas. John Wayne never squatted to sip mate, so why should one who does sound like him? I have
found that the only solution for such situations, and it is a difficult one to handle, is to invent, in this
case, a kind of artificial yet authentic-sounding gaucho or rustic speech in English. As I have said above,
when we try to find an existing equivalent we fall ever so short of the mark. I recall asking a Mexican
friend about certain purported Mexicanisms in the Spaniard Ramón del Valle- Inclán's novel of Latin
American dictatorship and revolution, Tir- ano Banderas. I wondered if he knew if any of these
expressions were really Mexican and his reply was no, but they could be. This is what the translator must
attempt in cases like these that go across cultures in their idioms.
The Puerto Rican novel La guaracha del Macho Camacho (Macho Camacho's Beat), by Luis Rafael
Sánchez, offered me many difficulties of a cultural nature. It is probably the most "Puerto Rican" novel
ever written in that not only the story, but the lan- guage itself is so tied to the culture. In the first place,
there was the problem of a title. The word macho, here an epithet (subsequently picked up by a boxer
named Camacho as life continues to imitate art) was no problem as it has become a solid fixture in
English and is used by many people who have no Spanish, to the extent that I have even heard machine
pronounced with a Germanic ch. The guaracha, however, is not as well known as the rhumba or the
samba and might even be confused with cucaracha, "cockroach," which is a derivation. So, with the
author's blessing, I adopted the suggestion put forth by my wife Clementine. Luis Rafael, or Wico, as his
friends call him, liked the idea that the word "beat" has a double meaning. It does a fair job of rendering
guaracha because it has a musical connotation, and it also can imply an itinerary, as in a policeman's
beat, and the novel is peripatetic as it traces a route through San Juan and environs. Implicit also could
be a Keroua- cian sense, but in a broader feeling because we are a generation removed, the aftermath,
as it were. This shift in the direction of the title shows once again that translation is an approach and not
an equivalency and that a word-for-word technique can often render the translation pallid and
ineffective. Faulkner's novel Light in Au- gust has been given in Spanish as Lue en agosto, perfect in a
direct matching of words. What Mr. Will had in mind, however, was the country expression used for a
cow who is "comin' in," expectant, "heavy in June, light in August." The Spanish is far from that meaning
and only preserves the mysterious lyricism of the title.

Sometimes censorship or bowdlerization will affect the trans- lation of a title. A recent example is the
novel South of Nowhere, by the Portuguese writer António Lobo Antunes. The book deals with Portugal's
colonial wars in Africa, which led up to the Apri revolution and the downfall of the post-Salazarian
dictatorship The English title conveys the idea of a faraway and dismal place but the Portuguese is ever
so much more direct and pithy: Os Cia de Judas (Judas's Arses). English has a closely parallel expression
also used quite often in military campaigns, "the asshole of the earth." It would seem, however, that
American sensibilities are still too delicate to see any of the "words" on a dust jacket, even though they
may be acceptable inside. Not only did the Portuguese strike a blow for political freedom, but they have
moved ahead of us in freedom of expression. (At this writing I am faced with a dilemma in the
translation of another novel by Antunes. It is a situation akir to the problem offered by the title of La
guaracha del Macho Ca- macho. Also using a local type of song, Antunes calls the book Fade Alexandrino.
As many, but not enough, people know, the fade is Portugal's national song, and the alexandrine refers
to versification. The problem is still unresolved.)

This matter of choice in translation always leaves the door open to that other possibility. We cannot be
sure of ourselves. Translation is a disturbing craft because there is precious little cer- tainty about what
we are doing, which makes it so difficult in this age of fervent belief and ideology, this age of greed and
screed. To paraphrase Villon in a way that would have suited Montaigne, "Où sont les que sais-je
d'antan?" The translator must be alert to that other possibility (or possibilities), even if it doesn't rise up
and bite him on the buttocks. He must assume the mind of the old Vermon- ter, who always sees that
other side. When asked by the evangelist, "Friend, have you found Jesus?" his perfectly logical reply is,
"God, I didn't know he was lost." The translator can never be sure of himself, he must never be. He must
always be dissatisfied with what he does because ideally, platonically, there is a perfect solu- tion, but he
will never find it. He can never enter into the author's being and even if he could the difference in
languages would pre- clude any exact reproduction. So he must continue to approach, nearer and
nearer, as near as he can, but, like Tantalus, at some practical point he must say ne plus ultra and sink
back down as he considers his work done, if not finished (in all senses of the word).