WHAT DOES A TRANSLATOR DO ALL DAY? Dr. Cathy Flick March 1993 John Woolman Enterprises cathyf@earlham.edu /\/\/\/\/\/\/\/\/\/\/\ Placed here with the author's gracious permission \/\/\/\/\/\/\/\/\/\/\/ Here are excerpts from some General Notes I provided for some inexperienced Russian-to-English scientific translators, to accompany detailed individual critiques of their work I prepared for a potential employer. This provides some idea about my own approach to the translation process. Although the focus is on Russian-to-English translation, much of the commentary is applicable to translation from other source languages as well. The translators in question were not native speakers of English, and the results of their efforts were certainly a graphic illustration of the universal professional advice to only translate into your native language as well as the crucial importance of access to the scientific literature in the target language (in this case, English). MY OWN TRANSLATION PROCEDURES Let me first explain how I would do a translation myself. I will detail the kinds of stylistic changes that must be considered when converting Russian to English, as well as the kinds of terminology errors that can arise. When I do a translation myself, I follow this procedure: 1. I first rapidly scan the original article. I look for names and specialized terms for which I will need to decide the proper English version; these are usually not in the bilingual dictionaries, or else there are several possibilities indicated for each in the dictionary (I will have to decide which one is appropriate in the given context). Sometimes I need to check my Russian science textbooks and dictionaries in order to clarify the meaning of the term. I will need to look at English sources (textbooks, abstracts, journals) to decide the proper term to use in English. I also check the bibliography for items for which I will need to find English versions of non-Russian names given in Cyrillic or to resolve other difficulties. I first look in my immediate resources (terminology/name card files, textbooks, computer-indexed old translation files) for resolutions to such problems. Any problem terms or names that I cannot find here are marked for future library research. 2. I then start typing the translation on the computer, looking at the Russian original while typing in English. If I am very familiar with the subject matter and do not run into more problem terms, I can probably type the first draft of a 1000-word Russian article within an hour. Usually about one-third of the time is spent actually typing, the rest "thinking". Less familiar material will require more "thinking time". Occasionally I will need to check a word in the dictionary or look in an English thesaurus if I am not satisfied with the dictionary offerings for the word. I make up more terminology cards for any specialized terms (not found in the initial scan) that must be checked in English sources either here in the office or in the library. 3. If after checking my immediate resources I still have problems to solve, then I do library research before doing the final editing. You can see that my pay per hour for each translation fluctuates considerably, depending on how many problems I need to solve and how easy (or difficult) it is to solve them. Very difficult and unfamiliar material may require general reading in the subject matter before even starting the draft, in order to get an idea of how specialists in the field talk about it. For an individual client, I would probably ask for help from other translators I know or experts in the field (including the client, if appropriate) before giving up entirely on a translation problem and providing a "translator's note". 4. Finally, I carefully check the translation draft against the Russian original for accuracy and make any necessary changes. This generally takes me about an hour for every 10-20 pages of English, typed double-spaced (about 2000-4000 Russian words). While I am doing this, I am also making stylistic changes in order to make the translation read as smoothly as possible. I also read the English straight-through again in hard-copy after this, making more corrections or changes as necessary. My goal is to make it sound as much as possible as though it were written by a native English speaker. Since Russian and English have very important differences, this might involve any of the following: --adding appropriate definite/indefinite articles and similar constructions (Russian and English handle such concepts quite differently, and often there are several choices here); --changing the tense (e.g., sometimes the Russian present tense should be changed to the English past tense or vice versa in order to sound natural, especially in scientific writing; the imperative is also good to use sometimes, particularly when describing procedures); --changing the voice (the passive voice is very natural in Russian, but often awkward in English); --changing prepositional phrases to adjectival forms (an English reader finds too many prepositional phrases in a row difficult to follow, but can tolerate incredibly long strings of adjectives); --adding nouns in place of pronouns (without the cues of gender and case endings, the English reader sometimes needs help in finding antecedents!); --changing person (e.g., even when there is only a single author, good scientific writing in English generally uses the first person plural in the active voice -- the "imperial we" -- in situations when the Russian often prefers third person passive); --eliminating unnecessary wordiness (sometimes good English requires a single word when the Russian uses several); --making necessary rearrangements for clarity, such as bringing subject and predicate closer together in long sentences, sometimes even breaking down very long or complex sentences into two or more simpler English sentences (English readers, with fewer linguistic cues, are easily confused....); --changing "noun" constructions to "verb" constructions (e.g., Russian often uses constructions like "the analysis of the spectrum was performed", which is better written as "we analyzed the spectrum" or "the spectrum was analyzed" in English); --making changes in the punctuation and use of parentheses (these are important ways to help an English-speaking reader navigate through complex sentences, again because we lack many grammatical cues indicating the function of words in the sentence). These stylistic changes take less time than it sounds, since I do most of it quickly in my head while I am typing the first draft. After 14 years as a professional translator, the rearrangements are almost "instinctive" to me. WHY DO I TRANSLATE ONLY INTO ENGLISH? Very few people are truly "bilingual", i.e., equally fluent reading, writing, speaking, and understanding spoken communications in two different languages. This is why professional translators worldwide recommend that the target language of our translations be our native languages. And even someone who is "bilingual" with respect to ordinary language is unlikely to be "scientifically bilingual". Let me explain the reason for this from my own experiences when translating into a language other than my native language. Then I will explain the kinds of errors in terminology found in your translations, giving a few examples. I am a native speaker of American English and have a B.S. in chemistry, an M.A. in physics, and a Ph.D. in chemical physics. I have also taught the entire undergraduate physics curriculum, ranging from introductory physics to quantum mechanics. I can speak French well enough for simple conversational purposes (if I've had some recent practice!). I can read both general and scientific French without much use of a dictionary and can understand spoken French in radio broadcasts. So while working on my doctorate, several times I was asked by colleagues to translate French articles in my research area into English. Even though I had no prior training in translation (my French classes were conducted entirely in French), I was able to do such informal translations (not for publication!) easily, essentially as rapidly as I was able to write the words down on paper. But several times I was also asked to write French abstracts for journal articles written by my own research group and others. This was very difficult for me. Even French scientific terminology was difficult to choose correctly (I am sure I made the same kind of errors I saw in your translations), although I certainly had no trouble understanding such terminology and translating it correctly into my native language when I read French articles. There are three reasons for this difference. First of all, there is simply a huge difference between reading a language and writing a language, even when discussing everyday topics. Different areas of the brain are actually involved; "recognition" memory is significantly different from the more active "usage" memory. Secondly, I had very little access to the French literature. In technical translation, it is important to have access to "models" of good writing in the target language which illustrate the proper usage of terms and phrases. Thirdly, it takes many years to become a competent scientist and technical writer in any language, which includes learning the specialized language of science. I had many years of training in the English version of scientific language and was immersed in an English- speaking scientific environment. I would need to have the same kind of experiences in French in order to easily do good, reliable scientific translations into French. TRANSLATION ERRORS Every translator (myself included) occasionally makes mistakes or unwise choices when translating ordinary words. We all even sometimes make mistakes when using scientific terminology. That's why we all benefit from an editor! ("Two heads are better than one.") But it is important for the professional translator to develop an ability to decide when scientific terminology needs to be verified in the English literature, and then to either verify the usage or tell the editor/client that the term is uncertain but all available resources have been exhausted. Dictionaries are simply not sufficient to resolve many terminology problems. This is where the native with extensive access to the English literature and coursework in English usually has a tremendous advantage over a non-native. There were various types of terminology errors I found in your translations. Let me call these first-degree, second-degree, and third- degree errors in descending order of seriousness. They all must be corrected prior to publication, of course. First-degree errors are errors in scientific terminology that involve very basic terminology in the field. For example, an English-speaking first-year college chemistry student would not say "single-wedge element cell" when talking about crystal structure. The correct term is "monoclinic unit cell", and you would find this in any introductory science textbook (chemistry, physics) that had even just a small section on crystallography. Likewise for "bond angle" rather than the Russianism "valence angle"-- although I admit that the British still seem to use "valence angle" on occasion, it sounds very "quaint" and obsolete to an American ear (and US usage is the standard for these journals). And an introductory organic chemistry text would tell you that the proper term is "equatorial plane", not "equator plane". Second-degree errors are errors in more specialized terminology. These might be terms that, even with my extensive interdisciplinary background, I probably would need to check myself in the current English literature unless I were already very familiar with the field. A specialist reader might understand an incorrect guess (and possibly be amused), but a non-specialist reader would be very misled! This might cause problems if they try to do a keyword search for more references... For example: "Solvationally divided ion pair" is an understandable guess, based on the Russian term, but actually we say "solvent-separated ion pair". (Yes, I had to look it up myself when I first encountered it in translation work.) Likewise, we say "cage effect", not "cell effect" in solution chemistry. Just about any English paper in the field using quantum mechanical methods would tell you it is the 4-31G basis, not the 4-31GF basis (this very commonly used basis set uses Gaussian functions; the Russian uses "GF", the English uses only "G"). It would be a little harder to track down the full correct name of this particular version: "split-valence shell 4-31G basis", not "valently split basis 4-31GF". But a look at any English paper or textbook on crystallography would tell you that we talk about "Fourier difference syntheses", not "Furieux differential syntheses". Third-degree errors involve the nuances of usage and are not so absolute. Often there are several conventions in use, so there may be a remote possibility that the "error" is acceptable, although uncommon. But when editing your work, I was obligated to change anything that sounds "wrong" to something that I am sure is "right", especially in view of the number of definite errors in terminology in your translations. It would not be cost-effective for me to spend extra time looking for possible verification of a usage I think is dubious. For example, in my experience an English-speaking writer is unlikely to say "stratified structure" when talking about crystal structure at the atomic level; in 14 years, I have only seen one instance where this was an appropriate translation in such a context. We do use the word "stratified" when talking about rock formations in geology or when talking about separation of immiscible liquids. But we would definitely say "layered structure" or possibly "laminar structure" when talking about crystal structure in the context of these papers. "Stratified" in this context would be understood, but would sound odd and "foreign". Other nuances: We generally say "chelate ring", not "chelate cycle", even though we do say "heterocycle" and "macrocycle" and "cyclic". We use "ring closure" and "cyclization" interchangeably. Stylistic nuances: An American scientist is very likely to say that a certain process is "energetically favorable", but very unlikely to say it is "not energetically of any advantage" in a journal article. It should be clear that the "ordinary" words in your translation must also be carefully scrutinized. The dictionaries give many possibilities, and deciding which one to use in a particular context is not always easy even for a native English speaker. Sometimes none of the dictionary offerings is really suitable, which is why my thesaurus is so well-used. But the problem of scientific terminology is undoubtedly the most crucial one. Any good editor can "fix" your ordinary language. But if your editor cannot rely on your terminology, then he or she must spend time verifying every unfamiliar term or sending out queries to other translators for help. This greatly adds to the cost (in time and money) of the editing process. HOW TO AVOID TRANSLATION ERRORS When I am translating an article on an unfamiliar subject, I often need to look at both the current and reference literature in order to learn common usage. Even an article in a familiar area may very well contain some difficult passages (especially the introduction, where the authors are explaining the reasons they chose the research topic). General note: Do NOT automatically trust abstracts or translations of non-English articles! Translators do make serious mistakes which unfortunately get published. I trust a translated book only if I know that the translator or technical editor is definitely an active expert in the field and has a native command of scientific English (this could include, for example, non-natives who are working in an English-speaking country; my best technical editor was a Japanese-born scientist who was teaching at a university in the US). By "active" I mean he or she is doing research, publishing papers, going to conferences, etc. Also be cautious when using English articles by non-native speakers of English unless they are associated with an English-speaking institution -- they sometimes use non-standard terminology. However, it might be safe if they are very well known internationally in the field and actually originated the theory, technique, or device under discussion. The second-degree and third-degree errors are probably the most difficult to avoid, even for experienced native translators. Experience is a real advantage, however. I do much less library research now than I did ten years ago, because I have already worked out so many problems of this sort through reading the related literature and have accumulated thousands of term cards. (The best advice I was ever given as a beginning translator was: "Write down all your translation problems and their solutions on index cards-- then file them!" I keep all my research notes on these cards, which is often helpful when I encounter the term again.) A good translator has a sensitive "internal alarm system" for all three types of potential errors, and will mark the term or phrase for further investigation. Often just looking at one or two English papers, a standard textbook, or various scientific encyclopedias resolves the problem of "how to say it". A very important way to avoid errors is to simply carefully proofread your finished translation. Compare each sentence with the Russian original. Then read the English again to make sure it really makes sense. Some of your translations were obviously not proofread after typing, or else you would have caught some simple mistakes.