Contraction and expansion in translated text
A while back I took part in a blog discussion that was attempting to establish a matrix for translation expansion and contraction by language pair. The secondary discussion in the thread interested me more: although the word count grows in many foreign languages when translated from English, it never seems to shrink in the other direction. Why?
At the time, I gave two main reasons: “I attribute it to the fact that translators, feeling tied to the source text, don’t always pare down enough, sometimes due to speed (it’s easier and faster to include all the words you see in the source), but mostly because they don’t feel they have the right not to reflect all of the words of the source by economizing into more geniune-sounding English.”
I realize now that the second explanation betrays my background in legal translation. Erring on the literal rendering of a text, even if the outcome is a bit stiff, has always been the best bet for law firms in my opinion. This is not to say legal translators do not need to get creative. They do. But it would not occur to them to report to a client “sorry, this just doesn’t work in English,” as it may to an advertising translator for whom the job is to convey the concept of an advertising campaign as if it had been created for an entirely different audience. The legal translator represents; the advertising translator recreates. Technical and literary translation have their own modes, too.
But this is off the topic of word count. Suffice it to say that someone whose translation will be used by an attorney does not have the same right to economize his language as others, thus a word count matrix would need to take into account not only language pair but audience.
8 responses so far ↓
1 B. Cleary // Apr 14, 2008 at 3:14 pm
I can’t speak with any authority about other language pairs, but for Chinese>English, word count certainly should shrink when going into English, and it should grow with translations in the opposite direction, for a couple of simple reasons:
1) Where the English lexicon relies on the aggregation of syllables to build longer individual vocabulary units, the Chinese language similarly strings its units of meanings (represented in writing by individual Chinese characters) together. However, because Chinese orthography does not employ spaces to separate these groupings, the only practical way to quantify volume is by counting characters. Thus, when translating into Chinese, the more five-dollar words an English text contains, the more Chinese characters the target text will have, and thus the greater the expansion rate will be. Conversely, when translating from Chinese, the more multi-character technical terms or otherwise complex terms occur in a Chinese source text, the lower the corresponding word count will be in English.
2) Because of the high degree of homophony among individual characters in Chinese, in order for it to be comprehensible, the vernacular language (and the modern written language upon which it is based) contains many compound expressions, even for words that in English are very basic monosyllables.
To complicate matters even further, historical/literary texts produced before the institution of the modern vernacular Chinese movement are written in what is commonly referred to as Literary Chinese, which, due to reasons of morphology and syntax, is very terse and in translation often requires the addition of a great many English words to render comprehensible.
Thus, in addition to audience, for translations from Chinese into English, any matrix would also have to take into account subject matter, style, and possibly other factors.
2 Administrator // Apr 15, 2008 at 7:54 am
B., thanks for adding some precision here. This is great. Please keep them coming. I’m more familiar with Romance languages than Asian. Chinese translators have taught me the rule of dividing characters by 1.5 or 1.6, but without further explanation of context
For English into Romance, it tends to be from 10 percent to 20 percent expansion. This is partly because Romance langauges are more nominal than English. We put all the meaning in the verbs. For example, in English, “he implemented the new program” would become something like, “he carried out the implementation of the new program” when translated into French.
Going into English, translators sometimes retain this wordiness, thus the lack of contraction. The second sentence above is grammatically correct, yet really French syntax with English words.
3 Kerilyn Sappington // Apr 16, 2008 at 4:22 pm
To piggyback on B.’s very instructive comments, I would add that many agencies use a 1.8 - 2.0 conversion rate for Japanese, which unlike Chinese is a combination of characters and syllabaries, and then go on to apply the same formula to Chinese. I divide the total number of Chinese characters by 1.3 when asked to estimate total English word count. That’s the factor I learned years ago as a Project Manager at Berlitz.
Also, the Chinese-to-English ratio varies by subject. In chemistry, for example, one English word can be 7-10 Chinese characters. Kinda cuts down on your profits for what is often highly technical material.
4 Glenn // Apr 17, 2008 at 12:10 pm
Kerilyn,
Thanks for refining the word count formulas. Chinese characters divided by 1.3 would result in a much higher English wordcount than from Japanese, therefore. Most translators I know of both language pairs hover around the 1.5 mark. Establishing better matrices by language pair and subject would be helpful, especially in the legal realm where we rarely get editable source files and always quote on the target wc.
Thanks again for reading!
5 Gandhi M. // Apr 18, 2008 at 8:58 am
I recently heard Russian - English translation pair expands both directions. How is it possible? I don’t get it. Can someone explain to me?
6 Glenn // Apr 18, 2008 at 9:28 am
Gandhi,
Thanks for reading. I have heard the same thing. Most of my experience is in going into English–because that is the standard fare of law firms– and I have really only seen Russian expand into English, sometimes quite a bit. Let’s see if we can get a Russian expert to weigh in.
7 David Russi // Apr 22, 2008 at 6:00 am
An inexperienced translator working into English from a romance language, say Spanish or Italian, might be tempted to pepper the text with unneeded prepositions .
8 Nominalizations de-noun-ced // May 19, 2008 at 8:33 am
[...] an issue I had touched on in my post about language contraction in English, is also the subject of a few of Wayne Schiess’ recent [...]
Leave a Comment