I recently came across this website, which compares English texts with their translations. It claims to give information on how 100 English words translate to other languages. I can’t attest to the truthfulness of most of the figures here, but as for the Swedish amount given (103), I have to disagree. Generally, an English text becomes shorter, in terms of word counts, in Swedish, because Swedish allows for longer words (created by putting two or more words together into one).
Despite the possible inconclusiveness of the numbers on this site, it relates to an interesting and important issue. Since translators get paid by the source word, should the pay be different depending on which direction a translator works in? How does, or should, this asymmetry in word count affect translators’ fees?
For example, if a Swedish to English translator gets 12 cents per source word, should an English to Swedish translator get a lower amount per word (since s/he will have more source words)? Or should the English to Swedish translator get paid per target word instead? Or is the system fair as it is?
#SundaySentence
18 hours ago
5 comments:
The figures on the website may be botched but text shrinkage and expansion is a well known problem, especially in DTP-intensive formats or those that have size limitations, such as software strings, flash files and even powerpoints.
In my opinion, the most important factor in pricing a translation is the level of difficulty found on the source text. It doesn't matter if it takes 12, 16 or 20 words to translate a difficult sentence that originally had 10. What counts is the amount of time it takes you to translate it.
An alternative route to decreasing your rate according to direction is charging per target word. Here in Brazil it's customary to charge based on target text word count. In fact, some translators turn out translations that are terribly wordy with a view to a larger check.
I don't charge based on target anymore because my direct clients want to know exactly what the final price is going to be. And I avoid translations that are too diffult or charge higher prices for them, whenever possible.
To be exact, the site you quote does not compare words but characters, and a larger number of characters does not necessarily mean a larger number of words. In German, for example, the translation nearly always gets longer (the number of 109/100 seems likely), but the number of words is often lower than English, we, too, have longer words. And - what a coincidence! - German translators very often charge by target lines (i.e. the number of characters in the target text). However, I think the reasons for this date back to pre-computer times when it was easier to count typed lines in a standard format.
Here in the U.S. people mostly charge based on target word. Easier because you often have a hard copy source and then you translate into MS Word and can just do a word count. You base your bid to the client on the estimated number of target words (calculating from the rough number of source words and the expansion rate). In my experience with Swedish to English translation, I've found that most texts expand about 15%, but some texts (especially legal) expand up to 30%! I just charge different rates per source word (e.g., $0.145) and per target word (e.g., $0.12) so that I end up getting paid the same amount no matter how the translation is billed. In my experience, agencies who want to pay you by source word are just trying to pay you less because they know full well about the expansion rates for different language combinations, whereas beginning translators aren't as aware of the phenomenon.
Every translator should know his or her average expansion rate and the general expansion rate for their language direction. An expansion rate is a technical term in translating that simply refers to the ratio between the number of words in the source document and the number of words in the target document.
You can calculate your own expansion rate by taking a few of your own translations and doing detailed word counts of source and target documents, then averaging them.
In general, the following language directions have the following expansion rates:
French -> English: 15%
Spanish -> English: 15%
Dutch -> English: 25%
German -> English: 25%
Swedish -> English: 12%
Different kinds of material may vary. For instance, legal documents tend to have higher expansion rates, whereas bullet-item lists will have low or even negative expansion rates.
As far as pricing per source or per target word: as a professional, you need to know what your expansion rate(s) are and charge so that you get approximately the same amount of money whether you bill by source or by target.
For instance, let's say you translate from German into English, and you usually bill by the target word (which is the standard in the United States). If you charge $0.14 per *target* word, and you have an expansion rate of 25%, then you should be charging $0.175 per *source* word. If your expansion rate is 10%, then you should be charging $0.155, etc.
These two rates should be set this way because, obviously, you want to get the same amount of money for the work whether you charge be source or per target word--and then there is no motivation to be "wordy" if you charge by the target word.
Incidentally, while I'm talking about rates, if your clients expect you to handle DTP-related issues, then you should be charging more: anything beyond translation per se is an added service and should be billed accordingly. If you go to the doctor's office for a blood test and she also endus doing an EKG, the doctor would bill you for the EKG, right? If you go to a lawyer to draw up a will and he also draws up a durable power of attorney, he charges you for both, right? Translators should be careful to be similarly professional: if you are doing any extra work like formatting, DTP stuff, localization, etc., the rate should be higher.
All of these points are very valid, however, they fail to address the issue of the client who has trouble paying the same unit costs for the words "a", "the" and "myxomatosis". It could also pose a problem should clients confer with one another and wonder why one is paying .18 a word while the other is paying .14 for what they see as the same service.
In addition, formatting and translation of PowerPoint/Excel documents, hidden text, metadata, etc. require specialized knowledge, and translators who work well with these formats are well worth the premiums they charge for the services they provide.
For this reason, it may be worth considering a switch to an hourly billing rate, at least for certain projects where there is an atypical level of complexity (many files to work with, changes to the source text in mid-translation, etc.).
If an hourly billing rate is used, everything can be billed in one total package, as opposed to billing for translation/typesetting/formatting.
When doing your translation estimate, it's a simple matter of being aware of how many words you translate an hour (or a day), then using your unit cost to arrive at an hourly rate. Then, adjust that hourly rate by a few dollars to take into account additional complexity.
Then, when you need to give a client an estimate, it's not a word-based estimate, it's an hourly estimate, and instead of adjusting the hourly rate per client, it suffices to adjust the number of hours in the estimate, so all of your clients are technically paying the same unit cost.
It doesn't work for all documents, but it is something to consider.
Post a Comment