| Ways of quoting Japanese-English translations || Jan 27, 2006 |
Sergio Juarez Garcia wrote:
It isn't so clear for me as yet, because characters are not units of meaning in Japanese as words (including these last ones compounds of one till x kanji). So I don't think it would be right.
I am not sure what you mean by "being right", but here is what I think:
If I understand your problem correctly, you have a Japanese text, that needs translation into English, and you need to quote (set the price) for the translation to the client.
(If you already have done the translation, and have not agreed on the pricing with the client beforehand, you may be in trouble, but that is another story.)
There are two cases regarding the source material:
1. If the Japanese source text is in hardcopy (fax, book, copy of a document, etc.). then there is no way to count the source text, unless you go ahead and do it by hand (who has the time for that?), so you will have to invoice AFTER the translation is done, based on the target English wordcount, or by the hour, as you agreed with your client.
2. If the Japanese source text is softcopy, meaning: in a file that is editable (Word, txt, etc.) - then there are various tools you can use for automatically counting the Japanese text. Microsoft Word's built-in Count function will do that, too.
When you count Japanese text, you count the characters, not "words" as others already mentioned. This is the standard way, as there is no clear algorithm to find where one word ends and the other starts.
MS Word's Count function will give you the character count. If you use Trados, the analysis function in Trados will also give you the counting results in characters for Japanese text.
You may found it strange, but you should establish your "per character" rate, and get used to using it, because this is the method used most often, and it is generally accepted.
There is a multiplier/divider that usually works for general text translation, and that is 2.5 . (I remember there were discussions about this figure previously in another thread on this forum.) This means you can estimate the target English wordcount by dividing the number of characters in the Japanese source text by 2.5 .
So, if you have a 1000 character long Japanese text, you will end up with approx. 400 English words after the translation.
There are some factors that can change this multiplier somewhat, such as having many katakana words in the text, or many alphanumeric strings (for example English product names, programming code, numbers, sizes for parts, statistics figures for clinical trials, etc.) that will stay the same in the translated text.
You can take some of your past translations and simply compare the number of source characters with the target word count, and see how the ratio works out for you, or for the particular type of texts you are dealing with.
There are some clients (I know at least one agency) that uses a slightly different counting method. They say they count "words" in Japanese text, and they using the figures MS Word would give for a Japanese text under "Words" in the Count window.
This counting method means that each Japanese character is counted as 1, and each continuous string that is written in alphanumeric characters is counted as 1. Here is a sentence as an example:
This would be counted as 18 characters, but only 13 "words" in Word.
So, you have to make sure that the rate you are quoting matches the counting method. In the example above, the more alphanumeric strings you have in the text, the less money you are getting in target word units. Whether it is fair or not, is another issue, the point is that both you and your client have to have a clear understanding on how the text is counted, what count will be the base of billing, and at what rate is used for the billing.
I hope this helps somewhat.
[Edited at 2006-01-27 16:49]
| || || |