|
|
I have been translating patent specifications, mostly from Japanese to English.
One part of my job that is very important but very tiring for me has always been "proofreading" in its broad sense. In this article, I use the term "proofreading" to refer to any post-processes to be performed before a translation is sent to a client, including "reviewing", "editing" and "proofreading" (in its literal sense) the translation. Particularly, I never liked the process of comparing the J and E texts while moving my eyes back and forth between the two documents laid out side by side, trying to spot errors. It takes a lot of time and energy.
In an attempt to improve my productivity in proofreading, I recently started "prooflistening" my translation instead of proofreading it, and I have found it quite a bit more efficient in spotting errors than proofreading. So, I thought I'd share the idea with those of you who haven't tried it.
Prooflistening is similar to, but a substitution for, "the side by side comparison".
To "prooflisten" my translation, I simply copy-and-paste the translation text (an English patent specification in my case) from MS-Word into a "speech synthesis" or "TTS" (text-to-speech) application to have it read the text aloud for me while I keep my eyes on a hardcopy of the Japanese original, waiting for any error or discrepancy to be "heard". Each time my ear catches an error, I pause the application reading the text, make changes on MS-Word, and resume the application.
As you can imagine, however, prooflistening doesn't replace my proofreading process completely. After the prooflistening process, I still need to literally "proofread" a printout in order to make sure my translation "looks" ok in terms of formatting, etc. There are some other drawbacks with prooflistening. For example, the TTS application often mispronounces words with more than one possible pronunciation, some abbreviations, special (e.g., Greek) characters, symbols, etc. Therefore, in most cases, I preprocess my text (search-and-replace special characters, etc.) so that it sounds as it is supposed to sound, though preprocessing may not be necessary because the application consistently mispronounces the same problem words/characters the same way throughout.
Despite all the drawbacks, however, prooflistening has been a much more productive method, at least for my case, in terms of eye fatigue, error-spotting accuracy and speed.
Another thing I like about prooflistening is that it can be nicely done even when you are out somewhere. For example, I can convert my translation text into an MP3 file (using freeware), and carry my MP3 player (loaded with the translation text), the original (Japanese) document, and a pen with me into a train or outdoors, where I can prooflisten my translation by marking up on the original document as my ear catches errors in the translation being played back by the MP3 player.
For those who want to give it a try, here's some information on the particular software I use.
The TTS application I use is "Say Hello! 1.0 for Windows 2000" (freeware), and I use it with Windows 2000 Japanese. I have tried some other (freeware) applications, and I like this one by far the best. All the others I tried seemed to be able to start reading only from the very beginning of a text; you cannot resume reading from anywhere in the middle of the text, which makes prooflistening practically impossible for medium to long documents.
Say Hello! can be downloaded from Vector here.
Say Hello! requires VB6 runtime. One of the latest VB6 runtime packages that seems to go along with any Windows OS can be downloaded from Vector here.
Say Hello! also requires TTS engine installed on your PC. If you use Windows 2000, it is already installed. Otherwise, you may have to download and install one. There seem to be many ways to do this, and I can't tell you which is the best way, but here is one way that worked on my friend's PC running Win95 (I can't be held responsible for any trouble caused on your PC; try at your own risk.).
Download and execute this file from Microsoft (Microsoft Speech API).
This should install TTS engine along with one voice ("Male Sam") on your PC. If you want to add more voices to your PC, download and execute this file from Microsoft (containing 15 voices).
I don't know much about Mac. However, as suggested on Honyaku mailing list, MS-Word for Mac (at least Word 2001 for Mac) has a TTS feature.
Ichiro Tsuji worked for four and a half years as a translator at a Kansai-based patent law firm, before he became a freelance J<>E patent translator in 2000.