Approaching and Executing a Japanese Patent Specification Translation

(by William Lise)

Seldom do translators actually go into much detail about what they do when they translate. We talk a lot about terminology, but seldom do translators actually reveal the steps they follow in doing their everyday work. I thought it might be useful for newcomers to patent translation to take a peek at what a not-so-new translator does before, during, and after the translation of a Japanese patent specification.

After receiving a job and identifying customer requirements I perform a physical check of the manuscript content.

Are all the pages there?

Are all the drawings there?

Has the bibliographic information been provided?

Mark manuscript with fluorescent pen

Then I make a step which might surprise some readers. I take a fluorescent pen and mark the end of all non-last sentences in all paragraphs. Why? There are several reasons.

To get a feel for the sentence length. This helps in judging the amount of time likely to be required for translation, since given the same number of total words a specification with a large average sentence word count is likely to take longer than one with shorter sentences.

Simply to identify the end of the sentences. This is useful because (1) my eyes sometimes can't quickly find the end of a sentence and (2) because of the structure of Japanese sentences, it is sometimes useful to "start from the end."

To avoid misunderstanding, I should note that I am not locked in the "sentence-for-sentence" translation mode. This is entirely unnecessary, and tends to result in strange English. For this reason, I combine or separate sentences, and change the paragraph breaks.

Terminology check

As I am marking up the sentence ends, I sometimes find terms that I would like to verify. This is more a matter of deciding what term I will use, than one of learning what term should be used or has been used, although terminology previously used by the inventor is something to consider.

Use of the bibliographic information

I use the bibliographic information on the first page of a patent application to search for similar patents granted to the currentinventors. Many inventors spend their entire careers working on one very narrow area of technology. If that is the case, it is very likely that you will be able to find similar patents that have been granted to the inventors in the US, for example. The USPTO website makes it very easy to search for patents based on inventor names. This capability - and the willingness to read or at least look at the relevant patents - can provide valuable background information and clues as to the terminology which is customarily used with the subject matter of a particular invention.

The above steps typically take me from 30 minutes to 2 hours, depending upon a number of factors (e.g., the volume of text involved and my familiarity with the subject matter of the invention).

Pre-reading

Here we have a point that very few translators wish to discuss. I myself very rarely read through an entire document before starting to translate. There are two good reasons for this.

1. The time requirements of commercial translation in Japan seldom provide enough time to pre-read.

2. With enough experience, a brief skim through the document is usually enough to discover errors.

One measure of an experienced translator is the ability to accurately judge when the translator should stop pre-reading and start translating. Too little pre-reading by an inexperienced translator results in false starts.

Translation

I now move into the translation phase. An important part of this for me is eliminating some of the drudgery of typing. Being a Word user, I use the simplest and closest (but certainly not most powerful) device available: Word's Autocorrect function. For me, this provides an on-the-fly expansion of keystroke sequences and the ability to quickly program new sequences as words come up in the manuscript repeatedly. I have a standard set of abbreviations that I keep in the dictionary, and new ones that I add and remove for each job. The two lists at the end of yesterday's work (a semiconductor patent specification) are given below.

Resting. When into the actual translation process, I tend to break for about 5 minutes each hour, to glance at e-mail for example, or make coffee. My lunch breaks usually run over an hour, but I work late in the day. A typical work day will see me working more than 10 hours net translating time.

Research while translating. Having now installed a broadband connection, I usually keep my Web browser window open for research while I am translating. That said, because I tend to work in fields I know, this research is quite limited.

Errors in the original. During the translation process, I often discover errors made by the author. These are usually things like incorrect reference numerals, incorrect references to drawings, and references to incorrect independent claims appearing in dependent claims. These errors are easy to discover when you are reading for understanding the invention while you translate the specification, something that not all translators apparently do.

Correct yourself and flag with a comment. If I am sure that an error exists and know the correction to be made, I work the correction into the translation and flag the corrected location. Most of my clients prefer that I do this, rather than leave the error in the translation.

Date: 2015-12-18; view: 591