![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() |
Possible
Strategies for the Translation of Foreign Patents This article was published in
the September 2001 issue of the ATA Chronicle,
Patent law firms often need English translations of patents for the
purposes of litigation aimed at preventing an infringement of patent
rights. In an effort to keep their costs down, sometimes these firms will
request that “only the claims” of a patent be translated. It should be
said that it is not really fair to ask a translator to translate only the
claims of a patent, since this portion is usually the most complicated
part. Some claims are written in a language that makes sense during the
first reading, but some claims are about as easy to interpret as the smoke
signals of a long lost civilization. To further complicate matters, often
only the claims portion of a patent is faxed to a translator with no
context or figures to help him or her understand their meaning. The
language of the claims is sometime deliberately ambiguous in order to
formulate a broad claim, even if the patent may typically provide only a
relatively small and incremental improvement. Moreover, the legibility of
a second or even third generation fax is often poor. This presents
enormous problems, in particular when it is impossible to determine with
any degree of certainty which characters are used in the Japanese text.
In
these situations it may sometimes seem as if the whole world is taking
part in a general conspiracy against translators. We are often asked to do
the impossible and are blamed for everything if we fail to produce a
translation that is less than perfect, regardless of the difficult
conditions under which we must work. As we all know, everything is always
the fault of the translator. The adage “shoot the translator” is
always the logical conclusion when things go wrong.
On the other hand, it is understandable that clients of patent law
firms, who are ultimately footing the bill, would not want to spend
hundreds or thousands of dollars on the translation of entire patents that
may or may not be relevant to a particular patented design. Despite this,
it is still difficult, or impossible, to translate a cryptic section of
text without any context. Fortunately, we can often find the context on
our own without any help from our penny-wise clients. Legible
Text Of Patents Is Available For Free on Internet The
complete text of most published unexamined and examined Japanese patents (Kokai
and Kokoku) can be downloaded or printed for free from the Japanese Patent
Office (JPO) Website (www.jpo.go.jp)
or from the European Patent Office (EPO) Website (www.ep.espacenet.com).
The search page of the EPO site makes
patents and utility models published in English, Japanese, German, and
French in Adobe PDF format available to anyone. If you find it difficult
to navigate the amazing quantity of information available on the European
Patent Office or Japanese Patent Office Website, you can access the
respective search pages on both sites from my Website (www.patenttranslators.comclick
on the links “Find Japanese Patents [JPO]” or “Find Foreign Patents
[EPO].” Thanks
to the online availability of patents in foreign languages, unlike a few
years ago, we can now go and take a look at the entire text, including
figures, whenever we are presented with a couple of poorly legible pages
containing nearly incomprehensible and/or illegible claims. We can also
quickly find other information online. The EPO Website, for example, not
only contains the text of patents in languages such as German, French, and
Japanese, but also a short English summary of the “Constitution of the
Patent” that was filed in Japanese, German, French, and other languages.
By simply typing in technical terms, we can do a field search for other
patents containing the same design. One can also search for patents
published by the same company, or under the name of the same inventor,
which may have been published in languages such as English, Japanese, and
German. This is very helpful not only for confirming the correct
translation of technical terms, but also when patent lawyers request the
translation of only the claim portion of a foreign language patent that
was filed initially in English in U.S. or Europe. Claims
usually need to be changed in Japanese to comply with specifications for
filing in Japan. These specifications are somewhat different from those in
U.S. or Europe, as claims are generally defined more narrowly in Japan and
more broadly in the United States. If we have the name of the inventor,
and we can always find out the name if we have the correct patent number,
we should be able to find the original patent that may have been filed in,
for example, English or German. However, translators are sometime also asked to translate the entire text of a Japanese patent that was filed originally in English and then translated by a Japanese patent lawyer (a benrishi) into Japanese. In these circumstances only the claim portion of the patent is usually changed, and the part “Effect (Operation) of the Invention” is sometime added to comply with Japanese patent laws. This probably happens either because U.S. patent lawyers don’t know that the rest of the Japanese patent will be essentially identical to the text in English, provided that the “benrishi” is a good translator, or because they want to know the exact wording in Japanese, regardless of the cost. One
Mistranslated Word In A Patent Can Cost A Manufacturer Millions While
most translations into Japanese are very good, or at least that has been
my experience thus far, it is understandable that it is important for a
law firm to try to find out as much as possible about the entire
translation of a certain patent. And while translation costs represent a
tiny fraction of the entire cost of patent litigation, a poor translation
can cause immense problems. This is evidenced by an often cited case
involving a patent infringement case, when the term “boron” was
mistranslated in a chemical patent into Japanese by an incredibly sloppy
translator (apparently a Japanese patent lawyer) as “bromine.” The
patent holder lost the infringement case as a result of this translation
error. Although the patent holder appealed the ruling, the appeal was
denied because under Japanese patent law, a patent examiner must recognize
from the invention’s description that the language used is clearly in
error in light of the rest of the application.
A quick search on the Internet that enables us to quickly find life
saving context for isolated patent claims, which may be next to
incomprehensible without the full text of the specifications and drawings,
may thus prove invaluable. For example, as a rule Japanese does not
specify singular or plural. Whether a certain term refers to one part or
several parts (or a plurality of parts, if we want to sound like U.S.
patent lawyers) can usually be determined only from figures showing
labeled parts. The EPO
Website is particularly useful to patent translators because it displays
the text of patents filed in English, Japanese, German, French, etc., with
a short summary in English. The
English summary (which, in case of Japanese patents, is incidentally the
same summary that is also available on the English part of the JPO
Website, along with machine-translated portions of some relatively recent
patents) is usually a translation of the patent part entitled
“Constitution.” This section is generally located on the first page of
the patent. This English summary is mostly written in fairly good or very
good English if the original patent was written in German or French, but
the English is often hard to understand and sometime quite hilarious when
the original patent was written in Japanese. The choice of English words
is often less than felicitous, and the structure of the English sentences
in these summaries often mimics Japanese grammar to an extent that makes
the meaning very hard to understand. Nevertheless, it is very useful when
we can see how another translator, apparently a native Japanese translator
with some background or experience in a given field, would translate
relevant terms. Most of the time, these relevant terms are correct.
Obviously, the problem is that native speakers of languages such as
Japanese, Chinese, or Korean, which are very different from English, lack
sufficient linguistic training or, judging from the results, perhaps even
a basic aptitude for foreign languages. The people who write these
summaries are probably not really translators by training, but rather
engineers, chemists, etc., who “also know English.” The dilemma here
is actually much bigger. How
Can We Be Native In Two Languages When We Live Only Once? A
good patent translator should be a native in both the source and target
language (and when we examine the meaning of the word “native,” we
find that what this really means is that he or she would have to be born
at least twice in one lifetime). Also, he or she should also be a trained
patent lawyer with a specialized degree, for instance a Ph.D. in chemistry
or biology. In a way, this explains the sometime hilarious language of
English summaries of Japanese patents published by official and respected
institutions. Although there are exceptions confirming the rule, a good,
fully bilingual patent lawyer (who can really translate Japanese patents
into good English) is likely to stick to the practice of patent law, which
tends to pay better than translation. Thus, the translation of patents
into English tends to be practiced mostly by people who are not really
qualified either as patent lawyers or as competent translators.
What has changed in the last few years is the fact that
translators, who are mainly linguists, now have immediate access via the
Internet to all kinds of information that used to be available only to
technical specialists in respective fields. It is exciting how the ease
with which specialized information can now be accessed in any language has
put a premium on the ability to access this information in more than one
language and to make it available to monolingual technical specialists.
The other part of a strategy that beginning patent translators may
need to employ in order to tackle initially unfamiliar texts is the
linguistic part. This area is generally more familiar to most translators,
as it is not very different from what they learned or should have learned
in school.
The basic rule that a relative novice can follow when translating a
foreign patent is the old golden rule: start from the simpler parts and do
the most complicated part at the end. After 15 years of translating
Japanese patents almost daily, with a generous sprinkling of German
patents and a few in French, Czech, Slovak, Russian, and Polish thrown in
for good measure, I still translate the claims, usually the most
treacherous part of the job, at the end. I think that it is best to skip
the claims and do them only when I understand the meaning of the text and
thus the design of the patent. The reason why claims are at the beginning
of Japanese patents but at the end of European patents, which is where
they should be, is again very simple. Everything is usually the other way
round in Japan, which includes the Japanese language and all things
Japanese in general. Therefore, when translating a Japanese patent, it
makes sense for us to somewhat modify the order and start from the title
page, but then to skip the claims and work our way through the description
of Prior Art (existing technology) and Preferred Embodiments (practical
examples) back to the claims. Nobody
Thinks Quite Like The Japanese -
With The Possible Exception Of The Germans Surprisingly,
many components of the linguistic analysis techniques that I find useful
for the translation of Japanese patents also work very well for German
patents. The verb is at the end of the construction in Japanese, as it is
in German. It is possible to string together several Japanese characters
that are not normally used as independent words in Japanese, which is very
similar to the way German compound nouns are used in technical terms. The
main problem, both in Japanese and German, is to make sure that the right
link is established between the right components, in particular that the
right nouns are connected to the right verbs. This link is indicated by
the case of the noun, whether it is singular or plural, tense, etc., in
German, but mostly only by particles (joshi) in Japanese. Sometime it is
not as simple as it sounds, either in German or Japanese. There is an
undeniable affinity between the psychology of the two languages and not
surprisingly, I have heard Japanese people comment that they find it
easier to get along with Germans than with other nationalities, including
Americans. In long, extremely complex sentences, it is best to identify
the subject or object first (the subject will be identified with particle
“ga” and the object with particle “wo” in Japanese and by the case
in German). If there is no subject in Japanese, it is tempting to use the
topic, or “wadai,” as the subject in English. However, this often does
not work because the “wadai” has a largely adverbial function in
Japanese. The best “direct” translation of “wadai A” would be
“as far as A is concerned.” This is also usually the worst possible
translation to use in English. In other words, the most faithful
translation is also the least attractive one, especially if this “as far
as” or “with respect to” construction is used several times. This is
a dead giveaway that the translator is an amateur who is not really sure
how the Japanese parts of speech really belong together or what a clearly
comprehensible English sentence looks like. If there is no subject
(particle “ga”) or topic (particle “wa”) in Japanese, the
“wadai” (topic) may be continued from a previous sentence. Sometimes
it will be found several clauses or sentences back on the previous page or
in the title of the entire section. But the topic can usually be found
quite easily in the Japanese text, unless the text is written by a
particularly poor and or/mean writer participating in the general
conspiracy against translators mentioned earlier in this article. “Patentese”
Should Make Sense - It’s The Law! Although
what the writers of Japanese and German patents produce may seem at first
glance completely incomprehensible, the text usually does make sense
provided that one can keep in mind the underlying objective of the
allegedly improved design. I find it very useful to create a mental image
of the main task to be achieved by a patentfor instance, the
design of a keyboard for a Japanese typewriter that crams about 3,000
Japanese characters on the keyboard of an English type of typewriter. I
then go back to this mental image I have created for myself, often with
the help of the figures contained in the patent, every time I seem lost in
long convoluted sentences. It may seem silly, but it helps me when I try
to connect the right verbs, hidden among the verbiage at the end of a
hideously long sentence, with the right nouns.
It
is perhaps understandable that it should not be the job of patent lawyers
to make their description simple to understand. If it was simple to
understand, inventors could decide to write their own patents for their
own inventions and patent lawyers might then have to look for a different
job. As my wife once put it, a Japanese patent lawyer is a “muzukashii
nihongo o kaku senmonka” (a specialist who specializes in writing
difficult Japanese.) Nevertheless, it is their job to write the
description in a way that makes sense, and our job is to figure out what
they actually mean. According to patent law, patents should be written in
such a way so that “a person with ordinary knowledge in the art” would
be able to understand the description. Therefore, what is required from
translators is “ordinary knowledge,” or an understanding of the basic
concepts, rather than a detailed knowledge of the field in question. To
understand patent lawyers, we have to think like them. Instead of
concentrating on words, words, and more words (as in how many words can I
translate per hour), we have to try to keep in mind the gist of the design
and the purpose of the patent. That is, a broad definition of a certain
technique that, if copied by the competition, will provide evidence for a
patent infringement lawsuit. In some cases, this so-called “ordinary
knowledge of a person in art,” may be a very specialized language that
is quite difficult to understand (in my case, for instance, the field of
biotechnology or dentistry). On the other hand, even a patent describing a
relatively complicated medical device (for instance, an angioplasty
device) may have relatively few specialized medical terms, such as the
names of heart muscles, while the remaining terminology will be quite
simple even for a translator who is not used to this field. Because
medical translation is not my specialty, when I am asked to translate a
medical patent, I first take a good look at the text and then either
translate the patent myself or, if the terminology is very complex, refer
it to colleagues who are specialists in this field. It’s
A Dirty Job But Somebody’s Got To Do It
The fact is that there are thousands of patents in foreign
languages that need to be translated quickly, and only a limited number of
translators who are able to translate them competently. Thus, the
translation of patents represents a fairly reliable source of work for
translators, especially those who translate from Japanese and German and
are willing to invest the time and energy required to overcome the initial
“learning period.” Patent translation is also very rewarding because
in addition to having a constant source of work, often at higher rates
than for other types of translation, patent translators are constantly
learning through patents about the amazing machines and technologies in
the world around them.
And it is kind of gratifying to know that when we, translators, are
learning in this way about the clever gizmos surrounding all of us
wherever we go, unlike “civilians” (nontranslators), translators
actually get paid for this.
|