The demand for technical and scientific translators has skyrocketed since I first became a translator in the 1960s. In those days, little, if any, formal training was available. Working in-house has always been an excellent steppingstone to a freelance career, so it is a great pity that it is almost unavailable to the fledgling translator. Even in the '60s in-house positions were few and far between, with only a very few large companies, such as Shell and Unilever, having translation departments. In the 1980s, the accountant mentality swept through multinational companies, and translation was the first department to be downsized. If it were allowed to continue at all, it was simply under a single person, often a secretary, whose job was merely to outsource the work.
Yet there is no substitute for a properly resourced, efficiently run translation department, stocked with all the specialist technical dictionaries that are so hard to find and cost so much that most individual translators simply cannot afford them. The sort of mistakes that can arise when work is subcontracted to a translation company whose only concern is how cheaply they can hire a translator can be extremely costly in the long run. Unfortunately, the solution companies have found is not one that benefits translators. They are using bilingual or multilingual secretaries as translators, and they also assume that Translation Memory tools, such as Trados, will help them produce technically accurate translations. It would take a very enlightened management to reintroduce the translation department, and most corporate management is far from enlightened. Even wealthy multinationals would rather spend the money to hire external management consultants, who will tell them they don't need a translation department! In any case, those companies that had such departments have long ago sold off their--sometimes irreplaceable--assets (dictionaries, software, etc.).
Fortunately, today, training opportunities for would-be translators are far more widely available, and as a result external translation services have become far more efficient. Nowadays, many universities offer degree courses in translation studies, and there are M.A. programmes and doctorates in translation. Some universities also offer interpreting courses. Many translation and interpreting courses are part-time, postgraduate courses, held in the evening, that would be ideal for the science graduate who has a gift for languages or rather for translation. Make no mistake, translation is a gift: It is not just a matter of speaking several languages. How the brain works to be able to put the two languages together--and in the case of conference interpreting the connection between the two is made almost simultaneously--is something even scientists have not yet worked out.
Most people think of translation when they think of scientific work, but interpreting is just as important a skill. Professional simultaneous interpreters are given time to study the documentation before a conference, but unless they have a profound knowledge of a complex subject they will find themselves unable to cope. Indeed, so many scientific applications now involve the use of computers that full-time interpreters who make little use of them are falling behind in scientific knowledge and are finding it harder and harder to interpret at technical conferences. Scientific knowledge is therefore equally important in the interpreting field.
Unfortunately, most linguists are arts graduates. Yet the biographies of some of the veteran technical translators reveal that they were science graduates first and translators second. Between 1933 and 1939, the English-speaking countries benefited greatly from the mass exodus from Nazi-occupied Europe, when scientists of all ages flocked westward and eked out a living translating vital scientific papers from what was then the supreme language of science: German. Ironically, the scientific brilliance of these refugees also shifted the primary language of scientific expression from German to English.
Yet there remains a shortage of scientific translators both into and from English. Every translation company and the documentation departments of every large multinational complain of how rare it is to find translators who fully understand what they are translating!
How does one go about getting started as a scientific translator? The first thing to do is to ensure that your knowledge of the languages in question is good and that you know your mother tongue particularly well. Scientists who cannot write well, who make mistakes in spelling and grammar in their own language, are useless as translators. Try to find translation courses that will help you practice and hone your skills. Even while you are learning, you could try to find work, perhaps in your own company if you are employed full-time, or helping students translate papers or theses in your scientific subject.
Once you have gained some experience, the next step is to join a professional body of translators and interpreters, as an associate member perhaps. There is always a professional body in your own country, and you can find the address on the Web (and many are listed in the feature Resources  page). Try to take the qualifying examination of the appropriate professional body, so as to become recognised as a fully fledged translator, and then offer your services to translation companies--you will find plenty on the Internet.
No doubt you have heard something about MT (machine translation) and TM (translation memory). These are great aids to translators who are tackling a large project, because words are constantly repeated and it is helpful to have an instant, automatic reminder of how you translated the same word in a previous passage. TM is very useful for "localization," the translation of series of manuals that are updated annually or periodically; such changes are often minor and occur at random, so translation clients do not want the expense of having the whole manual retranslated when only a few passages need translating. Instead you can run the translation through the TM program, and it will tell you where words match and where there is new, untranslated material. But these are simply aids to translation. It is no more possible for translation and interpreting to be automatic than for Shakespeare to be written by the three monkeys plugging away endlessly at their typewriters. That is because language is creative--each speaker and writer of each language has his or her own means of expression--and only a creative mind could turn that expression into the equivalent in another language.
Another dimension to working as a scientific translator is that you may find yourself getting paid for learning the sort of information that you otherwise would have to pay to find out! For example, I was asked recently to translate a review of LCD screens just when I was about to buy one for myself! In certain areas of science, you could find the knowledge gained invaluable. In any case, as a translator and interpreter you find yourself learning all the time: learning new words and expressions, learning about new advances in your scientific subject--the information gain is limitless.
I hope this has provided a brief insight into the opportunities for linguist-scientists and that it will encourage as many people as possible to consider translation and interpreting, either as a part-time or a full-time career.