One day machines will be as smart as human beings.

Never. What do you mean by smart in relation to a machine?

Originally written by Jane Lamb-Ruiz on March 21, 2009 6:48 AM Hello Jonathan, One aspect of translators that is elusive is improvement over time. Time is measured and dissected in just about every discipline. I think that one becomes a better translator (it's a reasonable hypothesis) over time. In other words, one's knowledge increases over time and one is better able to formulate one's sentences, etc. This is closely related to speed, also. Another critical issue I find is target language knowledge. I found that the more I could speak a source language, the better I could translate from it into my target. So efforts at boosting source knowledge, especially the spoken language, add to improving one's renditions in the target. Conversely, bookish knowledge (from reading) also helps improve both. And in this day and age, many young people are non-readers, therefore they are exposed to fewer narratives. Both, say, in the novelistic sense as in the legal or business sense. What I mean is that all disciplines have their narratives.  And I would add here, these speaking/reading exercises that help improve translation needn't be performed in a translation setting. One might go spend six months in a country of the source language, for example, and during that time, eschew contact with target speakers. When one returns from such a sojourn, the effects are felt down the line time-wise. Judging from my years spent answering translation questions, the biggest issue I see in terms of knowledge is that translators often have not interiorized their the major structures of their source language. Just a few thoughts...

Hi Jane,

On the point about narratives, modern text-linguistics now actually examines the narratives you are talking about. It has recently been realised that a translation can be 100% accurate in terms of "meaning" or "words" but can miss the mark but not taking into account these narratives. It can read fine as a text but not work as a contract or in the fashion it was meant to.

Target and source language knowledge are often overlooked, especially target language. People assume that because it is "their" language, they don't need to develop it. I would love to see studies on the difference between people who work at their TL and those who don't.

Yes, Jonathan...narratives are built through reading. So that reading literature or getting a business degree can both build richness of narratives.....insofar as every academic endeavour, the way I see it, is basically a story. Even texts written to read, for instance, contain narrative. Frankly, narrative for me is opposed to mathematics. Now, even math contains narrative when a mathematician presents a paper..or develops a theory...

But returning to the issue of time. At the ESIT in Paris, first year students are often told to go spend a year in the country of one of SLs. This "year" of course is time. So, presumably,  one way of developing narrative is time spent in country. So, perhaps, one can distinguish types of narrative that might be useful. 

I think time is vitally important. For instance, a 20-year old may not know enough of a language to interpret from it. BUT, by the time they reach 30, they may . Many simultaneous interpreters get their degrees much later than students in other disciplines. I do remember a case of a brilliant girl at the ESIT who finished her two years straight. They did not require her to spend a year abroad. But they didn't give her the diploma either. Because they decided she was too young. In other words, she was good enough to "get the meaning mostly right" but she didn't have the maturity to understand what she was really saying...interesting, huh?

As for working/working on TL (for NS).  Most people, as they start./study translation or interpreting, do it almost unconsciously....it is hoped. They are told, for simultaneous high-level interpreting, that they must be au courant of everything around them. Obviously, the more one reads in that setting the more one develops TL speciallized knowledge. Right? So, you are not really developing language knowledge ie structures of for instance. You are really developing area knowledge (concepts and vocabulary). One already has the language blueprint as it were. You just building different types of buildings...for example, you would already know even the trickiest aspects of "standard structure" albeit "instinctively".....as most interpreters have already finished university when they begin to study interpretation. In other words, the specialized knowledge conceptually gets plugged into existing knowledge blueprint...which is for a NS is pretty well set....Philosophically, I'm interested in the knowledge issue....that too has not been very much written about. And of course, in my case, I'm interested in how the unconscious operates in all this...