The usual controversy that has existed in the translation field for some years now. Will machines replace blood and flesh translators?
I am not going to add any new insight that has not been already discussed, I just want to add some thoughts for all those who still think translating is little more than write the same thing, but in another language.
Translating is more than just that, more than just looking up some terms on the dictionary, or in the internet.
At a linguistic level, grammar, spelling and syntax skills are necessary. Besides, another thing to bear in mind is the technical terminology for each sector: a satellite, for instance, is not the same thing for astronauts than for workers of industrial detergency machinery.
Having said this, it will be logical to think that a machine does not have the necessary skills to tell apart contexts and amniguities, however vast its database may be.
In order to illustrate this (and with this I do not intend to criticize the software, which I consider quite effective, within its limitations), I have turned to Google Translate .
I have typed the following in English, to obtain results in Spanish and Galician:
"Translation costs will add up to a total amount of 100"
As we can see, the translation from English to Spanish is quite accurate and correct, while the result of the translation to Galician is far from acceptable: its syntactic order is chaotic.
Another test, this time from Spanish to English:
"Las cotizaciones bancarias sufren cambios diarios" has resulted in the following English sentence: "Prices changed daily banking". As we can see, in this case the translation changes the message completely: "Los precios cambiaron la banca diaria". Nothing to do with the source.
To spot that type of errors, to tell the difference between the ambiguities a text may show are part of the tasks a translator carries out unconsciously and which software is very far from achieving, at least for the time being. It is true that in certain language pairs, specially if they are similar, the results are amazing, but in languages whose syntaxes are very different results are often completely
Ver ese tipo de incorrecciones, distinguir entre las ambigüedades que puede presentar un texto, son parte del trabajo que un traductor realiza inconscientemente y que el software está muy lejos de conseguir, al menos de momento. Bien es cierto que en ciertos pares de lenguas, sobre todo si son similares, el porcentaje de correccion es sorprendente, pero en lenguas cuya sintaxis presenta diferencias notables los resultados son a veces totalmente unintelligible.
And it is precisely this kind of thing that it is hard to explain a client who says that hiring you is very expensive, and that anyone can do it with online translation software. It is difficult to explain that software cannot tell apart between social, cultural or religious differences, that it does not have into account the grammatic cathegory of words, etc. The client is not interested in that: what he/she is interested in is getting the best result with the least inversion possible.
The ideal situation would be to educate the client in the importance of prioritizing quality over cost, but it is something difficult to achieve, specially in a moment like this. Even so, we will have to keep trying.
More than once a client has remarked, 'Translator charge too much for something you can do on Google. with some time, I could do it myself.' How do you face this attitude? Do you have any success stories of 'conversion' of this client profile into a client appreciating the perks of professional quality translation?