With naturalness, naturally

Since I’m a Brazilian translator, I'll naturally restrict my examples to the pair English-Portuguese, in either way.

Unnaturalness in translation can be lavishly exemplified, by resorting to anything translated automatically by any of the tools now available therefor. Bad human translators will also contribute a lot of these unhappy examples.

I know a blogger whose stuff is read by too many people from too many places. He has his posts translated automatically. Anyone can see the texts so translated are invariably clumsy, hard to understand, full of absurdities, incoherent, in short, they simply never sound natural. No wonder. Anyone but those who believe something automatically translated is adequate to be ‘served’ to a multilingual readership can see how texts are impoverished by direct machine translation, without at least a competent review.

Delivering the raw translation made by Google Translator, for instance, to readers of another language may seem almost unthinkable to anyone who has the slightest idea of what a well written text is. However, that’s just what some people out there are doing. No kidding. You’ll certainly get plenty of sentences which sound now stilted, now unintelligible, nonsensical and utterly ridiculous.

These automatic translating resources do not work in much the same way as we humans (the ‘natural’ translators) use to. For one, most of them, as far as I know, work for free. We humans also make mistakes, of course. The great difference I see lies in the fact that we, at least, stand a chance of ever feeling ashamed for such mistakes.

Onn the other hand, there is no point in praising a machine translator for an outstanding performance, is there? or in telling it off on account of a shabby, unspeakably ridiculous rendering. It is supposed to react with exactly the same indifference to either the highest praise or the harsher scolding, aren’t I right?

What is the point in expecting any naturalness from a machine? It can’t react to any opinion, it can’t understand anything at all, it can’t make any sensible choice to fit situations or circumstances, it can’t consider what it takes in order to choose how to say what to whom, it can’t translate with minimum taste or expertise. Only we humans, who have the experience of using a language as a tool for real communication, can supply any degree of naturalness in translation.

There is much to be considered when you translate, if you want to sound natural. First of all, you have to understand very well what is being conveyed in the original text, and you should also be able to guess certain intentions on the part of the author. For instance, when an author says/writes something only to be funny - no matter how successfully - if his translator just can’t guess this, and worse still, treats the would-be joke as something serious, something important will certainly be lost in translation. No naturalness will be attained. The language used in a given situation may be full of metaphors that give color to the original, but hell to the translator, unless if it is a machine, who isn’t anyway supposed to care for naturalness, adequateness, taste, etc...

Many a time I saw expressions like ‘How far is far?” translated into Portuguese as “Quão longe é longe?” True, it is a literal, word-for-word translation. The problem is that no Portuguese speaker would ever think of asking such a question this way. It is on the whole unnatural. There are human translators who don’t seem even to suspect how unnatural “quão longe é longe” sounds. These will never be good translators before they manage to make such an obvious realization. Machines will never be good translators at all, in this sense, before they can understand what is being said and react accordingly.

I don’t think a cold machine is likely to make the right decision when ‘good morning’ can be translated literally as ‘boa manhã’ (if that’s what these words really mean, for example, in “I spent a very good morning with them”), or when to translate it as “bom dia”, a usual greeting. As a rule, human translators are.

Expressions like “both A and B are …”, according to my experience, are too often translated as “ambos A e B são …”. Laughable, to say the least.

In order to translate even objective, predominantly referential texts with naturalness, the person (or machine, if it were possible) should possess both a lot of experience with the target language handling and a sound knowledge of the source language. In the absence of both, it’s simply impossible, for either man or machine.

Workout

I'm giving private classes again.
A new friend's wife had picked my phone number somewhere, so he gave me a ring a couple of days ago.
After a short introduction, we appointed a first class which took place at his house.
He is in his early thirties and wants to acquire a working command of the English language for professional reasons. He is in marketing, has a university degree and now he has to learn English, otherwise he will be doomed to stick to his present position in the business.
I had sent him some exercises by email, which were used in our first training session. He began to understand and use basic structures of the language by manipulating them himself in the way I had shown him.
He was glad to see by himself that the whole stuff works.
This kindled in him justifiable hope. He had already tried 'fashionable' methods which are expensive but led him to no progress worth mentioning.
Let's see what comes next.