W3C home > Mailing lists > Public > www-international@w3.org > April to June 2001

RE: [OT] Does language model thought!? (was RE: Re[2]: Business C ase for i18n?)

From: Marco Cimarosti <marco.cimarosti@essetre.it>
Date: Thu, 14 Jun 2001 16:29:50 +0200
Message-ID: <27E7FB58F42CD5119C0D0002557C0CCA03F929@XCHANGE>
To: "'webmaster@befrienders.org'" <webmaster@befrienders.org>, www-international@w3.org
Eric Jarvis wrote:
> The Italian words that come immediately to my mind are "subito" and
> dopo" both of which have English equivalents that can sometimes be
> substituted directly and sometimes not...they cannot be directly
> translated by the words "sudden/suddenly" and "later", that would
> lead to circumstances where the translation lost at best a whole
> layer of meaning.

"Subito" is more "immediately" than "suddenly", and the most general
translation for "dopo" is "after".

You are right that these words may have different translations. But this is
true for all words in any pair of languages. Otherwise, translating would be
a mere matter of substituting words.

But I still don't see how this creates a particular problem with the
Italian/English pair and, especially, with expressing time.

When talking about translations, you should always consider *sentences*,
rather that isolated words. Is there a single sentence in English that
cannot be translated in Italian or vice versa?

Based on my daily experience, I claim that the answer is a definitely "no",
for any pair of languages and any sentence we can utter.
There is, in fact, a marginal and well-known exception to this: words used
because of their sound rather than their meaning, for humor or mistake. For
instance a joke like "human bean" cannot be translated because, in other
languages, there is no similarity between the words for "been" and "bean"
(Italian, e.g., "essere" and "fagiolo").

> I was under the impression, after time working in Italian schools,
> that some of the more precise past tenses in English have no simple
> Italian equivalent.

I had exactly the opposite impression, for several years. Then my English

> Where there is a common Latin root then there is
> usually a direct equivalent. Where there is not it can sometimes be
> very difficult to conceive of an exact translation. I can speak
> forms of both English and Italian that use almost exclusively words
> taken from the same original root. I have no doubt this can be done
> also with Dutch or French and English. It's not using the complete
> range of the language though.

I don't agree. It depends how you look at things. If you compare paradigms
of verbal conjugations, I am afraid that you are summing apples and pears.

By the morphological point of view, the verbal flexion of Italian is
markedly more complex than English. English basically only has two personal
tenses: present (we walk) and past or praeteritum (we walked) and no moods.
Italian has many more forms, e.g.: camminiamo, (we walk), camminammo (we
walked), camminavamo (we were walking), cammineremo (we will walk),
camminassimo (should we walk), cammineremmo (we would walk), camminiamo!
(let's walk!).

From this list one should conclude that Italian has a variety of moods (such
as the conditional) and tenses (such as the future) that is unknown in
English. But this is clearly non-sense, as is immediately evident by the
translations in parentheses. In fact, some tenses and moods that Italian
expresses synthetically (by means of morphological variations) are expressed
in English analytically (by means of syntactical construction involving
auxiliary verbs). But, in practice, the two languages are functionally

Again, to judge translatability, the only unit of measurement is a sentence
or a longer passage with a well-definable meaning.

> This explains differences in vocabulary. It doesn't explain
> differences in grammar. In many ways it is grammar that shapes
> thinking, or at least the verbalisation and hence communication of
> that thinking. (What Sophocles thought is of no value to me, what
> has been passed along the ages of what he communicated is important)

I don't know what you mean. Vocabulary may be different, grammar may be
different but, at the end, all languages are equally sophisticated means to
express our thoughts.

The only cases when a language can be at odds expressing thoughts is when it
lacks words for a certain fields. But this problem is relatively easy to
solve: if we are inventing something new, we will shape a terminology for
it; if we are importing some new fashion from abroad, we will adopt (or
adapt) the terminology from another language.

Clearly, it must not be easy to talk about mathematics in Inuit. But not
because something in the Inuit grammar is intrinsically less sophisticated:
just because there is no Eskimo math terminology. But you may be assured
that if, for some reason, mathematics becomes very popular in the University
of Nunavut, the local mathematicians will soon develop a local terminology,
perhaps taking it from English or French, perhaps coining new words.
> It isn't the everyday use that varies. It is the more complex ideas
> that depend both on grammar and the culture at the time that aspect
> of the language was being developed.

I don't see how ideas may depend on grammar. Modern English and a modern
Japanese have totally different grammars, but if you have a manual of rocket
science written in one of the two languages you can easily translate it in
the other one. Ditto for a thrilling novel. Ditto for the instruction of a
microwave oven.

The same text would be very hard or impossible to translate in the English
or Japanese spoken in 1600, regardless of the fact that the grammar of both
languages has changed very little since then.

> My argument is EXACTLY the opposite. It is intended to give a reason
> why standardising to English is a very bad thing in the long term.

OK, sorry, I may have misunderstood. What I meant is that it is a bad thing
also in the short term.

Most people who work in information technology speak English, either
natively or as a second language. This fact is positive (this mailing list
would not exist otherwise) but can lead us to the wrong perception that this
is also true for the rest of the population, for the end users.

This is by far not true. All my colleagues (Italian programmers, like me)
would be able read this e-mail, and most of them would be able to write it.

But most of my friends and relatives out of here (I leave in Italy) would
hardly be able of grasping the general topic. Many of them are educated
people, and can and do use computers. They do speak English, but not at the
level of fluency that is customary in the computer field. What they need is
asking for the toilet or the post office, not discussing algorithms.

It is important to consider this, because it is an easy mistake for people
in this professional micro-cosmos to assume that "everybody speaks English",
even for we non native speakers.

> > By the way, Eric, if you really want to standardi*z*e [...] ;-)
> Actually it's only the language of part of one of them. Not even the
> part my family originate from. [...]

OK. You certainly have noticed the smiley at the end of this paragraph. I
learn English in UK, and it took me several years in an USA firm to come to
accept American spelling.

_ Marco
Received on Thursday, 14 June 2001 10:30:17 UTC

This archive was generated by hypermail 2.3.1 : Wednesday, 21 September 2016 22:37:20 UTC