Skip to content
loading...

re: How did linguistics influence programming? VIEW POST

TOP OF THREAD FULL DISCUSSION
re: Interesting discussion, but it seems that somehow implies no relation between natural languages (refered in the actfl link as world language) and c...

Thank you for the reply.

I did not suggest that natural languages did not influence computer languages, only that linguistics (ie the field of study that deals with natural languages) does not seem to have considered computer languages as belonging to its field and thus has likely not influenced their evolution much. Even if, as was mentioned earlier in the discussion both language groups share "meta" concepts like "grammar", "syntax", etc.

But the thing is, we don't describe computer languages with terms like "verb" or "object" (at least not in the grammatical sense of the word), and most of the terms we use to describe them are borrowed from mathematics (function, argument, etc.) which is another field that has little to do with linguistics (and certainly does not want to be seen as being influenced but such a "soft" science :).

Both language groups deal in symbolic expressions that are or are not allowed a given degree of expressiveness, and computer languages, as products of the symbolic work of humans are bound (for now) by the expressiveness of human languages.

As far as evolution is concerned, most computer languages (Lisp dialects are an exception, I don't know of others) are not allowed to easily create new structures that belong to that language, unlike natural languages.

There is a field that I am very interested in right now, which is language learning, that also does not seem to share much between the two groups. Computer languages are mostly taught through "rote learning", while natural languages, especially when communication is a requirement, are not. In other words, for computers, we need to learn (and are practically limited to) what is "grammatically correct" while for humans we need to learn what conveys meaning, which is limitless and includes only a small part of what is grammatically correct.

To conclude this reply, artificial languages are influenced by natural languages but I have yet to find an influence of linguistic research in the area of artificial language "design", which seems to me the question asked by the OP.

But the thing is, we don't describe computer languages with terms like "verb" or "object" (at least not in the grammatical sense of the word), and most of the terms we use to describe them are borrowed from mathematics (function, argument, etc.)

Of course that the grammatical is limited by practicals term. The languages include just what is needed at the syntactic and grammatical level in order to express the subset of semantics that is needed in the problem domain for which the language is designed.

I do not know in details all the computer languages in the list but are you sure there is no programming paradigm with such grammatical stuffs?

In practical terms, what we do is that we read the code and understand the meaning. That illuminated act is some kind of a matching at the semantic level.

If you read for example:

   order.add(item)

you have subject (order), predicate (add) and object (item).

... artificial languages are influenced by natural languages but I have yet to find an influence of linguistic research in the area of artificial language "design", which seems to me the question asked by the OP.

They gave you a generic way to design any language by mean of its grammar and generate automatically compilers and parser. Isn't that one influence ?

Also you may find you will find a lot more in the level of Semantics. Remember that linguistics also covers Semantics.

The past decade there was a lot of work in the field of Semantic Computing, check that work also.

...for humans we need to learn what conveys meaning, which is limitless and includes only a small part of what is grammatically correct.

Then you'll like the book about knowledge representation. When you see which semantics are behind every word you can better select which is the proper word to avoid ambiguity.

You might also find a lot of useful stuffs in General Semantics and Institute of General Semantics

Mathematics and computer science having little to do with linguistics seems to be more often the position of people with an interest in the former two than a generally shared perspective. It's true that programming languages are formally mathematical, but they're still languages with grammars and parsing and even room for implication and ambiguity and authorial expression now and then; and math or no, learners often find framing, for example, object orientation in terms of nouns and verbs more intuitive than staring at proofs.

Manufactured languages, software architectures, and specialized interfaces are much more limited in scope than natural languages, texts, and grammars. But there are still rules they conform to and tendencies they exhibit, so it's rather a waste of time to ignore the vocabulary and theoretical toolkit already developed for us by linguists and semioticians. And of course, treating natural languages themselves in mathematical terms is a topic of no small interest on either side of what dividing line exists.

I totally agree with you. I'm a failed mathematician and an unsuccessful linguist :) But I think the fair reply to the OP is that there are no proofs that linguistics as a scientific field has influenced computer languages anywhere close to the way mathematics have.

I'm re-reading all the replies and yes, there are plenty of proofs, as was referenced in the exchange. I was just stuck in my vision/experience of linguistics.

My understanding of the OP's question was more in the field of programming language design (as Ben seemed to hint) where an influence of linguistic research would have contributed to produce languages closer, in expressiveness and structure, to native languages. So I was considering only a very narrow aspect of both computing and linguistic research.

Thank you Dian and Yucer for the really interesting comments.

code of conduct - report abuse