Wednesday, April 24, 2024
HomeOpinionTechnology development for our languages and linguistics

Technology development for our languages and linguistics

Here we discuss the case of Odia. But many of the observations made here on the subject would apply to many of the other languages as well.

Of late, our language experts and intellectuals have been talking about the need for Odia becoming the language of knowledge and of economic opportunity, although there hasn’t been much talk about a viable plan of action for this. Now, what has not often been suggested is thatOdia must become the language of the Internet as well. Whether one likes it or not, much communication, information storage and access, etc. are going to take place, in not too distant future, on the online platforms. The digital age has quietly arrived. As of now, for many people of Odisha, English is the language of the Internet. Those who do not know English are excluded. The common man in the State will be empowered if he (she /they) is enabled to use the online platform effectively and this can happen if Odia becomes the language of the Internet. More people will then have access to information. Education and culture canbenefit more directly iflanguage technologies such as search engine, machine translation for information-centric prose material, text summarization, text-to speech and speech-to text conversion, spell checker, grammar checker, style checker, etc. are devised.

In order to develop these tools, we need to create the relevant knowledge.  Think of machine translation: there is need for computational bi-(in fact, multi-) lingual dictionaries, comparative computational grammars, comparative morphological analyzers, etc. As for the search engine system, Odia would require an Odia computational dictionary and a computational grammar, a morphological analyzer, a morphological synthesizer, a part-of-speech tagger, word sense disambiguation system, wordnet, etc. As of now, in this language, we do not have a grammar and a dictionary, which are machine-usable, which is eminently understandable, because there was no need so far for the same.“A sentence is a group of words that expresses complete sense” may be fine for a commonsense definition of a sentence but it is not machine-usable. The system would not know what “knowledge” and “complete knowledge” constitute.The problem is not with the machine but with us. Can we rigorously define “knowledge” and “complete sense”? We cannot, the machine cannot too. The machine works on the knowledge the humans have implemented on it.

Creating this knowledge is a big challenge but a challenge which has to be met. As already mentioned, this would not only enable the language to function in a new domain but also empower the people in this information age.

We need manpower – a hundred persons to start with, over a period of five years – trained well in computational linguistics, more specifically, natural language processing (NLP). It is indeed very heartening that the new language university in Odisha, namely, Odia University, offers Linguistics and NLP as one of its courses. This is very probably the first time in our country that a course with this title is offered at a university. From the title of the course it is not clear whether the course would cover only those topics in linguistics that are directly relevant to NLP or would be more broad-based. Some think this should not be so, if doing NLP work is the main objective. An old-timer linguist like me is more likely to think that an NLP person should have a broader understanding of language and its uses than a “linguistics-for-NLP”course would give him.  In any case, one hopes that the focus will be on the analysis of Odia language, in both the linguistics component to the extent possible and the NLP component, which must be entirely Odia language – centric.

Teaching material for Odia linguistics course, in some areas at least,needs to be created for this to happen; what exists has to be augmented. To take just one instance, our existing grammars do not deal with structural ambiguity. The sentence “flying planes can be dangerous” is ambiguous between the readings: the act of flying planes can be dangerous and the planes flying (i.e., not at rest) can be dangerous. It is ambiguous not because one or more words in it are ambiguous. The ambiguity is due to the structure. Our grammars do not deal with structural ambiguity. This gap has to be filled because ambiguity, including the kind mentioned, is part of day-to-day language use. This apart, NLP has to address the issue of disambiguation of ambiguities. Ellipsis resolution is another issue NLP has to be concerned with. There is, as far as I know, no detailed treatment of in our grammars. General purpose grammars have to be enriched by the addition of new material. In fact, a good general purpose grammar can be the basis for the preparation of a good computational grammar.

The problem is that linguistics hasn’t received the necessary encouragement in Odisha. Our Odia departments have been almost exclusively literature – oriented. Linguistics is viewed as basically grammar, phonology and morphology and the teachers of literature think all these would not contribute to the study of literature. But linguistics is a great deal more than grammar. It studies language use in society, the way power structure is expressed in language, the way language reinforces or questions the existing social systems and cultural practices, how one uses language to advance his self-interest, the nature of literary language, creative use of language in literary works and much else. These certainly have a place in the study of literature, in particular, drama, fiction and essay. In the University linguistics departments (there are just two, by the way), the focus does not seem to be on Odia linguistics.

Going by the pronouncements of the State Government, it is committed to encourage the use and the study of Odia. In my view, linguistics has an important place in this project and it deserves support – not just from the Government, but from the teachers, in particular, faculty members of the Odia departments and the intellectuals as well.

(The views expressed are the writer’s own)

Prof. B.N.Patnaik

Retd. Professor of Linguistics and English, IIT Kanpur

Email: [email protected]

(Images from the net)

- Advertisment -

Most Popular