At its simplest, Computational Linguistics (or CL) is all about trying to create a computer which uses language in the same way as a human does. The Holy Grail would be a human having a conversation with a computer but thinking they were talking to a another human.
But this is just the tip of a giant linguistic dark continent which has never been fully mapped and since it began CL has spread into many other fields and disciplines and expanded the boundaries of what it is doing as more is learned about language.
The easiest way to understand what CL is all about you is to go back to its origins in the 1950s when computers were just beginning to be more widely used.
The goal then was to write a computer program which could translate between two languages. On the surface this seemed eminently possible and groups of computational scientists and linguists came together to start work on it. However, it soon became apparent that language was far more complex than had previously been thought and the job was a lot more difficult than imagined and in the 60 or so years we haven’t perfected it by any means.
Branches of CL
There are two general branches of CL (notwithstanding the overlaps between CL and various other disciplines).
On one side Theoretical CL concentrates on using computers to analyze language and tries to formulate rules defining how language works. To take a simple example this is the kind of problem they face:
Time flies like an arrow.
Fruit flies like a banana.
Both are grammatically correct and almost identical on the surface but at the same time they are very different. In the first, the verb is flies and in the second the verb is like. But how can a computer know this when the two sentence structures appear identical? Theoretical CL tries to discover the underlying rule which separates these two sentences.
Meanwhile there is Applied CL where computer scientists and linguists are trying to create software which emulates human language production. (This is known as NLP or Natural Language Processing.)
They have come a certain way to do this, for example software now can emulate the reasonable human voice and one can have a “conversation” of sorts with a computer. One well known application here is the Turing Test named after Alan Turing, the esteemed computer scientist. The test involves a human user trying to identify between a conversation with a computer and conversation with another human.
An example of this kind of program is ELIZA first developed some 40 years ago. An online version allows you to “talk” to chatbot which analyzes your input and answers you appropriately. (See the links below.)
CL and TEFL
CL has affected TEFL as well. Mainly this involves taking the finding from the analysis of language and applying them to teaching.
This is not so much involved in trying to understand how humans learn language but rather how language works. The in-depth analysis done by theoretical CL has led to Corpus Linguistics which looks at how language is really used (as oppose to how we thought it was used). This in turn has led to the development in the way dictionaries are put together, the kind of language we teach in class, the order in which language is taught and so on.
For the TEFL teacher at the chalkface, however, CL is a distant land, not often (if ever) thought about.
CALL – Computer Assisted Language Learning – a general look at the more practical side of computers and TEFL
Corpus (pl Corpora) and TEFL – corpus linguistics