CALL and computational linguistics are separate but somewhat interdependent fields of study. The basic goal of computational linguistics is to “teach” computers to generate and comprehend grammatically-acceptable sentences… for purposes of translation and direct communication with computers where the computer understands and generates natural language. Computational linguistics takes the principles of theoretical linguistics with the aim of characterizing a language with computational applications in mind. [2]
A very simple example of computers understanding natural language in relation to second language learning is vocabulary drill exercises. The computer prompts the learner with a word on either the L1 or target language and the student responds with the corresponding word. The computer “understands” the input word by comparing it with a stored answer and gives feedback to the user. Cloze tests work on a similar principle, where the computer compares the words/phrases provided by the learner to a database of correct answers. [2]
On a superficial level, the core issue for humans and computers using language is the same; finding the best match between a given speech sound and it corresponding word string, then generating the correct and appropriate response. However, humans and machines process speech in fundamentally different ways. Humans use complex cognitive processes, taking into account variables such as social situations and rules while speech for a computer is simply a series of digital values to generate and parse language. [2] [3] For this reason, those involved in CALL from a computational linguistics perspective tend to be more optimistic about a computer’s ability to do error analysis and other pedagogical tasks than those who come into CALL via language teaching. [4]
The term Human Language Technologies is often used to describe some aspects of computational linguistics, having replaced the former term Language Engineering. There has been an upsurge of work in this area in recent years, especially with regard to machine translation and speech synthesis and speech analysis. The professional associations EUROCALL (Europe) and CALICO (USA) have special interest groups (SIGs), respectively devoted to Natural Language Processing (NLP) and Intelligent CALL (ICALL). See Module 3.5 at the ICT4LT website for further information.
