Since we all speak and understand human languages (like English), it would be handy if computers did too. After all, computers are the gatekeepers for vast stores of information, and even possess a certain fragile version of human expertise. Artificial languages - like programming languages and database query languages - are a big step up from the computer's native language of 0s and 1s, but they still are clumsy to use, even for the limited number of people who can use them effectively. So let's try to move the computer even further toward our way of communicating, and not just with canned phrases or templates, but with some of the flexibility that we humans bring to our communications. This is the aim of natural language processing.

Potential applications of natural language processing listed by Jurafsky and Martin in the preface to their 2000 text, used in this course, are spelling checking, text document search, speech recognition, web-page processing, part-of-speech tagging, machine translation, and spoken-language dialogue agents. It is worth noting, for perspective, that Allen, in his 1995 text, mentions document retrieval, extracting information from messages, translation, database question-answering, tutoring systems and automated customer service. Because the potential value of even restricted natural language processing is so great, and because there has been significant progress over the decades, Gazdar and Mellish observed in their 1989 text that, "almost every computing group and linguistics group in the world is urgently starting up courses in .. natural language processing."