Carnegie Mellon hosts national linguistics olympiad

The first-ever North American Computational Linguistics Olympiad (NAMCLO) will be held in Boston, Philadelphia, Pittsburgh, and Ithaca, as well as online for students who are unable to attend these venues, on March 29.

The local competition in Pittsburgh will be hosted by Carnegie Mellon; professor Lori Levin of Carnegie Mellon’s Language Technologies Institute is the general chair of the national linguistics competition.

Computational linguistics is the study of natural language from a computational perspective. The field involves studying applications of computers in language interpretation, as well as analysis of languages in a logical and systematic form.

“In short, computational linguistics aims to make computers ‘think’ more like human beings, so that interacting with a computer is more like interacting with a real person,” said professor of linguistics Thomas Payne, the second general chair of the competition, who teaches at the University of Oregon.

Payne said that the goals of computational linguistics include machine translation between natural languages, artificial intelligence, handwriting and voice recognition, and text analysis and processing.

The problems presented during NAMCLO will be representative of the challenges computational linguists face.

One of the example problems, for instance, involves decoding the meaning of individual words in an unknown language, given a few scrambled and unmatched sentences. Another problem explores the most efficient way to search a phone book for common last names.

“I enjoyed the problems thoroughly — [in past linguistics contests] we had to decipher texts in obscure languages, figure out the Japanese calendar system, and ‘discover’ vowel harmony in Hungarian and Turkish,” stated associate professor Dragomir Radev, who teaches in the department of linguistics at the University of Michigan.

The competition’s goal is to unearth new talents in linguistics among high school students, as well as to heighten student awareness of linguistics.
Additionally, the competition is aimed at helping interested students pursue higher education in relevant fields such as computational linguistics and human language technologies.

Local prizes include a copy of The Cambridge Encyclopedia of Language for the top three contestants.

The four best students nationwide will represent the United States at the International Linguistics Olympiad, which will be held in St. Petersburg, Russia, in early August.

The competition is intended to draw students in grades nine through 12, but younger students who are interested in the field and capable of outstanding performance are welcome to participate. Currently, there are no plans to expand this competition to include post-high school students.

NAMCLO will be an “all new and self-sufficient” competition, meaning that the questions will not require any prior training or special knowledge in linguistics to complete, and they concern newly developed phenomena.

“The problems are ‘self-sufficient’ in the sense that no outside information (dictionaries, courses in the languages) is needed to solve them. They are based on pure logic and analytic reasoning, at a level that is totally appropriate for high school students,” said Payne.

He added, “It is entirely possible to solve any of the problems with no linguistics or foreign language background. In previous trials, we have had sixth graders solve some of our most challenging problems. In many cases, the younger students can solve the problems more easily than can adults, because younger people have more flexible minds, and are more willing to ‘enter into’ the logic of an unknown language than many adults are.”

Another activity held by NAMCLO is “War of the Words”, which is a post-competition activity in which the fictional United Nations Space Protection Agency needs to decipher an alien language in order to save the world.

For more information, visit