Indiana University Research & Creative Activity

Humanities, Then and Now

Volume XXIX Number 1
Fall 2006

<< Table of Contents

Photo courtesy Michael Gasser

Michael Gasser
Michael Gasser
Photo © Tyagan Miller

Equality's Grammar

by Karen Grooms

In a small building in a remote region of western Guatemala, a group of Mayan children and adults, dressed in their culture's traditional multicolored attire, switch on a bank of computers and go online. They send e-mail messages, visit chat rooms, and complete interactive lessons. Once again, the old meets the new--that's not so remarkable these days. But what is extraordinary is that all of these electronic transactions are conducted in Quiché, a language that is spoken by a million people but until recently was virtually nonexistent in cyberspace.

Michael Gasser was heartened to learn about this educational center that empowers members of the Guatemalan Quiché community to join the Information Revolution on their own linguistic terms. He sees it and similar efforts as part of the "democratization of knowledge," a movement that Gasser, as researcher and activist, is working to expand. "Currently, about 70 percent of the world's Web sites are in English," he observes. The predominance of English on the Internet, Gasser says, reflects its speakers' disproportionate wealth and power in a world where they are nearly equaled by speakers of Hindi, one of the Net's many marginalized tongues.

An associate professor of cognitive science, computer science, and linguistics at Indiana University Bloomington, Gasser views the problem of the "digital divide" from an interdisciplinary perspective. As never before in his career, he is combining his scientific, computational, and linguistic areas of expertise in the development of a project known as machine translation. Building on ideas from artificial intelligence and models of human language acquisition, Gasser aims to make it possible for computers to translate to and from as many as 100 underrepresented languages spoken in developing countries. He calls his project L3 (for Learning Lots of Languages), and he proposes that L3 can help realize the hope many have held out for information technology--that it can make the world a more equitable place.

Gasser gained thorough acquaintance with some of the world's inequities during the late 1960s and early 1970s, when he spent four years teaching mathematics and English as a Peace Corps volunteer in Ethiopia. Though tragically impoverished, Ethiopia is home to a rich diversity of languages that fascinated Gasser. He became fluent in Amharic, the country's official language, and grew to appreciate the vast difference between its grammar and that of English. (For example, each Amharic verb is based on a sequence of consonants, usually three, which take on hundreds of different forms through the addition of vowels between the consonants and prefixes and suffixes before and after them.)

At the end of his Peace Corps assignment, Gasser returned to the United States and entered the English as a Second Language Program at the University of California, Los Angeles. At UCLA, he was exposed to the field of artificial intelligence. "The concept was mentioned in an education class that I was taking," he recalls. To undertake research for his doctoral dissertation, Gasser learned to program computers and constructed a computer model of sentence generation in first and second languages. This novel work in artificial intelligence led to his appointment, in 1988, to IUB's Department of Computer Science and, in 1989, to the campus's newly formed Cognitive Science Program.

Gasser spent most of the next 15 years, he says, trying to figure out "where human language comes from, especially as it develops over the long term in the language learner, but also as it emerges over the short term in the course of communicative acts." His experimentation with computer models soon caused him to dissent from some prevailing theories about language's origins, especially those advanced by prominent scholars such as Noam Chomsky and Steven Pinker.

According to Gasser, "one very influential position is that we are all born with a significant component of language (or of the concepts that language is about) already in our brains." So all we need to do is "figure out the details of the particular language or languages we're being exposed to." This notion of being "born with" much of our language relies on the premise that "human languages can be shown to share a sizable set of properties," Gasser writes on his Web page, "and much of the research within this framework has been dedicated to finding that set of properties and showing that they are already in place as language is acquired."

But Gasser argues that an innate "language component," or instinct, is not necessary to language acquisition because "there is considerable regularity in the environment, including regularity in the things language is about, regularity in language input itself, and unconscious strategies on the part of adults and older children to make language accessible." Language learners, he says, can rely on "powerful, general-purpose statistical learning mechanisms, . . . sophisticated perceptual and motor control systems, and . . . a propensity for social interaction." Cognitive science research has already made it obvious that "humans are sophisticated statistical learners," he says.

Gasser's L3 project will also be a "statistical learner," capable of adapting to and incorporating new information somewhat like a child learning a language. To be successful, Gasser says, the software will depend on huge amounts of human intervention. He looks to the massive online encyclopedia Wikipedia as a model for the level of collaboration, innovation, technical sophistication, and sense of community necessary to achieve his software's stated objectives. Gasser--a self-described Wikipedia addict who both continually consults the resource and regularly writes entries for it--says that the fact that Wikipedia (and other similarly constructed sites) allows users to add, remove, and edit content quickly and easily makes it an ideal medium for democratizing information.

Gasser plans to test his ideas in two different contexts: Guatemalan schools where Mayan children are learning to use Quiché and Spanish, and Ethiopian schools where students use the three main Ethiopian languages and English in their classes.

The past year has represented a turning point for Gasser--a stage, he says, when he has seen increasing opportunities to pursue his research interests in ways that he hopes can change the world. In addition to the L3 project, he has been collaborating with former Department of Computer Science student Stephen Hockema (now at the University of Toronto) and current Ph.D. student Matt Kane on the Information Customs Project, an electronic aid for critical reading of online information sources. This work is inspired by cognitive scientist George Lakoff, of the University of California, Berkeley, who has popularized the concept of "framing," a process whereby words and speech are selectively controlled, and differently understood, through the use of rhetorical devices.

"One example that Lakoff gives is the idea of taxation. It can be 'framed' as a burden, or it might be framed as a civic responsibility," Gasser says. "Often, through word choice, the writer or speaker 'smuggles' in presuppositions and implications to support a particular viewpoint about complex issues. Another example is the U.S. immigration debate. A conservative writer is more likely to use the word 'illegal,' while a liberal writer is more likely to use the word 'undocumented.'"

Information Customs is foreseen as a software application that will work in conjunction with a Web browser. Like L3, it will be a statistical learner, analyzing patterns of usage from large amounts of data--in this case, Web pages on a particular topic, such as immigration. It will analyze word choices on pages selected by users and indicate possible biases in the content by highlighting the polarizing words. On his Web site, Gasser explains: "Clicking on these words would give more information about whether these words are used by particular writers on the topic and not others, or whether these words are used by most writers but with different senses. Readers interested in further information could see examples of these words from other writers and the names of well-known writers who use the words as in the current document." This innovation, Gasser says, has great potential for helping consumers of Web-based news and opinions to make informed political judgments.

As he is developing L3 and Information Customs, two products intended to enhance democracy, Gasser is deepening his own political involvement. He is a core member of the Bloomington Progressive Faculty Coalition (PFC), which was founded in the days following the September 11, 2001, attacks in New York City and Washington, D.C. The coalition, which has approximately 140 members, sponsors public forums on topics that have included the logic of withdrawal from the war in Iraq, privacy and the USA Patriot Act, and the political role of university professors in state and national elections. Gasser has used the PFC as a springboard for his interest in exploring "scholar activism" in the context of the burgeoning social forum movement. He organized events on scholar activism at the Fifth World Social Forum last year in Porto Alegre, Brazil, and at this year's Midwest Social Forum in Milwaukee, and he is on the program working group for the First U.S. Social Forum, which will take place in Atlanta next summer. Gasser is also active in CubAmistad, a sister-city project that links Bloomington with Santa Clara, Cuba. Despite numerous restrictions, he has visited Cuba five times to accompany local delegations, meet with university officials, and attend conferences, and he is currently helping to plan a course that he hopes can be offered in Cuba to IUB students.

Profoundly committed to teaching, too, Gasser has written an online textbook, How Language Works, for his undergraduate course Introduction to the Study of Language. To accompany the book, he is developing software that enables students to experiment with artificial "mini-languages," allowing them to concentrate on fundamental properties of language without being distracted by the complexities of actual languages. He has also taught Rhythm and Cognition, in which students learn about basic concepts and methods of cognitive science and "about the role of rhythm in the mind."

Rhythm is often on Gasser's mind; he plays saxophone in the Bloomington-based band Afro Hoosier International, which performs popular music from various African countries and also features IUB faculty colleagues Robert Port of the Department of Linguistics, Patrick McNaughton of the Henry Radford Hope School of Fine Arts, and Eden Medina and Larry Yaeger of the School of Informatics.

As his research evolves, Gasser increasingly questions the value of the knowledge it produces. "What does it mean for society?" he asks. He offers an answer in his introduction to How Language Works: "The implication is that the study of what all languages share is also the study of what it is to be human."

Karen Grooms is a senior editor in the Indiana University Office of Creative Services in Bloomington.