Unique languages, universal patterns

Thu, 23 Feb 2012 18:30:08 GMT

Under the surface, English and Japanese have deep similarities, as MIT linguist Shigeru Miyagawa argues in his new book, Case, Argument Structure, and Word Order.

Samuel Jay Keyser - Interview No. 2 - Dec. 2, 2010

Fri, 03 Jun 2011 21:38:06 GMT

Second of three oral history interviews with Samuel Jay Keyser, Professor Emeritus in the Department of Linguistics and Philosophy at MIT, jazz trombonist, poet. Interviewed by Forrest Larson, MIT Lewis Music Library. Topics include: technical aspects of trombone playing, the King 3B trombone, trombonists Tommy Dorsey, Carl Fontana, Phil Wilson, Jack Teagarden, Bill Watrous, improvisation and rules, a linguistic perspective on jazz improvisation, bebop, jazz musicians from MIT and the Boston area, Roy Lamson, Warren Rohsenow, Herb Pomeroy, Tom Lindsey,(trumpet), Bill Youngren (piano), Jeff Stout (trumpet), Perry Lipson (guitar), Steve Pratt (bass), George Poor (trumpet), Dave Broderick (bass), Rich Orr (trombone), Mark Harvey (composer, trumpet), Everett Longstreth (band leader, arranger, trumpet).

Samuel Jay Keyser - Interview No. 3 - Dec. 17, 2010

Fri, 03 Jun 2011 20:35:31 GMT

Third of three oral history interviews with Samuel Jay Keyser, Professor Emeritus in the Department of Linguistics and Philosophy at MIT, jazz trombonist, poet. Interviewed by Forrest Larson, MIT Lewis Music Library. Topics include: scientific and artistic creativity, historical state of linguistics, phonology, argument structure, poetry analysis, generative grammar, jazz and linguistics, music and language, poetry, free verse poetry, Theory of Evolution and the arts, linguistics at the MIT Research Laboratory of Electronics, writing poetry and short stories, dixieland jazz, Noam Chomsky, Kenneth Hale, Sylvain Bromberger, Morris Halle, Leonard Bernstein, Ray Jackendoff, Ernie Clark (trombone), Bobby MacInnis (trumpet), Dan MacInnis (banjo), Dave Whitney Orchestra, New Liberty Jazz Band, Everett Longstreth (band leader). Professor Keyser reads two of his poems, and plays two tunes on the trombone.

Celebrating Science and Engineering Breakthroughs III

Tue, 29 Mar 2011 04:00:00 GMT

03/29/2011 10:30 AM KresgeSallie Chisholm, Lee and Geraldine Martin Professor of Environmental Studies and Professor of Biology, MIT; Regina Barzilay, Associate Professor of Electrical Engineering and Computer Science, MIT:; Anette Hosoi, Associate Professor of Mechanical Engineering, MIT; Nergis Mavalvala, PhD '97, Professor of Physics, MITDescription: Although these three speakers travel in quite disparate worlds -- natural language processing, mechanics of tiny organisms, and violent cosmic events -- they convey a comparably infectious enthusiasm for their research. In the early days of artificial intelligence, "people had the na've idea that if you took a computer, and fed it with enough human knowledge, it could eventually understand human language," relates Regina Barzilay. In the 1990s, after this "spectacular failure" in methodology, a new approach evolved: training a machine to infer knowledge from piles of data. Some successes are already in evidence (IBM's chess" and Jeopardy"playing programs). Barzilay wants to move this learning further, toward "grounding interpretation of language in the real world." She describes algorithms that could take over the often exasperating grunt work of installing new Windows software on computers, by reading common instructions and executing the actions. Her system learns to map words with increasing accuracy by using feedback. She is also investigating how to get a machine to understand and act in a much more complex environment, such as the Civilization computer game. Using simulations, her algorithms make predictions about the best possible moves (e.g., invade another nation), and can even absorb instructions from the game's manual, which should make the system "good enough to play against humans." Her ultimate goal: "enabling computers to function competently in a world rich in unstructured information." For tiny organisms in a fluid environment, a key challenges is viscosity, says Anette Hosoi . In particular, Hosoi focuses on eukaryotic cells with spherical heads attacked to flexible tails. Remarkably, the diameters of these tails are the same across all species -- between 250"400 nanometers. Whether a hair in a lung, or a cell in green algae, these tails are all made of microtubule structures that can slide and bend in similar ways. By analyzing these common properties, Hosoi can predict optimal morphology and kinematics of comparable tailed microorganisms. For instance, analyzing sperm cells, Hosoi's team determined that the best ratio of tail to head for swimming efficiency is 12 _ and not just for sperm. "I don't care what the species is, what it's made out of, the tail should be 12 times as long as the headGetting something as clean as that is very exciting!" One outlier in the study: the Bandicoot, whose sperm's fat tail did not share the same radius as all the others. This kind of optimization research, says Hosoi, can only improve, as the computational costs of analyzing vast repositories of biological data continue to drop. When you begin to understand underlying principles in biological structures, she says, "you can move on to inform engineering designs." Nergis Mavalvala takes her audience "to a slightly uncomfortable side of the universe, warped and violent." She searches the cosmos using "a completely new messenger -- gravitational waves that travel to us from distant sources." Mavalvala credits Einstein (by way of Newton) for first proposing these waves. He developed a picture of space and time "as a fabric" that can be dented by massive objects, exerting a gravitational pull. Massive objects bouncing around or vibrating caused "ripples of space time itself" -- hence gravitational waves. To produce these waves, says Mavalvala, you need lots of mass and rapid acceleration, explosions and collisions _ produced by compact stars slamming together, merging black holes, and the conditions that immediately followed the Big Bang. Using special detectors (laser interferometers), Mavalvala has been trying to detect gravitational waves. The first measurements taken by these detectors found a gamma ray burst explosion, but no gravitational waves. A next"generation detector is on the way that makes it possible to "listen to more distant sounds," but Mavalvala notes that at this level of "exquisite measurement," you pay the price of "quantum uncertainty." However, she is certain the elusive gravitational wave will finally be captured, and "we will be testing general relativity: the first direct observation of ripples of space"time." About the Speaker(s): In addition to her other appointments, Penny Chisholm currently serves as co"director of Terrascope, an MIT learning community for freshmen. She is also a visiting scientist at the Woods Hole Oceanographic Institution. From 1988"1995, she served as the MIT Director of the MIT/Woods Hole Joint Program in Oceanography. Chisholm received the 2005 Huntsman Award for Excellence in Marine Science, and is a Gordon and Betty Moore Foundation Investigator in Marine Science. She has published papers in PNAS and Nature. She received her Ph.D. in Biology in 1974 from S.U.N.Y. Albany.Host(s): Office of the President, MIT150 Inventional Wisdom

Samuel Jay Keyser - Interview No. 1 - Sept. 22, 2010

Fri, 25 Feb 2011 20:45:29 GMT

First of three oral history interviews with Samuel Jay Keyser, Professor Emeritus in the Department of Linguistics and Philosophy at MIT, jazz trombonist, poet. Interviewed by Forrest Larson, MIT Lewis Music Library. Topics include: musical experiences childhood through college, college and graduate studies, old English literature, early linguistics career, Noam Chomsky, Morris Halle, hearing jazz in the 1950s & 60s, becoming a jazz trombonist, playing in the MIT Concert Jazz Band, Everett Longstreth, jazz at MIT, Warren Rohsenow, Roy Lamson, trumpet player Herb Pomeroy, Mark Harvey.

20th Killian Award Lecture (1992) - Noam Chomsky, "Language: The Cognitive Revolutions"

Mon, 31 Jan 2011 15:36:03 GMT

Professor Noam Chomsky delivers the 20th annual James R. Killian, Jr. Faculty Achievement Award and Lecture, titled "Language: the Cognitive Revolutions," on April 8, 1992. The Killian Award was established in 1971 to recognize extraordinary professional accomplishments by full-time members of the MIT faculty. A faculty committee chooses the recipient from candidates nominated by their peers for outstanding contributions to their fields, to MIT and to society. [T1704, T1705]

Humanities in the Digital Age

Wed, 20 Oct 2010 04:00:00 GMT

Reports of the demise of the humanities are exaggerated, suggest these panelists, but there may be reason to fear its loss of relevance.

Rebuilding Haiti

Tue, 23 Feb 2010 05:00:00 GMT

02/23/2010 4:00 PM Bartos theaterCherie Moit Abbanat, Lecture, Department of Urban Studies and Planning and the Department of Architecture, MIT; Michel DeGraff, Associate Professor of Linguistics, MIT; Erica James, Associate Professor of Anthropology at MIT; Dale Joachim, Visiting Scientist, MIT Media LabDescription: In the aftermath of the January 2010 earthquake, four panelists with strong personal and professional ties to Haiti share their insights about the different paths to rebuilding and reconstructing the country. Erica James begins with a view of Haiti's history of "ins_curit_", a term used to describe "cycles of political violence, crime, and economic deterioration that have accompanied periods of political and economic upheaval, foreign occupation, dictatorship, and continued environment decline." She believes the transition from emergency to reconstruction must deal with the challenges of repeated cycles of psychosocial trauma. Her concern is that international organizations, in attempting to alleviate the suffering of earthquake survivors, will give rise to practices that undermine the effectiveness of their interventions and create even more victims and victimization-unintended, and unwanted, consequences of their help. For James, the issues of population management-the regulation and distribution of resources, identity, and accountability-are important considerations in reconstruction and rehabilitation efforts. Cheri Miot Abbanat taps into her American and Haitian networks to find out what survivors need and want immediately to help rebuild their lives and their country. While governments and NGOs bring in traditional support-technology, medicine, food, housing-Abbanat suggests first "seeing it with Haitian eyes." She asks that aid organizations respect what is already in place in Haiti: homegrown knowledge, the language, what already works. Although fragile, existing support systems could be bolstered by international aid organizations instead of being replaced by them. Dale Joachim recognizes that "technology doesn't solve everything, but it solves a lot of things." His vision for rebuilding Haiti focuses on energy, the environment, and communications. By addressing Haiti's serious energy imbalance and by "bootstrapping" the public health, education, and rural enterprise systems with a robust communication infrastructure, the path to reversing the breakdown of the environment-in particular, Haiti's massive deforestation-will lead to far greater long"term recovery for the country overall. Using a series of overheads comparing several different countries of similar sizes and densities, he shows how the imbalance in Haiti's energy input/output has a pervasive impact on the Haitian infrastructure. Resolving the energy problem will help resolve issues of education, deforestation, and public health concurrently. Michel DeGraff uses language and linguistics "as a lens on [Haiti's] history." Without recognizing and resolving the complicated socio political stratifications created by language and economics, Haiti will be "rebuilt for the 5% who have always been well off," leaving the other 90%-those who speak Creole-no better off than they were before. DeGraff asserts that Haiti still suffers under brutal class and race inequities brought about, in part, by the power held by those who speak French over those who speak Creole. He believes that by changing the school system, which has been used to maintain these inequities, and by using Creole as the language of all Haitians, the system of language apartheid would be minimized and allow more Haitians access to economic power. A Q&A session follows. Host(s): School of Humanities, Arts & Social Sciences, Center for International Studies

Computers with Commonsense: Artificial Intelligence at the MIT Round Table

Sat, 06 Jun 2009 04:00:00 GMT

06/06/2009 11:30 AM KresgePatrick Henry Winston, '65 SM'76, PhD '70, Ford Professor of Artificial Intelligence and Computer Science; Description: Visiting the San Diego Zoo's orangutans and chimpanzees inspires Patrick Henry Winston to ponder what makes humans different from our primate cousins. His field of artificial intelligence extends that question to thinking about how humans differ from computers. Winston's goal is to "develop a computational theory of intelligence." Bridging the gap from people to machines requires a complex understanding of how we think. Winston asserts we think with our eyes, our hands, our mouth. Humans rely upon visual, motor, and linguistic faculties to learn and solve problems. Perceptual powers enable naming, describing, categorizing and recalling. In the aggregate, these processes are "commonsense," a hallmark of cognition that Winston aims to vest in computer programs -- to endow transistors with the nuanced capabilities of neurons. Crucially, we also think with our stories. Throughout childhood and formal education, we are taught via fairy tales, myths, history, literature, religion, and popular entertainment. Professional disciplines like law, science, medicine, engineering, and business are conveyed through stories too. Recognizing patterns, relationships, and mistakes, as well as abstract concepts like revenge or success, helps us explain, predict, answer questions. The delicate processes of extracting knowledge and capturing meaning may appear seamless or instinctive in the evolved mind, but must be parsed syntactically to "teach" a computer to achieve the same ends. What might be practical applications "for systems that understood stories"? Winston suggests that decision"making in business and military strategy would benefit. And no less, comprehending cultures. If a computer program could derive clues from context, perhaps it could determine why "what plays in Peoria" doesn't translate to Baghdad. Early efforts to build a computational theory of intelligence focused on "symbolic integrationWe figured out how to make programs do calculus by 1960but computers remained as dumb as stones," Winston says. When we progressed to building robots -- "things that move" -- language was still lacking. "We forgot that the distinguishing characteristic of human intelligence is that linguistic veneer that stands above our perceptual apparatus," he remarks. A paradox emerging from Winston's study of how humans think is that "computers make us stupid." For instance, when students are freed from taking notes, absence of "forced engagement" with the material hinders learning. He cautions that teachers confuse the "presentation of information with the delivery of information." Too many words on a slide (or talking too fast) "jams the language processor" and impedes digesting content. Winston summarizes with an appealing prescription for becoming smarter. "Take notesdraw picturestalk and imaginetell stories!" The very act of explaining to another elucidates a lesson for oneself. About the Speaker(s): Patrick Henry Winston has been affiliated with MIT for five decades: from undergraduate and graduate education, to faculty appointment, to deep involvement in the life of the Institute through numerous committee memberships. In 1967, he joined the Artificial Intelligence Laboratory, serving as director for 25 years, and continuing in the successor Computer Science and Artificial Intelligence Laboratory. Winston focuses on integration of vision, language, and motor faculties to explain intelligence. His current research, the Human Intelligence Enterprise, is an interdisciplinary confluence of computer science, systems neuroscience, cognitive science, and linguistics. He also pursues an interest in the intriguing field of "computational politics," uniting computer scientists and social scientists toward an enlightened understanding of thinking in many cultures. Beyond academia, Winston cofounded Ascent Technology, Inc., a company that develops A.I. applications in resource planning and scheduling for airports and the Department of Defense. He is in his third term on the Naval Research Advisory Committee, for which he studied how to utilize technological advances for an all"electric Navy. This work was recognized with a Meritorious Public Service Award. Winston is past president of the American Association for Artificial Intelligence. He has written or edited 17 books, including texts on programming languages and artificial intelligence, as well as anthologies of A.I. research. Host(s): Alumni Association, Alumni Association

"Ideal" Language Learning and the Psychological Resource Problem

Fri, 19 Oct 2007 04:00:00 GMT

10/19/2007 3:50 Wong AuditoriumWilliam Gregory Sakas, Associate Professor, Department of Computer Science, Hunter College; Ph.D. Programs in Computer Science and Linguistics, The Graduate Center, City University of New York Description: Some linguists study what can be learned in principle, but William Gregory Sakas asks "the feasibility question -- how efficient learning takes place." This talk focuses on such research, its historical antecedents, and issues that trouble Sakas and his colleagues. Sakas provides a swift conceptual survey of modeling parameter"setting, from Chomsky, through Yang and Lightfoot. In "the spirit of Pinker," Sakas believes that any computational model of a feasible learner must be compatible with the psychological resources of a human child. So Sakas's lab tries to zero in on "what is needed in the way of psycho"computational resources in a learner to converge on the target grammar on the basis of a limited sample of sentences." The lab has created "a large, artificial but linguistically motivated domain of parameterized languages for evaluating learning models," with more than 1.6 million parsed sentences. Sakas and his colleagues compare the efficiencies of different parameter"setting models, attempt to solve such modeling problems as noise and over"generalization, as well as evaluate richness of the stimulus claims. Underlying this work _ "the whole parameter setting enterprise" -- says Sakas, is the effort "to limit the resources to what can reasonably be attributed to young children," so as to reduce the complexity of innate knowledge -- generally limit the amount and complexity of input and processing of each sentence. The sticky problem remains of finding a "psycho"computationally palatable way" of modeling a process of fitting grammar to multiple sentences. "We feel alone in this endeavor," says Sakas. While there's lots of interesting recent work on modeling syntax acquisition, mathematical, and statistical/probabilistic learning, "we haven't been able to take it into our models, because it doesn't seem to be concerned with resource limits." The ideal learner other linguists discuss "is not a learner concerned with resource issues." Sakas asks if "this work is intended to mirror psychological reality." He notes, "The effective richness of the stimulus for a child language learner depends on the child's non"ideal capacity to extract information from heard utterances." The point is to model what a child really does with the linguistic input available to her. Sakas concludes with a question: "Is there anyone out there trying to solve the psychological resource problem? About the Speaker(s): William Gregory Sakas has been a professor at CUNY/Hunter College since September 2000. He has published a book chapter with J.D. Fodor in Language Acquisition and Learnabilityand co"authored a textbook, The Core Guide to PPL Programming. He received his A.B. from Harvard College in Economics and his Ph.D. in Computer Science from the City University of New York.Host(s): School of Engineering, Laboratory for Information and Decision Systems

Explorations in Language Learnability Using Probabilistic Grammars and Child"directed Speech

Fri, 19 Oct 2007 04:00:00 GMT

10/19/2007 12:50 PM Wong AuditoriumJoshua Tenenbaum, PhD '99, Paul E. Newton Career Development Professor, Department of Brain and Cognitive Sciences, MITDescription: How do kids manage to figure out that the word "dog" applies to a whole category of animals, not just one creature? Joshua Tenenbaum wants to understand how children and adults manage to solve such classic problems of induction. Throughout cognition, wherever you look, he says "we see places where we know more than we have a reasonable right to know about the world, places where we come to abstractions, generalizations, models of the world that go beyond our sparse, noisy, limited experience." Tenenbaum's goal is to come up with "general purpose computational tools for understanding how people solve these problems so successfully." He's creating a set of hierarchical, probabilistic models that will help explain how humans make inductive leaps _ how abstract knowledge that "guides and constrains our inferences" helps us acquire language from our earliest days. While his models can apply to many areas of cognition, Tenenbaum focuses on recent work with syntax. From very simple data, children manage to turn a complex declarative like "The girl who is sleeping is happy," to a complex interrogative: "Is the girl who is sleeping happy?" They don't say, "Is the girl who sleeping is happy?" Tenenbaum suggests that humans somehow identify the hierarchical phrase structure of language, and use this as an "inductive constraint to guide acquisition of a particular piece of syntax." Tenenbaum and his colleagues have built representative grammars using data from child"directed speech --2300 sentences that correspond to 20 thousand"plus utterances. He deconstructs these sentences so that each word is replaced by a syntactic category. "The baby bear discovers Goldilocks in his bed" becomes "det adj n v prop pre adj n." He's explored these grammars for their capacity to balance complexity, generalize appropriately, and ability to fit the data. His results indicate that "by having the right kind of inductive bias, the idea of hierarchical phrase structure, you can make generalizations which you have no evidence for" By probing what seem to be "innate domain general capacities," says Tenenbaum, "we're trying to formalize these arguments and use them as a tool to diagnose what has to be innate, or what is more or less plausibly part of Universal Grammar." Tenenbaum sees his use of statistical inference methods as bolstering classical linguistics in its attempt to map out how humans learn from real data, and helping devise machine systems that might approach the capacities of human learners. About the Speaker(s): Joshua Tenenbaum is also a member of the Computer Science and Artificial Intelligence Laboratory. He received his Ph.D. from MIT in 1999 and after a brief postdoc with the MIT AI Lab, he joined the Stanford University faculty as Assistant Professor of Psychology and (by courtesy) Computer Science. He returned to MIT as a faculty member in 2002. He currently serves as Associate Editor of the journal Cognitive Science, and he has been active on the program committees of the Neural Information Processing Systems (NIPS) and Cognitive Science (CogSci) conferences. He held a Howard Hughes Medical Institute (HHMI) Predoctoral Fellowship from 1993"1998, and won the Outstanding Paper Award, IEEE Conference on Computer Vision and Pattern Recognition, 1997, for "Learning bilinear models for two"factor problems in vision", with William T. Freeman. Host(s): School of Engineering, Laboratory for Information and Decision Systems

Have We All Been Right? Looking Backwards at Linguistic Theory, Statistics, and Language Acquisition

Fri, 19 Oct 2007 04:00:00 GMT

10/19/2007 4:45 PM Wong AuditoriumCharles Yang, SM '97, PhD '01, University of Pennsylvania; Jean"Roger Vergnaud, PhD '74, USC; Anna"Maria di Sciullo , University of Qu_bec; Norbert Hornstein, UMD; Robert Freidin, Princeton University; William Gregory Sakas, Associate Professor, Department of Computer Science, Hunter College; Ph.D. Programs in Computer Science and Linguistics, The Graduate Center, City University of New York Description: It was uncertain by the end of this panel if linguists and computational scientists could find meaningful common ground. As conference organizer Michael Coen initially stated, "The issues we're discussing are as religious to people as the Red Sox." The two disciplines view their shared territory in distinctive ways, leading, in this panel and subsequent discussion, to some friction. Moderator Charles Yang sums up the preceding talks, describing how presenters explored such issues as whether statistical models could adequately capture psychological and linguistic complexity, and whether the learning models fit the developmental data. He cites continued conundrums, such as "How does a child do something that is so apparently in contradiction with what's in the data," which he would like to see addressed in discussions of statistical learning of syntax. Robert Friedin comments, "What I noticed in the presentations of modelers was that syntactic representations put forward were not syntactic representations that I would accept. There is an assumption in linguistics that language has a particular syntactic structure and not another. If you have a theory of grammar that gives you the right set of syntactic representations, you might want to say, let's take that and now let's see what else do we need to add to explain other things on the periphery." Jean"Roger Vergnaud is "puzzled by the approach" of some models that look at the distribution of data for the purpose of inferring grammar. He says, "I think there is a problem with standard treatments that purport to derive phrase structure or consistent structure just from examining strings." Norbert Hornstein says, "I was amused that poverty of stimulus here was considered a problem." Many people in this conference looked at it as a thing to solve, and "in my part of the world, it's an extremely effective tool, not a problem -- a given, we know it exists." He said that computationalists "seem to think we're people who generate phrase structure grammars. Frankly these are peripheral issues." He notes that many syntacticians are interested in the nature of the initial state of the language faculty, and suggests it might be useful to ask how current statistical techniques could study this question. William Sakas repeats his request for "discussion about how statistical models might be scaled down to feasibly be embodied in a child." Anna Maria Di Sciullo says, "Probabilistic models have been said to be the models of language acquisition. If we look at human possession and acquisition of language, whether words, sentences or text, a human tends to have different behavior with respect to different sorts of structures." Also, children don't acquire language instantaneously, and instead go through a set of errors. She seems dubious that a model based on probability would be able to account for the kinds of nuanced patterns found in human language acquisition. The question and answer period includes some energetic exchanges among panelists and conference participants, including Josh Tenenbaum , Lila Gleitman, Chris Manning, Amy Perfors, and Partha Niyogi. About the Speaker(s): Charles Yang received his Ph.D. in Computer Science from MIT in 2000. Before his appointment at Penn, he taught linguistics and psychology at Yale University. His research and teaching interests include, language acquisition and change; morphology and the mental lexicon; computational linguistics; and the evolution of language and cognition, Most recently, he authored the book Infinite Gift: How Children Learn and Unlearn the Languages of the World. Host(s): School of Engineering, Laboratory for Information and Decision Systems

Human Simulations of Language Learning

Fri, 19 Oct 2007 04:00:00 GMT

10/19/2007 9:30 AM Wong AuditoriumLila Gleitman, Professor Emerita, Department of Psychology,University of Pennsylvania; Michael Coen, '91, SM '94, PhD '06, University of Wisconsin"MadisonDescription: This workshop, explains Michael Coen, is an effort to engender temperate, collaborative discussion of a matter that inspires hot dispute: whether machine learning helps explain how humans acquire language. In particular, says Coen, machine learning advocates believe they have evidence against Noam Chomsky's "poverty of stimulus argument," which in essence states that language is built into us, that "children don't receive enough linguistic inputs to explain linguistic outputs." Coen, who doesn't think much of such claims, worries about a deeper problem, that scientists have "begun to discuss engineering at the expense of science." He describes 13"year"old Bobby Fischer's astonishing match with a world chessmaster, where Fischer managed to look 16 moves ahead -- eliminating about 10 to the 30th board positions. We had no way to represent his thinking process then, and we don't today, although scientists have built a machine, Deep Blue, that can topple any human chess champion. It seems there's nothing left to say about chess, yet we know absolutely nothing about how humans play chess, says Coen. "If you're an engineer, this may be fine, but if you're a scientist, that's deeply troubling." One problem with machine models, says Lila Gleitman, is that "they don't try to learn what the human already knows," and we really aren't sure "how big a piece of the pie that is in the first place." Gleitman distinguishes between acquiring language, and acquiring *a* language, like French or German. In her years of researching how children learn language, and specifically children who have been deprived of linguistic input entirely, Gleitman does not find a blank slate: "Children don't just sit there; they start to make gestures." Gleitman reviews various studies that describe a basic sequence in language acquisition that holds true regardless of specific 'inputs.' If researchers make models that are to be "of any interest, they ought to take into account the fact that you may not have to learn some of this." Gleitman has conducted simulations with adults, giving them incomplete scenes on video or paper (dropping words or substituting Lewis Carroll type doggerel) to see how we acquire the meaning of common nouns and verbs through contextual clues and inference. The more sources of evidence people get in these tests, the better they do. But such language acquisition "doesn't scale up" to higher level categories of words," such as "think." Says Gleitman, "It's crazyto suppose there's no biological given in a language learning situation. There's plenty. Some of it is maybe the substance of language and some of that is about the sophisticated learning procedures themselves." So any kind of "informative statistical modeling requires a matrix of conspiring cues, intrinsically ordered in time of appearanceRealistic models of incremental learning will incorporate what the learner brings to the task." About the Speaker(s): Lila Gleitman's research has helped to define current investigations into language learning. She examines the mechanisms that drive the process of language learning. In her work, she looks for the biases children bring to the learning situation and the processes through which children extract meanings from the input they receive. Gleitman has served as president of the Linguistic Society of America and is a Fellow of the Society of Experimental Psychologists and the National Academy of Sciences. She co"founded the Institute for Research in Cognitive Science. Michael Coen received his S.B., S.M., and Ph.D. degrees from MIT, where his doctoral work on self"supervised machine learning received the Sprowls Dissertation Award. His primary academic interests are developing biologically inspired approaches to machine learning and reciprocally, to using these approaches to better understand learning in biological systems. Coen is the author of "A Similarity Metric for Spatial Probability Distributions",a CSAIL Technical Report, MIT, 2007, "Multimodal Dynamics: Self"Supervised Learning in Perceptual and Motor Systems," Ph.D. Dissertation, MIT 2006. Coen worked on Wall Street from January 2000 through May 2004.Host(s): School of Engineering, Laboratory for Information and Decision Systems

Machine Learning of Language from Distributional Evidence

Fri, 19 Oct 2007 04:00:00 GMT

10/19/2007 10:45 AM Wong AuditoriumChris Manning, Associate Professor of Computer Science and Linguistics, Stanford UniversityDescription: Christopher Manning thinks linguistics went astray in the 20th century when it searched "for homogeneity in language, under the misguided assumption that only homogeneous systems can be structured." In the face of human creativity with language, rigid categories of linguistic use just don't help explain how people actually talk and what they choose to say. For every hard and fast rule linguists find, other linguists can determine an exception. Categorical constraints rise, then come crashing down. Manning argues for acceptance of variable systems of language, and for searching for structure in these systems using probabilistic methods. Manning applies quantitative techniques to sentence structure, digging for the frequency, probability and likelihood that people will use specific turns of phrase in certain real"world contexts. Looking at distributions in the ways people express ideas in a language "can give a much richer description of how language is used." Indeed, Manning finds that certain typical constraints on sentence structure in one language "show up as softer constraints and preferences in other languages." Manning looks at raw data, like sentences from the Wall Street Journal, and gleans such information as typical word associations that begin to "tell us about the dependencies of verbs and arguments." He looks for dependencies between words, the distance between them, and at a sentence's flow from left to right. Classes of words emerge, and clusters, yielding distributionally learned categories. Certain classes of syntax naturally fall together. Manning builds nested phrase structure trees, and branching structures, and derives simple probabilistic models that help explain "gradual learning and robustness in acquisition, non"homogeneous grammars of individuals, and gradual language change over time." Manning says computational linguistics is also proving useful in such applied fields as information retrieval, machine translation, and text mining. About the Speaker(s): Christopher Manning's research concentrates on probabilistic models of language and statistical natural language processing, information extraction, text understanding and text mining, and other topics in computational linguistics and machine learning. Together with Dan Klein, he received the ACL 2003 best paper award. He received a B.A. in mathematics, computer science and linguistics from the Australian National University in 1989. He earned a Ph.D. from Stanford in Linguistics in 1995. He previously served as an Assistant Professor at Carnegie Mellon University in the Computational Linguistics Program and as a lecturer in the University of Sydney Department of Linguistics. Manning's "bestseller" is Foundations of Statistical Natural Language Processing, Manning and Schôtze,(MIT Press, 1999). Host(s): School of Engineering, Laboratory for Information and Decision Systems

Statistical Natural Language Parsing: Reliable Models of Language?

Fri, 19 Oct 2007 04:00:00 GMT

10/19/2007 3:15 PM Wong AuditoriumSandiway Fong, SM '86, PhD '91, Associate Professor, University of ArizonaDescription: The statistical natural language linguist owes much to the University of Pennsylvania's famous Treebank project. But this giant corpus of one million words _ actually, 49 thousand sentences from the Wall Street Journal all carefully labeled for their syntactic and semantic components -- is actually both a "blessing and a curse," says Sandiway Fong. This "gold standard" list of parsed sentences, the result of more than a decade of work, has become "the only game in town,"according to Fong. Linguists developing natural language algorithms often rely on the complex Penn Treebank to construct and train probabilistic, context"free grammars, and Fong acknowledges the Treebank's revolutionary impact on the field. But he also thinks it' sworthwhile to examine how systems that rely on Penn Treebank actually perform. He has been exploring three basic questions: Do such systems attain cognitively plausible knowledge of language, such as distinguishing between grammatical and ungrammatical components of sentences? How brittle are these systems, so that if you misspell a word or flip one part of the sentence, the system will "give you back some parse? Can these systems learn non"natural languages? Fong has unearthed some interesting issues. For instance, two well"known parsing systems couldn't score more than 50% figuring out the right way to pronounce the word "read" in eight sentences that deployed the past and present tenses (e.g., The girls will read the paper; The girls have read the paper). And the two systems didn't get the same sentences wrong. Fong wonders if "reading the Wall Street Journal is not a good way to learn how to pronounce 'read' or 'red.'" Fong also demonstrated that a parsing system could be turned on the presence (or absence) of a single example involving the phrase "milk with 4% butterfat," calling in question whether such systems are truly robust. While Treebank"based parsing systems demonstrably perform well on Treebank"like sentences, one cannot infer they have necessarily achieved grammatical competence nor linguistic stability. We must understand, says Fong, that 40 thousand training samples do not really provide enough parameters to provide the broad range of linguistic cases for computational systems that ordinary people pick up nearly effortlessly. "We expect statistical systems to be able to deal with noise. But they are extremely fragile, despite their statistical nature and training over a large data set." About the Speaker(s): Sandiway Fong received his B.Sc. in Computing Science, at Imperial College of Science and Technology, University of London. He received an S.M. in 1986 at MIT, where he worked in the Artificial Intelligence Laboratory. After working at IBM's Watson Research Center, he returned to MIT for his Ph.D. In 1991, he joined the NEC Research Institute to work on natural language processing, and machine translation. In 2003, he moved to the University of Arizona, where his research interests are at the intersection of computer science and formal linguistics, with a focus on multilingual parsing, ontolinguistics, computational lexical semantics and computational morphology.Host(s): School of Engineering, Laboratory for Information and Decision Systems

Structure Dependence, the Rational Learner, and Putnam's "Sane Person"

Fri, 19 Oct 2007 04:00:00 GMT

10/19/2007 1:20 PM Wong AuditoriumHoward Lasnik, PhD '72, Distinguished University Professor, Department of Linguistics, University of Maryland; Juan Uriagereka, Professor, University of MarylandDescription: Young children say many surprising and funny things _ funny, often, because how they say it is not quite right in an endearing way. "My friend goed to the playground," and "I ated two desserts" both demonstrate errors that we readily understand, sympathize with, and are confident will go away with further listening and speaking. But there are other kinds of errors that children just don't seem to make. In his pathbreaking work on transformational grammar, Noam Chomsky has written extensively about sentences like "The dog in the corner is hungry." By applying a formal operation Chomsky described in detail, we can form the question "Is the dog in the corner hungry?" But confronted with "The dog that is in the corner is hungry," we do not end up asking "Is the dog that in the corner is hungry?" Instead, we apply the transformational rule in a different, more complex way, to ask "Is the dog that is in the corner hungry?" Chomsky draws two conclusions from close study of many such cases. First, he says, this shows that the transformational grammar rules we follow are "structure"dependent," that is, they apply to phrases, not simply to a string of words in sequence. Second, because a person can go through life without recognizing or even encountering some structure"dependent cases _ and yet make the correct choice when presented with alternatives _ this aspect of grammar has deep implications for human psychology. In fact, Chomsky claims, this is an argument for the existence of invariant principles of language, a universal grammar. Howard Lasnik cites evidence for a different interpretation: Chomsky's "poverty of the stimulus" scenario may not be relevant. By examining a large collection of speech (drawn from the CHILDES database), and applying a Bayesian model of grammar induction _ making use, in other words, of the speaker's knowledge of prior probabilities _ it is possible to show that a rational learner could in fact learn that transformational linguistic rules depend on phase structure. Lasnik's former student, now colleague, Juan Uriagereka, broadens the argument. Drawing on a startling range of examples _ from animal behavior to protein folding, Uriagereka wonders if the structural properties of grammar are unique to human language, or extend to other forms of human cognition, including music, mathematics, and complex planning. Structure dependence may be true, it may be specific to language or at least to human thought but how did it get there? Where does structure come from? These are the bold questions Lasnik and Uriagereka believe that contemporary linguistic cognitive science has to address. About the Speaker(s): Howard Lasnik is the author of A Course in Minimalist Syntax(2005), with Juan Uriagereka, and of Minimalist Investigations in Linguistic Theory (2003). He previously taught at the University of Connecticut, and has been a Fellow of the Center for Advanced Study in the Behavioral Sciences. He is also a Fellow of the Linguistic Society of America. Lasnik received his M.A. In English at Harvard University in 1969, and his Ph.D. in Linguistics at MIT in 1972. Juan Uriagereka is author of Rhyme and Reason; and co"author of A course in GB Syntax: Lectures on Binding and Empty Categories; and editor of Derivations: Exploring the Dynamics of Syntax. He previously taught as a Visiting Professor at Konstanz University in Germany, and at Wolfson College, Oxford University. He has served as a Visiting Chair, Basque Philology, University of the Basque Country. Uriagereka received his Ph.D. in Linguistics from the University of Connecticut.Host(s): School of Engineering, Laboratory for Information and Decision Systems

The Biology of the Language Faculty: Its Perfection, Past and Future

Fri, 19 Oct 2007 04:00:00 GMT

10/19/2007 2:00 PM Wong AuditoriumNoam Chomsky, Institute Professor, MITDescription: Noam Chomsky, around whose work much of the Syntax series revolves, gives listeners a glimpse into the evolution of his own thinking, with an emphasis on areas of linguistics where computational considerations play a major role. Chomsky briefly outlines the key components of a biologically based linguistics that began to emerge 50 years ago: first, a genetic language endowment (Universal Grammar), which interacts with the external environment, and second, the individual's development and learning strategies. While UG has been called "controversial," says Chomsky, the "alternative is magic," since something has to account for the fact that "my granddaughter picked out part of her environment as language related, and almost reflexively developed a language while her pet kitten, a chimp or songbird, exposed to exactly the same data, didn't take the first step and couldn't conceivably take the second." Chomsky links a third factor of language involving architecture and the principles underlying data acquisition to natural laws that may apply generally in biology, and not specifically to language. Research suggests that between 50 and 100 thousand years ago, humans made an abrupt evolutionary leap forward in cognitive capacity. Language seems to have emerged at this time. While long"term evolution can lead to great complexity, a sudden leap like this, says Chomsky, tends to yield something "simple, almost perfect -- a perfect solution to design problems imposed by circumstances and conditions prevailing at the time of emergence..." This proposal has been dubbed the Strong Minimalist Theory (SMT), and offers a plausible approach to studying the complexity of language, believes Chomsky. It might prove profitable to "examine the range of phenomena that fall under what's loosely called language," and try to "disentangle them so some parts of them conform more or less to SMT." And here, says Chomsky, issues of computational efficiency play perhaps an overwhelming role. Chomsky links SMT to transformational grammar, a long"standing component of his linguistic theory. He states that "a simple form of transformational grammar is just the optimal system, and if you don't have it, you'd have to have an argument as to why you don't." Well"designed systems should have simple, sensible properties. He recommends "chipping away at the stipulated properties of Universal Grammar, and technologies proposed to deal with particular problems to see how closely you can show that language does approximate to the perfect design that would be a natural expectation in light of what appears to be evolutionary history." About the Speaker(s): Noam Chomsky has written and lectured widely on linguistics, philosophy, intellectual history, international affairs and U.S. foreign policy. A brief sampling of his prolific work includes: The Logical Structure of Linguistic Theory; Aspects of the Theory of Syntax; Language and Mind; American Power and the New Mandarins; Reflections on Language; Rules and Representations; Knowledge of Language; The Culture of Terrorism; Manufacturing Consent (with E.S. Herman); Understanding Power; Hegemony or Survival: America's Quest for Global Dominance; and most recently, Imperial Ambitions: Conversations on the Post"9/11 World, (with David Barsamian). Chomsky received his Ph.D. in linguistics from the University of Pennsylvania in 1955. He joined the staff of the Massachusetts Institute of Technology in 1955 and in 1961 was appointed full professor in the Department of Modern Languages and Linguistics. During the years 1958 to 1959 Chomsky was in residence at the Institute for Advanced Study in Princeton, NJ. In the spring of 1969 he delivered the John Locke Lectures at Oxford; in January 1970 he delivered the Bertrand Russell Memorial Lecture at Cambridge University; in 1972, the Nehru Memorial Lecture in New Delhi, and in 1977, the Huizinga Lecture in Leiden, among many others. Chomsky has received honorary degrees from universities around the world, and is a Fellow of the American Academy of Arts and Sciences and the National Academy of Science. Host(s): School of Engineering, Laboratory for Information and Decision Systems

The Computational Nature of Language Learning

Fri, 19 Oct 2007 04:00:00 GMT

10/19/2007 11:15 AM Wong AuditoriumPartha Niyogi, SM '92, PhD '95, Associate Professor, Department of Computer Science; Associate Professor, Physical Sciences Collegiate Division; Senior Fellow, Computation Institute, University of ChicagoDescription: Language learning research becomes more robust when it incorporates insights from evolutionary theory, Partha Niyogi demonstrates. The principles of natural selection and variation in a population come into play not only when exploring how children learn language but how languages alter over time. All languages are learnable and more or less uniformly learnable, says Nyogi: "It doesn't take 25 years to learn Chinese and two years to learn Bengali." But in the last 30"odd years, Nyogi says, there's been a great deal of debate about the most useful models of learning theory, with efforts to explain how the human language faculty makes use of linguistic input at different developmental stages. One model of language acquisition depicts the child, armed with a basic grammar "map," extrapolating from data (interactions with adults), and assembling the components of language by some algorithm. Analysis has been conducted as if there were "a target grammar, which produces data, and an algorithm which is trying to acquire this target grammar." But, says Niyogi, "that's not true in the world." A child is exposed to lots of variation from within its population; parents and others all produce different grammars, different data sets. Niyogi believes that an "evolutionary trajectory" links how acquisition happens at an individual level, and how variation in language springs up from one generation to the next. But rather than inheriting the grammar of your parents, you have to learn it. Examining language variation over time as if it were genetic variation, "you get a different mathematical structureand probabilities start playing an important role." Small differences "can have very subtle consequences giving rise to bifurcation in nonlinear dynamics of evolution." For instance, 1000 years ago, the English were speaking a language that's unrecognizable to us today. How has it come to be that "we have moved so far from that point through learning which is mimicking the previous generation?" Niyogi explains that within a single population two varying languages may be in competition (say, a German and an English"type grammar). While a majority may speak the dominant variant, some children will likely be exposed to a mixture of the two. There's a "drift" in language use, "and suddenly, what was stable becomes unstable." In the next generation, even more learners pick up the minority variant. It's possible to determine the probability of learners in successive generations using new expressions, and tracking the evolutionary transformation of language. The "ubiquitous fact of languages is that they change with time," concludes Nyogi, and "even a slight effect of frequency can wipe out something that looks stable." About the Speaker(s): Partha Niyogi studied Electrical Engineering at the Indian Institute of Technology, New Delhi. He also studied at the Laboratory for Computer Science at MIT, where he completed a master's thesis on speech recognition. He received a Ph.D. in learning theory in the Department of Electrical Engineering and Computer Science at MIT. He was briefly a postdoctoral fellow and research associate in the Brain and Cognitive Science Department, MIT, then worked for Bell Laboratories. He joined the the University of Chicago in 2000. Host(s): School of Engineering, Laboratory for Information and Decision Systems

Discourses on Iraq and the Middle East

Wed, 04 May 2005 04:00:00 GMT

Noam Chomsky, Institute Professor, MIT Description: U.S. actions in Iraq get a thorough thrashing in this final chapter of the Reconstructing Iraq series. First, Yosef Jabareen sprints through editorial page cartoons from Arab print media, which represent the U.S. as immoral, abusive, greedy and above all, hegemonic. The drawings depict George Bush burning the world or swallowing up Arab nations, and a map of Iraq morphs into a division of the U.S. energy department. One cartoon shows the United Nations watching passively as the globe commits suicide. In these images, Arab leaders are corrupt puppets of U.S. policy and Iraqi insurgents are brutally oppressed heroes. Noam Chomsky paints his own cynical picture of the conflict in Iraq. "The U.S. goal certainly had nothing to do with stopping atrocities," he says, and even less to do with advancing political freedom. "The U.S. promotes democracy when it's in our strategic and economic interests and opposes democracy when it's not." Chomsky continues, "It's almost inconceivable that the U.S. could permit a sovereign, democratic Iraq. The reasons are transparent." Iraq, he predicts, would form an alliance with Iran, helping foment Shiite rebellion in Saudi Arabia, leading to "a Shiite alliance controlling most of the world's energy." Even more worrisome, Iraq would "rearm and develop weapons of mass destruction as a deterrent." Chomsky notes, "The one thing the U.S. invasion taught everyone is you better have WMDs to protect yourself from U.S. attack." Poses Chomsky, "Would the U.S. sit by and allow this? The chances are zero." So contrary to our own "messianic vision" of implementing democracy, the U.S. will try to "run Iraq." Chomsky's alternative: pay Iraq billions in reparations for having supported Saddam Hussein, for years of painful sanctions, and hand the country over to the Iraqis as soon as possible. About the Speaker(s): Noam Chomsky has written and lectured widely on linguistics, philosophy, intellectual history, international affairs and U.S. foreign policy. A brief sampling of his prolific work includes: The Logical Structure of Linguistic Theory; Aspects of the Theory of Syntax; Language and Mind; American Power and the New Mandarins; Reflections on Language; Rules and Representations; Knowledge of Language; The Culture of Terrorism; Manufacturing Consent (with E.S. Herman); Understanding Power; Hegemony or Survival: America's Quest for Global Dominance; and most recently, Imperial Ambitions: Conversations on the Post-9/11 World, (with David Barsamian). Chomsky received his Ph.D. in linguistics from the University of Pennsylvania in 1955. He joined the staff of the Massachusetts Institute of Technology in 1955 and in 1961 was appointed full professor in the Department of Modern Languages and Linguistics. During the years 1958 to 1959 Chomsky was in residence at the Institute for Advanced Study in Princeton, NJ. In the spring of 1969 he delivered the John Locke Lectures at Oxford; in January 1970 he delivered the Bertrand Russell Memorial Lecture at Cambridge University; in 1972, the Nehru Memorial Lecture in New Delhi, and in 1977, the Huizinga Lecture in Leiden, among many others. Chomsky has received honorary degrees from universities around the world, and is a Fellow of the American Academy of Arts and Sciences and the National Academy of Science. Host(s): School of Architecture and Planning, Department of Urban Studies and Planning

Vietnam Remembered

Sat, 30 Apr 2005 04:00:00 GMT

In this bitter commemoration of the end of the Vietnam War, the speakers dispel any comforting notion that Americans have absorbed lessons from that bloody time, much less sought the truth.

Noam Chomsky & Elizabeth Spelke: The Idea of Universality in Linguistics and Human Rights

Tue, 15 Mar 2005 05:00:00 GMT

Noam Chomsky, Institute Professor, MIT; Elizabeth S. Spelke, Professor of Psychology Co-Director, Mind, Brain, and Behavior Inter-faculty Initiative, Harvard University. If humans have a common, in-born capacity for language, and for such complex behaviors as morality, might the faculties be somehow linked? Noam Chomsky perceives a mere thread of a connection. At breakneck speed, Chomsky leads us through a history of language theory, concluding with the revolutionary model he championed: a universal grammar underpinning all languages that corresponds to an innate capacity of the human brain. While scientists may now have a "clearer grasp of the universals of language," says Chomsky, notions of universality grow murky as we move "into domains of will, choice and judgment." Chomsky cites the 1948 Universal Declaration of Human Rights as one example of "broad cross-cultural consensus." But he brandishes examples of how "our moral and intellectual culture '.forcefully rejects universal moral judgments" -- such as continued U.S. refusal to approve anti-torture conventions. In contrast, Elizabeth Spelke forcefully links "universals in human nature to some of the developments in bringing about a greater balance in human rights." Thirty years of cognitive and cross cultural research show that humans universally structure their world in terms of objects, have a universal capacity to represent numbers, and to represent other people as "intentional, goal-directed agents whose freely chosen actions are subject to moral evaluation." Variation among humans flows from another universal capacity: to "freely combine concepts from different core systems." Spelke speculates that "humans might be gripped by a tremendous illusion that different members of different groups really are fundamentally different" _ an illusion that might drive us to conflict and rights abuses. These aspects of human nature pose a major challenge, but, Spelke concludes, a more fundamental faculty "holds the potential key to remedy" our capacity to "articulate deeply entrenched notions, criticize and get beyond them." ABOUT THE SPEAKERS: Noam Chomsky has written and lectured widely on linguistics, philosophy, intellectual history, international affairs and U.S. foreign policy. A brief sampling of his prolific work includes: The Logical Structure of Linguistic Theory; Aspects of the Theory of Syntax; Language and Mind; American Power and the New Mandarins; Reflections on Language; Rules and Representations; Knowledge of Language; The Culture of Terrorism; Manufacturing Consent (with E.S. Herman); Understanding Power (New Press, 2002); and most recently, Hegemony or Survival: America's Quest for Global Dominance (Henry Holt and Company, 2003). Chomsky received his Ph.D. in linguistics from the University of Pennsylvania in 1955. He joined the staff of the Massachusetts Institute of Technology in 1955 and in 1961 was appointed full professor in the Department of Modern Languages and Linguistics. During the years 1958 to 1959 Chomsky was in residence at the Institute for Advanced Study in Princeton, NJ. In the spring of 1969 he delivered the John Locke Lectures at Oxford; in January 1970 he delivered the Bertrand Russell Memorial Lecture at Cambridge University; in 1972, the Nehru Memorial Lecture in New Delhi, and in 1977, the Huizinga Lecture in Leiden, among many others. Chomsky has received honorary degrees from universities around the world, and is a Fellow of the American Academy of Arts and Sciences and the National Academy of Science. Before arriving at Harvard University, Elizabeth Spelke was a professor in the Department of Brain and Cognitive Sciences at MIT, and in the Department of Psychology at Cornell University. She is a Fellow of the American Association for the Advancement of Science, and a member of the National Academy of Sciences and the American Academy of Arts and Sciences.

Among her numerous honors, Spelke was among Time Magazine's America's Best in Science and Medicine. She received the William James Award from the American Psychological Society. Spelke earned her B.A. from Radcliffe College and her Ph.D. from Cornell University in 1978.
Host(s): Office of the Provost, Program on Human Rights and Justice Event date: 03/15/2005

Pinker's Farewell

Thu, 25 Sep 2003 04:00:00 GMT
09/25/2003 5:00 PM BartosSteven Pinker, Johnstone Family Professor of Psychology; Harvard University; ; Jay Keyser, Peter de Florez Emeritus Professor of Linguistics Description: In this personal and reflective event, Pinker looks back at twenty plus years at MIT and shares his deep appreciation for the place where "ideas and content always come first."Recalling his earliest work at the MIT Center for Cognitive Science, he describes the maddening problem of how children learn to use verbs correctly. You can splash the wall with paint and can splash paint on the wall; you can spill water on the floor but you can't spill the floor with water. Pinker theorized that children unconsciously divide the world of actions into categories like geometry and force, and that humans have evolved a grammar based on this intuitive physics. Pinker discusses Noam Chomsky's "enormous" impact on him, as well as his profound differences with Chomsky concerning the evolution of humans' innate ability to acquire language. In spite of jibes from outsiders (often journalists), Pinker says he reveled in teaching MIT's introductory psychology course. Finally, he describes many sleepless nights while pondering the "most agonizing choice of my career" his decision to leave MIT for Harvard. ABOUT THE MODERATOR: This discussion is moderated by Professor Samuel Jay Keyser--Peter de Florez Emeritus Professor at MIT and an emeritus member of the Linguistics and Philosophy faculty. He is currently special assistant to the Chancellor at MIT. Keyser's most recent book is The Pond God published by Front Street Books.About the Speaker(s): ABOUT THE MODERATOR: This discussion is moderated by Professor Samuel Jay Keyser--Peter de Florez Emeritus Professor at MIT and an emeritus member of the Linguistics and Philosophy faculty. He is currently special assistant to the Chancellor at MIT. Keyser's most recent book is The Pond God published by Front Street Books. ABOUT THE SPEAKER: Pinker is the Johnstone Family Professor of Psychology at Harvard University. He returned to Harvard in September 2003 after 21 years at MIT, where he was most recently the Peter de Florez Professor of Psychology in the Department of Brain and Cognitive Sciences and a MacVicar Faculty Fellow. A native of Montreal, he received his B.A. from McGill University in 1976 and his Ph.D. in psychology from Harvard in 1979. His scholarship has brought him awards and election to the American Academy of Arts and Sciences. Many more awards and worldwide recognition have come from several popular science books, including The Language Instinct, How the Mind Works, and most recently, The Blank Slate: The Modern Denial of Human Nature. Host(s): School of Humanities, Arts & Social Sciences, Communications ForumTape #: T17464

Words and Rules: The Ingredients of Language

Fri, 13 Jun 2003 04:00:00 GMT
06/13/2003 11:00 AM 3-170Steven Pinker, Johnstone Family Professor of Psychology; Harvard University; Description: Why does a three year-old say "I went," then six months later start saying "I goed"? When you first heard the word "fax," how did you know the past tense is "faxed"? And why is it that a baseball player is said to have "flied out," but could never have "flown out"? After fifteen years of studying words in history, in the laboratory, and in everyday speech, Steven Pinker has worked out the dynamic relationship _ searching memory vs. following rules _ that determines the forms our speech takes. In one of his final lectures at MIT Pinker gives the ultimate lecture on verbs, in a rich mixture of linguistics, cognitive neuroscience, and a surprising amount of humor. If you've ever wondered about the plural of Walkman, or why they are called the Toronto Maple Leafs and not Leaves, this lecture provides answers to these and other questions of modern language. About the Speaker(s): Pinker is the Johnstone Family Professor of Psychology at Harvard University. He returned to Harvard in September 2003 after 21 years at MIT, where he was most recently the Peter de Florez Professor of Psychology in the Department of Brain and Cognitive Sciences and a MacVicar Faculty Fellow. A native of Montreal, he received his B.A. from McGill University in 1976 and his Ph.D. in psychology from Harvard in 1979. His scholarship has brought him awards and election to the American Academy of Arts and Sciences. Many more awards and worldwide recognition have come from several popular science books, including The Language Instinct, How the Mind Works, and most recently, The Blank Slate: The Modern Denial of Human Nature.Host(s): School of Science, School of ScienceTape #: T16626

Recent Videos tagged 'Linguistics' on MIT Video

Unique languages, universal patterns

Samuel Jay Keyser - Interview No. 2 - Dec. 2, 2010

Samuel Jay Keyser - Interview No. 3 - Dec. 17, 2010

Celebrating Science and Engineering Breakthroughs III

Samuel Jay Keyser - Interview No. 1 - Sept. 22, 2010

20th Killian Award Lecture (1992) - Noam Chomsky, "Language: The Cognitive Revolutions"

Humanities in the Digital Age

Rebuilding Haiti

Computers with Commonsense: Artificial Intelligence at the MIT Round Table

"Ideal" Language Learning and the Psychological Resource Problem

Explorations in Language Learnability Using Probabilistic Grammars and Child"directed Speech

Have We All Been Right? Looking Backwards at Linguistic Theory, Statistics, and Language Acquisition

Human Simulations of Language Learning

Machine Learning of Language from Distributional Evidence

Statistical Natural Language Parsing: Reliable Models of Language?

Structure Dependence, the Rational Learner, and Putnam's "Sane Person"

The Biology of the Language Faculty: Its Perfection, Past and Future

The Computational Nature of Language Learning

Discourses on Iraq and the Middle East

Vietnam Remembered

Noam Chomsky & Elizabeth Spelke: The Idea of Universality in Linguistics and Human Rights

Pinker's Farewell

Words and Rules: The Ingredients of Language