Gödel/Escher/Bach Symposium

registration (for lunch) is closed, but the lectures are open to everybody!

GEB Douglas Hofstadter's Pulitzer-winning 1979 book, Gödel, Escher, Bach - an Eternal Golden Braid, has been a major source of intellectual joy and scientific inspiration for many of us in the cognitive sciences. In fact, themes touched upon by GEB - logic, visual arts, music, paradoxes, computation, thinking - are also appealing to a much wider audience.

Certificate of solving the SOUND OF ONE HAND CLAPPING, by Hakuin The Cognitive Studies Program at Cornell is announcing a symposium focusing on interdisciplinary study of mind, as exemplified by the spirit of GEB. The symposium will include a presentation by Prof. Hofstadter, and will consist of popular lectures on a range of topics, matching the range of interests of the faculty and students associated with the Program:

Come and help us tune in to the sound of one hand clapping in January!

January 19, 2001, 9am-6pm, Uris Hall Auditorium

Presentations:

9:00-9:05 Introduction.
9:05-9:45 Zoltan Szabo (Philosophy, CU) Thinking of Things Not Thought of
9:45-10:25 Carl Ginet (Philosophy, CU) The Surprise Exam: Knowing Less by Knowing More
10:25-10:55 Break.
10:55-11:35 Kaushik Basu (Economics, CU) Paradoxes of Rational Behavior
11:35-12:15 Jon Kleinberg (Computer Science, CU) Computability: Three Millennia and Counting
12:15-1:45 Lunch.
1:45-2:25 Ronald Hoy (Neurobiology and Behavior, CU) Rhythm and Mode in the Music of Animals and People: some Messiaenic Ruminations
2:25-3:05 Carol Krumhansl (Psychology, CU) Cognition, hierarchies, and the music of J. S. Bach
3:05-3:45 Shimon Edelman (Psychology, CU) A Vision of Battlements: from Verbum, via Castrovalva, to Babel
3:45-4:15 Break.
4:15-5:45 Douglas Hofstadter (Cognitive and Computer Sciences, Indiana University) From Eye to `I': High-level Perception and the Emergence of a Self

Supplementary information:

 Zoltan Szabo. Berkeley had many arguments against the existence of matter. But he was prepared to set all of them aside and rest his case entirely on a simple and perplexing line of reasoning. It goes - roughly - like this: It is inconceivable that there could be things "without the mind", for if we were to conceive of such mind-independent things, then by this very conceiving they would exist in some mind, namely, in our own. The argument is a fallacy. But it is one of the most intriguing fallacies in the history of philosophy. I will argue that it can still teach us important things about the nature of thought.
Carl Ginet. This talk will present an analysis and solution of the Surprise Exam puzzle, derived from a lecture by Saul Kripke given at Cornell many years ago.
Kaushik Basu. In much of the social sciences and game theory, it is assumed that human beings choose rationally. Quite apart from the empirical validity of this assumption, questions have been raised concerning the internal consistency of the very concept of rationality. The talk will illustrate this problem with the help of examples and give some hints of ways out of some of these paradoxes of rational behavior.
Jon Kleinberg. Modeling what we mean by computation was a crucial step in the development of computer science as field; but it grew out of mathematical notions that reach back over the past two thousand years. A precise definition of computability has consequences that are stronger than one might have expected: that programs and data are essentially indistinguishable, that computers are a universal technology, and that many natural problems we wish to solve are uncomputable. More recently, these issues have been combined with the notion of efficient computation, leading to the surprising discovery that problems very similar at structural level can behave very differently when we study their computational tractability.
Ron Hoy. I propose to talk about animal communication sounds ("songs") and their perception by the animals themselves and by human observers. In the animal world, in singing insects and frogs, the temporal structure, or rhythm, seems to be privileged over complexity in the tonal dimension. In birds, tonal complexity is obvious, but rhythmic structure is also apparent. In both perception and "recognition" of acoustic communication signals, animals employ various neurosensory implementations that are both similar and different from our own. If we take an indifferent stance about their salience as communication signals but look at animal calls as musical sounds, how do they "stack up" against the"real" music of human culture? I will talk a little about how human observers have dealt with listening and coding what they believe to be the salient features of insect and bird "songs." Birdsongs were especially fascinating to Olivier Messiaen, an important 20th C. composer/music theorist, who devoted large amounts of time and thought to accurately transcribing bird songs and encoding his results into the instruments of the orchestra, including the piano, and incorporated them into his compositions, among them, "Catalogue d'oiseaux." Messiaen was also fascinated by the rhythmic complexities of the music of India, and these musings found their way into his musical language. I will consider Messiaen's work from my perspective of animal bioacoustics.
Carol Krumhansl. Hierarchies are manifest at many levels in the music of J. S. Bach - and they play a central role in how his music is represented in cognition. Tonal hierarchies enable listeners to orient quickly to the initial key and track modulations, that is, changes of key in the music. This process is called tonality induction, and it can be modeled with a self-organizing network that generates a dynamic spatial representation showing the strengths of different keys. Statistical distributions of tones support both tonal and metrical (rhythmic) hierarchies. Perceptual studies show their additive and independent influences on the formation of musical phrases. Finally, musical tension relates to hierarchies that express the relative structural stability of musical events, forging a possible link between musical structure and emotion.
Shimon Edelman. In GEB, hierarchically embedded formal structures, recursion, and multiple levels of interpretation (which often compete with each other) together weave a "golden braid" linking Escher's creations to those of Gödel (and Bach). The last of Escher's prints to appear in GEB is Verbum, a design that embodies a strict symmetry and the kind of intricate and precise figure-ground tessellation that became one of his trademarks. What can we learn from such formal structural characteristics of Escher's work about the perceptual modality that mediates our appreciation of visual art? The experience of several decades of vision research suggests that understanding our perception of the mountainous landscape of Escher's Castrovalva is much more of a challenge than understanding the perception of Verbum. This talk will draw upon that experience in surveying some of the central open issues in visual object recognition and scene interpretation.
Douglas Hofstadter. Toward the end of GEB, a sketch was given of an AI architecture for solving Bongard problems, which are visual-perception problems somewhat reminiscent of IQ-test problems; they involve low-level vision, high-level vision, abstraction, analogy-making, and some natural language, all of which, taken together, we refer to as "high-level perception". This rudimentary architecture for high-level perception was the germ for several sophisticated AI programs developed over the next 20 years in the Fluid Analogies Research Group at the University of Michigan and Indiana University; these programs include Copycat, Metacat, and Letter Spirit. One of the key lessons we have learned from these architectures' strengths and weaknesses is how crucial it is for a system to be aware of itself in many different ways, in order for its intelligence to feel at all human-like to an outsider. But the presence of a sufficient level of genuine self-awareness amounts to the presence of consciousness, or in other words, it entails the emergence of something that merits, at least to some extent, its usage of the word `I'. Thus a sufficiently profound eye entails the existence of an `I'.

BACK to top

Shimon Edelman <se37@cornell.edu>
Last modified on Thu Jan 18 07:45:30 2001