Hampshire College

Fall 2008 CS Wednesday Talks

December 3rd
Joanna Morris, associate professor of cognitive science
How to opticize revelment without engendering legalous victimity: an ERP examination of complex word formation.
In this talk I will present evidence about how the brain recognizes complex words. Do we decompose complex words into their component parts in order to recognize them and reassemble them in production? Or are they stored intact in our mental dictionary? In this study I examined event-related potential responses to simplex targets (e.g., work) preceded by masked primes in which the target was embedded. Primes were (a) true derivations of the target (worker), (b) non-words consisting of the target and a non-morphological ending (workel), (c) morphologically complex pseudowords consisting of an illegal combination of the target and a suffix (workness), and (d) unrelated primes (musical). I use these data to test two models of the lexicon: models that explain facilitation among morphological relatives by means of morpho-semantic lexical representations and those that do so by positing sub-lexical morpho-orthographic representations.

Joanna Morris, assistant professor of cognitive science, holds a Ph.D. in psychology from the University of Pennsylvania. She completed the M.Phil. at Oxford University as a Rhodes Scholar, and an M.A. at Penn and B.A. at Dartmouth, both in psychology. She was a recipient of a pre-doctoral fellowship from Penn's Institute for Cognitive Science. She is interested in the structure of lexical representations  and in applying ERP and chronometric techniques to the investigation of cognitive phenomena. Her thesis work investigated the relationship between orthography and lexical stress in English. Her current research is focused on examining the nature of morphological representation and using ERP methodology.

November 19th
Lee Spector, professor of computer science
Past and Future Origins of Artificial Life
In this short talk I will discuss research on the origin of life that has been conducted using the tools and methodologies of computer science. I will briefly review some of artificial life's greatest hits and present some of my own recent work on the open-ended and "from scratch" evolution of computer programs and virtual life forms.


November 12th
Piercarlo Valdesolo, Amherst College Keiter Fellow in Psychology
Beyond Emotion and Reason: The Social Function of Morality
"In-group morality" has been posited as a fundamental moral intuition. Little is known, however, regarding the nature of these moral intuitions subserving group-level social identities and how they might contribute to ingroup cohesion as well as intergroup discord. The present studies provide an initial examination of these phenomena, revealing the underlying processes as well as contextual sensitivities of moral judgments regarding transgressions against ingroup members as well as transgressions committed on the part of ingroup members.

By designing a methodology inducing in vivo moral transgressions in the lab we have identified a fundamental bias in moral judgment, which we term moral hypocrisy. Moral hypocrisy, as we have demonstrated, results when individuals evaluate their own transgressions much more leniently relative to others' identical transgressions. We have shown that this hypocrisy readily extends to the group level, affording even tangentially similar others the same moral latitude with which we judge ourselves. We also present evidence suggesting that this phenomenon does not result from self-serving automatic intuitions, but rather from the effect of self-serving motivated reasoning which operates in direct competition with selfless intuitions.

Using the same paradigm, we have studied reactions to transgressions against ingroup members, revealing a higher sensitivity to such actions relative to neutral others. We present evidence showing that this sensitivity is directly driven by enhanced feelings of empathy for ingroup victims, and motivates not only harsher judgments against transgressors but also significantly more costly prosocial behavior.


November 5th
Sarah Partan, assistant professor of evolution and cognition, Andrew Fulmer, and Maya Gounard
Upgrading a Robotic Squirrel for Studies of Massachusetts Squirrel Behavior
Four years ago the results of the human presidential race may have been tight in Florida, but in the rodent world, squirrels responded unambiguously to Rocky the robotic squirrel. "Rocky" is a mechanical squirrel that we designed to help us understand the behavior of live gray squirrels studied in St. Petersburg, Florida. Fast-forward to the present, where we are upgrading Rocky for a study of Massachusetts squirrels. We are interested in how the squirrels use acoustic and visual signals to transmit information regarding danger. Does one signal component, such as a vocalization, suffice to signal danger, or is a visual component also required? Many animals use multiple signal components when they communicate, but we are only beginning to understand how these components work together to create a complete message. We will discuss preliminary findings on Massachusetts squirrels, as well as our recent upgrades to the mechanical model.

Sarah Partan is an assistant professor of animal behavior at Hampshire. She and her students study the communication behavior of mammalian subjects using observation, video playback, and presentation of mechanical subjects as stimuli. Two of the Div II students who are currently working on the squirrel project, Maya Gounard and Andrew Fulmer, will be co-presenting with her.


October 29th
Carol Trossett, director of institutional research
Measuring Institutional Culture: Hampshire College seen through Grid-Group Theory
Grid-Group theory is an anthropological theory of social organization. Trosset applies this theory to Hampshire students and the culture of Hampshire College, using two sets of recent student survey data. She also places Hampshire in the context of data from other liberal arts colleges.

Carol Trosset (B.A. Carleton College, Ph.D. University of Texas at Austin) is Hampshire's director of institutional research. She is a cultural anthropologist and has held faculty positions at several other colleges and universities. Before coming to Hampshire she directed institutional research at Grinnell College.

October 22nd
Jay Trudeau, visiting assistant professor of psychology
Pain: It Hurts. But Is It Really All In Your Head?
The medical and psychological professions often have radically different approaches to the problem of pain.  Physicians typically approach pain as an issue of objective physical sensation, while psychologists see it as a subjective perceptual experience. The reality is that pain, as most of us understand it, is a complex interaction of mental and physical processes. This talk addresses where pain comes from, what cognitive factors frequently modulate painful experiences, and how we should measure it.

Jay Trudeau is a Hampshire alum with a Ph.D. in experimental psychology from the University of Connecticut. His research in cognitive psychology is primarily in perception (visual and painful) and psycholinguistics (reading models). He is a visiting professor of psychology at Hampshire College and the director of research psychology for a private research firm, Analgesic Research.


October 15th
Jonathan Westphal, visiting professor of philosophy
How to Measure Time
How can we measure time, since none can be found to be measure? The past is gone, the future doesn't exist yet, and the present is vanishingly small? I give an answer to this question of Wittgenstein's, applying his own commonsense techniques, which he seems not to have been able to deploy on this example. He called the problem one of "extreme difficulty," but I will suggest it is so easy as to seem correspondingly difficult. Like Mozart's music, the philosophical problem of the measurement of time is too difficult for beginners and too easy for experts.

Jonathan Westphal’s work in philosophy has centered on questions at the intersection of philosophy of mind; metaphysics; philosophy of science; logic and philosophy of language; and aesthetics. He wrote a ground-breaking book on the philosophy of color, but has recently become interested in issues in the philosophy of time, and in the understanding of human freedom, particularly the weakness of arguments for determinism. In the history of philosophy, he has published work on Wittgenstein and Leibniz. He has also published on optical logic and diagrammatic logic systems. Since 2004 he has directed a $2M/year federal project on conservative optical logic devices.

October 8th
Gary Marcus, NYU professor of psychology
Kluge: The Haphazard Construction of the Human Mind
In fields ranging from reasoning to linguistics, the idea of humans as perfect, rational, optimal creatures is making a comeback – but should it be? Hamlet’s musings that the mind was “noble in reason ...infinite in faculty” have their counterparts in recent scholarly claims that the mind consists of an “accumulation of superlatively well-engineered designs” shaped by the process of natural selection (Tooby and Cosmides, 1995), and the 2006 suggestions of Bayesian cognitive scientists Chater, Tenenbaum, and Yuille that “it seems increasingly plausible that human cognition may be explicable in rational probabilistic terms and that, in core domains, human cognition approaches an optimal level of performance,” as well as in Chomsky’s recent suggestions that language is close “to what some super-engineer would construct, given the conditions that the language faculty must satisfy."

In this talk, I will argue that this resurgent enthusiasm for rationality is misplaced, for three reasons. First, I will suggest that recent empirical arguments in favor of human rationality rest on a fallacy of composition, implicitly but mistakenly assuming that evidence of rationality in some (carefully analyzed) aspects of cognition entails that the broader whole (i.e. the human mind in toto) is rational. In fact, establishing that some particular aspect of cognition is optimal (or perfect, or near optimal) is not tantamount to showing that the system is a whole is; current enthusiasm for optimality overlooks the possibility that the mind might be suboptimal even if some (or even many) of the components of cognition have been optimized. Second, I will argue that there is considerable empirical evidence (most well-known, but rarely given due attention in the neo-Rationalist literature) that militates against any strong claim of human cognitive perfection. Finally, I will argue that the assumption that evolution tends creatures towards rationality or “superlative adaptation” is itself theoretically suspect, and ought to be considerably tempered by recognition of what Stephen Jay Gould called “remnants of history,” or what might be termed evolutionary inertia.

I will close by suggesting that mind might be better seen as what engineers call a kluge: clumsy and inelegant, yet remarkably effective.

October 1st
Wm. Josiah Erikson, systems support specialist
Providing top-shelf services for rock-bottom prices, or: how do we have all this cool stuff?
In this talk, I will explore the many ways that I and others at Hampshire use whitebox hardware and free software; custom-written solutions; reused and recycled or handbuilt products where appropriate; and also top-of-the-line, rather expensive hardware in other places, as well as where it actually makes sense to do so and where it just happened. The purpose of this talk will not be to bore you to death with minute details of the incredibly geeky stuff I do (though there will be some of that; this is a CS talk after all), but to give you an overall picture of what I am trying to provide and the philosophy behind it, as well as the how. Question and answer period to follow.