Interdisciplinary workshop on Generics
Provisional Workshop Program
- 8:30 - 9:00 Breakfast
- 9:00 - 9:15 Introductory remarks (Michael Henry Tessler)
- 9:15 - 10:00 Michael Henry Tessler (Stanford, Psychology)
Communicating generalizations in computational termsGeneralizations are the foundation of abstract thought and are naturally formalized in Bayesian models of cognition. The language of generalizations (e.g., Birds fly.), however, has received comparatively little attention by formal psychological models, despite its ubiquity in everyday discourse and child-directed speech. One major issue in formalizing genericity is that such language is extremely flexible in its usage. Using a state-of-the-art probabilistic model of pragmatic reasoning, I explore the hypothesis that the meaning of such linguistic expressions is simple but underspecified, and that general communicative principles can be used to establish a more precise meaning in context. To test this theory, I examine 3 case studies: generalizations about events (e.g., John runs), causes (e.g., The block makes the machine play music.), and categories (i.e., generic language, e.g., Birds fly), and find that the model explains both the endorsement and interpretation of generalizations in language. The model suggests that central phenomena of genericity emerge from the interaction of diverse prior beliefs about properties with general communicative principles.
- 10:00 - 10:45 Rachel Sterken (Oslo, Philosophy)
- 10:45 - 11:15 Coffee break
- 11:15 - 12:15 Andrei Cimpian (NYU, Psychology)
Generics and StereotypesStereotypes are typically defined as beliefs about groups, but this definition is underspecified. I will ask whether stereotypes are better characterized as generic beliefs about groups or as quantified beliefs. In addition to clarifying the cognitive structure of stereotypes, the answer to this question bears on the longstanding debate about stereotype accuracy: Whether stereotypes can be said to be accurate depends in part on what sorts of beliefs they turn out to be.
- 12:15 - 1:00 Lunch
- 1:00 - 2:00 Student and Postdoc talks
- Arun Chaganty (Stanford, Computer Science)
How much is 131 million dollars? Putting numbers in perspective with compositional descriptionsHow much is 131 million US dollars? While this is a precise number, it is often difficult for people to comprehend the scale of large (or small) absolute values. Rather, studies have shown that providing anchored comparisons, or perspectives, such as "$131 million is about the cost to employ everyone in Texas over a lunch period" significantly improves comprehension when measured in terms of memory retention or outlier detection, despite being gross generalizations. In this talk, we will discuss the role that perspectives can play in the understanding of numbers and present a computational approach to generating said perspectives using concepts of familiarity, numeric proximity and semantic compatibility.
- Nadya Vasilyeva (Berkeley, Psychology)
Structural interpretation of category propertiesA large body of research has documented pervasive internalist biases in reasoning about social categories: observable features of category members are often thought to stem from deep, stable, inherent properties. We examined one potential alternative to internalist thinking, which we call "structural thinking," drawing upon an emerging literature in philosophy on structural explanation. A hallmark of structural thinking is locating the object of explanation within a larger structure and identifying structural constraints that act on components of the structure to shape the distribution of outcomes for each component. For example, a structural explanation for women's underrepresentation in STEM might appeal to women's socioeconomic role as opposed to properties intrinsic to women. We provide initial data supporting the existence and unique profile of structural thinking in adults and children, and discuss implications of these findings for our understanding of categorical representation.
Ellie Chestnut (Stanford, Psychology) TBA
- Arun Chaganty (Stanford, Computer Science)
- 2:00 - 3:00 Susan Gelman (Michigan, Psychology)
This time it's personal: What "you" reveals about generic conceptsPrior psychological work on generics has focused on full NPs with category labels (e.g., "Lions have manes"; "A bird lays eggs"). In the social domain, however, some languages express generics with a 2nd-person pronoun ("You win some, you lose some"; "You eat ice cream with a spoon"; "You never know what might happen"), using you to refer to a broad social group (people in general). In this talk I focus on why a word that is typically characterized by specificity, context-dependence, and focus on the addressee is also used generically to convey broad generalizations that extend beyond time or place. I will discuss a series of experiments with adults and young children (2-10 years of age) suggesting that generic concepts are deeply linked to norms from early in development, and can be used to make meaning out of personal experience.
- 3:00 - 3:30 Coffee break
- 3:30 - 4:15 Bernhard Nickel (Harvard, Philosophy)
- 4:15 - 5:00 Discussion Everybody
Co-organised by Michael Tessler (Stanford, Psychology), Bernhard Nickel (Harvard, Philosophy) and Rachel Sterken (Oslo, Philosophy).