Presentation on theme: "Women, Fire, and Dangerous Things Women, Fire, and Dangerous Things: What Categories Reveal About the Mind."— Presentation transcript:
Women, Fire, and Dangerous Things Women, Fire, and Dangerous Things: What Categories Reveal About the Mind
Why Care About Categories Categorization is basic to experience. We categorize things automatically in order to make sense of the world. The majority of our words and concepts represent categories. These terms, representing categories, are how users navigate websites.
What the PB Book Lacks.... A cohesive theory of how people form categories. Definitive principles on what makes a good category, regardless of context. What if we could apply the key findings from experiments throughout the history of category theory in order to predict how users will label the content in a website, saving time and money in usability research?
George Lakoff "Our brains take their input from the rest of our bodies. What our bodies are like and how they function in the world thus structures the very concepts we can use to think. We cannot think just anything - only what our embodied brains permit." Professor of Cognitive Linguistics at the University of California at Berkeley since Most famous for his ideas about the centrality of metaphor to human thinking, political behavior and society. Began his career as a student of Noam Chomsky at MIT.
Criticism “Lakoff’s book is a tremendous piece of scholarship and an intellectual achievement of the first order.” – Raymond W. Gibbs, Jr. “The latest offering from one of the foremost linguists of our time should not be missing from any library that claims to have holdings in the humanities, cognitive science, or education.” – Terence M. Odlin “Stop! The! World! For! A! Moment!” - Alexander Johannesen “Lakoff's Germanic thoroughness comes as no unmixed blessing; the lay reader, at least, may find interest in the author's ideas drowning in a sea of repetition and restatement...” – Raymond W. Gibbs, Jr.
Classical Categories Defined by shared properties. Have clear boundaries. Conceptualized as abstract containers with things either inside or outside. Reflect categories in nature. Set theory – reality characterized by abstract entities, properties of entities, and sets of entities. Categories formed only through union, intersection, and complementation of sets (Venn diagrams)
Limitations of Classical Categories According to classical category theory: No member of a category has any special status. All levels of a hierarchy are important and equivalent. Wittgenstein: central and non-central members Austin: some categories don't show family resemblances Berlin: a single level of classification (the genus) is psychologically basic because human capacities for perception are utilized in the same way.
Basic-Level Categories Basic in Four Respects: Perception: overall perceived shape; a single mental image; Gestalt. Function: interaction with the world. Communication: Shortest, most commonly used terms; first words learned. Knowledge Organization: Most attributes of category members are stored at this level. Categories that are cognitively basic are at the middle of the general-to- specific hierarchy.
Rosch Prototype Theory Classical theory also holds that: If categories are defined only by properties that all members have, then no member should be a better example of the category than another. But, Eleanor Rosch found that categories, in general, have best examples, called prototypes, and that human capacities – such as imagination – play a role in categorization.
Prototype Effects Assymmetries within a category Prototypes, stereotypes, or otherwise "typical" examples of things Demonstrated empirically (people are slower and less certain when categorizing "atypical" items (e.g., Is a penguin a bird? Is a lamp a piece of furniture?)) Imply not that some members are “less” members, but that category has additional internal structure, central to meaning and inexplicable by a (fuzzy) set theoretic model. o
Idealized Cognitive Model (ICM) Thesis: We organize our knowledge with complex structured wholes (gestalts) called Idealized Cognitive Models. Category structures and prototype effects are by-products from attempts to fit these ungraded ICM's to the world. Bachelors? men in long-term unmarried couplings? boy abandoned in jungle to mature in isolation? the Pope? Combine to form clusters more psychologically basic than individual models
Structuring ICMs Kinesthetic Image Schemas are concepts that metaphorically structure complex concepts (ICMs) ICMs + basic schemas explain category structures: radial (central member + extension principles) metaphoric (mapping from experiential to abstract domain) metonymic (part stands for whole) foster mother birth mother adoptive mother surrogate mother genealogical nurturance birth marital genetic
“More is up, less is down” In addition to structuring complex ICMs, image schemas are directly understood structures of their own Container, part-whole, link, interior-exterior, boundary, center- periphery, source-path-goal, up-down, front-back, linear-order Experiential basis Structure our experience preconceptually Metaphors map schemas to abstract domain, preserving logic Set theory = interior-exterior, boundary, container + basic logic + metaphors
Women, Fire, and Dangerous Things “On those remote pages it is written that animals are divided into (a) those that belong to the Emperor, (b) embalmed ones, (c) those that are trained, (d) suckling pigs, (e) mermaids, (f) fabulous ones, (g) stray dogs, (h) those that are included in this classification, (i) those that tremble as if they were mad, (j) innumerable ones, (k) those drawn with a very fine camel's hair brush, (l) others, (m) those that have just broken a flower vase, (n) those that resemble flies from a distance.”
Case Study: Dyirbal Classification Bayi: men, kangaroos, possums,bats, most snakes, most fishes, some birds, most insects, the moon, storms, rainbows, boomerangs, some spears,etc. Balan: women, bandicoots, dogs, platypus, some snakes, some fishes, most birds, some fireflies, scorpions, crickets, the hairy mary grub,anything connected with water or fire, sun and stars, shields, some spears, some trees, etc. Balam: all edible fruit and the plants that bear them, tubers, ferns, honey, cigarettes, wine, cake Bala: parts of the body, meat, bees, wind, yamsticks, some spears, most trees, grass, mud, stones, noises and language, etc.
Rethinking Category Theory (and Linguistics, Philosophy, AI, IA...) Categorization in language is more often radial and metaphorical than hierarchical, and experientially-based Regardless of category type, general principles persist: centrality chaining experiential domains idealized models specific knowledge the other no common properties motivation
PB Book on Organization & Labeling Organization schemes are often ambiguous, and should include topical access to content. Polysemy can make topical categorization difficult! Hierarchies form the basis for organization. Balances must be struck between: exclusivity and inclusivity breadth and depth When designing a labeling system, consistency is stressed, and various sources, including social tagging, are suggested for ideas.
PB Book on George Lakoff Make use of already-persistent image schemas and metaphors to structure navigation of websites (link, source-path-goal, up-down, center-periphery, part-whole) Construct basic-level categories (think gestalts!) Cleaner, more orderly, clean borders among categories may be drawn by choosing prototypical elements to represent categories Design navigation aids to move users to basic-level categories as soon as possible.
Example: University of Michigan Library
Jorge Luis Borges, Information Architect If more than one classification scheme exists for a given set, how do we choose one over the other? How do we evaluate the internal coherence of a given classification scheme? Classification schemes can't be evaluated abstractly, a priori. As miscellaneous categories do exist in our minds, even if their use in a navigation system may be problematic, choice and coherence need to be assessed in respect to an empirical paradigm: context, goals, users, the cultural climate from which the classification stems. Instead of top-down and bottom-up design models, try an up-and-down model where classification process starts in the middle, from the basic-level categories, grouping them in super-categories and splitting them in subordinate, more specific classes.