Concepts Flashcards
(25 cards)
ROSCH 1973
Typicality ratings method - eliciting participants ratings of the typicality of particular instances of a category. She provided evidence for an internal structure of categories in which typicality seems to play a central part (for example - how typical a fruit is an apple?). She found high agreement on typical examplars, frequency of words was not an issue, task was natural and easy.
CLASSICAL VIEW
All members of a category must possess all of the common properties to satisfy defition of the category (having common properties is both necessary and sufficient for categorization). Concepts provide definitions of their corresponding category.Criticism of this view include: typicality, borderline cases, intransivity of categorization and the lack of definitions. However it gives a account of categorization when we have to appeal to definitions (i.e. court cases)
TYPICALITY
Criticism of classical view of concepts. Geometric definitions are clear cut, but everyday concepts are difficult to specify in terms of defining properties. I.E. furniture - there is no combination of properties that refers to all chairs. Functional, rather than perceptual properties play a part in defining them (i.e. is doll-house chair a chair?). If typicality reflects graded membership (some definite members, some def. non-members and some in between cases) then classical view cannot be correct. Classical view does not explain typicality, it would need supplementing to include it
ROSCH AND MERVIS (1975)
Property-, attribute-listing method. Participants were asked to generate lists of properties for a series of category instances (robin and penguin). Less typical instances (penguin) shared fewer characteristics with fewer category members. It shows that categories have rich internal structure
RIPS et al (1973), ROSCH (1975)
Category/sentence verification - measures time it takes participants to verify sentences that express categorization judgements. Faster RT for more typical exemplars, faster RT for nearer on hierarchy, fewer errors for more typical exemplars.
SMITH, SHOBEN & RIPS (1974)
Feature comparison model - criticism of Collins and Quillian’s (1969) hierarchical network model (the basis of classical view of concepts). S, S & R showed that distance in the hierarchy doesn’t always predict RT, as there was variation in response times across same no of levels i.e. participants responded faster to ‘is chicken a bird’ than to ‘is chicken an animal’, but slower to ‘is dog a mammal’ than to ‘is dog an animal’.
COLLINS AND QUILLIAN (1969)
Network model - concepts are represented as hierarchies of interconnected concept-nodes. Some concepts are superordinate (i.e. food) or other, subordinate (i.e. fruit, veg) concepts. Subordinates inherit properties from superodinate + have their own. General, common features are stored at higher level. Exceptions (ostrich -can’t fly) at lower. Methodology used to create this model is sentence verification - it takes longer to verify sentences that require two levels than one (is canary an animal takes longer than is canary a bird)
MCCLOSKEY AND GLUCKENBERG (1978)
Categorization judgement method - respond yes or no to category membership questions & rate them for typicality. Participants agreed between eachother and remained convinced over time on highly typical and atypical instances, but disagreed on some items of intermediate typicality i.e. is bookend furniture. Evidence of BORDERLINE CASES in categories. However the pattern of disagreement might reveal a lack of knowledge (i.e. a tomato is a fruit)
HAMPTON (1982)
Intransivity of categories - car seats are a kind of a chair - chair is a kind of furniture - car seats are not furniture
WITTGENSTEIN
Many categories have no defining characteristics - ie. board games, card-games (solitaire), ball games, Olympic games - form a family, have overlapping set of family resemblances
PROTOTYPE VIEW
Another similarity view. Prototype is the central tendency of a category. Whether or not an instance is an category exemplar depends on how similar it is to the prototype. Unlike in classical view, an instance might fall into a category if it mismatches the prototype in several characteristics. Some features are more important than others though (CUE VALIDITY). Informative and distinctive features have high cue validity. Typicality is ‘degree of similarity’ to prototype represented by the number of high cue validity features. This approach is consistent with fuzzy boundaries of categories.It explains for example RT to ‘is robin a bird’ and ‘is penguin a bird’ - the criteria for similarity to the prototype are likely to be matched with just a few properties for robin and the most attributes matched will have higher cue validity features. For low typicality instance such as penguin, many attributes will mismatch or be low weighted so more matches (longer RT) will be required to reach the criterion. Prototype theory may be useful in explaining categorization when it needs to be done rapidly, or under uncertainty, maybe when only superficial features are available
ARMSTRONG et al (1983)
Used definitional concepts such as ‘even numbers’, ‘odd numbers’, ‘female’ and ‘plain geometry figure’. Even with these concepts which have definite answers, they found some instances were rated as more typical. They proposed a ‘dual process model’ - concepts posses a core, used to judge membership (classical view) and a set of identification procedures used to identify instances of category on particular occasion - often rapidly (prototype view).
MEDIN AND SHOBEN
Instability of typicality. Participants rated metal spoons as more typical than wooden spoons, and small spoons more typ. than large spoons. however small metal spoons were rated as less typical than large wooden spoons - typicality is context dependent. Thus Rosch’ claim that prototypes correspond to a stable cluster of correlated features cannot be true. Also contributions to typicality of different properties (size and material) are mutually dependent.
COMMON-SENSE THEORIES
Categorization involves large knowledge structures called theories. Useful when a more reflective and considered judgment is required, i.e. explaining something
MURPHY & MENDIP (1985)
Plums and lawnmowers can be seen as having a lot in common, but don’t belong to the same category. Therefore for similarity, as in sharing properties, there must be a knowledge of which properties are meaningful for categorization - common sense ‘theories’ of categorization - it depends on dimensions of comparison. In their view concepts are explanation-based.
RIPS (1989)
Disassociation of similarity and categorizations. 3 objects presented to p. - a pizza (can have variable size), a quarter (fixed size) and asked to guess what the third object - medium sized - was most likely to be (pizza), and what is most similar to - a quarter. The two judgments were disassociated. BUT the choice was forced & p. might think it was neither of the two
KROSKA & GOLDSTONE (1996)
Emotions as concepts. Complex scenarios including phrases ‘threat of harm or death’ (central of FEAR), ‘being accepted or belonging’ (charac. of JOY) and ‘experiencing highly pleasurable stimuli or sensations’ (charac. of JOY). Part. categorized the scenario as an instance of fear (member of the ‘fear’ category) but more similar to instance of joy. Category membership was influenced by properties considered central to a category, judgments of similarity - by characteristic properties. Judgements of category membership can dissociate from judgement of similarity.
KEIL (1989)
Categorization in development. Hybrid animals - looks like a zebra, but both parents were horses and has horse’s insides: 4yr olds - zebra, based on superficial characteristics (categorizing based on appearance), 7 yr olds - horse, based on its biologically relevant properties (lineage). So called characteristic defining shift. BUT it’s possible younger children don’t know enough biological categories to work out which properties are characteristic and which are defining.
PSYCHOLOGICAL ESSENTIALISM
Explanation-based view, category members have essential properties, however most people will not know what they are, but will still believe that there are some. Speculation and vague idea of what these properties might be, a ‘place-holder’ for essence. It might be useful when explaining how we categorize when we want to be consistent with expertise and a scientific knowledge of the world.
GELMAN & WELLMAN
Even 4 & 5 yr old children believe the insides are more important than their outsides in determining category membership (not what Keil found!). According to G&W children are essentialist - they believe something hidden, something internal is responsible for category membership
MALT (1994)
People’s concept of water - if H2O is the essence of water than their categorization of liquids as water should be influenced by how much H2O is in them. However Malt found people’s categorization was influenced by the source of water, its location and its function. I.e. pond water was judged to be only 78.8% H2O and tears were judged not to be water but contain 88.6 H2O. Categorization depends on the context and the goals of the perceiver.
BRAINSBY et al (1996)
Tibby, and all the other cats, was discovered to be a robot controlled from Mars. Tibby is a cat though we were wrong about her being a mammal (essentialist). Tibby is not a cat though she’s a robot controlled from Mars (non essentialism). Half of participants thought the first statement was true, the other false. Many thought both were true or both were false. Concepts change content according to perspective and context.
MALT and in-between objects
Participants presented with objects, were told were inbetween the two categories, i.e. between oak and maple, or boat and ship. to decide how would they decide which category an object belongs to they could: ask an expert, call it whichever you want, tell which one it is if they had long enough to think about it. For natural categories, such as robin - sparrow, 75% would ask an expert. People might therefore be essentialist for natural categories as they recognize experts might be best equipped to categorize. For artefacts such as boat-ship 63% said it was ok to call it whichever you liked.
SMITH & SLOMAN
Duplicated Rips expriment (with a quarter and a pizza). They found the same results (dissociation) only when participants were asked to think aloud whilst making their decision and articulate reason for their judgements. Smith and Sloman interpreted these results as showing there are two modes of categorization - similarity-based and rule-based. People will focus either on similarities or on underlying rules, depending on how the task is presented. Evidence for changes to categorization depending on the purpose.
