The horizontal dimension
The horizontal dimension of the categorisation system (recall Figure 8.1) relates in particular to the principle of perceived world structure which we introduced earlier. This principle states that the world is not unstructured, but possesses correlational structure. As Rosch points out, ‘wings correlate with feathers more than fur’ (Rosch 1978: 253). In other words, the world does not consist of sets of attributes with an equally probable chance of co-occurring. Instead, the world itself has structure, which provides constraints on the kinds of categories that humans represent within the cognitive system.
One consequence of the existence of correlational structure in the world is that cognitive categories themselves reflect this structure: the category prototype reflects the greater number of correlational features. Recall that categories often exhibit typicality effects, where certain members of the category are judged as ‘better’ or more representative examples of that category than other members. Members of a category that are judged as highly prototypical (most representative of that category) can be described as category prototypes. This feature of category structure was investigated in a series of experiments reported in Rosch (1975), which established that prototypical members of a category were found to exhibit a large number of attributes common to many members in the category, while less prototypical members were found to exhibit fewer attributes common to other members of the category. In other words, not only do categories exhibit typicality effects (having more or less prototypical members), category members also exhibit family resemblance relations. While for many categories there are no attributes common to all members (not all members of a family are identical in appearance), there is sufficient similarity between members that they can be said to resemble one another to varying degrees (each having some, but not all, features in common).
Goodness-of-example ratings
In order to investigate the prototype structure of categories, Rosch (1975) conducted a series of experiments in which subjects were asked to provide goodness-of-example ratings for between fifty and sixty members of each category, based on the extent to which each member was representative of the cate gory. Typically, subjects were provided with a seven-point scale. They were asked to rate a particular member of the category along this scale, with a rating of 1 indicating that the member is highly representative, and a rating of 7 indicating that the entity was not very representative. Presented in Table 8.8 are the highest- and lowest-ranked ten examples for some of the categories rated by American under graduate students. It is worth observing that the experiments Rosch employed in order to obtain goodness-of-example rating were ‘linguistic’ experiments. That is, subjects were presented with word lists rather than visual images.
Family resemblance
Rosch argues that prototype structure, as exhibited by goodness-of-example ratings, serves to maximise shared information contained within a category. As Rosch puts it,‘ prototypes appear to be those members of a category that most reflect the redundancy structure of the category as a whole’(Rosch 1978:260). In other words, the more frequent a particular attribute is among members of a particular category, the more representative it is. The prototype structure of the category reflects this ‘redundancy’ in terms of repeated attributes across distinct members, or exemplars. This entails that another way of assessing prototype structure is by establishing the set of attributes that a particular entity has (Rosch and Mervis 1975). The more category-relevant attributes a particular entity has, the more representative it is.
In order to investigate this idea, Rosch and Mervis (1975) presented twenty subjects with six categories: FURNITURE, VEHICLE, FRUIT, WEAPON, VEGETABLE and CLOTHING. For each category, the experimenters collected twenty items that were selected to represent the full goodness-of-example scale for each category, from most to least representative. The subjects were each given six items from each category and asked to list all the attributes they could think of for each item. Each attribute then received a score on a scale of1–20, depending on how many items in a category that attribute had been listed for: the attributes that were listed most frequently were allocated more points than those listed less frequently. The degree of family resemblance of a particular item (for example, CHAIR in the category FURNITURE) was the sum of the score for each of the attributes listed for that item: the higher the total score, the greater the family resemblance. Rosch and Mervis’s findings showed a high degree of correlation between items that received a high score and their goodness-of example ratings. Table 8.9 illustrates these ideas by comparing some of the attributes common across the category BIRD against two members of the cate gory: ROBIN (judged to be highly representative) and OSTRICH (judged to be much less representative).
This table illustrates that the number of relevant attributes possessed by a particular category member correlates with how representative that member is judged to be. Robins are judged to be highly prototypical: they possess a large number of attributes found across other members of the BIRD category. Conversely, ostriches, which are judged not to be very good examples of the category BIRD, are found to have considerably fewer of the common attributes found among members of the category. Therefore, while OSTRICH and ROBIN are representative to different degrees, they nonetheless share a number of attributes and thus exhibit a degree of family resemblance. The claim that category members are related by family resemblance relations rather than by necessary and sufficient conditions entails that categories are predicted to have fuzzy boundaries. In other words, we expect to reach a point at which, due to the absence of a significant number of shared characteristics, it becomes unclear whether a given entity can be judged as a member of a given category or not.
