68 Meaning in Language 3.2.4 Dynamic construal Underlying most work on conceptual categories there has been an implicit assump- tion that mature, normal human beings possess an inventory of discrete and more or less stable categories with permanent mental representations, which are activated in appropriate circumstances. (Of course, new concepts can be learned, but these will have the same stable character as those already in the inventory.) The goal of conceptual studies, it is assumed, is to elucidate the nature of these mental entities, Conceptually oriented studies of word meaning typically assume that a word is associated with a specific concept (or, in the case of ambiguity, with a specific set of such concepts). However, the notion of fixed categories has not gone unchal- lenged. For instance, experiments by Barsalou, a cognitive psychologist, have shown that people are capable of forming ad hoc categories with no conventional names, such as THINGS TO TAKE ON A PICNIC, THINGS ON A DESK THAT COULD BE USED TO POUND IN A NAIL, SUITABLE BIRTHDAY PRESENTS FOR A TEN-YEAR-OLD, and so on. These categories showed the same characteristics as established categories, with graded centrality and characteristic features. These and similar results suggest that categories are generated online as and when required. However, all is not flux and chaos--that would be fatal for communication. The suggestion is that previous experience has a permanent effect on our tendency to construe concepts in particular ways. The psychologists Smith and Samuelson have this to say: Critically, the accrual of these long-term changes provides a source of stability in a continually changing system. If there are statistical regularities, patterns, in our experi- ences that recur over and over again, then as each moment of knowing is laid on the preceding moments, weak tendencies to behave and to think in certain ways will become strong tendencies sometimes so strong that they will not be easily perturbed and thus might seem fixed. (Smith and Samuelson 1997: 175-6) In other words, concepts may have any degree of fixedness/fluidity. The conse- quences of this view for word meaning will be taken up in Chapter 5. Discussion questions and exercises 1. Prototype features Suggest a set of prototype features for one or more of the following conceptual categories (or select your own example(s)): CLOTHES FRUIT MUSICAL INSTRUMENT HOBBY BUILDING HOUSEHOLD APPLIANCE For each category, draw up a list of possible members, including some marginal cases, and ask another person to assign Goe ratings. Consider to what extent the ratings can be accounted for in terms of your suggested features. Concepts and meaning 69 2. Category levels Assign the following categories to superordinate, basic, or subordinate level: BIRO TEASPOON SANDAL UNDERWEAR SEAGULL DAISY GRASS BULLDOZER BUS MOUNTAIN BIKE SELF-RAISING FLOUR WALNUT SUGAR ARMCHAIR DELICATESSEN SUPERMARKET PETROL STATION TOWN HALL PARK MOTORWAY ROAD CANAL POLICE STATION BUILDING GROCERIES WINE CHAMPAGNE BEVERAGE MILK Suggestions for further reading That meaning is essentially conceptual in nature is one of the central tenets of cognitive linguistics. For elementary introductions to cognitive linguistics see Ungerer and Schmid (1996). Ultimately, a reader interested in this approach will eventually want to tackle the foundational texts. The 'bible' of the cognitive approach is Langacker's two volume Foundations of Cognitive Grammar (1987 and 1991a). However, this is not an easy read; fortunately, many of the basic topics are expounded in a much more accessible form in Langacker (1991b). The inter- ested reader will also find articles on a wide range of cognitive linguistic topics in the journal Cognitive Linguistics. See Millikan (2000) for a thorough discussion of the notion of kinds'. An alternative conceptual approach to meaning can be found in the works of Jackendoff; Jackendoff (1983) provides a good introduction. An interesting com- parison between Jackendoff's approach and the cognitive linguistic approach (including a contribution from Jackendoff himself) can be found in Volume 7.1 of Cognitive Linguistics, which also gives a fairly full bibliography of Jackend- off's later work. Cruse (1990) provides an introduction to prototype theory as applied to lexical semantics. (The volume which includes this article also contains many other articles on the topic.) A fuller account is to be found in Taylor (1989); Cruse (1992C) is a critical review of this. Ungerer and Schmid (1996) has an interesting chapter on categorization. For a more psychological view of the prototype approach to meaning, see the articles in Schwanenflugel (1994), especially those by Hampton and Murphy. Cruse (1995) attempts to apply prototype theory to lexical relations.
Purchase answer to see full attachment