Semantic: Meaning and Cognition I: Categorization and Cognitive Semantics

Meaning and Cognition I: Categorization and Cognitive Semantics

1. The Semantic of Categorization

Categorization is something that humans and other organisms do: "doing the right thing with the right kind of thing." The doing can be nonverbal or verbal. For humans, both concrete objects and abstract ideas are recognized, differentiated, and understood through categorization. Objects are usually categorized for some adaptive or pragmatic purpose. Categorization is grounded in the features that distinguish the category's members from nonmembers. Categorization is important in learning, prediction, inference, decision making, language, and many forms of organisms' interaction with their environments.

Categorization is an important topic in semantics because language can be seen as means of categorizing experience. A word like flower, for example, categorizes an indefinitely large number of different entities in the world as all examples of a single kind of thing, the category FLOWER. The actual types of flower vary widely – think of the difference between a tulip, a carnation and a sunflower – but these differences in no way affect the categorization of all types as flowers. The same is true of other lexical categories.

1. Classical Categorization

According to the classical Aristotelian view, categories are discrete entities characterized by a set of features that are shared by their members. In the classical view, categories need to be clearly defined, mutually exclusive and collectively exhaustive. This way, any entity in the given classification universe belongs unequivocally to one, and only one, of the proposed categories.

Standard logical approaches to language are two-valued approaches. This means that they only recognize two truth values, true and false. On this approach, any proposition must either be true or false. There is no room for the proposition to be partly true and partly false, or true in some respects but false in others.

2. Problems with Classical Categories

Many semanticists rejected the classical view of categorization since it seems unable to account for basic semantic phenomena, such as the following:

• There are categories in which some members are better exemplars of the category than others.

• There are categories in which the boundaries of membership are not clear-cut: it is not always possible to say whether or not something is a member of the category.

3. Prototype Categorization

A prototype is a cognitive reference point, i.e the proto-image of all representatives of the meaning of a word or of a category. Thus, a robin or a sparrow can be regarded as a prototype or a "good example" of the category bird, whereas a penguin or an ostrich is a rather "bad example" of this category. Accordingly, the members of a category can be graded according to their typicality. A "good" example is only rated as such by virtue of its features. Defining a prototype as the bundle of typical features of a category, we can thus imagine birds as 'creatures that are covered with feathers, have two wings and two legs, and the majority of which can fly'. Therefore, a penguin is a less "good" bird, as it lacks some of the typical features, such as the ability to fly. Features themselves can also be more or less typical, for example 'twittering' is less typical and specific to birds than 'flying'.

If an item shares at least some central features with the category prototype, we consider it as an example of this category. As a consequence, word meanings contain all the properties of cognitive categories: We can distinguish between central and more peripheral meanings of a lexeme, and word meanings are not rigid, but there are often gradual transitions and fuzzy boundaries between them. Thus, prototype semantics is a 'more-or-less semantics', as opposed to the 'all-or-nothing' approach of structure-oriented feature semantics. However, this does not weaken the usefulness of a feature-based classification: The features belonging to a prototype of category are the ones that are relevant for categorization.

4. Problem with Prototype Categories

In each case, a quotation from early prototype studies is added to illustrate the point.

(i)                 Prototypical categories cannot be defined by means of a single set of crite-rial (necessary and sufficient) attributes:We have argued that many words ... have as their meanings not a list of necessary and sufficient conditions that a thing or event must satisfy to count as a member of the category denoted by the word, but rather a psychological object or process which we have called a prototype (Coleman and Kay 1981: 43).

(ii)               Prototypical categories exhibit a family resemblance structure, or more generally, their semantic structure takes the form of a radial set of clustered and overlapping meanings:The purpose of the present research was to explore one of the major structural principles which, we believe, may govern the formation of the prototype structure of semantic categories. This principle was first suggested in philosophy; Wittgen stein (1953) argued that the referents of a word need not have common elements to be understood and used in the normal functioning of language. He suggested that, rather, a family resemblance might be what linked the various referents of a word. A family resemblance relationship takes the form AB, BC, CD, DE. That is, each item has at least one, and probably several, elements in common with one or more items, but no, or few, elements are common to all items (Rosch and Mervis 1975: 574–575).

(iii) Prototypical categories exhibit degrees of category membership; not every member is equally representative for a category:By prototypes of categories we have generally meant the clearest cases of category membership defined operationally by people’s judgments of goodness of member ship in the category ... we can judge how clear a case something is and deal with categories on the basis of clear cases in the total absence of information about boundaries (Rosch 1978: 36).

2. Language and Conceptualization: Cognitive Approaches to Semantics

1. Commitments of Cognitive Semantics

The label cognitive semantics covers a variety of quite different approaches. In general, however, these approaches are characterized by a holistic vision of the place of language within cognition. For many investigators, this involves the following commitments:

• A rejection of a modular approach to language

• An identification of meaning with ‘conceptual structure’

• A rejection of the syntax–semantics distinction

• A rejection of the semantics–pragmatics distinction

1. Rejection of modularity

Many cognitivists share a commitment to understanding language as governed by the same cognitive principles at work in other psychological domains. This holistic approach to language structure contrasts with the strongly modular approach promoted by investigators in the tradition of Chomsky (1965) and Fodor (1983). Modular research assumes that language is one of a number of independent modules or faculties within cognition, each of which has a structure and principles independent of those at work in other cognitive domains. On a modular vision of the mind, the principles of the language module will be entirely distinct from those of the others, like vision, memory, reasoning or musical cognition.

2. Meaning as conceptual structure

Most cognitivists share a rejection of the dictionary–encyclopedia distinction (see 3.3): in the words of Jackendoff (2002: 293), ‘we must consider the domain of linguistic semantics to be continuous with human conceptualization as a whole’. In other words, studying linguistic meaning is the same thing as studying the nature of human conceptual structure a cover-all term for our ‘thoughts, concepts, perceptions, images, and mental experience in general’ (Langacker 1987: 98). The meaning of a word like house simply is the concept we have of houses; as discussed in 3.3, any aspect of the knowledge we have houses can become linguistically relevant. Cognitive approaches to semantics aim to describe the full knowledge structures that are associated with the words of a language. As a result, conceptualist descriptions of the meanings of words are considerably more rich,complex and open-ended than in other varieties of semantic analysis.

3. Rejection of the semantics–syntax distinction

Just as language as a whole is not seen as a distinct cognitive capacity in conceptualist frameworks, so too the language internal division between semantics and syntax often recognized in linguistics is typically rejected. Evans and Green (2006) identify a number of fundamental cognitive principles which do not respect any division between these two domains: prototype effects, polysemy (4.3) and metaphor (see below), for example, seem to exist not only in the domain of word meaning, but in morphology and syntax as well.

4. Rejection of the semantics–pragmatics distinction

Rejection of the semantics–pragmatics distinction Cognitive linguistics also typically rejects the distinction between a purely semantic level of word meaning and a nonsemantic level of language use. This means that facts which in other frameworks might be attributed to inferences based on literal meanings (see Chapter 4) are assumed to reflect aspects of the actual, literal meaning of the words concerned. We will see in the sections that follow how this and the other related commitments of cognitive semantics affect the analyses of semantic content proposed in these frameworks.

2. Idealized Cognitive Models

Inspired by the work of Fillmore (1982), Lakoff (1987) proposed that prototype effects are by products of the fact that our knowledge is organized into structures stored in long-term memory that he calls idealized cognitive models (ICMs). The notion of ICM is meant to capture the contribution of encyclopaedic knowledge to our understanding of concepts. ICMs can be thought of as theories of particular subjects the implicit knowledge we have about the objects, relations and processes named in language (Lakoff 1987: 45). Lakoff introduces the notion of ICMs with the example of the English word Tuesday. The meaning of Tuesday, he says, can only be represented by specifying the underlying knowledge English speakers have of the organization of time into days and weeks, and the place of Tuesday within this organization.

3. Embodiment and Image Schemas

A conceptualization is embodied is to draw attention to its origin in basic physical experience. Johnson (1987) pointed out that much language use reflects patterns in our own bodily experience, particularly our perceptual interactions, movements and manipulations of objects. Particularly basic patterns of repeated experience give rise to the conceptual categories which Johnson called image schemas, such as CONTAINMENT, SOURCE-PATH-GOAL, FORCE, BALANCE and others. These ‘operate as organizing structures of our experience and understanding at the level of bodily perception and movement’ (Johnson 1987: 20), and thus also underlie the conceptual categories deployed in language.

4. Metaphor and Metonymy

Metaphor is stressed in much cognitive semantics as ‘an inherent and fundamental aspect of semantic and grammatical structure’ (Langacker 1987: 100). Metaphor was originally a category of literary and rhetorical analysis, not of linguistic description (see e.g. Ricœur 1975).

Lakoff and Johnson noted that far from being unusual or atypical, metaphorical utterances like those in (3) are actually the basic, ordinary way in which obligations would be described in English. They also observed that these expressions all express a common underlying idea, which we could label OBLIGATIONS ARE PHYSICAL BURDENS.

Second, Lakoff and Johnson propose that it is not simply a random linguistic fact that English has so many expressions along the lines of (3). There is a reason that obligations are described as though they were physical burdens, and not in any other of the innumerable ways we could dream up.

The idea that some concepts can have metaphorical structure is referred to by Lakoff and Johnson as the conceptual theory of metaphor. This theory focuses on metaphor as a cognitive device which acts as a model to express the nature of otherwise hard-to-conceptualize ideas.

Notice the difference between metonymies and metaphors: in metaphor, there is a relation of mapping between two concepts, with the structure of one concept (JOURNEYS, PHYSICAL BURDENS) being imposed onto another (LOVE, OBLIGATIONS). Metonymies do not serve to structure one concept in terms of another: it is not possible to articulate the detailed mappings we established in the love and obligation cases. Instead, they draw on the associations within a single conceptual ‘domain’, allowing one part of a concept to convey another. We will see further examples of this in the next section.

5. Radial categories in word meaning

Semantics cognitive identifies meaning with conceptual structure, the network of stored representations in our memory involved in thought and language . A word can be seen as an entry point to a certain 'region' of our conceptual structure. Using the ideas discussed in the preceding sections, we can now sketch the way in which cognitive semantics models the conceptual knowledge structures underlying meaning.

The ICM of head contains such information as the fact that the head is at the top of the body, that it contains the brain, the fact that ears, eyes, mouth and nose are located on it, the fact that it is mostly made of bone, that thinking happens inside it, and so on. Perhaps, as suggested by some investigators, one aspect of the conceptualization associated with head (as with any other non-abstract word) is a visual/spatial element, encoding such features as the referent's typical shape, colour and overall appearance (Jackendoff 2002: 345-350).

The ICM of head determines the way in which ordinary sentences involv ing it are understood. For instance, we know that the expression shake one's head refers to a particular back and forth movement of the head rather than to an action in which one takes one's head in one's hands and shakes it. Similarly, we know that if asked to turn one's head we should turn it horizontally from one side to another, not move it in a fixed circular motion without allowing it to come to rest. Facts like these are part cf our understanding of head, and must therefore be represented in conceptual structure.

Note also that the ICM is modelled on the human head. The closer a creature's head is to a human head, the more appropriate it is to describe it as having a head. Thus, there is nothing odd about describing monkeys, dogs, cats and many other types of animal as having heads, whereas it seems more strained to describe the corresponding bodyparts of worms, whales, spiders, snails and starfish.

6.      Problems with cognitive semantics

Cognitivist analyses of meaning focus on metaphor and metonymy, clopaedic meaning description and semantic extension, and enable a much more detailed representation of semantic content than is possible in more formal or componential approaches. The radial network models and image-schema diagrams allow a rich description of meaning that seems to make contact with perceptual and cultural aspects of language- aspects which are easily left out in other types of analysis. In spite of these attractions, however. Cognitivist analyses of meaning have been criticized for a number of reasons. We will consider three:

1.    The ambiguity of diagrammatic representations;

2.    The problem of determining the core meaning; and

3.    The  indeterminate and speculative nature of the analyses.

            Cognitivist theories are often criticized for their arbitrary and Speculative character. The model of categorization proposed in cognitive semantics is offered as a psychologically realistic model of conceptualization, but has not yet been subjected to significant psychological experimentation, although this is beginning (Boroditsky 2000; Boroditsky & Ramscar 2002; Matlock et al. 2005). The analyses have largely been based on linguistic evidence a problematic state of affairs for a theory which wants to develop a psychologically realistic model. Kamp and Reyle point to the circularity of any attempt to explain meaning ('content') by way of the mental representations lying behind uses of language: it won't do to base whatever one has to say about mental representations of content solely on what can be learned from studying the linguistic expressions through which these contents are publicly expressed, and then to offer mental representation as explaining the content of the corresponding expressions of the public language.