Prototype Theory and Classical Theory: An Explanation and Comparison
The Classical Theory which identified concepts with a set of necessary and sufficient conditions has played a prominent role in mainstream twentieth-century linguistics. However, since the Classical view has its defects and was greatly criticized in 1960s, other theories including the Prototype Theory were later proposed as alternatives. This paper aims to discuss the main characteristics of the Prototype Theory and explain how it differs from the Classical Theory. 1. The Classical Theory This section is devoted to give an overview of the Classical Theory, including its origin, its characteristics and its problems. 1.1 Origin of the Classical Theory The Classical Theory emerged in Greek antiquity and continued to dominate psychology, philosophy, and linguistics throughout the twentieth century. The theory was rooted in Metaphysics authored by Aristotle, in which the discussion on “man” have decisive influence on how things are classified for the following generations. Following are the discussion on “man” in Metaphysics(Aristotle 1993). If “man” has one meaning, let this be “two-footed animal” . By “has one meaning” I mean this: if X means “man”, then if anything is a man, its humanity will consist in being X. (Aristotle 1993) If anything can be truly said to be “man”, it must be “two-footed animal”; for this is what “man” is intended to mean. (Aristotle 1993) Therefore, according to Aristotle, the concept of MEN is the set of semantic features which distinguish it from other concepts. To say X is a man, is to assign an entity X to the category MEN. And we do this by checking off the properties of X against the features which defines category Y; our knowledge of this set of features characterizes our knowledge of the concept MEN. This is also the main reason why the Classical Theory are often called the checklist theory. According to the discussion in Metaphysics, Aristotle singled out two defining features of the concept MEN, namely “two-footed” and “animal”. Therefore, the concept MEN can be presented as follows: MEN=two-footed, animal X is a man iff L. (L is a list of features) We can see that X can be a man if and only if X has the lists of features L. On one hand, if any of the defining features is missing, then the X can not be a member of MEN. Therefore, L is the necessary conditions. On the other hand, if X exhibits each of the features in L, then X is no doubt a member of MEN. Therefore, L is the also the sufficient conditions. 1.2 Characteristics of the Classical Theory Under the influence of Metaphysics, the first characteristic of the Classical Theory can be summarized as follows(Geeraerts 1983, Taylor 2001): (1) A set of necessary and sufficient conditions are needed for something to be an instance of a category. This characteristics means that the classical theory views concepts as lists of bits of knowledge: the necessary and sufficient conditions for something to be an example of that concept(Saeed 1997). Further influence of Aristotle’s thought on the Classical Theory came from the law of contradiction and the law of excluded middle(Aristotle 1993). The law of contradiction states that a thing cannot both possess a feature and not possess it. Neither can it both belong to a category and not belong to a category. The law of excluded middle states that a thing must either possess or not possess a feature, must either belong to or not belong to a category. Hence: (2) Features are binary. It is a all-or-none phenomenon. A feature is either involved in the definition of a category or not. An entity possesses this feature, or does not. In any given instance a feature is either present or absent, and it can take on only one of the two values, either [+] or [-](Taylor 1995). Therefore, MEN can be presented as follows: MEN=[+TWO-FOOTED], [+ANIMAL] X is a man iff L. (L is a list of features) The third and fourth characteristics came from the second. (3) Category membership is clearly bounded, based on a yes/no distinction. A category, once established, divides all entities into two sorts, namely, those that are members of a category and those that are not. It is either “yes” or “no”, and there is not any “in-between” or any ambiguous cases. There is not any entities which “in some degree” or “in a way” belong to the category but in another way do not. It is the “yes” or “no” distinction instead of “more” or “less” relationship. (4) All attributes or features have a full and equal degree of membership. All members of a category have equal status and any entity which exhibits all the defining features of a category is a full member of that category and is equal to any other members of that category whereas any entity which does not exhibit all the defining features is not a member. There is no distinction in degrees of membership in a category and all entities of the same category are as good a member as any others. According to the four characteristics stated above, the concept WOMEN can be presented and explained as follows: WOMEN=[+HUMAN], [-MALE],[+ADULT] X is a woman iff L. (L is a list of features) X is a woman if and only it has it list of attributes L. One can see these attributes as conditions: if something must have them to be a woman, they can be called necessary conditions. In addition, if we can find the right set, so that just that set is enough to define a woman, then they can be called sufficient conditions. And all features are binary, for instance, [+Human] is either involved in the definition of the category WOMEN or not whereas the entity X either possess a feature or not. The boundary of being a member of WOMEN is clear, this is a “either-or” situation and there are not any cases where “in a way” X is a woman and in another way X is not. Besides, the three defining features [+HUMAN], [-MALE],[+ADULT] has equal status. 1.3 Problems of the Classical Theory Because of the overwhelming effect of the classical theory, much work in phonology, syntax and semantics roots in the four characteristics of the theory and thus assumes categories are logical bounded entities, membership in which is defined by an item’s possession of a simple set of defining features in which all instances possessing the defining attributes have a full and equal degree of membership. However, such highly idealized binary thought was greatly criticized by the philosopher Ludwig Wittgenstein. In his work Philosophical Investigation(1978), he addressed the questions of how to define the word “game”. He notes that it is difficult to find a shared set of common properties on which games can be clearly distinguished from non-games. Among members of the categories GAME, A is similar to B and B is similar to C, meanwhile, C is similar to D and D is similar to A. Some member share some attributes while others share other attributes. There is no common sets of attributes. According to Wittgenstein(1978), the relationship among members of GAME was more like a family resemblance. Therefore, he came to the conclusion that the boundary of category is fuzzy and not based on shared defining features as expected by the classical theory, but rather based on a overlapping network of similarities. Wittgenstein’s philosophical thought has provided support for the emergent of the Prototype Theory. Another problem with the classical view is that it’s hard to define the necessary and sufficient conditions for even nouns which identify concrete and natural kinds like dog or cat. Following is an example take from Semantics written by Saeed(1997). We might agree on some attributes: Zebra is animal has 4 legs is striped is a herbivore... However, if we find in a herb of zebra, one that is pure white or black, we might still want to call it a zebra. Or if by some birth defect, a three-legged zebra comes into the world, it would still be a zebra (Saeed 1997). The third problem with the classical view is Putman’s (1975) observations about ignorance. According to Putman(1975), we often use words (and understand) without knowing the true nature of things. For example, we don’t necessary know the chemical make-up of glass but we know the word glass and also use it. Due to all these problems, the classical view is greatly criticized in the min-1950s and other theories emerged as alternatives. 2. The Prototype Theory This section aims to give an overview of the Prototype Theory, including development of the theory, definition of the prototypes and characteristics of the theory. The comparison between characteristics of the Classical Theory and Prototype Theory will also be made in this section. 2.1 Development of the Prototype Theory Due to the many inadequacies of the classical view of concepts explained above, the Prototype Theory was proposed as one of the alternatives. This theory was labeled “non-Aristotelian”, as opposed to the Classical Theory. The real development of prototype theory is triggered by anthropologist Berlin and Kay(1969)’s study on colors. They found that most colors are actually decided by some focal or basic colors. And one’s use of color vocabularies obey such sequences: “WHITE/BLACK--RED--GREEN/YELLOW/BLUEBROWN--PURPLE/PINK/ORANGE/GREY.” Such result is further proved by Glass and Holyoak’s study(1986: 161). The study of colors soon aroused scholars’ interests in focal instance of a category. Many studies (Battig &Montague 1969; Rips, Shoben &Smith 1973; Rosch 1973)proved that: (1) We also regard certain member of a category as the most typical one. (2)When citing members of a category, some members are cited more frequently than others. (3) It’s quickly to decide the truth value of a sentence involving a typical member(A robin is a bird) than a sentence involving a non-typical member(A chicken is a bird. The studies on focal instance of category finally triggered the establishment of the Prototype Theory. In the mid 1970s, the American psychologist Eleanor Rosch and her co-worker discovered prototype effects of categories and then suggested the notion of prototype(e.g. Rosch 1973, 1975; Rosch and Mervis 1975; Rosch et al. 1976; Simpson and Miller 1976). Ever since then, more and more linguists, psychologists, anthropologists and scholars in other fields made joint efforts to establish the Prototype theory and apply the theory to cognitive and linguistic studies. The new Prototype Theory was theoretically and empirically supported by Coleman Fillmore (1978), Labov(1978), Kay (1981) and Geeraerts (1984) etc. In Fillmore(1978)’s study, he gave an example that demonstrated the deficiency of the Classical Theory: if climb contains the features or components “clambering” and “ascending”, the acceptability of both The monkey is climbing down the flagpole and The snail is climbing up the flagpole cannot be explained. For the prototype approach climb also has two criteria, but either of them may be missing. In Labov(1978) and Rosch(1973), they discovered that the attributes are not the binary constructs of the classical approach. Instead, they are generalization from focal(i.e. Prototypical) exemplars. In Geeraert(1984), he summarized and justified the characteristics of the Prototype Theory. | 