Page 520 - Encyclopedia of Philosophy of Language

Key Figures
difficulties led to his removal from this post. His last 20 years were spent in poverty and isolation writing continuously on logic and philosophy. His friends, notably William James, arranged for him to give some lectures at Harvard during this period which developed his most famous doctrine, pragmatism. A systematic philosopher, much influenced by Kant, Peirce contributed to most areas of philosophy. Inde- pendently of Frege, he introduced quantifiers into logic in the 1880s; he made important contributions to the philosophy of science and metaphysics; he wrote on categories, on perception, on mathematics and, in later years, on the nature of religious experience. At the center of his work was a sophisticated and dis- tinctive theory of language and representation: he once wrote that he could not approach any topic 'except as a study of semiotic' (the general theory of signs); and his other philosophical views all reflect his innovations in that area.
1. Semiotic: The Sign Relation
Peirce's theory of signs was present in his earliest lectures and publications, but his ideas became stead- ily more sophisticated over the ensuing 50 years. His correspondence with an Englishwoman, Victoria, Lady Welby, contains many later thoughts on the subject (Peirce 1977). Since he insisted that all thought and sensation involved signs, and he used his theory in developing an account of the nature of logical and mathematical notations as well as in defending a com- plex theory of language and meaning, it is easy to appreciate its importance for his philosophy. His fun- damental claim was that the sign relation is irreducibly triadic: a sign can represent an object only by being interpreted in thought as standing for that object; an interpreting thought mediates between sign and object. Hence the focus of his work is the process of sign interpretation or 'semiosis.'
Interpretation need not involve merely judging that the sign has a distinctive object: inference can be involved, the sign being 'developed' through interpret- ation. If I already believe that Peter is either American or Canadian, I can interpret the announcement that he is not American by inferring that he is Canadian. My understanding is manifested by my drawing that inference. Or my interpretation of the claim that salt is soluble could be the acquisition of a tendency to be surprised if a sample of salt does not dissolve. Science can then be viewed as an attempt to arrive at ever richer and more stable interpretations of scientific assertions, adding information and removing error so that one arrives at a complete and accurate speci- fication of the object of the sign.
2. Classificationsof Signs
The core of Peirce's theory was a complex, somewhat baroque system of classifications of signs, objects, and interpretants, much of which depended on his theory
of categories. The most famous of these classifications concerns the connections between sign and object which enable the former to represent the latter: it distinguishes 'icon,' 'index,' and 'symbol.' An icon resembles its object: they share a property which either could possess even if the other did not exist. An index stands in a real 'existential' relation to its object: a weather vane is an index of wind direction, a pointing finger is an index of the object at which it points. A symbol represents its object only because there is a conventional practice of so using it: other 'replicas' or 'tokens' of the same 'type' have represented the same object in the past. Peirce is skeptical that there are any 'pure' icons and his discussion focuses on 'hypoicons': these are conventional signs, but the convention does not itself fix the sign's object but merely determines how the sign is to be used as an icon. Maps are thus hypoicons, as are systems of logical and mathematical notation. Analogously, expressions like 'that,' 'now,' 'here,' and T are conventional signs, the conventions determining how they are to be interpreted as indices rather than fixing their objects unaided. So they are conventional indices rather than symbols.
From the 1880s, Peirce insisted that an adequate descriptive language must contain signs of all three kinds. Unless it contained symbols it would lack gen- erality: if wisdom is a property that many can share, tokens of that type must be usable in different assertions; and reasoning involves using general stan- dards, claiming that all tokens of certain inference or assertion types are correct. Moreover indices are required if one is to refer to external objects: ordinary proper names, demonstratives, and quantifiers are all (conventional) indices. Finally general terms or predi- cates are (hypo)icons. Since icons share properties with their objects, one can learn more about the object by examining the icon: for example, learn about the terrain by studying a map. The systematic relations between the predicates in the different sentences one accepts provides a kind of map of the relations between the corresponding properties in the world; so reasoning and reflection, experimenting on and observing icons, can increase knowledge of the objects of thought.
3. Pragmatism
Peirce's pragmatist principle was introduced in his 1877 paper 'How to make our ideas clear' as a device for clarifying the meanings of words, concepts, and sentences and for identifying those which lacked meaning:
Consider what effects, which might conceivably have practical bearing, we conceive the object of our con- ception to have. Then, our conception of those effects is the whole of our conception of the object.
(Peirce 1982-86:266) As his examples make clear, the claim that something
498