Key points are not available for this paper at this time.
Like most other literary-critical terms, character, figure or person are polysemic and ambiguous. For the purpose of this essay, character or person in narrative will be understood as designating a human or human-like individual, existing in some possible world, and capable of fulfilling the argument position in the propositional form DO(X) - that is, a Narrative Agent (=NA), to whom inner states, mental properties (traits, features) or complexes of such properties (personality models) can be ascribed on the basis of textual data. This particular explication of the term has the advantage of capturing the reader's intuitive understanding of the term, while at the same time enabling a disciplined and explicit study of this phenomenon within narratology. The ascription of individual mental traits to an NA may be called and the ascription of complexes of traits character-building or portraiture. The two activities are logically and substantively different. The first is the primary one and is based on inference drawn from individual acts of the NA, details of his looks and setting, etc. (see below). Character-building, on the other hand, comes later and involves several additional operations. These include the accumulation of a number of traits from several successive acts of the NA, setting, or formal patterns; a generalization concerning their extent in terms of narrative time; the classification or categorization of these traits; their interrelation in terms of a network or hierarchy of traits; a confrontation of traits belonging to successive acts in order to infer second order traits such as inconsistent; and finally, an attempt to interrelate the traits or trait-clusters into a unified stable constellation (configuration, pattern, Gestalt, personality model) of some duration in terms of narrative time. Character-building consists of a succession of individual operations of characterization, together with second order activities of continual patterning and repatterning of the traits obtained in the first order operations, until a fairly coherent
Uri Margolin (Wed,) studied this question.