1. Two outstanding essays of Jakobson's on the subject of metaphor and metonymy are “Linguistics and Poetics,” in Style in Language, ed. Sebeok, Thomas (Cambridge: MIT Press, 1960), pp. 350–377Google Scholar, and “Two Aspects of Language: Metaphor and Metonymy,” in Fundamentals of Language, ed. Jakobson, and Halle, Morris (The Hague: Mouton, 1956);Google Scholar rpt. in European Literary Theory and Practice: From Existential Phenomenology to Structuralism, ed. Gras, Vernon (New York: Dell, 1973), pp. 119–131.Google Scholar

2. Evidence from studies of aphasia continues to recommend positing the existence of two primary verbal orientations. Individuals suffering from aphasia tend toward either similarity or contiguity disorders depending on which of two separate, identifiable parts of the brain have incurred injury. Examples of extremely valuable work building on Jakobson's ideas include the following: James Irby's “The Structure of the Stories of Jorge Luis Borges” (diss., University of Michigan, 1962) provides an example of a study that explores in depth the role of metonymy in specific texts. Wilden's, AnthonySystem and Structure: Essays in Communication and Exchange (London: Tavistock, 1972)Google Scholar relates the metaphor/metonymy concept to a wide range of fields, including Lacanian psychoanalysis, cybernetics, mathematical and logical theory, and Marxism. Le Guern's, MichelSémantique de la métaphore et de la métonymie (Paris: Larousse, 1973)Google Scholar presents a direct expansion of Jakobson's theoretical framework. My discussion of metaphor and metonymy in Agnon draws on material from my dissertation, “Spatial Form in the Social Novel” (Princeton, 1980). Works that draw on but attempt to fundamentally modify Jakobson's ideas include the following: Rhetorique Generate by Dubois, J., Edeline, F., J. M. Klinkenberg, P. Minguet, F. Pire, and H. Trinon (Paris: Larousse, 1970)Google Scholar, suggests a tripartite system of metaphor, metonymy, and synecdoche as the basis for a more complete understanding of tropes. This view fails to persuade me, since the essential features of metonymy, formulated in terms of context by Le Guern, apply to synecdoche as well and so reconfirm the soundness of the bipartite model. Substitutions of a part for a whole most naturally form a subspecies of metonymy and not an independent, equally important category of language. In Metahistory: The Historical Imagination in Nineteenth Century Europe (Baltimore: John Hopkins University Press, 1977), White, Hayden proposed a fourfold system of tropes that revitalizes an approach to figurative language widely accepted in the Renaissance. White adds irony to metaphor, metonymy, and synecdoche as basic rhetorical figures. Jakobson's model still challenges White's, since, as White himself acknowledges, irony is a phenomenon of a different order entirely from metaphor and metonymy. Not properly a linguistic operation at all, it is a perceptual one which may arise from either metonymic or metaphoric discourse. Gérard Genette, like these others, has attempted to counteract the confining nature of Jakobson's binary opposition, and he does so by calling for the revival of a plethora of rhetorical terms. See Figures II (Paris: Seuil, 1969), “La rhetorique restreinte,” pp. 21–40, and “Metonymie chez Proust,” pp. 41–63. Contrary to what Genette claims, it seems to me that his suggestions are most helpful to the extent that they refine rather than challenge Jakobson's ideas.Google Scholar

3. “Spatial Form in Modern Literature,” in The Widening Gyre: Crisis and Mastery in Modern Literature (New Brunswick: Rutgers University Press, 1963), pp. 3–60.

4. 'Omanul hassippur šel 'Agnon (Jerusalem: Ha-Kibbuts ha-MƏutiad and Keter, 1976), pp. 228–278.

5. For discussion on the change in Agnon's style, see, e.g., Shaked, “'Al šƏloša nusakhim ušƏloša slabim bƏhitpatkhut hassippur ‘Yatom bƏ 'almanah,’” in 'Omanut hassippur šet 'Agnon, pp. 137–150; Alter, Robert, After the Tradition (New York: Dutton, 1969), pp. 141–142;Google ScholarBand, Arnold, Nostalgia and Nightmare: A Study in the Fiction of S. Y. Agnon (Berkeley: University of California Press, 1968), pp. 62–63;Google ScholarKaspi, Joseph: A Study in the Evolution ofS. Y. Agnon's Style (Chicago: Spertus College of Judaica Press, 1969);Google ScholarDe Vries, Benjamin, “Vav ha-mƏhapakh,” in L'Agnon Šai (Jerusalem: Mercaz, 1959).Google Scholar

6. No discourse, of course, is entirely metaphoric or entirely metonymic. At stake are not entities but relations between terms; depending on how we observe any phenomenon in language, we will discover links of both contiguity and similarity. This overlap is evident in the simple fact that to form a sentence one must select lexical items from among the possible choices available (that is, from among things in some way equivalent) and then combine those items together syntactically. The words and phrases in the resulting utterance necessarily maintain relationships of both likeness and dependence to other elements in the language. Recognizing this flexibility in Jakobson's work makes his theory of metaphor and metonymy elastic enough to account for contrary narrative impulses without becoming ensnared in reductive polarities.

7. Quotations from 'Oreah nata lalun are from the Schocken edition (Jerusalem, 1976), and the English translation by Misha Louvish (New York: Schocken, 1968).

8. “Agnon Before and After,” in Prooftexts 2, no. 1 (January 1982): 78–94.

9. I am drawing here on discussion by Le Guern. A simple schematization, drawn from Rhétorique Générate (p. 118) may illustrate most clearly the difference between metaphor and metonymy. Metaphor appears as the overlapping of two circles, that is, as the copossession of semantic elements between two contexts. Metonymy appears as the coinclusion of two terms within a single ensemble of semantic components.

10. Convertibility is Burke's, Kenneth term from A Grammar of Motives (Berkeley and Los Angeles: University of California Press, 1969), p. 508. It seems a very apt description to me, though Burke uses it in connection with synecdoche and not with metonymy. 1 would consider synecdoche a subclass of metonymy. Burke defines metonymy as the result of several metaphoric substitutions, but since the one trope can be explained in terms of the other in this framework, the definition seems less economical to me than Jakobson's dichotomy.Google Scholar

11. This consideration of metaphor and metonymy does not, of course, pretend to exhaust discussion of Agnon's prose in A Guest for the Night, since so many factors contribute to the resilience of this writing. I do not think, however, that the complexities of the language discredit my principal arguments here, nor that my basic points are incompatible with the linguistic analyses by Shaked, Band, De Vries, and Alter.

12. In “Ha-mƏsaper kƏsoper” Shaked discusses how the Guest's faulty logic in like manner undermines his credibility.

13. The key motif has been discussed extensively by critics. See, e.g., Kurzweil's, BaruchMasot 'at sippurei Šai 'Agnon (Jerusalem: Schocken, 1963) and Band's Nostalgia and Nightmare.Google Scholar