I would like to thank Leo Treitler and Robert Bailey, who in the early stages of work for this article offered their valuable comments and advice, and to Janet Levy for her encouragement and support. For the final version I am grateful to the anonymous readers and, especially, to Nicholas Cook for their meaningful questions and suggestions; and to David Metzer for his attentive and constructive reading.Google Scholar

1 Hayden White, ‘The Value of Narrativity in the Representation of Reality’, On Narrative, ed. W. J. Thomas Mitchell (Chicago, IL, and London, 1981), 1–23 (p. 1).Google Scholar

2 Leo Treitler, ‘What Kind of Story is History?‘, Music and the Historical Imagination (Cambridge, MA, and London, 1989), 157–75 (p. 173).Google Scholar

3 Roland Barthes, ‘Introduction to the Structural Analysis of Narratives’, Image–Music–Text, trans. Stephen Heath (New York, 1977), 79–124 (p. 79).Google Scholar

4 See, for example, Prince, Gerald, A Dictionary of Narratology (Lincoln, NE, 1987), 58; and idem, Narratology: The Form and Functioning of Narrative (Berlin, New York and Amsterdam, 1982), 4.Google Scholar

5 See, for example, Seymour Chatman, ‘What Novels Can Do that Films Can't (and Vice Versa)’, On Narrative, ed. Mitchell, 117–36 (p. 118), and his book on narrative, Story and Discourse: Narrative Structure in Fiction and Film (Ithaca, NY, and London, 1978); Shlomit Rimmon-Kenan, Narrative Fiction: Contemporary Poetics (London and New York, 1989), 44: ‘time in narrative fiction can be defined as the relations of chronology between story and text [i.e. discourse]'; Gérard Genette, Narrative Discourse: An Essay in Method, trans. Jane E. Lewin (Ithaca, NY, 1980), 33 (quoting Christian Metz: ‘one of the functions of narrative is to invent one time scheme in terms of another time scheme'); and so on. As is well known, the terms ‘story’ and ‘discourse’ are transformations of the Russian Formalists' original distinction between ‘fabula’ and ‘sjužet’.Google Scholar

6 Genette, , Narrative Discourse, 34.Google Scholar

7 Rimmon-Kenan, Narrative Fiction, 44.Google Scholar

8 Goodman, Nelson, ‘Twisted Tales; or Story, Study, and Symphony’, On Narrative, ed. Mitchell, 99–115 (p. 111). Goodman, however, recognizes and gives examples of cases in which ‘not every narrative will survive every reordering’ (ibid).Google Scholar

9 Smith, Barbara Herrnstein, ‘Narrative Versions, Narrative Theories’, On Narrative, ed. Mitchell, 209–32 (pp. 224, 213–14).Google Scholar

10 Kermode, Frank, ‘Secrets and Narrative Sequence’, On Narrative, ed. Mitchell, 79–97 (p. 82), and Paul Ricoeur, ‘Narrative Time’, ibid., 165–86 (pp. 174–5). Both arguments are much more complex, but I have selected here minimal information related to the topic of the double-time structuring.Google Scholar

11 See his chapter entitled ‘Voice’ in Narrative Discourse, 212–62.Google Scholar

12 Bal, Mieke, Narratology: Introduction to the Theory of Narrative (Toronto, 1985; 2nd edn, 1997), 142–3, 157–8.Google Scholar

13 See Berger, Karol, ‘Diegesis and Mimesis: The Poetic Modes and the Matter of Artistic Presentation’, Journal of Musicology, 12 (1994), 407–33 (pp. 431–2).Google Scholar

14 Newcomb, Anthony, ‘Narrative Archetypes in Mahler's Ninth Symphony’, Music and Text: Critical Inquiries, ed. Steven Paul Scher (Cambridge, 1992), 118–36 (pp. 118–19). For Adorno's formulation, see Theodor W. Adorno, Mahler: Eine musikalische Physionomik (Frankfurt am Main, 1960), 100–1; see also the entire Chapter 4, ‘Roman’ (pp. 85–111), and Chapter 8, ‘Der lange Blick’, esp. pp. 200–16; trans. Edmund Jephcott as Mahler: A Musical Physiognomy (Chicago, 1992).Google Scholar

15 Maus, Fred Everett, ‘Music as Narrative’, Indiana Theory Review, 12 (1991), 1–34 (p. 14). He also provides a good discussion of the problems surrounding the notions of ‘story’ and ‘discourse’ as applied to music; see esp. pp. 21–4.Google Scholar

16 Samuels, Robert, Mahler's Sixth Symphony: A Study in Musical Semiotics (Cambridge, 1995), 155.Google Scholar

17 Samuels, Robert, ‘Music as Text: Mahler, Schumann and Issues in Analysis’, Theory, Analysis and Meaning in Music, ed. Anthony Pople (Cambridge, 1994), 152–63 (p. 153).Google Scholar

18 Abbate, Carolyn, ‘What the Sorcerer Said’, Unsung Voices: Opera and Musical Narrative in the Nineteenth Century (Princeton, NJ, 1991), 30–60 (p. 52). Jean-Jacques Nattiez, ‘Can One Speak of Narrativity in Music?’, trans. Katharine Ellis, Journal of the Royal Musical Association, 115 (1990), 240–57 (p. 257); originally published as ‘Peut-on parler de narrativité en musique?’, Canadian University Music Review: Alternative Musicologies/Les musicologies alternatives, 10 (1990), 68–91. Samuels ‘flatly contradict[s] one of Carolyn Abbate's best-known and most productive observations that music “seems not to have a past tense”’ by citing Mahler's Fourth Symphony, which Adorno hears ‘in the past tense'. See Samuels, ‘Music as Text’, 154. The past tense as essential for narrativity has also been questioned by certain literary critics. Prince observes that ‘the preterit in a fictional narrative is not primarily an indicator of time’, since the ‘past tense in which the events are narrated is transposed by the reader into a fictive present’ (cited from Mendilow). See Prince, Narratology, 28–9.Google Scholar

19 Abbate, , ‘What the Sorcerer Said’, 48, and ‘Music's Voices’, Unsung Voices, 3–29 (p. 29).Google Scholar

20 Kramer, Lawrence, ‘“As if a voice were in them”: Music, Narrative, and Deconstruction’, Music as Cultural Practice 1800–1900 (Berkeley, CA, and London, 1990), 176–213 (p. 189).Google Scholar

21 Kramer, Lawrence, ‘Dangerous Liaisons: The Literary Text in Musical Criticism’, 19th Century Music, 13 (1989–90), 159–67.Google Scholar

22 Nattiez, , ‘Can One Speak of Narrativity in Music?‘, 244.Google Scholar

23 See Maus, , ‘Music as Narrative’, and Leo Treitler, ‘Language and the Interpretation of Music’, Music and Meaning, ed. Jenefer Robinson (Ithaca, NY, and London, 1997), 23–56 (esp. pp. 45–50).Google Scholar

24 See, for example, Samuels, Mahler's Sixth Symphony, 135, but also Hayden White, ‘The Value of Narrativity’, and Wendy Steiner, Pictures of Romance: Form against Context in Painting and Literature (Chicago, IL, 1988).Google Scholar

25 Abbate, , Unsung Voices, xi.Google Scholar

26 Kramer, Lawrence, ‘Musical Narratology: A Theoretical Outline’, Classical Music and Postmodern Knowledge (Berkeley, CA, 1995), 98–121 (p. 101).CrossRefGoogle Scholar

27 My choice of the ‘Pastoral’ Symphony as representative of Classical stylistic features should not imply a statement about Beethoven's style in general, since, of course, other of his symphonies such as the Third or the Fifth are closer to the Romantic ideal. Implicit in my choice is the idea that if narrative characteristics were to appear in a Classical symphonic piece, they would most likely show themselves at their strongest in a programmatic piece. Although I do not consider the programme here, the piece serves better than others my purpose of illustrating how the differences in musical characteristics between the ‘Pastoral’ Symphony's Classical features and Mahler's Ninth influence the different degrees of their narrativity. Particular pieces by Mozart or Haydn might present higher degrees of narrativity than the piece I have chosen, but these other cases should be studied on their own.Google Scholar

28 Such premisses include biographical, psychological and contextual situations that might enter into narrative interpretations of music.Google Scholar

29 Nattiez, , ‘Can One Speak of Narrativity in Music?’, 245. As mentioned above, he believes that narrative interpretations of music are artificial constructions of the critics, ‘superfluous metaphors’ triggered by what he identifies as their ‘narrative impulse’. See also note 18 above.Google Scholar

30 Rose Rosengard Subotnik, ‘Romantic Music as Post-Kantian Critique: Classicism, Romanticism, and the Concept of Semiotic Universe’, On Criticizing Music: Five Philosophical Perspectives, ed. Kingsley Price (Baltimore, MD, 1981), 74–98 (pp. 84–5). In a different context, Leonard B. Meyer has made a similar observation. He distinguishes between ‘primary parameters’, which he calls ‘syntactic’ because they depend on syntactic constraints (melody, rhythm, harmony), and ‘secondary parameters’, which he calls ‘statistical’ because they can have only a ‘statistical’ characterization (dynamic level, tempo, texture, timbre, rate of activity, register, etc.). Thus he writes: ‘As rejection of convention led to a weakening of syntax … secondary parameters became more and more important for the generation of musical processes and the articulation of closure.’ See his Style and Music: Theory, History, and Ideology (Philadelphia, PA, 1989), 14–16. Also relevant is his following statement: ‘Complementing the trend toward syntactically weakened harmonic and tonal relationships was an increase in the relative importance of secondary parameters in the shaping of musical process and the articulation of musical form’ (p. 303).Google Scholar

31 Leonard Meyer has also recognized this: ‘The increasing importance of motivic relationships during the nineteenth century was in part a response to a gradual attenuation of the structuring provided by tonal syntax.‘ See ibid., 271.Google Scholar

32 This is not to say that music prior to Mahler's was semantically empty. Kofi Agawu has shown an unsuspected richness and variety of topics in, for example, Mozart's instrumental music. See his Playing with Signs: A Semiotic Interpretation of Classic Music (Princeton, NJ, 1991). Also, as James Webster observes: ‘In the eighteenth century, these “gestural” aspects of music were understood as part of a more general quality that has since become unfamiliar to us: that of rhetoric. … [E]very instrumental work was composed and understood within a context of genre, Affekt, and “topoi” (topics), which in principle enabled its ideas and gestures to be located within a network of traditional associations, including dance types and distinctions of social status.’ See Webster, James, Haydn's ‘Farewell’ Symphony and the Idea of Classical Style (Cambridge, 1991), 125. For other analyses of the variety of topics present in Mozart's, Beethoven's and Haydn's music, see Ratner, Leonard, Classic Music: Expression, Form and Style (New York, 1980); Wye J. Allanbrook, Rhythmic Gesture in Mozart (Chicago, IL, 1983); and Robert Hatten, Musical Meaning in Beethoven: Markedness, Correlation, and Interpretation (Bloomington and Indianapolis, IN, 1994). The difference, as I see it, between Mahler's and Classical music's referentiality is that while most topics in Classical music are an intrinsic part of the contemporary vocabulary, and therefore their recognition and interpretation are most often direct and univalent, some of the topics used by Mahler are anachronistic within the prevalent contemporary musical styles, so that their relationship with their context is more complex and thus generates multivalent levels of semantic meanings.Google Scholar

33 Although these characteristics are most valid for Mahler's music, the origins of these changes can be observed earlier in the century, for example in Beethoven's late works and in the music of Schubert, Berlioz, Liszt and Wagner.Google Scholar

34 Dahlhaus, Carl, Nineteenth-Century Music, trans. J. Bradford Robinson (Berkeley, CA, 1989), 256.Google Scholar

35 Leonard Meyer uses the distinction between ‘scripts’ and ‘plans’ to characterize the difference between Classical and Romantic discourse. Script is a structure which forms an interconnected whole, in which the individual parts are very dependent on one another. Plans, on the other hand, are ‘repositories for general information that will connect events that cannot be connected by use of an available script or by a standard causal chain expansion’. He then uses the opposition between the Classical (syntactic) ‘scripts’, based on primary (learnt, conventional) parameters of music, and Romantic (statistical) ‘plans’, based on secondary parameters which ‘shape experience with minimal dependence on learned rules and conventions’. See Meyer, Style and Music, 245, 208–9. According to this view, the syntactic script of sonata form stipulates tonal and functional relationships more than plans do. On the contrary, in Romantic music forms and processes are increasingly shaped by secondary parameters and, therefore, are based on plans, not on scripts (p. 246). When plans are preferred, script constraints become burdensome and, therefore, non-coincidence between the ‘statistical’ and syntactic' climax is a characteristic of Romantic music (p. 308).Google Scholar

36 Subotnik, , ‘Romantic Music as Post-Kantian Critique’, 82, 84. In what I see as an argument akin to Subotnik's observation of a ‘shift’ to a ‘linguistic ideal of musical meaning’, Daniel K. L. Chua traces the origins of the turn of ‘music into language’ to the end of the sixteenth century in the shift of music from the ‘quadrivium to the trivium, that is, from the immutable structure of medieval cosmos to the linguistic relativity of rhetoric, grammar and dialectics’. This, in turn, he states, led the ‘Romantics to reverse the process by turning language into music’, thus establishing the new ontology of instrumental (absolute) music. It was the ‘stylistic relativity inherent in the trivium’ and the ‘heterogeneous form of discourse’ of instrumental music that prompted Friedrich Schlegel's analogy between ‘the method of the novel’ and ‘that of instrumental music’. See Chua, Absolute Music and the Construction of Meaning (Cambridge, 1999), 34–5, 68–72.Google Scholar

37 This dichotomy originates in the work of the Russian Formalists. For various definitions see, for example, Victor Erlich, Russian Formalism: History, Doctrine (The Hague, 1965); Seymour Chatman, Story and Discourse; Genette, Narrative Discourse; Prince, Narratology and A Dictionary of Narratology; Jonathan Culler, The Pursuit of Signs: Semiotics, Literature, Deconstruction (Ithaca, NY, 1981), esp. Chapter 9, ‘Story and Discourse in the Analysis of Literature’, 169–87; Rimmon-Kenan, Narrative Fiction; and Bal, Narratology.Google Scholar

38 Culler, , The Pursuit of Signs, 171. An example of events in the ‘story’ would be: mother kills father; mother is arrested; son commits suicide. It can be argued that composers themselves conceive thematic materials as the basic events from which they build a narrative, as often they sketch those materials and then work out how to present them in sequence.Google Scholar

39 Rimmon-Kenan, Narrative Fiction, 3. One example of discursive emplotting of the previous events would be: ‘Right before the son committed suicide, his uncle told him that his mother was in prison because she had killed his father.‘Google Scholar

40 For other general comparisons between the nature of Beethoven's and Mahler's music, see Bekker, Paul, Gustav Mahlers Sinfonien (Meisenheim, 1921; repr. Tutzing, 1969), 11–23; and Adorno, Mahler: Eine musikalische Physionomik, 87ff.Google Scholar

41 Adorno situates the ‘classical conception of the symphony’ with its ‘well defined, well circumscribed diversity’ in the concern for economy originating in the Aristotelian and Cartesian rationalism. See Mahler: Eine musikalische Physionomik, 100.Google Scholar

42 Arnold Schoenberg, Fundamentals of Musical Composition, ed. Gerald Strang and Leonard Stein (Boston, MA, and London, 1967), 8ff. For further discussions of the principle of developing variation, see Frisch, Walter, Brahms and the Principle of Developing Variation (Berkeley and Los Angeles, CA, 1984), and Michael Musgrave's review of the same book in Journal of the American Musicological Society, 38 (1985), 631.Google Scholar

43 Schoenberg, Fundamentals, 103.Google Scholar

44 This corresponds to Meyer's ‘script’.Google Scholar

45 Adorno, , Mahler: Eine musikalische Physionomik, 117, 118.Google Scholar

46 In this respect they resemble more Schoenberg's definition of ‘melody’, which ‘extends itself by continuation rather than by elaboration’; ‘All these restrictions and limitations produce that independence and self-determination because of which a melody requires no addition, continuation or elaboration.‘ Schoenberg, Fundamentals, 102.Google Scholar

47 For definitions of denotation and connotation in the sense used here see, for example, Roland Barthes, Elements of Semiology, trans. Annette Lavers and Colin Smith (New York, 1967), 89–94; idem, S/Z: An Essay, trans. Richard Miller (New York, 1974), 6–11; and Umberto Eco, A Theory of Semiotics (Bloomington, IN, 1976), 54–7. I am adapting the sense of these terms by considering ‘denotative’ the most basic syntactic musical features that any trained musician would recognize. It is not always possible to draw a strict delineation between the syntactic and semantic levels, since some of the concepts we use in describing the specifically musical syntactic processes consist anyway of metaphors or analogies to the outside world. While in some situations we must content ourselves with acknowledging the ambiguity, most important in defining the signification of the observed phenomenon is making a decision about the relative weight of the purely musical components (tonality, rhythm, melody, etc.) and the referential ('extramusical') ones. (On this subject see, for example, Fred Everett Maus, ‘Music as Drama’, Music Theory Spectrum, 10 (1988), 56–73.) For other views on semantic signification in music see, for example, Jean-Jacques Nattiez, ‘Y a-t-il une diégèse musicale?’, Musik und Verstehen, ed. Peter Faltin and Hans-Peter Reinecke (Cologne, 1973), 247–57; Monroe C. Beardsley, ‘Understanding Music’, On Criticizing Music, ed. Price, 55–73; and Kofi Agawu, Playing with Signs, 23–5. Agawu's concept of ‘introversive semiosis’ (borrowed from Jacobson) would correspond to my morphological and syntactic (denotative) levels, while his ‘extroversive semiosis’ would correspond to my semantic (connotative) level.Google Scholar

48 Barthes, S/Z, 8.Google Scholar

49 Music lacks denotation in the traditional sense of the word; it has mostly ‘syntactic’ denotation. Charles Rosen also recognized this when he wrote: ‘Musical phonemes act directly without first being strained through an abstract system of denotation.’ See his ‘Art has its Reasons’, New York Review of Books, 17 June 1971, 38. However, I would disagree with Rosen when in the same review he states that ‘what it [music] lacks is a vocabulary’. I hope that this will become clear in the discussion below.Google Scholar

50 See Hawkes, Terence, Structuralism and Semiotics (Berkeley, CA, 1977), 89, 158. Chomsky's concepts of competence/performance coincide with those of langue/parole used by Saussure in his linguistics.Google Scholar

51 For definitions and applications of inter textuality, see, for example, Julia Kristeva, Semiotike: Recherches pour une sémanalyse (Extraits) (Paris, 1968), 52–5; Michael Riffatterre, ‘Intertextual Representation: On Mimesis as Interpretive Discourse’, Critical Inquiry, 11 (1984), 141–62; Culler, The Pursuit of Signs, 100–18.Google Scholar

52 George Grove, Beethoven and his Nine Symphonies (3rd edn, London, 1898; repr. New York, 1962), 192–3.Google Scholar

53 In fact, such associations are not ‘iconic’ or ‘isomorphic’, in the sense that the shapes of the music do not correspond somehow to actual objects in nature (the closest Beethoven gets to iconic signs is through the birdcalls and the thunder and lightning imitated later in the symphony). They, too, depend on a process of conventionalization, whose origins might be found in older ‘pastoral’ music. The most immediate associations, then, are of music with other music; in other words, they are intertextual connotations. Beardsley, following Nelson Goodman, pointed out that music exemplifies properties rather than objects. See Beardsley, , ‘Understanding Music’, 68, and Nelson Goodman, Languages of Art: An Approach to a Theory of Symbols (Indianapolis, IN, 1968), 52.Google Scholar

54 My understanding of the pastoral topic corresponds closely to that discussed by Hatten in Musical Meaning in Beethoven, 91–111. For definitions of topics, see the references to Agawu, Ratner and Webster, in note 32 above.Google Scholar

55 See in note 32 above references to the variety of topics in Classical music. For my purpose here I chose an extreme case where, probably for programmatic reasons, Beethoven insists on only one topic. The wider variety of topics one might find, for example, in instrumental works of Mozart does not invalidate the points made here about narrative, even though the degree of narrativity in those works might, paradoxically, be higher than in this programmatic work.Google Scholar

56 Diether de la Motte, ‘Das komplizierte Einfache: Zum ersten Satz der 9. Sinfonie von Gustav Mahler’, Musik und Bildung, 3 (1978), 145–51 (pp. 145–6).Google Scholar

57 See also Adorno, , Mahler: Eine musikalische Physionomik, 201 (trans. Jephcott, Mahler: A Musical Physiognomy, 155): ‘Telling of the past, the wholly epic voice is heard. It begins as if something were to be narrated, yet concealed.‘Google Scholar

58 Newcomb detects a similar quality: ‘The transformation of experience by memory is in fact one of the essential messages of narrative and of Mahler's Ninth’. See his ‘Narrative Archetypes’, 132.Google Scholar

59 Adorno, , Mahler: Eine musikalische Physionomik, 85 (Jephcott's translation emended). Donald Mitchell ascribes a similar incantatory, ritualistic function to the opening musical gesture of Der Abschied in Das Lied von der Erde. See his Gustav Mahler: Songs and Symphonies of Life and Death (Berkeley, CA, 1985), 355.Google Scholar

60 For a detailed consideration of the conditions that entered into the formation of what I call the ‘farewell myth’ of Mahler's Ninth Symphony, see my ‘The Farewell Story of Mahler's Ninth Symphony’, 19th Century Music, 20 (1996–7), 144–66.Google Scholar

61 Prince, , Narratology, 149.Google Scholar

62 See ibid., 61.Google Scholar

63 Nattiez, , ‘Can One Speak of Narrativity in Music?‘, 244.Google Scholar

64 See Prince, , Narratology, 62–3: ‘Events can be defined as stative, when they constitute a state …, or active, when they constitute an action. … The proportion of active and stative events in a narrative is an important characteristic of that narrative.‘Google Scholar

65 According to Gerald Prince's A Dictionary of Narratology, there are two types of narrative statements with which the discourse states the ‘story’, corresponding to the stative and active events: stasis statements and process statements (pp. 77, 90).Google Scholar

66 Chatman, , Story and Discourse, 19, and, for example, Rimmon-Kenan, Narrative Fiction, 3. Other definitions include: ‘the discursive presentation or narration of events’ (Culler, The Pursuit of Signs, 170); ‘the symbolic presentation of a sequence of events connected by subject matter and related by time’ and ‘a text which refers, or seems to refer, to some set of events outside itself’ (Robert Scholes, ‘Afterthoughts’, On Narrative, ed. Mitchell, 200–8 (p. 205)); and ‘the representation of at least two real or fictive events or situations in a time sequence, neither of which presupposes or entails the other’ (Prince, Narratology, 4).Google Scholar

67 I will deal later with the claim that the ‘causality’ and/or chronology linking the events of the ‘story’ in a literary narrative is lacking in music.Google Scholar

68 Prince, , Narratology, 145ff.Google Scholar

69 I do not mean to say that Beethoven's movement lacks expressivity, or that the various sections of the development do not bring in various moods and emotions. It is just that in terms of what the materials are ‘about’, the referentiality to the outside world is reduced to a very few references. In Subotnik's terms, this movement presents a universe closer to the ‘abstractly logical … ideal of meaning’.Google Scholar

70 Both Nattiez and Abbate support the idea of music miming or imitating narrative, rather than being one. See, for example, Abbate, Unsung Voices, 27, and Nattiez, ‘Can One Speak of Narrativity in Music?’, 257. I, however, think that texts are not narratives but, rather, may have and/or project ‘narrative qualities’ or characteristics, and in this respect music is not that different.Google Scholar

71 Unlike Adorno and Newcomb, who see the themes and their transformations ‘as characters in a novel, as beings that are constantly evolving and yet constantly identifiable with themselves, … they shrink, expand, even grow old’ (quoted from Adorno in Newcomb, ‘Narrative Archetypes’, 119), I find that the personification and anthropomorphization of the themes stretches too far the ‘narrative impulse’, as the analogy is too general and human behaviour could then be extended to any themes. I feel that my concept of ‘worlds’ is more concretely grounded in the semantic connotations of the materials themselves, and can thus more easily be historically and intersubjectively justified.Google Scholar

72 The same union between remnants of older forms (strophic variation, sonata form) and an irregular form (the combination of ‘strict and free’, as Donald Mitchell calls it) is active in Der Abschied. See Mitchell, Gustav Mahler, 344ff.Google Scholar

73 Smith, Barbara Herrnstein, Closure: A Study of How Poems End (Chicago, IL, and London, 1968), 99.Google Scholar

74 Meyer, Leonard, Emotion and Meaning in Music (Chicago, IL, and London, 1956), 151.Google Scholar

75 See Barthes, , ‘Introduction to the Structural Analysis of Narratives’, 93–7.Google Scholar

76 See Princeton Encyclopedia of Poetry and Poetics, ed. Alex Preminger (Princeton, NJ, 1974), 542.Google Scholar

77 Mahler seems to have been thoroughly aware of this potential of motive Y from the earliest stages of conception. The early orchestral draft of the symphony (probably finished on 2 September 1909) shows the Y motive used throughout the movement precisely with this function and only subsequently added to the introduction as an afterthought.Google Scholar

78 Barthes, , ‘Introduction to the Structural Analysis of Narratives’, 95. Shlomit Rimmon-Kenan calls these two types of events ‘kernels’ and ‘catalysts’ (Narrative Fiction, 16), while Seymour Chatman calls the second type ‘satellites’ (Story and Discourse, 53).Google Scholar

79 See Genette, , Narrative Discourse, 33–160.Google Scholar

80 Fred Maus seems to make a similar point when he writes: ‘[Jonathan] Kramer's gestural-time, the order that can be reconstructed from the qualities of musical actions, amounts to ordering in a story; piece-time is ordering in the discourse.’ See Maus, ‘Music as Narrative’, 31. He is referring to the terms employed by Jonathan Kramer in his article ‘Multiple and Non-Linear Time in Beethoven's Opus 135‘, Perspectives of New Music, 11 (1973), 122–45.Google Scholar

81 Genette, , Narrative Discourse, 40.Google Scholar

82 Newcomb, , ‘Narrative Archetypes’, 123.Google Scholar

83 To my knowledge, this quotation from Brahms has not previously been pointed out.Google Scholar

84 White, Hayden, ‘Commentary: Form, Reference, and Ideology in Musical Discourse’, Music and Text, ed. Scher, 288–319 (p. 296); repr. as ‘Form, Reference, and Ideology in Musical Discourse’, Figural Realism: Studies in the Mimesis Effect (Baltimore, MD, and London, 1999), 147–76 (pp. 151–2). The page numbers in subsequent references refer to the first publication.Google Scholar

85 Ibid, 293–4.Google Scholar

86 For a detailed discussion of the layers of meaning informing this ending, see my ‘Is Mahler's Music Autobiographical? A Reappraisal’, Revue Mahler Review, 1 (Paris, 1987), 47–63 (pp. 52–3, 60, and Examples 2–4).Google Scholar

87 By comparison, Schubert, who in his late Piano Sonata in B♭ can be said to have originated such lingering gestures, is less daring than Mahler in ending a piece with such mimetically disintegrating connotations.Google Scholar

88 Here I agree with Abbate and Nattiez that music does not constitute ‘a narrative’ in the sense of literature, but rather that it may at times mimic or create the illusion of a narrative. Yet, unlike them, I do not see ‘narrative’ as a superfluous metaphor in analysing music.Google Scholar

89 White, , ‘The Value of Narrativity’, 1.Google Scholar

90 White, , ‘Commentary’, 294.Google Scholar

91 Ibid., 293. Here White is in agreement with Nattiez, who writes: ‘In music, connections are situated at the level of the discourse, rather than the level of the story.’ See ‘Can One Speak of Narrativity?’, 244. I disagree with both.Google Scholar

92 White, , ‘Commentary’, 293.Google Scholar

93 Prince, , A Dictionary of Narratology, 64.Google Scholar

94 Steiner, Wendy, Pictures of Romance: Form against Context in Painting and Literature (Chicago, IL, 1988), 9. For ways of articulating the degree of narrativity, see, for example, Prince: ‘a passage where the signs of the narrated [the story] are more numerous than the signs of the narrating [the discourse] should have a higher degree of narrativity than a passage where the reverse is true’ (Narrativity, 146).Google Scholar

95 I am by no means the only one proposing that wordless music has meanings. See, among others, the works by Treitler, Maus, Agawu, Kramer and others cited throughout this article.Google Scholar

96 See Abbate, , Unsung Voices, 52–4.Google Scholar

97 Prince, , Narratology, 40, and A Dictionary of Narratology, 76. Barthes's ideas appear in ‘Introduction to the Structural Analysis of Narratives’, 98–100.Google Scholar

98 Barthes, , ‘Introduction to the Structural Analysis of Narratives’, 98.Google Scholar

99 Prince, , Narratology, 4.Google Scholar

100 See Bal, , Narratology, 143.Google Scholar

101 See my ‘The Farewell Story of Mahler's Ninth Symphony’.Google Scholar

102 Adorno, , Mahler: Eine musikalische Physionomik, 86: ‘Nicht Musik zwar will etwas erzählen, aber der Komponist will Musik machen, wie sonst einer erzählt‘Google Scholar

103 See Smith, Herrnstein, ‘Narrative Versions’, 213–14.Google Scholar

104 Johnson, Julian, ‘The Status of the Subject in Mahler's Ninth Symphony’, 19th Century Music, 18 (1994–5), 108–20 (p. 109).Google Scholar

105 Kramer, , ‘Musical Narratology’, 99.Google Scholar

106 Samuels, , Mahler's Sixth Symphony, 140.Google Scholar

107 Ibid, 133.Google Scholar