The Tristan Chord in Historical Context: A Response to John Rothgeb


William Rothstein

AUTHOR: Rothstein, William
TITLE: The Tristan Chord in Historical Context: A Response to John Rothgeb
KEYWORDS: harmony, recitative, intertextuality, Tannhaeuser, Purcell, Bach

William Rothstein
Oberlin College
Conservatory of Music
Oberlin, OH 44074

Return to the MTO page

John Rothgeb's analysis of the "Tristan chord" engages a large intertextual network, stretching back to the Baroque and centering on recitative. Examples of a specific figure of recitative, usually associated with the asking of a question, are presented and analyzed. Examples include passages in Wagner's operas before Tristan.

[1] Most of the discussion of Professor Rothgeb's article so far has focused, understandably, on enharmonic issues (e.g., can the TC be legitimately described as a half-diminished seventh chord?). In the process, it seems, the originality and cogency of Rothgeb's analysis of the TC in its original context--mm. 1-3 of the Tristan Prelude--has gone largely unremarked.(1) When I first read his essay, I found Rothgeb's analysis of these measures instantly convincing, and far superior to the conventional analysis of the TC's G#4 as an appoggiatura to A4 (the last eighth note of m. 2), or to William Mitchell's reading of a voice exchange within a composed-out V of A minor. The central feature of Rothgeb's analysis (see his Ex. 2) is his claim that G#4 properly belongs to the V chord of m. 3, but that a rhythmic shift--specifically, a measure-long anticipation--moves G#4 back to the downbeat of m. 2, thus displacing A4, the note that "properly" belongs to the French-sixth chord. This latter chord Rothgeb (along with many others) understands to be the chord "behind" the TC.(2) Unlike earlier authors, however, Rothgeb does not identify the displaced A4 with the eighth note at the end of m. 2. That A4, and the A#4 following it, Rothgeb reads as part of a slide embellishment--i.e., a passing motion--connecting the anticipated G#4 and the B4 of m. 3; the latter note arises from the voice-leading technique known as reaching-over (Schenker's Uebergreifen).

[2] Rothgeb's Ex. 2 isn't as clear as it might be about the rhythmic process that, in his analysis, gives rise to the TC. My Ex. 1 paraphrases his Ex. 2 in such a way that rhythmic issues are brought to the fore.

Ex. 1a is essentially the same as his Ex. 2a, except that chords are identified by their harmonic functions (T=tonic, P=pre-dominant, D=dominant), and the reaching-over-cum-voice-exchange is shown. Example 2b shows a rhythmic shift in the soprano, anticipating G#4 (the resolution of A4) and shifting B4 back to the downbeat of m. 2. (I will return to this form of the progression later.) Example 1c compounds the rhythmic shift by giving G#4 the entire value of the augmented-sixth chord; the resulting vertical sonority is, of course, the TC. Example 1d fills out the soprano's minor-third leap with passing tones, yielding Rothgeb's slide. Given the added passing tones, the alto's E4 would cause parallel fifths--the earlier "hidden fifths" having become open fifths--so D4 has been substituted; this is a common voice-leading contraction (a leading tone descends directly to a passing seventh, a chromatic semitone lower, instead of resolving upward as usual and then passing downward by whole step). Finally, Example 1e shifts the second passing tone in the slide, A#4, to the downbeat of m. 2. It will be noticed that Ex. 1e is virtually a durational reduction of Wagner's mm. 1-4; the principal difference is that m. 1 is represented in the example by a complete tonic triad.

[3] I have followed Rothgeb's harmonic analysis thus far, but I don't really hear m. 1 as a tonic harmony in A minor. F4 sounds to me like a consonant chord tone, not an appoggiatura. Along with Deryck Cooke, I hear the ascending sixth A3-F4, in upbeat-to-downbeat-rhythm, as an arpeggiation from the fifth to the third of a D-minor triad, which proves to be IV of A minor (3); the other two ascending sixths, in mm. 5 and 8, are similarly arpeggiations from the fifth of a triad up to its third. I have added Example 1f to represent this hearing of m. 1. The bracketed sixth in the example represents the cellos' ascending sixth. The bass and soprano notes in parentheses are, to my ear, implied by the larger context; notice the chromaticized voice exchange between bass and alto. I hear G#4 as coming from the implied A4 (so does Rothgeb, but he hears the elided A4 in terms of I, not IV). The entire progression thus represents a motion from pre-dominant (IV) to dominant (V), with an augmented-sixth chord (an altered pre-dominant) in between. The alto's E4 is a passing tone, analogous to the eighth-note A4 in the soprano. It is a simple matter to substitute the first chord in Example 1f, IV, for the tonic triad in Examples 1a-1e; the progression from level to level remains essentially the same.

[4] I now return to Example 1b, which, I here claim, forms part of the pre-history of the TC. Devotees of vocal music will quickly recognize the latter part of this example--everything, that is, but the opening tonic chord--as a common figure in recitative. With its half cadence and its rising melodic gesture, this figure was typically used by composers to set lines of text ending with a question mark. This "question" figure was taken over by Wagner in his early operas; it plays an especially important role in Tannhaueser. The same figure is transformed into the so-called Fate motive in the Ring. I was aware of the "question"/Tannhaeuser/Ring connection before I read Rothgeb's article, but his analysis of the TC adds Tristan to the intertextual network.

[5] The earliest example of the "question" figure that I am aware of--but almost certainly not the earliest that exists--is the very opposite of a question: it is Belinda's announcement of Aeneas's first entrance in Purcell's Dido and Aeneas (see Ex. 2).

Belinda's words are "how godlike is the form he bears." The example is in C major, although the succeeding passage is in G. The figure involves a bass motion from ^6 to ^5--here in major instead of the usual minor--with the downward melodic resolution, C5-B4, anticipated in the same way as in Example 1b. D5, at the end of the passage, comes from E5, the third of the tonic triad, which was itself reached by arpeggiation in the first two measures of the recitative. Thus the melodic structure is based on a pair of unfolded thirds: E5-C4 (mm. 2-3) is answered by B4-D5 (mm. 3-4).

[6] Many examples of the "question" figure, and of figures closely related to it, occur in the cantatas of J. S. Bach. (Curiously, I have found no examples in Bach's Passions.) In each case, the bass progression is ^6-^5, usually in minor; often, but not always, the figure concludes a bass progression descending by step from I to V. The harmonic progression, naturally, is from pre-dominant to dominant, with IV6-V the usual choice. The melody, at the half cadence, features either an ascending leap or an ascending passing motion from ^#7 to ^2. Those examples that include melodic passing tones, naturally, recall the Tristan progression most closely.

[7] Example 3, from Bach's Cantata No. 82 ("Ich habe genug"), is especially striking in connection with Tristan.(4)

The middleground harmonic progression, beginning in m. 2, is I-IV-V in C minor. At the fourth quarter of m. 4 the bass moves from the root of IV, F2, to its third, A-flat2; at the same time, two tones of the V triad--B-natural and D--are anticipated (the continuo figure is 6/4/2+). The vocal line connects B3 and D4 with a passing tone, C4; in most editions (notably that of the Bach-Gesellschaft), D4 is embellished with an appoggiatura from below, so that the passing C4 is heard twice, once before and once after the bar line. (Example 3, from the Neue Bach-Ausgabe, does not show this appoggiatura, but a singer would probably add it--also an upper appoggiatura on the word "Erde.") If the lower appoggiatura were made chromatic--C#4 instead of C4--we would hear a progression very much like Wagner's, lacking only the augmented sixth in the 6/4/2+ chord and the seventh in the V.

[8] Example 4 is a voice-leading interpretation of Ex. 3.

The passage is complicated, as one comes to expect of Bach. Level A, a foreground graph, omits the registral play of the vocal line, which descends into its lower octave to illustrate the "cool earth" ("kuehler Erde") into which the singer longs to be buried. Level B, a middleground layer, shows more clearly that the passage as a whole is governed by a series of parallel tenths between the bass and the vocal line; the bass leads.(5) The last two tenths of the series are displaced by anticipations and by an unfolding in the bass (shown at level A). C3--C4 in the graph--is anticipated at the end of m. 3, although it is sustained by the keyboardist through the first three beats of m. 4. B3 is anticipated on the fourth beat of m. 4, as indicated above. While F2 in m. 4 is the root of the IV harmony, the controlling bass tone within the larger progression is A-flat2, reached only at the fourth beat of the measure--precisely the moment when B3 is anticipated. Thus the tenth A-flat2/C4 is never explicitly represented at the musical surface (remember that the last sixteenth note of m. 4 is passing). The anticipation of B3 leaves time for the reaching-over of D4, a note that connects back to E-flat4 in m. 2.

[9] I do not know whether Wagner took the "question" figure from Bach, some of whose music he knew very well, or from later composers. Wherever he got it, he used it in Der fliegende Hollaender (I haven't looked for it earlier than that), and especially in Tannhaeuser.(6) In both operas, the pre-dominant chord used is generally an augmented-sixth chord of some kind; the minor third from ^#7 to ^2 is expressed as a leap, not a slide. If the local key is major, ^6 in the bass is typically flatted, as it must be if an augmented-sixth chord is to be used.

[10] The figure appears frequently in Act 3, Scene 3 of Tannhaeuser. While it is not always used to express a literal question, the figure's searching quality captures nicely the restless unfulfillment of Tannhaeuser's quest. Interestingly, almost all appearances of the figure in this scene are in either A minor or C major, precisely the two keys used at the beginning of Tristan. Ex. 5 shows two of the figure's occurrences. The first, in A minor, sets a question from Wolfram to Tannhaeuser: "Zogst du denn nicht nach Rom? (Didn't you go to Rome?)"

The second, in C minor/major, is of special interest because--in the vocal score, at least--it briefly introduces a transposition of the TC, on the last eighth note of the first measure (in fact, C4 is sustained by half the violas, so the chord is literally a French sixth). B-natural is, of course, an anticipation. Tannhaeuser is telling here of the mortifications he visited upon himself on his way to Rome (the text at this point is "vergoss mein Blut ich zu des Hoechsten Preis").

[11] The final link in the chain is the Ring. By the time he composed Tristan, Wagner had completed Das Rheingold and Die Walkuere, and had composed the first two acts of Siegfried. In Die Walkuere, the "question" figure appears most notably as the Fate motive; it also ends the "Todesklage" motive. The two motives are closely linked in Act 2, Scene 4, the "Todesverkuendigung." Whereas all earlier examples of the "question" figure have involved a melodic leap from ^#7 to ^2--the third and fifth, respectively, of the V triad--the basic form of the Fate motive involves a leap from ^2 to ^4, the fifth and seventh of V7.(7) This variant of the figure is made possible by the nineteenth century's acceptance of V7 as a legitimate goal for a half cadence. In Wagner's syntax, of course, such a V7 does not require conventional resolution, although the listener should (I think) still feel a yearning for such resolution.

[12] Ex. 6 shows two consecutive occurrences of the Fate motive in the "Todesverkuendigung."

As in the Purcell example (Ex. 2), the text contains an exclamation, not a question: Siegmund asserts that he will not follow Bruennhilde to Valhalla ("zu ihnen folg' ich dir nicht!"). The passage as a whole is in F# minor, but the first statement of the motive sounds like A minor. The progression moves from a French-sixth chord to V; the anticipated leading tone, G#, forms the TC at its "home" transposition level. The following statement, in F# minor, represents the original form of the Fate motive, with a progression from an incomplete diminished-seventh chord, 6/4/2+ (cf. the Bach passage in Ex. 3), to V7. Wagner's spelling here suggests that the apparent D-minor triad is not the "real" chord in the motive's first measure, but that A is an accented passing tone resolving to G#.(8) (Hearing A in this way also makes the outer-voice parallel fifths easier to accept.) Notice that, in the final measures of Die Walkuere, Wagner changes his spelling from E# to F-natural, emphasizing the Phrygian progression from F-natural to E and thus the plagal--as opposed to pre-dominant--function of the Fate motive at the end of the opera.

[13] I have tried, in this response, to support John Rothgeb's analysis of the TC indirectly, by drawing a thread back through history. The conceptual evolution of the TC shown in Ex. 1 mirrors, to a remarkable degree, the historical evolution of the "question" figure, much of which took place within Wagner's oeuvre. Wagner's seeming preoccupation with this figure supports the notion that, in the opening measures of Tristan (if nowhere else), ontogeny recapitulates phylogeny. The gradual dissolution of tonal syntax that Tristan did so much to further made continued evolution of this sort extremely problematic. In terms of tonal syntax--harmony and voice leading--Tristan represents a station very near the end of the line.

Return to the MTO page


1. In this response, I will use "TC" to mean exclusively the first verticality in m. 2 of the Tristan Prelude. Unlike Rothgeb, I accept the notion that this chord participates in a network of collections--all of them reducible, in the abstract, to the collection [3, 5, 8, 11] and its transpositions--extending over the entire opera. To my mind, it is this network that ultimately constitutes the "Tristan chord," not the specific manifestation in m. 2. However, a thorough understanding of the language of Tristan requires that each occurrence of the chord/collection be analyzed independently for its harmonic, contrapuntal, and (ideally) timbral context. It is the harmonic/contrapuntal analysis of mm. 1-3 that I find most exciting about Rothgeb's article.(return to text)

2. A work cited in Allen Forte's recent response to Rothgeb (MTO 1.2 [1995]), the Harmonielehre of Louis and Thuille, accepts the TC as an independent augmented-sixth chord, adding it (and two others) to the inventory of augmented-sixth chords. I thank Daniel Harrison for bringing this fact to my attention.(return to text)

3. See Deryck Cooke, "The Creative Imagination as Harmony," in Robert Bailey, ed., Wagner: Prelude and Transfiguration from Tristan and Isolde (New York: Norton, 1985), pp. 169-76. This essay is excerpted from Cooke's The Language of Music (Oxford, 1959).(return to text)

4.. Readers who own the Dover volume Bach: Eleven Great Cantatas (New York, 1976) can find other examples of the "question" figure on p. 61 (m. 4 and mm. 8-9 of the tenor recitative "Wie hast du dich, mein Gott" from Cantata No. 21) and p. 157 (mm. 7-8 of the tenor recitative "Der Heiland ist gekommen" from Cantata No. 61).(return to text)

5. On the concept of leading and following linear progressions see Schenker, Free Composition (Der freie Satz), trans. and ed. by Ernst Oster (New York: Schirmer Books, 1979), pp. 78-80.(return to text)

6. In Der fliegende Hollaender, see Erik's question "Welch' hohe Pflicht?" in his dialogue with Senta.(return to text)

7. Several occurrences of the Fate motive in Die Walkuere explore still other tonal relationships. See, for example, Bruennhilde's question to Siegmund, "So wenig achtest du ewige Wonne?", in the "Todesverkuendigung."(return to text)

8. At the beginning of this scene, where the Fate motive appears at the same transposition level, A is a suspension. Its tendency to resolve downward is intensified there by the fact that its preparation is a dissonant seventh (V7 of E minor).(return to text)