Of all we find in Webern’s early atonal works that engages our ears and our analyses, surely among the most mysterious and cryptic are those moments where the piece slows to a halt, only to be led away by a final gesture that breaks off into the void. These tiny “flyaway” endings, two of which are the subject of this essay, have a distinctly Webernian ring to them, and coalesce with his earlier style, often said to be fragmentary or aphoristic.[1] As endings to ensemble pieces performed by a single voice, they enact a kind of loneliness, a disengagement from the communal atmosphere of various voices interacting. Musical closure in this environment becomes a complicated matter, especially since the atonal idiom lacks an a priori criterion for closure, something native to the tonal system.[2]

While it is rare to find an analysis geared specifically towards these moments, analysts do offer perspectives on the issue.[3] Some of these accounts will be considered below, but the analyses proffered in response were constructed with a view to problematizing the endings, showing them to resist closure within their contexts at least as much as they can be understood to close them. And by aspiring – at least – to arriving at rich and elaborate descriptions of the contexts in which the endings take part, I hope, too, that a few additional aesthetic perspectives on early Webern can be gained.

Vier Stücke für Geige und Klavier, Op. 7, No. 4

The ending of Op. 7, No. 4 is a peculiar creature. It moves in only one direction – downwards – as it outlines a broad sweep across <D, Bb, Eb, B, G, C, F#>. It also emerged from primordial obscurity, existing for its initial ten years as an un-pitched glissando: “A sound more like noise,” wrote Webern of the figure to violinist Arthur Hartman. “The sliding [should be] fairly quick, so that the harmonics are not too distinct. In any case: the most extreme delicacy; a barely perceptible sound must be produced.”[4] Originally accompanied by the direction, “über die ganze G-Saite leicht rutschen” and a long, descending squiggle, Webern’s original ending literally vanishes into thin air, sliding off the instrument, while in both drafts maintaining the same metrical starting points.[5]

With respect to the figure’s role qua the ending of both a piece and a short cycle of pieces, commentators Allen Forte and Felix Meyer and Anne Schreffler highlight the continuity afforded by the pitched version and its coherence as an ending. As a glissando, Meyer and Schreffler argue, the figure follows a textural pattern in the violin that begins with the virtuosic opening gesture, and then dematerializes through successive stages of harmonics (mm. 7-9, 10-12), pizzicato (m. 9), col legno (mm. 9-10) and lastly the glissando itself, which, when situated at the end of this process, “subverts our normal expectations of a final movement by denying any form of closure.” Pitched, and with its specific rhythm, the figure both connects with other significant sets of the piece, and, by virtue of its very determinedness, provides more suitable closure for the movement and the cycle:

It also sums up in a single gesture the main registers in which the violin part has been active … By substituting pitched for non-pitched sound and specific rhythms for an amorphous shape, Webern strengthens the sense of closure, appropriately enough for a passage that concludes a whole cycle as well as a single piece.[6]

Forte’s analyses, the other hand, yield little in the way of aesthetic commentary, but do point out various correspondences between set classes found in the ending and other parts of the piece. For instance, he focuses on the initial four pitches of the ending as a T₉ reordering of the opening figure of the movement with D as a common tone.[7] In addition, Forte mentions the importance of [0, 1, 4, 5] as well the presence of the set class formed by first six notes [0, 1, 3, 4, 5, 8] as a cadential sonority in the second piece and the last six notes as [0, 1, 2, 5, 6, 9], the set associated with Schönberg’s name.[8] Although the T₉ reordering, for instance, is “by no means a replica”[9] of the opening gesture, presumably the implication is that this is at least a relatively conflict-free ending since, “if a segment produces a pitch-class set that does not occur elsewhere in the music, it is probably not an effective structural component,” “effective structural component” being about the most stable thing in the world for Forte.[10]

None of these accounts provide a robust portrait of the ending. While Forte tends toward the taxonomic side of pitch matters, Meyer and Schreffler, who do offer insights into non-pitched aspects of the piece, do not go far enough when they argue for a conclusive ending on the grounds of the presence of pitches alone. A more sensitive portrayal of the ending requires not only a journey through the final movement, but also a look at the earlier events of the piece (beyond set classes), for which the final movement functions as a refractive recapitulation. The much-discussed opening and third movements are ripe with possibilities that are taken up by the last and it will be necessary to refer to them as we go along.

Casting off the mute, the violin begins the final movement with a bold gesture, a widely spaced {D, G, F#, D#} [0, 1, 4, 5]. Certain properties of this set, and its particular unfolding, are of some importance. The set can be viewed as a perfect fourth with two interior notes, each a semi-tone off the boundary pitch classes. As the opening interval, <D, G> already picks up on an important trace of the first movement, which, with its conclusion, constructs this particular game.

One of the most striking features of the opening movement is the violin’s cool, aloof insistence (col legno, weich gezogen, slurred and with tenuto markings) on the repeating C#/Eb figure in mm. 5-8. This figure undergoes a ritard, coming to rest on the D resting lying between C# and Eb.[11] What follows is the famous on Eb major triad in 6/4 position. Never discussed is the melodic detail <D, G> that arises in the upper voices – and it is this detail that will have legs in the fourth movement. Heard in this frame of mind, a peculiar descending fifth – the “wrong” fifth for a tonal ending – unfolds with a completed triad hanging underneath its second pitch. The “tonal” ending thus disturbed, the <D, G> motive allows the triad to glow in a strange new light, with the guiding melodic fifth looking forward.[12]

Other evidence from the first movement supports an emphasis of ic5 space and the pitch classes D and G. For instance, notice the piano’s Tristan-like gesture beginning on the low Eb of m. 3.[13] Although a Tristan chord does not follow this detail, the upper line in the piano’s right hand does articulate an ic5 space, <F#, G#, B, Bb, A>.[14] As Robert Morris has pointed out, the pentachord outlined by the violin in mm. 3-5 (not including the violin’s C#) is the same set-class [0, 1, 2, 3, 5] as the upper line of the piano’s right hand just discussed.[15] Significantly, the violin’s first and third pitches outline a fourth from D to G, separated by the E that falls in between. Looking ahead to the ending with an eye and an ear for melody, we see that once the violin steps off its Eb, the pitches E, D and G are brought into a rapport with one another, adding meaning (by remembering the melodic origin of these pitch classes) to hearing the end as a melodic succession whose final pitch happens to have the rest of a triad hanging off it.[16]

An examination of the fourth movement shows these elements of the first movement receiving further treatment, quasi-recapitulatory in nature. Following the opening motive already discussed, the piano conflates the violin’s C# and Eb into a vertical sonority, maintaining the same register. The phrase then draws to a close in mm. 4-5 with another near-tonal moment, this time articulating a kind of altered dominant on G, with C mixed in. The pitch classes of the initial Tristan chord are all present in the right hand. After the opening motive, this time under T₁₁, sounds again in the piano, the piano emanates from G, the goal of the rising motive begun in m. 5.[17] From G the pitches are arranged in nearly symmetrical fashion with the crucial pitch G very audibly emphasized. Example 1 illustrates the symmetry.

The piano’s first pitch in m. 5 is G, the initial note of another [0, 1, 4, 5] {G, Ab, C, B} motive, this one extending exclusively upwards.[18] With the B that concludes this set, the upper line of the piano commences a dramatic recollection from the first movement, indeed the key memory. The violin B is part of a slur from the previous bar – a slur from G to the high D. Once D is achieved, the top line of the piano – sounding very much like a melody unto itself given its extension from the previous B – toggles between C# and D until C# moves to Eb, both marked tenuto, in m. 9.[19] This is a reversal of the state of affairs of the first movement, where the violin’s incessant C#/Eb movement resulted in a dissolution on D. Here, the music works its way through D in the struggle once again to regain C# and Eb, accompanied by a few thwacks on B, which infuse the moment with the memory of the conclusion of the second movement.

The final line of the score is mysterious. Following the regaining of C# and Eb in the melody, the piano strikes a peculiar chord that lingers into m. 10 with a fermata. Inscrutable, the chord is comprised of two familiar configurations from the first movement and the local symmetry suggests D, instead of G, as the axis of symmetry. This is illustrated in Example 2a, b.[20] As shown in the example , the C# harmonic in the violin interacts with the D axis when it enters. That same C# also participates in the formation of [0, 1, 5, 6] {1, 8, 7, e}, the most prominent sound of the third movement. The violin A of m. 12 almost sounds wrong, since it completes a distinctly diatonic tetrachord [0, 2, 3, 5] {0, 7, e, 8}, but the saturation of recapitulated elements continues: remaining within the sound world of the third movement, the A combines with left hand G# to recall the violin pitches from mm. 10-11 of that movement.

Two shadowy figures overlap as the violin articulates the first of the two flyaway endings as haze gives way to haze. The piano’s <C#, D, Ab, G> is the only genuine attempt at a true conclusion that acts fully as a summary; the pitches D and G receive emphasis by neighbors a semi-tone away in a kind of “resolution” on two pitch classes that have received so much emphasis. Yet this would be too facile, and Webern has complicated the ending. For one, {C#, D, Ab, G} make [0, 1, 6, 7], a sound that has received rather less emphasis at this point, save the combination of C# with the final three pitches of the first flyaway motive. And in addition, the reliance on a stable sonority for closure might have been overkill in a work that featured a triad at the end of the first movement.

As a set, [0, 1, 6, 7] concludes a gradual stretching of the initial set: [0, 1, 4, 5] at the beginning of the movement has become [0, 1, 5, 6] in mm. 10-1, which itself had been stretched by another semi-tone in mm. 13-4. Ostensibly stable, the piano G of m. 14 creates a platform for the first three pitches of the final flyaway motive and supports only these three. The combination of pitches here tantalizingly produces an Eb major7 chord, a sound whose catalyst is the pitch succession of G, D (piano, violin). This is supported by Webern’s choice of dynamics; the decrescendo does not start until the B in the second iteration. In addition to giving us another angle on these two pitches, it functions as a recollection of the ending of the first movement, whose Eb major triad had been colored by a melodic link between D and G. Here the pitches are in reverse order and they ascend. The entire chord has become a linear event, dematerializing the solid chord into a series of fleeting, ghost-like notes. And yet for all this concentrated recollections, the final four pitches ultimately disengage from it all. Unaccompanied by any other sounds – unlike the first flyaway motive – the music draws to a close on F#, dislodging from the security of G, which sounds almost like an afterthought in <B, G, C>, T₉ of <D, Bb, Eb>. With F#, it also offers us a new continuation following the transposed pitches, stretching the previous <Eb, B> into <C, F#>. Despite the final four notes’ formation of [0, 1, 5, 6], a sound with which we are by now familiar, simply appealing to set class does not capture what the ending does for us, namely to offer a reflection on much of what has come earlier in the piece only to drop it altogether, gazing outside the frame, as it were, and uncomfortable even with itself..

Fünf Sätze für Streichquartett, Op. 4, No. 4

Although Op. 5, No. 4 falls into three sections, the thematic arrangement does not match that of Op. 7, No. 4, mainly because the quartet movement explicitly revisits material from earlier in the piece, whereas thematic restatement as such was not a feature of Op. 7, although, as discussed, there are many other modes of recapitulation at work. An opening section (mm. 1-6) slowly unfurls new pitch classes starting with two motto chords that become linear events passed between violins and cello, before new thematic material and the first of three flyaway motives (mm. 5-6) draw the section to a close. A delicate and chilling middle section, not obviously related to the first, is itself concluded by another sounding of the flyaway motive, this time at T₅, before the material from mm. 5-6 sound in a much higher register. One final flyaway (T₈) marks the end of the piece.

Analyses of the Op. 5, No 4 usually point to the ending’s capacity to complete, both of terms of processes already underway or with respect to the transpositional level, as it appears at a total of three levels of transposition. Charles Burkhart, for instance, considers the ending both to complete two consecutive aggregates (the last three pitches, however, are extraneous in this regard) and to associate, finally, pitch classes Eb and F#, who had been eyeing each other from across the room, so to speak, but who were not yet on speaking terms.[21] The latter interpretation is very much in line with the flüchtig of the ending, but combined, the two interpretation tend to overstate the extent to which the ending is truly conclusive.

David Lewin, whose main motivation is to come up with an intellectual rationale for the issue of why T₈ was selected as the final level of transposition, points to the balancing quality of the transposition with a view to seeing the motive as a serial event and not just a pitch one.[22] Example 3 is a reproduction of Lewin’s Example 1, where he illustrates the idea. L is the first instance of the flyaway figure. By reading its first interval backwards from the first pitch, he generates a new flyaway, L₈, beginning on G#. Similarly, he reads the last interval of L₈ backwards and generates a third flyaway, L₅, by spelling it backwards from the newly attained initial pitch, in this case Eb. Sharing intervallic relations and common tones at the beginning and end of L₈ gives it the balancing quality that so interests Lewin here and in other publications.

Not yet “intellectually convinced,” he scours the rest of the piece for evidence of affinity with his diagram. He finds it in a figure that presents the same “G# extended backwards idea,” along with the explicitly stated G#-C in the first violin in m. 9. The extended backwards idea comes into play at the beginning of the contrasting middle section, where an [0, 4, 8] sound is present in general. Here in m. 9, the pitch D precedes the onset of the set {Bb, Gb, E, B}, which, as it is often pointed out, is an inversion of the first four pitches of the first flyaway sharing three common tones. He is after the idea of a major third preceding the opening intervals of a flyaway motive. The final {C, Eb, F#} sound of the last two pitches of L₅ and L₈ also signal the relevance of the balancing effect.

It is not at all clear, however, why this particular correspondence contributes to L₈’s having “a balancing and resolving centric function suitable for closing the piece, formally somewhat analogous to the centric function of a tonic with respect to its dominant and subdominant.”[23] Presumably the inversional relationships results in the ending sounding as a kind of “answer,” but to complete his analogy, wouldn’t that answer be expected to surround the focal pitch C, or at least G#? It is also a challenge to imagine how such a rigorously abstract concept – it is this concept, after all, that intellectually convinces Lewin of something by which he has been intuitively convinced, namely the choice of transposition of the ending flyaway – can be a stand-in for something as familiar as the resolution to a tonic, which has been accessed via its subdominant and dominant. Perhaps he intends this as some sort of recondite, composerly game employed by Webern, but by ascribing the notion of “function” in his analysis of T₈ along with an appeal to the signal aspect of tonality, surely perception – and to a high degree – is something he is working towards.[24]

The idea of balancing is at least abstractly vivid in Lewin’s pointing to the beginning and ending intervals of the figure as transpositional motivators, but there are further complications with hearing the serial relations at the end of the figure. Are we, for instance, to imagine the flyaway figure as a backwards entity? If so, there are no obvious features of the piece that suggest we should do this. Moreover, he doesn’t seem to have an “extending after” explanation from elsewhere in the piece that would enable us to begin hearing this property of Example 3 other than the idea of a diminished triad, in which case the example lacks a good deal of specificity, precisely the virtue evinced by the D-preceding example. In any case, though intriguing and characteristically original, Lewin’s analysis poses a problem, for if the aural aspect of the concluding function cannot be more firmly established outside of the world of imagination and analogy, what does this say about the relationship between our intellectual faculties and musical intuitions? An alternate account of the movement will hopefully provide us with a richer, hearable, more contextual portrait of the ending.

My reading turns on the seizure of a tiny detail, though not the detail that is usually upheld as germinal. It is customary to note the minute pitch alternation between the chilling opening chords, where one can observe the melodic either detail E-F# or the semi-tonal motion F-F#, which represents the pitch swap.[25] In Burkhart’s analysis, the T₂ represented by E-F# kicks off a process that results in T₂ of the second chord in mm. 5-6, {C#, D, F#, G}. It is undoubtedly true that the set-class perspective, along with the appropriate segmentation, will uphold this relationship. If “total intervallic palette, transposed or inverted” is the principal category in your conception of sound, this will do that work for you. But even with respect to pitch-class, sound can be more than an interval vector. Notice that although the segmentations reveal identical set classes, the constituent pitches are doing different things. In the opening they are verticalities. At im Tempo they become discrete, ordered arpeggios, quietly completing themselves before (or coterminous with) the arrival of the next arpeggio. In mm. 5-6, on the other hand, the point seems rather to be D going to C# and F# going to G. With this in mind, the sum total of these four pitches becomes a different entity – two motions of a semi-tone as opposed to a set per se. We can discuss relatedness as a function of intervallic similarity or even in terms of composing out a property inherent to the opening two sets, for which the semi-tone is one of their signature properties, but in an important sense these bits of music represent different events, which are not adequately summarized as T₂. The pitches in mm. 5-6 are at most a fragmentation of [0, 1, 6, 7] and perhaps something more remote.[26] The subsequent measure – the first appearance of the flyaway motive – supports the idea of gradual fragmentation; opening chords appear solid, then in linear groups of four accompanied by a viola line derived from the opening, then linear groups of two before the airiness of a single line takes over entirely. The point is that in the analytical context of motion between individual pitches, the motion from C to D in m. 9, reminiscent of the E-F# motion of the opening, is far too much for C# and D to countenance and that from this disturbance, which C# promptly extinguishes, arise far-reaching consequences that influence our story of the movement.

Timidity characterizes the opening of the movement, with its gradual accrual of new pitch-classes. We have already noted the addition of F# in the second measure and we observe, further, that it creates an [0, 1, 6, 7], which, in this configuration, is hardly a departure from the initial [0, 1, 5, 6]. But even this microscopic change has exacted a toll: the disposition of chord tones has undergone a voice-exchange, where minor ninth B to C has become major seventh C to B. The result is small but is not difficult to hear. With this level of hyper-sensitivity, the low Eb in the cello seems the most crass, intrusive transgression imaginable; it certainly contributes to the mild upsetting of voices in the upper chords just mentioned. Perhaps its ability to sneak in only minimally detectable – by listeners and analysts alike – despite its potential to disrupt is precisely the point, but in any case, it is quickly assimilated into the sound-world of the opening, as it provides a major seventh platform for the viola E-F# to meld with.

Meanwhile, the melodic detail <E, F#> has begun a story of its own. First emerging as tremulous and hesitant in the form of the top voice of the opening chords of the piece, E and F# begin to gain life in measure 2, calling attention to itself more explicitly by emanating out of the major seventh created between the viola and cello, the figure takes on some security as the viola’s only pitches in that measure. Then appearing atop the texture once again, E and F# constitute the first violin’s highest pitches in the pizzicato chord at the end of m. 3 and the first pitch of the arpeggio in m. 3. Thus E and F# have “mastered” the upper and middle registers of the texture. Moreover, the arpeggio is crucial to the idea of the piece being “ready” to progress as the result of its “security,” since one of the opening sets is now comfortable enough to be arpeggiated. By the time the viola sounds E and F# in this measure, they have become routine, the piece poised to carve out more territory. And carve it does; a lone cello C# joins the fray on the last eighth note of m. 3. Combining with the F# of the viola, the C of the first violin and the cello G on the downbeat of m. 4, the intervallic results thrown up are nevertheless conservative, as even progress is attempted with the utmost care. The product of this combination is {C#, G, C, F#}, or yet another [0, 1, 6, 7] which provides a platform for another arpeggiated motive. Yet although the new arpeggio does exhibit three pitches (B, E, F) that are familiar, the new pitch, Bb, along with the new tri-tone Bb-E and the liberating from having to play am Steg is perhaps a bit ambitious at this point. As a corrective, the cello, without being given a platform as the second violin had and returning to playing am Steg, arpeggiates the original second set at T₀ and the new material beginning with the violin C at the end of m. 4 seems to being saying, “enough of that.”

The new line in the violin in at the end of m. 4 combines with itself (the cello follows on the next eighth note) along with the viola – here acting as a kind of tenor line – to form numerous instances of [0, 1, 5, 6] and [0, 1, 6, 7], and has already been reported on in some detail in the literature.[27] What concerns us here is the introduction of a competing rising semi-tone <C, D>, the only other figure thus far. The emphasis with which the music seems to resist this challenge has already been commented on. The key detail here is, of course, the incommensurability of C and D with E and F#, which have been now established their dominance as the premiere whole-tone thus far. As the second D dissolves into C# in the cello in m. 6, the first flyaway motive occurs. The first four notes struggle to re-establish the four pitch-classes with which the first violin began; struggling because of the interruption the music has undergone after experiencing the material in m. 5 with its cease and desist order and rival whole-tone. And as Burkhart notes, three of the new pitch-classes that appeared in the first five measures now appear as the last three notes of the flyaway motive in the order of their first appearance.[28] Thus the flyaway figure is a capsule summary of the pitch class territory attained so far. Missing, of course, is the “Other,” the D.

D, however, returns in the second phrase as the very next pitch in the viola, a reflection of its banishment from the first phrase. Measures 7-10 are drastically different from what has come before and perhaps the main differences occurs first. The augmented triad in the viola is the characteristic sound of the second phrase and is markedly absent from the first, with its emphasis on ics 1 and 5. Conceivably a link could be forged by a whole-tone collection as a superset – the one that would include C, D, E and F# – but this is suppressed in the first phrase. Enacting what Julian Johnson identifies as a feminine and angelic trope in Webern[29], the second phrase is delicate and brittle. Analyses typically begin by identifying the pitches {Bb, Gb, E, B} as an inverted form of the first four pitches of the flyaway motive that shares three pitches. While this surely plays a role in motivating the effect of the section, there are other observations to be made, specifically regarding symmetries.

Although the sets that began the piece are highly symmetrical – especially [0, 1, 6, 7] – symmetry did not play any obvious role in the first phrase. Here, there are two clearly articulated axes of symmetry, both shown in example 4a.[30] The Bb is heard as the center of the augmented triad in the viola. The cello’s E lies a tri-tone from Bb and thus fits into the Bb scheme. The second violin’s B-natural is more difficult to assimilate, but it may have been designed to participate in the first violin’s axis of C/B, the “major” and “minor thirds” of the first violin’s quasi-triadic melody. The B does return to maximum disturbance during the most delicate moment, the verklingend in m. 9, where the first violin comes to rest on C. The Bb axis even catches the first two pitches of the second flyaway motive, as illustrated in example 4b.[31]

Part of the alienating effect of the second phrase is this conflict of axes, which are promptly dropped in m. 10 following the re-assertion of E and F#, the effect of which is one of disengagement; the low F and A are the last vestiges of the Bb-symmetry (the pitch-interval answer to the high Eb and B) and their passing marks the exit of the world of the second phrase, though its influence will be felt later. I also attribute an imaginary or hypothetical quality to the phrase, which goes to the heart of the E/F# and C/D story. Simply, Webern is inviting us to envision a world in which these two prominent and complementary melodic dyads know each other in a scenario that was all but impossible in the first phrase.

The suggestion of a whole-tone collection is strong, and the only candidate is the collection that includes all four pitches. Except that F# has been respelled and its context inverted (remember {Bb, Gb, E, B} as inversion of the first four notes of the flyaway), applying some new color to the dyad. Moreover, C comes at a crucial point (verklingend) in the phrase. At the beginning in that measure (m. 9), E and Gb, with the enharmonic respelling, give way to C and D at the end of the measure (violin 1 and viola). Only C and D have been blown up into a minor seventh. Perhaps most imaginary of all is the unadorned E major triad found on the first half of the second beat in m. 9, triads being all but anathema to the reality of Op. 5. All of these factors – the inclusion of all four pitches in a whole-tone collection, the inversion, the respelling and the new interval between C and D, the presence of a triad – touch on aspects of the first phrase that we have been tracking but give them an even more mysterious dimension. The pizzicato notes don’t resemble any of the other pizzicato moments, which are instead simultaneous chords and have a tenuto marking instead of a staccato. Even the expressive markings are ephemeral: zart means gentle, but can also mean flimsy or frail and verklingend suggests a fading or dying away.[32]

As already shown, then, the flyaway motive that bridges the gap back to reality, asserting E and F# once again, after involving F and A in the inversional scheme about Bb.[33] The material from m. 5 is recapitulated in an even wispier form in mm. 11 and 12, the viola and first violin in a sense switching roles. The end of the piece turns on the last pizzicato chord. Usually considered some form of transposed or inverted [0, 1, 6, 7], its most salient property for the end of the story is the D atop the chord, the very D from the first phrase. The last flyaway is a direct outgrowth from this chord as well as the two first violin pitches marked verklingend: the G# from the bass, which, when initiating a flyaway, is followed by C and D, which here assert themselves unencumbered by E and F#, the complementary possibility of the first phrase, where C and D appeared but were immediately suppressed. This time given a pulpit from which to speak – instead of having to interrupt – in the form of the pizzicato chord, C and D rule, and the music speeds up (flüchtig, quintuplet sixteenths), concluding on Eb and F#, which, like the opening material to which it refers, has become memory. Rather than conclude or resolve this conflict, the ending simply asserts an alternate possibility and walks away from it (it even increases the pace), with only the faintest whiff of the beginning allowed to linger. Thus by combining a common response to this piece (E and F# as motivating subsequent events) with a new perception (C and D as intrusion and challenge), we have worked our way to an analysis in which intuition and reason have hopefully been drawn into a more mutually informative relationship.

As a final thought on Op. 5, No. 4, there is another distinctly Lewinian take on the choice of transposition of the flyaway motives that integrates well with the present analysis. The choice of {0, 4, 6, e, 1, 7, t} [0, 1, 2, 3, 6, 7, 9], or T₀ of the flyaway motive, appears motivated by the desire to catch the majority of pitches classes of the opening phrase – except D, as noted – and, significantly for the present analysis, to catch E and F# right next to each other. There are two pairs of whole-tones in the arrangement of this set class used by Webern in this piece, and since there is no way to get four consecutive whole-tone collection members out of the set class, clearly no transposition can contain both E/F# and C/D. There are only two ways each to get one or the other pair of pitch-classes. Since the two whole-tones dyads are ic5 apart, there is a transposition that turns B and C# (from T₀) into E and F#. That transposition is T₅, which is what Webern uses in m. 10. Drop T₀ {C, E, F#, B, C#, G, Bb} down by a major third and {E, F#} will become {C, D}, and this is what happens with T₈ of the figure in mm. 12-13. The last transposition (which is not used) is T₁, which results from sliding the initial transposition up a semi-tone, which will turn {B, C#} into {C, D}.[34]

There is something to this. Lewin arrives at his rationale for the final transposition by treating the set as an ordered set and balances the final transposition between the first and the second by treating the outer intervals as the transpositional generators. But there is another way of deriving the transpositional relationships. Notice the prominence of the rising fourth and falling major third in the material at mm. 5 and 11-12, {C, F, Db}, the passage that was so crucial for having introduced the conflicting C/D. While Db as a starting point for the flyaway motive does not occur, the prominence of these intervals do seem related to the choice of transpositions; the ic5 between C and F is captured in the first transposition and the ic4 that succeeds it (F to Db) is reflected in the starting points of C and G#, the first and last transpositions. However, it would have been possible for Webern to accomplish some of the same goals – as I see them – by using a transposition beginning on Db, or T₁ of the initial flyaway motive. Starting on Db would yield {Db, F, G, C, D, G#, B}, which, like the transposition beginning on G#, catches C and D, and even throws up G# and B at the tail end, which could be interpreted as growing out of the violin melody in the second phrase.[35] This is not to suggest that the Db option would have been just as good, but it does point to a family of four transpositions that includes T₁, which, together, form the basis of a transformational story based on an important melodic detail in an influential passage that both provides another rationale for the final transposition (Lewin’s original question) and uploads fairly easily into this analysis.

The foregoing analyses have hopefully shown how two of Webern’s flyaway endings complicate the notion of closure. That they do so in different ways is signal, for we should be prepared to discover any compositional strategy and any response it may engender. In the case of Op. 7, No. 4, Webern ups the ante by saturating the score with recollections from the previous movements and posits something akin to a stable structure – emphasis on pitches D and G – and walks away from it entirely. With Op. 5, No. 4, Webern stages a conflict between two salient whole-tone motions, at first asserting the primacy of one pair, then inviting us to imagine a scenario in which they interact, only to end by having the competition win the day. And these conclusions were arrived at by seizing upon a salient quirk and following through those observations for the remainder of the analysis. But if the idea of “imagining” sounds overly subjective, it only speaks to the role of listener response in constructing the analysis and the importance of concept formulation. The second phrase of Op. 5, No. 4 is not imaginary because I’ve hallucinated the notes; it’s imaginary because under the influence of a mindset that says, “E and F# comprise the premiere whole-tone, as borne out by their thorough extinguishing of C and D,” the coexistence of these entities further produces the response, “this can’t be.”

By way of conclusion, the methodological upshot is that we ought to treat tools such as set class and transformations with some flexibility in analyzing this literature. Burdened with too much interpretive responsibility, the familiar notion of set class might have caused us to stop, or at least not scratch much further, upon noticing that the flyaway motive that concludes Op. 7 contains sets already heard before. Likewise, tracking aggregate completion in Op. 5, No. 4 or invoking the idea of inversional balance are interpretively thin, and tend to answer the question “why does the piece end this way?” and much less “what is it like for the piece to end this way?” Attempting to answer both is not a bad approach for atonal or other musics.