The linear style, which is often associated with Prokofiev, was characteristic of the mainstream of Russian music. Two techniques which are frequently connected with, though not the exclusive prerogative of the linear style, are neighboring-tone progression and the ostinato principle. These techniques are of basic importance in Russian music; they are fundamental to Russian folk song and were extensively employed by Modernist composers.

The Linear Style

Linearity may exist on any one of three levels: (1) It may be an obvious surface feature of the score. Many of the examples from the works of Prokofiev quoted in chapter 4 illustrate this type. (2) It may exist in the background in the form of extended lines, which move across the texture but are not essential to an understanding of the surface features. An instance is the opening of Joseph Shillinger’s “Heroic Poem,” No. 1 of Five Pieces for piano Op. 12 (Example 5-1). (3) The linearity may not be immediately apparent but it is nevertheless essential to the logic of the passage (see Example 5-2).

Shillinger was a pupil of Nikolai Tcherepnin (who admired the work of Skriabin); he also came under the influence of Prokofiev, a fellow pupil at the St. Petersburg Conservatory. Shillinger’s early works reflect something of the style of both composers. The opening of “Heroic Poem” is like Skriabin in the firm movement of the bass (though not in the use of the 5ths) and in the complex and sometimes octatonic harmony. The music is also reminiscent of the linear style of Prokofiev in the movement of the upper texture, which occasionally seems to be out of phase with the bass. Underlying the surface features of the score are many long and often chromatic lines not immediately obvious to the listener. Shillinger’s strange notation may be intended to draw attention to them; double-sharps indicate an ascending line, flats a descending one. A comparison of the voice leading in Example 5-1 with Example 3-5a will make this clear.

The “out-of-phase” effect mentioned above, produced by independent movement of lines, is typical of the linear style. A related effect, which may be described as one of gradual harmonic fluctuation, is found in bars 5-7 of the second movement of Shebalin’s Sonata No. 2 (Example 5-11). From the last beat of bar 5 this is basically a I-V-I progression in B-flat minor. A move toward the dominant is made at the beginning of bar 6, but the B-flat in the treble, which sounds here like a pedal, and the D-flat and B-flat in the bass on the second beat ensure that there is still a marked feeling for the tonic, and the move toward the dominant is not clarified until the end of the bar. In similar vein the descending bass line does not complete its drive toward the dominant until the second eighth note in bar 7, by which time the inner parts are already moving back toward the tonic. Although the ensuing B-flat in the bass is emphatic, its chord is not confirmed before the harmony again begins to change on the third beat. The effect is of a lush undulation of sound, in and out of certain harmonic areas, in which the various parts are never completely in phase with one another.

The interplay of lines can also result in compression of the harmonic rhythm, as in bar 14 of the same movement. The harmonies have been changing at a leisurely rate, but in bar 14 several passing chords may be heard before the first beat of the following bar (Example 5-1la). This initiates a “crescendo sostenuto” buildup toward the climax in bar 22 (Example 5-1lb). The compression is reconciled with the slower-moving context by the similarity of harmonies at the beginning of bars 14 and 15, which are heard as a gradual clarification of the chord of C. Thus the rapidly changing harmonies are comprehended as movement within a wider span.

On a larger scale this kind of writing can lead to very free linear movement over long periods, supported by “pier” chords, which give strength and cohesion to the structure. An example is found in the first fourteen bars of Feinberg’s Sonata No. 5 Op. 10 (Example 5-3). The structurally most important chord here is the B triad, which is heard in the right hand at the beginning of bars 1, 2, 3, and 5 and at the end of bars 11, 12, and 13, supported by the bass B in bars 11 and 12. The spaces between these chords are filled in by the right hand with chromatic progressions and linear movement of augmented triads.

Such reference points need not necessarily involve all the strands or bring them into agreement at one time. In Anatoly Aleksandrov’s Sonata Op. 18 there is a passage, commencing at bar 18, where modal centers in the tenor and treble do not agree (Example 5-4). A melody in the left hand centered on A, over a D-sharp pedal, is associated with chromatic formations above leading to repeated occurrences of the C major triad (spelled enharmonically). The A and the C triad differentiate the modal leading of the two lines. The logic of the passage is clear if it is heard in linear terms.

Compatibility of parallel lines may take the place of specific reference points. This is illustrated by the second subject, beginning at bar 40, of Miaskovsky’s Sonata Op. 19 (Example 3-4a). In the bass is a series of chromatically descending tritones—D-sharp/A, D/G-sharp, C-sharp/G. Above this, at first forming part of the tritones but later becoming independent, is an internal pedal on A. Above that, another pedal, the tritone F/B, is occasionally ornamented by its neighboring tones, F-sharp and C (bars 42 and 43). In the upper texture the treble and alto voices move freely. The leading of these high voices brings them, at the beginning of every bar, to notes of the dominant-structured chord on G. The G and D at the beginning of bars 40 and 41, along with the tenor, form the same chord, creating a general agreement between the high voices and the internal pedal F/B.

The tune, which is in the alto in the first half of bars 40 and 41 and elsewhere in the treble, at first emphasizes a descent from D to C-sharp and then, in bars 42 and 43, a similar chromatic descent to A, returning to D/C-sharp in the following two bars. This leading, in the middle of each bar, to A or C-sharp also creates a compatibility between the high voices and the other internal pedal on A, and between both of these and the augmented 6th chord formed by the treble C-sharp/G and the bass D-sharp/A on the second beat of bars 40 and 41. The bass line ends in bar 45 with C-sharp and G, making an interval consistent with many of the events which have been occurring in the higher voices.

The Ostinato Principle

Example 3-4, from Miaskovsky’s Sonata, illustrates the ostinato principle in its use of internal pedal points. The ostinato is especially useful in the linear style because of its ability to bring coherence to otherwise loosely connected textures. It is also employed in Russian works of a nonlinear character, where the music is often modal rather than traditionally tonal.

It is impossible to impose absolute limits on a definition of the ostinato because sometimes a passage will combine the character of an ostinato with some other function. Generally, for a pattern to become an ostinato, it is necessary for it to be stated three times at its original pitch, though sometimes two statements are sufficient in the bass. Transposition or other variation will weaken the effect of ostinato but does not necessarily destroy it. In chapter 4 it was stated that an ostinato could be a means of strengthening the tonic, but it has many other uses. In the context of Russian music of the Modernist period an ostinato can obscure the tonic, create friction, and generate new harmonies; it can have a cadential function, provide local color, or be used as the basis of structure of an entire piece; it can also serve a pictorial purpose. The ostinato may consist of one or two pitches, in which case it is more frequently referred to as a pedal; it may be ornamented by neighboring tones or consist of a note cluster; it may be a melodic, rhythmic, or harmonic progression. It may be diatonic, chromatic, or whole-tone or be based on symmetry. It may be unchanging, or it may consist of the same tones arranged in varying patterns or of a single pattern involving different tones. Often it will form part of a layered texture the logic of which lies chiefly in its linear aspect.

The melodic ostinato is illustrated in Example 5-5, taken from the beginning of the second movement of Anatoly Aleksandrov’s Sonata No. 6. Example 5-6 shows an ostinato from the “March Grotesque,” No. 6 of Ornstein’s Dwarfs’ Suite Op. 11, in which the rhythmic element is dominant. The harmonic ostinato is represented in Example 5-7 by a passage from Anatoly Aleksandrov’s Sonata No. 3. Although the complete pattern is stated only twice, the static underlying harmony and the contrasted linear movement above lend a distinctly ostinato effect. An ostinato may be formed from a symmetrical pattern: a harmonically symmetrical ostinato is illustrated in Example 5-8, taken from Deshevov’s Meditations Op. 3 No. 7; and a melodically symmetrical one in Example 5-9, from Pieces Without Title No. 5 by Aleksandr Tcherepnin (1899-1977).

The “obstinacy,” or persistence, which is the essential characteristic of the ostinato principle, may be vested in a single intermittent note. The chief means of cohesion in the opening section of Ornstein’s “Funeral March of the Gnomes” Op. 11 No. 3 is the melodic B-flat which occurs seventeen times on the beat in the first thirteen bars (Example 5-10). Although part of the tune, this B-flat has sufficient independence to be heard as an intermittent ostinato or pedal. Another example of a discontinuous pedal—where it has a quite different effect—is the slow movement of Shebalin’s Sonata No. 2 (Example 5-lla). Here a recurring (mostly internal) F in the first eleven bars creates a friction against the prevailing harmony at the beginning of bar 4, on the third beat of bar 8, and, in the treble, at the commencement of bar 9 and on the third beat of 10. These clashes set the style and prepare the way for similar frictions in the treble line in bars 19, 23, 40, and 41 (Example 5–11b and c).

Whether an element functions as an ostinato depends in part on context. In bars 1-11 of Leo Omstein’s “Dance of the Dwarfs” Op. 11 No. 2, the left-hand part—despite variations of pitch—has sufficient regularity of pattern to be an ostinato (Example 5-12a). However, at bars 84-87 (Example 5-12b) the chaotic nature of the left-hand chords means that the music relies heavily for coherence on the right hand, including the tune, which itself now assumes something of an ostinato character. This duality of function is assisted by the repetitive nature of the rhythm and melody.

A further instance of the modifying effect that ostinato and context have on each other is in Prokofiev’s Visions fugitives No. 4 (Example 5-13). At bar 9 the repeated bass pattern reinterprets the tonal implications of the right-hand part, first heard in the opening four bars. The music tends alternately toward D-flat and F minor before settling on D-flat minor at bar 12. Another example is in No. 2 of the same set (Example 5-14), where the cadential effect in A-flat is brought about by the formation of the upper parts in bars 12-13, despite the E pedal of the preceding six bars. A particularly ambivalent example of the relationship between ostinato and context is in the Seconda part of “Napolitana,” No. 4 of Stravinsky’s Five Easy Pieces for piano duet. Here the ostinato either supports or denies the E modal center of the Prima according to whether the player emphasizes the right hand or the left hand on the beat (Example 5-15).

Omstein’s music has many examples of the ostinato note cluster. Example 5-16 summarizes those that occur in his Op. 11 No. 2, between the passages quoted in Examples 5-12a and b. Omstein’s Wild Men’s Dance (Example 5-17) makes extensive use of the whole-tone cluster F-sharp, G-sharp, A-sharp, sometimes accompanied by C (or occasionally B); despite the general cluster style of the whole piece, this particular group is clearly discernible in many places. In bar 2 it is isolated in the left hand; it occurs again (enharmonically) in the first subject, commencing at bar 13, where it forms an internal ostinato (joined here by the C), which lasts until bar 44. In the second subject, beginning at bar 69, it is heard most clearly when it forms the lowest group in the left hand, as in bars 73, 78, and 81-82. The rhapsodic middle section has the same cluster, high up on the top stave in bars 90-95 and in the accented chords in the right hand in bars 108-13, accompanied now by B. It returns at the end in the prestissimo coda, where it occurs on top of all the chords having dotted time values and the final two chords. It is one of the main unifying factors in this otherwise rather undisciplined work.

The left hand of bars 13-28 has another cluster ostinato, comprising six different one-bar patterns arranged against four-bar right-hand phrases, commencing with a rhythm derived from bar 1 (Example 5-18). The left-hand patterns are arranged in such a way that every time the right hand of bar 13 or bar 16 is repeated it is accompanied by a different group in the left. The six similar right-hand bars (bars 22 and 24-28) are accompanied by three different patterns, which are then repeated (bars 22, 24, 25: 26-28). Some internal pattern is also discernible within the left-hand part, as shown in the following diagram, where the numbers correspond to the successive statements of the two left-hand chords:

Neighboring tones play an important part in the ostinatos in Sonata No. 2 by Aleksandr Mosolov (1900-1973), shown in Example 5-19. The first movement begins with a pedal in which each of the two notes, A and D, is sounded together with its neighbor to produce a pianissimo background halo of sound. This serves to modify the otherwise clearly expressed D major in the middle texture. The ostinato is varied slightly when the E and D are separated, in bar 9, and this coincides with the point where the E-sharp is given more emphasis in the middle texture and the music begins to move toward F-sharp at bar 17 (Example 10-18). After a dominant preparation in bars 66-72 the exposition ends over a pedal D commencing at bar 73 (Example 5-20), ornamented by the neighboring tones E- flat and C-sharp (D-flat). At first the pedal is the chief means of emphasizing the tonic D, but after bar 80, when the tenor melody leads modally to D, the bass has alternating minor 3rds on C and C-sharp, an ostinato which is intended to cloud the pitch center until D returns in bar 88.

The ostinato occasionally has a cadential role, as in No. 3 of Ornstein’s Dwarfs’ Suite (Example 5-21). The descending 3rds in the alto of bars 23-25 and the major 3rd steps in the left hand stabilize the music at the end of the section and, as well as forming an ostinato, substitute for the traditional cadence. The ostinato at the end of Lourie’s Nocturne of 1928 (Example 5-22) also plays a cadential role. The low G occurring from bar 158 onward may at first appear to be a single-note pedal; it is however an unstable note and it is more meaningful to regard the whole of this end section as a freely constructed ostinato on the notes G, B-flat, D, F, C-flat, the effect of which is to emphasize the instability of the G. The expectation that G will eventually fall to F is so great that the F is almost as strongly felt as if it had been sounded in the low bass after bar 172. This creates an illusion of a V-I progression at the end.

Complex ostinatos consisting of more than one line are sometimes used. One of these is in the March from Stravinsky’s Three Easy Pieces for piano duet (Example 5-23). The Seconda part at the outset relates harmonically to the tune, while the left hand of the Prima has a neighboring-tone relationship to the other two lines.¹ Another example is at the beginning of Shillinger’s “Eccentric Dance” Op. 12 No. 4 (Example 6-37). Here the treatment makes use of a number of modal techniques. The right-hand melody centers modally on D, which is treated as part of a G chord with G emphasized as an internal drone. Against this the left hand has an ostinato pattern built around the neighboring chord of F-sharp, which is the tonality of the piece. This in turn is ornamented by its neighboring tones C and A, creating a common ground that facilitates the move to F in bar 5.

There are two ways in which the ostinato is employed to control the overall plan of a work. One is for each main section to have its own ostinato which recurs when that part is repeated. Anatoly Aleksandrov’s Sonata No. 6 is largely planned in that manner throughout its three movements. The other method, more suitable for smaller pieces, is for the ostinato to bring together the main tonal elements of the work. Feinberg’s Prelude Op. 15 No. 3 illustrates this type. In the music reduction (Example 5-24) the bar numbers indicate the more stable passages. Bars 1-3 contain three tonal areas: an F pedal, a tune which centers on B-flat, and a background of E-flat harmony. The second, contrasting idea has notes of the C major triad in the bass with a shifting chromatic texture above. The bass of the recurring ostinato sections, shown in the last three “bars,” progressively draws together the main pitch elements: B-flat, F; E-flat, B-flat; and finally all three. In retrospect the C major triad may be heard as having a dominant relationship to the final F.

Lastly, ostinato can be used for pictorial purposes. In the descriptive piece Rails Op. 16, by Vladimir Deshevov (Example 5-25), the composer often obscures the harmony by additions which render the chords ambiguous, as in bars 19, 25, and 29. Thus the harmony becomes a vehicle for carrying rhythmic patterns, instead of being a functioning entity in its own right. The result is that much of the piece seems to consist of a series of short rhythmic ostinatos, which accurately match the composer’s descriptive intentions. In this connection it may be noted that one of Ornstein’s most successful compositions, Chinatown, conveys, with kaleidoscopic effect, aspects of life in a busy city. The difference is that Deshevov considerably modified his usual technique to achieve the effect in Rails, whereas Ornstein for once found a title which gives point to his frequently ostinato style of writing at that time.

A quite different descriptive use of the ostinato is found in Nikolai Obukhov’s Berceuse to words by Balmont (Example 5-26). The often distraught expression of the music is contrasted with a gentle rocking motion of a minor 3rd in the bass, usually between E-sharp and D. Toward the end this motion is emphasized: in the last four bars the E-sharp-D is heard first at the bottom of the texture and then in the alto, and the rocking movement of a minor 3rd dominates the whole harmony at the cadence.

Neighboring-Tone Technique

Several of the ostinatos examined in the last section have either neighboring-tone ornamentation or a neighboring-tone relationship to other strands in the music. Neighboring tones may be employed for melodic decoration, in which case they are often left unresolved. They are also used harmonically to give weight or pungency to a chord; here they must be used with caution, for the texture has to be clear and transparent so as not to obscure the harmony. Despite the cluster-like effect of the left-hand sixteenth notes in Example 5-27, the low bass leaves no doubt as to the identity of the chord. Structural use of neighboring tones occurs when they are employed to give emphasis to an important pitch.

Sonatina No. 3 (1917) by Arthur Lourié (Example 5-28) illustrates the decorative use of neighboring tones against an ostinato bass; some of them are left unresolved for several beats. The central part of the Etude, No. 6 of Aphorisms Op. 13 by Shostakovich, is a further example of this decorative play. It is unusual in that the decoration occurs in the left hand, which adds neighboring tones a tone or a semitone away from the 5ths in the right hand—except in bars 16 and 21, where some of the left-hand notes complete the chord (Example 5-29).

The use of the chromatic focus to emphasize a structurally important pitch was a favorite device of Lourié’s. Example 5-30 illustrates the use of this technique in Nocturne written in 1928, by which time the composer had considerably simplified his style. The opening section of this piece is based on the dual pitch center of B-flat and G. The connection is suggested right at the beginning by the grace notes in the bass, and neighboring tones introduce both pitches in bars 3-5, 6-7, and 12. After bar 15 the music begins to move toward the F-sharp of the second section. G has been prominent in the bass since bar 12, and, from bar 19, G and E are taken together as neighboring tones of F-sharp. This is a little unusual for Lourié in that the E is not a chromatic neighbor.

The technique of building a piece on a decorative interplay of neighboring tones is illustrated in Shillinger’s Dance from his Op. 12 (Example 5-31). The piece begins with the triad of F-sharp decorated by its neighbor G and an additional neighboring tone A entering in the second bar. In bar 7, the right- and left-hand groups tend to separate into distinct harmonies, whereas in the following bar—by a change of layout—they coalesce into a single chord in which only the initial F-sharp can be heard as nonharmonic. By replacing the original major triad of bar 1 with a minor triad in bar 9 a new relationship is set up between the two hands so that the highest note of bar 10 is now a harmony note. In bars 13-14 the G-flat chord of the right hand is at first decorated by neighboring tones in the left, but a rearrangement and the addition of a low F in bar 15 make the G-flat chord a neighboring-tone decoration of the triad on F. Following that, the two hands combine once again to form a single harmony on E-flat in bar 16.

A particularly striking use of neighboring-tone technique is to be found in Lourié’s Syntheses No. 2 (Examples 5-32a and b). The vertical pitch collections in this piece have a marked similarity to traditional chords, and the interval immediately above the bass is most often a 3rd, a 5th, or a 6th. Intervals discordant with the bass move to more concordant relationships—although the harmony may at the same time change, making the note of resolution a new discord. Much of the radical sound is due to the placement of neighboring tones against their corresponding harmony notes, which produces some quite strong voice leading even when the neighboring tone does not actually proceed to its resolution.

Example 5-32c is an attempt to trace some of the real or potential points of resolution in this piece and their effect on voice leading and pitch center. From bar 3 to the end of bar 6, where the F-sharp center is reached, the discords form (or suggest) rising lines, though notes of resolution are not always sounded. In this example the notes of resolution are written in brackets, and implied resolutions that form especially significant lines are marked by dotted beams. The one in the second bar of the example is an imitation of one that appears in the first bar (bars 3 and 5 of the score). After the F-sharp has been reached in bar 6 the resolutions have a generally falling movement and therefore a more-relaxing effect. This is also true of bars 16 and 17, at the beginning of the last section. In the coda the movement of the melodic line toward C, whether by rising or falling resolution, is especially marked, but the generally downward movement, particularly in the bass, relaxes the tensions of the piece. (For further discussion of Syntheses, see chapter 11.)

Lourié makes some subtle play on constantly varying relationships between neighboring tones in his experimental piece Forms in the Air (Example 11-19). Each of the three movements is composed of strongly characterized cells. Cell 5 of the first movement ends with an A chord, enharmonically written with a D-flat. This note is then supported by G-flat in cell 6, so that the C sounds like a neighboring tone of D-flat in a quasi-trill effect. This is reversed in cell 7. As the D-flat is no longer supported by G-flat, and C becomes the lowest note, the D-flat and B both begin to sound like neighbors of C. The effect is confirmed when the D-flat finally ceases, leaving E-flat and C together. The triad of E-flat minor is now presented, in cell 8, with neighboring tones E and D left unresolved at the end. In cell 9 the play continues with E joined by G-flat as neighbors of F, against a background of D-flat harmony. [Forms in the Air is considered further in chapter 11.)

Deshevov’s Op. 3 No. 3, bars 1-10 (Example 7-6), like Rails, makes considerable use of neighboring-tone relationships. G-flat is at first emphasized both by its insistence in the bass and by the modal leading of the treble line as far as bar 10. Harmonically the opening bars are based on the tritone G-flat/C and the neighboring tones D-flat and F-flat, which also suggest a possible resolution of the tritone. The trill is unusual in that its two notes—as in many other places in this piece—relate to the two underlying harmonic groups. There is strictly no auxiliary in the trill; rather it is an alternation of two harmonic notes.

In Deshevov’s Rails (Example 5-25), bars 19 and 23, neighboring tones create an ambiguous common ground between two harmonies. Prokofiev sometimes exploits this characteristic to create common ground between two tonalities. An example is his Op. 22 No. 6 (Example 5-33). Throughout this piece there is a rivalry between C major and A minor. The treble sets off as if heading for the note C; the bass tries to subvert this attempt and convert it to A minor, but it is frustrated by the unexpected B-flat and C-sharp in bars 2 and 3. In bar 7 the music seems to arrive at C major followed immediately by a close in A—which is rather indecisive because of the lack of a leading tone. The conflict continues in the following bars, facilitated by the persistent A-flat and E-flat, which are neighboring tones of the chords of both C and A. The device is also turned to comic effect because the addition of C—the only tone common to the chords of A minor, C major, and A-flat—leads to the triad of A-flat, creating a kind of “dead end,” and the whole process has to start over again.

The use of neighboring tones to generate harmony is demonstrated in the last movement of Shillinger’s Five Pieces Op. 12 (Example 5-34). The notation of the right-hand part at bars 5-9 emphasizes its neighboring-tone relationship to the G chord underneath. However, the right-hand part is carefully separated from the left in register and texture, thus drawing attention to its independent existence as an E-flat triad. The first piece of the set ended in E-flat (Example 7-19), and there was some emphasis on G at the very beginning (Example 3-5a), so the splitting of the texture into G and E-flat harmonies here has a structural significance. It occurs again at the recapitulation in bar 71 and in bars 105-107, at the final cadence. The whole piece is an example of the contrast between neighboring tones used as decoration of a single chord and their use to generate independent harmonies. Where there is no differentiation of timbre or register the neighboring tones are heard as related to a single root, which is thus emphasized. Where the texture is thinned out and the two parts are characterized by texture and rhythm, they are heard as separate harmonies.

Attention has now been drawn to two of the main strands in the tapestry of Russian music: Skriabin and his followers and Prokofiev and the linear style. Chapter 6 deals with techniques of more general application and after that consideration is given to another main strand, the influence of folk music.