WTC I/23 in B major -- Prelude

    from Siglind Bruhn
J.S. Bach's Well-Tempered Clavier
In-depth Analysis and Interpretation

I/23.1.1 The prelude-type

The material of this prelude derives almost entirely from a short ornamental figure introduced in the first half bar; in fact there is only one instance in the piece where this motive is momentarily absent! It can thus safely be claimed that this prelude belongs to the motivically determined compositions. In its texture, the prelude is polyphonic, with three parts which are amplified only in the final bars (after a splitting of the upper voice on the middle beat of bar 16) to four parts.


I/23.1.2 The overall design of the prelude

The very first harmonic closure occurs - one could say implicitly - on the middle beat of bar 2. Here the F#7 chord, represented in the upper and middle voices above a sustained tonic pedal, resolves onto the tonic. The tonic is embodied by its third (D#) and the "silent" root B which the listener hears in the middle-voice rest as a melodic continuation of the rising line before. This cadential close, however, is structurally not decisive. Reasons are, above all, that the harmonic progression coincides with the first melodic phrase (made up by the motive, two sequences and a final note), that neither of the other two voices has yet taken part in the polyphonic display of material, and that the bass has not even begun to participate in any harmonic development.

The first structural caesura appears at the subsequent perfect cadential close. As the frequent E#s from bar 3 onwards indicate, the prelude modulates to its dominant F# major, in which key it closes on the downbeat of bar 6. This cadential close is firmly established in bars 5/6 with a cadential-bass pattern (ii-V-I) and a melodic closing-formula do-si-do (keynote / leading-note / keynote). Similar cadential features mark the next structural break which occurs, after a modulation to the relative key (G# minor) on the middle beat of bar 10, as well as the return to the tonic in bar 15. (The bass patterns in bar 4: C#-F#-B and in bar 13: F#-B,-E which both seem to support V/V-V-I progressions, are not convincing as structural closes since in both cases the upper voice remains unresolved.)

The prelude contains altogether three sections and a short coda.


bars 1 - 6d

tonic to dominant
II bars 6 - 10m dominant to relative minor
III bars 10 - 15d relative minor back to tonic
coda bars 15 - 19 confirmation of tonic

There is only one short structural analogy:

bars 4 - 6d

correspond with

bars 13 - 15d

including the

  including the
melodically hampered cadential   melodically hampered cadential
close in B major (tonic)   close in E major (subdominant)
followed by the   followed by the
structurally relevant cadential   structurally relevant cadential
close in F# major (dominant)   close in B major (tonic)


I/23.1.3 Practical considerations for performers

The basic character of this prelude must be interpreted as rather lively. This is supported both by the ornamental character of the principal motive and by the very even overall impression of the rhythmic pattern. The tempo allows for a brisk pace in the quarter-notes; the sixteenth-notes should sound ornamental - i.e. without emphasis on each single note - but by no means hasty.

The general pattern of articulation includes legato for the sixteenth-notes and non legato for all other note values. Going into more detail, a gentler, dynamically shaped detached style in the melodic quarter-notes (see e.g. bars 1, 6) and equally melodic eighth-notes (see e.g. bars 5, bars 10m-12) should be distinguished from a more neutral non legato touch in the cadential-bass patterns in quarter-notes (see bars 3/4, 5/6, 13, 14/15, 18/19) and eighth-notes (see bar 10). Conversely, one should single out those among the larger note values which demand absolute legato; this is the case in the melodic closing-formulas (see bars 5/6 upper voice: F#-E#-F#, bars 9/10 middle voice: G#-Fx-G#, bars 14/15 and 18/19 upper voice: B-A#-B.

The phrasing before or after the principal motive, however, deserves detailed discussion. The question whether or not phrasing should be expressed by a slight cut in the sound flow arises mainly in two cases:

(a) wherever the motive's after-beat beginning is preceded by an on-beat note belonging to a previous melodic line (as e.g. in bar 4 beat 3, lower voice, and in bar 5 beat 3, middle voice);
(b) wherever the principal motive, originally only seven sixteenth-notes long and omitting the strong beats, is extended to include a final on-beat note, and is then sequenced (as e.g. in bar 6 beat 3, lower voice, and in bar 7 beat 3, upper voice).


Any all-too-obvious jump between notes is definitely inappropriate in this piece, but, for performers with good skills in subtle articulation, slight interruptions certainly constitute the more plausible solution than simple legato continuation. (The interruption may be more noticeable in all cases (a), but hardly perceptible in all cases of (b)).


I/23.1.4 What is happening in this prelude?

A description of the material Bach uses here must begin with the principal motive. It consists of seven sixteenth-notes circling around a center note (in the first half of bar 1, e.g., this center note is B). The circling motion resembles an inverted turn (B-A#-B-C#) and is followed by a repetition of the three ascending notes (A#-B-C#). The figure thus ends one note above what was recognized as its center. Harmonically, this ascent is already established in the middle of the motive (on beat 2); the step to the higher pitch, accompanied by a step upwards in the middle voice, represents a progression also on the harmonic level. The motive is thus not just "self-centered" and circling without aim, but can be said to comprise an active step. In terms of melodic tension, this active step is expressed in a delicate crescendo through the initial notes up to the middle note - the note which is also metrically the strongest in this group. The three notes which complete the motive provide the complementing diminuendo, so that the motive ends as softly as it had started.

For two reasons is it essential to delineate the tension-curve within this motive very clearly. Firstly, due to the short duration of the motive in relatively fast tempo, the figure may easily come out as mere finger work, i.e. with no melodic expressivity whatsoever. Secondly, the fact that Bach later extends the motive by adding another sixteenth-note which then falls on the strong beat, easily traps performers into an awkward accent on this extra note, thus distorting the original shape of the figure.

Apart from the extension just mentioned, the principal motive suffers two other modifications. Four times, a slight change in one of the intervals (see bars 22 and 122; bars 51 and 141) eliminates the ascending step inherent in the original shape. In addition, inversions convey a resolving tendency and are almost exclusively found in the coda (see six times in bars 15-18 and once in bar 12).

Besides this principal motive, there are two characteristic note-groups which recur.

One is the ascent in quarter-notes which is introduced in the middle voice of bar 1. It describes a gentle dynamic rise which strings the consecutive figures of the principal motive together in one overall direction. This quarter-note motive recurs in the first bar of section II (see bar 6: middle voice) as well as in double notes in the coda (see bar 17, split upper voice).
The other secondary motive graces only section III of the prelude. It moves primarily in eighth-notes (the three motives thus each favor a different note value) and is conceived as a question-and-answer pattern between the outer voices (see bars 10/11, U: D#-G#-F#-E# / L: C#-B-A-G#-F#; sequenced in the subsequent bar). One very convincing way of molding this motive-pair is to choose complementing dynamics for the complementing segments: crescendo for the upper-voice "question" and diminuendo for the lower-voice "answer".


The development of tension in each of the sections is fairly straight-forward as it follows largely the overall pitch outline. In section I, an initial increase of tension (bars 1/2) is followed by a long decrease (bars 2m-6). In section II, proportions are reversed with a much more extended increase (bars 6-9m) and a short relaxation (bars 9m-10m). Section III begins in minor mode and with a new motive - both reasons to create a little contrast in color (bars 10m-12m), followed then by a long decrease which corresponds with that of the first section (bars 12m-15). In the coda, the tendency of withdrawal is thwarted by three features: the replacement of the principal motive by a figure with an ascending-scale component (bar 15 second half), the increase in texture after a splitting of the upper voice (bar 16m onwards) and the recurrence of the ascending-quarter-note motive, now in double thirds (bar 17). The prelude thus ends in a assertive mood.


WTC I/23 in B major - Fugue

I/23.2.1 The subject

The subject of the B major fugue is two bars long. It commences with an up-beat gesture after an initial eighth-note rest and concludes after an ornamented C# (representative of the dominant harmony) with a return to the keynote on the downbeat of bar 3.

The pitch pattern exhibits almost exclusively seconds, interrupted only for the interval C#-F#. This leap of a perfect fifth is not of expressive quality, and thus does not immediately match the stepwise motion around it. Upon closer inspection of the pitch pattern in the subject it becomes clear that what occurs here is a change in pitch level rather than an interval between two notes. The initial ascent, launched from the keynote, breaks off with a quarter-note on the first strong beat, only to start afresh from the lower F# and climb even higher up. The subject thus consists of two subphrases which relate to one another in such a way that the first appears as an abandoned attempt of what the second then completes in a more powerful format.

A look at the rhythmic design confirms this assumption. Considering that the longest note value, the half-note C#, is ornamented by a trill and thus sounds very animated, the quarter-note C# at the beginning of the subject is in fact a powerful rhythmic interruption and marks the point where the regular motion comes to a halt, to start newly from F#.

The harmonic background of the subject is difficult to determine as Bach harmonizes it differently in almost each statement. It seems, however, safe to claim that there is a two-fold progression, with an interrupted cadence either on the downbeat of the second subject bar (as e.g. in bar 6) or slightly later (as e.g. in bar 30), and the perfect cadence taking place in the two final notes (ex. 49):

In a subject with two subphrases one can obviously expect two climaxes. As it was already established that the second subphrase completes the aborted efforts of the first, the weighting between the components is obvious. The first climax is easy to determine since the melodic rise (B to C#), the harmonic movement (tonic to dominant), the rhythmic value (quarter-note) and the metric position (middle beat) all support the final note of the first subphrase. Determining the climax of the second subphrase is, however, not so simple. Here the pitch motion reaches its peak on E, a note which is neither harmonically nor rhythmically or metrically supported. By contrast, the ornamented C# represents two important steps of the cadence, apart from being in a rhythmically and metrically stronger position than the off-beat eighth-note E. As a result, a climax on E will give the subject (and, with it, the entire fugue) a more virtuoso touch by stressing superficial features (pitch), while a climax on the trilled C# gives the subject more depth by emphasizing its structural traits.


I/23.2.2 The statements of the subject

The fugue contains altogether twelve statements of the subject.

1. bars 1 - 3 T 7. bars 18 - 20

S (inv)

2. bars 3 - 5 A 8. bars 20 - 22

A (inv)

3. bars 5 - 7 S


bars 21 - 23 B
4. bars 7 - 9 B 10. bars 24 - 26 T
5. bars 11 - 13 T 11. bars 29 - 31 A
6. bars 16 - 18 A 12. bars 31 - 33 S

(ex. 50)

There are two kinds of modifications occurring in the subject statements throughout the fugue, one at the beginning and the other at the end of the subject.

The initial intervals of both subphrases are adjusted in all tonal answers (see bars 3, 7, 31), while in the inverted answer, only the beginning step of the first subphrase is enlarged (see bar 20).

The final resolution appears delayed (bars 7, 31), diverted (bars 18, 20) or omitted (bar 22). In one instance, the subject ending is varied without causing a change in the essential steps (bar 26).

Parallel statements do not occur; nor do true strettos in which a crucial segment of one subject entry is overlapped by the beginning of the subsequent statement. The only instance where an entry commences at less than two bars' distance from the beginning of the previous one, occurs in bar 21, i.e. in connection with the one entry in the fugue which omits the resolution. (Thus the process one is hearing - as opposed to seeing in the score - can be said to be concluded by the time the subsequent statement enters).


I/23.2.3 The counter-subject

Bach invents one counter-subject for this fugue. It is introduced against the answer of the subject in bars 3-5. Beginning two eighth-notes later than the subject itself, the counter-subject also displays two subphrases separated by a change in pitch level (see bar 3 beat 4). The phrasing falls one eighth-note after that in the subject, and, interesting enough, in its first appearance the counter-subject even ends with a metrically delayed resolution, i.e. one eighth-note after the subject (see bar 5 beat 1).

Besides this similarity in phrase structure, the counter-subject is also related to the subject in pitch pattern as it displays an overwhelming majority of seconds and a scalar ascent in the second subphrase. Independence in dynamic design is also limited. Whether a performer follows the first subphrase in its pitch outline with a diminuendo (and thus links the tension decline over the phrase cut between the subject's two subphrases), or whether the performer expresses an active gesture in a downwards crescendo (and thus imitates that in the subject's first subphrase), not much individuality can be gained. In the second subphrase, both pitch and rhythm favor the syncopated highest note - with the result that this climax very nearly coincides with that of the subject. There is, to conclude, fairly little challenge for the subject.

Furthermore, the counter-subject is not a very faithful companion. In its complete range it recurs only three times (see A: bars 5-7, S: bars7-9, A: bars 31-33). Additionally, the second subphrase appears once without the first (see A: bars 12/13). Finally, there is an even more truncated second half of the second subphrase (see S: bar 17).

Here are two possible ways in which the counter-subject may be heard against the subject - depending on the performer's preference for a more virtuoso (pitch-oriented) or more structurally supported interpretation of both components:

(ex. 51a) .

(ex. 51b)



I/23.2.4 The episodes

The fugue contains five subject-free passages.


bars 9 - 11m

E3 bars 23m - 24d
E2 bars 13m - 16d E4 bars 26 - 29d
  E5 bars 33 - 34

The first episode introduces the listener to several motives which, as they are independent from both subject and counter-subject, must be regarded as genuine episode motives. When attempting to distinguish these motives, closer inspection reveals that they all derive from a single common source. The different versions share the shape of a concave curve in which the longest note (which provides both the harmonic and the dynamic climax) falls on the lowest pitch. The example below gives the three versions of this one motive; see bars 9-11 and 26-28 (ex. 52):

In the three more substantial episodes (excluding the merely half-bar long E3 and the final cadential close), M1 can be found at every opening. In each case, M1 is accompanied by M1a in such a way that these two run in parallel thirds or sixths throughout their "tail". Also in each of the three cases, M1 is sequenced by the variation M1b, which is then followed by an extra "tail". Before this sequence and in stretto to the original M1, there appears a variation in which the quarter-note is replaced by two eighth-notes in octave displacement. (In E2, this octave jump even involves a swap of voices; see bar 14: from tenor to bass.) All three episodes then end with separated "tails" and "heads" of the motive.

The following table visualizes this play with the episode motive.



  M1b   tail


    tail tail
      M1bvar head head  
E1 bars 9/10: S+A B 10/11: S A+B S+A+B
E2 bars 13/14: A+S T/B 14/15: A S+T A+S+T
E4 bars 26/27: S+T B 27/28: S T+B S+T+B

Both the short E3 and the final episode E5 only feature the "head" of M1b (see B: bars 23/24, and T: bars 33/34). In the former, the two upper voices continue with sequence and imitation drawn from the fragmentary counter-subject figure in the soprano, so that this episode appears more like an extension to the preceding subject statement - an extension which serves, one might say, to re-establish the original metric position of the subject after the premature entry of the tenor in bar 21. In the final episode, the three surrounding voices state closing patterns; this episode appears not as an extension but as a traditional cadential bar.

As regards the relationship between the episodes in this fugue, the above table demonstrates the analogy between E1, E2 and E4. The remaining two episodes, while both featuring the same segment from the motive variation, fulfill structurally different purposes and should therefore not be interpreted as related. While the motivically determined episodes constitute a noticeable color contrast in the fugue (each of them with a slight diminuendo in the descending sequence of M1b and a slight crescendo in the ascending sequences of the "head"), the two shorter episodes are integrated into the level of the primary material - one bridging between two subject statements, the other providing the final relaxation, but both without too much of a dynamic development.


I/23.2.5 Character, tempo, articulation, ornament realization

The predominance of stepwise motion, coupled with the variety of note values, determines this fugue as a piece in rather calm basic character. The pace is serene, neither hasty nor hesitant. The relative tempo of the prelude to the fugue is complex. This is mainly owing to the fact that both pieces are in common time and based on the same note values, so that a simple proportion would create an effect of dullness in the succession of the two pieces. This is the translation of one tempo into the other:

three quarter-notes

correspond with
two quarter-notes
in the prelude

in the fugue

(Approximate metronome settings: prelude beats = 108, fugue beats = 72.)

The overall articulation is legato. (Exceptions occur in the cadential-bass patterns of bars 13, 17/18 and 33/34, as well as in other consecutive quarter-note or eighth-note leaps, e.g. bar 24: B, 31/32: T.)

The trill in the subject poses some problems. It is, without any doubt, conceived as an integral part of the thematic phrase - so much so that playing the half-note without an ornament would sound extremely dry. (This is certainly the impression we get when we play the voice alone. It is also what any string or wind player, performing this fugue in a quartet, would feel. Only pianists, busy with figures in other parts but still in their own ten fingers, sometimes choose to ignore the single-voice demands, with the excuse that "there is already enough happening in this bar".) How, then, should the trill in its original setting be executed? As it is approached in stepwise motion, it commences on the main note. As the fastest note values in this piece are sixteenth-notes, it shakes in thirty-second-notes. And as its resolution appears - at least in the two initial statements - in proper metric position, it ends with a suffix. This long trill thus contains altogether fifteen notes (see below in ex. 53), to be played very regularly and, what is even more often neglected, with a dynamic shading that reflects the decrease of tension in the subject at this moment.

Whatever problems arise with the trill in this fugue are caused by several irregular endings - those which were earlier mentioned as "statements with delayed resolution". In all these cases, the trill begins in the same manner but ends prematurely and without a suffix, stopping short on the last main note before the bar line. (The only statements which forgo the trill are those with omitted resolutions or varied endings: bars 17, 21 and 25.)

To provide those performers who now decide against playing this fugue - just because of the trills - with an incentive for trying, the following examples give the most prominent occurrences spelled out, with some suggestions for appropriate fingering in the trickier cases (ex. 53):



I/23.2.6 The design of the fugue

When trying to determine how this fugue is structured, one cannot rely on any of the data which normally guide such an analysis. Except for the final bar, there are no explicit cadential formulas outside the confines of the subject statements. (The only obvious cadential-bass patterns mark the endings of the fifth and sixth statements respectively; they thus follow one another too closely to indicate section endings.)

The texture is unusual insofar as there are only two subject statements (among the twelve of the fugue) which appear in full ensemble. All other entries either include one resting voice or maintain the full four-part texture only for a short span (see bar 7 first half, bar 21m-22m, bar 31). The harmonic design is quite atypical insofar as no note-worthy modulation takes place throughout the entire composition; except for one entry in the subdominant, the statements alternate regularly between tonic and dominant positions of the home key.

However, the analogies observed earlier in the design of the episodes continue as one inspects the structural layout of the entire composition. The pattern of subject statements and episodes is as follows:

4 consecutive statements (bars 1-9) 4 consecutive statements (bars 18-26)
  (if one regards E3 as an extension)
E1 (bars 9-11) E4 (bars 26-29)
2 additional statements 2 additional statements
interrupted by E2 rounded off by E5

As both the fourth entry in bar 9 and its correspondent in bar 26 close with a perfect cadence while the ensuing episodes feature open endings leading into the next statements, one has to assume that these episodes open sections - rather than close them, as in so many other fugues. The result of these investigations leaves us with four sections, of which

the first and third each consist of four consecutive subject statements and end in a perfect cadence (bar 9: F# major, bar 26: C# major);
the second and fourth each encompass an opening episode of equal material and design, two subject statements and one additional episode (incorporated into the second section but closing the fourth section).


For a sketch showing the design of the fugue in B major, see ex. 54.



I/23.2.7 The development of tension

Concerning the dynamic design valid in the sections of this fugue, two basically different patterns can be distinguished.

On the one hand, there are the first and third sections which are made up exclusively of subject statements. Within the first section, the tension rises gradually along with the usual increase in the number of voices. Due to the fact (already mentioned above) that this section does not truly establish the expected four-part texture, the growth should be restrained in such a way as to avoid the sensation of a powerful climax. In the third section, the two initial statements appear in inversion. Given the particular shape of this subject with its two rising motions which are now converted to falling ones, the inversions sound much less cogent and self-confident than the original. As a result, the four entries of this section also give the impression of curbed tension.

On the other hand, there are the second and fourth sections which encompass alternations of episodes and subject statements. The prominent musical message expressed in the structure of these sections is obviously that of color contrast. A very light and delicate touch in the episodes, conveying both playful character and melodic openness, is set against a much more assertive touch in the subject statements which, by comparison, are clearly directed towards their goal and assuredly closing. If one wished to weigh the two entries in each section against each other, one would find that in the fourth section, the final subject statement surpasses the preceding one, both because of the compelling rise in these consecutive entries from the alto to the soprano and because of the temporary four-part texture. In the second section, however, where the statements are separated by a substantial episode, the question of any dynamic relation between the statements seems beside the point.