WTC II/24 in B minor - Prelude

II/24.1.1 The prelude-type

This prelude, the only one within the Well-Tempered Clavier to carry a tempo indication, is marked Allegro and metrically conceived in alla breve time. This indication of swift tempo in large beats is matched by strikingly regular phrase structure; all patterns take up exactly four bars, whether completely filled by thematic material or made up of a motive of three and a half bars complemented by a half-bar link. The only interruption of the dance-like regularity occurs in bars 57/58. These two bars are not only "extra" (see the four-bar phrases in bars 49-52, 52-56 and 59-62) but further halted by a pause.

The consistent two-part texture of this piece features voice-splitting only in the final bars (see bars 62-66). The degree of independence between the two voices varies; there are phrases with obvious accompaniment patterns (see e.g. bars 49-52) as well as genuine contrapuntal settings (see e.g. bars 33-36).

Several motives are presented in the course of this prelude. The main motive, the only one to recur in immediate imitation, is evenly distributed between both voices. Other motives, by contrast, display a strong preference for the upper voice and do not make use of imitation, thus introducing a more homophonic style. The accompanying figure of each motive can take various forms but contrapuntal interplay among the motives themselves does not occur. The prelude can thus be described as determined by motives in two-part setting.


II/24.1.2 The overall design of the prelude

Small-scale harmonic closures occur at the end of most four-bar phrases (sometimes even within) and therefore cannot be regarded as relevant in the overall structural layout. Thus the tonic is reached in V-i cadence on the downbeats of bars 3, 5, 7 and 9; only then is the first modulation launched.

In the absence of strongly determining harmonic processes, structural analogies help to present an overview of the layout. There are several corresponding phrases:

bars 1-4

recur in

bars 17-20 (transposed)

and in

bars 41-44

(transposed and varied,

      with two bars in inverted voices)
bars 5-8
recur in bars 29-32 (transposed, voices inverted)
bars 21-24 recur in bars 59-62 (transposed, ending varied)
bars 9-13d recur in bars 45-49d (transposed)
bars 13-17d recur in bars 49-53d (transposed, lower part varied)

On the basis of these analogies, the prelude can be defined as made up of four sections. The table below states the harmonic progression as established by the occurrences of the main motive (see U: bars 1-4m etc.):

I bars 1-16 main motive twice on the tonic (B minor)
II bars 17-40
main motive twice on the relative key (D major)
and twice on the subdominant (E minor)
III bars 41-58 main motive once on the dominant (F# minor)
IV bars 59-66 main motive once on the tonic (B minor)


II/24.1.3 Practical considerations for performers

Given that the composer has prescribed the character he desired for this piece very clearly, the first questions performers usually face when dealing with Bach's music are already answered: the tempo is swift and the basic character is rather lively. What remains to be pondered is the question of articulation of the different note values.

The three note values determining this piece are quarter-notes, eighth-notes and sixteenth-notes. Close inspection reveals that almost all sixteenth-notes are written-out ornamental figures. There are turns in different metric positions (see e.g. bars 1-3; bars 10, 12; bars 21-23), inverted turns (see e.g. bars 9, 11), and off-beat trills with spelled-out diverting suffixes (see e.g. bars 7, 13-16). When settling for an appropriate articulation, one therefore has to consider mainly the non-ornamental values, i.e. eighth-notes and quarter-notes. The eighth-notes, being the faster values among the two, must be played in a light, transparent legato touch, while the quarter-notes should be rendered in non legato. There are, however, a number of exceptions. They occur as follows:

Within the eighth-notes, the light legato should be interrupted where escape notes separate slurred note-pairs (see e.g. in U: bar 26; the F# does not form part of the main melodic line). On the other hand, quasi legato must give way to dense legato where this is explicitly indicated (see U: bar 58).
Particularly easy to overlook are interruptions of the eighth-note flow due to phrasing (see e.g. U: bars 13-16 after the downbeats, bars 23, 61 before the last eighth-note where the last of three almost identical sub-phrases must be separated from the typical upbeat + appoggiatura-resolution group).
In the quarter-notes, the usual non legato touch must be suspended for those note-pairs which constitute appoggiatura-resolution. (These may or may not be indicated as slurred: compare U: bar 8 and L: bar 32 with U: bar 24; see also the syncopations in U: bar 37 with sequences and imitations in bars 38/39 as well as similar figures in bars 53/54, and bar 64: E#-F#, D-C#, B-A#.)
Other quarter-notes which must be played legato are those in the traditional closing-formula (see U: bars 65/66: keynote / leading-note / keynote).

The quarter-notes which are topped with wedges (see U: bars 21/22 and 59/60) pose another question. As these note values are detached anyway, one may wonder what is meant by this additional information. One good way of interpreting the symbols is that Bach wishes the performer to play the high notes not heavily weighted, as climaxes in themselves, but with a certain springing power towards an overall climax which, after having been twice avoided (see the downbeats after the wedged notes), is finally reached in the appoggiatura-resolution group after the third sub-phrase. (The wedges would thus be read as indicating attitude - forward-directedness - rather than simply technique.)

In addition to the ornamental figures written out in sixteenth-note values, the prelude contains a number of embellishments. The mordents on the weaker part of the appoggiatura-resolution pairs in bars 8, 24 and 32 should only be played by performers who can trill extremely softly without disturbing the relaxation. The same holds true for those mordents which, in U: bars 34 and 36, fall on notes in the middle of a gradual decrease. If executed, all these mordents commences on the main note since they are approached stepwise, and contain three notes. The inverted mordent in bar 31, like all mordents mentioned before printed in brackets and thus obviously deriving not from the manuscript but from am early copy, might be interpreted as a means of emphasis; one can easily imagine the composer-teacher adding it in the score of a student who failed to properly accent the syncopation in the lower part. (If played here, consistency requires that the ornament be added in any corresponding bar, i.e. in U: bar 7.)

The turn in U: bar 32, if played - it is also given in brackets - begins on the main note and thus encompasses five notes: four thirty-second-notes and a final sixteenth-note. The grace-note in U: bar 57 designates an appoggiatura of eighth-note duration, possibly with an implied pause already for this note; its resolution, now also reduced to eighth-note value, is definitely lengthened by the fermata.


II/24.1.4 What is happening in this prelude

The thematic material of this piece consists of three motives. They represent lessening degrees of importance, diminishing polyphonic independence, and gradually reduced complexity.

M1 is the prelude's main motive; it marks the beginning of all four sections. This motive, first presented in U: bars 1-4m, consists of two almost identical subphrases (compare bars 1-21 with bars 3-41) which are linked (in the second half of bar 2) and completed with a tail (see the descending broken chord in bar 4). In both subphrases, the initial bar is composed as a circling motion around the note D, and the climax falls on the following downbeat C#. Among the two climaxes, the first is more important; the second, a mere repetition, is slightly weaker. Harmonically, M1 contains an interesting design in which each subphrase ends on the dominant; the tonic is only transitorily regained at the beginning of the second subphrase (see bar 3d). At the end of the motive, a link is required to return to both the downbeat-position and the tonic.

M1 is accompanied by four different contrapuntal companions (CM is here used for counter-motive); two of them are remotely related to one another.

CM1a is introduced in bars 1-4m. Its phrase structure matches that of M1, with two subphrases of almost identical design (compare L: bars 1-2d with bars 3-4d). The dynamic development also consists of two curves of which the second, a repetition, is weaker. The climaxes fall on the tied notes F#, followed by a relaxation until the first or second quarter-note in the next bar. (Beats 3/4 in these bars are links.) CM1a recurs in bars 17-20 and, shortened, in bars 41/42. It thus accompanies the main motive in crucial situations such as the beginnings of the prelude's three larger sections.
CM1b is first presented against the imitation of M1 in bars 5-8. It is essentially different in structure, consisting of three sequences (the last of them varied) that build a single overall arch and thus do not share the binary design of M1. In bars 5 and 6, the upper voice contains three falling quarter-notes after downbeat rests; in bars 7/8, the falling broken chord is ornamented with a trill+suffix figure and extended to include an appoggiatura-resolution figure. Each of the three descents comes with a natural decrease in tension. Among the three peaks which are organized in ascending order, the highest (see bar 7: F#) is additionally enhanced by its syncopated rhythm and thus offers itself as the evident overall climax. This contrapuntal companion recurs in L: bars 29-32.
CM1c is introduced in U: bars 21-24. It shares with CM1b the downbeat rests, the ascending peak notes, and the concluding appoggiatura-resolution figure. It is distinguished from CM1b in that the subphrases are not falling but rising, with peaks on the fourth quarter-note in the bar. Also, the third subphrase is not simply extended but, as it were, metrically complemented by a fourth subphrase. This contains an upbeat + appoggiatura-resolution, the overall climax of the four-bar unit towards which the wedged peaks were pointing. This contrapuntal figure also recurs once (see bars 59-62).
CM1d appears only once, in U: bars 25-28. Commencing as a parallel to the first bar of M1, it continues as a free play of the various elements which were characteristic for the other companions. Interestingly, its phrase structure is even more independent of that in M1 than was the case in CS1b and CS1c. In its initial subphrase (see bars 25/26), the M1-parallel leads to a first upbeat + appoggiatura-resolution figure (E-E-D#) followed by an escape note and another appoggiatura-resolution (B-A); the climax within this subphrase falls on the downbeat E, a secondary climax on B. The second subphrase, only one bar long, is conceived as a varied partial sequence. It begins with an ornamented version of the upbeat + appoggiatura-resolution figure (F#-F#-E) and concludes similarly in the lower register (A-A-G). The final subphrase begins with one more upbeat + appoggiatura-resolution figure (E-E-D#). Its rhythm is contracted and thus prepare a climax on the syncopation following it. This high A serves also as the overall climax within four-bar unit of CM1d.

The sketch below attempts to sum up the details of phrase structure and dynamic design in the prelude's main motive and its contrapuntal companions (ex. 65).


follows the two initial statements of M1 after a link (see bars 8m-9d). Like M1 it consists of two subphrases; here the second subphrase is a lower sequence of the first one (compare bars 9-11d with bars 11-13d, each beginning after the initial eighth-note in a bar). Within each subphrase, the second half is an almost exact inversion of the first (compare bars 9-10d with bars 10-11d, each beginning one eighth-note after the strong beat). The most likely spots for climaxes in this very complex rhythmic pattern are the high-pitched, syncopated notes F# and E which, together with the low-pitched syncopations one bar later and the resolutions on the final downbeats of each subphrase, form dramatic lines (see F#-D-C#, E-C#-B).
is accompanied by quarter-note motion (CM2a). As the accompaniment does not claim any attention of its own, the listener's understanding is facilitated if these quarter-notes also describe two-bar dynamic curves, with a crescendo towards bars 10d and 12d respectively, followed by diminuendo until bar 11d and 13d.
(ex. 67)

M2 recurs twice. In L: bars 45-49d it is supported again by CM2a. In bars 33-37d, however, it is set against a contrapuntal line (CM2b), similarly designed as a two-bar subphrase with sequence, although very independent in detail. Climaxes in the two subphrases fall either on the high-pitched middle-beat notes (G in bar 33, F# in bar 35) or, possibly better in the polyphonic design, on the downbeats A and G. Phrasing is most convincing if aligned with M2; after a link (see D# to E in bars 32/33), the model spans from B in bar 33 to bar 35d, followed by its sequence from A in bar 35 to bar 37d.


is introduced in bars 13-16. It is designed in four ascending sequences. Each commences on the syncopated second eighth-note of a bar, culminates on another syncopation and relaxes through a trill-plus-suffix (similar to that in CM1b) until the following downbeat. As the sequences ascend, so does the dynamic intensity. The accompaniment to this motive (CM3) is completely devoid of characterizing features and does not even invite phrasing.
M3 recurs once (see bars 49-53d). Apart from two details - the final note does not continue the descent but bends back, and the accompaniment contains larger intervals - there are no essential modifications.

The layout of the prelude as established by the motives is summed up in the table below.

Section I Section II Section III Section IV
M1 + CM1a M1 + CM1a M1 + CM1a var M1 + CM1c
M1 + CM1b M1 + CM1c M2 + CM2a + 4 bars
M2 + CM2a M1 + CM1d M3 + CM3a  
M3 + CM3 M1 + CM1b + 4 bars  
  M2 + CM2b + 2 bars  
  + 4 bars    

Section I is taken up completely by the presentation of all three motives. M1, as the most important motive in this prelude, occurs twice and with two different contrapuntal companions. It is followed by the still quite complex M2 and rounded off with the much simpler M3. The lessening complexity and, as a result, intensity of the motives, together with the descending order of sequences in M2, creates a gradual decline of tension throughout the first twelve bars of the section, followed by a straight-forward increase in the ascending sequences of M3.

Section II is set up like a development. M1 recurs four times, modulating and taking four kinds of companions. M2 is also modified as it comes with a polyphonically active companion. The final four bars of the section develop the upbeat + appoggiatura-resolution figure in a syncopated version (see U: bar 37 C#-G-F#, sequenced and freely imitated in bars 37-39, and liquidated for retransition in bar 40).

Section III seems conceived as a kind of recapitulation. The two M1 statements at the beginning of the prelude are here contracted into one which experiences an inversion of voices half-way through the motive (see U: bars 41/42, L: bars 43/44). This is followed, as in section I, by M2 and M3 with their original accompaniments. Four bars with a descending bass line (B-A-G-F#-E-D-C#-B-A#, see L: bars 53-56) and free development in the upper voice are followed in a fifth bar by a suspended V7 chord (with fermata and "general pause"), after which a cadential formula supplies the return to the tonic. With regard to dynamics, the slight decrease in M1 and M2 is answered here by a stronger increase, due to the modification of M3 which causes the sequences to ascend not in steps but in fourths. Bars 53-57 reply with a decrease which is not yet complete when it is interrupted. The resolution follows all the more stringently in the ensuing cadence.

Section IV appears as a coda; it consists only of a single statement of M1 followed by four cadential bars in hardly disguised homophonic texture. Interrupted by three further "general pauses" (see the middle beats in bars 62, 63 and 64), these bars pick up the appoggiatura-resolution featured prominently in the middle section of the piece and lead to a final climax on the downbeat of bar 64.



WTC II/24 in B minor -- Fugue

II/24.2.1 The subject

The subject of this fugue spans five and a half bars, from the upbeat to bar 1 to the downbeat of bar 6 where the tonic is regained in its third, D. The combination of simple upbeat, downbeat ending and three-eight time immediately evoke a lightly swinging character. The phrase structure, however, is irregular. The initial two-bar subphrase consists of a falling tonic chord complemented by the leading-note and a return to the keynote. The second subphrase, four bars long, can read as an elaborate version of the descent G-F#-E-D.

The rhythmic pattern in the subject itself as well as in the entire fugue is simple, consisting predominantly of eighth-notes and sixteenth-notes. The pitch pattern contains not only broken chords (as in the first bar of the subject) but also numerous melodic octave jumps (see bars 3/4).

The harmonic design of the subject can be regarded as a simple cadential progression (see ex. 68a) with harmonic ornamentation (given in ex. 68b):

(ex. 68a)

(ex. 68b)

The subject's dynamic design follows very naturally from a combination of the features mentioned above. Within the first subphrase, the upbeat enhances the downbeat of bar 1, while the downbeat of bar 2, representing only the return to a note already heard, is completely relaxed. In the second subphrase, the pitch line coincides with the harmonic design; the climax falls on the downbeat of bar 3 which represents both the harmonically active subdominant and the beginning of the descent.

The second of the two climaxes is stronger than the first. In terms of phrase structure, the subject thus presents itself as consisting of a smaller "preliminary gesture" followed by a larger "main gesture".


II/24.2.2 The statements of the subject

The fugue contains nine complete subject statements as well as an incomplete one.

1. bars 0-6d M 6. bars 44-50d L
2. bars 6-12d U 7. bars 54-60d M
3. bars 15-21d L 8. bars 70-76d L
4. bars 26-32d M 9. bars 81-87d U
5. bars 35-41d U 10. bars 96-98d Mincomplete

(ex. 69)

Apart from the usual adjustments in the answer, the subject contains only a single modification in the varied (and extended) upbeat in bar 70. Neither inversions nor grouping into strettos or parallels occur.


II/24.2.3 The counter-subjects

Bach has given this subject two counter-subjects which accompany the subject alternately. The polyphonic density thus does not surpass that of two thematically shaped voices.


is introduced against the second subject statement (i.e. the answer). It contains three characteristic features:
It begins with an imitation at the octave of the subject's first subphrase (compare U: bars 6-8d, B-A-F#-E#-F# with M: bars 7-9d. The exact beginning of CS1 is difficult to determine. One reason is that the notes against the first bar of the subject are different each time (compare M: bar 7 with M: bar 16 and U: bar 27). More importantly, the sixteenth-notes linking the first subject entry to the second in bar 6 certainly never recur and cannot be regarded as part of the counter-subject. Instead, they must be interpreted as an episode, in keeping with Bach's obvious design in this fugue to separate all subject statements by episodes. This episode, commencing thus on the second sixteenth-note of bar 6, could be understood as concluding on the downbeat of bar 7 or on the second beat of that bar. All we can safely claim with regard to the beginning of CS1 is therefore that its initial imitation of the subject's first subphrase comes with a varied upbeat of at least two descending sixteenth-notes which might, however, be preceded by one or three more sixteenth-notes in varying shape.)


This is followed by a trill with anticipated main note and ascending resolution; a sequence of same trill concludes CS1 (see M bars 9/10: G#-G#tr-A, and bars 11/12: E#-E#tr-F#)


The two trills are linked by a scalar descent (see bars 10/11: B to D#).

The dynamic design of CS1 commences with an imitation of the process in the subject's first subphrase (see bars 7-9: an actively increasing upbeat in bar 7, a climax on bar 8d and a relaxation until bar 9d).
The trill prolongs the relaxation for another bar (until its resolution in bars 10d). The descending scale then serves as an upbeat to the second trill which, on a dynamically lower level, repeats the climax/relaxation and thus completes the counter-subject. In all, the two subphrases in CS1 sound predominantly in diminuendo and thus build a good contrast to the subject.
CS1 only accompanies the subject statements in bars 6-12, 15-21 and, partially, that in bars 26-32. As soon as the second counter-subject emerges, CS1 breaks off and disappears. (This retreat of CS1 at the arrival of CS2 is repeated in bar 73).


is first presented in L: bars 29-32d. Its features are most unusual for a contrapuntal figure in Bach's polyphonic style. Moving in regular sixteenth-notes with constant large jumps back and forth, this counter-subject resembles an accompaniment pattern in homophonic texture. One easily detects the hidden two-part structure which comprises the bass line B-B-A-A-A-G#---C#-F# and, as "background", an ornamented descent A-G#-F#.
The dynamic representation of this counter-subject must be as simple as its structure: a single decrease of tension which thus presents no polyphonic contrast whatsoever to the subject

The sketch shows the phrase structure and dynamic design in the subject and its two counter-subjects (ex. 70):


II/24.2.4 The episodes

The ten subject statements in this fugue are interspersed with nine subject-free passages. Some of them are extremely short ("hardly worth giving a name", as a student once remarked), others are up to fourteen bars long. There are, however, good reasons to take even the very short links seriously - not only on paper when analyzing but, more importantly, when performing. Both the particular design of this fugue with its alternation of statements and episodes, and the observation that the subject's initial subphrase is imitated at the beginning of the first counter-subject (which thus enters "late") require this concept. Furthermore, this "late arrival" of CS1 causes frequent overlapping of episode material and subject entries. The episodes are:

E1  bars 6/7 (only M) E6 bars 50-55d (54/55 only L)
E2 bars 12-16 (U + M) E7 bars 59-69d

(59 only L)

E3 bars 21-27

(U + L)

E8 bars 76-82d (82 only L)
E4 bars 32-36d (35/36 only L) E9 bars 87-100  
E5 bars 41-44d        

Material occurring in the episodes includes motives deriving from the subject's second sub-phrase (below referred to as Ms), the first and second counter-subjects (Mcs1, Mcs2), as well as a number of independent motives. In addition, there are false entries (subject: see M: bars 96-98d; CS1 see L: bars 973-99d). Interestingly, the motives stemming from primary material occur almost exclusively in the initial third of the fugue, while those in the following third introduce more and more independent features as secondary material, and those in the final third review material introduced earlier.

A unifying feature among many of the motives is a sixteenth-note run in various shapes, always beginning after the downbeat and ending at the beginning of the next bar with one or three sixteenth-notes. This run may appear on its own as in E1 (see M: bars 6/7 F# to C#), introduce an episode (see U+M: bars 41-42d) or connect between two episode segments (see L: bars 91-92d, 96-97d). Moreover, it may actively launch a motive (see below Ms, Msa, Mcs1a) or passively conclude it (see below M1b, M2, M2a).


consists of the above-mentioned run in increasing tension, followed by the ending of the subject. It is first presented in E2 (U: see bars 12-14d, sequenced in bars 14-16d). Clear phrasing between the model and its sequence is vital for clarity.
Msa is used for different variants in which the interrupting eighth-note in the middle of Ms is avoided by an additional sixteenth-note. The resulting two-bar sixteenth-note motive appears in various shapes, always in the lower voice (see in E3: bars 25-27d) and, surprisingly, not only in episodes but also against the subject in those statements where CS1 is absent (see L: bars 36-38d, 55-57d, 82-84d). Although the pitch direction in the runs changes in each variation, Msa should retain the dynamic shape of its forerunner, with one bar crescendo towards the downbeat in the middle and a following diminuendo.

Mcs1 focuses on the trill - with prefix and suffix - that characterizes the first counter-subject. This motive is introduced in E2 (see M: bars 12-14d).
Mcs1a integrates the sixteenth-note run in E3 (see L: bars 21-23d, 23-25d, and U: bars 22-24d, 24-26d). In both versions, the trill with its resolution must retain the decreasing dynamic shape it had in the counter-subject. The motive describes a curve, with a climax on the downbeat before the trill.

follows the first appearance of CS2 from which it develops as a kind of extension (see L: bars 32-34d, 34-36d). It is conceived in hidden two-part structure; the melodic component, in the lower part, is a rising scale which expresses rising tension. The motive recurs in L: bars 87-89d, 89-91d.

is introduced in E4 (see M: bars 32-34d; U: bars 33-35d; M: bars 34-35). Its prominent trait is a syncopation which, owing to changing harmonic surroundings, turns into an appoggiatura as it is suspended into the next downbeat and consequently resolves downwards. The resolution is followed by descending eighth-notes. The dynamic layout of the motive is that of a curve in which the climax falls on the syncopation. M1 recurs in U/M bars 87-91d and, without the stretto imitation, in U: bars 42-44d.
A variation of M1 isolates the prominent syncopated appoggiatura with its resolution (see M: bars 493-542 and bars 913-962). In both cases, the two accompanying voices display new figures. M1a (see U: bars 50-52d, 52-54d and 92-94d, 94-96d) represents rising chromatic lines (in crescendo), while M1b (see L: bars 50-52d, 52-54d and 92-94d, 94-96d) describes a new version of the dynamic curve in two bars of uninterrupted sixteenth-notes.

is again based on a syncopation which appears here combined with the sixteenth-note run to a single diminuendo gesture (see L: bars 592-61d, bars 612-63d, bars 632-65d, bars 662-68d; and, with a fourth upbeat and in complementary voices, U/M: bars 602-62d, bars 622-64d, M: bars 642-66d).

M3 and its contrapuntal companion M3a are developed as extensions of the preceding subject/CS2 combination. M3 begins with a vivid pattern around a chromatic descent (see U: bars 76/77 D#-D-C#) followed by the final bar of CS2 (see U: bars 77-78d); the companion M3a modifies the last octave jump in the subject with a syncopation before adding the subject's final bar (see L: bars 76-78d). This model is then sequenced, both faithfully (see bars 78-80d) and in modification (see bars 80-82d, particularly M3 in L).

The following chart gives an overview of the material presented in the episodes of this fugue, and the relationship among the episodes.

E1 E2 E3 E4 E6 E7
link Ms/Mcs1 Mcs1a M1/M1/Mcs2 M1/M1a/M1b M2/M2a
  contrapuntal imitative imitative contrapuntal


      + accompaniment    


        link + M1    




E4 + link +
E6 + S/CS1
incomplete + cadence  


+ accompaniment

The design and order of the episodes in the final fugue of the Well-Tempered Clavier thus invites two very interesting observations:

while the structure of alternating subject statements and episodes seems simple, the introduction of new material is most intriguing;


E8 appears related to E2 (both are extensions of the preceding statements), and E9 reviews E4 and E6. The entries surrounding the two final episodes may thus be considered as a kind of structural recapitulation.


II/24.2.5 Character, tempo, articulation, ornament realization

The basic simplicity of the rhythmic pattern, toge

ther with the thematic broken chords and octave jumps, indicate rather lively character. The three-eight time signature supports this interpretation, hinting at fairly swift whole-bar beats.

Probably the best tempo proportion between the prelude and the fugue is achieved by relating the larger beats of the two pieces:

one half-note
corresponds with
one dotted quarter-note
(half a bar)
(one bar)
in the prelude
in the fugue
(Approximate metronome settings: prelude beats = 54, fugue beats = 162.)

Adequate articulation comprises quasi legato for the sixteenth-notes and non legato for the eighth-notes. The most significant exception - and one that is bound to cause errors for all but the very careful performers - lies in the pairs of longer note values which constitute appoggiatura-resolution. Several of them occur against "normal" (i.e. non legato) eighth-notes in another voice and therefore need careful fingering; moreover, they often develop out of suspended (tied) notes and are thus easily overlooked.

The appoggiatura-resolution pairs in this fugue, in the order of their appearance:

bar: 15 18 20 33 34 35 43 50 51 52 53 54
voice: M U U M U M U M M M M M
notes: B-A# F#-E D-C# C#-B F#-E B-A E-D D-C# C-B E-D# D-C# #-E#

bar: 77 79 87 88 89 90


93 94 95 96  
voice: L L M U M U M M M M


notes: G-A F#-E C-B F#-E B-A E-D G-F# F-E A-G# G-F# B-A#  

Ornaments in this fugue include the trills in CS1 (and consequently in the episode motives Mcs1 and Mcs1a) and the grace-notes in the final bar.

All trills begin on the upper neighbor note, shake in thirty-second-notes (i.e. twice as fast as the sixteenth-notes which are the shortest regular note values in the piece) and end without a change of speed in the suffixes as spelled out by Bach. Each trill thus comprises eight notes.
The grace-notes in bar 100 represent appoggiaturas. As their notation in brackets indicates, they derive not from the manuscript but from a copy. Ending the fugue without these appoggiaturas gives the piece a very straightforward close, while playing them results in a more complete ending of the fugue and recalls its emotional qualities. The duration of the appoggiaturas is that conventionally used in connection with dotted notes: one third of the ternary value. This gives the appoggiaturas one eighth-note and leaves one quarter-note to the resolutions.

One word about the alternative readings given in the Urtext for bars 16 and 21: The beginning of CS1 in M: bar 16 would sound meaningful in both versions, while closing the Ms sequence in U: bar 16d without a link to the tied F# seems structurally more logical. In bar 21, the alternative reading is strictly logical in the terms of the episode which it opens, but may be perceived to interrupt the particularly close link between the preceding subject entry and the episode.


II/24.2.6 The design of the fugue

The structural layout of the fugue in B minor is supported by two groups of factors which complement each other to provide a clear pattern.


The use of the counter-subjects distinctly marks the line between the first and second sections; it also helps to determine the partition between the second and third sections:

While the first three subject entries are accompanied exclusively by CS1, the fourth statement (see M: bars 26-32) and all further full entries feature CS2.


The subject statement in L: bars 70-76 is unique in that it is the only one with a varied beginning and, more importantly, the only one to be preceded by its first subphrase in another voice (see M: bars 69-71d).


The same voice which anticipated the entry continues with a sequence which, owing to its relative position to the subject, now appears as the initial subphrase of CS1. This subject entry, accompanied in its first half by CS1 and in its second half by CS2, thus corresponds with the one in M: bars 26-32 which opened the second section.


The episodes and their relationship both to the primary material and between each other support the first demarcation and also determine the line between the second and third sections:
* The characteristic motives in the episodes following the first three subject statements (E1, E2, E3) all derive from primary material which does not recur later in the fugue (see Ms, Mcs1 and Mcs1a).
* The episodes following the next four subject entries (E4, E5, E6, E7) introduce genuine episode motives (see M1, M1a, M1b, M2 and M2a) in addition to a motive derived from the newly introduced second counter-subject (Mcs2).

The remaining two episodes (E8, E9) constitute, as has been shown, recapitulations: E8 uses material derived from subject and counter-subject (thus recalling E2 from the first section), and E9 combines the motives from E4 and E6 with "fake" entries of the imitative initial subphrases from subject and CS1 (see bars 96-99d) before it closes the fugue with a final cadence.

The fugue thus clearly consists of three sections: Section I covers twenty-six bars and contains the first subject statement in each of the three voices. It ends, in the upper and lower voices, on the downbeat of bar 27, while the middle voice overlaps, with the next subject entry commencing in bar 26. Section II is forty-two bars long and encompasses four subject statements. It concludes on the downbeat of bar 69.

Section III begins in reduced ensemble after the downbeat of bar 69. It covers thirty-two bars and contains two complete entries. While it structurally retraces much of the layout presented in section II, it also features traits from section I (see E8 and the recurrence of CS1 at the beginning and end of the section).

For a sketch showing the design of the fugue in B minor see ex. 71.


II/24.2.7 The overall dynamic outline of the fugue

The same characteristics which distinguish the three sections structurally by means of a change in the counter-subjects and the episode material, also influence the varying intensity.

The first section is characterized by a relatively high degree of polyphonic sophistication. On the one hand, the imitation of the subject's initial subphrase in CS1 evokes the momentary impression of stretto. On the other hand, the close relationship of the episode material in E2 and E3 allows for no real color contrast throughout the section. The tension thus rises gradually, subdued only shortly in E1 and in the dynamically decreasing E2 with its descending sequences.

Section II begins, overlapping with the end of section II and growing out of the ascending sequences of E3, on a fairly elevated level of tension. As soon as CS2 emerges and CS1 is abandoned (which means that a polyphonic companion is substituted by a virtuoso accompaniment pattern), the earlier intensity gives way to extroverted playfulness. This is well supported both by the episodes (which invite contrasts in color, because of their unrelated material, and in mood, because of their graceful suspension-resolution figures) and by the key of the subject statements, two of which are in major mode (see U: bars 35-41 and L: bars 44-50).

Section III begins with a return to the more energetic character of the first section. Commencing in a setting of reduced voices and without overlapping with the previous material, it represents a genuine new beginning. In the further course of this section, the return of both the counter-subject in the form of a homophonic accompaniment pattern (CS2) and the episodes in contrasting color which are characteristic of section II brings the more outgoing mood. E9 combines a diminuendo in the descending sequences of the original E4 with a crescendo in the ascending lines of the original E6. This inverted dynamic curve is then topped off by the incomplete subject statement which retrieves, for a last time, the initial phrase of CS1. The fugue thus ends in a very confident mood.