WTC I/10 in E minor – Prelude 

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

I/10.1.1 The prelude-type

The category to which this prelude belongs can most clearly be defined by looking at its forerunner, the Little Prelude in E minor from the Notebook for Wilhelm Friedemann. This original version is only 23 bars long; it displays a sixteenth-note bass-line accompanied exclusively by block chords in recitativo style (chords in eighth-note length followed by three-eighth-note rests, on the first and third beats of each bar). In this earlier setting, the prelude clearly represents a harmonically determined composition.

The prelude as it is found in the Well-Tempered Clavier consists of altogether 41 bars. The first half (bars 1-22) rest on a sixteenth-note bass line which has been kept almost unchanged*. The recitativo-style block chords have also been retained; they appear now in the middle layer of the texture and often reduced to only two notes. These two layers are topped by a cantilena in highly elaborate fashion which is reminiscent of a violin-obbligato part in a baroque oratorio aria. This line sounds so unique that it might make the listener overlook the fact that it is only an ornamented variation of the highest notes in the original block chords. In other words, in its initial twenty-two bars the E minor prelude is essentially a harmonically determined piece wrapped in luxuriant adornment.

* [These are the notes in which the Wilhelm Friedemann version differs from the Well-Tempered Clavier version:


The G on the downbeat of bar 9 avoids the seventh jump downwards by appearing one octave higher; the line is thus smoother, but the bass-note descent from bar 6 to bar 9 is less satisfactory without its expected target in the lower octave.


In bar 13, the re-sharpening of the F already occurs in the second half of the bar and not, as in the later version, only in bar 14.


In the second half of bar 15, the original version features the harmonically logical figure G-B-C-D-C-D-C-B which, in the adaptation, is modified to G-A-B-C-B-C-B-A in favor of an assimilation with the ensuing bars.


Bars 21/22 are erected on a repeated tonic pedal; as the reworked prelude for the Well-Tempered Clavier is destined to continue with a weighty second half, these bars are changed to support the intended modulation.]

The second half of the prelude (bars 23-41) carries the tempo indication Presto which, although literally only indicating a change in the pace of the prevalent pulse, leads performers and listeners alike into expecting entirely new material. It may therefore come as a surprise that these nineteen bars rest on a sixteenth-note figure which, on closer inspection, reveals a striking relationship with that of the first half.

What is new is that here, the right-hand part also appears in sixteenth-notes which, for the most part, move in parallel to the bass but nevertheless create the impression of ornamented broken chords. Although the final nine-bar portion features several details which will deserve special mention later, we can safely maintain that this part of the prelude is also, above all, harmonically determined.


I/10.1.2 The overall design of the prelude

The first harmonic progression concludes in bar 4 (see bar 1: i, bar 2: ii7 inverted, bar 3: V9 inverted, bar 4: i). This harmonic close comes with a melodic caesura (with phrasing in bar 4 after the middle beat). On the last three eighth-notes of bar 4, a melodic upbeat prepares the beginning of the ensuing phrase. The next harmonic progression modulates to the relative key (G major) which is reached in bar 9m. This progression coincides again with a melodic phrase.

Throughout the entire prelude one can distinguish 8 structural sections. The group into two more embracing parts:



bars 1 - 4m

(complete cadence in E minor)


bars 4m - 9m

(modulation to G major)


bars 9m -15d

(modulation to A minor)


bars 15m -21d

(modulation back to E minor)


bars 21m -23d

(modulation to A minor)



bars 23 -26m

(complete cadence in A minor)


bars 26m -28d

(modulation back to E minor)


bars 28 -41

(confirmation of E minor)

There are several portions in the first half of the prelude which recur in the Presto.


The initial four-bar cadence in E minor reappears, at the beginning of the Presto, transposed to A minor (compare bars 1-4 with bars 23-26; see particularly the faithfully corresponding bass line).


The diatonically descending bass line of bars 14-17 (C-B-A-G-F#-E-D#) is taken up in bars 27-30 of the Presto (G-F#-E-D-C-B-A), surrounded by an analogous harmonic pattern. 


The return modulation from the end of the first half is used to close a phrase in the middle of the Presto (compare bars 19-21 with bars 31-33).

To sum up, one can state that, the final eight bars aside, the Presto presents a recapitulation in a chain, without links, of three separate passages from the first half of the prelude:


bars 23-26

correspond to

bars 1-4

bars 27-30

bars 14-17

bars 31-33

bars 19-21


I/10.1.3 Practical considerations for performers

The choice of tempo in this prelude is tricky because it has to accommodate the change of speed in the middle of the piece in such a way that the second half develops as naturally as possible from the first. The simplest and the most balanced way of achieving it is to play the Presto sixteenth-notes twice as fast as the sixteenth-notes in the first half of the prelude.

With regard to the articulation one can assume that the first half of the prelude, with its cantilena in highly complex rhythmic pattern, represents a rather calm basic character while the Presto obviously indicates a contrasting character. The sixteenth-notes in the first half should therefore sound in true legato, while those in the second half benefit from a crisper touch. The melodic line in the first half is also very much legato (Bach's paired slurring in bar 3 and the uncommon interruptions of the trills just before the suffixes in bars 10 and 12 take care of some deliberate variety in this pattern), while the remainder of the original block chords in the middle voice fits best if played in a neutral shade of non legato articulation. In the Presto, as in any lively piece, longer note values should be detached; yet apart from the leap in bar 24 (upper voice) and the half-notes in bars 36-37 (middle voice) and 38-39 (upper voice), there are no other notes to which this would apply.

The first half of the prelude contains five ornaments. As all of them are long trills, the paramount concern will be the tempo of the shake. The immediate choice of performers familiar with the rules of ornamentation in Baroque polyphonic music would be to shake in notes twice as fast as the shortest appearing values (i.e. with four notes to a left-hand sixteenth-note). For those who find this technically too demanding or musically too congested, there are, in this particular piece, indicators why a trill speed in slower values might be not only acceptable but, some believe, even advisable:


The fastest occurring note values, the thirty-second-notes, all appear in a strictly ornamental context; these groups could be interpreted as spelled-out unaccented ornaments and embellishments (see e.g. bar 1: an upbeat turn; bar 3: an embellishment of the step A-G; etc.). It would therefore make sense to play those ornaments which appear in sign notation in the same motion.


Four of the five trills end in suffixes which are also spelled out. This is either because an interrupted trill is normally not expected to conclude with a suffix (as in bars 10 and 12), or because Bach desires suffix notes different from those according to the rules (as in bar 14). Note, however, that the very regular suffix in bar 1 is also written out. These suffixes appear in thirty-second-note notes, a fact which supports the rendition of the entire trills in thirty-second-notes.

Having settled the question of "how fast?", the next issue is "how to begin?". Four of the five trills (bars 1, 10, 12 and 20) are approached in note repetition and thus commence on the upper neighbor note. The fifth (bar 14), preceded by stepwise motion, begins on the main note which is held for a sixteenth-note.

Finally, the "how to end the trill?"-question is special in this prelude. Only the trills in bars 1 and 14 feature regular motion right into the written-out suffix. The others are either interrupted by a sudden rest (which just cuts out one thirty-second-note pair but leaves the remainder of the trill intact; see bars 10 and 12), or end in an anticipated resolution (see bar 20) which requires a point d'arrêt stop shortly before the anticipation note. The complete rendition of these trills with their respective approaches can be seen below in the slower of the two possible speeds.

(ex. 37)



I/10.1.4 What is happening in this prelude?

The question of dynamic development is a special issue in this prelude. In the first half of the prelude, it needs to be addressed separately and in detail for the harmonic background and the cantilena because of occasional contradictions of harmonic progression and melodic features.

Inside the closed cadences, the harmonic climax falls on that beat where the subdominant harmony first emerges (in bars 1-4, on the downbeat of bar 2; in bars 23-26, on the downbeat of bar 24). In bars 23-26, both hands support this design. While this is presented, in bars 1-4, by the left-hand line and the chords in the middle of the texture, the cantilena is obviously headed towards a melodic climax on the middle beat (A) of bar 3.

In passages where the original prelude featured clearly discernible sequences, separate rules which override those of the immediate harmonic relationships apply. Such a sequence pattern can be found in bars 5-8 (see both the bass notes E-D, D-C, C-B, B-A and the paired pattern of the double thirds in the middle). In the model (bar 5), the chord in the first half of the bar relaxes towards the chord in the second half of the bar (the inverted B minor chord being the resolution of the inverted F#7 chord). Throughout the four-bar descent, this pattern of tension/resolution should therefore be repeated, each time on a slightly softer level. The final chord of this section, the G major harmony in bar 9, thus sounds as a soft ending. The cantilena enhances this by preparing each of the seventh chords with an elaborate upbeat while sustaining a long note at each of the moments of resolution.

Another succession of sequences appears in bars 9m-13m. Here the model is two bars long; the bass line G-G-F#-E is sequenced as E-E-D-C, and the harmonic progression of a modulation from G major to E minor is sequenced as a move from E minor to C major. The climax within the sequence model falls on the step which actively leaves the momentary tonic, i.e. on bar 10d, after which the tension subsides gradually until bar 11m. This dynamic pattern is then taken up, slightly softer because of the generally descending direction, in the sequence. The cantilena supports this pattern again by approaching these climax points in large leaps (see bars 9/10 = octave, bars 11/12 = minor sixth).

Similarly, in passages where a clearly distinguishable bass-note pattern creates a superimposed line, this line will outrank any small-scale increases and decreases in the harmonic tension. In bars 14-17 and bars 27-30, the bass descends in diatonic half-bar steps. The superimposed harmonies represent an increasing tendency, except for one definite resolution at the respective ends of the short modulation (onto A minor on bar 15d, onto E minor on bar 28d). The bass line Bach composed reflects this break in the dynamic line by a break in the straight pitch line (see in bar 15 where the A appears an octave higher than expected, and in bar 28 where the same holds true for the E.)

This descent is prolonged with a chromatic descent which, as it surpasses the boundaries of the home key and brings with it very high-tensioned harmonies, creates an even stronger increase in tension (see bar 17m to bar 20d). The ensuing resolution onto E minor brings a relaxation.

The corresponding phrase in the Presto features an extension over a diatonic bass line which reaches its climax on the chord which corresponds to the one mentioned above (i.e. that on bar 32d). The ensuing relaxation is not quite as complete, both because the dominant appears in minor mode (thus lacking the leading-note) and because the A minor chord in bar 33 is inverted. In all these developments, however, the right-hand part supports the general harmonic design.

Finally, what remains are the final bars of each half of the prelude. In the first half, bars 21/22 modulate actively towards A minor and therefore represent an increase in tension. In bars 33-41, a rather abrupt increase (see bars 33/34) leads to the considerably softer beginning of the pedal note B. During the length of this pedal (see bar 34m until bar 39), the dynamic growth is that generally associated with an extended pedal note: it is very smooth and gradual. However, it manifests itself here not only in the obstinate repetition of the bass note but also in an intensification of the texture from three to four voices (compare bar 34 with bar 38. Note that this increase of texture does not change the essential figures in the least). The logical target for this crescendo is the interrupted cadence in bar 40, after which the home-key tonic is approached in an overall diminuendo.

The simplified outline of the prelude in E minor given in ex. 38 tries to capture these processes.



WTC I/10 in E minor – Fugue


I/10.2.1 The subject

The subject of this fugue is slightly over two bars long; it commences on the downbeat of bar 1 and ends on the second eighth-note of bar 3. As it modulates one fifth up to B minor, the closing cadence is represented by the dominant of this target key, i.e. the F#-A#-C# in the latter part of bar 2, and its resolution falling on the first beat of bar 3. This subject consists of a single, indivisible phrase.

The reason why the second eighth-note in bar 3 must be regarded as the final note, thus creating an unusual overlapping of the end of the first subject statement with the beginning of the next one, lies in the particular melodic structure of the subject. The pitch pattern is unusual in that the single line is in fact composed in hidden two-part structure. In an attempt to determine which notes belong to the melodic part of the structure and which to the (harmonic) background, the following layers are revealed:


The melodic part of this structure comprises an initial E (represented by either or both of the key notes in the first chord), followed by a chromatic descent to the B on the downbeat of bar 2; the end of the line is then provided by the A# on the second beat in bar 2 and the two upbeat groups G-F#-G and F#-E-D.


The harmonic background is represented first by the chord notes of the tonic harmony, and then by a keynote pedal (in an offbeat repeated-note pattern). This part of the harmony is concluded by the ornamented keynote pedal (E-D#-E) after the downbeat of bar 2. The modulation to B minor shows on this "background" level in the offbeat notes C# and A# (5th and 9th sixteenth-notes in bar 2) which find their resolution onto the target key on the equally offbeat B (second eighth-note in bar 3).

A written-out version of this hidden two-part structure with possible variations in the first beat would look as shown here (ex. 39).


Because of this particular structure, a regular analysis of the pitch pattern by describing the occurring intervals would be meaningless for, as can be seen in the example, successive notes hardly ever belong to the same hidden line and therefore transcend the original concept of intervals.

The rhythmic structure shows only sixteenth-notes and eighth-notes in the subject. Observing the variety of note values throughout the composition one finds that these two basic values are doubtlessly predominant, and the third rhythmic unit, the tied-over quarter-note in the counter-subject, does not change the fundamentally simple rhythmic pattern.

The harmonic structure of the subject is very simple; beside the original tonic and the two chords representing the modulation it only contains a rudimentary subdominant on the third beat of bar 1 (ex. 40).


The swift motion permits no particular emotional involvement in melodic details. At the same time, neither the simplicity of the harmonic design nor the rhythmic pattern in uninterrupted sixteenth-notes offers any climax. Therefore, the only feature in a position to influence the development of tension in this subject is the straight descent in the melodic part of the structure. This leaves the subject with a very straightforward dynamic gesture: an energetic beginning followed by a gradual decrease in tension.


I/10.2.2 The statements of the subject

There are altogether eight complete entries of the subject. These appear as follows:


bars 1 - 3



bars 20 - 22



bars 3 - 5



bars 22 - 24



bars 11 -13



bars 30 - 32



bars 13 -15



bars 32 - 34


(ex. 41)


The only variation in the subject is the omission of the unaccented second eighth-note at the end; this slightly shortened form (in which the "background" within the hidden two-part structure does not resolve properly) can be observed in the lower-voice entry of bars 13-15 and in the upper-voice entry in bars 32-34, i.e. in the fourth and eighth subject statements in the fugue which just obtain a distinct hint of structural correspondence.

Neither parallels of the subject nor strettos are used in this fugue.


I/10.2.3 The counter-subject

As this fugue is conceived in only two voices, one is not surprised to find only one counter-subject. It is introduced against the second subject statement but is shorter at both ends, beginning after beat 2 in bar 3 and ending on the downbeat of bar 5.

The predominant note value in the counter-subject (as in the subject) is the sixteenth-note. The only longer note appears in the middle; it is a quarter-note on the downbeat which is tied into an additional sixteenth-note on beat 2. Following this interruption, the original sixteenth-note motion resumes.

It seems worth noting that the counter-subject is not composed in hidden two-part structure but in a regular one-track pitch pattern consisting of various ornamental figures. An analysis of the melodic structure reveals that each of these four-note groups consists of three neighboring notes, one of which is repeated as an axle note. These figures are different from one another. The design depicts their various shapes by placing the axle note in each figure on the line (ex. 42).


The choice of a climax poses no problem whatsoever; it falls on the sudden long note value which, in addition, represents a strong metric position.

The two sketches show the phrase structure and dynamic design in the E minor fugue. The first one represents the two-part result as it is apparent in the written version (ex. 43):


 A more accurate depiction of the effect created in a listener would look like this (ex. 44):



I/10.2.4 The episodes

The eight subject statements in this fugue are very neatly grouped in four pairs, each of which is followed by a subject-free passage. The episodes can thus be found in


bars 5 - 10


bars 24 - 29


bars 15 - 19


bars 34 - 42

The second half of E4 recalls segments of the subject (compare U: bar 39 and L: bar 40 with the first subject bar, as well as U: bars 41/42 with the second half of the subject). E1 and E3 quote the counter-subject; instead of the expected final note, however, one hears a very surprising large jump downwards, followed by the rising sixth which completed the subject.

Against this hybrid episode-motive (which could be described as "counter-subject body with subject tail") Bach places the first genuine, i.e. independent episode motive (see U: bars 5-7). M1 consists of two rising broken chords, followed by a four-note ornament, a descending scale and a long ending note. In this episode motive, two dynamic designs are possible. The climax can either fall on the peak note of the broken chord sequence (e.g. the G in bar 5), or on the rhythmically exposed tied note (e.g. the D in bar 6). The first rendition allows the motive to appear more independent while the second choice, because of the analogous length of the climax note with that in the counter-subject, makes it sound like a free imitation of the other voice.

The other two episodes, E2 and E4, provide two more motives. The lower voice in bars 15/16 introduces M2 which is characterized by a simple broken chord (complemented by the repetition of the third and octave) in eighth-note motion. The upper voice contrasts this with M3 which consists of two descending scale sections in uninterrupted sixteenth-note pattern. A very balanced effect in these two motives is achieved by subtle means: the two ascents in M2 are contrasted in M3 by two descents; while in M2 the four-note broken chord was followed by the shorter two-note jump, M3 begins with the shorter scale section which is then followed by the longer one.

The presentation of M1 and the hybrid "counter-subject body with subject tail" motive is followed by a descending sequence and rounded off by two bars of almost complete parallel structure in the two voices (see bars 9-11 and 28-30). Similarly, the combination of M2 and M3, covering only a single bar, is followed by double imitation and a descending sequence of original plus imitation; thereafter, this episode-type too is rounded off by parallel motion of the two voices (see bars 19/20 and 38/39).

The relationship between the episodes is thus a very straightforward one: E1 recurs in E3 as an inversion of the voices; E2 is equally recapitulated, also hands inverted, in the first half of E4 (compare bars 15-19 with bars 34-38).

The role which the episodes play in the development of the composition is only slightly more complex.


E2 and the first half of E4, consisting entirely of material not related to the subject or its counter-subject, provide a certain contrast of color. This is enhanced by the fact that M2 is the only component in this fugue to appear in a longer succession of eighth-notes. The descending direction of the sequences and the ensuing double descent in the parallel-motion bar create a distinct effect of relaxation.


E1 and E3 are more closely related to the primary material of the fugue and thus create less contrast. The descending sequence certainly depicts a release in tension, but the ensuing parallel-motion bars build a powerful curve which brings about its own virtuoso climax, so that this episode-type ends on a level similar to that from which it commenced.


The fugue's three final bars (see bars 39-42) serve as a coda with a clearly relaxing tendency. This effect is due not only to the decreasing tendency in the subject which is partially quoted here, but also to the softening shift towards the major-key close and to the graceful ornamental arpeggio (see bar 42) which replaces the more austere sixth jump originally closing the subject.


I/10.2.5 Character, tempo, articulation, ornament realization

The simplicity of the rhythmic pattern, combined with the ornamental structure in the sixteenth-notes and the leaps and broken chords in the eighth-notes, leaves no doubt about the rather lively (most probably even very lively) basic character of the E minor fugue.

The tempo requires moderate beats in the 3/4 time signature, so that the eighth-notes will sound swift and the sixteenth-notes truly ornamental. (Ornament symbols do not occur in this fugue.) The articulation encompasses few non legato notes; these include only the final leap in the subject (and the equivalent in the hybrid episode figure), the eighth-notes in M2 and the left-hand eighth-notes in the second last bar. All sixteenth-notes as well as the quarter-notes tied into a sixteenth-note are quasi legato, i.e. played with a basic technique of legato but with "crisp fingers".

The relative tempo of the prelude to the fugue is in this case certainly determined by the continuity of the sixteenth-note motion. As the character of the fugue allows for a fairly swift tempo, the pulse of the Presto-section in the prelude can be continued directly into the pulse of the fugue. The overall proportion is thus:

one eighth-note


1 quarter-note


1 quarter-note

(prelude, basic tempo)

(prelude, Presto)

in the fugue

(Approximate metronome settings: prelude:
basic beats = 60, Presto beats = 120; fugue beats = 120)


I/10.2.6 The design of the fugue

The design of the E minor fugue is reflected unequivocally in its analogies. The analogy of, on the one hand, E1 with E3 and, on the other hand, E2 with E4 has already been expounded in detail. In addition, the first two pairs of subject statements are faithfully recapitulated (in inverted voices and different keys) in the last two pairs. This perfectly symmetrical structure with two corresponding sections (compare bars 1-19 with bars 20-38) is then complemented by the three-bar coda which, as has been shown above, was the only episode portion to state larger segments of the subject.

The harmonic outline of this fugue comprises a very active movement from one tonal area to another. This is caused by two facts: the subject itself is modulating, and the answer is conceived as a real (not a tonal) one, thus modulating further away from the original tonic.

As the modulation in each entry pair leads inevitably two fifths up, one wonders how and when the composition can smoothly find its way back to the home key. Bach solves this ingeniously by commencing the two subject pairs in the second half of the fugue from the subdominant (one fifth down) and the double subdominant (two fifths down) respectively, so that the modulations will logically return to the tonic at the end of the very last entry. (For a sketch of the fugue, see ex. 45.)


The tonal areas as represented by the subject entries are as follows
(D/ = dominant of, S/ = subdominant of):


bars 1-3

E minor (tonic)


B minor (minor dominant)


bars 3-5

B minor (minor dominant)


F# major (D /dominant)


bars 11-13

G major (relative major)


D major (D/ relative major)


bars 13-15

D major (D/ relative major)


A major (D of D/ relative major)


bars 20-22

A minor (subdominant)


E minor (D/ subdominant) (= t)


bars 22-24

E minor (D/ subdominant)


B major (D of D/ subd.) (= D)


bars 30-32

D minor (S/ subdominant)


A minor (D of S/ subd.) (= S)


bars 32-34

A minor (S)


E major (D/ subdominant) (= T)



I/10.2.7 The development of tension

The dynamic outline also displays a symmetrical design on several levels.

Within each pair of subject statements, the second statement is slightly enhanced in comparison to the first. Within each section, the second pair of subject statements represents a higher level of tension than the first: :it is composed like a rising sequence (compare bars 1-4 with 11-14 and bars 20-23 with 30-33), and the tonal area in the second pairs represents once a switch to the major mode (section I), once the return to the long-abandoned tonic (section II).

The two episodes which link the pairs in each section (i.e. E1 and E3) serve as bridges in the dynamic design. The initial relaxing tendency in the descending sequences is counterbalanced by the energetic curve in parallel motion. The two episodes which conclude the symmetrical sections (i.e. E2 and E4a) are composed in such a way that they bring forth a natural release in tension, commencing in the four-bar imitative pattern and continuing into the descending parallels of the final episode bar.

The coda may commence with another small impulse; this should, however, not sound so assertive as to be mistaken for the beginning of a third section.