WTC II/4 in C# minor – Prelude 

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

II/4.1.1 The prelude-type

The texture of this prelude consists of three polyphonically independent voices among which both imitation and contrapuntal techniques are used. The original thematic idea and its contrapuntal companion, which characterize the beginning of this piece, recur several times in the course of the composition. In-between, passages are determined by independent motives.

What distinguishes this prelude from others which one would not hesitate to describe as "written in the style of an invention" is, above all, the phrase structure within the thematic material. The main idea which will be referred to as "main theme", to distinguish it from the concept of thematic units in imitative compositions appears as strongly rhetorical. Its phrase of four bars length (see upper voice, bars 1m-5d) contains at least three very individual gestures. Their particular melodic meandering may perhaps remind the listener of the expressive vocal lines in a Baroque aria. The high degree of ornamentation which even appears varied in the recurrences of the same thematic line adds to this impression, as does the required slow tempo and the gently swinging 9/8 time. We are thus dealing here with what could be described as an instrumental aria in polyphonic three-part setting.


II/4.1.2 The overall design of the prelude

The "main theme" occurs four times, in two paired entries:

bars 1m-5d


bars 33m-37d


bars 7m-11d


bars 39m-43d


This design creates a strong impression of "recapitulation" from bar 33 onward. It is therefore worthwhile to look for further correspondences which may support or alter the emerging structural concept.


The analogy of bars U: 1m-5d with M: 33m-37d extends also to the contrapuntal line: compare L: bars 1d-5d with L bars 33d-37d. The harmonic development in both, however, shows a slight but interesting difference. In bars 33-37, the "main theme" entry and its contrapuntal companion are transposed to the subdominant. This is a function traditionally used at the beginning of Baroque recapitulations as it guarantees an ending of the piece in the tonic without the need of adjustments in the modulation: I-V becomes IV-I. It comes therefore as a surprise that the imperfect cadence at the end of the "main theme" (see bars 4/5: G#-C#-G#) is changed here (see bars 36/37: C#-F#-G#7).
The surprise continues when we discover that the transposition of the melodic lines continues (see e.g. U: bars 5m-7d which recurs in transposition in M: bars 37m-39d), but that the changed harmonic circumstances allow the second statement of the "main theme" to enter in the original key (compare bars M: 7m-11d with U: bars 39m-43d. Again the end of the phrase is harmonized differently: the downbeat of bar 11 represents V/V while that of bar 43 is the simple imperfect cadence.
If we add these observations regarding the differently harmonized phrase endings to the fact that the entering voices of the main theme in "exposition" and "recapitulation" are exactly exchanged, we suddenly see a striking cross in the symmetrical portions:

bars 1-5

main theme in U
ending in V-I-V


bars 33-37

main theme in M
ending in V-I-V/V7

bars 7-11

main theme in M
ending in V-I-V/V7

bars 39-43

main theme in U
ending in V-I-V


Other structural correspondences can be discovered, yet they do not continue straight away from the above-mentioned passages:


bars 5-7 recur once more in bars 43-45

compare all three voices in identical position and key


bars 11-17 recur in bars 56-62


U: bars 11-16d with M: bars 56-61d


U: bars 16/17 with U: bars 61/62


bars 27-31m correspond with bars 50-54m


U: bars 27/28 with U: bars 50/51

M: bars 28/29 with M: bars 51/52

L: bars 29/30 with L: bars 52/53

all three voices bars 30m-31m with bars 53m-54m


Structurally relevant harmonic closes with perfect cadences can be found

in bars 16/17 (conclusion in G# minor, i.e. in the minor dominant),
in bars 26/27 (conclusion in E major, i.e. the relative major key),
in bars 32/33 (conclusion in F# minor, i.e. the subdominant),
in bars 38/39 (conclusion in C# minor, i.e. in the tonic),
in bar 56 (in weak-beat position) C# minor, and in the final bar 62.


II/4.1.3 Practical considerations for performers

The aria-like character of the prelude can best be expressed in a calm mood combined with high expressivity. The meter must be read as indicating compound time, so that the main pulse determining the tempo is the dotted quarter-note, with the beats grouped in slow triple meter.

The appropriate articulation in such piece is overall legato. This will include the slow broken-chord figures in the "main theme", in its contrapuntal companion and in any further motivic material. Only definite cadential bass patterns must be taken non legato; this holds true above all for the left-hand part (beats 2/3) in bars 4, 16, 26, 32, 42 and, as far as possible, in bar 61. Cuts in the melodic flow occur only because of phrasing; these, while crucial for a discerning interpretation, must be very gentle so as not to disturb the mood by an exaggerated focus on eloquence.

The manifold ornaments deserve and need – in-depth investigation. A few general comments might help the performer to reach independent yet stylistically coherent judgments:

Ornamentation in compositions of arioso style was often executed on the spur of the moment. This practice derived from pieces in which the melodic lines appear enveloped in homophonic or at least less intricate texture, so that there was little risk of clashes even if the lines were given to different players in an ensemble; there was generally only one line, at any given moment, that invited embellishments. In a polyphonic composition, the need arose for a certain agreement – between the players of an ensemble at any rate. In the case of a keyboard composition with a single performer, particularly in the case of a work which the composer could expect to either perform or teach himself, the notation of these conventional ornaments was treated with some degree of negligence. Copies of various pupils show differing distribution and density of ornaments in the same piece – a fact which proves (a) that more than one solution seemed acceptable in many instances, and (b) that Bach took into consideration the particular disposition (emotional as well as, probably, technical) of each player.

It may be due to these same considerations that the composer, who would expect performers in strictly imitative works like fugues to imitate thematic ornaments regardless of awkward fingerings, was more lenient in works of this gentle nature. He thus often avoided the more difficult middle-voice ornaments and substituted them with an embellishment in the upper voice. Today's performers, whose technique has been sharpened in tasks of quite different nature, may therefore in many cases choose between following either the convention of such adjustments or observing a more logical imitative use of embellishments.

Lastly, it is vital to understand that there are a few apparently different ornaments which express a similar melodic attitude: that of an indirect approach to the target note (appoggiatura-resolution). These comprise mainly the written-out grace note (which creates a strongly accented, longer appoggiatura) and the mordent (which equally contains the on-beat appoggiatura, although in shorter value and repeated pattern). To understand their relationship is important as Bach uses the two alternatively in similar circumstances, depending on whether a longer or shorter appoggiatura fits better into the surrounding texture.

In devising an ornamentation for this prelude, it is a good idea to begin by comparing the four statements of the "main theme", considering the possibility of retaining or interchanging ornaments depending on their changed surroundings.


The inverted mordent on the second note (U: bar 1, C#) does not depend on harmonic preconditions. As retaining it in each entry is definitely conducive to the listener's early recognition of the theme, this inverted mordent should be transferred to both middle-voice entries (bar 7: C#, bar 33: F#). In bar 7, this ornament had then best replace the immediately preceding one in the upper voice, to avoid congestion.


The same applies to two subsequent inverted mordents (see bars 2, 3); if both of them are played in the initial entry of the main theme, there is no musical reason why they should be omitted later. This implies: if the second inverted mordent is added in M: bar 9 on F#, it replaces the ornament indicated for the simultaneous upper-voice note.


Another case needs more courage. In the fourth entry of the "main theme", an additional appoggiatura (see U: bar 41d) seems to disturb rather than beautify the theme. On second thoughts, however, it is easy to appreciate why Bach added a grace-note here: due to the longer note-values in all three voices, the unembellished version contains a sudden interruption of the rhythmic flow. In each of the previous statements, this continuous motion has been taken care of by one of the other voices; see in bar 3 the appoggiatura to the lower voice, in bar 9 the eighth-note (which represents a written-out appoggiatura) in the lower voice, and in bar 35 the appoggiatura in the accompanying upper voice. One may discover that in bar 41, too, an appoggiatura in the lower – rather then the upper – voice would benefit the flowing motion while leaving the thematic line intact.



The fourth note of the "main theme", in bar 2d preceded by an appoggiatura, is found in different circumstances in each of the four statements. While in bar 2, the eighth-note of the ornament resolves in a movement of parallel tenths with the lower-voice line, a similar ornament in bar 8 would result in impossible parallel sevenths. In this case the note can either be left unornamented or decorated instead with a mordent. In bar 34, however, a middle-voice appoggiatura which highlights the line of the "main theme", is definitely preferable to the written upper-voice appoggiatura.


The appoggiatura towards the end of the theme (see bar 4 on E) has in bar 10 been replaced by a mordent. Due to the eighth-note motion in the lower voice this represents the only possible solution and should not be changed.

Other corresponding motivic details must be treated in the same way; in this context they can be mentioned only very briefly:


The eighth-note appoggiaturas in U: bars 5/6 cannot be retained in M: bar 37; in this bar it should be substituted with a mordent or, better here because of the approach from below, an inverted mordent (as indicated by Bach). The eighth-note appoggiatura of bar 7d can be transferred to bar 39d. In the lower-voice line accompanying this motive, ornaments may or may not be adjusted to highlight the sequence; thus the inverted mordent in bar 6 might be displaced from beat 2 to beat 3, as suggested also in bar 44.


In U: bar 11, either of the two ornaments should be selected. Since an eighth-note appoggiatura causes the B to sound in octave with the lower voice, a mordent seems preferable. A mordent is also possible in the analogous position (see M: bar 56) while the eighth-note delay is here excluded. Later in the same line, both grace-note and inverted mordent can be played in U: bar 13 and M: bar 58, while the simultaneous grace-note in U: bar 58 may be omitted. The inverted mordent in U: bar 14d should be transferred to M: bar 59 where it sounds much better than the noted upper-voice grace-note. The cadential mordent in U: bar 16 should definitely be transferred to U: bar 61 – this kind of ornament needed no mention in Bach's times.


In bars 17-23, ornamentation in the three-part imitation is consistent apart from the additional inverted mordent on the ending-note in L: bar 23; the brackets already suggest that there were second thoughts, and omitting it is certainly conducive to the clarity of the structure. In bar 18, the accompanying lower voice features a mordent instead of the inverted mordent conventionally used in such patterns; this might be a miswriting on Bach's part. In these bars, the secondary voices might also benefit from less ornamentation; thus it is possible to omit the grace-notes in M: bar 21, L: bar 25 and U: bar 26.


In the subsequent three-part imitation (bars 27-30, recurring bars 50-53), bracketed ornaments can safely be spared (see beat 2 in U: bar 27, L: bars 29, 30) whereas the characteristic inverted mordent on the longer note (beat 3) should be retained and also transferred to the middle voice entry (M: bar 28) as well as to the corresponding passage (U: bar 50, M: bar 51, L: bar 52). The ornaments in L: bars 31 and 53, although in parenthesis, are an asset since they decorate partial sequences.


Besides the many grace-notes which, due to their stepwise resolution and harmonic independence, represent appoggiaturas, there are some which must be read as acciaccaturas. These are notes of very short duration which, while falling on the beat, immediately give way to the main note which they highlight in a virtuoso (and not in a harmonic) manner. Such notes can be recognized either as pitches belonging to the same harmony as the main note, or as chromatic "sparkles" without harmonic value. In this prelude we find acciaccaturas in U: bars 6, 7, 44 and in M: bar 38.


Finally, there are several long trills. Three of them occur in chain and, uncomfortably, in the left hand (see bars 14/15, 19/20, 59/60). In all three instances, the second trill is approached stepwise and thus begins on the (prolonged) main note. The same holds true for the first trill in bar 19, while the fact that the other two embellish chromatic progressions also suggests particular emphasis on the main notes. As a result, all trills may begin on the main note, hold this for one sixteenth-note, shake four times in thirty-second-notes with the upper auxiliary, and end in a suffix as indicated. The three remaining trills, occurring in U: bar 31, U: bar 54 and L: bar 50 respectively, have exactly the same shape – except that the latter, very softly, shakes twice as long.


II/4.1.4 What is happening in this prelude

The first section (bars 1-17d) contains three phrases and a free continuation ("codetta"). The leading features are the "main theme" with its imitation, the contrapuntal accompaniment with its later free variation, and an interlocked motive (M1) with its own contrapuntal line.

The "main theme" contains three subphrases. The first is one bar long and has its climax on the appoggiatura on bar 2d. The second subphrase consists of a four-note ascent preparing the climax, and a subsequent double note repetition which provides some relaxation. The third subphrase begins like a varied sequence: it reaches its climax in an immediate jump, its double note repetition is rhythmically extended, and a one-bar tail complements the phrase. Of the three climaxes, either the first or the second can be regarded as prevalent in the theme. Their metric placement is intriguing as it creates the impression of a hemiola (as they fall consecutively on beats 1, 3, 2). After the main climax, the impression of gradual decrease should prevail, and the tail is best kept free of any further accent.

The contrapuntal accompaniment to the "main theme" contains two subphrases: one ends on the long B in bar 2 and has its climax on the tied C#, while the second is characterized by a protracted decrease after a climax on the appoggiatura (bar 3d). A third voice fills in the texture in bars 1-3d but displays more independence thereafter. It creates a tension-increase towards the appoggiatura (bar 4d) before it resolves together with the other two voices.

On the occasion of the second "main theme" statement, the contrapuntal accompaniment is strongly varied. With only its second climax retained (bar 9d) it describes a simple curve. The third voice keeps a low profile during the statement but then adds a shortened imitation of the third subphrase from the "main theme" (see U: bar 10-12).

The first independent motive (M1) is introduced in bars 5-7. Its two subphrases had best both peak on the downbeat-appoggiaturas (see U: bar 6d and 7d). The accompaniment in the lower voice (M1a) retains many features of the "main theme", particularly the division into two subphrases and the obvious similarities at the beginning of each subphrase. The climaxes thus fall on bar 5, beat 2 and bar 6, beat 2 respectively.

To distinguish the free development in the codetta from the preceding tight-knit phrases, the dynamic lines should be kept as simple as possible there. A convincing solution is a long and gradual build-up (from L: bar 11 and U: bar 12 respectively) which peaks in bar 15 and is followed by a relaxation until the close of the section on bar 17d.

The second section begins, over an accompaniment which is launched once more from an ascending broken chord, with a second motive (M2). This motive, again, contains two subphrases. The first is closely related to that of M1 (compare U: bars 5/6 with bars 17/18), and even in the second we can discover, with some imagination, traces of the second half of M1 (compare bars 6/7 with bars 18/19). Not surprisingly, climaxes fall on the respective downbeats. As this section is characterized by the immediate imitation of M2, the performer's attention will be captured by one voice at a time. From bar 23 onwards, the lower voice dominates with partial sequences of the last- heard motive statement; climaxes of gradually lessening intensity fall on bar 233 and bar 243. The section is rounded off with a two-bar cadential close (see bars 25m-27d). The gradual decline which begins with the third imitation of M2 and continues through to the end of the section is underpinned by a descending peak-note line in the upper voice (see U, bars 21: F#, 22: E, 23: D#, 24: C#, 25: B, 26: A-G#-F#, 27: E).

The third section begins similarly with a motive (M3) in three-part imitation. While M2, as has been shown, recalled M1, M3 (see U: bars 27-28m) appears as a condensation of M1a (imagine L: bar 51 followed by bars 62-7m). Like the second subphrase in M1a, M3 reaches its climax after the broken-chord ascent on beat 2 and continues in a single unbroken decrease. As in the previous section, the third motive statement is taken up in two partial sequences (see L: bars 30m-32d) and rounded off by a short cadential close (see bars 32/33).

It has been mentioned earlier that a large portion of section IV (bars 33-45d) is conceived as a (albeit somewhat irregular) recapitulation of the thematic passage of section I, and that another passage (bars 50-55) very faithfully takes up section III. Between the two, the lower voice retrieves the habit achieved in the two preceding sections and continues with partial sequences (see L: bars 45-48d). The other two voices recall fragments of earlier motives in variation (see e.g. M: bars 45/46 from M1, partly sequenced but differently varied in bars 47/48). A descending peak-note line in the upper voice creates once more the effect of gradual decline (see U, bars 44: A, 45: G, 46: Fx-F#, 47: E#-E, 48: D#-C#-B#-B, 49: B-A-G#-F#-E, 50: D#).

After the weak-beat cadence-ending in bar 56, the final seven bars represent the coda. As was mentioned earlier, Bach had already used this coda in the conclusion of the first section.

The inner design of this prelude is thus far from simple. It contains enough repetitions to serve the intended mood of calm expressivity, and enough variation to provide for constant surprises.
















M3 "codetta"


















WTC II/4 in C# minor – Fugue


II/4.2.1 The subject

The subject commences on the downbeat of bar 1 and ends on the middle beat of bar 2. The dominant-seventh chord (represented by D#, G# and F# in bar 2 beat 2) resolves onto the third of the tonic (see E, bar 2 beat 3).

The rhythmic pattern within the subject is strikingly uniform, consisting exclusively of sixteenth-notes. These note values and their relationship to the pulse of the piece deserve careful consideration. The fugue is notated in the compound time signature of 12/16. Such an indication may seem confusing for anybody who has learned that time signatures designate how one should count, or how the conductor should beat. The meter in this fugue, however, is nothing but a hidden 4/4 time with triplet patterns in each beat.

Following this insight, the pitch pattern will have to be interpreted mainly as ornamental. If one assumes a "simplified tune" underneath the ornamented surface, it would therefore read C#-D#-G#---F#---E. This (skeletal) tune possibly tells more about melodic tension than the busy and elaborate sixteenth-note motion.

The harmonic background of the subject is ambiguous: Bach harmonizes each subject entry slightly differently. The first bar may come as a I-V-I curve or, particularly towards the end of the fugue, begin on the subdominant and resolve into the major tonic. In the second bar, one usually finds a complete cadence with the subdominant (or its representative) on the downbeat, followed by the dominant-seventh and the return to the tonic on beats 2 and 3 respectively; in other entries, the dominant may already be reached at the beginning of the subject's second bar. The following example only mentions two of the various possibilities.

(ex. 42)


The dynamic outline of the subject follows from the above observations. The tension grows from the initial keynote through D#, onwards through the low G# which builds the basis for the large spanning interval, and up to F#. This F# on bar 2d, reached thus in a one-bar crescendo, captures all possible tension-enhancing features. Harmonically, it represents the active step to the subdominant; melodically, it is the target of the high-tension seventh interval (in the "simplified" tune), or the summit of a powerful thrust upwards expressed in the ascending scale. The climax is then followed by a gradual relaxation, through the 4-3 step downward in the simplified tune. (Great care should be taken not to play "Romantic waves" – dynamic curves which follow the ups and downs of the pitch instead of expressing a clear sense of purpose. Thus G#, although it is the lowest sound, is dynamically powerful as it propels the run up to the climax.)


II/4.2.2 The statements of the subject

The subject appears sixteen times in the course of the fugue.


bars 1/2



bars 28/29



bars 2-4



bars 30/31



bars 5/6



bars 48/49



bars 16/17



bars 53/54



bars 17-19



bars 55/56



bars 20/21



bars 61/62



bars 24/25



bars 66/67



bars 26/27



bars 67-69



(ex. 43)

Of the sixteen subject statements, four are inverted, but none appears in stretto or parallel. The subject receives a real answer without interval adjustment.

Modifications occur only rarely; they exclusively affect tonality, not shape, and materialize only towards the end of the fugue. In bar 53, the thematic fifth interval is diminished to a fourth, and the two notes after the downbeat of bar 54 are shifted one tone higher. In bar 55, Bach introduces an additional accidental on the note before the climax, and in bar 62, he raises the final note to the major third. The most drastic effect among these small alterations is reached in bars 67/68 where the beginning of the subject answer sounds for once not in G# major but on the fifth degree of C# minor. The fifth jump is altered significantly into a diminished fifth. As this subject entry also picks up alterations that had been heard separately before (the raising of the note before the climax and of the final note; see B# bar 68/69), this final entry of the fugue sounds harmonically somewhat erratic.


II/4.2.3 The counter-subjects

The only counter-subject that Bach invents for this fugue holds two surprises: It takes several entries and different attempts until the counter-subject reaches what can later be recognized as its final shape. Moreover, there are three statements of the counter- subject in the long subject-free passage in the middle of the fugue; they behave like entries in a section and have thus led some analysts to the assumption that this is a double fugue. While we can appreciate that the three statements of the counter-subject in bars 35-39 create this effect, we must also admit that this thematic phrase was presented before – it is not "exposed" here, nor anywhere else, for that matter.

Let us take the most frequently occurring version as the "final" shape (see bars 30/31:L, 35/36: U, 48/49: L, 55/56: U, 61/62: L, 66/67: U, 68/69: M). In this version, the counter-subject describes a clear diminuendo throughout its phrase. The first four notes are perceived as melodic, while the final jumps give a strong impression of a cadential-bass pattern. Comparing the other statements of the subject against it, one finds the following variations.


The first statement of CS (see bars 2-4: L) enters late, thus shortening the initial note. This note is tied and thus launches a diatonic descent, instead of the chromatic one characteristic of the final version. The high pitch Fx – needed here to complement the harmony – replaces the expected D#.


The second and third statements of CS (see bars 17-19: M, 20/21: U) already come close to the final version. The former still omits the chromatic- ism by suspending the initial note, while the latter introduces the chromatic descent but features a variation (three-note group instead of jump) towards the end.


The sixth and seventh statements (bars 36/37: M, 37-39: L) begin a quarter-note late; in addition, the former entry omits the final note.


Due to the harmonic modifications in the final subject entry, the partnering entry of CS (bars 68/69: L) begins and ends half a bar late.

The contrapuntal dynamics of subject and counter- subject are very simple, as shown here in ex. 44:



II/4.2.4 The episodes

The fugue contains thirteen subject-free passages. Some of them are very short, only filling up the half-bar between the end of one subject entry and the beginning ensuing statement commences on the next downbeat. These filler-episodes which join what are conceived as consecutive entries will be indicated below with an asterisk.


bar 4


bar 25*


bars 49m-52


bars 6m-15


bar 27*


bar 54*


bar 19


bar 29*


bars 56m-60


bars 21m-23


bars 31m-47


bars 62m-65


bars 69-71

Within these episodes, sequences of the subject's ending play a major role; they will be referred to as Ms (motive derived from the subject). Several other components recur frequently and therefore deserve to be pointed out. One is the conventional melodic closing-formula consisting of syncopated keynote, leading-note and downbeat keynote (see e.g. L: bars 4/5) which in this fugue is frequently used outside cadential conclusions. For convenience, this formula shall be referred to as "the close".

Two independent motives enrich the subject-free passages. M1 can be observed in U: bars 10m-12m. It consists of a three-note descent jumping up a fourth to two descending syncopations, followed by another three-note descent and two more descending long notes (now not syncopated). M2 is introduced in U: bars 13/14 (from B to B – a convex curve characterized by a syncopation before a weak beat), and M3 in U: bars 44/45 (from E# to A – another convex curve, this time with the climax on the weak beat, followed by a fifth jump before the relaxing stepwise descent).

The very short episodes (see table a) and the two very long ones (shown in table b) are built in symmetry.

(The descriptions "descending" and "ascending" refer to the connection of the motive with the previous subject ending.)

Table a

E1 = Ms 2x descending in U
close in L close in U,

E3 = Ms 2x descending in M
cadential steps in L

E5 = Ms ascending in U
ascending sequence in L

E6 = Ms ascending in M
ascending sequence in U

E7 = Ms ascending in L
ascending sequence in M

E10 = Ms ascending in M

The role each of these short episodes plays in the dynamic design of the fugue follows from the direction of their sequences. Thus E1 and E3 are decreasing, but E5, E6, E7 and E10 are increasing. All of them link consecutive subject statements.

Table b






Ms (2 x, descending) in M
syncopations in U
jumping dotted notes in L



Ms (2 x, descending) in M
syncopations in U
jumping dotted notes in L
M3 in U, ascending dotted notes in
L+M/L+U, CS in U, M, L



Ms (3 x, descending) in L
imit. 2x in U, 1x in L, 1x in M
close + M2 in M
M1 imitated 1x in U, 1x in M
E major cadence



Ms (2 x, descending) in L
imit. 2x in U, 1x in L, 1x in M
close + M2 in M
M1 imitated 1x in U, 1x in L
E major cadence


M2 in U
ascending dotted notes in M/U
fake subject entry in L/M


M2/M3 in U, imitated in M
M3 (3 x, ascending) in U
imitated (2 x) in M

The dynamic layout of these two long episodes is more complex. The peak notes in the upper voice feature two conspicuous descents, each marked at its end with a close and a cadence (see e.g. in E2 bars 6-9: A-G#-F#-E-D# in U, close in M, cadence in C# minor; bars 10-13: F#-E-D#-C#-B-A-G# in U, close in M, cadence in E major). In these segments, the tension is obviously diminishing. In the final segments, however, the direction of the sequences is ascending and, with rising tension, prepares the following subject entry.

As for the remaining episodes:


E4 is equally divided: bars 21-22m diminish with descending sequences in all voices, while bars 22m-23 increase again.


E9 is simple in structure, with only decreasing tendency.


E11 sets off with a cadential close in C# minor (bars 56/57), overlapping with the beginning of an Ms stretto and a dynamic curve formed by ascending M2 sequences followed by cadential resolution; all this occurs over a dominant pedal.


E12 too is two-fold in its dynamic shape, featuring a decrease (until bar 64m) followed by an increase which prepares the next pair of entries.

II/4.2.5 Character, tempo, articulation, ornament realization

The ornamental nature of the subject clearly suggests a rather lively character. This concept is supported both by the consecutive jumps which mark the final shape of the counter-subject, and by the gigue-like rhythm that accompanies the inverted subject entries in bars 24-29. The tempo can be fairly swift; its upper limit depends on the individual performer who must still be able to clearly shape all motives.

The relative tempo of prelude to fugue can be a simple one as the change from nine-eight to twelve-sixteen compound time (representing a change from triple to double beat) and the shift from very calm character to the boisterous runs of the fugue speaks for itself.

a dotted quarter-note

corresponds with

a dotted quarter-note

in the prelude

in the fugue

(Approximate metronome settings: 60 for the dotted quarter-note in both prelude and fugue.)

The appropriate articulation entails non legato in the dotted eighth-notes and gigue-patterns (eighth-note + sixteenth-note) and legato in the sixteenth-note figures. It is possible – but rather demanding – to distinguish a tighter legato in the inverted-mordent figures from a crisper quasi legato in all other sixteenth-notes. The chromatic segments in the counter-subject can be rendered in either legato or gentle non legato. Exceptions from the non legato touch in the longer note values occur above all in the keynote / leading-note / keynote closes. (Some of these melodic-closing formulas might be overlooked in the drive of the piece, so here is a list: L: bars 4/5, 5/6; M: 8/9, 12/13, 39/40, 43/44, 48/49; U: bars 19/20, 32/33, 53/54, 56/57, 70/71.)

The score indicates four ornaments. All are note-filling trills shaking in values twice as fast as the shortest notes in the score (i.e. in thirty-second-notes) and resolving after a suffix onto the target note.


The first trill (bar 26) is approached by step; it thus commences on the main note and contains only one longer and four shorter notes.


The second trill appears in the context of an imperfect cadence (see bars 32/33). It begins on the upper auxiliary B and encompasses twelve notes (including the suffix) before resolving onto G#.


The other two ornaments appear in consecutive bars (bars 60, 61); both are weak-beat trills on Fx followed by a middle-beat resolution onto G#. This resemblance makes them appear as an imitation, although the melodic context in the two voices is otherwise different. A possible interpretation is that the ornament in bar 60 (which, as the brackets indicate, was added later) helps, through its relationship with the next trill, to tie the subject statement in bars 61/62 to the previous section. Both ornaments should be rendered as note-filling trills; both begin on the upper neighbor note and resolve with a suffix from the melodic minor scale (G#-Fx-G#-Fx-E#-Fx-G#).


A last comment concerns the final chord. It is surprisingly short: only a dotted eighth-note instead of the expected (and, out of negligence, often played) dotted half-note of full-bar length. This rather unceremonial ending does not invite a pronounced ritardando. A mere hint of a relaxation would be most adequate.


II/4.2.6 The design of the fugue

Determining the design of this fugue is not too easy. In the absence of obvious cadential closes and reductions in the texture, interpreters and analysts have to rely on the logic brought forth by the order of the entries and on symmetries.

The striking analogy of episodes E2 and E8 relates bars 1-15 to bars 24-47. The latter portion appears as an enlarged modification of the former: Both sections commence with three connected subject statements (bars 1-6, 24-29), and both end with ascending motives and scale portions (bars 13-15: M2 + ascents in M, U; bars 44-47: M2 + ascents in M, L + M3).

Between these two sections lies a shorter second section which, for once, begins with one voice resting. More convincing even is the direct symmetry of the beginning of the first and second sections: subject on tonic (compare bars 1/2 with bars 16/17), answer on dominant (compare bars 2/3 with bars 17/18), one-bar episode with identical notes (compare bar 4 U with bar 19 M and bar 4 L with bar 19 U). The second section also features three entries (the third one appears on the relative major) and a concluding episode which, at its very end after an initial descent, musters new energy to prepare the advent of the next section (see the ascending lines in bars 22/23).

The fourth section begins in bar 48 with the return to the tonic. Its three initial subject statements all seem, at first glance, to be rooted in the tonic. Yet their individual surroundings define their harmonic background quite differently and reveal that the return to the tonic occurs only very gradually (see bars 48/49: F# minor to C# minor; bars 53/54: A major beginning, unresolved ending; bars 55/56: C# minor).

The longer episode E11 with its ascending sequences of M2 seems to announce the conclusion of this section (see bar 59), but the above-mentioned trill imitation leads the listener to include a further entry, the answer in bars 61/62, into this round. Once more an episode ends with rising tension (see bars 64/65) to prepare the forthcoming section.

The short final section embarks once more in F# minor, thus picking up the beginning of the fourth section. Its two subject statements appear in the keys and metric distance of the initial entries in the first and second sections of the fugue, and even the ensuing episode-bar is very similar (compare bars 66-69 with bars 1-4 and 16-19). The fifth section then concludes with a final cadential ending in E13.

The structural balance achieved by the composer in this fugue is worth noticing.


The five sections group into three blocks:
sections I + II, section III, sections IV + V.


Both sections I and II and sections IV and V feature very closely related beginnings.
In both cases the shorter follow-up sections seem to fulfill the harmonic purpose left open by the larger sections – in section II the modulation away from the tonic, in section V the firm re-establishment of the tonic.


The third section provides some contrast in its newly introduced subject inversions, gigue-rhythm and the independent "round of entries" of the counter-subject.


The three-part layout is further supported by the length of the three portions:

sections I + II = 23 bars

section III = 24 bars

sections IV + V = 23 1/4 bars

For a sketch showing the design of this fugue, see ex. 45.



II/4.2.7 The overall dynamic outline of the fugue

This fugue is very playful in character; thus dynamic increases and decreases are slight rather than dramatic. Most influential in the overall design are the extensive episodes. Their dynamic impact, based on descending or ascending sequences or peak-note lines, has been mentioned above.

What remains to be decided is the relationship between consecutive subject statements. Within the three entries which open the first and second sections, the tension rises gradually, supported both by the increase in the number of voices and, in the third entry of section II, by the modulation to the major key. Tension is diminished, however, by the relaxing attitude of the linking episodes (E1, E3). Section III, by contrast, sets out with a much lighter character. This is softened even further in the two-part statement (bars 26/27) but recovers a little in the following lower-voice entry. The fourth statement returns to the uninverted shape of the subject; nevertheless, since it appears in the weak middle-voice position, it reaches only moderate intensity.

The fourth section returns to the original touch and color but fails to create similarly persuasive dynamic groupings. Only the short final section with its compact design reaches the density of the two initial sections.