WTC I/15 in G major - Prelude

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

I/15.1.1 The prelude type

In the G major prelude, the overall picture is dominated by sixteenth-notes. They appear in triplet groups, with eight such groups to each 4/4 bar. (Bach's two-fold time signature in the first bar makes things look more complicated and treacherous than they really are since it is obviously not the treble part alone which features the triplets.)

These sixteenth-note triplets appear in three different grades of melodic quality, and these are indicators for the basic interpretation of the prelude. A large portion of the triplets displays broken chord patterns (see bars 1-3 and 6-8, as well as the bass part from bar 11 onward); here the prelude is harmonically determined. In other instances, the sixteenth-notes present a hidden two-part structure (see bars 4/5 and 9/10); melodically rather neutral peak note lines emerge here. One step further on, the sixteenth-notes appear in one-track melodic patterns (see treble part from bar 11 onwards); this is the realm where small motives might surface.

The prelude cannot therefore be described with a single general term. It sets out as a harmonically determined piece, but tentative melodic features soon arise and develop more and more into independent entities.


I/15.1.2 The overall design of the prelude

The first harmonic progression already concludes in the second half of bar 2. However, as the pattern continues uninterrupted, this cadence should not be regarded as a structural break. Instead, the end of the first section follows with the subsequent abrupt modulation to the dominant. As this concurs with a visible change of pattern from bar 4 onwards, one can safely assume a structural caesura here (see U: bar 3 before the final eighth-note D; L: end of the bar).

The second section sets out in bar 4. A first miniature cadence which already resolves in bar 5 does not interrupt the progression of the material. The next harmonic close confirms the dominant key of this section on the downbeat of bar 11, thus concluding the second section of the prelude.

The prelude consists altogether of four structural sections:

I bars 1-3 tonic confirmed and modulation to dominant
II bars 4-11d dominant
III bars 11-131 dominant
IV bars 132-19 return to the tonic

No larger segment of the piece recurs. Although the eye is caught by the obvious similarity between bar 5 and its transposition in bar 9, the position of both bars in the structural context is entirely different. Bar 5 is the second in a pair of hidden two-part structure bars and represents a harmonic resolution, while bar 9 is the first in the pair and harmonically active.


I/15.1.3 Practical considerations for performers

The basic character of this prelude is easy to determine as rather lively, due to, on the one hand, the obviously simple rhythmic pattern with only two note values and, on the other hand, the frequent broken chords (in the triplets) and leaps (in the eighth-notes). Like the fugue which it precedes, the prelude is downright virtuoso and should be played accordingly, in a very brisk pace. It does not contain any ornaments.

With regard to articulation one should differentiate the coarse level of simple contrast - non legato for the eighth-notes and legato for the sixteenth-notes - from the level of subtler shades. This includes:

non legato in a light, neutral touch for the accompanying eighth-note jumps
non legato in a more expressive touch for the eighth-notes in the hidden two-part structure bars, i.e. those in predominantly stepwise motion
quasi legato with a crisp touch for the sixteenth-notes in the broken chords
legato in a moderately expressive touch for the sixteenth-notes in the hidden two-part structure bars
legato with some melodic expressivity in the sixteenth-notes which form motivic figures
legato with high melodic intensity for the eighth-notes in U: bars 11 13 as they represent pairs of appoggiatura / resolution


I/15.1.4 What is happening in this prelude?

The G major prelude can be regarded as consisting of two larger sections, each of which embodies a shorter introductory portion followed by a longer one which develops and complements the material introduced before.

The first of these encompassing sections includes bars 1-11d. It is based on a harmonic idea introduced in the right hand with virtuoso broken chord patterns, accompanied by a pedal note with rhythmic grouping and octave displacement. Inside the first simple cadence (bars 1/2), the descending pitch direction of the broken-chord pattern is so dominant that it overruns all considerations for shaping along harmonic lines. Faithful to the character of this virtuoso figure, the piece thus begins with a rather energetic tone color for the first downbeat followed by a plain diminuendo through two bars. The third bar is composed as a harmonic sequence; it takes up the V7-I pair from bar 2 and relocates it to D major. The equivalent to this feature in performance is a sequence also in the dynamic process, i.e. a repetition of the second half of the previous diminuendo. The question whether the dynamic level in bar 3 is generally louder or softer than that in bar 2 can be solved with the help of three observations: (a) the pitch level in the second half of bar 3 is slightly lower than in the second half of bar 2, and descending sequences are usually played in diminishing line; (b) the second half of bar 3 features rests in the right hand part, and this thinner texture adds to the effect of decreasing intensity; (c) bar 3 was earlier recognized as the end of the first short section, and a definite relaxation serves to underscore this fact in performance.

A new pattern is presented from the last eighth-note beat of bar 3 onwards. Retaining the virtuoso quality of the beginning, bars 4/5 nevertheless expand this idea by changing the texture into that of a hidden two-part structure. In U: bar 4, a line with stepwise motion in eighth-note rhythm builds the melodic foreground, while the background consists of a pedal on D (the tonic of the newly established key) which is ornamented regularly by its leading note C#. In bar 5, the foreground line continues while the background figure is transferred to the left hand. (This should obviously be executed without any audible difference between the two bars.) The melodic part is further paralleled in the left hand of this bar.

Bars 6-8 return to the texture and material of the first bars but vary it slightly. Harmonically, bars 6/7 and 8/9 recall the earlier sequences with their dominant-seventh / tonic resolution. The chords appear here extended to whole-bar lengths and feature varied accompaniment figures. Very important for the understanding and performing of this section is that the harmonic functions in the leading bars of these pairs represent a sudden and very powerful increase in tension: the chord in bar 6 is vii7 of A (the dominant in D major); the chord in bar 8 is vii7 to E (step ii or the relative to the subdominant in D major).

Bars 9/10 return to the hidden two-part structure. The pedal note change between the two bars reveals the cadential steps: E-A which are steps ii-V of D major. The melodic parts of both voices set out in parallels and only separate in the second half of bar 10.

The dynamic equivalent to the processes described above for the first half of the prelude could be rendered somewhat like this:

bars 1 2 3 4 5 6 7 8 9 10 11
  f + mf + mp+ p+ mp - p mp mp+ poco f mf f - mp mp+ mp

The second encompassing section, covering the remainder of the piece (see bars 11 19), is based on motivic figures in the treble part which introduce a new expressive quality. Already in the first little motive (see bar 11 beats 2/3, with three sequences until the E D on bar 13 beat 1), the melodically conceived sixteenth-notes followed by the appoggiatura-resolution pair in the eighth-notes add charm and grace to the prelude. (Note the dynamic shaping with U: crescendo to the appoggiatura, L: diminuendo from each strong beat onwards. Well intentioned imitations would thus be ill placed here.)

The second motive can be seen as a development of the first. It is twice as long, more elaborate in its triplet figures, and the appoggiatura is now an indirect one (see bar 14: D E embedded in D E F E). This motive also incites sequences (see bars 142-152 and from there to bar 162.) The accompaniment is much more intricate here. In the first motive it had contained old material, combining the broken chord figure and the accompanying octave jump from bars 1-4. Now it sets out with the varied broken-chord figure from bar 6 and continues it freely, albeit in a very virtuoso way.

The section ends with a figure of only one quarter-note length (see U: bar 163), complemented in the left hand by a unit which, because of its rhythm, has to be regarded as overlapping (see bar 16 beats 3/4 C-A-F-D etc.). Both figures are sequenced many times, the treble figure right through to the end, whereas the bass figure gives way, in the bar 18, to a cadential-bass pattern.

The development of tension in this second larger section is determined entirely by the direction of the sequences. Thus the first motive (which commences in something like mf) is followed by a diminuendo in its descending sequences, the second motive (which sets out with more vigor anyway) experiences a crescendo in its ascending development, and the third little figure compensates this when - commencing with a climax of less melodic intensity than its predecessors - it describes a continuous relaxation up to the end of the prelude.


WTC I/15 in G major - Fugue

I/15.2.1 The subject

The subject of this fugue appears confusing with regard to its length. Asked where it ends one might find three possible answers; yet each of them is, for one reason or another, not entirely convincing.

Performers who regard the return to the keynote G (and with it, possibly, to the tonic harmony) on the sixth eighth-note of bar 4 as the end of the subject, must admit that this conclusion is metrically not too satisfactory.
For performers who decide to include the downbeat of bar 5, the subject closes with an imperfect cadence, a harmonically unsatisfactory solution.
Some performers interpret the subject as concluding on the downbeat of bar 4. In this case, the subject would appear both harmonically incomplete (i.e. without the subdominant which, substituted by its relative minor, materializes only in bar 4) and melodically unconvincing, with an end on the fifth. (Only in the inversion does this solution of ending the subject in fact sound very persuasive. Bach must have felt the same, since he states the inversion often in this shortened scope; see below ex. 24b).

With this complex answer and no simple solution at hand, it will prove necessary to give each subphrase a name tag and then state very clearly which of them appear in the statements of this fugue. Following either the first or the second of the three concepts mentioned above, three subphrases must be distinguished in this subject.

The first subphrase (a) consists of the ascending turn figures in bar 1 and includes the subsequent eighth-note descent and the syncopated seventh leap in bar 2.

The second subphrase (b) is confined to bar 3 and contains a varied partial sequence of the first subphrase: the eighth-note descent and the syncopated seventh leap recur, but they are linked in a smaller interval than they were before.

The third subphrase (c) then consists of the downbeat eighth-note in bar 4 and the two fold scalar descent (with or without the note in bar 5d).

If one follows the third of the above-mentioned concepts and assumes the closure of the subject already on the downbeat of bar 4, there would only be two subphrases, (a) and (b), with (b) complemented by a resolution note.

The pitch pattern consists primarily of stepwise motion, with only three leaps (two in bar 2, one in bar 3). Two of the leaps appear as consecutive intervals (see bar 2: G-D-C). Among the seconds, the longer note values (see the eighth-notes in bars 2/3) have melodic value while the shorter are either ornamental (see the two turns at the beginning, from A to A in bar 1 and from B to B in bars 1/2) or constitute scale passages (see the two five note groups at the end of the subject, from E to A and from C to F# in bars 4/5).

This brings us to rhythm. The subject features three values: sixteenth-notes, eighth-notes and (syncopated) quarter-notes. The same three note values constitute also the predominant rhythmic pattern throughout the entire composition. A particular attribute of the rhythmic pattern in this subject is that it displays (within the confines of bars 1 4) perfect symmetry. If one were to give a name tag to the rhythmic patterns of each bar, the result would read "x-y-y-x".

When analyzing the harmonic background of the subject it seems worth while anticipating an important trait of this fugue; i.e. that the subject is frequently used in inversion and that this inversion follows a harmonic outline which is considerably different from that underlying the original shape. The following examples demonstrate this. (Harmonizations are taken from bars 11-15 and bars 20-24 respectively.)

(ex. 24a)


(ex. 24b)

In our search for the intended climax we may, as was demonstrated above, not be able to rely on harmonic features. However, both the rhythmic pattern and the pitch outline provide congruous guidelines. Concerning the rhythm, the two syncopations obviously capture special tension. As both of them coincide with high tension leaps, there can be no doubt that the quarter-note C in bar 2 represents the climax of the first and the quarter-note E in bar 3 that of the second subphrase. Between the two, the second is stronger than the first both because it is composed in an ascending sequence and (in case another reason should be needed) because it is harmonized in a dominant ninth chord, as opposed to the simpler dominant seventh chord in bar 2. After these two energetic climaxes, the third subphrase appears as little more than an after thought; this may well be one of the reasons why the doubt whether or not it "belongs" survives. The dynamic tendency follows the descending motion; thus the D at the outset of this subphrase is comparably louder than all that follows. (But calling it a climax seems, nevertheless, to misinterpret its ancillary function.)


I/15.2.2 The statements of the subject

This fugue contains 16 statements of the subject.

1. bars 1-5d U 9. bars 51-54d U*
2. bars 5-9d M 10. bars 52-54d L*-
3. bars 11-15d L 11. bars 60m-63 M*
4. bars 20-24d Minv 12. bars 61m-64d U*
5. bars 24-28d Uinv 13. bars 69m-73d Linv
6. bars 28-31d Linv 14. bars 77-80d Minv/orig
7. bars 38-42d U (15. bars 78/79 Linv**)
8. bars 43-47d Minv 16. bars 79-82 U -

(ex. 25)

Several of them are incomplete in one way or another. Only one is a genuine "false start"; in the table above it appears in brackets. Those among the entries which are shortened at the end have been marked here with a hyphen ("-"); those which are shortened in the middle carry an asterisk (" * ") symbol. It must be stressed that in this fugue, these abbreviations of the subject are not arbitrary but have particular reasons:

The attenuation of the third subphrase arises from the fact that this is perceived more like an afterthought, and that it leads to a harmonically imperfect ending which, particularly after the perfect cadence included in the inversion, makes it appear somewhat superfluous.

The compression of the first and second subphrases into one occurs only in stretto positions in order to avoid the octave parallels which would otherwise result. This abbreviation uses in the seventh leap the pitches of bar 3 but incorporates them already into bar 2.

The reductions of the subject aside, there are not many other modifications. In two instances (including the "false start"), the beginning is delayed by one sixteenth-note and the subject thus appears in slightly altered rhythm (see bars 51 and 78). On another occasion, a statement sets out as an inversion but later seems to be "convinced" (or dragged along) by the entering stretto partner to change its ending into that fitting the original shape (see bars 77-80). Finally, one inverted subject finds its closing note displaced an octave lower (see bars 72/73), and another features an ornamental variation of the third subphrase (see bar 46).

There are three stretto passages in this fugue, occurring respectively in bars 51-54, bars 60-63 and bars 77-82. Parallel statements are not used. However, a short but very effective three fold parallel (two voices in double thirds, another in the same rhythm but contrary motion) is created in the final stretto. Here the "false entry" is continued in sequences, and the middle voice statement which sets out in inversion is "persuaded" by the upper voice entry to transform its last subphrase into the turn-figures of the subject beginning.


I/15.2.3 The counter-subjects

This fugue contains only one counter-subject. Its rhythmic structure allows division into three distinct segments, yet these structural components do not function as separate subphrases but only as consecutive portions in the pursuit of a single purpose.

The scope of the counter-subject is slightly blurred at its beginning, as was that of the subject at its end. Thus the F# in U: bar 5 can be regarded as the first note of the counter-subject (if the subject is understood as ending on the last eighth-note of bar 4) or as the point of overlapping phrases, belonging at the same time to both elements. Finally it is also possible to perceive the counter-subject as beginning on the second eighth-note of the bar.

Taking a closer look at the segments one finds that, faithful to the requirements of independence between contrapuntal "partners", their confines do not coincide with those of the subject subphrases. The first segment consists only of eighth-notes and ends on the downbeat of bar 6. It shares with the beginning of the subject the pattern of half bar model and ascending sequence; therefore its independence is a limited one. In this analysis it will be referred to as (d).

The second segment within the counter-subject is two bars long. Moving exclusively in sixteenth-notes, it describes a continuation of the previous ascent, by relying on an ornamental figure which is sequenced at half bar distances. This substantial portion of the counter-subject will here be named (e). Finally, the third segment (f) is scarcely one bar long. It begins with a scalar descent which sets out as a parallel to the subject but then bends backwards and reaches the high G, thus disclosing its nature as yet another section under the overall purpose of a peak note ascent. From the beginning of the counter-subject until this target note the underlying ascent thus progresses through an entire octave. This directedness distinguishes it sharply from the subject which, in its concluding portion, returns to the level from which it was launched.

Having observed all this, the choice of an appropriate dynamic outline is as easy as could be since the tension seems so obviously to be rising throughout the entire length. The opposite, a continuous diminuendo, will then apply in the inversion of the counter-subject.

The following sketch shows phrase structure and dynamic design in the primary material of this fugue (ex. 26):


I/15.2.4 The episodes

There are eight subject-freepassages within this fugue.


bars 9-10


bars 47-50


bars 15-19


bars 54-60m


bars 31-37


bars 64-69m


bar 42


bars 73 – 76

In addition, the five bars which follow the last subject statement (E9) present the cadence which re establishes the tonic (bars 82 83d) and a coda on a tonic pedal (bars 83 86).

The episodes are almost exclusively made up of three independent motives. While the first two motives are related (for details see below) and of only moderate melodic intensity, the third motive introduces a truly lyrical quality and is thus the most conspicuous among the three.


is first presented in U: bars 9 10d. Its rhythm is characterized by uninterrupted sixteenth-notes which begin immediately after the first beat of one bar and end on the downbeat of the next. The pitch pattern is conceived in hidden two-part structure: one of the parts remains, as an indirect pedal, on a repeated note; the other part describes a scalar motion. (Note that this melodic part, while moving basically in regular eighth-notes, begins with a sixteenth-note note of upbeat character. This note is mistakenly too often played as part of the background.) This scale points downwards in the original appearance of bar 9 but is just as often used in ascending direction (see e.g. M: bars 10-11d); those cases will be referred to as M1 inversions. The concluding note on the following downbeat serves as a harmonic resolution. It is interesting to see that in the original version, the resolution appears on the pitch level of the indirect pedal (see bars 9/10), whereas in the inversion it sounds as a continuation of the melody (see bars 10/11).

M2 is introduced as a companion to M1, with the original in M: bars 9-10d and the inversion in U: bars 10-11d. As a companion it is by no means polyphonically independent but composed as a parallel to the melodic part of M1. Consequently its rhythm moves in regular eighth-notes.

emerges only in the second episode (see U: bar 17 and sequences). Unlike the two preceding episode motives it is metrically built from the downbeat onwards, ending somewhat indistinctly before the next bar line or on the ensuing downbeat in that case overlapping with the beginning of the sequence. Its lyrical quality arises both from the melodic content of the sixteenth-note line and from the expressive syncopation on the second eighth-note of the bar.

As to the dynamic outline of the three motives, both M1 and M2 represent simple lines which follow the pitch direction: crescendo in all ascending and diminuendo in all descending scalar motions. The distribution of these motives in the eight episodes of the fugue is very dense. Based on the use of this material it is possible to distinguish three episode types:


are all built on M1 and M2. The shortest episodes, E4 (one bar long) and E1 (two bars long), consist entirely of the juxtaposed first and second motives. E6 presents a less regular picture. After a first bar with the inverted M1 in parallel motion and a second bar also featuring parallels (in a figure which does not have any importance outside this episode), the remaining four bars are all dominated by the inverted M1 which enters almost unaccompanied, is then joined by M2 and later met by sixteenth-note figures in two fold contrary motion. Finally, the coda recalls M1, quoting it in only half its length and with that variation which already appeared from E5 onwards but which now grows even more elaborate.

E3,E5,E7,E8 all neglect M2 in favor of the other two motives. E5 and E7 each set out with one bar of M1 inversion in non motivic surroundings, followed by three or four bars respectively with M3/M1 and their sequences. E3 seems to go the reverse way, commencing with three bars of M3/M1 followed by four bars in which M1 or its inversion sound against whole bar scales. Finally, E8 is only left with these four bars in which the M1 inversion is contrasted with whole bar scales.

is related to all of the others. It sets out with M1 in non motivic surrounding; the following bar combines an M1 inversion with an M2 parallel, and the remaining three bars present sequences of the M3/M1 pair, complemented by a neutral middle voice.

As is obvious from this table, none of the episodes exhausts its function as a cadential close. In one instance, however, a closing formula determines the material of a final segment: see the do-si-do figure in the upper voice as well as the typical bass pattern in the first half of bar 69.

The relationship between episodes is evident from the above table. E1 is related to E4 and also, though less closely, to E6; E5 and E7 seem to be similar, and E8 shows analogy to the second segment of E3. E2 can be divided to reveal several relationships in its first two bars to E1 and E4, in its remainder to E5 and E7.

The role which each of the episodes plays in the development of the composition also stems directly from its material. Those of the first type, based only on M1/M2, act as bridges, while those which end with descending M3/M1 sequences clearly have concluding character. The two episodes which end with the M1 + scale combinations lie in-between; they provide more contrast to the primary material than the first type but at the same time lack any features of conclusive force one could say that they insinuate a gap which they nevertheless span.


I/15.2.5 Character, tempo, articulation, ornament realization

The basic character of this fugue is best interpreted as rather lively. This decision is supported mainly by the pitch pattern with its jumps and ornamental figures. The rhythmic pattern contains four melodically relevant note values - three already in the subject, and additional thirty-second-notes in the variation of M1 (see from bar 47 onwards), but all these values fit smoothly into a generally simple rhythmic structure.

The tempo of the G major fugue is swift, playful if not outright virtuoso, but allowing for an energetic and non superficial touch. The articulation demands non legato for the eighth-notes and quarter-notes, and different kinds of legato for the sixteenth-notes and the few thirty-second-notes.

As to these different kinds of legato, truly melodic quality seems appropriate for M3, and an equally dense sound, though for different reasons, suits the hidden two-part structure of M1. Those sixteenth-notes which are ornamental, like the turn figures and scale portions in the subject and the longer scales in episodes 3 and 8, are best rendered in quasi legato.

An ideal proportion of tempo between the prelude and the fugue is found in the larger pulse units of each:

one quarter-note
corresponds with
half a bar

in the prelude

in the fugue

Approximate metronome settings: 80 for both the beats (quarter-notes) in the prelude
and the compound beats (dotted quarter-notes) in the fugue.

The fugue contains a number of ornaments, namely in bars 22, 25/26, 64, 69, 78. Before pondering on any single one of them, the general question of the tempo of ornaments in this fugue must be settled. If they are to shake "twice as fast as the shorter note values", the dispute arises as to which of the values, sixteenth-notes or thirty-second-notes, are to be regarded as "the shorter" rhythmic units. In accordance with both the virtuosity of the piece and the ornamental character already found in some of the sixteenth-notes, it seems preferable to regard these as the values to be doubled in the trills. The thirty-second-notes will then appear as additional written-out slides and turns.

The ornaments listed above can be grouped as follows:

The mordent symbols in both bars 22 and 78 designate long trills, as the spelled-out suffixes indicate. The first is approached stepwise, therefore commences on the main note and allows for five trill notes (one sixteenth-note, four thirty-second-notes) before the two printed suffix notes. The second begins on the upper neighbor note and moves in six regular thirty-second-notes before the suffix.

The two trills in the inverted subject statement (bars 25 and 26) pose certain problems. The notes they ornament find, in regard to both pitch and metric position, possible resolutions; however, these supposed resolution notes do not belong to the same subphrases! The conclusions drawn from this predicament by performers (analysts, to my knowledge, refrain from taking positions on this case) include three options: (a) to ignore the original phrasing and link the trills, with a proper suffix, to their subsequent notes; (b) to ignore the trill symbol and play the subject as it was introduced, i.e. unornamented; (c) to respect both ornament and phrasing by playing a short, mordent like embellishment (with five notes).

The two trills in bars 64 and 69 share the same ending: they do not resolve on any appropriate strong beat but terminate in a tied note. Both therefore conclude without a suffix and stop before the beginning of the tie. When exactly each of them halts requires particular consideration. Convention suggests that the halt be as late as possible before the last bar line; however, it is unlikely that Bach had this in mind as in the second case it is technically altogether impossible to continue the trill motion in the second half of bar 70 where the right hand has to take care of the counter-subject notes. A feasible as well as musically convincing solution is to play both trills equally long. The one in bar 63, covering slightly less than a bar, would give the example for the one in bars 69/70 which would then come to a halt before the middle voice enters. (By the way, these interrupted trills do not, despite their beginning on the main note, prolong the first note as they are not "note filling" ornaments.)

The following examples show these two interrupted trills (ex. 27):


I/15.2.6 The design of the fugue

There are several indicators which are very helpful for the determination of the structure.

The order and shape of the subject entries allow definition of some of the boundaries between sections. It seems reasonable to assume (a) that the first three statements belong to one section, (b) that consequently the three statements of the subject inversion which follow establish the second section, and (c) that the appearance of the first stretto marks the beginning of a new section.

Two subject statements appear in reduced number of voices, namely those which begin in bars 38 and 51. These are therefore likely to constitute section openings.
The cadential closing formula in the first half of bar 69 also indicates the end of a section.

The role played by each of the episodes (as recognized above) sheds additional light on the question of structure. These facts confirm what has already been stated:

* the bridging E1 connects the second and third statement of section I;
* the concluding E5 rounds this first section off on bar 20d;
* E3, "spanning a larger gap", connects the three inverted subject entries of section II with the beginning of section III, marked by the entry in reduced ensemble (bar 38). Sections II and III, while clearly distinct for several of the reasons stated above, thus build a group due to the episode which links them;
* the bridging E4 connects the two statements - one original and one inverted - of the third section;
* the concluding E2 rounds this third section off on bar 51d, where the fourth section begins with another entry in reduced ensemble;
* the bridging E6 connects the two strettos of the fourth section;
* the concluding E7 rounds this fourth section off on the middle beat of bar 69, additionally enhanced by the closing formula;
* E8, "spanning a larger gap", connects the single entry which opens section V with the following stretto. The analogy of this episode with E3 suggests that the fifth section is to be regarded as structurally corresponding to the joint second and third sections which, as shown above, build a group.

As this detailed listing reveals, structural correspondences in this fugue are complex. The first two sections contain an obvious analogy in the number of their entries and the consistency in which these are presented. (The impression is that of a double exposition: first of the subject in its original shape and then of the inversion. Another "exposition" occurs in the fourth section which introduces the strettos.) The second large scale analogy is that between the second and third sections on the one hand and the fifth section on the other.

The harmonic outline of the composition is very clear. Both the first and second sections remain on the tonic, the third section is in E minor (the relative to the tonic), the fourth section in D major (the dominant), and the fifth section returns to G major which is approached from an entry on the dominant and crowned by the unusual statement on the third of the home key (see U: bars 79 82).

For a sketch showing the design of the fugue in G major, see ex. 28.


I/15.2.7 The development of tension

In each of the five sections, the tension rises from the first to the last entry; this is due either to the increased number of voices (see sections I, III and IV), to the enhanced presentation of the subject (section V, from single entry to stretto) or, in the second section, only to the effect of repetition and the analogy with the first section.

Overall, the second section appears dynamically increased owing to the fact that its three statements are launched in full ensemble and follow one another without interruption; the third section falls back both because of its reduced range and its minor mode, appearing almost as a softer appendix to the third section. The fourth section takes on yet a higher level of tension. Finally, the fifth section inverts the situation in the pair to which it showed analogy (see above: sections II/III correspond to section V) by setting out from the softer level of the single entry and traversing a powerful increase towards the "stretto with parallel tail".