II/11.1.1 The prelude type

The prelude in F major is a meditative piece, determined by metric features. This is expressed in many details, particularly in the continuous eighth-note motion which is not once interrupted during the entire piece, the equally continuous presence of half-note beats in at least one of the complementary voices, and the abundant use of ties which collect notes from among the eighth-note motion into chords.

There are up to five voices, but the part-writing is not entirely consistent. Although the main melodic motive changes between the player's two hands, it can hardly ever be allocated to any one voice. It might even be preferable, in this composition, not to talk about voices or parts at all, but to perceive the texture as eighth-note lines embedded in four- to five-part homophonic chords.

The initial phrase of the prelude could thus be represented as follows (ex. 59):

 

II/11.1.2 The overall design of the prelude

The initial harmonic progression concludes in the middle of bar 8. Here the tonic is regained by gradual resolution of one voice after the other, in a metrically weak position on the fourth quarter-note. This is a structurally relevant cadence. The next eight bars modulate to the dominant key C major. The striking correspondence of bars 17 and 1 suggests a "new beginning on the dominant" and thus reveals a larger structural grouping of two harmonic progressions into one section.

Altogether the F major prelude contains four such compound sections:

I	bars 1-16		Ia	bars 1-8		tonic confirmed
			Ib	bars 9-16		modulation to V (dominant)
II	bars 17-32		IIa	bars 17-24		dominant confirmed
			IIb	bars 25-32		modulation to iii (dominant relative)
III	bars 33-56		IIIa	bars 33-40		modulation to vi (tonic relative)
			IIIb	bars 41-46		modulation to V (dominant)
			IIIc	bars 47-56		modulation to iii (dominant relative)
IV	bars 57-72		IVa	bars 57-64		modulation to IV (subdominant)
			IVb			bars 65-72 return to the tonic

There are several very extended analogies:

Shorter correspondences are composed as sequences of the entire ensemble:

II/11.1.3 Practical considerations for performers

The meditative character of this prelude requires a calm tempo and overall legato playing. As the slurs in bar 1 indicate, the eighth-notes must not be perceived as active notes but grouped in such a way that only the metrically relevant ones carry some weight.

Dynamic shaping can easily be misinterpreted in this piece. The issue lies not in melodic events – in whatever voice – but in the interplay of extended harmonies woven into arabesque-like threads. Note-to-note increases or decreases would therefore gravely mislead the listener. Very gentle curves are built on two levels: by the metrically emphasized notes within the eighth-note-threads, and by the harmonic development within each phrase.

The prelude contains no ornaments except a single grace-note in bar 66. Its appearance in brackets indicates that Bach may not have notated it in his original manuscript. It adds, however, greatly to the smoothness and beauty of the bar. As the grace-note represents an appoggiatura, it must be played for half of the main note value (i.e. appoggiatura B^b = half-note, resolution A = half-note).

II/11.1.4 What is happening in this prelude

The weaving eighth-note lines in this prelude build mainly three patterns. All commence on the first eighth-note after a strong beat with a three-eighth-note ascent which serves as an upbeat to the following strong beat. Furthermore, all patterns are made up of little four-note groups and include a varying number of sequences. The metrically strong final notes of each four-note group may or may not be tied over to form sustained chords.

We will refer to these patterns as P1, P2 and P3 to remind us constantly that we are not dealing with melodic motives.

P1	is introduced in bars 1/2. The upbeat-group is an inverted turn, followed by four turn-figures. At the end of bar 2, a renewed upbeat seems to trigger an imitation which, however, sounds much more like a continuation of the same pattern. The metrically enhanced notes throughout bars 1-3 are: F, D, Bb, G, E; C, A, F and C. Two further imitations of the turn-figure in higher register add the metric accents G and E. Dynamically, P1 represents a protracted tension decrease.
P2	is first presented in bars 5-7d. It consists of three inverted-turn figures in ascending sequences complemented by two turn-figures in descending sequence. (The third of the turn-figures is extended and slightly varied: see instead of a simple A-Bb-C-Bb in bars 5/6 A-Bb-C----A-Bb). The metrically enhanced notes of P2 build this curve: D, F, (C)- Bb, G, E. The pattern is immediately sequenced one note lower (see bars 7-9d). The curve modeled by the clue notes is expressed also on the dynamic level: P2 is expressed in a one-bar crescendo followed by a one-bar decrescendo.
P3	is first heard in the cadential close of the first section (see the upper right-hand part in bars 143-16d). It is characteristic only at its beginning, since it commences not with a written-out ornament but with a straight four-note ascent. The immediate imitation can be distinctly perceived here as a different level in the polyphonic structure, unlike the imitations in P1 and P2 which were integral parts within a single complementary line. The relaxation which follows the double ascent abides by no particularly regular pattern. A variation of P3 occurs in section III of the prelude. The pattern establishes a model in which the initial ascent and its imitation are complemented by a triple turn (see bars 41-43d). This variant will be referred to as P3a; it is followed by two sequences. In both cases this pattern, like that of P2, builds a curve. The rise and the decline in P3 are both slightly steeper than in P2.

The three above-mentioned patterns build the entire prelude. The relevant features of the structural layout are listed below in table format, so as to allow easy comparison and overview.

bars		pattern		clue-notes		structure
Section I
1-4		P1		F-D-Bb-G-E-C-A-F-C + G-E		complete phrase
5-7d		P2		D-F-(C) Bb-G-E		two-bar model
7-9d		P2		C-E-(Bb)A-F-D		two-bar model sequenced
9-11d		P1		D-Bb-G-C-A-D		short version, phrase 1
11-143		P1		F-D-B-G-E-C-A-F-D-B + E		variation of phrase 1
143-16d		P3		C / A -- C / C		closing-formula
16-17d		P1		G-E-C		link
Section II
17-20		P1		C-A-F-D-B-G-E-C-G + D-B		complete phrase
21-23d		P2		A-C-(G)F-D-B		two-bar model
23-25d		P2		G-B-(F)E-C-A		two-bar model sequenced
25-27d		P1		A-F-D-G-E-A		short version, phrase 1
27-293		P1		D-Bb-G paralleled by F-D-G + l.h. C-A-D, Bb-G		development of phrase 1
29-322		P1		G-E-C#-A-F-D-Bb-G		variation of phrase 1
32-33d		P1		A-F		link
Section III
33-37		P1		D-Bb-G-E-C-A-F-D-B + D-G#		development of phrase 1
37-38d		P2		E-B-C#		development of pattern
38-40d		P3		A / E -- E / D		development of pattern
40-41d		P1		F-D-Bb		link
41-43d		P3a		Bb / G -- A-F-D		two-bar model
43-45d		P3a		D / Bb -- C-A-F		two-bar model sequenced
45-47d		P3a		F / D -- E-C-A		two-bar model sequenced
47-49d		P1		F-D-B-G#-E-C		development of pattern
49-51d		P1		E-C-A-F, F-D		free imitation bars 47/48
51-54d		P1		A-F-D, F-D-Bb, D-Bb-G#		partial sequences
54-56		P1		E-C-A-E, B-A-E-C#		cadential close + link
Section IV
57-60		P1		F-D-Bb-G-E-C-A-F-C + G-E		complete phrase, see 1-4
61-623		P2		D-F-(C) Bb-G		two-bar model, see 5/6
623-642		P3		Bb / Eb --D / Bb		free development
643-67d		P1		Eb-Eb, G-Eb-C, F-D-G		free development
67-702		P1		Bb-G-E-C-A-F-D-Bb-G-E		variation of phrase 1
703-72d		P3		F / D -- F / F + A-F		closing-formula

II/11.2.1 The subject

Before one considers the scope of the subject, the time signature of this fugue invites a comment. Bach's 6/16, like many other of his compound time signatures, does not really fulfill the task later generations expected of this information: it does not indicate that one should count six small beats in each bar, but is a way of notating triplet figures without the tedious bracketing they would normally require. Having understood this, one is advised to imagine what a conductor would beat. (A 2/4 bar, seems the only reasonable answer.) This should be kept in mind when considering questions of tempo, phrasing, small-scale dynamics etc..

The subject is simple only at first glance. It encompasses four bars: from the middle of bar 1 to bar 5. Here, the relevant return to the tonic actually occurs already on the downbeat. In many cases, however, the two ensuing sixteenth-notes have to be included as a passive extension – or as a completion of the "larger beat". In other cases, these two notes clearly form the beginning of the subsequent melodic unit. Finally there are those subject endings where these two sixteenth-notes do not remain within the harmony, and therefore cannot belong. We are thus dealing with a subject which is conceived with a "female" (weak-beat) and a "male" (strong-beat) ending.

The rhythmic pattern of the subject is very simple, comprising only sixteenth-notes and eighth-notes (or: triplets and entire beats). A third, very characteristic rhythmic figure which is introduced outside the subject (see, from bar 5 onwards, the eighth-note + sixteenth-note groups) gives the fugue the distinctive touch of a gigue.

The subject's pitch pattern includes inverted-mordent figures (the three-note groups in bars 1 and 2), two larger intervals, and consecutive seconds. While the direction is clearly upward in bars 1-4 and downward in bars 4/5, the former is interrupted by two restarts. Both of them follow interruptions of the melodic flow (emphasized by the combination of rest + wedge on bars 2d and 3d) and thus combine to deliver an explicit message regarding the phrase structure. We can therefore describe this subject as consisting of the (very regular) structural pattern of 1+1+2 bars.

The harmonization of this subject is interesting particularly at the beginning insofar as almost all listeners interpret the unaccompanied version in a much simpler pattern than it later turns out to be. The example below gives the solution Bach uses most frequently in the statements of this fugue and, underneath in brackets, the imagined simpler harmonization (ex. 60).

 

Any decision regarding the development of tension in this subject will have to take into account the allusion to the gigue. In light of this virtuoso dance and the exemplary simplicity of phrase structure, the three consecutive upward thrusts (F-C, A-D, C-F) take the lead over any considerations of harmonic tension. The dynamic outline of the subject thus consists of three short crescendo upbeats, each surpassing the preceding one, followed by an unbroken decrease through the octave from F to F.

II/11.2.2 The statements of the subject

This fugue features only eight entries of the subject.

(ex. 61)

 

The subject suffers very few changes. In the tonal answer, the first jump is modified from a fifth to a fourth. In three instances, the originally female ending of the subject – i.e. the melodic ending on a weak beat – is changed to a male ending. (See entries 4 and 8, bars 25 and 89 respectively, and bar 95 where the harmonic resolution occurs on the downbeat but then gives way to a new active movement in the following sixteenth-notes.) No inversion, stretto or parallel of the subject are used in this fugue.

The two final statements in the fugue feature, in addition to the modification of the ending, more striking changes. The second-last entry fluctuates between the major and minor modes: its beginning is surrounded by the F minor chord but the fifth note of the subject is A natural, followed then again by D^b which in turn gives way to D natural (see bars 85-87). This entry also surprises the listener by its thickened texture: bars 86/87 contain two five-part chords and one with six voices. The final entry is considerably lengthened by way of additional sequences of its head motive (see bars 89-93). It is further set apart from the remainder of the fugue by the accompaniment in thirty-second-note runs which do not occur anywhere else in the piece.

II/11.2.3 The counter subjects

Bach does not give this subject any regular companion.

II/11.2.4 The episodes

As the list of the subject entries reveals, by far the largest portion of this fugue is taken up by passages in which the subject is absent. Only 34 of the 99 bars are taken up by subject statements; the remaining two thirds are divided between six episodes.

Material from the subject plays a vital role in the construction of episodes. The third subphrase (Ms) is used frequently, with or without the ascending half-bar upbeat.

-	In E2 we find three imitations of Ms without its upbeat: bars 18/19 (M), 19/20 (U), 20/21 (L).
-	In E3 Ms appears five times extended by a the additional interval of a fourth: bars 29/30 (L), 30/31 (U), 31/32 (M), 32/33 (U), 36/37 (U).
-	Still in E3, Ms (now without the additional skip but with its original upbeat) occurs four times in the lower voice: see bars 44-52.
-	The last recurrence of Ms can be heard in E5, bars 70-72, in the lower voice as a partial sequence of the preceding subject statement.

Other characteristic material includes one motive and several sequence-models. (It is common to call a melodic unit a motive if it recurs in more than one place and in more than one voice. Sequence-models, by contrast, are short-lived and remain unchanged in their texture. In some cases, however, a motive can be part of a sequence-model; see the first model mentioned below.)

M1 is introduced in E1 (see bars 9 B^b - bar 10 E^b); its imitations fill the episode entirely. It recurs in the middle of E3, both in its original shape (see M: bars 37/38 and U: 38/39) and in a variation with a suspension of the third note (see M: bars 39-44 and U: bars 39-45). This variation of M1 is sequenced twice in ascending direction in both the middle and upper voices. At the same time, the lower voice also presents a group of notes (see bars 38-40: E-F) which recurs twice in ascending sequences. The three voices together thus form Model 1.

E4 contains two models, both of them short and occurring only in U+M. In bars 57/58, M imitates U in a three-note ascent; this combination is sequenced twice in a rising direction. Similarly, the same two voices establish another short model in bars 62/63 which is sequenced three times in falling direction. E5 also contains two such models; a very generous eye might even detect a relationship with those in E4, last but not least because the two episodes share the pedal note feature. The first model in E5 appears in bars 72/73: it is again an imitation between U+M with three notes (rhythmic pattern as before) but including a note repetition. Three falling sequences follow here.

The second model in this episode materializes in bars 78/79. It is more complex since the middle-voice figure is longer. Once again ascending sequences follow. Finally there is one figure which only ever appears in the lower voice but presents its sequences without the support of coinciding sequences in other voices – it is thus neither a real model nor a true motive. This sequential figure first appears in bar 25/26 (F-F) where it is followed by two ascending sequences (until bar 28d). Exactly the same order of events recurs in bars 95-98d.

In searching for cadential closes one discovers that the fugue contains only one truly significant cadence; it resembles the final close to a remarkable degree (compare bars 25-29 with 95-99) and must thus be interpreted as dividing the longest episode into two structural segments (E3a, E3b).

Summing up, one can distinguish three types of episodes in this fugue:

-	motivically determined episodes including imitation (E1, E2)
-	episodes built mainly on sequence-models (E4, E5)
-	cadential episodes (E3a, E6)

Then there is the composite construction of E3b with motives and models which, with its twenty-three bars, is almost as long as the four initial subject entries together.

II/11.2.5 Character, tempo, articulation, ornament realization

The character of this fugue is determined by its metric and rhythmic allusion to the gigue. Like all gigues, the F major fugue is lively. Sixteenth-notes mostly constitute ornamental figures or runs; they should be articulated accordingly, i.e. legato for the written-out inverted-mordent in the first and second subphrases of the subject, quasi legato or even leggiero for those in runs(as in the subject's third subphrase). The typical eighth-note + sixteenth-note pattern is decidedly bouncing, with a hint of heavy-light, heavy-light.

Also from the gigue comes the combination of the two rhythmic patterns in one bar (the triplet leading into the eighth-note + sixteenth-note group, see e.g. U: bar 8) in which the eighth-note is traditionally unaccented. Dotted eighth-notes are non legato; exceptions occur only where a tie forms another pattern typical for the gigue: the suspension followed by sixteenth-notes (see e.g. M: bar 23/24). In such cases, the suspension acts as the first of a sixteenth-note group and is therefore not separated from the others.

The tempo of this fugue can be as lively as the thirty-second-notes towards the end of the piece allow without sounding blurred. Ornament symbols do not appear in the score. The relative tempo to the prelude is in simple proportion:

II/11.2.6 The design of the fugue

The two blocks of four subject statements each make a first appreciation of the structure easy. There can be no doubt that the episodes E1 and E2 act as bridges between consecutive entries in the first section, and that E4 and E5 fulfill the same function for the second round of entries.

There is, however, one open question – and one which has a strong impact on the performance. The very conspicuous correspondence between E3a and E6 strongly suggests that the first section might close in bar 29. As, however, the second round of subject entries definitely begins not earlier than in the middle of bar 52, we find ourselves in the following predicament:

Either	we assume that Bach conceived three sections in this fugue. This concept is unorthodox with regard both to a gigue (which, as we all know, is in binary form) and to a fugue (in which sections are defined by the statements of the subject).
Or	we regard the entire episode E3 as belonging essentially to the first section. The piece then appears as consisting of two halves (of almost equal length), but the cadential close in bar 29 and the correspondence between E3a and E6 would seem to be ignored.

For a sketch showing the design of the fugue in F major, see ex. 62.

II/11.2.7 The overall dynamic outline of the fugue

The gigue-like character of this fugue determines the mood to such an extent that increases and decreases of tension play only a subordinate role, mainly in shaping the episodes in direct relationship to the direction of the sequences (ascending sequence means increase, descending sequence means decrease). Furthermore, the dynamic layout of this fugue is very closely related to the interpretation of the design (see the predicament mentioned above). A compromise which evades the problem – without solving it, though – is to play the first section as consisting of two halves.

-	The first half is determined by subject entries and episodes which connect (E1, E2) or close (E3a), and therefore sounds in the usual touch for gigues (mf-f, or register I). The second half of section I consists exclusively of secondary material and therefore sounds like a shadow, or musing afterthought, of the preceding (p, or register II).
-	Consistency then requires that the second section be interpreted as an intertwining of the two levels. Thus E4 and E5 are played as "afterthoughts" following each subject statement separately – and therefore soft (register II), unlike E1 and E2 with whom they no longer correspond in this option. Of the episodes in the second section, only E6 will then sound on the dynamic level I.