This cycle is designed to transmit some very different messages at the same time. And this ambiguity is reflected in the number of compositions: the three contrasting songs are at the same time three similar parts of one song. It would only be consistent to expect the amount of bars in this song(s) to adapt itself accordingly. Which means that the results of the recount could be somewhat less certain than the inflexible laws of mathematics predict.
Of course a regular piece-to-piece counting of written bars will never produce a deviating result. So, in this case this is not the proper procedure. To get somewhere we must start from following paragraph from Part One:
Therefore it is no surprise to recognize in the partsongs a connection between the words and their musical expression that is as close as in the music of Bach. The most obvious example of this imitative style, is the music’s slow dying away on ‘die’, but it is also applied on the more obscure details. And even the totally invisible one: Virtue’s textreduction from 16 to 12 lines is reflected in the choise of measure. Choises actually: in the end Sweet Day is still telling a complete story, and so the 3/4 beat is in the two final bars replaced by a 4/4.
One of the more obscure details is the fact that Sweet Day covers exactly 49 out of its 50 bars. As a rule, every written bar is counted for a full one, even when (most of) it represents silence. But to be truely imitative, the music must add up to a netto score of 49 bars as well: 16 2/3 for verse 1; 17 1/3 for verse 2; and 15 for verse 3. Allowing the score to copy the text in reducing virtue a little more as a first glance reveals.
Applied on the verses of O Mistress Mine, this same method produces a score of 13 1/2 + 13 1/2 = 27 bars. Related to her size, a significant larger reduction than Sweet Day’s, but then, the reduction of her virtue is according to Part Two far more spectacular as well.
In contrast to her companions The Willow Song suffers no reduction of virtue. Lack of virtue is her main concern – Shakespeare did not insert some explicitly sexual remarks in the song for nothing – and what is not there, cannot be reduced. Another thing lacking is a division in verses, so the music flows on uninterrupted; covering all 33 bars from begin to end. What the song does have, however, is a division into story and refrain.
Considering that in these songs the story is everything, we could leave the refrain out. But neither Shakespeare nor RVW give any certainty whether the ninth line is the only complete line from the refrain, or a part of the story. Leaving the options open to delete either 13 or 18 bars. On the other hand we are by now aware of the great similarity between songs. The outer pair being identical twins is good reason to delete at this point some information from the story, and to insert a refrain instead. And by reducing the centre to its four or five refrain lines, we arrive at a total of five different options: 33; 20; 18; 15; and 13 bars.
Instead of a fixed total of 112 bars, we are now facing (in proper order) 2 x 5 x 2 = 20 different ways to count them, producing 15 different results, ranging from 89 to 112. And this is only the beginning: the music’s imitation of the text includes OMM turning into a dance on the word ‘trip’. This dance involves in bars 8 and 10 the appearance of triplets; 3/8 motives in the timespace of a quarter, in this case consisting of a stressed crotchet (1/4) followed by an unstressed quaver (1/8): long-short-long-short-long-short. In short: this is a walz.
Which means that RVW has managed to spirit away eight complete bars: walzes are usually written in 3/4. All metrical accents in bars 8 and 10, and in the second verse’s bars 22 and 24, therefore are in fact downbeats of separate 3/4-measured bars. This results in four options: 27; 29; 35; and 37, and 2 x 5 x 4 = 40 totals.
The first multiplier will also raise, so we now are in acute danger of drowning in data which relevancy is still questionable. Therefore it is from this point onward very important to keep in mind that this first song, the one this procedure is supposed to set apart from the Shakespearean couple, has reduced Virtue by four lines; from sixteen to twelve.
As mentioned, Sweet Day’s verses cover respectively 16 2/3; 17 1/3 and 15 bars. The third verse, however, concludes with two bars in a deviating 4/4 beat. Such a change is not uncommon in music, but at this occasion it is rather peculiar: while handling with superior control the irregularities in the meters of two lines in this song, RVW allows himself not even the slightest irregularity in his chosen measure. And now, dealing with an immaculate regular meter, he introduces an irregularity of his own making.
The mathematical correct way to express the presence of different kinds of bars in the calculation, is the placing of the third verse between brackets. With verses one and two sharing bar 18, the regular way of counting now looks like this:
17 + 1 + 17 + (13 + 2) = 50
The netto sized alternative runs:
16 2/3 + 17 1/3 + (13 + 2) = 49
Brackets are very usefull to isolate a part of the calculation. Making it possible to introduce changes without disturbing the formula as a whole. And because mathematics is a necessary part of a composer’s professional education, RVW must have been aware of the fact he created such an opportunity. This being his very intention, would make a rather plausible explanation for a seemingly lighthearted abandoning of a beat he previously considered worth a real effort in maintaining. And indeed; applying between brackets all four basic routines will prove very instructive:
Varying the formula with (13 + 2); (13 – 2); (13 x 2), and (13 : 2), produces 8 different results: 40 1/2; 41 1/2; 45; 46; 49; 50; 60; and 61. Leading up to 8 x 5 x 4 = 160 different ways to calculate an overall score, with 70 different results, ranging from 80 1/2 to 131. But no musicologist will ever accept fractions for the overall size of a piece of music. And RVW seems to have carefully avoided such a breach of rules to happen in both TWS and OMM. Musically isolating SD from a combination of TWS and OMM by a fractioned number of bars must therefore be considered impossible. Which makes the calculations resulting in fractions for SD’s total, superfluous.
This not only leads to a marvellous regularity in the formula that defines the number of alternative calculations: 6 x 5 x 4. It also reduces their number from 160 to 120 by removing the fourty superfluous ones.
This in exact parallel of the reduction of Virtue to Sweet Day.