A few days before that, I wrote, in connection with menuets in Lully operas:
A look at these menuets shows -- by negative example -- how closely Caplin's analysis model and terms fits music of the later 18th and early 19th centuries. When ideas are three bars rather than two and clearly demarcated, there is no problem. Six-bar phrases are common enough in 17th century music, and especially in menuets. The difficulty arises at the next level, as these six-bar phrases consisting of two three-bar ideas are often treated as the first strain in a small binary form. In the analyses here, I have in most instances understood the entire piece as a prolonged theme. Since the distinction is based on length, however, that is not entirely satisfactory as a method. (link to post)Rebecca Harris-Warwick has explored the issue of varying and irregular phrase lengths and groups more thoroughly in a chapter for the volume Lully Studies (citation at the end of this post). She "demonstrates that only a small proportion of Jean-Baptiste Lully's dance music—roughly a quarter of his binary pieces—is based on consistent four- or eight-bar phrases, a point that remains valid even for generic dances such as the minuet, gigue, or gavotte" (from the RILM abstract).
For example, in the category she refers to as "celebratory dances" (marked as and performed as dances), "the affect may be predictable, but the phrase structure of the dances is not." In several examples she cites, "the reigning affect is joy, and nothing suggests that Lully automatically assigns five-bar phrases to shepherds or nine- and eleven-bar phrases to river gods. On the contrary, the same set of characters may perform back-to-back dances with completely different metrical structures" (49). Even in some of the most formal dance types, such as the theatrical entrée, highly flexible phrase designs reign. Two surviving choreographies of this type, "both . . . for a man and a woman[,] draw upon [a characteristically] intricate and technically demanding step vocabulary. Both demonstrate not only that irregular and varied metrical structures pose no impediment to choreography, but also that such structures are perfectly compatible with dances in the noble style. Lully [composed them] as representatives of the art of dance. The art they exhibit is subtle, and varied, not restricted to predictable, square phrases" (54).
In still another example, Harris-Warwick notes that the music for a dance may appear irregular if heard on its own, but it may well gain a strong sense of symmetry when combined with the dance (as in the mirroring patterns commonly found in menuets) (56). The implications for research are obvious: "dance music operates in conjunction with the visual and the kinetic. A study of the phrase structures of Lully's dances needs to take these intersecting systems into account."
And the historical generalization she derives from all this is as follows: "Lully [was not] alone in this flexible approach to the construction of dances; seventeenth-century dance music in general exhibits a much freer approach to phrase structures than does the dance music of the eighteenth century, one that embraces irregularity and sees square phrasing as only one possibility among many. The notion that stretching a four-bar phrase into six represents a deliberate distortion of an underlying dance pattern may be tenable in regard to a Haydn minuet, but not to one by Lully" (55-56).
All this seems consistent with my own assertions in blog posts about Lully menuets (link to the first of three), for example, "Form descriptions of Lully menuets from the operas show the fluidity (and uncertainty) of design analysis based on Caplin's model and terminology. We saw in the vocal and instrumental menuets from Atys and Proserpine that the design can appear to be simultaneously a small form and a prolonged theme, and that in the latter case an initial phrase is followed by two continuation phrases" (link).
Reference:
Rebecca Harris-Warwick, "The Phrase Structures of Lully's Dance Music." In John Hajdu Heyer, ed., Lully Studies, pp. 32-56. (Cambridge: Cambridge University Press, 2000).
