The following is a tentative draft of a short section from “Introduction to the Steps,” chapter 13 of the work in progress The Technique of Eighteenth-Century Ballet.


8.5 Tempo

Perhaps one of the most debated issues in reconstructing eighteenth-century dance has been the choice of fitting musical tempi, an important issue since the speed of accompanying music greatly determines the range and character of a dancer’s movement. And thus an understanding of how certain pieces of music were typically performed in our period of study should, it has been argued, tell us something about the nature of eighteenth-century dance. One of the major hurdles here is the fact that the eighteenth century did not possess any widely-accepted practical means of communicating precise degrees of speed, for the practice of giving standard metronome markings was not known until the nineteenth century. Clearly some individuals in our period felt that time-beating devices would have been useful. Rousseau (1768: 99-100), for example, writes that

many claim, however, that having such an instrument is most desirable in order to set precisely the tempo of each measure in [100] a piece of music; one might thereby more easily preserve the true tempo of tunes, without which they lose their character, and which, after the death of their creators, can be known only from a kind of tradition very liable to change or die out. It is already a complaint that we have forgotten the tempi of a great number of tunes, and it is to be believed that all of them are slower now.

Despite a number of different designs for “chronometers,” or metronomes, that came to light in our period of study, none of these early attempts achieved any wide or lasting success (Fallows 2001: 535). One of the main objections to such time-beating devices was the fact that tempo in good music-making was flexible, especially in French practice, wherein a strict tempo was often not observed in the interests of expressiveness, at least according to Rousseau. As the latter again writes (1768: 100-101), connoisseurs of music

“will object,” says Mr. Diderot (Mémoires sur différents sujets de mathématiques [(1748) 1987: 62]) “to any chronometer broadly, since in an air, there are perhaps not two bars that are of the exact same duration, two things necessarily contributing to slacken the one and hasten the other: taste and harmony in pieces with many parts; taste and a sense of harmony in solos. A musician who understands his art will not play four measures of an air without entering into its character and being wholly wrapped up in it; it is but the pleasure in the harmony that holds him back. He wishes the concords to be striking here, hidden there; that is to say, he sings or plays with lesser or greater slowness from one bar to another, and even from one beat or quarter of a beat to the next.”
To be sure, this objection, which is of great weight in French music, is of no consequence in Italian, which is irremissibly subject to the most strict time: indeed, nothing can show better how utterly opposed these two musics are, since what is a beauty in the one would be the greatest flaw in the other. If Italian music draws its energy from that slavery to the strictness of the measure, the French seeks its own in mastering the same measure as it so pleases, by hastening or slackening it, as taste in singing so demands, or the degree of flexibility in the singer’s voice. . . .
[101] I will add that whatever instrument may be found to set the duration of a measure, it cannot possibly have a place in practice, no matter how fine its working might be. Musicians, those confident, forming, like many others, the rules of good taste after their own, would never adopt it; they would forget the chronometer and govern themselves only according to true character and the true tempo of the tune. Thus, the only good chronometer to be had is a skillful musician with taste, he who has read well the music to be performed and knows how to beat its time.

Indeed, the view that the musical tempo within a given piece ought to be expressively variable predates our period and finds utterance in Mersenne (1636).


8.5.1 Tempo Descriptions

When the “typical” tempo of different French dance movements is described in the sources, the movement in question is not uncommonly defined in relation to another, such that this one movement is somewhat slower than that one but faster than another. Such relative definition is evident in Tomlinson’s brief discussion of musical forms (1735: 148), among others. After asserting that the courante is “very solemn and grave,” for example, he adds that “the next grave Movements are Sarabands, Passacailles, and Chaconnes, each of three Crotchets to a Measure, and every one a Degree lighter than the other,” and then “the next Minuets and Passepieds are still brisker.”

It is apparent from a variety of sources, however, that there was a fair range of speeds possible for at least many, if not all, of the French dance movements, enough that references to a characteristic tempo may in fact be contradictory. Türk (1789: 401), for example, writes that “the passacaille (passacaglio [sic]) is played somewhat more slowly, or as others would have it, almost a little faster than the chaconne.” Indeed, while the sources commonly attempt to give brief characterizations of dance movements partly in terms of speed, it is clear from a comprehensive consideration of various descriptions and an examination of pieces of music written in a given form that tempo was not the most significant defining feature for such pieces. Consider, for example, D’Alembert’s remark (1779: 210) that “the tempo of the gavotte is either slow or gay but never extremely fast nor excessively slow.” The same ambiguity is apparent in Brossard’s sketch of the gavotte (1705: 29): “sometimes gay, sometimes grave.”

Thanks to its popularity during our period and thus to the existence of a fair amount of information about its character, the minuet affords a good example of how a given movement could vary in tempo. The sources from the whole century inconsistently describe this movement as typically slow or modéré (‘moderate’) or gai (‘gay’) or even vite (‘fast’). It should be noted that these French terms were roughly the equivalents of the Italian markings andante, allegro, and presto respectively, at least according to Rousseau (1768: 303):

There are five main changes in tempo, which, when ordered from slow to fast, are expressed by the words largo, adagio, andante, allegro, presto, and these words are rendered in French by the following: lent, modéré, gracieux, gai, vite. It should be noted, however, that, tempo always being much less strict in French music, the words denoting it [in French] have a much vaguer sense than in Italian music.

Masson (1699: 7), for example, speaks of the minuet as “fast” (vite), while Brossard (1705: 43) characterizes the movement as “always very gay [fort gai] and very fast [fort vite].” To Taubert (1717: 615), the minuet was “a right merry dance,” Sol (1725: 51) similarly writing that “to my mind, there are no grave steps in the minuet, for the movement in the minuet is gay.” Pauli (1756: 63) states that “the movement in the minuet is always serious although the accent is animated by the expression in the tune, which is more or less brisk. It always keeps a strong harmony [i.e. beat] by virtue of its ternary measure.” In his 1773 German translation of Charles Burney’s The Present State of Music in France and Italy, Bode notes that the “rapidity in the minuets of all modern overtures [i.e., symphonies] renders them ungraceful in an opera, but in a church they are indecent” (translation in Hogwood 2002: 237).

Other sources, in contrast, describe the minuet as being only moderate or even slow in tempo. Mattheson (1739: 2/224), for example, writes that the minuet expresses “no other affect but moderate gaity” (mässige Lustigkeit). Rousseau (1768: 277) writes that the minuet is

a dance tune of the same name, which the Abbé Brossard tell us comes from Poitou. According to him, this dance is very gay [fort gai], and its tempo is very fast [fort vite]. On the contrary, the character of the minuet is one of elegant and noble simplicity; the tempo is more moderate [modéré] than fast [vite], and one can say that the least gay of all the kinds of dances used in our balls is the minuet. It is otherwise on stage.

D’Alembert (1779: 208) writes that “the minuet is a tune in three time at a moderate [modéré] tempo.” Türk (1789: 401) in like manner speaks of the minuet’s “noble and charming character,” which proceeds “moderately fast” (mässig geschwind). In contrast, Busby (1786?: “minuet”) describes the dance as “of slow and graceful motion.”

Tomlinson (1735: 106) acknowledges that there was more than one possibility here, such that the ballroom minuet could be “performed faster or slower, according to the Tune that is played, which the Dancer is obliged to follow.” Indeed, Pemberton (1729?) published a notated dance called the “Slow Minuet.” According to the research of Zaslaw (1989: 496-497), the tempo of the minuet, based on extant chronometric markings or surviving mechanical musical instruments from the period 1705-1775, could range anywhere between dotted half note = MM 46 to 77.


8.5.2 Tempo in Music for Dancing

It is also clear from a variety of sources that a given musical form could be performed faster or slower depending on whether it was intended for dancing or only for listening. As the ballet master Fabiani (1770) writes:

There is a great difference between chamber, symphonic, and ballet music. As far as the latter is concerned, it is not just the notes that matter, but first and foremost the direction by the ballet master, how long one step [pas] or another, as well as the pauses, has to be held. If the orchestra does not have this advantage, the performance is inevitably threatened with confusion and its quality diminished. (Trans. in Grave 1996: xvi)

Indeed, Quantz (1752: 269) intimates that the tempo of dance music broadly was to be altered to suit the dancer’s needs: “Fairness thus demands that the orchestra accommodate [the dancers] as much as possible, which can easily be done if one gives heed just now and then to the coming down of the feet.” This general practice of altering the tempo of a dance movement depending on its use is apparent from a number of remarks from the period concerning again the minuet. Saint Lambert (1702), for example, speaks of a difference in tempo between minuets intended for dancing and those for listening. Türk (1789: 401) fleetingly notes that “in some areas, minuets, when not meant for dancing, are played much too fast.” Busby (1786?) also contrasts the tempi of minuets for dancing with those only for listening:

MINUET. A movement of three crotchets or three quavers in a bar; of a slow and graceful motion, and always beginning with a beating note. This is the dancing minuet, and is said to have been invented at Poitou: but there are other minuets of a time somewhat quicker, and which were formerly much used as concluding movements of overtures, sonatas, &c.

In a letter to his sister from Bologna, dated 24 March 1770, Mozart reveals that a “theatrical minuet,” that is, one to be danced to on stage, could in practice proceed more slowly than those meant only for listening:

I shall soon send you a minuet which Mr. Pick danced in the theatre and which everyone danced to afterwards at the feste di ballo in Milan, solely in order that you may see how slowly people dance here. The minuet itself is very beautiful. It comes, of course, from Vienna and was most certainly composed by Deller or Starzer. It has plenty of notes. Why? Because it is a stage minuet which is danced slowly. The minuets in Milan, in fact the Italian minuets generally, have plenty of notes, are played slowly and have several bars, e.g., the first part has sixteen, the second twenty or twenty-four. (Trans. in Anderson 1985: 121)

The expression tempo di minuetto (‘minuet tempo’) could be used in our period to mark minuets which were to be performed with a slower speed suitable for dancing (Hogwood 2002: 247).

Thus, it is fairly clear that the needs of the dancer were far more important in determining the tempo of dance music than any systematizing by music theorists.


8.5.3 Stylistic Considerations

The sources intimate, moreover, that the style of dance cultivated also had a great influence on the tempo of the accompanying music. Ballroom dancers, it seems, typically danced to faster tempi than their counterparts in the theater, particularly those in the serious style, given the more contained range of movement used by amateurs and the need for them to avoid having to maintain more taxing positions, such as balances (if they bothered to do them at all). It is clear at least that ballroom contredanses proceeded at very brisk paces. Lecointe (1752: 63) writes that the tempi in such dances could be so swift that the dancers were at times obliged to run after the music, which only encouraged the fashion for artful carelessness in ballroom technique:

People begin to give Way to their natural Propension to Negligence. Country-dances come again into Vogue, and the Complaisance which Masters have to teach these silly Dances, tends to destroy the few Rudiments which the English retain of this Art . . . for how is it possible for a crowd of young People to observe Rules, who have their Heads full of Nothing but how they must keep Time to a precipitate Measure, which hardly gives them Leisure to describe the Figures, tho’ they exert their utmost Activity [i.e., nimbleness]?

The author of “A New Treatise on the Art of Dancing” (March 1785) similarly writes that “the cotillon must be considered, as it really is, a French country dance, and consequently should be danced, as already said, with uncommon vivacity.”

Mention should be made here of a collection of dances recorded in Beauchamps notation (Paris, Bibliothèque de l’Opéra, ms. 817), which contains six dances apparently bearing chronometric markings (Harris-Warrick 1993: 13-15). These markings more or less match those given by L’Affilard and d’Onzembray based on dance type (e.g., bourrée, sarabande, etc.), as Harris-Warrick notes. While the dances in question are styled “theatrical,” the collection perhaps belonged to a master who taught dance to amateurs – thus, a collection of danses d’exercice (Pauli 1756: 54-57) – and so may be a window into amateur practice. (It is well documented that amateurs of our period did in fact study more difficult – even theatrical – dances (Fairfax 2003: 317-333)). As Harris-Warrick notes, some of the dances, when reconstructed, proceeded quickly, the fast tempi resulting in mincing movements, again in agreement with the characterization of ballroom dance technique as “miniaturized” and at odds with the well-documented characterization of theatrical dance as “exaggerated” (Fairfax 2003: 15-79). Of course, it is impossible to be certain that these few markings were really directed towards dancers: The manuscript could also function as simply a collection of tunes.

The music meant to be danced to by theatrical performers in the serious style clearly proceeded very slowly in order to give the executants enough time to do their expansive movements. As discussed earlier (§3.10), all the evidence points to leg gestures being commonly done at the height of the hip or thigh in the theater, and in the wake of Louis Dupré’s precedent in the 1730s, even higher in the serious style. Indeed, references in the sources to tempi employed in the grave operas at the Paris Opéra stress that the music was commonly very slow moving. In his retrospective look at earlier performances there during the first half of the period, Noverre (1807: 2/104-105) alludes to a slow pace of movement, such that Camargo was a welcome relief with her sprightliness: “in a spectacle where everything was gloomy, drawn-out, and languorous, it was a joy to see a dancer so animated and whose liveliness could draw the public out of the drowsiness into which monotony had plunged it.” Suard (1753?: 33-34) also notes that music in French serious opera was typically very slow:

First of all I showed that the music of the French was intimately bound up with the dance which they were pleased to call ‘grave’ and that no alteration to the character of one could be done without effecting the same changes in the other. I proved that the softness and monotony of this dance stemmed from the same taste that was responsible for the slowness of the music.

The entrée grave, a form proper to the serious style, was particularly languorous, Tomlinson (1735: 143) writing that “from the former of these [i.e., common time] flow very slow Entrees containing two steps in each Measure call’d, Quadruple or of two Times on Account of their Slowness or admitting of a suppos’d Bar in the Middle of the said Measure.” Anne Heinel (1753-1808), who according to Noverre (1807: 2/118) was “the most perfect model for serious dance,” was noted for her smoothness of movement in the slow-moving grave genre: “She moves as gracefully slow as Pygmalion’s statue when it was coming to life, and moves her leg round as imperceptibly as if she was dancing in the Zodiac” (Walpole 1904: 8/76 = 25 Aug. 1771). The serious dances created by Charles Didelot, a traditionalist who upheld the system of distinct styles at the end of our period and into the next, typically used music described as adagio and marcia, according to Glushkovsky (1940: 163-164). For further discussion, see Fairfax (2003: 86-103).

Biomechanics suggest that the maximum workable range of tempi for the traditional serious style, regardless of musical form, was roughly largo to marcia (in current practice), which would allow a dancer to move in an expansive unhurried manner.

The following is a MIDI mock-up of the first few bars of the passacaille from Lully’s Armide performed at quarter note = MM 55, included here to give the reader a sense of the slow tempo suitable for the serious style in light of the foregoing and descriptions of the movement quality and range of movement typical of the traditional grave style (Fairfax 2003: 86-103). I have more or less followed Quantz’s prescription (1752: 270) that in such music “the eighth notes that follow the dotted quarter notes must be played not according to their literal value but in a very short and sharp manner,” so as to create the requisite “slow, starchy movement of a passacaille” (Angiolini 1765). Incidentally, a handful of dances choreographed to this passacaille are in fact extant from the early eighteenth century (Little & Marsh 1992: 58-59). (Click on the white arrowhead on the left of the bar in order to play.)


As discussed in detail elsewhere (Fairfax 2003: 103-188), the traditional half-serious, comic, and grotesque were airborne styles that typically moved to faster tempi. Biomechanics suggest that for these styles, the maximum range was roughly andante to allegro moderato (in current practice).

Performing music faster than the maximum ranges given above necessitates treating two bars of music as one – as far as the dancer is concerned – such that one pas composé (normally to be executed in one bar) would need to take up two measures rather than one. All of the extant dances in notation, however, found in the early-eighteenth-century published collections and in the Ferrère manuscript, regularly show one composite step fitting within one bar of music, except in the slow 4/4 of the serious style, wherein two composite steps fit within one bar of music. (This does not take into consideration such steps as the pas de menuet, which normally take up more than one bar of music, or isolated précipité executions.) In fact, Magri (1779: 1/23-24) notes that in the more virtuoso airborne dancing of his day, the trend was to “put more steps into one bar,” rather than the other way around, and this cannot be done to very fast music without the movements becoming laughably shrunken or the dancer running after the music.

It is clear, however, that in some cases, fast tempi were used, which did indeed necessitate treating two bars as one, from the dancer’s perspective. The music for the grotesque roles of the Furies and Winds, for example, is said to have proceeded at a very fast clip. As Bonin (1712: 165) writes, “the Furies are dances the speed of which can scarcely be imagined, for not only the music itself but also the feet must move in and out like lightning, such that the latter, because of the speed, can hardly be told apart.”  As to the Winds, Bonin (1712: 196) again writes, “the dance itself goes likewise very fast.” A concrete example is afforded by the extant annotated score of the Vestris/de la Borde pantomime ballet Médée et Jason (1770), which marks the dance of the Furies, beginning at bar 518, as “très vif” (‘very brisk’). Given that the Furies “go beyond the norm in everything” (Magri 1779: 1/111), i.e., employed highly exaggerated movement, the only real way to reconcile large gesture and high elevation in jumps with very fast music is to treat two bars of music as one. Indeed, any grand jump would need to be treated thus, regardless of style.

Very fast tempi were unproblematic in scenes wherein pantomime alone was performed, or nearly so. A further concrete example can be found in another scene from Médée et Jason, that wherein Medea appears in her chariot with her children, one them already slaughtered, beginning at bar 418: Here the music is again marked “très vif.” And finally, “dances” consisting of a series of poses theoretically could be performed at any tempo as well (Fairfax 2003: 172-174).