V

Articulation of
Quantitative Meter

THE PERCEPTION OF quantitas intrinseca, or “good” and “bad” notes, gave essential information to performers about standard articulation patterns. Instead of relying on markings for slurs, staccato marks, sforzandos, and accents, seventeenth- and eighteenth-century performers interpreted their unmarked scores through habits and formulas learned as part of their elementary instruction. No matter how subtle and polished a performer eventually became, articulation determined by the meter of the music was embedded in his or her technique.

Some metrical articulations are immediately audible and define a style of performance as well as the beat, as in the lilting quality of the French practice of notes inégales. Other articulations may be less overtly apparent to the listener, but they also persuade the ear to a greater appreciation of metrical structure.

Articulations are sometimes specific to a particular instrument, for instance, minute silences result from keyboard fingering patterns that require some fingers to lift early in order to make way for the placement of others. Tonguings for wind instruments and bowings for stringed instruments use varying degrees of sharpness of attack. Articulations are sometimes explained as a means of producing an instrumental equivalent to pronunciation that can be subtly and infinitely varied to fit different musical circumstances.

When editors attempt to help musicians today to perform seventeenth- and eighteenth-century music by supplying phrasing and articulation marks, they are all too often based on the articulation practices of a later style of performance.

Patterns of articulation that do not depend on accent or dynamic stress are discussed in this chapter.

Notes Inégales

Notes inégales is a performance practice associated with French style in the seventeenth and eighteenth centuries. It imparts a graceful swing to quickly moving melodic notes and enhances meter through lengthening good notes and shortening bad notes.¹ Any degree of inequality, from the slightest (such as 9:7) to the greatest (3:1), is in marked contrast to equal performance. Perception of the beat and also the sense of style are influenced by the degree of inequality.

The practice of performing quick-moving conjunct equal notes unequally can be traced to the sixteenth century, when it was not considered to be specifically French. It was one of a number of liberties exercised by performers in ornamenting and improvising. The first mention of this practice is found in a pamphlet by Loys Bourgeois in 1550. He states that notes performed unequally are restricted to those moving at one-quarter of the speed of the tactus and are sung “two by two, staying a bit longer on the first than on the second, as if the first had a dot.”² Diego Ortiz describes notes inégales as among the simplest techniques of ornamentation (Ex. 5.1). ³

EX. 5.1

Sometimes ornamentation was initially conceived as melodic elaboration in even notes and later complicated by rhythmic alteration. A section from Tomás de Santa Maria’s Libro llamado Arte de tañer fantasia describes how to enrich the practice of improvising ornamentation by performing notes unequally. ⁴

Musicians continued the tradition of metrical freedom and the practice of diminutions in the early seventeenth century, but with a change in attitude that is expressed by Giulio Caccini.⁵ According to him, exuberance of decoration ideally is tied to a specific affect or to the articulation of a phrase. Perhaps it was at this time that rhythmic alteration began to be appreciated for its ability to enhance the awareness of metrical organization. From the viewpoint of the twentieth century, however, even the comparatively precise rules of the seventeenth and eighteenth centuries allow performers considerable freedom.

After 1650, unequal performance of notes written as equal became associated primarily with French practice. Although Bacilly,⁶ Marais,⁷ and Jean Rousseau⁸ all discuss inequality, the first thorough account is in Etienne Loulié’s Elemens of 1696. Loulié’s rules governed the practice of French musicians until the demise of notes inégales in France, apparently in the second half of the eighteenth century.

The rules for inequality are combined with explanations of time signatures and gestures for beating time. Since the relation of note values to the beat determines which notes are to be performed unequally, this arrangement was a logical one. Here is Loulié’s explanation of two-beat meters:

La mesure à deux Temps ne se bat que d’une maniere: Un Frapper & un Lever. Le 1. et le 3. quart de chaque Temps sont plus longs que le 2. & que le 4. quoy qu’ils soient marquez égaux, dans quelque mesure que ce soit.⁹

The beat of duple meters is given in only one way, one downstroke and one upstroke. The first and third quarters of each beat are longer than the second and fourth quarters, even though they are written as equal, no matter in which measure they occur.

Each temps of the mesure à deux temps is a beat, within the comfortable physiological and psychological tempo range required, and each beat is divided into two pulses. Unequal notes are eighth notes in 2 and , and sixteenth notes in 2/4.

Loulié continued with a description of triple measures:

3/1, 3/2, 3/4, 3/8, 3/16, & 3

La mesure se bat de trois manieres: 1. 2 frappers & un lever pour les mouvements lents. 2. 1 frapper qui vaut 2 temps & un lever pour les mouvements plus vites. 3. 1 frapper qui vaut trois temps pour les mouvements très vites. Dans quelque Mesure que ce soit, particulièrement dans la Mesure à trois Temps, les demi- temps s’excutent de deux manieres différentes, quoy que marquez de la même maniere. 1. On les fait quelque fois égaux. Cette maniere s’appelle détacher les Nottes, on s’en sert dans les chants dont les sons se suivent par Degrez interrompus. 2. On fait quelque fois les premiers demy-temps un peu plus longs. Cette maniere s’appelle Lourer. On s’en sert dans les chants dont les sons se suivent part Degrez non interrompus. On appelle Degré interrompu lorsqu’un son est suivy d’un autre son que est au 3. ou 4. degré & plus du 1. soit en montant, soit en descendant, comme ré, la, fa. Il y a une troisiéme maniere, ou l’on fait le premier demi-temps beaucoup plus long que le deuxiéme. Mais le premier demi-temps doit avoir un point. On appelle cette 3. maniere Piquer ou Pointer.¹⁰

The measure is given in three ways: (1) two downstrokes and one upstroke for slow tempos, (2) one downstroke, lasring two times, and one upstroke, for faster tempos, (3) one downstroke, lasting three times, for very fast tempos. In any measure, and particularly in triple measures, the notes equal to a half beat are performed in two different ways, even though they are written the same. (1) They are sometimes played equal, this is called “detached,” and it is appropriate when notes succeed one another disjunctly. (2) Sometimes the first part of the half beat is played a little longer, this is called lourer and it is used when melodic progression is conjunct. Disjunct signifies that one note is followed by another after an interval of more than one pitch, either ascending or descending. There is a third way; the first half beat is played much longer than the second half beat, but the first must have a dot after it. This is called dotted.

Much later in the book, Loulié added another explanation:

On avoit oublié de dire dans la 2. partie en parlant des signes de mesures de trois temps, que les premiers demi-temps s’excutent encore d’une quatriéme maniere, sçavoir en faisant le 1. plus que le 2. ainsi [Ex. 5.2].¹¹

I forgot to mention in the second part, speaking of time signatures of three beats, that the first half beats are played in a fourth way; with the first shorter than the second, thus [Ex. 5.2].

EX. 5.2

Loulié and many other French writers demonstrate that lourer, or notes inégales, became an accepted custom in the performance of all French music and remained so until late in the eighteenth century. The rules of notes inégales were restated with further refinements and particulars by subsequent theorists.

The question of whether notes inégales was a performance practice of musicians outside of France is difficult to answer conclusively. By far the greater number of treatises discussing it are French, yet it is defined and discussed in a number of German instruction books. The practice of notes inégales would have been spread by French musicians who had careers outside France, for example, in the famous orchestra at Dresden that had many French performers: J. B. Volumier was the leader of the violins, Pierre Gabriel Buffardin the flutist, and François LaRiche the oboist. At least a dozen French oboists were employed in Germany in various cities, and Bach’s bassoonist at Cöthen was J. C. Torlée.¹² French musicians were active in London; the names of Jacques Paisible, P. J. Bressan, J. B. Loeillet, and Charles Dieupart come readily to mind.

Georg Muffat lived in Paris from 1663 to 1669, from age ten to sixteen, and studied with Lully. His preface to Florilegium secundum (1698) describes the customs of the “Lullists” to his German contemporaries and offers one of the most thorough descriptions of French performance available from this period. Muffat’s directions (in Latin, German, Italian, and French) cover many details of performance—bowings, ornamentation, descriptions of dances—and include the following discussion of notes inégales:

3. Les notes diminuantes du premier ordre, telles que sont les doubles cro- chuës sous la mesure à quatre tems; les crochuës sous celle à Deux, ou alla breve; celles qui vont de la moitié plus vîte qu’une partie essentielle de la mesure aux triples un peu gays, & aux proportions, étant mises de suittes ne se jouënt pas les unes égales aux autres, comme elles sont marquées; car cela auroit quelque chose d’endormy, de rude, & de plat; mais se changent à la Françoise en adjoutant à chacune de celles qui tombent au nombre non pair comme la valeur d’un point par le quel devenant plus longue, elle rend la suivante d’autant plus courte.¹³

The small notes of the first class, which are sixteenths in C, the measure of four beats; eighths in 2 or allabreve, the measure of two beats; those notes that are twice as fast as essential parts of the beat in the faster triples and proportions (provided these notes are conjunct)—all these are not played equal to each other as they are written. This would be sleepy, inelegant, and flat. Rather, the odd numbered notes are played as if they were altered by the addition of a dot in the French style, and by becoming longer they cause the following notes to become shorter.

In Walther’s Lexikon, notes inégales, under the name of lourer, is similarly defined and noted as French in origin.¹⁴ The next, and the best-known, German description of notes inégales is that of J. J. Quantz, a flute student of Buffardin’s. Quantz did not identify notes inégales as French or refer to it as a foreign practice, but at the French-speaking court of Frederick the Great, where Quantz spent most of his life, perhaps French performance practice was not considered foreign.

Ich muss hierbey eine nothwendige Anmerkung machen, welche die Zeit, wie lange jede Note gehalten werden muss, betrifft. Man muss unter den Hauptnoten, welche man auch: anschlagende, oder nach Art der Italiäner, gute Noten zu nennen pfleget, und unter den durchgehenden, welche bey einigen Ausländern schlimme heissen, einen Unterscheid im Vortage zu machen wissen. Die Hauptnoten müssen allezeit, wo es sich thun lässt, mehr erhoben werden, als die durchgehenden.¹⁵

I must in this connection make a necessary remark concerning the length of time that each note must be held. A distinction in performance must be made between principal notes, also called “struck” or “good” notes, after the Italian custom; and passing notes, which are called “bad” notes by some foreigners. Wherever possible, the principal notes must be brought out more than the passing notes.

This was not just a general introduction to the remarks that followed. The terms—“struck notes” and “good notes”—are those used to define quantitas intrinseca. These notes were to be “brought out” (mehr erhoben). However, “struck” or “good” notes are not the same as notes inégales:

Dieser Regel zu Folge müssen die geschwindesten Noten, in einem jeden Stücke von mässigem Tempo, oder auch im Adagio, ungeachtet sie dem Gesichte nach einerly Geltung haben, dennoch ein wenig ungleich gespielet werden; so dass man die anschlagenden Noten einer jeden Figur, nämlich die erste, dritte, fünfte, und siebente, etwas länger anhält, als die durchgehenden, nämlich, die zweyte, vierte, sechste und achte: doch muss dieses Anhalten nicht so viel ausmachen, als wenn Puncte dabey stünden. Unter diesen geschwindesten Noten verstehe ich: die Viertheile im Dreyzweitheiltacte; die Achtheile im Dreyviertheil- und die Sechszehntheile im Dreyachttheiltacte; die Achttheile im Allabreve; die Sechzehntheile oder Zwey und dreissigtheile im Zweyviertheil- oder im gemeinen geraden Tacte: doch nur so lange, als keine Figuren von noch geschwindern oder noch einmal so kurzen Noten, in jeder Tactart mit untermischet sind; denn alsdenn müssten diese letztern auf die oben beschriebene Art vorgetragen werden. Z. E. Wollte man Tab IX Fig. 1. die acht Sechzehntheile unter den Buchstaben (K) (M) (N) langsam in einerley Geltung speilen [Ex. 5.3]; so würden sie nicht so gefällig klingen, als wenn man von vieren die erste und dritte etwas länger und stärker im Tone, als die zweyte und vierte, hören lässt. Von dieser Regel aber werden ausgenommen: erstlich die geschwinden Passagien in einem sehr geschwinden Zeitmaasse, bey denen die Zeit nicht erlaubet sie ungleich vorzutragen, und wo man also die Länge und Stärke nur bey der ersten von vieren anbringen muss. Ferner werden ausgenommen: alle geschwinden Passagien welche die Singstimme zu machen hat, wenn sie anders nicht geschleifet werden sollen: denn weil jede Note von dieser Art der Singpassagien, durch einen gelinden Stoss der Luft aus der Brust, deutlich gemachet und markiret werden muss; so findet die Ungleichheit dabey keine Statt. Weiter werden ausgenommen: die Noten über welchen Striche oder Puncte stehen, oder von welchen etliche nach einander auf einem Tone vorkommen; ferner wenn über mehr, als zwo Noten, nämlich, über vieren, sechsen, oder achten ein Bogen steht; und endlich die Achttheile in Giguen. Alle diese Noten müssen egal, das ist, eine so lang, als die andere, vorgetragen werden.¹⁶

In accordance with this rule, the quickest notes in every piece of moderate tempo, or in the adagio, though they seem to have the same value, must be played a little unequally, so that the struck notes of each group, namely the first, third, fifth and seventh notes are held somewhat longer than the passing notes, which are the second, fourth, sixth and eighth, although this lengthening must not be as much as it would be if the notes were dotted. Among the quickest notes I include the quarters in the three-two measure, the eighths in the three-four measure and sixteenths in the three-eight measure; the eighths in allabreve; and sixteenths or thirty seconds in two-four or common time; but this is true only so long as no figures of still more rapid notes or doubly quick ones are mixed in, in whatever time signature, for then these last named have to be performed in the manner described above. For example, if one were to play the eight sixteenth-notes under letters (K) (M) and (N) slowly and evenly [Ex. 5.3], they will not sound as pleasing as if the first and third were held somewhat longer and played somewhat louder than the second and fourth. This rule has the following exceptions: fast passages in a very fast tempo, when there is not enough time for them to be performed unevenly and therefore only the first of each group of four notes can be given (extra) length and force. Also excepted are all fast passages for the voice, when they are not to be slurred, for every note in this type of vocal passage must be marked and made clear by a gentle expulsion of air from the chest, and thus unevenness has no place in such passages. Further exceptions are: when notes have dashes or dots over them, or when there are several successive notes of the same pitch; also when there is a slur over more than two notes—that is over four, six or eight; and finally eighth notes in gigues. All of these notes must be played even, that is, one as long as the other.

EX. 5.3

Quantz’s instructions are more detailed than those of Loulié, as they combine a consideration of notes inégales with that of “good” and “bad” notes, or quantitas intrinseca. This description remained in both later editions of the Versuch (1780 and 1798). Notes inégales is mentioned in the Klavierschule of Daniel Gottlob Türk¹⁷ as one of several techniques for varying ornamentation, somewhat the same as in Ortiz’s treatise of 1553.

Explanations of the practice of notes inégales were less frequent in the second half of the eighteenth century. In France the theorists were no longer in agreement: An unknown musician of the Comédie-Italien wrote a Noveau manual musical contenant les elements de la musique, des agrements, du chant, cited and dated as 1781 by Jane Arger,¹⁸ that does not mention notes inégales. French writers had long recognized that the practice of notes inégales was not appropriate to the performance of Italian music, which was written according to the true values of the notes. ¹⁹

Italian style invaded France in the early eighteenth century through the sonata and the cantata. Italian opera gained a resounding success with Pergolesi’s La serva padrona, Italian musicians came to France, and Italian style became familiar to French musicians. The practice of notes inégales may have lost favor as musical taste changed. Pierre Marcou wrote in 1782 that there was no general agreement on notes inégales and that its practice was better left to the individual performer.²⁰ Although it was a particularly French style of performance, it seems to have been known wherever French music and musicians were popular.

Keyboard Fingering

The study of keyboard fingering technique shows that the performer’s choice of fingers depended on the position of notes in the measure. Performers usually began their study of the instrument with simple pieces in which there was a close connection between the choice of fingers and the meter of the music. Explanations of fingering patterns in keyboard music were often illustrated by scales or scale passages derived from division-style ornamentation. This basic technique was adapted by accomplished performers when they met the demands of different styles in more difficult music. Fingering that enhanced the perception of notes grouped by twos or fours, through crossing the middle finger over either the index or ring finger, often remained a part of advanced technique.

A somewhat detached articulation of all notes is specified by a number of writers from Tomas de Sancta Maria in 1565 to Friedrich W. Marpurg in 1755. Marpurg wrote:

Sowohl dem Schleifen als Abstossen ist das ordentliche Fortgehen entgegen gesetzet, welches darinnen besteht, dass man ganz hurtig kurz vorher, ehe man die folgende Note berühret, den Finger von der vorhergehenden Taste aufhebet. Dieses ordentlich Fortgehen wird, weil es allezeit voraus gesetzet wird, niemahls angezeiget.²¹

The usual way is contrasted both to legato and staccato, and consists of lifting the finger from the preceding key just before you touch the following note. This usual way is not indicated by sign, since it is always presupposed.

Thus, a style neither legato nor exaggeratedly detached was considered usual in keyboard performance. This practice is confirmed and amplified in Türk’s Klavierschule, where clear distinctions are made between “playing in the customary fashion, neither detached nor slurred,” and articulation that is indicated by slurs or dots. In the “customary fashion,” the finger is “lifted a little earlier from the key than is required by the duration of the note,” and for slurred notes, “the finger should be allowed to remain on the key until the duration of the given note is completely past, so that not the slightest separation results.” ²²

Exceptions to this partially detached style were marked with slurs or dots, although such indications were rare until the mid-eighteenth century. The range of articulation also included super-legato fingerings, advocated by François Couperin and Jean-Philippe Rameau. This technique was done by holding one key down until after the next one was depressed, and only then releasing it.²³ Articulations for a finished performance must necessarily have encompassed a wide range of touch, from super-legato to highly detached.

Raising the finger before depressing the next key produces a minute silence between notes. Elementary fingering techniques that require the fingers (but not the thumb) to cross over one another in extended scale passages also result in silences. Such fingerings provide silences of articulation involuntarily and allow the performer to place his awareness on the fingering patterns rather than on the articulation that will result. Novice keyboard players from the sixteenth through the eighteenth centuries were taught such crossover fingerings by matching their fingers to metrically significant notes; the thumb or the fifth finger was used only at the beginning or at the end of scale passages.

Girolamo Diruta associated notes that were metrically good or bad with good or bad fingers. The good fingers were to be used on the first of paired notes and on dotted notes, and the bad fingers used on the second of paired notes and the notes following dotted notes. When a keyboard player had established these patterns of fingering, they would be the ones he would be most likely to employ in the performance of compositions that included scale patterns (Ex. 5.4).²⁴

EX. 5.4. B indicates buona (good) and C indicates cattiva (bad).

The toccatas in Il transilvano are filled with scale passages for which the repeated use of 3 and 4 on pairs of ascending notes and 2 and 3 on pairs of descending notes suggests metrical groups of two; the bad notes are connected to the good against the beat. Example 5.5 is from the intonazione nono tono by Giovanni Gabrieli.²⁵

EX. 5.5

Music suitable to novice players, such as the “Brande Champanje” in the Susanne van Soldt manuscript, sometimes specified fingering patterns that identified good and bad fingers differently than Diruta did. Articulation silences between notes bear a different relationship to the beat than in Diruta’s music. In Ex. 5.6, connections are made from good to bad notes, with silences coming just before the beat.²⁶

EX. 5.6

Harald Vogel comments on this passage:

At the beginning of the seventeenth century, the fingerings in use incorporated three different concepts. These are illustrated in the original fingerings at the beginning of the first composition in the Susanne van Soldt manuscript, the “Brande Champanje.” In the first two measures of the piece, fingerings from around 1600 are written in. In the first half of measure one, the third finger is the “good” finger, and is used consistently on the “good” notes. The second half of the same measure shows the second possibility: the playing of as many notes as possible from a single hand position, regardless of which fingers play which “good” notes. In the second measure, we find the third possibility: a combination of the first two. Most of the time the “good” third finger is placed on the “good” note, but when it is possible to play a long passage without a shift in hand position, this is done, even if it means using the fifth finger on the highest note of the passage. Thus, in order to better understand the discussion of fingerings, it is necessary that the application of “good” finger to “good” note is only part of the story. The other principle that was also considered very important was the idea of playing a figuration in one hand if possible. ²⁷

Julane Rodgers comments on sixteenth-century Spanish practice:

Scale fingerings found in the Spanish sources are of two types. The first type uses two fingers in alternation: in the right hand 3 4 3 4 ascending, 3 2 3 2 descending; in the left hand 2 1 2 1 ascending, 3 4 3 4 descending. This type of fingering seems to be the basic scale fingering and is the one more commonly used. It is perhaps older than the fingerings which use the thumb. The second type of fingering involves the use of three or four fingers consecutively, repeated as many times as necessary, such as 1234 1234. Some fingerings mix the two kinds, for instance the right hand ascending fingering, 1234 3434, which begins with four fingers consecutively, then continues alternating two fingers to the end of the passage.²⁸

Groups of either two or four notes would be followed by articulation silences, however slight, that would mark metrical groups. The selection of the fingering pattern is determined by the metrical position of the notes in relation to the tactus or to the bar line.

Robert Parkins shows that these fingerings continued to be taught in the Spanish treatises of the later seventeenth and early eighteenth centuries. In contrast to the Spanish sources, English, Italian, and Dutch treatises prefer the two-finger groups recommended by Diruta.²⁹

English scale fingerings in the early seventeenth century show a preference for 3 4 3 4 in the right hand ascending, and 3 2 3 2 descending. Left hand ascending uses 1 2 1 2, and descending 3 4 3 4.³⁰

The German keyboard performance tradition included scale fingerings that suggest grouping by pairs or by fours.

C. P. E. Bach deals with the octave scales of all twenty-four keys in the first chapter of his Versuch and thus shows signs of the pedagogic movement that was to result in the practising of multi-octave scales by all learners of keyboard instruments over at least the last century and a half. Thus it is reasonable to assume that when he gives three different fingerings for C major they should produce a similar effect, i.e. one of even touch modified by neither phrasing nor expression [Ex. 5.7]. But concerning the situation in 1715, three things in particular are still very uncertain: that scales as such were ever practiced, that the evenness required of them in later periods was already desired at that period, and that such evenness (though it may have been easier in some keys) was characteristic of all keys. If the 1 2 3 4 3 4 3 4 fingering was meant to produce an even line by means of controlled fingers, one can imagine this as a characteristic of scales as scales, not of music. If such a scale-like pattern occurred as a contrapuntal motif—as in the Orgelbüchlein chorale BWV 644—it could well be that 3 4 3 4 fingering resulted in a degree of pairing. ³¹

EX. 5.7

Practicing scales and scale exercises as a path toward virtuosic keyboard technique is not found in pedagogy until the late eighteenth and nineteenth centuries. However, since fingerings for scale patterns are indicative of elementary training in keyboard technique in the seventeenth and eighteenth centuries, we may assume that metrical groupings were embedded in every player’s fingers. This evidence says nothing about accentuation (although the clavichord could give an accent), and it does not prove rhythmic inequality, although these fingering patterns are compatible with notes inégales. These elementary techniques of fingering can be mastered so thoroughly that a player achieves remarkable smoothness, evenness, velocity, and flexibility through their use. Metrical groups may well have been apparent, but perhaps they were perceived almost subliminally by both player and audience.

The use of elementary fingerings in the performance of a large-scale composition is illustrated in Alessandro Scarlatti’s Toccata prima. This work offers a diversity of passages with fingerings; supposedly these are for the instruction of Scarlatti’s pupils, including his son, Domenico. Scale passages are in paired or four-note groupings, and many other passages are fingered so that an effect of metrical units of two or four would be difficult to avoid. It seems that the performance of this piece was intended to demonstrate a clear metrical structure resulting from the motions of the fingers. Rhetorical or lyric phrasing seems not to have been planned through fingering, but may well have been the goal of performance after the finger patterns were thoroughly mastered. ³²

Fingering patterns in a polished performance depend on much more than the elements learned through metrical fingering of scales. C. P. E. Bach stresses that every musical figure has its own proper fingering, and that new ideas in music therefore introduce new fingerings.³³ Fingerings for rhetorical passages such as quick runs leading up to high notes or down to low notes sometimes alternate the hands. This facilitates a smooth sweep of notes leading to the climax, and any kind of metrical subdivision within the run itself is ignored, as we can see in Ex. 5.8, an excerpt from Azzolino Della Ciaja’s Sonata op. IV, no. 2.³⁴

EX. 5.8. Bars 101-102 of Azzolino Della Ciaja’s Sonata op. IV, no. 2, from Sonate per cembalo, opera quarta (Rome, 1727).

Notable changes in fingering took place in the eighteenth century, many of them associated with the performance of J. S. Bach, who placed much greater demands on the technique of keyboard players than did his contemporaries. Peter Williams discusses what is known of Bach’s fingering technique, and the difficulty of interpreting his technique through his students’ writings and fin- gerings, most of which were published long after Bach’s death. The legends that Bach was the inventor of modern fingering and that he was the first performer fully to use the thumb have gained popular currency; however, these are falsifications or, at the least, exaggerations.³⁵

Comments and instructions that illustrate Bach’s keyboard technique are important for what they suggest, but they are open to question and interpretation on many levels. Fingerings, stemming either from Bach or from student copyists, illuminate specific passages rather than a basic technique. Metrical fingerings (which imply groups of two or four notes in scale-wise passages) as well as “modern” fingerings and two-hand alternations (which imply smooth unbroken passages) can all be justified in appropriate places. The musical style itself holds the essential clues on which a performer must base his interpretation, in that he must select those techniques appropriate to the nature of the music.

Daniel Gottlob Türk, organist of the Halle Frauenkirche from 1787 to the end of his life in 1813, bases the instructions in his Klavierschule on the clavichord. In this he follows a long tradition of organists who learned and practiced on that instrument. In Türk’s instructions, metrical patterns are marked by accents, not by silences of articulation. He outlines a style of fingering that makes considerable use of crossing the thumb under the longer fingers as well as crossing 3 over 4 ascending, or 3 over 2 descending in the right hand.

Putting the thumb under, and crossing over, both of which are most useful in fingering, must be practiced until they can be skilfully done without twisting the fingers and hands. Above all, even in skips, there should not be the slightest perceptible break in the legato, the keys should not be struck harder, etc. In short, one should not hear when and where the player makes use of these two devices.

Evidently, any articulation or metrical enhancement is done quite independently of fingering patterns, even though the notes were expected to be somewhat detached when not marked legato.³⁶

The treatises and music of the late eighteenth and early nineteenth centuries indicate a change of technique due to a different use of scales as technical exercises, as seen in the quotation from C. P. E. Bach cited by Peter Williams above. Czerny reported that Beethoven thought that the playing of Mozart’s time was “gehackte und kurz abgestossene”³⁷ (choppy and cut-off short), tempting us to believe that detached articulation was still a part of Mozart’s, and perhaps other late eighteenth-century keyboard players’, techniques. Czerny comments on the legato style of performance he learned from Beethoven’s approach to fingering.

Clementi advocated a legato style in articulation in 1801, perhaps influenced by Beethoven’s performance, and in contrast with the “normal” articulation advocated for keyboard performance throughout the seventeenth and eighteenth centuries. He describes the new style:

The best general rule, is to keep down the keys of the instrument the FULL LENGTH of every note. . . . When the composer leaves the LEGATO and STACCATO to the performer’s taste, the best rule is to adhere chiefly to the LEGATO; reserving the STACCATO to give SPIRIT occasionally to certain passages, and to set off the HIGHER BEAUTIES of the LEGATO.³⁸

Although an elementary keyboard technique that had endured for well over two centuries was losing favor, new ways emerged to make metrical structure evident to the listener. Both aesthetic sensibility that favored a different musical style and keyboard technique that suited the sensational new piano forte may have been involved in a change of fingerings and articulation in the late eighteenth century.

Wind-Instrument Tonguing Patterns

Tonguing patterns for wind instruments are also used to heighten the listener’s perception of musical meter. Tonguing patterns, like rhythmic inequality of equal notes, can first be found in sixteenth-century performance instructions for diminutions. The rhythmic and melodic formulas of diminutions were taught to performers in method books much like Hanon’s exercises for the piano, except that the Renaissance exercises trained both fingers and memory with formulas that were intended to be the basis for improvisation. The use of tonguing patterns was important, since every note of the diminution was expected to be articulated.

When these patterns were applied to ornamentation formulas, they fully supplied phrasings and articulations. To the modern performer, this “unmarked” music seems to require either slurring or a mixture of slurring and tonguing; however, this is a great deal of unnecessary interpretation. Richard Erig has presented a thorough summary of the wind-instrument articulations of the sixteenth and early seventeenth centuries, from which the later seventeenth- and eighteenth-century wind articulations developed:

All long notes, which in vocal pieces are also those which carry the text, are spoken with “single tonguing,” with the syllables te or de; each t or d creates a small accent, and thus a separation from the previous note, thereby setting off the notes from one another.

For the fast notes there were three types of “tonguings,” each of which joins two notes together: lere, tere, and teche (and variants thereof).

The most important of these tonguings, because of its particular gentleness, was the “lingua riversa” (lere); it is described as “dolce.” For Dalla Casa and Rog- noni, who point out its similarity to the “gorgia,” the virtuosic throat technique used in singing, it was the “prima lingua.” Rognoni writes that it is preferred by the “boni Sonatori” and does not mention the other two types. ³⁹

All three fast tonguings mentioned above group notes by twos and yet give each note a separate impulse. Rognoni’s illustrations are melodies in which tonguings are mixed, for example, sixteenth notes repeated on the same pitch are tongued te te te te, but ascending or descending sixteenths are tongued te re le re. Fantini’s trumpet tonguings even include words and sentences, for example, “da tondella butta sella,” which corresponds to a trumpeter’s battle calls.

The consonants used in the tonguing syllables undergo relatively small changes in the wind-instrument instruction books of the late seventeenth and eighteenth centuries, but the vowels e and a of earlier players are changed to ü and i in the method books of Freillon Poncein, Hotteterre, and Quantz. Erig suggests that the palatal r becomes an I in Romance languages and a d in German. For example, Quantz uses did’ll instead of der ler. ⁴⁰

The oboist Jean Pierre Freillon Poncein described the following varieties of tonguings:

Des coups de langue. & de la manière dont il faut les marquer. Il faut remarquer que generalement dans toutes les mesures où il y a trois, quatre noires ou croches, il faut marquer tu sur la premiere, re sur la seconde, & tu sur les autres, en passant la derniere un peu plus vîte que les autres, après avoir demeuré sur la précédente sur tout lors qu’elle est blanche ou d’égale valeur [Ex. 5.9].

Lorsque c’est à trois temps, il faut faire la blanche longue & la marquer tu & passer la seconde legèrement en la marquant ru [Ex. 5.10].

Aux mesures de 4 & 6, de 8 & 6, de 3 & 9, de 8 & 9, de 4 & 12, de 8 & 12, on marquera la seconde de chaque teḿs ru & les autres tu [Ex. 5.11].

Aux blanches, aux noires & aux croches, pointées de toutes sortes de mesures, il faut en user autrement car aprés fait tu sur la premiere & sur la croche qui la suit il faut faire ru sur la troisième & tu sur la quatrième [Ex. 5.12].

Les croches simples, doubles & triples lorsque le nombre est pair, c’est à dire lors qu’il n’y a point de noires pointées, ou de demy soûpir auparavant, on doit les marquer tu, tu, ru, tu, ru jusqu’à la fin; & quand le nombre est impair; tu, ru, tu, ru & continuer de même, & que celle où l’on finira soit marquée ru quand même elle seroit blanche [Ex. 5.13].

Les doubles des mesures du 2, du barré, du 4/8, 3/8, & les croches & doubles croches du 6/4 & 6/8, comme elles se passent vîte lors qu’il y en a 8, au plus on les doit marquer tu, ru, tu, ru, jusqu’à la fin, comme on le peut voir aux Ex. F [Ex. 5.14] & de même à l’égard des autres mesures qui doivent aller vîte pour avoir plus de liberté de la langue & pour les passer plus legerement. ⁴¹

Tonguing and how it is done: Generally in all meters containing three or four quarters or eighths, the first is played tu, the second ru, and tu on the others. The last is played a little faster than the others after having rested a little longer on the note preceding it, especially when it is a half note or a note of equal duration [Ex. 5.9].

EX. 5.9

In a meter of three beats, play the half note long, say tu and play the second quickly saying ru [Ex. 5.10].

EX. 5.10

In 6/4, 6/8, 9/3, 12/4, 12/8 measures the second (note) of each beat is marked ru and the others tu [Ex. 5.11].

EX. 5.11

For dotted half notes, dotted quarters and dotted eighths in all measures, tonguings are used differently. After using tu on the first and on the eighth that follows it, ru must be used on the third and tu on the fourth [Ex. 5.12].

EX. 5.12

When there are even numbers of eighths, sixteenths, and thirty-second notes, that is, when there are no dotted quarters or eighth rests before them, the tonguings should be tu tu ru tu, ru [beginning the next group], to the end; and when there are uneven numbers of them; tu ru tu, ru [beginning the next group] and continuing the same, even though the last note would be tongued ru if it were a half note [Ex. 5.13].

EX. 5.13

If they are played quickly and if there are as many as eight of them, the sixteenths in the measures of 2, , 4/8, 3/8, and the eighths and sixteenths in 6/4 and 6/8, use tu ru tu ru, until the end, as is shown in example F [Ex. 5.14].

EX. 5.14

The same should be followed in other measures played fast to have more liberty for the tongue and to be able to play more quickly.

Modern performers use a “double tonguing,” te ke (equivalent to tu ru), in fast passages when it is not possible to tongue te repeatedly and quickly. In present-day performance the ideal is to pronounce te and ke in such a way that no difference can be detected by the listener. However, there is no reason to suppose that eighteenth-century performers tried to eliminate the pronunciation difference between tu and ru.

Freillon Poncein’s tonguings provide for a differentiation between notes through articulation in addition to facilitating the performance of fast passages. Tu causes the reed of the oboe to be articulated by the tip of the tongue and ru causes the column of air to be stopped by the middle of the tongue. Most modern oboists cannot articulate notes except by touching the tip of the tongue to the reed, and therefore can use only tu. Both tongue strokes can be done on the flute and recorder, with good effect. Tu is precise, although it can be done very softly and gently, and ru may be strong but not as precise in its initial impact as tu. By accepting the difference between tu and ru, or even heightening it, Freillon Poncein’s tonguings reinforce metric grouping.

Hotteterre also advocates tu and ru tonguings, although he frequently uses them differently than Freillon Poncein. Hotteterre’s method is primarily for the flute, and Freillon Poncein’s for the oboe. The important point is that there is a difference between tu and ru, not that they are used in the same patterns.

Pour rendre le jeu plus agréable, & pour éviter trop d’uniformité dans les coups de “Langue, on les varie en plusieurs manieres; Par exemple on se sert de deux articulations principales; Sçavoir, tu & ru. Le tu est le plus en usage, & l’on s’en sert presque par tout; comme sur les Rondes, les Blanches, les Noires, & sur la plus grande partie des Croches; car lorsque ces dernieres sont sur la même ligne, ou qu’elles sautent, on prononce tu. Lorsqu’elles montent ou descendent par degrez conjoints, on se sert aussi du tu, mais on l’entremêle toûjours avec le ru, comme l’on peut voir dans les Exemples cy-après [Ex. 5.15], où ces deux articulations se succedent l’une à l’autre.

On doit remarquer qu le tu, ru, se reglent par le nombre des Croches. Quand le nombre est impair on prononce tu ru, tout de suite, comme l’on voit au premier Exemple. Quand il est pair on prononce tu, sur les deux premières Croches, ensuite ru alternativement, comme l’on voit dans le deuxième Exemple. On fera bien d’observer que l’on ne doit pas toûjours passer les Croches également & qu’on doit dans certains Mesures, en faire une longue & une breve; ce qui se regle aussi par le nombre. Quand il est pair on fait le premiére longue, la second breve, & ainsi des autres. Quand il est impair on fait tout le contraire; cela s’appelle pointer. Les Mesures dans lesquelles cela se pratique le plus ordinairement, sont celles à Deuxtemps, celle du triple simple, & celle de six pour quatre. On doit prononcer ru, sur la note qui suit la Croche quand elle monte ou descend par degrez conjoints [Ex. 5.16]. ⁴²

In order to make the performance more agreeable and to avoid too much uniformity in tonguings, they are varied in several ways. For example, there are two principal articulations used: tu and ru. Tu is most frequently used and everywhere, on whole notes, halves, quarters, and on the greater number of eighths, as when these last are [repeated] on the same line or when they move disjunctly. When they rise or descend conjunctly tu is also used, but it is mixed with ru, as can be seen in the example following [Ex. 5.15], where these two articulations come one after the other.

EX. 5.15

It must be noted that tu, ru is governed by the number of eighths. When the number is odd, tu ru is pronounced immediately, as in the first example. When it is even, tu is pronounced on the first two eighths, then alternatively with ru, as in the second example. It is well to observe that the eighths should not always be played as equal, in certain measures one must be long and the other short; this also is determined by the number. When it is even, the first is long, the second short, and so on with the others. When it is odd, they are done entirely the opposite, this is called pointer. The measures in which this is most usually done are the Deuxtemps (2) the triple simple (3 or 3/4) and the 6/4. Ru must be pronounced on the note that follows the eighth, when it rises or descends by conjunct motion [Ex. 5.16].

EX. 5.16

Directions for playing notes inégales were interspersed with those for various articulations. This proximity heightens the impression that they were considered useful in much the same way, even though no specific comment to this effect was made.

Hotteterre’s articulation of the 3x2 measure, marked 3/2, should be considered in contrast to the 2x3 measure illustrated in Ex. 5.16, marked 6/4:

Dans la Mesure du Triple double, on prononce tu, ru, entre les Noires, & ru, sur la Blanche qui est precedee d’une Noire, en montant ou en descendent, par degrez conjoints [Ex. 5.17].⁴³

In the triple double [3/2] measure, tu, ru is pronounced on the quarter notes, and ru on the half note that is preceded by a quarter note that rises or descends conjunctly [Ex. 5.17].

EX. 5.17

This example provides a means of differentiating between these two measures that supplements the articulations mentioned by Butler, Malcolm, and Rousseau.

Quantz’s Versuch discussed many different tonguing patterns, including ti, di, di ri, ti ri, and did’ll. Quantz specifically associated ti ri with the passages that “are played with some inequality,” and ti is said on the shorter, ri on the longer note. “Bey diesem Wörtchen tiri fäll der Accent auf die letzte Sylbe” (the accent in the word tiri falls on the last syllable), so that ri is said on the good note and ti on the bad. The resulting grouping is against the beat, from arsis to thesis. Did’ll has the opposite accentuation, with the accent falling on the first syllable. Most oboists cannot use did’ll because a relatively stiff oboe reed requires that the tongue come in contact with it in order to articulate sounds. ⁴⁴

Quantz is one of the first to discuss extensively the subject of slurs or ties in woodwind performance. Only the first of a group of slurred notes is tongued, using di rather than ti. If a strich or staccato mark is used on the note preceding a slur, the syllable ti is used on both that note and the first of the slurred notes. If passage-work must be played more quickly than tiri can be tongued, then either the first two notes or the last two notes of groups of four should be slurred together and the other two tongued. ⁴⁵

The most extensive use of slurs in Quantz’s musical examples occurs in the musical illustration for chapter XIV, “Of the manner of playing the adagio.” Slurs are used to link groups of notes in his Italianate, division-style, florid ornamentation and seem to be included as an illustration of how to perform the written-out melismas. The unornamented version of the adagio melody has conspicuously few slurs.⁴⁶

Articulation patterns that tongue every note in a passage were taught to performers throughout the eighteenth century by many later editions of Hotteterre’s principes⁴⁷ and Quantz’s Versuch.⁴⁸ However, a number of flute method books beginning about 1735 taught other tonguings and different patterns of articulation.

Corrette’s flute method (ca. 1735) disdained the use of the syllables taught by Hotteterre and suggested that flutists imitate the articulation of violin bow strokes. Indications of phrasing originally used for violin bowings were adapted for the flute, such as a dot over the note indicating staccato. The dot over notes had been used in French music to indicate that the notes were not played unequally, but now it came to mean a short articulation. Delusse (ca. 1761) described an imitation of the violinist’s “slurred staccato” as the “pearl stroke”; every note was to be articulated with the individuality and perfection of a pearl on a string. Delusse also advocated articulation by impulses of the breath alone, which he indicated by “hu, hu” beneath the notes. On repeated notes, this articulation was notated by a combination of dots over the notes with a slur connecting them all together.⁴⁹

Devienne (1795) recommends an alternation of two notes slurred and two notes tongued in fast passages when articulation is not specified, a practice that was first mentioned by Quantz for fast passages. It seems to have become popular with performers in the late eighteenth century and continues to be used extensively today.⁵⁰

From this brief survey of woodwind articulations, it is evident that wind players applied many patterns of articulation to the apparently unphrased music that we see in original notation. The variety of both patterns and tongue strokes provided many different shadings of articulation. The early eighteenth- century manuals teach articulations in order to group notes and define measure organization, rather than to heighten particular melodic ideas or introduce a dramatic effect, as modern articulations often do. The various systems of tonguing patterns differ in specific details but testify to the importance attached to distinguishing between notes according to their relative position in the measure.

Stringed-Instrument Bowings

Wonderfully rich and varied possibilities for shaping and inflecting sound are available to the players of bowed stringed instruments. These include shadings of tone quality, a wide dynamic range, and control over intonation as well as techniques of articulation to clarify metrical structure.

The most basic and elementary distinction made in using the bow is whether it is drawn up or down. On the violin, this entails moving the weight of the arm, hand, and bow by or against the force of gravity, and on the viola da gamba, moving the bow toward or away from the body.

The basic rules for guiding the choice of up- or down-bow strokes depend on the metrical position of notes in the measure. The down-bow stroke on the violin is stronger, more precise and definite in effect, and the up-bow stroke is lighter and more delicate. “Up-bow” on the viola da gamba, actually a stroke from right to left, toward the body, is the equivalent of down-bow on the violin, and “down-bow,” from left to right, is equivalent to up-bow on the violin. This is obvious to players, of course, and was mentioned by Marin Mersenne as a clear distinction in technique between the instruments.⁵¹

Rules for choosing up- or down-bow on the viola da gamba are given by Jean Rousseau⁵² and summarized and discussed by Hans Bol.⁵³ Rousseau’s rules were not new at the time but are a summary of bowing practices that can be seen in the examples of Ganassi in the sixteenth century⁵⁴ and in the descriptions of Mersenne,⁵⁵ Simpson,⁵⁶ and Danoville.⁵⁷ These are remarkably similar to the rules for choosing violin bowings given by Georg Muffat, derived from the practice of Lully’s performers, except for the opposite direction of drawing the bow.

The differences between the advanced techniques of the viola da gamba and the violin are many and subtle and deserve careful delineation beyond the scope of this chapter. Let us consider elementary bowing technique in relation to the violins in the later part of the seventeenth century.

Violin technique was discussed in the writings of J. A. Herbst⁵⁸ and Gasparo Zannetti,⁵⁹ and by Georg Muffat late in the seventeenth century.⁶⁰

Violinists distinguished between a great variety of possible bowings, although the effect of the down-bow or up-bow was the most basic. The “Rule of Down-Bow” was best defined in detail by Georg Muffat in the preface to Florilegium secundum:

However it is well known that the Lullists, whom the French, the English, those from the Low Countries, and many others follow, all observe an identical way of bowing, even if a thousand of them play together. They all observe the same way of playing the principal notes in the measure, above all those that begin the measure, those that define the cadence, and those that most clearly emphasize the dance rhythm. . . .

1. The first note in each measure, when there is no rest or pause, should be played down-bow, regardless of its value. This is the principal and almost indispensable rule of the Lullists, on which almost the entire secret of bowing depends and which differentiates them from the others. All subsequent: rules depend on this rule. In order to know how the other notes fall into place and are to be played, one must attend to the following rules.

2. In common time, which the theorists call “tempus imperfectum,” the measure is divided into equal parts. Odd-numbered notes (1, 3, 5, etc.) are played down-bow, and even-numbered notes (2, 4, 6, etc.) should be played up-bow [Ex. 5.18]. This rule applies also in triple meter, or any meter when the beats are diminished equally by half. I call diminutions all those notes that are faster than those values indicated in the time signature [Ex. 5.19]. This way of counting equal divisions of beats is similarly observed if rests of the same value appear instead of notes [Ex. 5.20]. All the finest masters agree readily with the French on this second rule.

EX. 5.18

EX. 5.19

EX. 5.20

3. Since, according to the first rule, the first note in the measure is down-bow, the second of three equal notes (which comprise a complete measure in triple time) is always up-bow, and the third is once again down-bow, at least when one plays rather slowly; therefore in beginning the measure following, one must play down bow for the second time in succession [Ex. 5.21]. More often, however, the second and third notes are played in the same up-bow stroke, divided distinctly in two. This is called craquer. It allows the measure to go a little faster with greater ease [Ex. 5.22].

EX. 5.21

EX. 5.22

4. In Proportione Sextupla, the measure is divided into two basic parts [Ex. 5.23]. In Proportione Nonupla the measure is divided into three parts [Ex. 5.24]. In proportione Duodecupla the measure is divided into four parts [Ex. 5.25]. In these distributions, each part contains three of those values indicated in the time signature. The first of the three equal notes is almost always played down-bow, for a clearer sound, even if the group does not begin the measure, and the two others are played in an up-bow stroke, divided in two. If there is a rest instead of the first note, the following note should unquestionably be played down-bow in triple measures (a) and other proportions (b) [Ex. 5.26].

EX. 5.23

EX. 5.24

EX. 5.25

EX. 5.26

5. When several notes, each of one measure duration, appear in succession, each should be down-bow [Ex. 5.27]. In six, or in twelve, several successive notes of equal value should be played alternately down-bow and up-bow according to whether they fall on an odd or even note, as explained in the second rule [Ex. 5.28]; but in nine they follow the first aspect of the third rule (triple meter) [Ex. 5.29].

EX. 5.27

EX. 5.28

EX. 5.29

6. Several equal successive syncopated notes usually require down-bow and up-bow alternatively. This is all concerning notes of like value [Ex. 5.30].

EX. 5.30

7. As far as mixed note-values are concerned, the first of the smaller values following longer values should be counted the odd number, so that it can be played down-bow, if it should come out that way [Ex. 5.31], or divided up-bow, if it should occur this way [Ex. 5.32]. The first two beats of smaller value are played with divided up-bows [Ex. 5.33]. If other smaller values follow after that, they are played up-bow and down-bow alternately. As far as pauses and rests are concerned, they can be counted as notes [Ex. 5.34].

EX. 5.31

EX. 5.32

EX. 5.33

EX. 5.34

8. When subdivisions of the measure consist of three notes, and the first has a dot after it, it is ordinarily down-bow [Ex. 5.35].

EX. 5.35

9. Several successive notes, each completing the measure (or subdivision) after a pause or rest, should be down-bow and up-bow alternately, regardless of the said rest or pause [Ex. 5.36].

EX. 5.36

10. The little note by itself before the beginning of the measure (20a) as well as the one that passes quickly after a dot or after a short rest (20b) and likewise the smaller note that follows a larger syncopated one, (20c) should always be up-bow [Ex. 5.37]. If the longer syncopated note is also up-bow, it will be necessary to divide the up-bow stroke in half, thereby adding the following note to it (indicated by) [Ex. 5.38].⁶¹

EX. 5.37

EX. 5.38

The use of bowings, therefore, to establish metric relationship among notes was basic to this school of violin playing. This bowing discipline allowed any number of violinists to play in unanimity in orchestral ensembles. It seems likely that the differences between the sound of down-bow and up-bow were subtle, as is the case with wind articulations, so as not to overwhelm the flow of the music. The effect of different bow strokes on an individual instrument is magnified by the unison of an orchestral section.

Corelli’s musicians in Rome sometimes numbered as many as one hundred and fifty. Georg Muffat commented that the concerti were “performed with the utmost accuracy by a great number of instrumentalists.”⁶² Boyden comments that while Muffat’s bowing directions are primarily for the performance of dance music, “it is likely that at first the sonata players adopted the same basic principles . . . considering dance bowings carefully before rejecting or amplifying them. We do not know in what respects, if any, the sonata players before 1650 deviated from the basic bowing principles of the dance violinists.”⁶³ As there are no treatises that consider Italian bowing technique in the late seventeenth and early eighteenth centuries, for all we know, the “Rule of Down- Bow” could have been used to produce the unity of Corelli’s orchestra as well as of Lully’s.

French violinists seem to have retained Muffat’s “Rule of Down-Bow.” Dupont⁶⁴ and Montéclair⁶⁵ continue to teach the basic rules but give more examples in which exceptional bowings are explored. Marc Pincherle comments that Corrette (1738), Leopold Mozart (1756), and the Abbé le fils (1772) continue, at least in part, to teach the “Rule of Down-Bow.”⁶⁶

Although bowing directions similar to Muffat’s are found in Quantz’s Versuch, he remarks that “present musical writing” requires an equal strength from the up-bow and the down-bow in orchestral violin playing.⁶⁷ Perhaps this remark should be taken to mean that before “present musical writing,” upstroke and downstroke were not even; that is, that Muffat’s bowings did not give the effect of evenness. Leopold Mozart also comments that upstroke and downstroke must be played alike in order to achieve the results he desires.⁶⁸ His bowings are almost identical to those of Muffat.

Referring to scale passages, Geminiani (1751) warns: “There it must be observed, that you are to execute them by drawing the bow up and down, or down and up alternatively; taking care not to follow that wretched Rule of drawing the Bow down at the first note of every Bar.”⁶⁹ Geminiani continues:

So in playing Divisions, if by your Manner of Bowing, you lay a particular Stress on the Note at the beginning of every Bar, so as to render it predominant over the rest, you alter and spoil the true air of the piece, and except where the composer intended it, and where it is always marked, there are very few Instances in which it is not very disagreeable.⁷⁰

Perhaps the proper assumption is that only an inept violinist would mechanically accent the first note of every bar. Stress or accent was mentioned but not featured in descriptions of measure organization in the first half of the eighteenth century, although many instruments were capable of it. Geminiani’s warning against accent seems in accord with the general attitude of the time.

David Boyden summarizes the elementary technique of both violin and viola da gamba bowing in the seventeenth and early eighteenth centuries as the “Rule of Down-Bow,”⁷¹ although for the viola da gamba it should be called the “Rule of Up-Bow.” He equates both the down-bow stroke and good notes with accent or stress, despite the care used by writers in the earlier sources not to use the words “accent” or “stress.” In order to facilitate his explanation, Boyden has adopted the usual twentieth-century idea that meter is identical with stress. Unfortunately, this interpretation distorts the effect and intent of bowing technique in the seventeenth and eighteenth centuries.

The variety of individual bow strokes discussed in late eighteenth-century treatises is much greater than those mentioned in the seventeenth-century treatises of Zannetti, Herbst, and Muffat. If the discipline of the “Rule of Down-Bow” was preserved in the late eighteenth century, it was interpreted with a new richness in variety of sounds, colors, and nuance. Robin Stowell describes three strokes: the staccato stroke (called détaché in French treatises), the “bounding” stroke (called sautillé, spiccato, or “flying staccato”) for bravura passages, and the legato bowing strokes that imitate the voice. Many kinds of slurred bowings were used as well as special strokes such as bariolage (playing the same note alternately on two strings) and ondeggiando (moving the bow back and forth across two or more adjacent strings).⁷² All of these techniques were described and used on the bows preceding the introduction of the more rigid and powerful Tourte bow in 1761.

The viola da gamba bow stroke, whether up or down, was capable of much subtlety, and John Hsu has distinguished six different bow strokes that were mentioned by Hubert Le Blanc⁷³ and described in some detail by Etienne Loulié.⁷⁴ The basic stroke is divided into three parts, the première impression (which gives almost the effect of plucking the string), a prolongation of the sound, and the release of the tension of the bow on the string, which results in the sound’s fading. Loulié’s other bow strokes expand the volume of sound after the première impression, sustain the sound at the same level, stop the sound after the première impression, go quickly into the preparation for the next note, or articulate more than one note while moving the bow in the same direction. ⁷⁵

The employment of such a repertory of articulations, based on the practice of Marin Marais and Antoine Forqueray, gave great variety and subtlety to the performance of the French school of viola da gamba players in the late seventeenth and early eighteenth centuries.

Elementary bowing practices agree fundamentally on the importance of enhancing the metrical structure of music in performance. The “Rule of Down- Bow” is clearly identified with musical meter, and varieties of bowings are tied to various time signatures. The results from even the most meticulous recreation of performance based on the “Rule of Down-Bow” are dependent on many kinds of up- or down-bow strokes, many ways of continuing the sound after the initial moment, and different ways of terminating the sound. Almost all pedagogical or theoretical discussions of musical performance are only sketches; imagination must infuse them with the subtlety of a finished performance.

Engramelle’s La tonotechnie

The effect of bowings, fingerings, and tonguings is difficult to specify, since it probably varied from performer to performer and certainly varied according to the style of the music. However, there is one source of information about performance that is uniquely precise about the effect of metrical articulation. Directions for making a mechanical organ in Father Marie-Dominique-Joseph Engramelle’s La tonotechnie (1775) give such accurate information about performance that it has been possible to make a recording of the melodies he describes.⁷⁶ The mechanical organ cylinder illustrates two aspects of quantitative articulation with absolute clarity: longer and shorter durations of equal notes, and notes inégales. Perhaps Engramelle’s directions can illuminate some of the less precise instructions of other musicians.

Engramelle goes into more detail than the usual descriptions of performance technique because such specificity is required in order to record music for a mechanical organ, where pins, or bridges linking pins, are embedded in a rotating cylinder to trip the keys. In the middle of plate 6 there is a drawing of the cylinder, below which are various sizes of pins and bridges and a gauge for measuring them. Plate 7 shows the cylinder in place with a handle to turn it, the gears that engage with the cylinder, and a dial that measures the space on the cylinder. The noteur arranged the pins by measuring the space between notes according to divisions of the dial.⁷⁷ Plate 8 is the frontispiece from La tonotechnie, which shows two craftsmen, one working on a serinette (canaryorgan), the other working on a larger cylinder. The room is filled with mechanical musical instruments, including a mechanical harpsichord, Vaucanson’s mechanical flute player, a mechanical organ (over the doorway), and a musical clock.

Forty turns of the serinette’ s handle make one complete revolution of the cylinder, and each turn of the handle may be divided into as many parts as necessary to mark each note and the subdivisions required for silences of articulation and for ornaments. Each subdivision of the dial, a module, is equivalent to the smallest note value or silence in the music, usually the length of one of the battements in an ornament. The length of the piece, its tempo, and its character or expression determine what number of divisions of the dial is most suitable.

Engramelle uses the singing voice as his primary model for articulation in notage by simulating silences in pronunciation before hard consonants, the articulation of a syllable without a tongue-stroke, and pauses to take breath. These are mixed with articulations that are purely instrumental in concept, such as the separation of notes from one another required by raising the finger from one key in order to strike the next:

Si, par exemple, il se rencontroit deux noires de suite, dont la seconde dût être frappée d’un coup de langue, la première de ces deux noirs n’étant tenue qu’à demi, ou ne parlant que pendant la première moitié de sa valeur, l’autre moitié restant en silence, seroit un détaché qui équivaudroit à un coup de langue sur la seconde de ces deux noires; & plus on augmenterait le silence de la première, plus le coup de langue sur la seconde seroit ressenti & détaché.

PLATE 6. Serinette cylinder and tools. (plate CIV from L’Art du facteur d’orgues by Dom Bédos de Celles.)

PLATE 7. Serinette. (Plate CIII from L’Art du facteur d’orgues by Dom Bédos de Celles.)

PLATE 8. Frontispiece of Engramelle’s La tonotechnie.

Si au lieu de deux noires, il se rencontroient deux croches, dont la seconde dût être marquée par un coup de langue, en ne faisant qu’une tactée de la première, le silence qui resterait à la suite produiront sur la seconde l’effet d’un coup de langue.

Aux notes tenues, qui ne doivent pas fournir des coups de langue sur leurs suivantes, leur silence doit être plus court pour ne faire qu’un simple détaché, afin qu’elles ne se confondent pas avec celles qui les suivent immédiatement; ce silence de détaché doit être de la valeur d’une double croche, à moins que ces notes ne soient marquées liées; pour lors leur silence ne seroit que d’une triple croche ou environ.

Pour les reprises d’haleine, les silences sont ordinairement d’une noire pointée, même quelquefois de la valeur d’une blanche, suivant l’espace qui se trouvera depuis le commencement d’une note à l’autre, ce qui ne se peut déterminer que consé- quemment au genre d’expression qui convient à la piéce de musique qu’on veut exécuter. ⁷⁸

If, for example, the second of two successive quarter notes is to have the effect of a tonque-stroke, the first of the two should be held only half of its value with the other half silent, this makes the second detached in a manner equivalent to having a tongue-stroke. The greater the silence after the first note, the more the tongue- stroke is perceived as detaching the first note from the second.

If, in place of two quarter notes we have two eighths, of which the second is to be marked by a tongue-stroke, by making the first tactée, the following silence produces the effect of a tongue-stroke on the second.

For held notes not preceding tongue-strokes, the silence should be shorter to give the effect of a simple detached note that does not merge with the following note. This silence of detachment should be the value of a sixteenth note, at least if the notes are not marked slurred. If they are slurred the silence is no more than a thirty-second note, or thereabout.

For taking a breath, silences are ordinarily a dotted quarter note or sometimes a half note, according to the space found between the beginning of one note and the next. This can be determined only as a consequence of the kind of expression suitable to the piece of music one wishes to perform.

Without exception, says Engramelle, every note is partly sustained and partly silent, but notes may be identified as either tenues or tactées by the duration of the sustained part. A tactée is barely “touched” ; only the very first part of the note is sounded and the remainder is silent. A tenue, on the other hand, sounds through half or more of the length of the note, followed by a short silence. In the ordinary performance of a succession of eighth notes, and some- tunes quarters or sixteenths, tenues and tactées alternate in order for two-note groups to be discerned as though they were separate syllables.

In Engramelle’s musical examples, reproduced in plates 9, 10, and 11, marks are added above the notes to illustrate these principles to the noteur. Tenues are symbolized by a horizontal line and tactées by a vertical stroke. Longer pauses are indicated by dots over the tenue lines. In the Menuet de Zelindor (see plate 9), each dot is equivalent to a sixteenth-note triplet. The first of two tied notes is followed by only a thirty-second note silence, and the second note is tactée. All of the eighth notes are performed as equal.

PLATE 9. Music examples 1—4 from Engramelle’s La tonotechnie.

Ornaments are indicated by abbreviated versions of common French signs for agréments. With one important exception, the melodic shape of Engramelle’s ornaments agrees with French agréments in use from the rime of D’Anglebert. The exception is an ornament that begins and ends on the main note and alternates with the upper neighbor note, called a tremoletto by the Italians in the seventeenth and eighteenth centuries, and sometimes called an “inverted mordent” today. Engramelle calls it a martellement.

It is interchangeable with an ornament that alternates between the main note and its lower neighbor, a familiar pattern in the French music of the early eighteenth century. In his description of Mr. Balbastre’s Romance, in L’art du facteur d’orgues, Engramelle remarks on this ornament: “La premiere note de Ia seconde mesure est une blanche sur le Ia, martelee avec le si; on pourroit aussi Ia marteler avec le sol dieze; on choisira, elle seroit egalement un bon effet” (The first note of the second measure is a half-note a, ornamented with ab. It could also be ornamented with a g#, the choice is open, and it would sound as well).⁷⁹

PLATE 10. Music examples 5—8 from Engramelle’s LA tonotechnie.

Each note in the quick alternations of agréments is ordinarily equal to one division on Engramelle’s dial, or one module. In the Menuet de Zelindor, each module is equal to a single sixteenth-note triplet. Usually the number of battements in the ornament allows sufficient time for the main note to sound, followed by the necessary silence before the next note. However, cadential trills sometimes require more battements than there are modular divisions in the note. The usual method of noting these is temporarily to decrease the space on the cylinder for each module, so that they will fit into the available time.

PLATE 11. Music examples 9—12 from Engramelle’s La tonotechnie.

An alternate method of accommodating florid ornaments gives each module of trills a full division of the dial. This results in a ritardando, which En- gramelle employs for occasional final cadences, for instance, the cadential trill of the second part of the Menuet de Zelindor, where two or three sixteenth- note triplets are added to the duration of the penultimate measure. If a ritardando is used, the spaces on the cylinder cease to correspond to the numbers on the dial, and the exact meter of the piece is disturbed. Engramelle omits the numbers of the dial below the music at such points, as in La Marche du Roy, Menuet de Zelindor, and the Romance. You may wish to hear the recording of Le Menuet de Zelindor at this point.

Engramelle describes his analytical method for determining the silences of instrumental articulation:

Pour se convaincre de la nécessité de ces silences à la fin de chaque note, qu’on exécute sur un orgue, un clavessin, épinette, ou toute autre instrument à clavier que ce soit tel air qu’on voudra, & qu’on l’exécutant on fasse plutôt attention à l’exécution qu’à la manière dont on le note sur le papier, on s’appercevra qu’un doigt qui vient de finir un note, est souvent levé long-tems auparavant que de poser le doigt pour le note suivante, & cet intervalle est nécessairement un silence, et si l’on y prend bien garde, il se trouvera entre toutes les notes de ces intervalles plus ou moins longs, sans lesquels l’exécution seroit mauvaise: il n’est même pas de modules de cadences qui ne soient séparés par des très-petits intervalles très courts entre la levée & la pose de doigts sur les touches: ce sont tous ces intervalles plus ou moins longs, que j’appelle les silences d’articulation dans la Musique, dont aucune note n’est exempte, pas plus que la prononciation articulée des consonnes dans la parole, sans lesquelles toutes les syllables n’auroient d’autre distinction que le son inarticulé des voyelles.

Un peu d’attention dans la prononciation sur l’articulation des syllables sera appercevoir aisément que, pour produire l’effet de presque toutes les consonnes, le son des voyelles se trouve suspendu & intercepté, soit en rapprochant les levres l’une contre l’autre, ou en rapprochant la langue contre le palais, les dents, &c. toutes ces suspensions ou interceptions du son des voyelles sont autant de petits silences qui détachent les syllables les unes des autres pour former l’articulation de la Musique, à la différence près que le son d’un instrument étant partout le même, & ne pouvant, pour ainsi dire, produire qu’une seule voyelle, il faut que les silences d’articulation soient plus variés que dans la parole, si l’on veut qu’elle produise une espèce d’articulation intelligible & intéressante. ⁸⁰

In order to be convinced of the need for these silences at the end of each note, play whatever you wish on an organ, harpsichord, spinet, or any other keyboard instrument, and while playing, pay more attention to the performance than to the way it is noted on the page. It will be seen that after playing a note, a finger is often lifted a long time before it is placed for the following note. This interval is necessarily a silence, and if you notice it, you will find it between all notes in smaller or greater measure. Without these the performance is bad; even the notes of trills are separated by very short intervals between raising and lowering the fingers on the keys. These are the longer or shorter intervals that I call “silences of articulation” in music. No note is exempt from this rule, any more than is the pronunciation of consonants. Without these, all syllables would be no more distinct than the inarticulate sounds of vowels.

With a little attention to the articulation of syllables it is easily seen that in order to give the effect of almost all consonants, the sound of vowels is suspended and interrupted either by the lips coming together or by pressing the tongue against the palate or the teeth, etc. All these suspensions or interruptions are silences separating syllables from one another in order to articulate a word. It is the same in the articulation of music, with the difference being that the sound of an instrument is always the same, so to speak, and not able to produce more than one vowel. It is necessary that silences of articulation be more varied than in speech if they are to be intelligible and interesting.

Engramelle minutely considers how notes inégales are performed. He distinguishes between firsts, the initial note of a pair, and seconds, in successive eighth notes. Firsts are tenue and seconds are tactée, so that there is the effect of a tongue-stroke on all firsts. In addition, he advises noteurs:

Il est une observation essentielle à faire sur les croches qui s’exécutent souvent inégalement de deux en deux. Les papiers notés ne nous indique pas quelle est la valeur de cette différence; si elle est de la moitié du tiers ou de quart: il seroit cependant essential de la fixer; car si l’on suivoit exactement dans le notage la valeur des croches comme elles sont indiquées sur les papiers de Musique, elles seroient toutes égales en durée, en tenue & en silence, & c’est ce qui arrive rarement dans l’exécution, laquelle doit être la régie invariable du Noteur des cylindres: ainsi lorsque cette inégalité entre les croches doit avoir lieu dans l’exécution, pour soutenir le genre d’expression qui convient à l’air, il faut que le Noteur puisse l’apprécier pour la rendre comme il faut, ce qu’il ne peut faire que par l’exécution elle-même.

Ces croches inégales se marquent de deux en deux, c’est-à-dire, de noire en noire; les deux ensemble faisant la valeur d’une noire entière, dont la première qui fait la première partie de la noire, est plus longue, & la seconde qui fait l’autre partie de la noire, est plus courte: mais quelle est cette différence de la plus longue à la plus courte? Voilà la difficulté qui arrête le plus communément les Noteurs de cylindres.

Il est des cas où cette différence est de la moitié, ensorte qu’il faut exécuter les premières comme si elles étoient croches pointées, & les secondes doubles croches: d’autres où la différence est d’un tiers, comme si la première valoit deux tiers de noire, & la seconde l’autre tiers, d’autres enfin où cette différence, moins sensible, doit être comme de 3 à 2; ensorte que la première vaudra 3 cinquièmes de noire & la seconde 2 cinquièmes. On trouvera dans le chapitre XXV, en parlant du détail des airs, plusieurs observations sur la différence des premières & des secondes croches, qui suffiront en saisissant bien le mouvement nécessaire au genre d’expression qui se fait sentir dans l’exécution des piéces, pour apprécier au juste la différence dont je parle, laquelle est souvent dans le cas de varier dans le même air, si l’on veut exprimer certains passages d’une manière plus intéressante.⁸¹

There is an essential observation to make about eighth notes, which are often performed unequally, two by two. Notated music does not tell us what the value of this difference is, whether it is of half, one third, or one fourth. It is necessary, however, to determine this value because if in notage one follows the value of eighths exacdy according to the musical score, they would all be entirely equal in duration, sound, and silence. This happens rarely in performance, and it should be the invariable rule of noteurs that when inequality between eighth notes occurs in performance, in order to maintain the kind of expression suitable to the air, they must appraise it correctly in order to include what is done in the actual performance itself.

These unequal eighths are evident two by two, that is from quarter to quarter, the two eighths together make up the value of a quarter note. The first, which occupies the first part of the quarter note, is longer, and the second, which occupies the other part, is shorter. But what is this difference between the long and the short? This is the problem that usually stops most noteurs of cylinders.

Sometimes this difference is half, so that the firsts must be performed as though they were dotted eighths and the seconds sixteenths. Other times the difference is a third, as though the first were two-thirds of the quarter and the second one-third; and finally other times when the difference is less perceptible, it should be as 3:2, so that the first equals three-fifths of the quarter and the second two- fifths. In speaking of the details of airs in Chap. XXV, there are many observations on the differences between first and second eighth notes. These are sufficient to allow one to find the rhythm necessary to the proper expression of each piece, and to perceive the differences of which I speak. These often vary in the same air when one wishes to perform some passages in a more interesting way.

This information is considerably more exact than that in any other source. Engramelle’s directions for the notage of the twelve examples at the end of La tonotechnie are even more precise. The “least perceptible” inequality mentioned above, the ratio of 3:2, is given as 7:5 in Menuet no. 6 and in Le Bûcheron no. 7. The ratio of 9:7 is discussed for the inequality of eighth notes in La Barcelonette, a tune Engramelle gives as an example in L’Art du facteur d’orgues (Quatrième Partie, planche CXIV). It is instructive to hear inequality in the precise ratio of 7:5, as in the eighth notes of the second recorded example, Menuet no. 6, with first eighth notes tenue and second eighths tactée.

Engramelle’s comment that the degree of inequality may vary in a performance is the only written indication of this known to me, although such variation might well occur subconsciously as well as through a performer’s intent. It is possible that changing the degree of inequality was thought by most writers to be too obvious to mention.

Since many French writers state that inequality is confined to stepwise motion in equal eighths,⁸² perhaps arpeggiated passages, such as those in Menuet no. 6, should be performed as equal. While this approach is not a variation in the degree of inequality, mixing equal with unequal eighth notes produces an interesting contrast in the performance of these notes. The third recorded example is of Menuet no. 6, with the arpeggiated passages performed as equal.

One might suppose that inequality would become more pronounced approaching a cadence or an important point of a phrase. The ratios in measures 4, 7, 12, and 20 of Menuet no. 6 in the fourth recorded example change by pairs of eighths from 7:5 to 3:2 to 2:1. The arpeggiated eighths remain equal.

Engramelle states that the degree of inequality depends on the piece:

J’ai observé en notant des cylindres, qu’il est nombre de marches, entr’autres celle du Roi de Prusse, où la différence des premières aux secondes croches est de la moitié, comme 3 est à 1, c’est-à-dire, que les premières croches valent des croches pointées, & les secondes croches des doubles croches. Dans certains menuets, entr’autres le petit menuet Trompette, la différence est du tiers, comme 2 est à 1 ; ensorte que les premières croches valent 2 tiers de noire, & les secondes l’autre tiers; & enfin le différence la moins marquée, comme dans beaucoup de menuets, est comme de 3 à 2, de 7 à 5, &c. Il faut donc que le Noteur ait assez de goût pour apprécier ces différences au juste, & qu’il supplée en cela aux principes de musique qui ne les indiquent pas. ⁸³

I have observed in notating cylinders that there are a number of marches, among them that of the Roi de Prusse, in which the difference of the firsts to seconds is one half, as 3:1, that is, the first eighth equals a dotted eighth and the second equals a sixteenth note. In some menuets, as the petit menuet Trompette, the difference is one third, as 2:1, so that the first eighth equals two-thirds of a quarter note and the second the other third. Finally the least marked difference, as in many menuets, is that of 3:2 or 7:5, etc. It is necessary that the noteur have sufficient taste to understand these differences precisely, and that he supply the musical requirements when these are not indicated.

Although Engramelle mentions the ratio of 3:1, he does not use it in any of the examples in La tonotechnie. La marche du Roy uses 3:2 for the first version, and 2:1 for the second. In the Marche no. 10 the degree of inequality is not specified, although the five divisions of the dial for each quarter note suggest the ratio of 3:2.

Engramelle specifies that eighth notes are performed as equal in La badine d’Alarius and Les portraits à la mode. The Romance uses the ratio of 5:3, Le Bûcheron, 7:5, and La Fontaine de Jouvance uses sixteenths in the ratio of 3:2. From the number of divisions on the dial, we may assume eighth notes are unequal in the ratio of 3:2 in the Allemande no. 9, and 5:3 in the Menuet du Roy de Prusse.

Engramelle continues:

Il faut cependant observer que tout ce que je dis, dans le détail de tous ces airs, sur l’inégalité des croches, n’est que pour faire apprécier ces inégalités: c’est au bon goût seul à apprécier cette variété dans ces inégalités. Quelques petits essais feront rencontrer le bon & le meilleur, ou pour l’égalité, ou pour les inégalités: l’on verra qu’on peu plus ou un peu moins d’inégalité dans les croches change considérablement le genre d’expression d’un air.⁸⁴

It is necessary to observe that everything I say about the inequality of eighth notes, in giving details about these airs, is only to make an estimate of their inequality. Only good taste can correctly determine which variety of inequality to employ. Some small experience will acquaint you with the satisfactory and the preferable in equality or inequality. It will be seen that a little more or less inequality in eighth notes changes the expression of the air considerably.

Engramelle’s musical examples and ornamentation do not seem to be quite in the style of the eras of Couperin, Hotteterre, or Rameau, but his instructions regarding articulation and inequality differ from treatises of the earlier eighteenth century only in that they are much more specific.

Engramelle is not very successful in creating the effect of musical phrasing through his varied duration and inequality of notes. His ritardando of final cadences is so subtle that it fails to have much effect. Performers ordinarily take great liberty in ritards and in leaving space for taking breath between phrases, yet they manage to project rhythmic regularity to an audience. If these computer recordings seem overly mechanical, it is largely due to Engramelle’s underestimation of the performer’s liberty in taking time to clarify the connections between phrases.

Engramelle’s care and subtlety in using the duration of notes to heighten metrical organization, however, gives vivid results. The precision of his directions for performing equal notes unequally shows remarkable sensitivity to musical style as well as invention in clarifying meter. The reader may wish to listen to the remaining recordings of Engramelle’s music at this point.

Engramelle’s instructions for notage of the mechanical cylinder are given from a completely different point of view than instructions for the technique of performance on any other instrument. Engramelle tells us something of the effect that was desired even though the performance on his cylinder organ is excessively mechanical and limited in scope.

The performance he describes is applicable, with reservations, to keyboard instruments. Since his serinette, the organ, and the harpsichord are incapable of dynamic stress, they are entirely dependent on note duration as a means of gaining variety in articulation. However, keyboard technique can use subtle differences of attack and release to give shape to the beginning and ending of each note, as well as note duration to mark metrical units. This technique is beyond the capability of the serinette. The amount of separation between notes that Engramelle specifies for the serinette is, quite likely, exaggerated in order to make up for the lack of any other dimension in articulation. Therefore Engramelle’s examples should be studied rather than imitated, even in keyboard performance.

This survey of seventeenth- and eighteenth-century instructions may give the impression that meter was articulated with relentless intensity. Every note was bowed with a separate stroke, or tongued, with part of its duration silent in order to give clarity to the succeeding note. Such articulation enhances the perception of the beat and its subdivisions and groupings, but the rhythm of phrases and lyric melodic organization seem neglected. It is easy to envision that a performance that narrowly adheres to these instructions would be too rigidly metrical.

A balance between metrical order and larger rhythmic structure must be found in any successful performance. It is useful to mention some elements of performance that keep these articulation traditions from being excessively rigid. The first factor is the absence, or minor role, of the dynamic accent. Twentieth-century performers are so accustomed to employing dynamic accentuation that it is sometimes difficult for us even to imagine a performance in which it is not used. The careful avoidance of the word “accent” in a large number of the performance directions examined is not oversight but demonstrates the reliance of eighteenth-century performers on different means of enhancing metrical perception.

Some experimentation with tonguings, bowings, and fingerings as well as an acquaintance with the serinette music of Engramelle demonstrates that meter can be adequately clarified without accentuation. Even if a modern performer is intellectually convinced that adding a dynamic accent is unnecessary, he or she may have to retrain fundamental musical instincts in order to avoid this habit in performance.

The effect of tonguings, bowings, and fingerings is greatly variable. Bowing instructions don’t tell us how much or how little an articulation was intended to be evident. The degree of separation in tonguings is infinitely variable, and the ear of the listener may hear a nearly smooth legato even when the performer’s tongue separates every note. Both woodwind and stringed instruments incorporate a great variety of articulation techniques to vary the beginnings of notes, which may range from explosive to precise, to less definite, to smooth, and even to imprecise and soft. The same variety applies to note endings, although instruction manuals have little directly to say about these. The greatest difference between keyboard instruments and wind and stringed instruments lies in the ability of the latter two to give different sound colors and dynamic variety to notes by swelling and diminishing the sound.

Little is said about meter and singing technique in instruction manuals, except that both rhythmopoeia and quantitas intrinseca relate poetic meter and word accent to musical meter. A singer’s pronunciation is equivalent to the articulation of an instrumentalist and offers an ideal model. The musical instincts of most performers, perhaps then as well as now, are happiest with performance techniques that allow for the widest possible range of expressive possibilities.