Jump to ContentJump to Main Navigation
Classical and Romantic Performing Practice 1750-1900$

Clive Brown

Print publication date: 1999

Print ISBN-13: 9780198161653

Published to Oxford Scholarship Online: May 2008

DOI: 10.1093/acprof:oso/9780198161653.001.0001

Show Summary Details
Page of

PRINTED FROM OXFORD SCHOLARSHIP ONLINE (oxford.universitypressscholarship.com). (c) Copyright Oxford University Press, 2020. All Rights Reserved. An individual user may print out a PDF of a single chapter of a monograph in OSO for personal use.  Subscriber: null; date: 27 October 2020

Embellishment, Ornamentation, and Improvisation

Embellishment, Ornamentation, and Improvisation

(p.415) 12 Embellishment, Ornamentation, and Improvisation
Classical and Romantic Performing Practice 1750-1900

Clive Brown

Oxford University Press

Abstract and Keywords

This chapter examines a variety of types of embellishment — from rhythmic flexibility through the ornamental use of vibrato and portamento — and the arbitrary addition of appoggiaturas, trills, turns, and other small ornaments to elaborate fiorituras. It traces the range of approaches and the use of different types in specific repertoires at specific times. The relationship of such practices with a freer approach to notation in general is considered. Evidence is drawn from treatises, editions, critical writings, and early recordings by Patti, Moreschi, and Joachim.

Keywords:   embellishment, ornament, fioritura, appoggiatura, trill, turn, vibrato, portamento

Throughout the period 1750–1900 musical notation in European art music was generally viewed as something much more flexible with respect to pitch, rhythm, and embellishment than it has been for much of the twentieth century. But as the tendency for composers to specify their requirements with ever greater precision grew progressively during that period, performers became inclined to observe the letter of the notation ever more punctiliously. Between the middle of the eighteenth century and the middle of the twentieth century the scope for performers to modify and elaborate on the strict meaning of the notation became increasingly restricted, and the orthodoxy of the second half of the twentieth century (based largely on misconceptions of eighteenth- and nineteenth‐century practice), upon which scholarship has only recently begun to make a modest impact, is to regard Classical and Romantic composers' notation as literal and definitive, and to adhere to it as closely as possible in performance. For the vast majority of modern performers, the addition of a few cadential trills and prosodic appoggiaturas, together with minor modifications of phrasing and dynamics, is the most extreme alteration to the received text that they feel to be justified in the performance of late eighteenth‐century and nineteenth‐century music (apart from the employment of a patently unhistorical continuous vibrato, now almost ubiquitous on many instruments and in singing, which appears to be regarded as an essential element of the sound and therefore not dependent on the notation).

In fact, during the first half of the period the embellishment and elaboration of all kinds of music by performers was endemic and, in many respects, fundamental to the aesthetic experience of composer, performer, and listener alike. The alteration of attitudes during the course of the nineteenth century was slower than might be imagined. Change was most gradual in vocal music, (p.416) especially opera, and most rapid in German chamber music, yet by the end of the century it seems probable that in all these areas performers' practices, and their perceptions of what the notation implied, were still considerably closer to those of the late eighteenth century than late twentieth‐century performers' practices and perceptions are to theirs.

The present‐day musician who wishes to understand the ways in which, with respect to embellishment, eighteenth- and nineteenth‐century performers might have responded to the notation of their day, or the sorts of expectations that composers might have had about the interpretation of their notation, needs to be conscious of a number of important distinctions. At one extreme was the addition of more or less elaborate fiorituras to the given musical text, substantially modifying the melodic line or introducing new material at cadences: at the other was the application of various less obtrusive embellishments, ranging from vibrato, portamento, and subtle modifications of rhythm to the interpolation of arpeggiation, trills, turns, and appoggiaturas. The former type of embellishment was considered appropriate and necessary in specific circumstances and genres of music, especially in vocal or instrumental display pieces: the latter (though its precise nature was subject to changes of taste and fashion) was regarded as an essential aspect of musicianly performance in all circumstances, without which the music would be lacking in communicative power. The distinction that Spohr made so clearly in 1832 between a ‘correct’ style and a ‘fine’ style1 was the distinction between music rendered in a literally correct manner and music in which the performer subjected the text to a host of small modifications for the sake of expression. There were also a number of specific situations in which the performer was expected to see beyond the literal meaning of the composer's text. Sometimes, by generally understood convention, the given notes were recognized as standing for different ones, or a particular type of execution was implied by the musical context. Conventions of this sort applied especially to appoggiaturas and in recitative, but were also operative in respect of variable dots, the arpeggiation of chords in keyboard playing, the application of slurs and dynamics, and a whole series of consequences arising from metre and musical genres, which were discussed by German writers under the heading of ‘heavy and light performance style’ (’schwerer und leichter Vortrag’).2 The area in which these latter considerations applied, however, shrank gradually in the course of the period as notation became more prescriptive.

During the second half of the eighteenth century and the first decade or so of the nineteenth century all these kinds of embellishment were a prominent feature of musical life wherever solo performance was involved, but the type of performance in which a large element of fairly elaborate improvised ornament (p.417) was envisaged played a much greater part than it was later to do. In the last decades of the eighteenth century there was still a widespread view that it was the composer's duty to provide solo performers with an outline on which they could elaborate in a manner that displayed their own abilities to best advantage; and even where ornament signs or notated ornaments were included by the composer it was commonly taken for granted that performers were fully at liberty to substitute others that might suit them better. This attitude was clearly enunciated by many writers of the period. Anselm Bayly, for instance, observed in his Practical Treatise on Singing and Playing with Just Expression and Real Elegance in 1771:

Many composers insert appoggiaturas and graces, which indeed may assist the learner, but not a performer well educated and of a good taste, who may omit them as he shall judge proper, vary them, or introduce others from his own fancy and imagination. … The business of a composer is to give the air and expression in plain notes, who goes out of his province when he writes graces, which serve for the most part only to stop and confine the invention and imagination of a singer. The only excuse a composer can plead for this practice, is the want of qualifications in the generality of singers.3

Domenico Corri, approaching the matter from the opposite direction, made much the same point when he stated in the early 1780s that ‘either an air, or recitative, sung exactly as it is commonly noted, would be a very inexpressive, nay, a very uncouth performance; for not only the respective duration of the notes is scarcely hinted at, but one note is frequently marked instead of another’.4 However, as Corri's editions and his other writings suggest, he was concerned rather with the discreet embellishment of the melodic line demanded by ‘good taste’, and proper understanding of composers' notational conventions, than with the addition of elaborate fiorituras.

For many connoisseurs of that period the individuality of a performer's embellishment of the given notation was a vital part of the musical experience. Burney commented in 1778 on ‘La Bernasconi’ that she ‘has no great voice, but she has a very elegant style of singing, and many embellishments and refinements that are wholly new here’.5 And three years later he remarked approvingly of Teresa Maddalene Allegranti: ‘Indeed she seems to me original—her graces and embellishments do not appear to have been copied from any other singer, or to have been mechanically taught by a master.’6 These comments confirm that in Burney's opinion the ability to invent new modes of embellishing familiar music was an essential attribute of a successful solo singer. The continuation of similar attitudes well into the nineteenth century is suggested by the cellist J. J. F. Dotzauer, who observed in the 1820s: ‘There are a mass of (p.418) ornaments which fashion and the humours of virtuosos have increased to such a number that suitable names have not even been found for them.’7

By the early years of the nineteenth century the excess and misuse that inevitably arose from ignorance and incapacity on the part of less gifted solo singers and instrumentalists led to increasingly frequent criticism. Domenico Corri, whose own notions of the appropriate embellishment of vocal music are clearly illustrated by his publications of the 1780s and 1790s (see below), complained in 1810 about ‘the abuse of ornament’, considering it to be ‘of recent date’. He observed:

within my memory, those famous singers Farinelli, Cafarello, Geziello, Pacchiarotti, Millico, Aprile, David, Raff and others of the first eminence, sung compositions with little ornament, exerting their talents, on the parts appointed to them; nor were they permitted to introduce, at random, any graces, ornaments etc., as caprice directed, but in such places only as the composer had allotted.8

He added that the talent of these singers was principally shown by their portamento di voce, though his own publications suggest that, from our perspective, they appear to have modified the musical text to a considerable degree.

The extremely intricate embellishment that was typical of some of the leading performers of the early nineteenth century is exemplified in numerous printed and manuscript sources.9 As an illustration of the sort of treatment expected in a particular genre of aria, Anton Reicha gave examples by Cimarosa, Giordanello, and Lamparelli with extensive added embellishment. He was not, however, entirely happy about the predominance of the practice, commenting:

Singers, at the period of this decadence, only want songs to embroider [airs à broder]: and one might say that for almost forty years [i.e. c.1770–1810] we have lived in the period of musical embellishment, of which the three arias which we have cited may henceforth provide an idea, and serve as a tradition for the history of this art; for it may be presumed that this manner of singing, as a result of the abuse to which it leads, will pass out of fashion, or at least be restrained within reasonable bounds … one might see [these airs] as a special genre and distinguish them from others by calling them air à broder.

But he went on to ask, rhetorically, whether the composer could not write in all the required decoration. For instrumental music his answer was ‘yes’; for vocal music ‘no’, since he clung to the view, enunciated by Bayly, that the singer (p.419) would be bound to do it better. Only in bravura arias did he consider it proper for the composer to specify all the embellishment.10

There is much evidence of changing attitudes during the second decade of the nineteenth century. An anecdote about Manuel García senior provides a glimpse of the developments that were taking place, even in Italy, where the tradition in which composers provided merely the skeleton of an aria was giving way to one in which they assumed greater responsibilities for the detail of the music. Manuel García junior recalled an incident in about 1815 when, at the first rehearsal of a new opera by a composer of the older Italian school, his father was given his part to read at sight:

When his first aria had been reached he sang it off with perfect phrasing and feeling, but exactly note for note as written. After he had finished the composer said ‘Thank you signor, very nice, but not at all what I wanted.’ He asked for an explanation, and was informed that the melody was merely a skeleton which the singer should clothe with whatever his imagination and artistic instinct prompted … The elder García was skilful at improvising … he made a number of alterations and additions, introducing runs, trills, roulades and cadenzas … The old composer shook him warmly by the hand. ‘Bravo! magnificent! That was my music as I wished it to be given.’11

About the same time, Rossini had begun to notate embellishments more fully in his own scores, and he was clearly unhappy to trust his music entirely to the chance of finding a singer with sufficient skill, taste, and understanding to achieve the effects he desired; but there is no reason to believe that either Rossini or the singers with whom he worked would have regarded his embellishments as binding or exclusive. As indicated by the treatment of some of his arias by late nineteenth‐century singers, whose performances are preserved on early recordings, the tradition of elaborate and individual interpolated embellishment in this repertoire was far from defunct, even in the generation after Rossini's death. The cavatina ‘Una voce poco fa’ from Il barbiere di Siviglia (1816) is typical of the type of early nineteenth‐century display aria that lent itself admirably to the inventive embellishment cherished by nineteenthcentury singers and audiences alike. In the performance of such longestablished repertoire pieces, an element of tradition undoubtedly crept in, so that one might expect to find correspondences between the interpretations of different artists; but the desire for individuality was just as important, if not more so, even to the extent that many nineteenth‐century singers were renowned for their fertility of invention in elaborating the same aria differently on different occasions.12 Recordings of Rossini's ‘Una voce poco fa' by three great sopranos born during the second half of the nineteenth century reveal consistency in the choice of places that invited elaboration and in some (p.420) traditionally sanctioned changes, but they also show how other passages were regarded as an opportunity for the singer to introduce the types of ornamentation that displayed the individuality of their voices and technique to the best advantage.13 The modern approach to this repertoire is very different, and it is rare to hear more than one or two very minor conventional modifications and departures from Rossini's notated text.

Rossini's own attitude towards the embellishment of his music remains enigmatic; after Adelina Patti had given a particularly florid rendition of‘Una voce poco fa’ at one of his Saturday soirées he politely remarked: ‘Very nice, my dear, and who wrote the piece you have just performed?’14 Perhaps Patti's rendition seemed musically inappropriate to him, or perhaps he objected to the aria (which he had written for alto) being sung by a soprano. Yet Rossini would certainly have expected a degree of vocal display, if only at the allotted places for cadenzas. Whether ornamentation of the kind that is preserved on early recordings was enjoyed, merely tolerated, or even detested by the composer, however, does not alter the fact that it was a pervasive aspect of nineteenthcentury musical life.

The tendency for nineteenth‐century composers to provide examples of appropriate embellishment, if not obligatory fiorituras, was probably prompted less by concern about performers altering and adding to the musical text than by incorrect or inappropriate ornamentation on the part of inexperienced or unmusical performers. There is considerable evidence that, even in Verdi's generation, opera composers did not necessarily expect the singer to execute written‐out cadenza‐like passages literally; they provided them rather as a guide to length and correct positioning,15 whereas earlier composers had tended to leave such decisions almost entirely to the singer. Thus, although writers and musicians, throughout the Classical period, had inveighed against excessive embellishment and its application incorrectly or in the wrong places, such comments should not be taken to mean that these writers advocated no or even very little embellishment. As Reicha remarked:

One should not confound a thing with the abuse that is made of it; for there is always a great difference between the two. It is necessary also to distinguish between a singer of talent who embellishes a melody with a flexible and pleasant voice, and with exceptional tact and exquisite taste, with those bad mimics and pitiful caricatures who make something worse of it. And if the former has, in addition, enough spirit to place his embellishments in just the right manner, one should not confuse him with the latter who use them profligately.16

(p.421) Complaints about embellishment from this period seem often to have been misinterpreted. Frederick Neumann, for instance, in support of his contention that no embellishment should be introduced in Mozart's later operas, cites early nineteenth‐century objections in the Allgemeine musikalische Zeitung to the decoration of arias in Figaro and Die Zauberflöte by the bass I. L. Fischer (who had been the first Osmin in Die Entführung). These reports suggest, however, not so much that Fischer was taken to task because he added fiorituras and ornaments, but because he added them unskilfully. In 1802 a reviewer complained about his ornamentation of ‘In diesen heil'gen Hallen’ (Die Zauberflöte), where the harmony was complex and where Mozart had in any case written many notes; yet three years earlier, criticizing the same piece, the journal had merely asked that he employ ‘somewhat fewer embellishments’. The fact that Fischer's use of embellishment, despite his distinguished career, was by no means exemplary is illustrated by a report from Hamburg of his elaboration of a fermata in Haydn's Die Schöpfung, having criticized a French singer's harmonically false embellishment in the same work, the reviewer observed: ‘The lack of basic knowledge and the concomitant French levity can to some extent be credited; but how astonished I was when I heard the German singer who has been famous for more than thirty years do the following [Ex. 12.1] and introduce several similar embellishments and decorations.’17

Embellishment, Ornamentation, and Improvisation

Ex. 12.1. Allgemeine musikalische Zeitung, 4 (1801–-2), 14

Other complaints about embellishment from the same journal are revealing both of attitudes and of the types of interpolation that were common. In a report from Hamburg in 1799 a reviewer commented of Madam Righini that

her singing was clean and correct throughout; also she rarely made superfluous ornaments and embellishments [Manieren und Verzierungen], and when one heard these, they never‐theless fitted with the harmony and the accompaniment. How much good that did me, and how I rejoiced over it!—For the following examples may prove how right are my frequent complaints about the widespread mania for making bad and often harmonically quite incorrect embellishments and alterations which not only are tolerated by music directors, but even, to judge by the loud applause of the public, are taken for the non plus ultra of art.

He then gave several examples from Mozart operas. Herr Rau, as Tamino in Die Zauberflöte, sang embellishments as in Ex. 12.2 The reviewer added sarcastically that in the aria ‘In diesen heil'gen Hallen’ Herr Krug as Sarastro ‘every (p.422)
Embellishment, Ornamentation, and Improvisation

Ex. 12.2. Allgemeine musikalische Zeitung, 1 (1798–-9), 604

Embellishment, Ornamentation, and Improvisation

Ex. 12.3. Allgemeine musikalische Zeitung, 1 (1798–-9), 604

time sang the following excellent appoggiatura’ (Ex. 12.3). He continued: ‘All this cannot in any way be compared with the embellishment of a fermata in the last duet of Mozart's opera CosÌ fan tutte by which Mad. Lange was able to gain equally universal and loud applause. This went as follows:’ (Ex. 12.4).18

Embellishment, Ornamentation, and Improvisation

Ex. 12.4. Allgemeine musikalische Zeitung, 1 (1798–-9), 604

The kind of embellishment that was considered appropriate may be illustrated by a multitude of examples from the period. Instances by composers themselves are particularly enlightening and provide useful models for modern performers. Mozart's embellishments for arias by himself and J. C. Bach, evidently intended to assist an inexperienced singer, are readily available,19 though whether embellishment should ever be applied in quite the same way to his later operas is a legitimate question. Interesting examples in Salieri's (p.423)

Embellishment, Ornamentation, and Improvisation

Ex. 12.5. Salieri, Venti otto divertimenti vocali, no. 2

music are contained in a copy of his Venti otto divertimenti vocali in the British Library,20 where the composer himself has pencilled in simple embellishments to the vocal line. In no. 2, for instance, he changed the printed text from Ex. 12.5(a) to Ex. 12.5(b). Much later, Meyerbeer's acquiescence in the modification of display passages in his vocal music is implied by his own practice of providing different versions. This is nicely illustrated by comparison of a transposed version of the Page's aria from Les Huguenots with the original (Ex. 12.6).

There seems little doubt that, as implied by Reicha's treatise, during the early decades of the nineteenth century there was a growing reaction against the kind of butchery practised by Fischer and others. This was combined with the composers' assumption of greater control, especially in German music, something evident, for instance, in Weber's well‐known insistence that his singers abstain from embellishing their parts. By the 1830s and 1840s a widespread prejudice was developing against the addition of ornaments where the composer had not indicated them, particularly in music that was increasingly coming to be seen as ‘Classical’. But in 1844 the English critic Henry Fothergill Chorley advocated a broader view:

That the Vienna tradition of singing Mozart's operas does not bind the vocalist to a bald and literal enunciation of the text and nothing but the text, we have had proofs in the singing of Madam van Hasselt‐Barth, Madam Jenny Lutzer, and, most recently, Mdlle. Zerr—all vocalists formed in Mozart's own town, and who may naturally be supposed to possess some idea of the manner of executing his operas, sanctioned and provided for by himself. But this newly fashioned edict, in command of an utterly and servile plainness, which, if carried out, would utterly destroy all the singer's individuality in art, seems to me to receive contradiction from the music of Mozart itself—even if we had not tradition to confirm us—even if we did not know that Mozart wrote for singers, who were nothing (p.424)

Embellishment, Ornamentation, and Improvisation

Ex. 12.6. Meyerbeer, Les Huguenots, no. 6b, transposed German version from MS in the Meyerbeer Archiv, Staatliches Institut für Musikforschung Preussische Kulturbesitz, Berlin; original French version from printed full score (Paris, Schlesinger, 1836)

without their changes and their closes. That to apply ornament unsparingly would be an insolence—that to employ it out of place and out of style is a musical offence, to be repudiated by all musical people—are facts which by no means imply that to apply and employ it at all are cardinal sins. It is the promulgation of such a canon by modern Pedantry, which has caused one‐half of the transgressions found so nauseous by severe folks and purists who Embellishment, Ornamentation, Improvisation (p.425) are never so complacent as when they can ‘make those singers keep in their right places.’ Without some discretionary taste, delicately and scientifically exercised by the Susanna, the Zerlina, the Fiordiligi, and the Pamina, of Mozart's operas, I am satisfied that no performance of his music is classical—otherwise in conformity with his intentions.21

However, referring to ornamentation in the music of Gluck and Beethoven, he added, with dubious historical justification, especially with respect to the former, that ‘the addition of even an appoggiatura would be intolerable’. That Beethoven was generally opposed to ad libitum additions in his music is beyond dispute; but there were certainly occasions on which he might add ornaments himself, though Ferdinand Ries recalled that he did this ‘Very rarely’,22 or extravagantly embellish a fermata, as occurred in a performance of the Quintet for piano and wind op. 16, or even applaud a performer's initiative in similar circumstances, as happened during a performance of the ‘Kreutzer’ Sonata by Bridgetower.23

Of course, everything hung on what was tasteful embellishment and where it might be appropriate, but it is beyond question that the propriety of introducing a degree of embellishment, even, say, in some of Mozart's string quartets, was accepted by leading musicians of the day. The fact that Friedrich Rochlitz, for instance, considered a certain amount of embellishment permissible in the late Mozart quartets is indicated, in a rather roundabout way, during his discussion of the six quartets dedicated to Haydn, in which he observed: ‘His later quartets are more galant and concertante: in the former, however, every note is thought out; they must therefore be executed precisely as they stand, and no figure may be altered.’24

Most of the chamber music from Mozart onwards that still remains in the repertoire belongs to the kind in which ‘every note is thought out’ and which tolerates virtually no ornamental additions of the type under consideration here, but the dividing line between this kind of music and works in which some latitude for the performer was still envisaged is difficult to determine. Mendelssohn's music, for instance, belongs quite definitely to the mainstream German tradition, and what is known of the composer's strict views about unwritten rallentando and accelerando and other performance matters suggests that he would have set himself firmly against any alteration of his melodic line or addition of ornament. In most cases it is probably inappropriate to consider ornamenting his music, yet Mendelssohn is reported by Henry Chorley, who knew him quite well, to have permitted Staudigl to modify his part in (p.426) Elijah and to have smiled at the effect of an added trill in his ‘Frühlingslied’. He observed:

Mendelssohn… wrote so as to allow no space or exercise of fancy for the vocal embroiderer; and thus, to alter or add to his music, would be to injure it, by showing an arrogant disloyalty to the master's wishes and meanings. Nevertheless, I well recollect the quiet smile of pleasure with which even Mendelssohn used to receive a shake exquisitely placed in the second verse of his delicious ‘Frühlingslied’ (Op. 47); and it must not be forgotten, by all who desire to see the question fairly argued out, and illustrated by facts, not dogmas, that the first singer of Elijah in Mendelssohn's Oratorio—Herr Staudigl—was sanctioned, in one of the finest pieces of dramatic recitative which the work contains, to heighten the effect, by substituting one note for another—the upper G flat, I mean, in place of D flat—in the scene with Baal's priests, on the last repetition of the words ‘Call him louder.’25

In instrumental music in the middle of the century, there were still genres of contemporary music that required considerable embellishment, and few which wholly excluded it, as Charles de Bériot's treatment of the subject in his Méthode demonstrates. Discussing fiorituras, he remarked that

the melody which is best adapted to the type of embellishments we are discussing here is that which aims to please by its amiable, flowery, and graceful style, and of which the accompaniment is light and simple in harmony. But all melody that contains a very pronounced sentiment, whether profound, solemn, or serious, and of which the accompaniment produces complicated harmony, excludes, in part, all kinds of ornament.

Hence it comes about that German music, more bound by harmony than Italian music, lends itself less to embellishment. In proportion as this harmonic complexity has won over all the modern schools, ornamentation has become rarer, while the old melody, more simply accompanied, lends itself more advantageously to it.

These changes are due more to progress than to fashion. It is for the performer to accept these diversities of expression, and to adapt his playing with discrimination, only embellishing music of which the character is suitable: that is a matter of taste.26

It is clear that in mid- to late nineteenth‐century music, major elaboration which substantially changed the shape of the melodic line, or the interpolation of fiorituras, was confined to specific genres of music and particular circumstances, such as cadenzas in concertos (though Brahms was conservative in leaving this to the performer in his Violin Concerto). There was, nevertheless, an expectation throughout the century that performers would modify the written notation in a multitude of less obtrusive ways, which, although they involved departures from or additions to the strict meaning of the notation, were probably not seen as significant alterations to the composer's text any more than a modern performer regards continuous vibrato as an embellishment. These kinds of embellishment were seldom included in the elaborations of arias or instrumental pieces that were provided by composers or theorists as a guide to the inexperienced performer, but they are present in a few sources, (p.427) considered below, that specifically concerned themselves with the characteristics of a ‘fine style’. It is questionable whether even the most informed and imaginative modern musicians have more than a generalized and theoretical notion of the many small ways in which performers of late eighteenth- and early nineteenth‐century music might have deviated from the preserved notation in pursuit of a ‘fine style’. The weight of documentary evidence certainly suggests that musicians of the Classical era would have approached the repertoire of their own day with a freedom far greater than that attempted by the most adventurous present‐day period performers. Indeed, Corri's opinion that a piece of music performed exactly as it is written would be a poor if not a ‘very uncouth performance’ seems to have been widely shared by solo singers, and also by solo instrumentalists, for more than a century afterwards. Leaving aside more elaborate embellishment, it may be useful to consider in general the other kinds of discreet embellishment that were seen as essential to ‘fine’ performance at various stages during the period. By making a careful comparison of the instructions given by Corri, and other writers of the late eighteenth century and early nineteenth century, with the performances by singers and instrumentalists trained during the middle of the nineteenth century that are preserved on early recordings, we find some vital clues to interpreting the necessarily ambiguous verbal and graphic attempts to explain the ways in which ‘fine’ interpreters of the late eighteenth century and early nineteenth century were expected to manifest taste and understanding in their realization of the notated text.

It seems likely that many of the so‐called graces that Corri described and illustrated were still employed, though probably to a lesser extent and perhaps not in quite the same manner, by singers who were born in the generation after his death. Vocal effects, quite different from anything a twentieth‐century singer would produce, can be heard, for instance, in recordings by one of the oldest of the great singers on record, Adelina Patti (1843–1919), and many of them appear to correspond with Corri's graces. Such effects are considerably less evident in performances by opera singers of the next generation; they are much more prominent, however, in performances by one younger singer whose recordings, though they certainly do not belong in the first rank artistically, are in another respect unique: the castrato Alessandro Moreschi (1858–1922). Moreschi's recordings give a vivid impression of what, in aural terms, Corri's notational conventions may have been intended to convey. Although the vocal sounds produced by Moreschi and Patti27 may not be what we would naturally conceptualize from Corri's notation of the graces or his descriptions of their execution, there can be little doubt that they are very closely related to, if not directly derived from the types of ornament that he (p.428) described, many of which seem intended to indicate various portamento techniques.28 If a couple of these recordings are transcribed using the notational conventions employed by Corri (which seems a very reasonable way of notating the effects that can be heard there), the results look remarkably similar to examples in Corri's Select Collection, for instance in his version of J. C. Bach's aria ‘Nel partir bell’ idol mio' from his 1778 opera La clemenza di Scipione (Ex. 12.7). In this example the grace‐notes are given exactly as in Corri's edition, some with two and some with three tails. It is questionable whether a distinction was intended in most cases, for in his instructions (where the grace‐note has two tails) he gave the general instruction that they should be ‘so rapid that, while the effect is felt, the ear shall yet be unable to determine the character of the sounds or to distinguish them from the predominant note’ (for the full quotation see Ch. 13 and n. 60). Corri's specific instruction for the ‘Grace of more intervals’ (e.g. in b. 1), which in his later Singer‘s Preceptorhe called the ‘leaping grace’, was that it ‘is to be taken softly, and to leap into the note rapidly'. The grace ‘close after the note’ (e.g. at the ends of bb. 70 and 93), called the ‘anticipation grace’ in the Singer's Preceptor, was ‘to show that the time necessary for its execution is to be deducted from the last part of that note’, and he advised that ‘in executing it, it is necessary to swell the note into the Grace, and the Grace must melt itself again into the note following’.29 For the ‘Turn Grace’ (e.g. in b. 2), he instructed that it should be ‘taken strong, and melted into the note’. His two signs for breathing meant either a pause ‘about as long as that made by a Comma in reading’ (as in b. 2) or a pause that is to be made ‘as imperceptile as possible’ (as in b. 12). The sign above the fourth note of b. 4 is an accent.30

It may not be too far‐fetched to suggest that the pervasiveness of comparable ornaments in Moreschi's performances resulted from the preservation of older practices among the increasingly isolated traditions of the castrati in the nineteenth century. Many of the arias in Corri's collection were originally intended for castrati, and their unfortunate successors, as members of an increasingly marginalized tradition, may well have clung to the style and methods employed during the period of its final ascendancy in the eighteenth century, while changing tastes and fashions conduced to much more rapid stylistic development in the world of nineteenth‐century opera. Allowing for deficiencies in Moreschi's technique and in the recording conditions, a transcription of his 1904 performance of the ‘Crucifixus’ from Rossini's Petite messe solennelle might appear as in Ex. 12.8 It is not always possible to represent Moreschi's vocal effects in conventional notation. The present transcription uses Corri's notation where it seems probable that similar embellishments were intended, (p.429)

Embellishment, Ornamentation, and Improvisation

Ex. 12.7. J. C. Bach, ‘Nel partir bell’ idol mio', original edition (lower stave) and D. Corri's version from A Select Collection, i. 90–-4

Embellishment, Ornamentation, and Improvisation

Ex. 12.7. cont.

Embellishment, Ornamentation, and Improvisation
Embellishment, Ornamentation, and Improvisation

Ex. 12.7. cont

Embellishment, Ornamentation, and Improvisation
Embellishment, Ornamentation, and Improvisation

Ex. 12.8. Rossini, Petite messe solennelle, Crucifixus, original text (lower stave) and transcription from 1904 recording by the castrato Alessandro Moreschi

Embellishment, Ornamentation, and Improvisation
(p.436) though the starting pitches of the ‘leaping grace’ ornaments are not always distinct. Rhythmic differences from the standard text have been transcribed as accurately as is practicable with regular, conventional notation. Variations in tempo have not been indicated, though this recording and those transcribed in Ex. 12.9 and Ex. 12.12 exhibit considerable flexibility.

The recording of Mozart's ‘Voi che sapete’ (Le nozze di Figaro) that Patti made in 1905 (Ex. 12.9), which has also been transcribed using Corn's notation, shows a number of the same traits, though not all her portamentos correspond with the apparent implications of Corri's notation. It is interesting to note that many of the same characteristics are also found in Patti's performance of a piece much closer to her own time, ‘Ah non credea’ from the beginning of the finale of Bellini's La Sonnambula (1831), a transcription of which also looks remarkably similar to Corri's texts.31 Particularly striking is Patti's frequent use of the type of portamento indicated by Corri in bars 1,11, etc. of Ex. 12.7, in view of Manuel García's description of this type as an old‐fashioned ornament.32 But equally notable are the many other small deviations from the strict letter of Bellini's notation, especially with respect to rhythm, including the assimilation of both dotted figures and pairs of equal notes to the triplet rhythms of the accompaniment.33

It is evident from comparison of these transcriptions of Moreschi's and Patti's recordings with Ex. 12.7 that, apart from their similarity in respect of the graces, many of the slight and not so slight changes of rhythm and pitch in the recorded performances, which may at first seem merely capricious, are also paralleled in Corri. Vibrato and portamento are considered in detail below, but it may be appropriate to draw attention here to the frequent employment of the latter (corresponding, again, with what is implied in Corri and other late eighteenth‐century sources) and the discreet and varied use of the former, which on the earliest recordings is perceived more often as a fluctuation of intensity than as one of pitch.

Though vocal music, and opera in particular, may have exhibited the most extreme manifestations of artistic freedom in the late eighteenth century and nineteenth century, a similar approach was by no means excluded from the performance of solo instrumental music. Violin music, which allowed the closest approach to the eighteenth- and nineteenth‐century instrumentalist's ideal of emulating the human voice, provides some illuminating examples. Comparison of the original text of Rode's Seventh Violin Concerto with Spohr's version, included in the final section of his 1832 Violinschule, reveals many deviations, the reasons for which are partly explained in the accompanying (p.437)

Embellishment, Ornamentation, and Improvisation

Ex. 12.9. Mozart, Le nozze di Figaro, ‘Voi che sapete’, original text (lower stave) and transcription of recording by Adelina Patti

Embellishment, Ornamentation, and Improvisation

Ex. 12.9. cont.

(p.439) commentary and partly passed over in silence. Since Spohr had heard the concerto performed by its composer in 1804 and had, by his own admission, striven to perform it as much like Rode as he was able, it is tempting to believe that Spohr's text furnishes a hint as to the manner in which Rode himself might have played it at that time; but it is probably more representative of Spohr's own manner, and, in any case, Rode himself almost certainly varied his performance considerably with the passage of time. In fact, another version of the concerto, edited by Spohr's pupil Ferdinand David, includes the note: ‘The markings and ornaments are precisely those which the composer was wont to employ in performance of this concerto, and the editor thanks his late friend Eduard Rietz, one of Rode's most prominent pupils, for the information.’34 David's version is quite different from Spohr's with respect both to notes and to bowing, though much of the fingering is similar; in so far as it approaches what Rode might have played, it presumably represents the version of his later years, when Rietz studied with him. Spohr's and David's versions of the first movement of the concerto are compared with the text of the original edition in Ex. 12.10. Spohr marked four types of vibrato with different forms of wavy line, indicating fast (b.3), slow (not included in Ex. 12.10), accelerating (b.5), and decelerating (b.37). The following aspects of performance are indicated by notational changes or added instructions in Spohr's and David's texts:
  • 1. Rhythmic modifiation/agogic accent/tempo rubato. Spohr: bars 16–19, 25, 28, and 30 (Spohr's commentary (p. 185) reads: ‘The second half of the 28th and 30th bar must be so played as slightly to augment the duration of the first notes beyond their exact value, compensating for the time thus lost, by a quicker performance of the following notes. (This style of playing is called tempo rubato). But this acceleration of the time must be gradual, and correspond with the decrease of power’), 31–4, 58, and 60 (Spohr's commentary (p. 187) reads: ‘In the 58th and 60th bar, the ninth note (G natural) should be dwelt upon a little, and the lost time regained by increasing the rapidity of the following notes’), 66, 71, 81, and 83 (Spohr's commentary (p. 189) reads: ‘The last two quavers of the 81st and 83rd bar are to be slightly prolonged, yet so as not to occasion any marked difference in the time’). David: bars 4,12,16,17, 31, 33, 80–2, 83–5.

  • 2. Accent. Spohr: bars 10,11,15, 22, 26, 42, 44, 51, 66, 67, 69, 77. David: bars 14, 22, 42, 44, 60, 80–5, 90, 91.

  • 3. Articulation. Spohr: bars 5,13, 31–4, 80, 82. David: bars 5,13, 85.

  • 4. Embellishment (added notes). Spohr: bars 40 and 48 (‘leaping grace’), 69, 77, 92. David: bars 3, 7, 25, 50, 69, 74, 75.

  • 5. Realization of Rode's small notes as appoggiaturas or grace‐notes (where different from Rode's notation or from each other). Appoggiaturas: Spohr: bars (p.440) 8, ii, 21, 42, 77, 79, 85; David: bars 6, 11, 14, 21, 42, 77, 85. Grace‐notes: Spohr: bars 6, 14, 52, 54; David: bars 8, 52, 54, 79.

  • 6. Vibrato (not specifically marked by David). Spohr: bars 3, 5 (accelerating), 11, 13, 21, 22, 24, 26, 37 (decelerating), 43, 44, 52, 58, 60. David: bar 44? (><).

  • 7. Portamento, implied by fingering or specified by Spohr's commentary. Spohr: bars 5, 132, 36, 40, 48, 68, 69, 76, 77, 79, 90. David: bars 5, 13, 36, 68, 69, 76, 77, 79. (The portamentos in 68, 69, 76, and 77 occur between different notes in Spohr's and David's versions).

Embellishment, Ornamentation, and Improvisation

Ex. 12.10. Rode, Violin Concerto no. 7 in A minor, i, first solo section, original text (bottom stave) compared with versions by F. David and Spohr


Embellishment, Ornamentation, and Improvisation


Embellishment, Ornamentation, and Improvisation

Ex. 12.10. cont.


Embellishment, Ornamentation, and Improvisation


Embellishment, Ornamentation, and Improvisation

Ex. 12.10. cont.


Embellishment, Ornamentation, and Improvisation


Embellishment, Ornamentation, and Improvisation

Ex. 12.10. cont.


Embellishment, Ornamentation, and Improvisation


Embellishment, Ornamentation, and Improvisation

Ex. 12.10. cont.

Interestingly, a review of a performance of the concerto by a Russian violinist, Raczynski, in 1818, which referred to his ‘tasteless additions’ and cited the opening of the first movement as an example,35 indicates that some of the embellishment preserved in David's edition may to some extent have become ‘traditional’ at an early stage (Ex. 12.11). (The tied notation and accents in the first, second, and fifth bars, however, are meant to illustrate Raczinski's ‘poor management of the bow’, also referred to in the review.)

Embellishment, Ornamentation, and Improvisation

Ex. 12.11. Allgemeine musikalische Zeitung, 20 (1818), 317

Aural evidence of the discrepancy between text and performance, which confirms the testimony of Spohr's and David's editions of Rode's concerto, both in terms of personalities and musical similarities, can be heard in a recording of Joseph Joachim playing one of his own pieces. Joachim, widely seen as the ‘High Priest’ of Classical violin playing in the second half of the nineteenth century, studied principally with Rode's pupil Joseph Boehm, knew Spohr, and (p.449) was associated with David. His five recordings, made in 1903, when he was 72, include two pieces of solo Bach, which are free and individual in their interpretation of the written text, two performances of his own arrangements of Brahms's Hungarian dances, which also reveal considerable flexibility in the realization of the notation, and a performance of his own Romance in C major. Of all these pieces, the Romance is the closest to the mainstream nineteenth‐century Classical tradition and the one from which we are most likely to gain some idea of the manner in which he might have treated the musical texts of his friends Brahms, Mendelssohn, and Schumann. The many small deviations from the literal text and the occurrence of vibrato and portamento reveal remarkable similarities to the kinds of amplifications of Rode's text by Spohr and David and, in the broadest sense, to many of the features that are apparent in the vocal music discussed above.

Joachim's performances provide particularly revealing evidence of what a great musician of the nineteenth century felt to be essential in a composition and what he regarded as incidental. For Joachim the concept of ‘rubato’ clearly extended to allowing the executant extensive licence to alter rhythms within the framework of correct harmony and essential melodic contours. In terms of embellishment, too, he permitted himself considerable freedom. Comparison of a transcription of Joachim's performance of his C major Romance with the published text reveals many disparities; some of these may be genuine ‘second thoughts’ on the part of the composer, but most seem typical of the artistic freedom that was regarded as integral to a fine style (Ex. 12.12. 12). Similar principles have been followed in this transcription to those employed in Exx. 12.8 and 12.9, insofar as they are applicable to instrumental performance. However, clearly perceptible vibrato has been marked using Spohr's signs; in very many cases where vibrato is not marked the notes appear to be played with a completely still left hand. Prominent portamento has been indicated here by slanting lines above the stave. Rhythmic deviations are notated to the nearest standard pattern, although many subtle rhythmic nuances in Joachim's perfor mance defy notation.

Many similar things would undoubtedly also have been heard in solo performance on wind instruments. Keyboard music provided less scope for imitating the voice, since it entirely excluded vibrato and portamento (in the sense of an audible slide, though not as extreme legato) and many varieties of dynamic nuance on a single note; yet some keyboard players were noted for the vocality of their playing (for instance, Thalberg36 and Chopin), which became increasingly possible as the sustaining power of the instrument was improved. But some freedoms, such as tempo rubato in the strict sense, or modifications of notated rhythms, were just as prominently employed as in other solo instrumental and (p.450)

Embellishment, Ornamentation, and Improvisation

Ex. 12.12. Joachim, Romance in C major, original text (lower stave) and transcription of Joachim's 1903 recording

Embellishment, Ornamentation, and Improvisation
Embellishment, Ornamentation, and Improvisation

Ex. 12.12. cont.

Embellishment, Ornamentation, and Improvisation
Embellishment, Ornamentation, and Improvisation

Ex. 12.12. cont.

vocal performance; and some idiomatic keyboard techniques, such as arpeggiando, were widely applied where they were not indicated in the notation.37

More detailed consideration of some of the major types of ‘discreet’ embellishment is included in the following chapters. In addition, these chapters examine significant issues surrounding the most important classes of ornaments, as well as a few other related aspects of performing practice.


(1) Violin School, 181.

(2) See Ch. 16 below.

(3) pp. 47–8.

(4) A Select Collection, i. 2.

(5) Alvaro Ribeiro SJ ed., The Letters of Dr Charles Burney (Oxford, 1991), i. 264–5.

(6) Ibid. i. 334

(7) Méthode de violoncelle, 40.

(8) The Singer's Preceptor,3.

(9) Johann Adam Hiller, Sechs italienische Arien verschiedener Componisten (Leipzig, 1780); Douglas Lee, ‘Some Embellished Versions of Sonatas by Franz Benda’, Musical Quarterly, 62 (1976), 58–71; Corri, A Select Collection; Mozart's ornamented arias: Austin Caswell, ‘Mme Cinti‐Damoreau and the Embellishment of Italian Opera in Paris 1820–45’, Journal of the American Musicological Society, 28 (1975), 459–2; Baillot, L'Art du violon; Chappell White, ed., Viotti's Violin Concerto no. 27 (Recent Researches in Classical Music, 5; Madison, Wis., 1976) (containing Viotti's additions to the autograph).

(10) Cours de composition /VollstÄndiges Lehrbuch, ii. 498–9.

(11) Malcolm Sterling Mackinlay, García the Centenarian and his Times (Edinburgh, 1908), 34.

(13) Compare e.g. recordings of the aria by Marcella Sembrich (1858–1935), Luisa Tetrazzini (1871–1940), and Amelita Galli‐Curci (1882–1963).

(14) Cited by Richard Osborne in ‘Barbiere di Siviglia, Il (ii)’, The New Grove Dictionary of Opera, ed. Stanley Sadie (London, 1992), i. 311.

(15) David Lawton, ‘Ornamenting Verdi Arias: The Continuity of a Tradition’ (forthcoming).

(16) Cours de composition/VollstÄndiges Lehrbuch, ii. 491.

(17) Allgemeine musikalische Zeitung, 4 (1801/2), 14–15.

(18) Allgemeine musikalische Zeitung, 1 (1799), 604 f.

(19) See e.g. New Grove, ix. 46, art. ‘Imrpovisation’, §1, 3. An example of embellishment indicated by Haydn, in Ilritorno di Tobia, is also given there on p. 45.

(20) Pub. (Vienna, [c.1803]); Shelfmark K.7.c.26.

(21) Modern German Musk (London, [1854]), ii. 376–7.

(23) F. G. Edwards, ‘George P. Bridgetower and the “Kreutzer” sonata', Musical Times, 49 (1908), 302.

(24) Allgemeine musikalische Zeitung, 1 (1798–9), 52.

(25) Modern German Music, ii. 377.

(26) Méthode, 206.

(27) These singers may be heard on CD. Patti's recordings are available on The Era of Adelina Patti from Nimbus Records (NI 7840/41), and Moreschi's have been issued on Opal 9823.

(28) Discussed as such in Ch. 15.

(29) For further discussion of these types of vocal portamento see Ch. 15.

(30) A Select Collection, i. 8.

(31) Fora full transcription of this recording see Clive Brown, ‘Nineteenth‐Century Notation: Appearance and Meaning’ in Musical Theatre in Nineteenth‐Century Europe, Venice 1997 (forthcoming).

(32) See below, Ch. 15, especially pp. 567 ff.

(33) See below, Ch. 16, ‘The Variable Dot’, esp. pp. 618 ff. and Ex. 16.36.

(34) Concert‐Studien für die Violine, no. 7.

(35) Allgemeine musikalische Zeitung, 20 (1818), 317.

(36) His publication L'Art du chant appliqué au piano (The Art of Singing Applied to the Piano) aimed to aid the cultivation of this quality.

(37) See Ch. 16, ‘Arpeggiation’.