‘Interpretation’
This chapter is addressed to ‘modern performers’regardless of whether they sing with or play modern or period instruments. We may start by considering the question ‘What exactly should we try to do in performing music written before our own time?’In fact, the answer is essentially the same for early music as it would be for a new composition. Most performers would think in terms of being true to the work, of exploring its emotional content, of attempting to honour the composer’s intentions. We value imagination and originality in performers, but recognise that (normally) this serves the music they perform, helping to illuminate its character or make palpable its emotional content. By and large, we are not so happy when a performer’s imagination distorts or disguises the music on which it is exercised.
This process of realising a musical work in sound is generally called
‘interpretation’– though some composers have been very uncomfortable with all that this implies. Famously, Ravel proclaimed, ‘I do not ask for my music to be interpreted, but only for it to be played’, a remark echoed by Stravinsky when he wrote, ‘Music should be transmitted and not interpreted, because interpretation reveals the personality of the interpreter rather than that of the author, and who can guarantee that such an executant will reflect the author’s vision without distortion?’1Their complaint, clearly, was with wilful renditions in which performers’whims unjustified by anything in the score are conspicuous.
‘Interpretation’, however, ought to be a good word for what these com-posers wanted. The OED gives as one definition for ‘interpret’to ‘bring out the meaning of ’a musical composition by performance.2In this sense, it might be seen to parallel the responsibility borne by foreign-language interpreters not to misrepresent the meaning of what they are given to translate.
Thinking about the ambiguity of the term ‘interpret’is quite a useful way of distinguishing between performances which are true to the music being performed and those in which something else (showmanship, for instance)
18 peter walls
gets in the way. We could, in fact, call the first kind of performance ‘interpre-tation’(anchoring that word to its primary meaning) and the second kind
‘appropriation’(since the musical work has, in a sense, become a vehicle for the performer’s personal agenda). Interpretation, then, might become syn-onymous with trying to determine and to realise the composer’s intentions.
Score reading
All written music needs ‘interpretation’. A musical score awaits re-alisation in sound. And, just as with a foreign language, expertise in being able to read the music is a prerequisite for doing this. This is something we tend to take for granted until we come across an unfamiliar form of notation such as lute tablature (Example 2.1a), scordatura writing (Example 2.1b) or graphic notation (Example 2.1c). But we need to recognise that, just as the pronunciation and meanings of words change over time, apparently stan-dard musical notation has subtly different meanings depending on where and when it was written. In fact, it always needs contextualising as part of this process of interpretation. In 1717 Franc¸ois Couperin wrote about the particular problem with the music of his own milieu: ‘There are in my view some deficiencies in the way we notate music which relate to the manner in which we write our language. It is that we write something different from what we play. This is why foreigners play our music less well than we do theirs.’3What specifically he had in mind was that the first in every pair in a series of quavers moving by step would be subtly elongated (notes in´egales) despite the apparently equal notation.
Many elements in even quite a simple piece of keyboard music pre-suppose some sense of context. Example 2.2 is the second Gavotte from J. S. Bach’s ‘Ouverture nach Franz¨osischer Art’(Overture in the French Style) BWV 831. The Overture (meaning Suite in this case) was one of the few works by Bach published during his lifetime. It appeared in 1735 in Part II of the Clavier¨ubung (Keyboard Exercise), whose title page tells us that it was written for a ‘harpsichord with two manuals’. Immediately, the piano marking at the beginning assumes a distinctive significance since it is effectively an instruction from the composer to play on the upper manual where the jacks pluck only a single set of strings tuned at written (8) pitch.
In contrast, Gavotte I (which precedes and follows Gavotte II in the kind of relationship we associate with minuet and trio pairs) is marked forte, meaning that it is to be played on the lower manual with its coupled sets of strings, possibly even enriched by a third (4) set tuned an octave higher.
Of course, this alternation of manuals creates a dynamic contrast between
Historical performance and the modern performer 19
(a)
6 5
6 5
violin
violin basso continuo
basso continuo written:
sounding:
(b)
(c)
Example 2.1 (a) John Dowland, ‘Lachrimae’, bars 1–2, with realisation in upper system ( c 1974, 1978, 1981 by Faber Music Ltd. Reproduced by permission of Faber Music Ltd.
from The Collected Lute Music of John Dowland, 3rd edn, transcribed and ed. Diana Poulton and Basil Lam.) (b) Heinrich Biber, ‘Rosary Sonata’No. 7 (‘The Flagellation of Christ’), bars 1–2 (c) John Cage, Water Walk, beginning (Edition Peters No. 6771; c 1961 by Henmar Press Inc., New York. Reproduced on behalf of the publishers by kind permission of Peters Edition Limited, London.)
20 peter walls
6
Example 2.2 J. S. Bach, Gavotte II, ‘Ouverture nach Franz¨osischer Art’BWV 831, bars 1–10
the two gavottes, but more than that, it results in a significant change in timbre.
Such an instrument would have been ‘well tempered’. In other words, it would not have been tuned in equal temperament, which is the standard way on modern keyboard instruments of resolving a fundamental problem – namely, that a ‘circle’of acoustically pure fifths does not quite return to the point from which it started. In the eighteenth century, the preferred solution lay in a range of circular temperaments in which a few of the fifths (normally four or six) are slightly smaller than pure. This creates a kind of chiaroscuro (a term actually used by Tartini4in recommending one such temperament) between the sweet consonance of the most commonly used chords and the rather more tense sound of remoter harmonies. Consequently, each key has a distinctive character. There are many quite detailed instructions from the eighteenth century on how to set various circular temperaments.
Exactly which one Bach favoured is a matter of conjecture, but all circular temperaments share certain broad principles. One reasonably reliable rule of thumb for conceptualising the effect of these tunings is to regard each note on the keyboard as being tuned to suit its simplest form, meaning the form which is lower in the order of accidentals; hence the note immediately above F on the keyboard will sound better when it functions as an F rather than a G, and so on. Thus, D major – the tonic of Gavotte II – is likely to be a sweeter sounding chord than F major (the dominant of Gavotte I, which is in B minor). Within Gavotte II itself, a circular temperament will subtly underline the movement away from and back to D major.5
The indication that this work was written to be played on a two-manual harpsichord may be read as implicitly more specific than the actual form of words suggests. There is no definite information about Bach’s own harp-sichords (though the inventory of his estate shows that he owned five at the time of his death). In 1719 he was responsible for purchasing a
Historical performance and the modern performer 21
two-manual instrument by Michael Mietke for the court at C¨othen (where he was Kapellmeister), and this maker seems to have been well regarded in Leipzig. Bach also worked with the Leipzig-based harpsichord maker Zacharias Hildebrandt. Two years before the publication of the second part of the Clavier¨ubung, a Leipzig newspaper had announced the resumption of concerts by Bach’s Collegium Musicum which would be graced by ‘a new harpsichord [Clavicymbel] such as had not been heard here before’.6
Regardless of national type, there are features of such instruments that affect technique. The slightly shorter length of key makes orthodox piano fingering impractical (in particular, passing the thumb under the fingers in scale passages). Bach, according to his son Carl Philipp Emanuel, was forward-looking in his fingering methods and used the thumb more than had been normal – but this is still a far cry from standard modern piano fingering. Moreover, the fingering exercises included in the ‘Notebook’com-piled for his son, Wilhelm Friedemann, show that he expected him to practise the paired fingers then commonly used (see Example 2.3). Such fingerings lend themselves to the subtle lengthening of the first in a pair of conjunct quavers, resulting in notes in´egales.
Since Gavotte II is quite explicitly a French-style piece (and since Bach was well versed in matters of French performance practice), the stepwise quavers in Example 2.2 should perhaps be treated as notes in´egales. The ornament signs are also manifestations of a French stylistic trait. Bach, indeed, included an ornament table in Wilhelm Friedemann’s ‘Notebook’ which can be used to decode the signs used in this score. And, last but not least, it is obvious that the fact that this is a gavotte (even though clearly not a danced gavotte) will have a direct bearing on tempo and rhythmic character. Johann Joachim Quantz described the gavotte as being a little slower than a bourr´ee or rigaudon but sharing the gay and well-articulated character of those dances.
Example 2.4, a short recitative introducing an aria sung by Argante in Handel’s 1731 version of Rinaldo, presents a case in which it is all but impos-sible to read the notation accurately without some understanding of matters which fall within the province of performance practice. First, any trained
3434 343 1 343 4 1
Example 2.3 J. S. Bach, Klavierb¨uchlein f¨ur Wilhelm Friedemann Bach, Applicatio BWV 994, bars 1–4
22 peter walls
che l’in de gno tuo lac cio io vuo pur fran ge re.
T’ar re sta, oh Di o! non pian ge re;
Recitativo RGANTE
A
Continuo (Violoncello,
Cembalo)
Example 2.4 Handel, Rinaldo (1731 version), recitative ‘T’arresta, oh Dio!’
singer in Handel’s company would have realised that appoggiaturas should be introduced where the repeated notes on a stressed–unstressed pair of syllables coincide with the changes in harmony and are followed by a rest (in a so-called feminine ending). This is a way of doing justice to the accentu-ation of the words and has a syntactical (rather than decorative) significance.
Depending on the taste of the performer, an appoggiatura might also have been used on ‘T’arresta’. The continuo harpsichordist would add a 4–3 sus-pension on the dominant in the final cadence (bar 4) to avoid an abrasive clash with the singer’s appoggiatura. And the cellist playing alongside the harpsichord would have known not to sustain the long notes for their full value.
How would eighteenth-century performers have absorbed these princi-ples? Partly from their music teachers, partly from just hearing colleagues, partly from treatises, partly from written-out exempla. How do we assimilate them today? Directly, only from the last two sources (and indirectly from these sources as they are mediated by our own teachers or by the performers we most admire). Appoggiaturas of this kind are discussed, for example, by Francesco Gasparini in his L’armonico pratico al cimbalo (Venice, 1708).7 An even fuller discussion can be found in Johann Friedrich Agricola’s heav-ily annotated translation of the singing treatise by Pier Francesco Tosi.8 Alessandro Scarlatti habitually wrote exactly what notes he expected from the singer in cadences like the one ending the recitative in Example 2.4. The aria that follows assumes an equal familiarity with unnotated conventions – particularly in regard to ornamentation. Again, this is discussed by Tosi (and Agricola), but there are a few fascinating written-out examples of operatic arias which can be traced to Handel himself.9
Who should sing this music? For his 1731 production Handel had the famous castrato Francesco Bernardi, known as Senesino, singing the title role of Rinaldo. But Argante (another male role) was sung by the contralto
Historical performance and the modern performer 23
Francesca Bertolli. These days, high male roles in baroque opera tend to be assigned to falsettists (countertenors). The operatic countertenor is an invention of the twentieth century – an unnecessary answer to the problem of not having castrati available. When Handel found himself in a similar situation (that is, when he lacked castrati who were good enough), he simply cast the best female singer in his company. To be historically accurate, we ought to follow his example.
It is tempting to regard notation as prescriptive only, or at least primar-ily, in relation to pitch and duration. But the Bach and Handel examples considered here help us to identify a number of other parameters which are implicit in the way the music is written. We could thus create a check-list of basic questions with the aim of arriving at a thickly contextualised understanding of a musical score:
rWhat voice-types or instrument(s) are intended?
rHow will they be set up (tuned etc.)?
rWhat kind of technique is assumed and what are the implications of this for articulation, phrasing, tonal colour etc.?
rWould the composer have assumed familiarity with rhythmic or syntac-tical conventions not made explicit in the score?
rWhat kind of ornamentation, if any, is appropriate?
rAre there expressive devices which would have seemed natural when the music was composed but which no longer form part of the well-trained musician’s vocabulary?
rAre there aspects of modern expressiveness which are not suitable for this repertoire?
All these questions are so closely related to a properly nuanced understanding of the notation that I hestitate even to describe them as aspects of context.
But to this list of questions (the ‘A list’) others could be added (the ‘B list’) which might also affect the way we approach the performance of a particular work. For what circumstances was it written? Was the piece composed with a particular kind of acoustic in mind? And so on.
Being able to come up with ‘right’answers to all of these questions is never easy and often impossible; moreover, such answers are no substitute for the performer’s imagination. At best, the arena within which imagina-tion funcimagina-tions is one in which a set of received assumpimagina-tions is displaced by an alternative range of stylistic possibilities. That process of displacement can itself act as a stimulus to musical imagination. Carl Dahlhaus argued that what made a history of music – and, by extension, any history of the arts – distinct from other kinds of histories is its concern not just with things from the past (events, documents) but with an aesthetic present.10
24 peter walls
A similar statement could be made about the role of performers. The per-former, whose task it is to realise that score for contemporary audiences, is especially concerned with this act of mediation between an historic past and an aesthetic present.
The ‘right’ instrument
At this point, I want to look at the most obvious of the parameters outlined above – the choice of instrument – to consider how integral it is to a proper understanding of music written before our own time. This in turn leads us to a related question: to what extent should performers be free to ignore (or remain ignorant of) historically verifiable information about how the music they play originally sounded?
Interest in period instruments has been driven by a conviction that music of earlier eras is best served by recreating the instrumental timbres envisaged by the composer (in so far as they can be established). There is, of course, an alternative point of view, most eloquently put in the 1870s by Philipp Spitta, who proclaimed that the modern pianoforte ‘floated in the mind of Bach’.11 Spitta, incapable of appreciating the beauty of a fine eighteenth-century German harpsichord (and perhaps even unable to hear such an instrument in playing order), was convinced that Bach’s intellectual conception was ahead of instrument technology – in other words, that his keyboard works came into their own only once an instrument worthy of them had been developed.
It is, in fact, interesting to speculate about what, say, Mozart’s attitude might have been to a mid nineteenth-century piano or even a modern nine-foot Steinway. He might (in the manner of Spitta) have thought, ‘This is the instrument of my dreams; this is what I really wanted for my concertos’, or, alternatively, ‘If only I’d known about this wonderful instrument I’d have written some music for it.’On the other hand, he might have dismissed it as a crude and rather vulgar product of an industrialised society. We cannot know for certain which (if any) of those responses the piano would have evoked – but we do know that he was excited by the fortepianos that he encountered in the late 1770s. That in itself should be more than enough to make any performer who wants to play Mozart take them seriously.
The first thing one notices about Mozart performed on instruments of the kind he knew and loved is that the common view of his music as an expression of classical restraint and elegance tends to evaporate fairly quickly. Even a work like the Piano Concerto in B major K. 450, which seems to evoke a Figaro-like world of dramma giocoso, uses the resources of the piano to the full. Example 2.5 begins at the point where the piano takes over an idea first outlined in the opening bars in a beautiful dialogue between winds and
Historical performance and the modern performer 25
Vc., Bass Vn 1/ 2, Vla 3
3 83
3 3
3 79
Bsn Ob.
74
3 3 3
[ ]
Vc., Bass Orch.
[ ]
Vn 1/ 2, Vla
[ ] Pf.
[Allegro]
[ ]
71
Example 2.5 Mozart, Piano Concerto in B major K. 450, i, bars 71–87 (piano with orchestral reduction)
26 peter walls
strings. As the solo part approaches the cadence at bar 86 (where the strings effect a one-bar transition to the relative minor), it expands out to cover the full five-octave (F1 to f3) range of Mozart’s instrument. Furthermore, the transparent tone of the instrument suits both the double thirds so prevalent in this example and the agile semiquavers.
The autograph manuscript of Haydn’s Sonata in E major Hob. XVI:49 allocates it to the ‘Forte-piano’– the first sonata to be thus described by the composer. Moreover, three letters written to Maria Anna von Genzinger at the time of its completion in June 1790 make it quite clear that he had composed it with a specific instrument in mind:
This Sonata was destined for Your Grace a year ago, and only the Adagio is quite new, and I especially recommend this movement to your attention, for it contains many things which I shall analyse for Your Grace when the time comes; it is rather difficult but full of feeling. It’s a pity, however, that Your Grace has not one of Schantz’s [sic] fortepianos, for Your Grace could then produce twice the effect.12 The preference not just for a fortepiano but for a Schanz instrument is explained and reinforced in two subsequent letters. First, Haydn insists that this sonata really will not work very well on the harpsichord:
It’s only a pity that Your Grace doesn’t own a Schantz [sic] fortepiano, on which everything is better expressed. I thought that Your Grace might turn over your still tolerable harpsichord [Fl¨ugl] to Fr¨aulein Peperl, and buy a new one for yourself.
It’s only a pity that Your Grace doesn’t own a Schantz [sic] fortepiano, on which everything is better expressed. I thought that Your Grace might turn over your still tolerable harpsichord [Fl¨ugl] to Fr¨aulein Peperl, and buy a new one for yourself.