It was as a boy of twelve or thirteen that I played the harpsichord for the first time. My early experiences of ensemble playing with the instrument took place with my contemporaries, but primarily with my Dutch teacher at secondary school and his wife, Mr and Mrs Meijer. Both were amateur recorder players, and as a young amateur harpsichordist I would accompany them. At first I would diligently play the realized basso-continuo parts, such as the ones often found in twentieth-century editions. Sometimes it was clear that one could make changes to what was in the realizations, informed by one's own insights, but my insights were not formed at that stage and I did not know where to begin. Very soon I found these realized parts boring. One day, while drinking a glass of wine (one of my first), Mr Meijer asked me, ‘Have you noticed that under the realized part, there are often figures noted down? Do you dare to play from them?’. After many wrong chords, but in the company of kind people who had also made mistakes themselves, I began to understand the system of figures a little. We played together every Saturday, and very soon I found it more pleasant to play without the chordal realizations. Then one fine Saturday afternoon, I sat down at the harpsichord to find that Mr or Mrs Meijer had taped over the right hand of the realized continuo part with an empty staff! There was no more escaping it now: I was dragged back to the bass line and the figures. After many further Saturday afternoon sessions, the results became more acceptable.
When I entered the Conservatorium in Amsterdam, I heard that basso-continuo lessons were on offer. These were taught not by my principal-subject teacher, Gustav Leonhardt, but by the viola da gamba player Veronika Hampe. And with this, a new world opened to me. There were seventeenth- and eighteenth-century sources on continuo playing! Veronika had copies and facsimiles, and generously allowed students to borrow them. I began to read, full of enthusiasm.
At the Conservatorium, I learnt to play strict, four-part continuo. But I found the rules boring from the start: no octaves, no fifths. It was not difficult, but so uncreative. What interested me much more was how I could accompany well, without always being too loud. The stricture on four parts offered absolutely no possibility for making crescendos or decrescendos. At this point, I have to think of the traverso player and composer Johann Joachim Quantz, who, no doubt in irritation at continuo players who insisted on four-part (or more) playing, formulated one of his excellent rules which I adhere to to this day: ‘The universal rule of continuo playing is always to play in four parts, but if one really wishes to accompany well, it is often better not to follow this rule to the letter’ (‘Die allgemeine Regel vom Generalbaß ist, daß man allezeit vierstimmig spiele: wenn man aber recht gut accompagniren will, thut es oft bessere Wirkung, wenn man sich nicht so genau hieran bindet’; Versuch einer Anweising die Flöte traversiere zu spielen (Berlin: Voss, 1752), 223, §4). Now here was something that I could work with! Accompanying a traverso is a delicate matter, of course, and Quantz had experienced this all too often; as a traverso player, before you know it, you can be overwhelmed by a harpsichordist playing too loudly.
I formulated my own first rule of basso continuo: as a continuo player, you are the slave of the person you are accompanying. Happily, this doesn't necessarily mean that you have to play slavishly. The giving of impulses is essential, just as good articulation of the bass line, a feel for dynamics, rhythmic freedom and creativity must all be used to serve the music. Providing direction to a musical line, which is such an essential element of music-making in general, is an absolute must when it comes to continuo playing.
As a student, I read and re-read F. T. Arnold's The Art of Accompaniment from a Thorough-Bass as Practised in the XVIIth and XVIIIth Centuries (Oxford: Oxford University Press, 1931), and from this learnt about the existence of differences in style; continuo playing in early seventeenth-century Italy sounded different from that in eighteenth-century France, for example. I began to study sources: Heinichen, Mattheson, Niedt, Gasparini, Saint Lambert and so forth.
There are also sources attesting to J. S. Bach's superior continuo playing. Lorenz Christoph Mizler witnessed at first hand Johann Sebastian's ability to realize melodies in the right hand as if they were written out (Werner Neumann and Hans-Joachim Schulze, eds, Bach-Dokumente, volume 2: Fremdschriftliche und gedruckte Dokumente zur Lebensgeschichte Johann Sebastian Bachs 1685–1750 (Kassel: Bärenreiter, 1969), 321, No. 419). And C. P. E. Bach recounts that his father, when playing a trio sonata, had the facility to add in an extra melodic line just for the fun of it! (Hans-Joachim Schulze, ed., Bach-Dokumente, volume 3: Dokumente zum Nachwirken Johann Sebastian Bachs 1750–1800 (Kassel: Bärenreiter, 1972), 285, No. 801). Yet if someone dared to play in such a way today, the ‘authenticity purists’ would emphatically label it a disgrace!
What are we now to make of the basic idea that continuo playing should be four-part? In virtually all sources, this is the instruction that has been handed down to us. But aren't many of the continuo treatises actually conceived as a means of teaching the art of composition? In which case, learning to write in four parts is indeed a good and necessary place to start. Such is the situation with, for example, Johann David Heinichen's Der General-Baß in der Composition (Dresden: author, 1728).
It would be wonderful to peer over the shoulder of an eighteenth-century continuo player. How did they play? How did they accompany? What were the dynamics like, the voice-leading, the ambitus? What was fashionable and what was not? Realized continuo parts from the eighteenth century do admittedly provide us with glimpses, but these are rare and the question remains as to how representative they are. What are we to think, for example, of the extensive manuscript Regeln des Generalbaßes von dem Herrn Musico Heering, written by the pupil of Johann Friedrich Hering, Otto Carl Friederich von Voss (D-B, Mus. ms. theor. 348)? It contains four-part realizations of a large number of works, including trio sonatas by Corelli, overtures by Handel, Pergolesi's Stabat Mater and works by J. S. Bach, including the St John Passion and the trio sonata from The Musical Offering. (Concerning this manuscript see Jörg-Andreas Bötticher, ‘Generalbaßpraxis in der Bach-Nachfolge: Eine wenig bekannte Berliner Handschrift mit Generalbaß-Aussetzungen’, Bach-Jahrbuch 79 (1993), 103–125; and Bötticher, ‘Regeln des Generalbasses: Eine Berliner Handschrift des späten 18. Jahrhunderts’, Basler Jahrbuch für historische Musikpraxis 18 (1994), 87–114.) Such an extensive collection of realized continuo parts from the eighteenth century should, of course, constitute a wonderful source, but unfortunately this is not the case. Voss was still only a teenager when he wrote these realizations. His teacher Hering always demanded four-part realizations of his pupil, and on those occasions when Voss forgot something, he was required to add it in later. This is dry teaching and practice material from which to learn four-part writing, and not an example for us to emulate as players today. J. S. Bach would certainly have been horrified by it.
Along similar lines, there used to be a basso-continuo exam at the Conservatorium in Amsterdam such as the one I took in the 1960s. This consisted of a written paper requiring the realization of a bass line, which Gustav Leonhard had notated for the candidates in his elegant hand. Four-part writing was no longer required, but fifths and octaves were forbidden. Later, in the 1970s, Leonhard and I abolished this written exam in favour of a practical one.
Another example of a curious written-out continuo part from the eighteenth century is the set of realizations by cellist Antonio Tonelli for Arcangelo Corelli's Op. 5 violin sonatas, found in a manuscript in Modena (I-MOe, Mus. F.1174, available at imslp.org; see also Corelli: Sonaten für Violine und Basso Continuo, ed. Christopher Hogwood and Ryan Mark (Kassel: Bärenreiter, 2013)). A striking feature of Tonelli's realizations is his frequent use of thick chords of more than four parts. This is excellent for making crescendos, but it means that the accompaniment is far too loud for a violinist who wishes to make a decrescendo or play softly. What do we do with such a source? In the explanatory notes to a recording of the Corelli violin sonatas with these realizations (Corelli: 12 sonate a violino e violone o cimbalo, Op. 5, Trio Veracini, Novalis 150128–2 (1996)), I read that the harpsichordist, Lars Ulrik Mortensen, (fortunately) did not play exactly what Tonelli wrote, but sometimes reduced the number of parts and corrected the extremely high registers. Who is to say that such a realization is not authentic? Perhaps it was indeed played as written during the eighteenth century. Most likely a realization of J. S. Bach's own continuo playing would also raise eyebrows nowadays.
We know of one continuo realization that must have been seen, and possibly corrected, by Bach: that of his pupil Heinrich Nicolaus Gerber for a violin sonata by Albinoni. Philipp Spitta, who published this realization in his Bach biography, writes that since Bach did not make many corrections to Gerber's work, he was obviously satisfied with it (Joh. Seb. Bach, third edition (Leipzig: Breitkopf & Härtel, 1921), volume 2, 125 and Appendix 1). But we must consider this realization to be an exercise, another good example of a pedagogical assignment. The doubling of the solo violin melody at the start of the second, fugal movement suggests that the realization would have been played without violin during the lesson. In eighteenth-century orchestral music, a fugal opening would normally have been played by the continuo, but to suggest that the continuo would also play along with the melody of a solo instrument is taking things several steps further. In any case, such a practice would most certainly not have contributed to better intonation.
In the late works of Bach, such as the trio sonata from The Musical Offering or several movements of the Mass in B minor, continuo parts become more and more complicated. For these works, several eighteenth-century realizations are known. These are not practice exercises, but are designed for playing, since the realization of such complicated parts from the figures would have been within the capabilities of very few players.
In reading Augusta Campagne's book Simone Verovio: Music Printing, Intabulations and Basso Continuo in Rome Around 1600 (Vienna: Böhlau, 2017) I was reminded of a rule that I set down some time ago but which I still stand by: ‘If a suspension is resolved by another player in the ensemble, then the continuo player need not also resolve it.’ Unfortunately, I cannot find any eighteenth-century source offering corroboration. From the point of view of four-part continuo playing, this is of course incomprehensible, but in practice it works extremely well. It is more understandable when we think that in a larger ensemble context, a harpsichord is (and was) not always very audible. It is a fact that much of the artistic voice-leading of a continuo player will be lost if the harpsichord is too soft or too far away from the audience. This helps us to understand the realizations of the non-harpsichordist Tonelli, for sometimes one is obliged to take a heavy-handed approach to make the harpsichord heard.
Nowadays our ears are attuned to different playing conditions. I have devoted a large part of my musical life to making CD recordings, in which one's playing is extremely audible (if that is what the ‘Tonmeister’ wishes, at least). In the seventeenth and eighteenth centuries, there were of course no loudspeakers or monitors in opera houses to help the singers hear the continuo. It was for this reason that in 1660 two loud Italian harpsichords were imported for an Italian opera in Paris, along with the harpsichord builder who was to make a couple more instruments while there (Henry Prunières, L'Opéra italien en France avant Lulli (Paris: Champion, 1975), 243). Our role as continuo players in the context of a recording is decidedly different from our role in a concert, but the core remains: one must be at the service of the players one is accompanying. There is a nice rule from Austrian music theory that has stayed with me (although, unfortunately, I have forgotten the precise source): ‘Parallel fifths and octaves are sins that may be forgiven, but mistakes of musical taste are never so.’
Playing continuo is a creative discipline. Boring playing negatively affects those you are playing with. What is required is spirit, along with improvisation skills. Don't be too virtuous. Have fun, and then those who are listening will have fun, too! Change your role continuously from continuo ‘slave’ to soloist. Create a pleasant background, a warm bath, in which the soloist is free to find his or her way. Accompany, anticipate and be the supporting percussionist or rhythm section of the ensemble. Break those rules which can be discreetly broken. Be simultaneously the leader of the pack and the best friend of those you are accompanying. It is then that continuo playing becomes interesting and rewarding, and it is then that we can start to approach the excellent continuo players of the past.
