But if there is a need to rediscover it, how was it lost in the first place? There is much evidence that until Bach’s death in 1750 the musical tradition was very much continuous. The music played and sung in the Leipzig church services when Bach was Cantor (from 1723 until his death in 1750) was not confined to what we would now call baroque. Far from it. Interwoven into the service among Bach’s own compositions were chorales and plainsong chants going back one, two and three hundred years in an unbroken tradition.
After Bach’s death however, music took on a different style, and, perhaps for the first time in musical development, the older style was considered “unfashionable”. There was a major break with the past. The break was not complete of course, but it was such that baroque music could be “rediscovered” a hundred years later as something of a new revelation. Mendelssohn and his sister played Bach regularly in their home, their favorite works being the "48" Preludes and Fugues. And it was Mendelssohn's promotion of the St Matthew Passion in 1829 which marked the first public "revival" of Bach and his music.
This ground-breaking performance was given with a choir of some three or four hundred - certainly not at all as Bach would have known it. But this was not an intentional "romanticization" of Bach; rather it was something of a relief to Mendelssohn personally, and ultimately a triumphant re-birth.
Mendelssohn himself, though personally enthusiastic and dedicated to the revival of the St Matthew Passion, had been very dubious as to its reception by the public; indeed he was equally dubious as to his choir’s reception of it. If they didn’t like it they would slowly drift away. Would he be left with a “choir” of four voices by the end of rehearsals?! And would anybody be at the concert to hear it? In fact the opposite was the case. Such was the choir’s enjoyment of this work that its numbers swelled with every rehearsal as the word went around. Hence the rather over-large vocal section, which may not have been ideal for a Bach performance, but in the circumstances made Mendelssohn very happy!
Wilfred Blunt gives this account in his detailed and perceptive biography of Mendelssohn:
The choir of the Singakademie, which it had been feared would gradually fade away as rehearsals proceeded, grew ever larger and ever more enthusiastic under Felix's inspired direction: they noticed, too, that he knew the work so intimately that he dispensed with a score; his musical memory was extraordinary.
Excitement in the cultural world of Berlin mounted as the great day approached, for the choirs, three or four hundred strong, had passed the word round that the work was a revelation: that old Bach was after all capable of drama, passion and melodiousness; that here was 'an architectonic grandeur of structure' undreamed of by those familiar only with his smaller instrumental works.
Everything had gone so well it seemed impossible that on the day itself the standard achieved during rehearsals could be surpassed. Yet it was. 'Never', wrote a contemporary participant, 'have I known any performance so consecrated by one united sympathy. Our concert made an extraordinary sensation in the educated circles of Berlin. If only old Bach could have heard our performance!' The King and his whole Court were there, and the hall was completely full. More than a thousand people had been unable to get tickets, and two further performances which were called for followed almost immediately.
That performance marked the beginning of the move towards what we would now call a baroque revival. The Bach Gesellschaft (Society) began in 1850 the task of publishing all Bach’s works (all that could be found that is), a project which they completed fifty years later in 1900.
In the early 1900s, Wanda Landowska “re-invented” the harpsichord, which had been almost completely supplanted by the piano for home and concert performance. Her great "iron grand" was, to say the least, unlike anything built during the baroque period. But to have the prestigious Paris firm of Pleyel temporarily abandon their piano manufacture in order to attempt a re-creation of this peculiar antique instrument was a major pioneering achievement in the rediscovery of the baroque. Landowska’s performances, incidentally, though the recorded sound is not of today’s technical quality, are still exemplary, and Landowska's interpretations are rarely matched today in their insight and technical precision.
So the movement of rediscovery gradually progressed. In 1950 the advent of the long playing record created a new vehicle and a new public for classical music, followed in 1960 by stereo with the parallel improvement both in recording equipment, and in the standard of home sound reproduction.
During the 70s and 80s further valuable research was conducted into the music and performance of the baroque, applied in practical recording and concert performance. It was during this period that performances began to bear the title "authentic" or "on period instruments".
At the same time however, it should also be understood that performance to a major degree reflects the spirit of the times, and some of today's "authentic" performances have less to do with historical accuracy, attempting rather to produce a performance which, in John Eliot Gardiner's words, will "excite modern listeners". Thus it is that "authentic" performances, while aiming to please modern tastes, often make presumptions which lack historical authenticity and which fail to bring out the full potential inherent in the music.
Nor has "authenticity" fully percolated through to the recording and balance engineers, who are still failing to pay enough attention to contrapuntal clarity which requires very delicate balancing - the harpsichord particularly suffers consistently from adverse balancing treatment.
Tempi, balance, and instrumental timbre: these key issues of performance and recording practice may be reviewed one by one.
Many "authentic" performances of Bach's cantatas adopt a fast, almost racy tempo which would never have been considered or tolerated in the staid atmosphere of a Lutheran church service in 1730. Tempi if anything would have been slower and more deliberate than we today would probably want to accept. Likewise many “authentic” performances of orchestral and solo works adopt a tempo the speed of which may display the players’ dexterity but obscures much valuable and enjoyable detail. Many a time I have listened to a racy performance; having heard many slower and clearer performances and studied the scores I at least know what I’m missing, but I feel sad for those listening for the first time, who will miss so much wonderful detail. The tempo should never be faster than that which will allow the fastest (=shortest-value) notes to be articulated clearly.
Another issue of authenticity might also be considered in relation to tempi: the question of relative tempi as between movements of a concerto. Many believe that the ultra-slow middle movement contrasting with excessively fast and often hectic outer movements was a 19th century creation. The respected Romanian/French conductor/composer Georges Enesco believed that the three movements of a baroque concerto (or sonata for that matter) should be approximately equal in duration, that the slow movements should be faster than current practice, and the "fast" movements should be slower. He put this principle into practice in his wonderful Bach clavier concerto recordings, now long ago deleted. Internal evidence of the music itself suggests that the difference between the two outer, and the middle movement was one of character not speed. The outer movements would be lively and outgoing, while the center movement would be more introspective or lyrical. Thurston Dart, a major pioneer in the search for authenticity in performance during the 1960s, was also of this view.
As a simple rule, "slow" movements should move along gracefully, never drag, while "fast" movements should never express haste, and should always respect the player of the fastest notes, so that every note is distinct. As Alessandro Scarlatti wrote in a letter to the Grand Duke Ferdinando de' Medici: Where 'grave' is marked, I do not mean 'melancolico'; 'allegro' should be judged so that too much is not demanded of the singer. And the literal translation of the Italian Vivace is "lively" - not ultra-fast!
Many "authentic" performances also adopt unsteady tempi, so that the music seems to move in waves, or fits and starts, ignoring the fact that a regular tempo was universally accepted in baroque times when the major concern was keeping unruly players and singers together. Indeed it was quite usual for conductors to beat time with a heavy object on a desk, or, more commonly still, on the floor with a staff. The French composer Lully was conducting a Te Deum to celebrate Louis XIV's recovery from illness; he was banging loudly on the floor with a staff when he struck his foot with such force that it developed an abscess, from which the unfortunate Lully died shortly after. Slow, steady and deliberate tempi were the order of the baroque day. And clarity of contrapuntal line was paramount, which itself dictated a slow and deliberate rendition.
The importance of contrapuntal clarity leads to the issue of balance. Many recording engineers and studios will record a harpsichord concerto one session and a piano concerto the next; in both cases there is a keyboard soloist set against the orchestral background. Yet while the piano is given prominence in the piano concerto, the harpsichord will be pushed into the soundscape background for the harpsichord concerto! It is not always easy to find recordings of Bach's harpsichord concertos in which the harpsichord is given correct prominence; and as for the poor Fifth Brandenburg Concerto, often (and rightly) billed as the world's first Triple Concerto, it seems impossible to find one recording from the zillions available in which the harpsichord is given the same prominence as the violin and flute, the other two solo instruments. The harpsichord is doing lots of wonderful things, but balance and excessive speeds usually render most of the player's work inaudible. When the harpsichord emerges into its beautiful solo cadenza it is barely audible; indeed in one recent "authentic" recording the harpsichord volume is actually turned up for the solo, either during the recording, the mixing or mastering - a shameful practice hardly worthy of any self-respecting recording company!
The tradition of the Inaudible Harpsichord is probably based on the perception of the harpsichord as being solely a continuo instrument, there only to keep the rhythm and to fill in the background harmony. While this may be a true reflection of the harpsichord's major traditional role, this fine instrument was obviously much more significant to Bach, who pioneered its use as a solo-in-concerto.
Another example of poor balance which fails to reflect Bach's own view of the harpsichord's role can almost universally be found in his sonatas for violin and keyboard. These were written as 3-part Trio Sonatas, one part for each keyboard hand, and the third for the violin. But once again the harpsichord is generally relegated to the rear of the sound spectrum, the result being an almost solo violin with a faint tinkling in the background. Thus when the counterpoint moves from violin to harpsichord it is all but lost. The same applies with the sonatas for flute - or viola - and harpsichord. In choral music too, balance is often inappropriate musically, when for example the choir is given prominence over the instruments, although baroque composers generally and Bach in particular wrote equally for instruments and voices, taking the musical lines freely from one to the other. A good Bach performance and recording might be summarized in one simple objective: "if Bach wrote it, the listener should hear it".
In the case of Bach's cantatas and choral works, most performances and recordings use a small portable organ for the continuo, ignoring the church's main organ in the gallery. This arrangement, while no doubt preferred by conductor and recording engineer, may be adequate for the accompaniment of arias and recitatives, but not for the opening choruses in cantatas such as 29 and 146 which feature what amounts to a solo organ concerto movement. Here the thin, almost pitiful sound of the little chamber organ is simply not up to the task. This is one fine aspect (among many others!) of Karl Richter's cantata recordings for Archive, where the big organ is always used. In the accompaniment of concluding chorales too, the big organ provides what Bach would call gravitas. A document in Meissen Cathedral written by JF Doles (1715-97) gives detailed organ registration recommendations for the performance of Preludes, and for the accompaniment of chorales sung by the congregation. Since Doles was a pupil of JS Bach for five years (1739-44) his recommendations may be presumed to reflect Bach's own views. Full organ is recommended for congregational accompaniment, including the 16' Posaune in the pedal. Doles incidentally, after a spell at Freiberg Cathedral, took over Bach's old post as Thomaskantor in 1755 which he held until his death.
The third major issue to be considered is that of timbre, or sound quality produced by the instruments, in particular the violin and the harpsichord.
The harpsichord sound generally associated with most "authentic" performances is, in the words of one outspoken reviewer, “tinny and jangly”. Was this the sound Bach would have preferred from his own harpsichord? This is not a rhetorical or unanswerable question, for we can ascertain Bach's taste in harpsichord sound with some accuracy. To begin with, the harpsichords built today as copies of baroque instruments normally copy the lighter French and Flemish designs. German harpsichords of the baroque period however, were much heavier and more solid, giving a deeper, richer, rounder tone. Even this was not entirely satisfactory to Bach, whose ideal was a harpsichord more resembling the soft tone of the lute. In pursuance of this ideal, Bach had two "lute-harpsichords" custom-built with gut strings and other modifications rendering an even gentler, more rounded sound. Though an actual example of a lute-harpsichord has not survived, there are historical records of orders and specifications, and a wonderful reconstruction has been made for Gergely Sárközy (among others) whose equally wonderful performances of Bach can be heard on the Hungaraton label. It would seem fairly conclusive that this is the sort of sound Bach would have preferred. For more detail about the lute-harpsichord check the link at the bottom of this page.
Nor was Bach in any way opposed to the use of the 16' stop in harpsichord performance (as "authentic" performers imply); the inventory of Bach's possessions at the time of his death reveals that as well as two lute-harpsichords, he owned several harpsichords of which his main instrument had a 16' stop on its lower manual.
Also fashionable in "authentic" circles is to scorn such "bells and whistles" as foot pedals for registration changes, and - perish the thought - Venetian swell-shutters for volume variations. Once again however, "authentic" ideology disregards historical accuracy. Among the instruments in the Fenton House Collection, London, is a harpsichord built by Burkat Shudi and John Broadwood in 1770. In addition to its six hand stops, it also has three pedals to control the lute, machine and buff stops for quick registrational changes. When the machine stop is put into the 'on' position the upper keyboard commands the upper eight foot stop, and the lower keyboard the three sets of strings. On depressing the lute pedal this combination is changed to: upper keyboard, the lute stop: lower keyboard, the lower eight foot. The case also contains a Venetian swell, an inner lid consisting of eleven hinged shutters covering the whole soundboard area. These shutters can be opened by depressing the buff pedal, which permits crescendo and diminuendo, and also alters the tone color.
Similarly in the case of the violin, the research movement into "authentic performance" has totally overlooked one very important aspect of baroque performance on stringed instruments generally: the ability of the performer to produce true chords, a technique which required a type of bow widely used in German baroque performance.
The German baroque violin bow was quite different from its Italian counterpart, reflecting differences in German musical taste. The Italian bow was slim, light, almost straight, and very similar to those in general use today. The German bow was heavier and deeply arched; the strings were loose, and the tension was maintained by the pressure of the player's thumb which was placed under the bow strings. The more cumbersome method of holding the bow which this required, would have dictated slower performance speeds. But more significantly, the tension, being maintained by the player’s thumb, could be tightened for single-line melody, or loosened to play chords on three or all four strings simultaneously. This technique was expounded by one Emil Telmanyi who recorded Bach's solo violin sonatas and partitas for German Decca (Das Alte Werk series) many years ago using a modern reconstruction of the baroque German violin bow. His performance brought out for the first time the alternation between chord and solo line which is such an important feature of Bach's solo string writing - all other performances play broken arpeggios which are not the same as true chords.
There need be no doubt as to the historical validity of the arched German baroque bow, with its associated technique of using the thumb to control tension and play chords or single line as required. A cursory glance at the frontispiece to the Musikalishes Lexikon published in 1732 and edited by Bach's cousin Gottfried Walther clearly shows players using arched bows, their thumbs holding the tension of the bowstrings. And there is other documentation, as for example in the written comments by Georg Muffat (1698). A further point is that Bach was not the sort of slapdash musician who would write chords for an instrument incapable of playing them. His solo flute sonata has no “chords” which the player must replicate with arpeggios. Bach wrote chords in his solo string sonatas and partitas because chords were what he intended to be played and chords were what he himself would have played (he learned the violin at an early age and was very fond of the viola). In the absence of a revival of the baroque German bow and a fund of expertise in its use, the only way at present to render a truly authentic performance of Bach's solo violin and solo cello sonatas and partitas - authentic in the sense of how Bach visualized and would have heard them - would be to use a quartet.
Fuller detail on the German baroque bow together with the illustration mentioned above and several others can be found by following the link at the bottom of this page.
Another very prominent feature of string playing in today's "authentic" performances is the almost total absence of vibrato, resulting in a flat, plaintive and lifeless tone. It seems quite unclear as to where this aspect of "authenticity" derived from, since much evidence supports quite the contrary view. A star pupil of Corelli, Geminiani moved from Naples to London in 1714 and was to become the most important Italian violin virtuoso resident in Britain, also teacher, composer and the author of an immensely influential treatise addressed to advanced players, The Art of Playing on the Violin (1751) as well as several other advanced musical treatises. He published several challenging collections of violin sonatas which require dramatic flair from the player. Geminiani provided ornaments for both slow and fast movements as well as cadenzas in his treatise; he advocated the use of vibrato 'as often as possible', and the expressiveness of his playing was much admired by both Hawkins and Burney. Vibrato was also well known and much valued during baroque times in its application on the clavichord, known in German as Bebung. There are also many references in baroque musical literature, both to the importance placed on warmth and vibrato in vocal performance, and to the ideal in violin playing of replicating the human voice. A further, more practical consideration arises from the fact that musical instruments in those days were not maintained to the same high standard of tuning as they are today, since they were usually stored and often played in damp, cold conditions. It was recommended that instruments play in groups of at least three in order to minimize this problem, and vibrato likewise would have helped overcome the perception of imperfect tuning.
A word or two might also be said about the use of the term "period instruments". Some rare and unusual instruments have indeed been revived, but with very few exceptions, these are more relevant in the performance of mediaeval music rather than baroque. The use of the (wooden and much softer) baroque flute is important, as opposed to its more strident metal counterpart - this is a matter of balance, between the flute and harpsichord or other instruments. In the case of stringed instruments however, few stringed instrument players of any standing have ever used an instrument produced more recently than the mid-1800s. The violin, viola and cello players recorded in the 1960s, 70s and 80s were all proud owners of original baroque instruments. Thus the use of the term "period instruments", while it may be indisputably accurate, should not be taken to imply that this is an exclusively modern-day revival phenomenon.
Clarity of diction in arias and recitatives
It is important to understand that baroque listeners to cantatas and other sacred works were not there simply for the musical entertainment, as may be the case with today's listener. Bach's cantatas, for example, formed a major part of the lesson for the day, and his texts were expected to reflect the gospel and theme of the sermon. Arias and recitatives were an essential tool in the telling of a biblical story or the communicating of the lesson. Thus the diction would always have been clearly articulated, the voice projected from the front of the gallery for the whole congregation to hear. In many performances of vocal works heard today it is difficult to tell what language the singer is using, let alone the precise words. Such a performance on the part of the vocalist is neither authentic nor useful.
The Baroque Spirit
While the debate on authenticity in baroque performance will continue, certain essential characteristics should be present, if the performance is to reflect the true baroque spirit. The musicians must first and foremost show a respect and an affection for the music; this is most important. A violinist or singer performing with real sensitivity, even just for a few lines, immediately seizes one's attention. Tempi also are extremely important; if the tempo is too slow the piece drags; too fast and vital detail is lost as the musicians scramble to grapple with unnecessary challenges of physical dexterity. Too many performances today reflect this unseemly haste. Balance is vital too, so that everything can be heard.
In the performance itself it is very important, particularly in the works of JS Bach, to display the "architecture" of the piece, especially in his organ preludes and fugues, many of which are constructed in the form of an arch with side pillars at beginning and end, curves and a keystone at the top, with excursions into carved embellishments along the way. A good performer will study the architecture and reflect it in performance through changes in registration. All too often today such considerations are dispensed with, and indeed there is a school of thought which supports performance from start to finish on one manual with one selection of stops, overlooking the fact that much of Bach's organ-writing was produced specifically for and at the request of students and colleagues who wanted pieces which would show off the full power and also the individual sounds of newly built organs which they had been invited to test and approve.
There is a spirit to every age, every composer, and every piece of music. In baroque times secular and sacred life were very much inter-related, and music was to be enjoyed, but also respected as a spiritual gift. Bach spoke often of a piece, its performance, and the instrumentation or style with which it was performed as requiring gravitas. More importantly, the spirit of the baroque is characterized above all by clarity, for the music is very contrapuntal (fugal/canonic) and every note, every line has its place. Love and respect for the music, enjoyment in performance, and above all, clarity in the articulation, ensemble and recording balance. These are the true essentials of baroque music. If performance practices billed as “authentic” on "period" instruments can reveal these qualities and this spirit then that is true authenticity. If modern instruments can do the same, then that too is authenticity. It’s the spirit that counts.