Sound consists of such things as the instruments or voices used, the sound they make, and the number of performers (one of the most hotly contested performance practice issues recently, for example, has been the number of performers Bach had or would have wanted for his concerted sacred music). For the instruments themselves, surviving instruments are of immense value, providing us with concrete items to be measured, analyzed for materials and methods of construction, played, “copied,” learned from, and admired. The different qualities of instruments made at various times and places can be a revelation in terms of understanding the music they were intended for. We become more and more inclined to require the perfect instrument for each repertoire. Harpsichordists nowadays need a French double — that goes without saying — but an Italian instrument is nearly as essential, not to mention a Flemish single (or double) and a late German double, maybe a Zell, Hass, or Mietke — maybe all three! At about $15-20,000 a pop, it’s enough to give heartburn to keyboard spouses everywhere. Voices, naturally, are in a different situation, especially since certain types of voices are not available now…except perhaps by computer synthesis. (Anyone seen the movie, Farinelli?) The performance practice focus for voice has been more on historical pronunciation which can have a striking effect on the sound.

But even with surviving instruments many questions remain: Are they fake? If not, how were they set up originally, what materials were used for the strings, how has time changed the shape or dimension of wooden parts, at what pitch were they originally intended to play? Experts are much better at spotting fakes than they used to be. (A late eminent organologist went through the instruments at the Cleveland Museum of Art in the mid-1960’s and pronounced a late-nineteenth century, highly decorated “Franciolini special” lute a magnificent Paduan instrument from around 1600 while passing over a real but plain tiorbino truly from that period as negligible.) Bass bars, bridges, and necks have been altered in many bowed string instruments, keyboards have been replaced in harpsichords, fingerholes have been enlarged in many woodwind instruments. All this needs to be assessed by experts before we have a context for the instrument.

Sometimes, little bits of string turn up — wound on tuning pins for example — and these can be analyzed scientifcally so that modern makers may try to reconstruct them. Ironically, we sometimes know more about the way things actually were than players today are willing to live with. Stradivarius’s violin test strings survive, for example, so we know that his G string was gut with open-wound wire, but not many players use such strings now because they’re unhappy with the sound and the feel of them. Modern makers need to keep experimenting with historical methods of manufacture, just as players need to keep trying new old things and learn from them.

Some original instruments are too fragile to be played, or impossible to restore to playing condition without destroying much of the original material and workmanship. Makers and restorers today are going through a crise de conscience about how much to do to original instruments that come into their hands. Will we know more about what to do with them in another 50 years? Many restorations, even from the middle of this century, now look like ham-fisted amateur jobs, obliterating the details of the originals.

Pitch is something else that we can sometimes learn about through surviving instruments. Organ pipes are one reasonably reliable source although, winter to summer, they might have varied as much as half a semitone because of temperature changes in the churches. Woodwind instruments are another useful source for pitches. But exactly how long was the staple on the reed intended for a certain oboe, and where, if you could move it, was the original place for the plug in that traverso? The actual picture is one of a profusion of different pitches in use at various times and places. If you know where the instrument was used, and when, and for what purpose, you’re in business — provided, of course, you know what note the player was reading in order to sound the pitch you have in mind! Transposition practices and multiple pitch standards offer much pleasure to the masochistic performance practice student, and much cause for disagreement among performance practice specialists!

To give one example of a difficult case: Bach’s organ, string and voice parts to Cantata 106,Gottes Zeit ist die allerbeste Zeit, are in Eb while the recorder parts are in F. What pitch did they sound at? Well, the organ was tuned at approximately A=460 and the recorders at about A=415, so they met in the middle, sounding in E at A=440. To add to the confusion, in the early 17th century, Praetorius calls the higher pitch Kammerton and the lower pitch Chorton, but by Bach’s time those terms had been precisely reversed. Should all Bach’s organ and choir parts be up a half-step from written pitch? Ah, but Cantata 106 was written at Mühlhausen, where Bach resided for less than one year. Organs in other places were at different pitches. You get the idea.

Another controversial case is Monteverdi’s Vespers of 1610. There is some evidence that it may have been performed with some — not all — of the movements transposed down a 4th (though probably from the Venetian standard of A=460), but not everyone agrees which movements, if any, need to be transposed, nor how far down they should go. Arguments about this can get very unfriendly.

Situation includes knowing the kind of physical space the music was originally intended for and how the performing forces were disposed within that space. Here we can look at actual rooms where they exist (and sometimes floorplans when they don’t) as well as seating plans for opera orchestras. We can read descriptions of events and account books which tell us who was paid for what services. Situation also includes knowing the function. A sonata, for example, might be for domestic chamber performance or it might be for performance during a church service. Does the function affect the form? Are certain repeats necessary that are not evident from the score because they were simply taken for granted? Should concerted sacred works that originally had a liturgical context only be performed with such a reconstructed context, or is it OK to perform them in concert? Not everyone agrees about these things.

Tuning & Temperament refers to the variety of systems devised to get around the fact that acoustically pure 5ths and acoustically pure major 3rds, both of which are desirable musical sonorities, cannot be completely reconciled to the 12 pitches normally available in the keyboard octave. Numerous writers offered methods, mostly in treatises, which attempted to combine the virtues of serviceability, key differentiation, and ease of setting. Historical temperaments tend to give more repose to the consonances and more drama to the dissonances; they create more distinction among the various keys and thus give more meaning to the composer’s choice of keys beyond the simple differentiation of pitch.

If you’re unfamiliar with the effect of different temperaments, you might appreciate this visual analogy: the difference between equal temperament and historical tuning systems is a bit like the difference between a drawing and a painting, or between a monochrome monitor and and color monitor. The picture is there, but it’s not as colorful or exciting. Another way to think of it is like spice in cooking: the dish is palatable “as is” but with the spice it’s more piquant and you enjoy it more, even if you find it unfamiliar the first time you try it.

It’s really impossible to overemphasize the positive effect of historical tuning systems, so it’s something that should be of vital concern to all members of an ensemble, not just the keyboard player. (After all, the keyboard player typically chooses the temperament and sets the tuning, but everybody else has to live with it from that point on, actively putting the notes where they belong!)

With Articulation, bowings, tonguings, and keyboard fingerings (and pedalings) are the obvious areas of concern, although we should also include right-hand technique for lutes and guitars, and so on. Performance manuals for different instruments are the primary source for this kind of information, although the level of performer they address can vary widely. For keyboard instruments, there are in addition a number of works that survive with fingering indications in the score. These can be especially valuable, since applying a theoretical system to an actual piece can be frustratingly difficult.

Another aspect of fingering is the physical approach to the keys, and this is something which we try to learn about through iconography — depictions of people actually playing, usually in title-page engravings and the like. How high the wrist is held, for example, can affect the joining of the small groups of notes typically seen in early fingering systems.

Notation deals with the realization of the written music. Some composers use unique signs, or they use common signs in a unique way. We are fortunate that writings by the composers, either in treatises or short introductions to their publications, sometimes elucidate their usage of such signs. Notational conventions like overdotting and inequality were applied to notated music as a matter of course by musicians. But does the inequality convention, recognized as an integral part of French style, hold true for music in the French style composed in Germany or England as well? What about Italian style music played by French musicians? Scholars are debating such questions now.

Tempo is something we get from the tempo-indication at the beginning of a movement, right? Allegro means fast, Adagio means slow, and Grave means deathly slow, right? Actually, those indications might be more accurately described as affect indicators: Allegro means ‘cheerful,’ Adagio means ‘in an easy manner,’ Largo means ‘broadly,’ and Grave means ‘serious.’ Only rarely is there some actual tempo implication in the term: Lento does mean ‘slow,’ and Presto means ‘fast.’ But what comes where in the order? Is Vivace (lively) likely to be faster or slower than Allegro. What do you think? Different theorists had differing opinions on that. Telemann, for example, in the preface to his cantata collection Harmonische Gottesdienst places Vivace between Allegro and Largo, a choice that I expect not many modern performers would have made. Another way we can derive information on tempo is through the situation category mentioned above: what are the acoustical limits of the original room?

At least as important to baroque musicians in choosing the tempo was the meter. Some writers, like the woodwind theorist Freillon-Poncein (1700), actually rank the meters according to implied speed. For example, 3/8 was the fastest simple triple meter, followed by 3 (a popular French meter), 3/4 and 3/2. So, even without affect indications, we can get some idea of tempo from the meter sign itself.

Another obvious source for tempo are dances. Many baroque movements are based on dance rhythms even when they don’t bear dance titles, and we are fortunate to have a notational system which has allowed us to reconstruct many of the dances in the French style. This gives us a tempo range between what is impossibly slow and what is impossibly fast, although those limits necessarily depend on the skill of the dancer — and the question inevitably arises as to whether the composer was actually thinking of the dance when he/she wrote a work in a common dance form since few such movements were probably ever used by dancers. A number of writers, too, give what might be called metronomic indications for various dances, using the scientific method of a swinging pendulum of a certain length. Seemingly failsafe, scholars have nevertheless managed to interpret these tempos in two radically different ways, however, since they disagree over whether the time interval given by the theorists represents a whole period or a half period. Along this same line, there are also some mechanical devices, usually called “musical clocks” which have provided evidence on tempo.

Dynamics for baroque music were long thought to be exclusively of the “terraced” variety, based largely on the sparse indications in scores, on the solo-tutti contrast in Italian concertos and the like, and on the limited dynamic capabilities of instruments like the harpsichord. The late David Boyden published an article in 1957 that “blew that theory away,” but it is still widely held. Basically, dynamics were the responsibility of the performer rather than the composer, so unless there was some unexpected but necessary effect in the music, the composer would just leave the score alone. There are plenty of indications scattered through 17th and 18th century manuscripts to suggest that there was much dynamic variability, and that gradual progress from one level to another was quite common. We also know that a certain amount of dynamic shading of individual notes was going on: the messa di voce (crescendo-decrescendo on a single note) introduced Caccini in 1601 was still being used, some might say overused, by Quantz at the end of the baroque era. As for “limited dynamic capability,” expert harpsichordists know how to “fill” the texture with more notes in order to increase the volume.

Improvisation is present to a certain extent in virtually every baroque piece for more than one instrument since the performance of the basso continuo requires that the player create something out of very little. Nowadays, keyboard players, lutenists, and even harpists are learning to improvise accompaniments based on the original figured or unfigured bass melodies. For those wanting to pursue it, there are numerous treatises, a few apparently written-out examples, and a plethora of modern writings on the subject. What is sometimes forgotten is that there were other aspects to improvisation in the baroque era as well. I like to quote an audition announcement for the church of St. Jakob in Stockholm in 1673:

The items that an accomplished organist should and must know are the following:

1. He should know how to play a praeludium…on any of the keys given him by a knowledgeable musician.
2. He should have learned how to perform fugally, as is customary, any chorale or psalm tone…
3. He must be given the theme of a fugue by a knowledgeable musician which he must be able to elaborate with at least four voices ex tempore.

And besides preludes and fugues, we might count also the canzonas, ricercars, fantasies, and toccatas among the improvised forms for keyboard and lute players. Yet another form in this category is the ground bass improvisation which, besides the possibilities for solo instrument, might also have been practiced by ensembles. Witness the numerous surviving settings of Passamezzo antico and moderno, romanesca, Ruggiero, and La Follia, some of which you may have played thinking of them in the same way as sonatas — but they are not — next time, try making up at least one of your own variations and see how gratifying it can be. Such endeavors require “good brains,” a good hand,” and the “bravery of the player,” according to the theorist Thomas Mace, but the benefits of working on that level of creativity can be incalculable for the performance of written music from the baroque era. In class, as a kind of therapy, I make my students improvise according to certain rules I give them in advance. Many of them are terrified at the prospect even though I assure them they won’t be graded on their success — only on their participation. When all your life you’re taught to play exactly what is on the page — neither more nor less — and to play it perfectly, it’s a big leap to play something that isn’t in front of you.