Sonata, K. 322 Domenico Scarlatti

Diffi cult do you call it, Sir? I wish it were impossible! —Samuel Johnson

he transcriber looking through keyboard is played on the instrument for which it was Tmusic for suitable material must constantly originally written. It is particularly a problem ask herself a simple question: Is it even possible with music written between 1600 and 1817. Be- to play this on the guitar? Th ere can be a wide fore this span there was a theoretically reliable range of transcribers’ willingness to modify the correlation between time signature and tempo. original music, but all too often we are simply All tempos were numerically related to a basic, confronted with the limitations of the guitar presumably universal, beat—the tactus—which as compared with a keyboard instrument. An was about 60 to 70 beats per minute. Th is men- objective assessment of these limitations, while sural notation allows us to establish the tempos dismaying, is also necessary in producing use- of modern performances with a high degree of ful musical transcriptions. So how does the gui- confi dence. tar measure up as a solo instrument? Well, its In 1817 Beethoven was the fi rst well-known pitch range is rather narrow. Its dynamic range composer to use a recent invention, the metro- is weak. It can only produce four notes simulta- nome, to establish tempos for his music. From neously and has relatively modest contrapuntal this time on, a composer’s intentions about agility. Th ere are innumerable combinations of tempo could be accurately communicated to the notes that cannot be played because they are ei- performer. It turned out, however, that even this ther too close together or too far apart, or both. did not solve the problems. Beethoven himself is It cannot sustain notes and so can only suggest said to have radically changed his own previous connected melody. Its variety of timbres is less metronome markings and he once wrote that than the organ or voice. And, a few superhu- the metronome was “of no value.” man players aside, for most of us mortals the Many guitar transcriptions are from the pe- guitar can produce notes at only relatively slow riod with the least agreement about tempo, that tempos. Any one of these limits may eliminate is, the Baroque and Classical periods. After a keyboard piece from serious consideration 1600 tempo words were increasingly used, while for transcription to the guitar, but the last one, mensural notation declined. Tempo words car- tempo, is the focus of this article. Transcription ried little precision, however. Each theorist of inevitably involves modifi cation. Th e question I this era seems to have had his own table of tem- will explore is: How is the tempo of a piece af- po words. Not only do these tables contradict fected by transcription to the guitar? each other as to what speeds to attach to partic- ular words, sometimes they even disagree as to Tempo and Performance the order of quickness, that is, whether allegro is faster or slower than allegretto, etc. Compound- Th e selection of the tempo of a piece of music ing the problem were the overlapping defi ni- is hardly a trivial matter, even when the music tions of tempo words. Some were descriptions of

26 moods, as in allegro (cheerful). Others described musical continuity”—there are still real, practi- speed, as in lento (slow). Over time some mood cal limits to this principle. words came to have speed meanings, but others What of the other end of the spectrum? It did not. Th e result is considerable uncertainty. seems that there are reasons for a lower tempo Underlying most of the debates about tempos giusto on the guitar. First, there is the often was another concept that was thought to bring overlooked fact that the guitar is a low pitched some coherence to it all—the tempo giusto— instrument. Since guitar music is usually written but even this term brings confusion in that it is with an octave transposition—or with a small sometimes used to mean “strict tempo.” It is the “8” at the bottom of the clef—it is easy to forget other meaning to which I refer: Tempo giusto that the E on the fi rst string twelfth fret is only means the “right tempo.” Th e idea was that each the top space on a regular treble staff , and the piece of music has an implicit and discoverable open sixth string is below the bottom of a staff correct tempo. Leopold Mozart said that infer- with a bass clef. Th e overall pitch of an instru- ring the correct tempo from the score “infalli- ment has implications for tempo. Lower pitches bly shows the true quality of a musician.” Th e are slower to speak. Larger forces and move- implication was that the tempo words could get ments are required to set the string in motion, you to the right neighborhood and then your re- and the acoustic events that follow the plucking fi ned sense of the tempo giusto could take over of the string take longer to play out. As a result, for the fi ne tuning, so to speak. rapid notes have less clarity than they would if I will show you some hard evidence that bears higher pitched. In addition, the guitar’s shorter on this question a little further on, but we must string lengths, compared to, say, the harpsi- fi rst recognize that the foregoing has shed little chord, mean lower tensions and less clarity. light on the central question about tempo and A more subtle point relates to diffi culty. transcription. Th e question might be reworded Technically, the guitar presents monumental at this point: Does the tempo giusto of a piece obstacles to speed compared to a keyboard in- of music change when it is played on a diff er- strument. Each note played on the guitar usu- ent instrument? In fact, I have found in music ally requires two fi ngers. Successive notes re- literature no direct reference to this issue at all. quire extremely precise coordination of fi ngers Th e only remaining avenues for investigation on both hands. Th e point is not just that a gui- are unfettered, but informed, speculation and tar cannot be played as fast as a keyboard, it is observation of the actual performance practice that the perceived diffi culty, including speed of the best players. and complexity, is part of the total experience of the listener. To quote Kirkpatrick again, “Per- Tempo and Transcription haps the very opposition with which the artist is confronted by the tools and materials of his Th ere are several qualities of the guitar itself that medium is a stimulant necessary in some mea- may aff ect tempos. Certainly, as an instrument sure to the intensity and concentration of his that cannot sustain tones, the guitar has defi nite communication through domination of that lower limits of reasonable tempo in slow pieces. medium.” Most guitar music played on the pia- Bowed string instruments, organ or voice can no sounds thin—it presents little opposition, in shape a coherent line using tempos that would Kirkpatrick’s words—and it is natural (“giusto”) completely fragment the melody on the guitar, to play it faster. Conversely, the guitar gives us or the piano. Although—as more opposition, and hence, slower tempos may wrote about nonsustaining instruments, “It is fi t better. not sound but mental expectation that produces Earlier in this article I made unfl attering com-

27 parisons of the guitar’s capabilities, and of course performer. To cut down the size of the search I only told half the story. While on each of the di- focused on two composers: Domenico Scarlatti mensions I listed (pitch range, dynamics, com- and Enrique Granados. Both composers wrote plexity, etc.) the guitar is far surpassed by other keyboard music that has been transcribed and instruments, it may be that no other instrument recorded many times, and so provided lots of has as great a variety of these dimensions within data points from which to observe patterns and which one can create artistic nuances. It may draw conclusions. I later narrowed the search be the most versatile of instruments, in that it further when I found out which particular can produce distinctions in more ways. Th e vio- pieces had been recorded the most. While there lin is sadly lacking in contrapuntal ability, the were a few recordings in my own collection and harpsichord cannot vary loudness at all, and the at the local library that I could actually listen to piano always has the same timbre. Th e guitar and measure with the infamous metronome, I can do these things and more, simultaneously. needed more information. How I obtained this Well played guitar music presents the listener was a fascinating experience that I hope will be with a wealth of subtle variations and nuances a useful digression for readers. Th ose of you who of sound. Th is is a lot of information to process. are already connected to the Internet may fi nd Perhaps the tempo giusto depends on the rate of a few new resources mentioned here; those who information. If this is so, then slower tempos are have yet to connect may fi nd that there are good necessary in performing guitar transcriptions. reasons to do so. Of course, this assumes that the player is doing Without access to many actual recordings, I something artistic with all of the instrument’s resorted to the next best source of information potential. Sadly, we have all heard fast players about tempo—the timings that were listed for that were giving us only rapid notes. Th e British the music. Th e samples that I could listen to guitarist Stephen Kenyon wrote on this point, confi rmed that the overall timings of pieces gave “We are used to hearing a certain amount of a reliable measure for comparing average tempo. information—hopefully what the player intend- Th e advantage of using this measure was that it ed—and if they play too fast sometimes we do gave me access to many recordings through li- not recognize that and then subconsciously we brary catalogs and other people’s collections. In think ‘this player’s boring.’ Th is certainly hap- fact, all the data that I will describe was collect- pens with some very fast players I can think of. ed from my home computer. I have listed some Th ey proudly play perhaps Bach or Paganini at sites at the end of this article. exactly the tempi of the fi ddlers, but after the On the Internet I was able to get into the gasps of astonishment have subsided—assuming online catalog for a network of 13 college and they do—they are often, I fi nd, boring.” university libraries in Oregon and Washington (ORBIS). In one afternoon I also accessed the Sources of Information entire University of California catalog (MEL- VYL), the New York Public Library, and the Th e foregoing discussion is mostly speculation. United States Library of Congress. All of these I felt a need to fi nd some hard data in hopes electronic catalogs allow detailed, automated that it would provide a fi rmer foundation to searches. I simply had to type in what I was my theorizing. Leopold Mozart’s advice seemed looking for—for example, audio recordings of a logical way to pursue the question. I decided Sonata K. 322 by Domenico Scarlatti. Almost to fi nd as many examples as I could of music instantly I could scan all of the catalog entries played on both the guitar and keyboard to see that applied. Although only about 20 percent of how tempos varied with the instrument and the the catalog listings included the timings of indi-

28 vidual pieces, I could quickly fi nd a great deal of example. Th e CGML is deftly and diplomati- relevant information in a short time that I could cally administered by the Portland, Oregon gui- not have collected at all without the Internet. tarist and teacher, John Philip Dimick. A second source of information was an e-mail discussion group devoted to the topic of classical Evidence from Performances guitar. Group members contributed both opin- ions and data from their collections. Known as For this article I have collected the timings for the Classical Guitar Mailing List (CGML), this two pieces. Th ey are well known works that is a group of several hundred people from all have been frequently recorded. Th e fi rst is the over the world with a common love of the guitar Harpsichord Sonata in E Major, K. 380 by Do- who engage in lively and interesting discussions menico Scarlatti. Th e second is the Intermezzo through e-mail. Each message is automatically from Goyescas by Enrique Granados. Th e per- sent to all members of the group or “list” and formance data for these two pieces is typical of an average day includes about 30 messages. Th e all that I collected. Th e following tables list the exchange of ideas from players—both amateur performer, the instrument(s), and the timings. and professional—writers, luthiers, teachers, and students is stimulating, educational and a Harpsichord Sonata in E Major, K. 380, useful source of information, as in the present Domenico Scarlatti

Anthony Newman harpsichord 4:25 Wanda Landowska harpsichord 4:35 Trevor Pinnock harpsichord 4:45 Ida Presti and Alexandre Lagoya guitar duo 4:47 Ivo Pogorelich piano 4:59 Ida Presti and Alexandre Lagoya guitar duo 5:03 Eliot Fisk guitar 5:04 John Sankey MIDI harpsichord 5:04 Dinu Lapatti piano 5:12 John Williams guitar 5:16 Eliot Fisk (1985) guitar 5:16 Manuel Barrueco guitar 5:20 Dinu Lapatti piano 5:26 Groniger duo guitar duo 5:26 Kaare Norge guitar 5:35 John Browning harpsichord 5:42 Mikhail Pletnev harpsichord 5:44 Edoardo Catemario guitar 5:47 Igor Kipnis harpsichord 5:56 Rafael Puyana harpsichord 6:00 Christain Zacharias piano 6:04

29 Vladimir Horowitz piano 6:20 Meiko Miyazawa piano 6:21 Vladimir Horowitz (1968) piano 6:22 Carlos Barbosa-Lima guitar 6:22 Luciano Sgrizzi harpsichord 6:40

Th e fi rst table shows harpischords as the quick- are further back. Overall, there seems to be very est, but Presti and LaGoya and then Eliot Fisk little correlation between tempo and instrument. are very close. Other eminent harpsichordists

Intermezzo from Goyescas, Enrique Granados

Sergio & Eduardo Abreu guitar duo 4:03 Arthur Fiedler Boston Pops 4:21 Casals and Mednikoff cello and piano 4:28 John Williams and guitar duo 4:45 Trio Sonata Flute, oboe and guitar 4:53 Paul Brody and George Brough saxophone and piano 5:05 Raphael Frühbeck de Burgos New Philharmonia 5:16 United States Marine Band military band 5:23 Marek Jerie and Konrad Ragossnig cello and guitar 5:25 Fritz Reiner Chicago Symphony 5:31

Th e second table has a wide, and sometimes of the Scarlatti sonata are at 70 beats per min- improbable, array of instruments, but again, ute; the fastest is more than 105 beats, a full 50 there is no correlation with tempo that I can percent increase. see. Guitars and orchestras span the whole range My conclusion from all this is that as tran- of tempos at which the Intermezzo has been re- scribers and performers, we can, with ample corded. Th e objective evidence seems to show justifi cation, ignore speed contests and com- that there is not a correlation between instru- parisons, and instead confi dently select the mu- ment and tempo. sic for which we have an affi nity and fi ll the si- But can we also conclude that the mystical lence with the endlessly varied sound that can tempo giusto is maintained as a piece of music is be uniquely produced by our guitars and our transferred from instrument to instrument? Not selves. at all! And this is what I fi nd most striking about the data. Th ere is no evidence whatsoever for an Featured Transcription implicit tempo in this music. Th ere is a tremen- dous range, and smooth distribution, of tempos Th e transcription I selected for this issue is chosen for the same music by skilled and highly one I made in the course of my research for regarded performers. Th e slowest performances this article. It is one of Scarlatti’s most famil-

30 iar harpsichord sonatas, and one that has often play it no faster and no slower than that which been recorded on the classical guitar. In my makes perfect sense to you. search I found eight guitar and fi ve piano re- Two fi nal notes: Th e Longo number for this cordings; oddly enough, there were no harpsi- sonata is L. 483, but I have seen two references chord recordings. I thought it would be a good to it—one on a John Williams album cover—as piece for readers to try out with the discussion L. 485. Also, there is a marvelous website listed about tempo in mind. in the references (below) that has a huge num- While Scarlatti is best known for his more ber of downloadable MIDI fi les. Included are all than 500 harpsichord sonatas, it is impor- of the Scarlatti sonatas performed by the harp- tant to remember that his father and teacher, sichordist John Sankey. Th ese and other similar Alessandro Scarlatti, was a prolifi c opera com- sites are a magnifi cent resource for transcribers poser, and Domenico himself produced mostly who are listening for music that may be suitable operas until rather late in life. Th e sonatas bear for the guitar. all the signs of his early experience. Th ey are, As always, I look forward to hearing readers’ above all, theatrical. Phrases call and answer reactions, comments, and suggestions. each other, voices join in duets, hunting horns sound and horses gallop, bulls charge and thun- References and Suggested Browsing der crashes, all in contrasting episodes highly evocative of dramatic scenes. Th ere is tremen- Classical MIDI Archives, www.prs.net/midi. dous opportunity here for the broad use of tonal html contrasts and agogic accents. Th is music is the Cyr, Mary (1992). Performing Baroque Music, strongest argument I know to refute the stereo- Amadeus Press, Portland, Oregon. type of Baroque music as mechanical and cere- Donington, Robert (1982). Baroque Music: Style bral. and Performance, W.W. Norton, New York Th e Sonata in A Major, K. 322 is marked Kirkpatrick, Ralph (1984). Interpreting Bach’s “Allegro,” which, as we have already seen, means Well-Tempered Clavier, Press, diff erent things to diff erent people. Narciso New Haven. Yepes blazed through it at 111 beats per min- University of California Libraries, www.melvyl. ute; John Williams, Carlos Barbosa-Lima, Ra- ucop.edu chel Gauk, and Eliot Fisk were all clocked at U.S. Library of Congress, www.loc.gov about 100, Andrés Segovia came in at 90, and a pianist, Andras Schiff , played it at a leisurely 77 beats per minute. My suggestion is that you

31