Elliott Carter Studies Online

VOLUME 2 (2017)

Studying Carter’s Music through his Compositional Sketches: Two Instructive Examples

Jonathan W. Bernard

Elliott Carter, Allegro scorrevole, sketch page

Elliott Carter Collection, Paul Sacher Foundation. Used by permission.

The study of composers’ sketches and manuscripts is a fixture of historical musicology, and has been so for some time; but it is only within the past few decades that any significant attention has been devoted to the documentary record of composers from the relatively recent past or, indeed, of composers who are still very much alive. This state of affairs, of course, is not to be taken exclusively as a judgment upon the worthiness of the work of contemporary or near-contemporary composers as a subject of protracted study. Anyone who might have wanted to spend a significant amount of time closely scrutinizing the sketches of Edgard Varèse (d. 1965), for example, would have found it impossible to do so until his papers were placed in an archive and made available to scholars in 2004. But apart from questions of sheer availability, the neglect of what are usually called primary sources for many twentieth-century composers (and now including those who have lived into the twenty-first) can probably be attributed to the fact that the music of the twentieth century, in historical-musicological circles, was not even widely accepted as a fit subject for any kind of study until the century itself was nearly over – perhaps owing to a general reluctance to embark upon the study of any repertoire that was not yet sufficiently “historical.” And this reluctance was bound to be even more strongly evident in the case of music that was not yet, properly speaking, historical at all.

If this predisposition against the documentary study of music of the past century has largely melted away in recent years – starting with early twentieth-century giants such as Schoenberg, Berg, Webern, Stravinsky, and Bartók, among others; moving on to those whose careers were principally or wholly made after 1950 – we can give at least partial credit to two major developments. One is the modern rise of music theory as a discipline separate from historical musicology, starting in North America shortly before 1960, spreading during the next few decades to Great Britain and the Western European continent. From the beginning, in the United States at any rate, the field always numbered among its practitioners a significant group of scholars whose principal interests lay firmly in the twentieth century, and who were best prepared and motivated to explore the sketches of composers from this period as at length they became accessible at various libraries and other archival institutions. The inauguration of such study was still occurring in largely piecemeal fashion when, in 1986, came a second major development: the opening of the Paul Sacher Stiftung in Basel, Switzerland, the first and still the only institute in the world solely devoted to the study of twentieth- and twenty-first-century music. Over the past twenty years the importance of the Sacher has grown enormously as it has systematically acquired the papers of, to date, about 100 composers, many of them major figures of Western music since 1900. For most of the composers represented, the Sacher is the principal or exclusive repository: Stravinsky, Webern, Varèse, Boulez, Ligeti, Wolpe, Birtwistle, Gubaidulina, Lutosławski . . . and, last on this list but certainly not least, Elliott Carter.

My own deep fascination with Carter’s music goes back to my adolescence, but it was more or less by accident that I became involved in the study of his sketches – an experience that proved to be my entrée to sketch study in general. In 1980, a junior faculty member at the time in the Yale University Department of Music, I had the good fortune to hear David Schiff give a guest lecture on Carter, in the course of which he mentioned that the bulk of Carter’s sketches and manuscripts were on deposit at the Music Division of the New York Public Library, where anyone with a good reason to look at them could do so. I found a good reason the following year, when I carried out the first of several projects to date involving Carter’s sketches. This, of course, was before the Sacher Stiftung purchased the materials being held in New York and most, if not all of the collection at the Library of Congress, bringing them together into one collection under ideal conditions for study – ideal, that is, except for Americans, who now must make a much longer trip to see them. The New York Public’s facilities, however, were also excellently arranged for scholars, and I spent a pleasant ten days or so there with the sketches for the Piano Concerto, one of the big works of the 1960s.

Very quickly, I came to appreciate one of the principal characteristics of these sketches, which is that there are a lot of them. Schiff has called this body of material “a paper mountain,” which is no exaggeration; in fact, the sketches for even one of Carter’s larger pieces would amount to a small hill. The sketches for the Piano Concerto come in at a relatively modest 360 pages, but there are 800 pages of sketches for the Second String Quartet, 1250 for the Duo for Violin and Piano, about 3300 for the Concerto for Orchestra. The sheer logistical challenges posed by such a staggering amount of material are obvious. Just to leaf through the sketch record of a typical Carter work, according each page no more than a passing glance, can consume an appreciable amount of time; and many a single page is of sufficient density and intricacy to keep one occupied, potentially, for hours. Worse, how does one retain what has been seen, keep it fresh enough in mind to remember it when running upon something else that might closely resemble a previously viewed sketch? The researcher’s problems are compounded by the fact that, while a good many sketches are carefully dated, a good many others are not, and these dated and undated varieties tend to be thoroughly intermixed. For some works, it may never be possible to reassemble the sketches in a reliably chronological order, assuming that this would be a useful task to carry out.

Owing to these circumstances, I doubt that study of Carter’s sketches can, for the most part, proceed along the comprehensively meticulous lines that have become familiar for such composers as Webern and Stravinsky. For if some sort of order is to emerge in Carter’s case from all these documents, no matter how provisional or tenuous, a great deal must simply be set aside. Such a strategy is bound to make the researcher uneasy – what if some crucial fragment gets by unnoticed? – and yet it is necessary, as Carter himself warned me in a letter soon after I first got in touch with him about the work I had been doing at the New York Public:

You must realize that all those pages of sketches are truly sketches, in the sense that some technical problem having to do with interval or rhythm or figuration is considered and solutions are worked over until certain artistic demands like “expression,” “character,” “emphasis” are satisfied in a way that seems to me relatively “fresh” sounding. Solutions that sound “dead,” “dry” (unless specifically needed at a certain moment), are discarded. Sometimes no “lifelike” solution can be found so the whole premise is discarded. All that takes paper and time . . .

So one must pick and choose; and although it may sometimes be possible to assemble a coherent narrative of the genesis of a work that way – as John Link, for one, has done so successfully for Night Fantasies, the towering solo piano work completed in 1980 – this won’t always be the case. As Schiff has noted in his “Paper Mountain” essay, “Carter’s sketches, despite their dauntingly copious state, are as fascinating for what they do not show as for what they reveal” – and Schiff goes on to mention “the absence of evolutionary steps in the final material. After the huge amount of abstract speculation the piece seems to find its nearly final shape with ease.”

Nevertheless, it is sometimes enlightening to focus closely on individual documents plucked, so to speak, out of the great stream: some of them representing brief moments in the compositional process, others evidently representing an amalgamation of decisions or an accumulation of isolated, sporadic discoveries scattered elsewhere through the sketch record. In the remainder of this essay, I will focus on a few sketches for two of Carter’s works from the 1980s and 1990s, giving some idea, I hope, of what can be learned when such documents are willing, as it were, to speak to us.


If it weren’t for the dates that appear on many of Carter’s sketches, one might be inclined to put them in an order reflecting the degree to which their form and content approach those of the finished piece. That is, it would seem plausible that the more closely a given sketch resembles a passage in the published score, the later it was written down – until, eventually, the very term sketch no longer seems particularly appropriate. But it would be a mistake to assume that every scrap of material for a piece conforms to some generally valid teleological pattern. Where dates are stamped on the sketches, they sometimes reveal that certain passages were invented fairly early in the compositional process, almost precisely in their final form, and then held in reserve until Carter could find a use for them. A case in point is a few measures from the Fourth String Quartet, composed in 1985–86, which as far as I can tell first appear in the sketch record for this piece in the version shown in Example 1a.

Example 1a
Elliott Carter, String Quartet No. 4, sketch page

Elliott Carter Collection, Paul Sacher Foundation

As the reader can see, the notation is already quite precise. But the date is also relatively early: 10 October 1985. More than five months later, the same passage resurfaces in recopied form (16 March 1986; see Example 1b), slightly extended and with some tiny adjustments to the rhythm of the second violin part and to the durations of some notes in the other parts.

Example 1b
Elliott Carter, String Quartet No. 4, sketch page

Elliott Carter Collection, Paul Sacher Foundation

One notices that by this point the cello has been put on a footing of the septuplet divisions that become essential to its rhythmic profile throughout the Fourth Quartet – divisions that, although they do not ultimately survive in this location, are nevertheless an important indicator of their growing overall importance to the cello part.

Further adjustments have been made to this passage by the time it turns up in the finished work. The passage corresponding to the sketches just discussed (bars 296–300) appears as Example 2.

Example 2
Elliott Carter, String Quartet No. 4, mm. 296-302

© Copyright 1986 by Hendon Music, Inc., a Boosey & Hawkes Company

Here, the second violin’s rhythm has been further fine-tuned, most noticeably in the third measure – but the biggest change is in the pitch domain: the second violin begins a major third higher than in the sketches, then in the second measure shifts to just a major second higher, a shift to which the other three parts, entering at various other points in that same measure, then conform. Such transposition is not in itself surprising, for Carter himself has allowed that this is an integral part of his working method – a fact, by the way, from which Andrew Mead, in his article on the role of octave equivalence in Carter’s music, has drawn important insights. Here is Carter in a 1971 conversation:

The actual notion of “absolute pitch” is not significant in my pieces. The pitches are chosen registrally as a matter of instrumental practicality. In fact, I frequently transpose parts of my pieces up and down, when I compose them, to try and decide in just which register they would sound most characteristic, given the instruments that are playing them. I make a decision on this basis and in consideration of the degree of expressivity of the given passage as compared with the preceding and succeeding passages.

Just what considerations of “expressivity” might operate in the ultimate context – those preceding and succeeding passages – to mandate the transpositions that Carter decided upon is an interesting but rather difficult puzzle to solve. What can be readily noticed is the way in which the integrity of certain details of the passage itself has been retained in its revision. Despite the shift from major-third to major-second transposition, not only is the repertoire of intervals characteristic of the second violin preserved (those are: unordered pitch intervals 3, 6, and 11, or minor third, tritone, and major seventh), but the tetrachordal set class bridging the transpositional shift has been preserved as well: that is, the original notes B, C, E♭, and G♭, constituting set class (0147), have become G, B♭, C♯, and D, also set class (0147), an inversion of the original collection and therefore, of course, equivalent according to the reigning harmonic principles of Carter’s music.


In my experience, it is relatively rare to encounter the same passage or fragment over and over again in Carter’s sketches, recognizably “on its way” to becoming an actual passage in the finished work. This is not, however, necessarily because Carter comes up with exactly or even close to exactly what he wants on the first attempt at invention from the repertoires of intervals and chords that usually form his starting point. Sometimes, one gathers, he does get lucky, as the record of the passage from the Fourth Quartet just discussed seems to show. But that 10 October 1985 sketch may well appear to spring forth fully formed only because its predecessors are so sketchy or nebulous that the resemblance is not detectable – or simply because the intermediary stages never made it into the written record. As Carter put it once in conversation: “One decision piles on another, and then I begin to see what in a chord [chosen for the basis of a composition] would contribute to [the desired] effect. The sketches don’t show this, because there’s an awful lot of mental work that’s never put on paper.”

Sometimes, however, the written record is a little more revealing about the successive stages that a passage goes through as it is brought to its final form. One pertinent example is provided by a thematic line from Allegro scorrevole, the third and final movement of the Symphonia. This line passes through at least eight drafts, one of which branches extensively into more alternatives, from its first appearance in the sketches on 26 November 1995 until the last version I have been able to find, on 1 February 1996, before its incorporation into the score at a point beginning two-thirds of the way through the movement. Example 3 is a compilation of these drafts, along with the corresponding extract from the score as completed by Carter. (Needless to say, perhaps, these drafts don’t appear so neatly reconciled in the original sketches; they’re on separate pages for the most part. Some of the sketches are more fragmentary, and peter out fairly quickly; others keep going, beyond the limits of this example.)

Example 3
Elliott Carter, Allegro scorrevole, drafts of melodic line

Two aspects are being worked on in the portions shown. One is fine-tuning of the opening gesture, whose rhythm and contour are already well defined even in the earliest sketch, up to the first high, held note, but whose exact pitch content remains not completely settled until the final sketch of 1 February. Scanning the series of opening gestures from top to bottom in Example 3, one may witness the process as it unfolds: first, the A♭ and A♮ are interchanged (the ♮ and ♭written above bar 159 and pickup substitute in this transcription for Carter’s revision of the sketch by means of overwriting the accidentals on the staff itself); then the next five notes, through B4, are straightened out; then the high point and final note of the gesture is changed, first from F♮ to D♯, then to F♯, then briefly back to D♯, ultimately settling on a form that incorporates both D♯ and F♯ into a brief divisi in the violin parts that have charge of this line at first. The version in the finished score evinces one further change, moving E4 up to E5 (bar 159) in a revision of the contour that had up to this point remained unaltered.

The other aspect worked on in these sketches is the continuation flowing from the opening gesture, over which Carter evidently expended a good deal of effort to achieve the expressive quality he wanted. The first attempt shortly becomes rather stiff and jerky – a quality that persists throughout the dozen or so bars not shown – leading Carter to write this note to himself for future reference: “more eloquent – much larger range.” The 20 January version doesn’t appear to have pleased him to any greater extent, for he does not keep it going beyond the first few bars. 22 January brings an even more abortive attempt. On 29 January, Carter got farther into it, broke off and began again with the B4 a little ways back to provide a slightly different continuation, which likewise breaks off a short while later with a reminder to himself to “extend.” This is carried out at great length in Version A of 31 January, but Carter is still not satisfied: “IMPROVE!” he admonishes himself in capital letters, with the further comments “more resistance” and “should be longer.” Version B on the same date is not yet longer but employs a branching style of sketching (indicated by the arrows) that diverts the flow to keep it going whenever blockage threatens. From this point in the process, the line gets steadily longer, with the already evident incremental effect of the changes made so far further extended: that is, measure by measure, working through the passage in temporal order, the notes steadily fall into place, until, as is evident from the last version of 1 February, this part of the line is in almost precisely the form it will assume in the finished score.

Comparison of the last two lines of Example 3 will show that the 1 February sketch presents almost all of the final version, except for a concluding measure (174) and some rhythmic and pitch revision in bar 166. Rhythmically speaking, it would seem that in this final tweaking Carter sought a slightly greater degree of plasticity, avoiding two triplets in a row (which he has already in the previous measure) but also avoiding the interruption of the characteristic triple divisions of this line by quadruple for two beats in a row. The pitch changes are more extensive: only five different pitches appear in this measure of the sketch, but while in the revision the emphasis upon D5 is retained (heard four times), four more pitches – D♭, F, G, and A – are added.

This too may have been done to promote plasticity, at least in part. However, there is surely more to be observed about the effects of these pitch changes, in light of what is known about Carter’s intricate and highly developed harmonic/contrapuntal language. As the previous discussion shows, it is possible to view Carter’s process of revision from a more or less traditional, “intuitive” standpoint and learn a good deal about how passages like this take shape – but one does not see/hear everything that way, because another sort of process is going on at the same time, one involving some of his favorite set classes (or “chords,” as Carter called them, referring to collections that may be presented either as simultaneities, or as melodic segments, or as some combination of the two): specifically, the all-triad hexachord (ATH) (012478) and the two all-interval tetrachords (AITs), (0146) and (0137). Example 4 is a pitch graph of some excerpts from Example 3, accompanied by analytical overlay.

Examples 4a-d
Elliott Carter, Allegro scorrevole, melodic line analysis

In the earliest of the sketches (Nov. 26, 4a), Carter seems to be working with all three set classes; there are quite a few instances of (012478) to be found here, and both (0146) and (0137) are abundantly represented as well. By the time of the first of the 31 January sketches (4b), the presence of the all-triad hexachord is much reduced, with a counterbalancing efflorescence of the all-interval tetrachordal classes; the second 1 February sketch (4c) has but one instance of (012478), apart from its “oblique” presentation in the opening gesture of the passage – a feature that is carried over into the final version (4d). An overview of the three sketches shown here – early, middle, late – shows a steady increase in the intensity of AIT interconnection and an intensifying emphasis upon these tetrachords by way of use of repeated groups of notes, culminating in the passage as finally scored. Although any further remarks about these features must remain firmly in the realm of speculation, it seems at least possible that Carter began work on this melody with the idea of using both the ATH and the AITs but soon decided that composing mostly with the smaller units (AITs) gave him more flexibility as he sought something “lifelike” and convincing.

The successive stages of revision for this passage show very clearly and abundantly Carter’s concern for the continuous, tensile qualities of the “long line,” that concept instilled in his developing technique during his studies with Nadia Boulanger in the 1930s and never abandoned from then on. This example, typically for the late style and the “late late style,” as Schiff has dubbed it for Carter’s music beginning in the mid-1990s, privileges the long line to an even greater extent than formerly, in that it is sketched first in the compositional process, in fact save for all but the finest of adjustments is apparently composed in its entirety before anything else is added.

For the student of Carter’s music, being in the presence of his sketch collections may make one feel like the proverbial kid in a candy store. But readers of a certain age will certainly remember that the trip to the candy store was never an unalloyed pleasure. So many treasures; such limited funds! One cannot after all have everything one’s heart might desire. When visiting the Sacher collections, it is the currency of time that one must take the greatest care not to squander – for the typical researcher’s resources in this commodity operate even more crucially as a constraint than do the monetary, and given the circumstances that prevail there, in which photocopying of primary materials is not allowed and even microfilms may be viewed only on the premises, forcing one to copy everything by hand, there is never a moment to waste. Working under such conditions can sometimes prove frustrating; on the other hand, no researcher will find it difficult after that first experience to imagine new projects requiring further visits, for this institute is so beautifully set up that its sheer existence seems practically miraculous. I suspect many of us who work on the sketches of twentieth-century composers – not just Carter – will never run out of motivation to return there.

Brief Bibliography Pertaining to the Study of Carter’s Sketches

Aylward, John. “Considering Elliott Carter’s Harmonic Process through Sketch Study.” Mitteilungen der Paul Sacher Stiftung 24 (2011): 42–45.

________. “Metric Synchronization and Long-Range Polyrhythms in Elliott Carter’s Fifth String Quartet.” Perspectives of New Music 47, no. 2 (2009): 88-99.

Bernard, Jonathan W. “Elliott Carter: Concerto for Orchestra (1965–69).” In Settling New Scores: Music Manuscripts from the Paul Sacher Foundation, ed. Felix Meyer. Mainz: Schott, 1998, 121–126.

________. “The Evolution of Elliott Carter’s Rhythmic Practice.” Perspectives of New Music 26, no. 2 (Summer 1988): 164–203.

________. “Premises and Applications of Spatial Analysis.” Musikometrika 2 [Quantitative Linguistics 43] (1990): 241–277.

________. “Spatial Sets in Recent Music of Elliott Carter.” Music Analysis 2 (1983): 5–34.

________. “The String Quartets of Elliott Carter.” In Intimate Voices: Aspects of Construction and Character in the Twentieth-Century String Quartet, ed. Evan Jones. Rochester, NY: Rochester University Press, 2009, II: 238–275.

Carter, Elliott. Collected Essays and Lectures, 1937–1995, ed. Jonathan W. Bernard. Rochester, NY: Rochester University Press, 1997. See, in particular: “The Orchestral Composer’s Point of View,” 235–250; and “Music and the Time Screen,” 262–280.

________. Harmony Book, ed. Nicholas Hopkins and John F. Link. New York: Carl Fischer, 2002.

Coulembier, Klaas. “Multi-Temporality: Analyzing Simultaneous Time-Layers in Selected Compositions by Elliott Carter and Claus-Steffen Mahnkopf.” Ph.D. dissertation, University of Leuven, 2013.

Eisenlohr, Henning. Komponieren als Entscheidungsprozess: Studien zur Problematik von Form und Gestalt, dargestellt am Beispiel von Elliott Carters “Trilogy for Oboe and Harp” (1992). Kassel: Gustav Bosse, 1999.

Elliott Carter: Sketches and Scores in Manuscript. New York: New York Public Library, 1973.

Emmery, Laura. “Evolution and Process in Elliott Carter’s String Quartets.” Ph.D. dissertation, University of California, Santa Barbara, 2014.

________. “Rhythmic Process in Elliott Carter’s Fourth String Quartet.” Mitteilungen der Paul Sacher Stiftung 26 (2013): 34–38.

Heinemann, Stephen. “Melodic Practice in Elliott Carter’s Clarinet Concerto.” Mitteilungen der Paul Sacher Stiftung 14 (2001): 19–22.

Link, John F. “The Composition of Carter’s Night Fantasies.” Sonus 14, no. 2 (Spring 1994): 67–89.

________. “Long-Range Polyrhythms in Elliott Carter’s Recent Music.” Ph.D. dissertation, City University of New York, 1994.

Meyer, Felix. “Left by the Wayside: Elliott Carter’s Unfinished Sonatina for Oboe and Harpsichord.” In Elliott Carter Studies, ed. Marguerite Boland and John Link. Cambridge: Cambridge University Press, 2012, 217–235.

Rao, Nancy. “Allegro scorrevole in Carter’s First String Quartet: Crawford and the Ultramodern Inheritance.” Music Theory Spectrum 36 (2014): 181–202.

Rosen, Charles. The Musical Languages of Elliott Carter. Washington, DC: Library of Congress, Music Division, 1984. See, in particular, the title essay, 1–19.

Schiff, David. The Music of Elliott Carter, second edition. Ithaca, NY: Cornell University Press, 1998.

________. “A Paper Mountain: Elliott Carter’s Sketches.” In Settling New Scores, 115–118.

Schmidt, Dörte. “Das ‘bemerkenswerte’ Interesse an Alois Hába: Anmerkung zu Elliott Carters ‘Harmony Book.’” Mitteilungen der Paul Sacher Stiftung 6 (1993): 38–42.

Shreffler, Anne. “‘Give the Music Room’: Elliott Carters ‘View of the Capitol from the Library of Congress’ aus A Mirror on Which to Dwell.” In Quellenstudien II: Zwölf Komponisten des 20. Jahrhunderts, ed. Felix Meyer. Winterthur: Amadeus, 1993, 255–283.

Soderberg, Stephen. “At the Edge of Creation: Elliott Carter’s Sketches in the Library of Congress.” In Elliott Carter Studies, 236–249.

Vermaelen, Denis. “Elliott Carter’s Sketches: Spiritual Exercises and Craftsmanship.” In A Handbook to Twentieth-Century Musical Sketches, ed. Patricia Hall and Friedemann Sallis. Cambridge: Cambridge University Press, 2004, 161–175.

________. “Fonction de la polyrythmie dans la pensée musicale d’Elliott Carter.” Mitteilungen der Paul Sacher Stiftung 12 (1999): 32–35.