Sketch study may have originated in the mid-nineteenth century but interest in this field of enquiry burgeoned in the twentieth, when compositional process became regarded as something of an art form in itself, and, more importantly, when opportunities of access to musical sketches increased significantly. Key events that contributed to this rise in activity include Hans and Rosaleen Moldenhauer's acquisition of Webern's manuscripts: these were made available to researchers in 1953 and partly reproduced in facsimile editions and other books, such as the Moldenhauers' seminal Webern biography published in 1979. Another important milestone was the opening of the Arnold Schoenberg Institute at the University of Southern California in 1973. This active centre, which published its own journal and hosted regular conferences, provided the opportunity for scores of scholars to study Schoenberg's sketches; the archive eventually moved in 1997 to the Arnold Schönberg Center, Vienna. A further decisive moment occurred when the Swiss conductor, impresario, and private collector Paul Sacher opened his eponymous Foundation in Basle in 1986. Sacher had previously purchased the Moldenhauers' Webern collection and Stravinsky's manuscripts, the latter for a reported US$5.25 million. There are now over ninety collections in the Sacher archive, including manuscripts relating to Bartók, Henze, Varèse, Lutosławski, Boulez, Carter, Kagel, Kurtág, Ligeti, Berio, Birtwistle, Ferneyhough, Nancarrow, Feldman, Andriessen, Rihm, and, the most recent addition, Reich. Many of these collections are, of course, ‘live’, in the sense that the composers continue to send new material as it is composed. Since 1990 over fifteen other archives have opened elsewhere in the world, while public libraries continue to acquire new material, as in the case of the British Library's recent acquisition of Peter Maxwell Davies's manuscripts.

Analysis of twentieth-century music often benefits from knowledge of the composer's working methods: serial permutations are laid bare, or insights are afforded into compositional systems that are unique and therefore not explicable by existing theories. New narratives may emerge that differ from those mythologized by analysts or the composers themselves. Moreover, as sketches gained greater autonomy from the score, particularly in the second half of the century (with the increased use of charts, tables, and other funds of material), the equation of the ‘work’ with the score became increasingly untenable. Rather, it might be argued, the work is revealed in the dialectical relationship between the score, or its performance, and the composer's more abstract, pre-compositional concepts, as revealed by the sketches. In other words, twentieth-century musical sketches offer the analyst a tool with which to prise open the work-concept.

Given the growth of interest in musical sketches in recent decades and the lack of guidance available in this area at university level, the editors of A Handbook to Twentieth-Century Musical Sketches, Patricia Hall and Friedemann Sallis, rightly identify the urgent need to plug what they refer to as a ‘methodological black hole’ (1). Theirs is a wide-ranging, handsomely presented, pedagogical survey that is intended to explain ‘how scholars and students should work with and think about the composer's working manuscripts’ (dust-jacket, front inside).

This book enters a distinguished but contested field of study that effectively began with Gustav Nottebohm's monumental documentation of Beethoven's manuscripts in the 1860s and 70s, alongside work by Alexander Wheelock Thayer and Ludwig Nohl (though reproductions of the sketches had been published as early as the 1830s and 40s). In Nottebohm's time sketch study focused on issues such as dating and chronology, or what is often referred to as the ‘biography’ of the work. The crucial point that emerged from this work was an understanding that compositional processes are invariably messy: they rarely proceed in a predictable fashion from the earliest idea to musical coherence. In other words, the focus in Beethoven sketch study was by no means wholly, or even primarily, analytical. The 1920s and 30s, however, witnessed a new wave of scholars, which included Schenker, August Halm, Ernst Kurth and Paul Mies, all of whom asserted a link between the value and essence of a composition and the demonstrable integrity of its genesis. In the 1960s and 70s, a third wave of interest in Beethoven's sketching practices developed, led by Alan Tyson, Douglas Johnson, and Joseph Kerman. By contrast, sketch study also became associated at this time with the rise of formalist analysis, particularly through Allen Forte's The Harmonic Organization of the Rite of Spring, published in 1978, though in the same year Johnson expressed a very honest scepticism about how sketch study would inflect any reading by the type of analyst for whom only the integrity of the finished work matters.Footnote ¹ Johnson's observation and the robust responses to his position (by Sieghard Brandenburg and William Drabkin) illustrate that sketch study can bolster either a positivist or a critical approach. In other words, it is not the methodology that is positivist, as some commentators have argued,Footnote ² but the way the fruits of that methodology (the outcomes of sketch study) are interpreted.

This problem was either overlooked or ignored by Kerman when he suggested that sketch study should contribute towards his notion of ‘music criticism’.Footnote ³ Indeed, more recently it has been argued that sketch study ‘can create a powerful site for the kind of humanistic, interpretative criticism which is at the heart of Kerman's agenda.’Footnote ⁴ Hall and Sallis, however, call for old debates to be put to rest: ‘The question is not whether sketch material will be used for analysis, but rather how this should be done. Sketch studies have become and will no doubt remain a permanent fixture of both the history and analysis of music for the foreseeable future and it is within this perspective that the book was conceived’ (4). The way forward, they argue, is to adopt ‘a broadly based, holistic approach’ (3). These points are well made, but there are three potential problems that go unacknowledged here. Firstly, a significant amount of the knowledge and expertise on twentieth-century musical sketches exists on the European continent, where the debates referred to above have not necessarily been a central concern.Footnote ⁵ This seems potentially significant for a book in which the majority of its authors (eight out of thirteen) are based in Europe. Secondly, it is not precisely clear what ‘a holistic approach’ should, or could, amount to. If this means moving beyond analysis or basic description, such ambitions are potentially compromised by the content of the sketches, which tend to draw attention to matters of structure. Thirdly, the editors sidestep the potential for conflict between analysis and sketch study. Although the sketches may inform or even support analysis, a tension may well arise between one's fidelity to the sketch on the one hand and an analytical approach that is consistent purely within the terms of its own internal logic on the other.Footnote ⁶

The editors' rationale for their choice of content and the ordering of their chapters is cloudy. We are told in the introduction that the book divides roughly into two sections, the first, chapters 2–6, on ‘the knowledge and skills necessary to work efficiently in an archive’, written mostly by those who work for archives, and the second, chapters 7–14, on ‘issues and techniques pertinent to the study of sketch material’ (2). These definitions do not differ in any substantive way and the sense of a bipartite design is further undermined by the fact that all the chapters focus on specific case studies and methodological issues of one kind or another. From chapter 7 onwards the composer case studies appear to be grouped for the most part in pairs: Webern and Berg, Bartók and Stravinsky, Boulez/Stockhausen and Carter, Ferneyhough and Cage – the last, inevitably the most open-ended, is concerned with electronic media (computers and magnetic tape). The more obviously methodological chapters are 2–3 and 5–9. These discuss archival etiquette, handling manuscripts, classification systems, digital preservation, transcription, reconstruction, and dating, respectively. Chapter 4, ‘Coming to Terms with the Composer's Working Manuscripts’, by Sallis, considers the conventional terminology used to describe and classify manuscripts in relation to intriguing pages by Kurtág and Ligeti. However, this chapter also reintroduces more general concepts already discussed in the rather wayward first chapter, ‘Sketches and Sketching’, by Giselher Schubert and Sallis. In both instances the arguments are by turns enlightening and vague. For example, Sallis's metaphor of sketches as the ‘vast underside of creative activity out of which individual works emerge rather like the tips of icebergs’ (56) effectively highlights the potential existence of an intertextual network that may relate different compositions (this has particular relevance to Stravinsky and Kurtág). However, it is not absolutely clear why Umberto Eco's concept of ‘the open work’ is considered ‘necessary knowledge’ before visiting an archive. A clearer sense of purpose here, and elsewhere, would have been instructive. Also, the lack of a detailed chapter on text-based documents, such as correspondence, is surprising.

A real disappointment, however, is the book's failure to summarize key achievements in sketch study: no attempt is made to provide a critical overview of ‘where are we now’ in relation to such important composers as Stravinsky or Schoenberg, and, currently, no such survey exists. An opportunity has been missed to present a more exhaustive bibliography, one that would convey the breadth of the field (there is little evidence in the bibliography of the immense body of Stravinsky sketch study, which includes some of the most accomplished examples in the field, and reference to work on Schoenberg is limited).Footnote ⁷ Conversely, the individual chapters often fail to engage with existing sketch studies or suggest ‘key texts’.

Despite these concerns, the book remains an invaluable resource for anyone coming to sketches for the first time. The contributing authors strike a fair balance between practical guidance and specific research concerns, and there is a useful appendix that lists the addresses of thirty-eight archives and institutes that house twentieth-century music manuscripts.

Chapter 2, by Ulrich Mosch, offers sound advice to a scholar planning a first trip to an archive. A musicologist at the Paul Sacher Foundation, Mosch runs through some interpretational issues, patiently examining drafts and sketches for Berio's Requies (1983–5). In the process he highlights such problems as determining the chronology of sources and the order of events on a page, the mismatch between what is observed and a reader's preconceived ideas, and deciphering text and marginalia. (Here Mosch's suggestion that ‘we can always seek the advice of native speakers of Italian’ (27) is infinitely more encouraging than the editors' schoolmasterly injunction that ‘[f]oreign-language competence is essential’ (3)). In order to present his discussion, however, Mosch asks us to ‘assume that we have received a microfilm or photocopies of all the musical manuscripts’ (23) in advance of visiting the archive, when in fact, as he freely acknowledges, institutes rarely send copies of sketches, usually owing to composer rights (he does, though, offer sound advice on finding published sketch pages).

Thérèse Muxeneder, who is based at the Arnold Schönberg Center, provides another ‘insider’ perspective. Muxeneder discusses archival etiquette, but her principal concern is the preservation of materials since, owing to mass production processes, many nineteenth- and twentieth-century manuscripts are more endangered than those from earlier centuries, while magnetic tape may deteriorate even more rapidly. In the course of her essay, Muxeneder remarks on the possibility of making manuscripts available via the Internet. Although not explored here, the ramifications of this are surely worth considering. Does widening access necessarily benefit the discipline? It removes the relative comfort zone afforded to scholars whose findings are not easily verified and therefore often taken on trust. But time spent with originals, learning the habits of a composer, is surely invaluable, and, as Regina Busch discusses later in the book, a copy, however technologically advanced, will never reproduce an original entirely faithfully.

Erika Schaller's consideration of cataloguing at the Archivio Luigi Nono in Venice provides a fascinating window onto Nono's working practices and the tremendous diversity of materials pertaining to his music stored there, though a companion chapter on interpreting this material would have been useful. Determining a precise chronology of Nono's sketches is impossible, as he rarely dated his materials, but Schaller notes that visitors can, for future reference, leave their own comments, as well as ‘tentative’ transcriptions of illegible manuscript passages. To this reader this seems like an invaluable practice, too rarely adopted elsewhere. Certainly archivists should not be expected to interpret sketches, but archives need not be neutral repositories. With inventories often running to thousands of pages, it would surely save new researchers time if previous visitors were able to single out important pages.

Turning to the subject of sketch transcription, with specific reference to Webern's Five Canons on Latin Texts, Op. 16, Regina Busch emphasizes that any copy of a sketch is already a transcription, and that a transcription ‘always presents something of the adapter's ideas’ (87). (Evidently sensitive to such matters, she inserts a footnote distancing her own original German text from the edited translation used in the book.) Busch's chapter exposes the vast space between the ambiguities of a handwritten sketch and a handwritten copy or typeset example which, in many instances, simply cannot convey the nuances of the original, such as the distinction between harder and softer lines, or signs of hesitation. Busch, however, does not provide an example of ‘good practice’ in notation transcription herself, but restricts her examples to the transcription of verbal text. She also provides facsimile reproductions, but these plates (7.1–7.3) are faint and harder to read than those in other chapters.

On the reconstruction of sketchbooks, Patricia Hall's chapter on Berg's sketchbooks for Wozzeck is a model of pedagogical clarity. After referring directly to other scholars who have worked on Berg's manuscripts, Hall not only invites us to consider the size and colour of the pages in Berg's sketchbooks (here helpfully referring to Dumont's Farbenatlas as an invaluable source for describing the paper's precise shade) and their pagination and foliation (again with clear diagrams and advice on treating the manuscripts with care), but she even refers to their smell, differentiating between pre- and post-First World War paper, the latter reminding her of ‘old, cheap paperback books’ (103) (inadvertently, however, such observations highlight the fetish-like quality of sketch study, which some find disturbingFootnote ⁸). Advice is offered on deciphering verbal annotations, a problem that Hall candidly describes as one of the ‘curses of studying sketchbooks’ (108); Berg's illegibility often arose from his composing while travelling in cars and trains or walking. Here, once more, the collaborative nature of studying sketches is revealed in Hall's advice to work with other scholars faced with similar deciphering issues, as well as the need to exercise an analytical mind when determining the meaning or significance of particular passages of text.

Where the previous two chapters considered issues of identification and transcription in sketches by two composers of the Second Viennese School, the following two chapters examine fragmented compositional processes in sketches by Bartók and Stravinsky. László Somfai's chapter ‘Dating Bartók's Sketches’, which considers the First Violin Sonata, confounds its own logic. ‘Here we face the typical danger of sketch studies’, he writes. ‘One can place the first written notes under the magnifying glass and interpret them, but one cannot even guess what took place in the composer's head’ (122). In the literal sense this may be true, but all sketch study requires an attempt to think beyond the sketch into the composer's mind. Not even Somfai can ignore this necessity, and he chooses to go one step further by relating the sketches to what he terms the ‘hypothetical “concept” or “narrative”’ (128) behind the work. To do this he must determine the chronology of the thematic materials for the Sonata. Fragments of the three-movement piece are scattered throughout a sketchbook in such a way as to suggest that Bartók assembled the basic elements of sonata form in a kind of patchwork assemblage, developing and adapting the themes out of sequence. Somfai's task requires, among other things, familiarity with the strokes of a fountain pen (thinner, scratchy strokes indicate a newly-filled pen, thicker lines imply more or less continuous use), and knowledge of the composer's travels and letters that document early stages of the work's evolution.

Somfai concludes that the Sonata's unwritten concept consists of the opposition between the opening ‘art-music-style’ material and the finale's ‘invented pseudo-peasant fiddler dance theme’ (127), since these were both sketched, in thin, dried-out strokes, at a crucial early stage. The author's extensive knowledge of Bartók's sketches is amply illustrated in this chapter, and his observation that ‘sketches often point to the essence of a piece […] in a clearer way than the finished score’ (128) is well made, though the journey of discovery is a relatively complex one that involves numerous ‘digressions’, as Somfai himself refers to them.

Should we be tempted to over-emphasize the role of pitch, Mäkelä's chapter, ‘Defining Compositional Process’, in line with other recent work on Stravinsky sketches, reminds us that rhythm and sound-colour ‘also constitute essential compositional parameters’ (131). Mäkelä considers ‘idea and instrumentation’ in Stravinsky's Ragtime (1918) and Pribaoutki (1915). A number of surprises emerge here. For example, the sketches suggest certain hierarchies of idea that cannot be read from the score. Such important textural ideas, which often appear in short score, may have modest, isolated origins. The sketches also reveal that instrumental sonorities were altered through the creative process and only finalized after pitch and rhythm, even when the final choice had a decisive impact on the overall sound quality. Mäkelä's chapter follows on nicely from Somfai's, since they are both concerned with imprints and elaborations of generic ideas that were not conceived in the order in which they appear in the final score, and both composers mark the beginning of a break with more context-dependent tonal practices. With Stravinsky, however, additional complications include an over-abundance of ideas, or ‘building-blocks’ (145), some of which appear in different pieces or are not used at all, and the fact that there is ‘no correlation between the certainty with which a sketch is written down […] and its inclusion in the published work’ (142). Although this information is fascinating, we are not always told why. For example, a table (Fig. 10.1, 143–4) reveals the apparently random order in which parts of Ragtime were composed, but we are not told how this is significant for an appreciation of the piece, and although this fragmented perspective may contradict a listener's experience, or an analysis, it is surely not always possible to prove conclusively at what stage Stravinsky knew, or did not know, where his various fragments would end up. The sketches here seem to raise more questions than they answer.

Questions are, however, answered by Pascal Decroupet's chapter on the 1950s sketching practices of Boulez and Stockhausen, for here it is possible to see precisely how and where predetermined systems permit degrees of latitude, and examples are given of how such imaginative freedom is realized. This draws attention to what Ross Feller, elsewhere in the book, valuably refers to as ‘the human agent at the threshold between the generation and disintegration of systems’ (188). From my own perspective, I found the similarities to procedures adopted by Harrison Birtwistle, such as the use of filtering and rotation, a salutary reminder of the impact serial thinking has had on recent non-serial, but serially conditioned, contexts. Although Decroupet's chapter, more than any other in this book, pursues the positivist bias of twentieth-century sketch materials to their ultimate and not infrequently complex ends, there are good reasons for doing so, not least to demonstrate that it wasn't all a fraud: serialism really was taken on to a new level in the 1950s. Nor should an appreciation of such procedures necessarily be seen as subscribing to a formalist aesthetic; these sketches should take their place in a wider discourse for they have a significant narrative to tell. Yet Decroupet's chapter, for all its impressive grasp of the details of a wide range of works by Boulez and Stockhausen, privileges the composers' voices to such an extent that the reading amounts to an authorized interpretation. It is left to the reader to decide why any of this information is important. Despite – or perhaps because of – the music's rigour, this approach serves above all to reinforce the need to consider a work's genesis from a variety of critical vantage-points.

Seemingly strict formalist concerns are cast in a rather different light by Denis Vermaelen's chapter. Vermaelen considers Elliott Carter's scepticism – one shared more generally – towards the usefulness of his own sketches and the ability of sketch study and analysis to ‘explain why it is that the work, when heard, captures our attention, and what is so valuable about it musically’ (162).Footnote ⁹ When composing ‘Anaphora’, from A Mirror on Which to Dwell, Carter began with abstract limbering up exercises that included the manipulation of hexachords indexed in the ‘Harmony Book’ he assembled for himself in the 1960s.Footnote ¹⁰ Vermaelen proceeds to illustrate how such raw materials are transformed by ‘virtuoso’ compositional techniques – at one moment, ‘twenty-seven aggregates follow each other in the space of four seconds’ (175) – into music that precisely ‘captures our attention’. Vermaelen's chapter is persuasive and demonstrates effectively how Decroupet's approach may be extended towards the level of an aesthetic appreciation of the music, but without sacrificing methodological rigour or detail.

Our ability to trace such transformational procedures in more recent music, it seems, will be severely hampered when, as in the case of Brian Ferneyhough, computers are used as a tool to aid the compositional process. Not only will scholars need to understand the programs being used, but alterations will be erased by the composer's act of clicking ‘save as’. The fact that ‘e-sketches’ may become corrupted, or the necessary hardware become outdated, suggests that, at some point, electronically stored data, such as Ferneyhough's work in Finale and PatchWork, may even be lost altogether. Ross Feller's privileged insight into Ferneyhough's e-sketches, gained from extensive correspondence and access to the composer's extant computer files, is valuable and instructive. Computer-assisted composition speeds up certain aspects of Ferneyhough's work and presents him with more solutions from which to choose. However, were he to arrive at his ultimate ideal of a virtual, disembodied form of composition, whereby all the dimensions of a piece were calculated at the push of a button, his claim that computers bring him closer to creative spontaneity, and his trademark dialectical engagement ‘between formal and informal processes’ (184), would be lost.

Our notion of what sketches are and how they may be used undergoes a further transformation in the final chapter. Here the surviving materials – tapes, commentaries, sketches, and a 192-page unpublished score – relating to John Cage's Williams Mix (1951–3) are considered. Access to the source materials of ‘the first octophonic tape piece [using eight loudspeakers] in the history of electronic music’ (213), inspired Larry Austin, the author of this chapter, to collect nearly 600 new sounds and to design a computer programme that reruns Cage's compositional process. This resulted in a three-year project to produce a new work, Williams [re]Mix[ed], that lasts just four minutes.

At one moment in his exhaustive account, Austin performs an unwitting coup de main when he refers to his use of the sound of Cage's pencil as he writes and very quietly talks to himself while demonstrating his coin-tossing process. This musical sketch made audible is a reversal of the very notion – the inaudible sketch – that, perhaps rightly, troubles those who do not see the relevance of sketch study. But this moment also sounds a wake-up call. To write engagingly about sketches is not easy, and a pedagogical book is bound to come across as overly descriptive at times. Important observations are lost, however, if accounts of compositional process are too exhaustive. While it is true that those who are skilled in sketch study have an invaluable contribution to make at the initial ‘descriptive’ stage, the discipline has arguably been less successful when it comes to extrapolating from the raw data. True, there is a need to develop and refine our knowledge of twentieth-century creative processes, even for such high-profile composers as Stravinsky. But there is also scope to generalize about types of sketching practice. Is the term ‘elaboration’ pertinent to sketching practices in the second half of the century? Are ‘extensional’ modes of sketching, where attention is placed on the large-scale addition of basic units or themes, replaced by ‘intensional’ modes that focus on small-scale inflections of basic ideas or patterns (a building ‘inwards’ rather than ‘outwards’)? Or, as Vermaelen illustrates in relation to Carter, is there a dialectical ‘alliance of opposites’ (169) between these modes? Too often sketch studies are narrowly focused, as if examining and making sense of manuscripts were troublesome enough, without perceiving a need to search for further meaning, to couch findings in clearer terms, or contribute to, or even consider, ongoing academic debate. While this observation should not be levelled at the more pedagogic chapters here, it does apply to some of those with more critical ambition.

If this handbook is some kind of prescription for the future of sketch study then the horizons of that future seem limited. Firstly, there is very little encouragement to develop a reflexive approach that is honest about the ideology of sketch study – an understanding of the potential circularity whereby the reasons why we look at sketches and the things we look for in them come to determine what we think they tell us. Secondly, although the methodological tools used in sketch study are unquestionably of practical importance, we need to open sketch study out into (and demonstrate its relevance to) the whole host of questions explored within the humanities. While scholars press ahead with work on cultural theory, cultural transfer through different arts and media, linguistic and social theory, race, ethnicity, and gender, the evidence of this book is that sketch studies pay no heed. But this is not necessarily so.Footnote ¹¹ Let us not forget that it was to musical sketches that Richard Taruskin turned in order to move beyond what he perceived to be sterile analytical accounts of Stravinsky's music, which led to his interpretation of Stravinsky's music through the concept of ‘neonationalism’.Footnote ¹² In this respect an opportunity has been missed to argue more strongly the case that sketch study can have far-reaching implications for all aspects of musicology. Scholars who spend years ‘living’ with a composer's sketches – though by no means all those in this book – are apt to emphasize the ‘correct’ way to interpret them, but this is to confuse knowledge and descriptive interpretation with use and extrapolation.