Introduction: Some Preliminary Editorial Considerations

Helmut Lachenmann's Salut für Caudwell (1977) for two guitarists (hereafter Salut) was first published in 1985 by Breitkopf & Härtel and since then has been widely performed and recorded several times. A new edition was published by Breitkopf & Härtel in 2020, the work of Seth Josel, a contemporary guitarist-researcher whose other projects include a large database of contemporary guitar music,Footnote ¹ The Techniques of Guitar Playing,Footnote ² as well as an extensive and critically acclaimed performance practice. The publication is part of a research project hosted at the Orpheus Institute,Footnote ³ an Historically Informed Performance investigation carried out by Josel and Tom Pauwels for HIPEX (Historical Performance Practices of Experimental Music), an ongoing research cluster at this institution. The outputs of this research include not only the new published edition of the score, but also extensive archival work, including an article in TEMPO in 2019, based on the recompilation and study of various sources: sketches, drafts, various score iterations, as well as interviews with the composer and the work's dedicatees, the Cologne duo of Wilhelm Bruck and Theodor Ross. Josel has also published a comparative analysis of the discrepancies between different authors’ understanding of the formal segmentation of the work,Footnote ⁴ which informs part of this study.

Josel places the beginning of Salut's chronological timeline in 1976 with its commission by Bruck and Ross. In 1977, the autograph was finished and the piece was premiered in December at SWR; between 1978 and 1984 there were many performances by Bruck/Ross, Lachenmann wrote an essay on the work and a manuscript edition was published. Preparations for the print edition culminated in its publication on 9 April 1985 by Breitkopf & Härtel. The premiere recording by Bruck/Ross was released in 1986 by Col Legno (LP version in 1986; CD version in 1988) and a second recording was released by Kairos on 31 January 2008. Around this main timeline are other performances by both German and non-German-speaking guitar duos, commercial recordings and several live performances available on Youtube and Vimeo.

Salut's dedicatees, Bruck and Ross, have been highly influential not only in the conception and dissemination of the work (with a significant co-creative agency/authority/responsibility explicitly acknowledged by the composer, especially to Bruck) but also in the afterlife of a score which has undergone several revisions after its first performance during many hours of rehearsal supervised by the composer. The need for a new edition became evident for Josel after extensive study of the work's source materials within the HIPEX research cluster, as ‘Lachenmann made diverse alterations to the score over time, reflecting his experiences with Bruck and Ross in rehearsals and as an audience member in their concerts’.Footnote ⁵ These sources were ‘four different versions of the score: the manuscript edition, the annotated version of the autograph, the print edition, and the copy of the print edition (annotated), as well as Bruck and Ross's parts’.Footnote ⁶ In addition, Josel's rigorous archaeological work includes oral and performance histories: ‘interviews with Lachenmann, discussions with the first performers, Wilhelm Bruck and Theodor Ross, and, importantly, the performance materials used by that duo’.Footnote ⁷

Unlike Pression, whose two editions were both made by the composer and have different notational approaches, sections and bar numbers, both of the Salut editions present the same 533 bars. Among the most significant changes and features in the new edition – justifying a new edition rather than just an errata list – are: (1) the restoration of the traditional five-line staff for the right hand that Lachenmann originally used; (2) the (re)casting of the spoken section, discarding the use of the International Phonetic Alphabet (IPA) and restoring the text as it was originally written in Lachenmann's autograph (including his striking use of capitalisations of plosives and consonants – a most effective and user-friendly visual stimulus for performers); (3) a supplemental ‘Étude’, written by Josel and consisting of the rhythmic structure of the spoken section, together with the previously included IPA notation. This not only serves as a study aid but also facilitates performance by non-German speakers and provides a significantly improved translation of the text.

These improvements to the spoken section in the 2020 edition are indeed significant for the expressive content of Salut. This section crucially includes a quotation from Illusion and Reality: A Study of the Sources of Poetry (1937), a book by the young English Marxist Christopher Caudwell who died during the Spanish Civil War, and in his 1979 essay ‘Struktur und Musikantik’ Lachenmann described how he dislocated individual phonemes so completely that one perceives not words but the sound of each phoneme:

The 2020 edition also includes a discussion of stage positioning, offering two options. One is inspired by Bruck and Ross's performance practice of sitting as far apart as each performance space would allow, to enhance textures implying a stereophonic image, as well as several hoketus-like passages, by projecting two clearly separated sound sources. The other option resembles a more traditional chamber music arrangement, with the guitarists sitting closely together, suggesting the idea of a ‘meta-instrument’ with its own set of interesting aesthetic implications. Both options have the composer's approval.

As for tempo indications, the 2020 edition introduces some corrections and slight modifications. The 1985 edition omits the tempo change in bar 28 that indicates the dotted quaver from the previous subsection (crotchet at 80 BPM) should become a crotchet at 106.9 BPM, continuing until bar 186. This is a significant correction because a literal rendition of the 1985 edition from bar 28 to 186 would have maintained the crotchet at 80 BPM, a rather slow tempo for the remainder of the section, which includes the spoken passage with the Caudwell quotation. On the other hand, it is worth noting that, since Bruck and Ross's first recording, performers have not followed the 1985 edition literally, suggesting that, as is often the case, the dedicatees’ first recording has been taken as an authoritative text. Both editions indicate the tempo change from a crotchet plus a semiquaver into the new crotchet in bar 186, although the 2020 edition specifies that the new crotchet is at 85.3 BPM, evidencing the use of calculator.

This clarification has an impact on the tempo change at bar 208 where, in the 2020 edition, the crotchet + semiquaver turns into a new crotchet at 68.266 BPM; in the 1985 edition the new tempo at bar 208 is given as a quaver at 126 BPM (crotchet at 63 BPM). This modification is evidence of Josel's rigorous approach to editing the 2020 version. As the 2020 edition is not a completely annotated critical edition the source of this modification is not clear, however, although it is close to Bruck and Ross's performing practice as documented in their commercial recordings. It would perhaps have been useful to clarify whether this was the composer's intention, was introduced by Bruck/Ross or was simply a matter of neglect in the 1985 edition. As a performer myself, this would clarify to what extent the use of an exact calculation of BPM beats is part of the composer's intentions (Werktreue) or the notation itself (Texttreue), and would thus clarify the limits of the performers’ freedom.

The new edition presents other modifications. Several new dynamic indications are included, resolving discrepancies across the multiple sources; these need not be itemised here but involve either new dynamic indications or minor displacements of already existing ones. The same is true of some performance-action directions: there are new indications or clarifications as well as minor displacements of the exact point of execution, such as the actions of the (fore)arms dampening or releasing the strings, or modifications of exact features of attacks, such as the ‘volle Handfläche zur Griffbrettmitte’ in bar 488 of the final section, which subtly impact timbral qualities.

Lachenmann entrusted Josel fully with the creation of the new edition but declined his offer of a consultative role for specific modifications. This might be interpreted as the composer remaining ‘loyal’ to Salut's dedicatees and one is reminded of Lachenmann's statement that

This new edition constitutes a fundamental new epistrata (a Deleuzian term to which I will return later) that opens Salut to wider dissemination, especially in its re-casting of the spoken section, which makes it more accessible to non-German speakers. Together with the other research outputs of this project, including two articles and an interviewFootnote ⁹ and the video recording of the piece by Seth Josel and Tom Pauwels, these outputs also open the way for new performances and understandings, both for Salut and for other music or notational approaches close to it. Crucially, Josel's efforts clarify the ‘what it was’ of Salut and open up possibilities for its ‘what it could be’. I will now use the lens provided by recent developments in artistic research to examine some selected recordings of Salut – the ‘what it was’ – so that later I can turn to the ‘what it could be’.

Salut as Structure: A Brief Morphosyntactic Analysis

Although musicology devoted to the analysis of recordings is not as abundant in the realm of new music performance studies as it is in common-practice period music, a convincing analytical model can be found in Christian Utz's ‘morphosyntactic’Footnote ¹⁰ and ‘performance-sensitive approach to musical analysis’, which seeks to relate the spatial and the temporal aspects of music and aims

Utz's analytical focus is on the time-experience of music as performance and temporality is chosen as a principal field of inquiry because ‘the experience of musical time is often conceived to lie at the “heart of performance”,Footnote ¹¹ while appearing to be sidelined by structuralist analysis’.Footnote ¹² Utz is also influenced by Nicholas Cook's concept of music as performance.Footnote ¹³ Cook distinguishes two main performing styles in relation to musical time: the structuralist and the rhetorical.Footnote ¹⁴ The former relates to an architectural/synchronic view of musical form and is usually associated with literalist/modernist performance and the concept of reproduction. The rhetorical model is, in contrast, pre-modernist, characterised by treating musical materials as if they were topics, so that performance becomes a kind of semiotics practised in real time,Footnote ¹⁵ crucially resulting in a more flexible treatment of musical time as performers do not reproduce but create musical structure through performance. Respectively, the structuralist approach regards music as an ideal object, not inherently temporal but presented through time in performance – that is, music in time – while the rhetorical, by contrast, regards music as time. Utz's analyses are devoted to contemporary solo cello works, including Lachenmann's Pression, making his model highly pertinent to Salut, and they provide ample evidence for Cook's claim that ‘rhetorical and structuralist approaches represent complementary possibilities for construing music as thought and action’Footnote ¹⁶ rather than irreconcilable opposites which establish a linear historical causality.

Following the example of Josel's research on Salut's form, in which he undertakes a revisionist comparison of the discrepancies between several analyses of the score (including that of the composer himself), I will use the segmentation of Salut that Josel found in Lachenmann's own formal plan for the work in a sketchbook archived at the Paul Sacher Foundation. Josel observes that Lachenmann's original plan was to compose Salut in 11 sections. The sketchbook itself includes 12 sections, the third of which became the spoken section, but I will label the second and third sections as 2a and 2b, following the composer's original 11-part plan. I will not attempt a detailed consideration of differences in perception about where new sections or subsections begin or end, not only because Josel's article already discussed this in great detail, but also because formal segmentations are never truly vertical: that is, we could speak of a certain diagonality, in the sense that each section (or subsection) anticipates either gestural or metric modulations in the next section. Salut is an example of Lachenmann's practice as a ‘master of even the smaller-than-smallest transitions’,Footnote ¹⁷ as Adorno wrote of Alban Berg. Thus I adopt the segmentation found in the composer's sketchbook.

My purpose here is to compare notated and performed sound, measured in time, as my analytical starting point. Accordingly, Table 1 shows the temporal measurement of notated sound according to the 2020 edition; as I mentioned earlier this is more rigorous in terms of tempo indications and closer to Bruck/Ross performing practices. Brief fermatas have been included in this analysis, although one big obstacle in measuring notated durations is that bars 282 and 318 prescribe ‘ripetere ad lib’. Surprisingly or not, the Bruck/Ross recordings take a uniform approach to this: their first recording (1986) presents 12 repetitions of each bar and their second recording (2008) presents six repetitions of each bar. The video of Josel and Pauwels’ 2021 performance, an output of the research project which resulted in the 2020 score edition, also presents six repetitions in each bar. As I will examine all three performances, I have opted to include both options.

Table 1:

Sectional analysis of Salut based on the 2020 version.

As well as these considerations I have also included a general description of particular features of each section. I am not suggesting that Salut is configured as a collection of extended techniques, but rather as constellations of various sound families that, crucially, offer a cross-related reflection on the various aspects of guitar attack (with and without plectrum), resonance on both sides of the stopping hands (with and without slide), the corporeality involved in their production and perception, and some cultural allusions within a formal plan which embodies transition and transformation in every sound parameter.

The quantitative comparison of durations in Figures 1a and 1b shows that both of the Bruck/Ross recordings often differ from the notated durations – they play either faster or slower than notated – while the Josel/Pauwels durations reflect their efforts to reproduce tempo indications as faithfully as possible. Indeed, as can be seen in their video recording of the piece,Footnote ¹⁸ they use earphones, probably hearing a clicktrack.

Figure 1a:

Comparison of durations in the score (12 repetitions in bars 282 and 318) in Bruck and Ross's first recording.

Figure 1b:

Comparison of durations in the score (6 repetitions in bars 282 and 318) in Bruck and Ross's second recording and in Josel and Pauwels’ recording.

As for the comparison of ratios in Figures 2a and 2b, this again demonstrates a tendency in the Josel/Pauwels recording towards a literal rendition. It is noticeable that Bruck and Ross's second recording is closer to the score, whereas their first recording has more interpretive freedom, perhaps because it is chronologically closer to the work's conception and the extensive period of collaboration between the composer and the Cologne duo in the revision of the score revealed by Josel's archaeological research.

Figure 2a:

Comparison of ratios in the score (12 repetitions) and Bruck/Ross I.

Figure 2b:

Comparison of ratios in the score (6 repetitions), Bruck/Ross II and Josel/Pauwels.

Figures 3 and 4 also show that the deviation of both durations and ratios in each of these three recordings confirms the clear tendency of Josel/Pauwels towards a more literal rendition of temporal aspects of notation, whereas the Bruck/Ross recordings tend to a freer interpretation.

Figure 3:

Comparison of deviations of durations in the three recordings.

Figure 4:

Comparison of deviations of ratios in the three recordings.

In terms of performing styles, Cook's categorisation of structuralist and rhetorical styles is useful in comparing these recordings. Although it could be argued that each performance oscillates between these two poles, my brief quantitative analysis demonstrates a clear dominance of a rhetorical style in Bruck and Ross's two recordings and a structuralist one in the Josel/Pauwels recording. If one thinks of the dedicatees’ first recording as a kind of authoritative version, since they not only commissioned the piece but also had an active and explicit co-creative role in its dissemination across many years, one might argue that their approach embodies a kind of Werktreue whereas the Josel/Pauwels approach is more akin to a Texttreue. In both cases, however, the differences are evidence not of the performers’ technical capacities but of their interpretive approaches; the differences manifest themselves in faster and slower tempi, regardless of the technical difficulties of each section.

Utz's archetypes of musical temporality make possible another categorisation, based on the perception of music as time.Footnote ¹⁹ Specifically, Josel and Pauwels are closer to Utz's architectural (or spacialised time) model,Footnote ²⁰ which is mainly synchronic and sustained by literalist performance traditions that focus on the individualisation of morphological events. Their version of Salut manifests a high degree of contrast in terms of dynamics, articulation and overall gestural clarity and definition. As for Bruck and Ross's recorded performances, they seem to adhere both to Utz's transformation (or processual time) modelFootnote ²¹ – the idea that each sound event is part of a large chain across the duration of the entire work, pursuing the metaphor of organic growth, with the result that there is a minimisation of dynamic contrasts – and the presentist (or moment time) model,Footnote ²² which draws upon Stockhausen's idea of ‘moment form’, offering a sense of fragmentation through its tendency to irregular tempi and an impression that individual events are isolated from one another. Such sophisticated theoretical categorisations of performing styles within the musicology of performance have proved to be problematic, however, as Dorottya Fabian states in her study of Bach performance practice:

Note that Fabian explictly uses the term differenciation, a Deleuzian term which implies the actualisation of the virtual, and not differentiation, which ‘determines the virtual content of an Idea as problem’.Footnote ²⁴

Salut as Assemblage: Beyond Interpretation?

After the ‘what it was’ of my examination of these selected recordings of Salut, I will turn to a Deleuzian ontology of music, which is more likely to open up a discussion of ‘what it could be’, crucially approaching music not merely as stratified structure but as assemblage. Paulo de Assis, another researcher working at the Orpheus Institute that hosted Josel's research on Salut, has advanced the application of assemblage theory for music, developing ideas derived from Deleuze and Guattari's ideas into artistic research.Footnote ²⁵ Josel himself has described his rigorous archaeological work, collecting and analysing the documents relating to Salut, ‘as constituting an assemblage (to use Deleuze's term), [which] can alter the way we perceive various relationships and interconnections between them, enabling new modes of thought, expression and perception’.Footnote ²⁶ These various documents are categorised by Josel as ‘strata’, a Deleuzian term that is a principal component of the concept of assemblage, carefully mapped into music by de Assis in his Deleuzian ontology of musical works.Footnote ²⁷ The concept of assemblage, however, shifts from ontology to epistemology in its emphasis on the dynamic and process-oriented nature of knowledge. According to de Assis, assemblages are not static structures but fluid arrangements of material, semiotic and social flows. They are perceived and understood through concrete performative operations that reconstruct them anew each time they are encountered. This perspective acknowledges that musical works and practices – even knowledge itself – are not fixed, predetermined entities, but part of a constantly evolving and contingent process.

The concept of assemblage is closely related to the notions of strata and diagram as its basic components. Strata refers to those components of an assemblage that are characterised by a high degree of coding and territorialisation. They are concrete, material and historical formations that result from intensive processes of stratification and individuation. They represent the fixed and stable aspects of an assemblage. On the other hand, a diagram is an abstract machine that operates within an assemblage. It is an image of something to come, rather than something that is already there. The diagram is not a structure, but rather a tracing of the flows of forces enabling the continuous unfolding of these flows and the emergence of new configurations. In other words, strata represent the fixed and stable components, while the diagram represents the dynamic and transformative aspects. They interact and interrelate within an assemblage, shaping its dynamics and processes of becoming. The concept of assemblage, therefore, encompasses both the stable and the dynamic elements of a complex system.

According to Assis, a crucial aspect of the study of strata is to consider them from both an historical and a philosophical and artistic perspective, checking them against the grain of common sense (which emphasises an idealised past against the present), and revealing them as (fragmentary) building blocks of a time to come. In this sense, the strata might also be a subject for archaeology because ‘archaeology does not necessarily refer back to the past. There is an archaeology of the present’.Footnote ²⁸ But how this could be actualised in Salut today?

In the book in which Assis consolidates his work on assemblage theory for music, he also proposes a generic and circular methodology for artistic research that involves three crucial steps: archaeology (focused on sources, documents and things), genealogy (historiographic, analytic and comparative research) and problematisation (research in and through artistic practice).Footnote ²⁹ Within this methodological frame, one can understand Josel's 2019 article on Salut in TEMPO as an exemplary model of musical archaeology, his article ‘Form can wait’ as a genealogy (focused on Salut's form) and his 2021 recording as a problematisation. At the same time, the comparative analysis of some selected recordings of Salut in the previous section of this article (which can be seen as a brief example of a genealogy) shows Josel and Pauwels’ performance standing closer to a paradigm based on historically informed interpretation, consistent with the research cluster HIPEX (Historical Performance Practices of Experimental Music) within which it was conceived. The Orpheus Institute's research project MusicExperiment21 – which advocated a paradigm shift from interpretation to experimentation and which strongly influenced Josel's archaeological work on Salut – also invites us to think about how these historical strata can be actualised in a productive confrontation with its possible futures.

A seminal example of an application of assemblage theory in music performance is provided by Rasch^x:

These performances problematise not only the strata in play but also the concert situation itself, adding diverse elements on stage that allow the listeners the freedom to focus on different layers of perception, transforming familiar artistic objects into objects for thought.

In my own recent performance of Salut – performed by non-German-speaking performers for a non-German-speaking audience – we attempted to make the work more immediately comprehensible for the audience by projecting Caudwell's text, translated into Spanish, alongside Josel's étude for the speaking section (see Figure 5).

Figure 5:

The Chilean premiere of the new edition of Salut in July 2023 played by Erik Marroquin and Diego Castro Magas.

This staging did not attempt a radical rethinking of the concert situation, but by offering the audience the possibility to understand Caudwell's text and integrate it into their experience of the piece, a range of new possible actualisations can be anticipated. We decided to keep the slide with the last words of the quotation, ‘then, we shall say…’ on screen for the rest of our performance, as a suggestion that the subsequent sections somehow embody a continuation of the text. Later, I discovered that Anders Førisdal's discussion of the pieceFootnote ³¹ suggests that the spoken section can be seen as an introduction, as if the composition begins after this quotation, with the ellipsis operating as a monumental colon.

Approaching Salut as assemblage invites questions about its diagrams: what informal and operative functions are likely to destabilise, deterritorialise and decodify its sedimented strata. The Caudwell text is certainly likely to operate as diagram, deterritorialising the concert situation but also reterritorialising the work within a non-German-speaking country. And so on; the possibilities unfold almost to infinity.

Among many other possibilities, the portion of Caudwell text omitted in the quotation – which belongs to ‘The Future of Poetry’, the last chapter of Illusion and Reality – turns its interest towards the future:

Beyond the political content inherent in this text's focus on the problematisation of art and society, this layer of content might also be included as a visual element in a realisation of the piece (either entirely or slightly modified, just as Lachenmann modified the previous section of the text), introduced after the spoken section to reinforce (or speculate on) the composer's original intent.

As to the immanent aspects of the performance, the comparative analysis of recordings of Salut strongly suggests that a performing style which resonates with the composer's intentions is closer to a rhetorical than a structuralist performing approach. Utz's presentist (or moment time) model can also help to enrich the contrast and isolation of musical events and characters, as a counterpoint to fine sonic transitions, and this recalls Lachenmann's performing suggestions for Pression, as reported by Orning:

The use of visual elements in a concert performance also picks up an idea proposed by Lachenmann: ‘I like (I am a utopian), I like to think of a civilisation in which going to a concert is an adventure of observing’.Footnote ³⁴ Salut is a piece which challenges us to imagine ways in which our embodied perceptions of the work can contribute to its future, and an experimental approach to its performance practice is perhaps the most appropriate response to this challenge.