Group Performance Paradigms in Free Improvisation

Abstract

This article proposes four paradigms of group performance in free improvisation: 1) sound composition; 2) social communication; 3) ‘parallel play’; and 4) ‘one beast with many heads’. While these paradigms are identifiably different, they are often engaged flexibly and/or in combination; and, importantly, it is very possible for the same performance to be experienced or interpreted by different performers or listeners according to different paradigm perspectives. After providing some background and defining key terms, the four paradigms are introduced and discussed; connections with, and departures from, the existing literature are considered, followed by analysis of illustrative examples of each of the four paradigms drawn from the broader field of sonic and electroacoustic improvisation. To conclude, further larger-scale research potential for furthering this investigation is identified.

1. Introduction

Small group free improvisation is a particularly fascinating performance practice. In part this is because it is a kind of creative ‘hothouse’ that supercharges a number of key aspects of the creative process at once and displays these for audiences to see (or, in our case, hear). For example, group free improvisation grants us the opportunity to watch the evolution of musical form in real time, in the moment of its conception; fascinatingly, in small group improvisation this is a collective endeavour, so it also places explicitly before our eyes (and ears) the interactions and collaborations between performers that allow this real-time creation to be spontaneously navigated. While, in every improvisation, both of these are taking place, all of the time, these are nevertheless quite different perspectives on what we hear happening on stage before us: two sides of the same coin perhaps, but substantially different, nevertheless. It is therefore possible for both performers and audiences to be focused on substantially different aspects of any given improvisation, at any given time.

This article proposes four paradigms of group performance in free improvisation. These paradigms attempt to articulate several identifiably different perspectives that might be adopted by an entire group, or by different individual improvisors within a group, or by audience members, in experiencing and interpreting an improvised performance. These are not ‘fixed’ perspectives, but fluid: more often than not, more than one of these paradigms is being engaged by a performer at any given time, although it is possible – perhaps likely? – that one paradigm will receive more focus at any given moment. While it is certainly possible for an improvisation to maintain a focus on just one of these paradigms – or, indeed, for a given improvisor to focus on one approach throughout their practice and their career – it is rather more common for performers to shift freely and flexibly (and, very likely, unconsciously) between these paradigms as they perform.

The observations expressed through these paradigms are based on my own experience: first as an improvisor and then as a pedagogue. Both of these activities have heavily informed the points of view expressed here: my own performance experiences, my interactions (onstage and off) with fellow performers and with audiences, and my observations of many years of students in the early stages of their attempts to navigate free improvisation.

2. Context

To begin with, let us clarify the context with which we are dealing: we are concerned here specifically with group improvisation, within the broader practice often referred to as ‘free improvisation’. The former is relatively straightforward, but the latter requires some introduction and unpacking.

2.1. Group improvisation

To begin with, I would argue that solo improvisation and group improvisation are, substantially, two different practices and two different art forms. While both (at least nominally) involve the creation of musical form and materials in real time, the means by which this is done are very much distinct. The practice of collaborative real-time creation requires its own skill set, skills which are not activated at all in solo improvisation: the art of communicating musical intention to other performers through sound alone; the ability to respond to unanticipated shifts in musical direction with speed and dexterity; being adept at balancing the need to take creative responsibility through one’s own musical contributions with the need to leave ample space for other performers to do the same; etc. None of these are engaged in solo improvisation (or, if they are, then to a very limited extent); as a result, by comparison, solo improvisation appears somewhat one-dimensional. It is the multi-dimensional aspect of group improvisation that opens up the opportunities provided by the multiple paradigms described in this article. ¹

2.2. Free improvisation

The idea of ‘free’ improvisation can be difficult to pin down, and existing definitions are not always entirely satisfying. Personally, by ‘free improvisation’ I mean:

1. 1. Performance in which nothing has been discussed or pre-planned by the performers beforehand, at any level.

2. 2. Music that operates outside established formal or other genre conventions and therefore does not rely on shared and recognised formal or structural models. As a result, in ‘free improvisation’ the real-time imagining and creation of form and structure, as well as of musical materials, are central to the art form.

Notice, though, that both of these points define ‘free improvisation’ in the negative, that is, according to what it is not. This is fairly common – consider, for example, the following from improvisor Derek Bailey in attempting to define (or not) ‘free improvisation’:

Diversity is its most consistent characteristic. It has no stylistic or idiomatic commitment. It has no prescribed idiomatic sound. The characteristics of freely improvised music are established only by the sonic-musical identity of the person or persons playing it. (Bailey 1992: 83)²

The challenge of definition is further complicated by the choice of the word ‘free’, which is full to bursting with loaded connotations: ‘free from’ what, exactly? Does this mean that other forms of improvisation are necessarily less ‘free’? If we limit ourselves to the points outlined earlier – ‘free from’ pre-agreed or pre-arranged structures/forms/materials; ‘free from’ genre conventions or pre-established models and so on – then this holds up fairly well.³

2.3. Sonic and electroacoustic improvisation

Personally, I am particularly drawn to what I see (or hear) as substantial common ground between free improvisation and electroacoustic music. This connection can be quite obvious, as in for example the practice of ‘electroacoustic improvisation’ (EAI) – from the purely electronic, to the more ‘mixed music’ paradigms found for instance in the example from the Evan Parker Electro-Acoustic Ensemble in section 5.2. However, I would argue that the connection lies deeper than this somewhat simplistic overlap between the tools and instruments involved. The soundworld of free improvisation often gravitates towards a sonic palette of gestures, timbres and textures that are quite familiar to the electroacoustic listener, regardless of whether these emanate from electroacoustic resources or more traditional instruments. Perhaps more fundamentally, the close listening on which much free improvisation is predicated occupies a similar space to practices such as ‘reduced listening’, ‘deep listening’ and others that have emerged over the decades.

This is relevant because it informs the perspective revealed in the discussion of the paradigms proposed in the following sections – especially the first paradigm, which, as we will see, shares much in common with studio-based composition, but transposed to a scenario of creation in real time.

2.4. A caveat

I would like to stress that, while what follows may (hopefully) be useful as observation, as analysis, and as musicology, it is not intended as any kind of an improvisation ‘method’. The best improvisation relies on being unfettered, intuitive, spontaneous; attempts to encode improvisation practice, for example by formalising strategies, risks ‘cluttering up’ the mind and preventing the free and unhindered flow of ideas and expressions. I do believe, as an improvisor, that it is valuable to think about these things; however, in my opinion it is important to ‘turn off’ this kind of analytical or musicological thinking before one sets foot on stage.

The exception to this, is the kind of onstage situation that is all too familiar to many improvisors: the constant risk of finding yourself empty-handed, of standing in front of an audience with no ideas materialising, with nothing to say and nothing to play. In these instances, all bets are off, and most improvisors, consciously or unconsciously, have equipped themselves over time with a collection of strategies for moving forward in such a situation, before panic sets in. Any tool that might prove effective as a mental or creative ‘crutch’ to get out of these ‘deer in the headlight’ moments should remain on the table; so, if deploying these paradigms as a ‘method’ proves useful in such a context, one is welcome to hold on to this as part of the ‘panic button’ toolbox.

3. Four paradigms of group free improvisation

When free improvisors perform in a group, there are several identifiably different modes, perspectives or positions that they might choose to take – either from moment to moment, or across entire improvisations or sets. When audiences experience live improvisation, they are equally likely to engage with these same perspectives – either in their own experience of the music, or in their interpretation of what they think the performers’ focuses might be, or both. I propose four paradigms that demonstrate different positions regarding key facets of group free improvisation, such as creation, composition, communication, collaboration and so on.

While these paradigms represent differences in priorities – sometimes substantial – they are not mutually exclusive. On the one hand, it is possible to shift from one paradigm to another, flexibly and fluidly, throughout an improvised performance; and it is equally possible to simultaneously hold positions rooted in more than one paradigm (in which case, we might perhaps consider the balance or weighting of the paradigms at play in any particular moment). Except in extreme cases, it is often possible, as audiences and listeners, to interpret, analyse or describe the same improvisation from contrasting paradigms, much as it is possible for performers and audiences to arrive at different conclusions regarding the most appropriate paradigm for a particular improvisation, or indeed for different performers to similarly arrive at different conclusions. Far from being a problem, this demonstrates the strength of these paradigms – or, at least, indicates that they are realistic: free improvisation is fluid and multidimensional, so this should be kept in mind should we be tempted to take too rigid a position in the application of these paradigms.

The four proposed paradigms are:

1. 1. Sound composition.

2. 2. Social communication.

3. 3. ‘Parallel play’.

4. 4. ‘One beast with many heads’.

We will discuss these one by one, then compare and contrast them with similar propositions from existing literature, and finally use repertoire examples as likely or potential demonstrations of each.

3.1. Sound composition

This paradigm is closely connected with the idea of improvisation as ‘real-time composition’. The focus is on the performance as an evolving piece of music and on collectively shaping that piece of music as it unfolds. Thinking here is not dissimilar to compositional thought while composing in the studio, with the primary exception that ideas must be expressed at the same time as they are conceived, without recourse to correction or refinement after the fact. Form, or the ‘shape’ that the music takes over time, is of particular importance in this paradigm; however, again, there is a special case here in that form, once expressed, cannot be re-ordered or re-arranged – it can only be created in a linear fashion, without recourse to second thoughts. ⁴ In this paradigm, what matters is the performance as an evolving composition, which should be judged on the merits of its musical outcome alone.

3.2. Social communication

In this paradigm, the focus is on the interactions between improvisors. Small-ensemble improvisation might include any combination of solos, duets, trios and so on, in any subgrouping up to and including the full ensemble. More importantly, it can incorporate any number of forms of interaction: dialogue, call-and-response, conversation, interruption, ignoring somebody and so on.

For example, in a hypothetical performance, one player begins then a second player responds, evolving into a dialogue. A third player engages, expanding the conversation. After a period of three-way discussion, a fourth player loudly interrupts. The first three players now must choose whether to acknowledge this interruption and engage, starting a new conversation, or whether to ignore the interloper and continue with their original conversation and so on.

The focus is therefore entirely on the social aspect of the evolving improvisation: interpreting the performance as a web of evolving musical interactions between the individual performers.

3.3. ‘Parallel play’

I have taken the name of this third paradigm from one of the six stages in the development of play in young children, proposed by Mildred B. Parten (1932, 1933), that can be observed as children grow and develop from infants to five year olds. Early play, for example at two years old, is classed as ‘solitary play’: the child plays alone and is unaware of or unengaged with other children who may be nearby. Eventually children arrive at ‘cooperative play’, in which children are able to fully engage with one another in mutually organised game-playing. In between, there is a category known as ‘parallel play’, in which children play alone, but in the shared company of other children who are themselves playing their own individual games: ‘[t]hey are aware of the presence of peers – in fact, the presence of others obviously has some meaning for them – but each child is still playing separately’ (Hughes 2009: 101–2). ‘Parallel play’ is therefore a transitional stage, leading from ‘solitary play’ to ‘cooperative play’ (albeit with a few other steps identified along the way).

In free improvisation, ‘parallel play’ is intended to denote a paradigm in which each performer focuses primarily or entirely on their own musical output, without concerning themselves too much with the details of the contributions of others or with the collective result – instead trusting that the sum total of these musical contributions will be rich and interesting. While the ‘parallel play’ paradigm can be engaged at any time, there are particular circumstances in which it is to some extent imposed: improvising at volume levels that preclude close listening between performers, or performing at so rapid a pace that detailed response to other performers is at least curtailed. This works best among improvisors who have a degree of familiarity with one another’s performance styles, as this paradigm is built on a foundation of faith that, whatever the details of a particular performance might be, the combination of these individual improvisors performing simultaneously will collectively produce something worth listening to. Examples include improvisors closer to the ‘noise music’ end of the spectrum – whether this be any of a variety of electronics, or powerfully amplified acoustic instruments (such as perhaps some performances by saxophonists Peter Brötzmann or Dror Feiler), presenting situations in which performers might not be able to hear each other very well, or at all – as well as situations where the density and/or the intensity of improvisation (whether, for example, at the ‘climax’ of a performance, or more broadly as the primary style of playing) is such that the performers cannot fully register and respond to the finer details of the performances of their fellow improvisors. In both cases, each performer’s focus is largely or entirely on their own output, in contrast with the ‘sound composition’ paradigm (in which each performer’s focus is on the global, collective output) and the ‘social communication’ paradigm (in which each performer’s focus is on their detailed interactions with other performers). ⁵

3.4. ‘One beast with many heads’

It could be argued that this final paradigm is the ‘holy grail’ in all free improvisation: it is the goal, the ‘ideal state’, but, unlike the other three paradigms, it cannot simply be engaged by choice – one can only open the door and leave room for this paradigm to materialise. It is a special condition in which individual performer identities are subsumed, and the group becomes in essence a single ‘hive mind’: one performer with many heads and many hands. ⁶ For the audience this paradigm may perhaps be indistinguishable from a very successful improvisation under other categories, but for performers it is a very special case: one ‘loses oneself’ in the performance, and when the improvisation is over, it is like coming out of a trance – one is not quite sure what happened, one’s memory of the performance that just took place is fuzzy at best, one is not sure how long one has just performed for, and so on. Almost by definition, the very finest improvisation is found here. The ‘social communication’ of paradigm 2 focuses on close interaction between performers; in paradigm 4, inter-performer engagement is so close as to become unconscious, with all performers performing together ‘as one’ – which equally offers the ideal state for delivering the ‘sound composition’ paradigm, with all performers working inextricably towards a unified goal. Unlike the other paradigms, the ‘beast with many heads’ can only be reported by the performers themselves; audiences and listeners can only guess at its possible appearance.

3.5. The paradigms in action

Recognising the possible distinctions between these paradigms may help in bridging possible communication gaps between performers, or between performers and audiences. To help illustrate this, and to highlight the differences between paradigms, incidents from my own experience as an improvisor will be drawn upon to provide an example of the paradigms in action.

Improvisor one listens to the evolving improvisation, and decides that the overall sounding result is a bit ‘thin’. Improvisor one therefore decides to add a layer of shuffling ‘texture’ to ‘fill out’ the sound a bit. This textural layer is relatively stable and static, albeit with a good amount of internal differentiation; it is not intended to draw attention to itself, but simply to complement the overall sounding result.

At the same time, improvisor two is engaged in musical conversation with fellow performers, drawing out individual performers in a series of dialogues. Improvisor two attempts to engage improvisor one in dialogue but this is not successful: despite offering several short musical suggestions to improvisor one to try to engage them in conversation, improvisor one’s output does not change – improvisor one appears to be ignoring improvisor two.

In this example, improvisor one is primarily engaged with the ‘sound composition’ paradigm, while improvisor two is primarily engaged with the ‘social communication’ paradigm. From the perspective of the ‘sound composition’ paradigm, it is important that improvisor one maintain a steady textural layer; however, from the perspective of the ‘social communication’ paradigm, the fact that improvisor one is not responding to improvisor two’s invitations to dialogue suggests that improvisor one is either ignoring improvisor two, or simply is not listening. From one perspective, improvisor one is providing a valuable compositional contribution; from the other perspective, improvisor one is proving to be a poor – or, at least, thoughtless (which is likely the same thing) – improvisor. Without recognising this difference between the paradigms engaged, tension, friction or hostility can result.

On the surface, ‘parallel play’ appears to offer the likeliest grounds for miscommunication. All other paradigms are predicated, either implicitly or explicitly, on close listening to one’s fellow performers; ‘parallel play’ is not – or, at least, in ‘parallel play’ the fundamental importance of close listening is significantly reduced. This may be uncontroversial at moderate performance levels; however, as noted earlier, it is precisely at more extreme levels that ‘parallel play’ is most likely to be invoked. Thus, for example, if one performer launches in at either a volume level or a level of density and intensity that appears to ‘shut out’ close engagement with their fellow performers, this can seem a hostile act. However, as described earlier, ‘parallel play’ is predicated on a strong degree of understanding and trust between performers; one would only typically engage this paradigm if one were already confident that it would be effective for that particular combination of performers – that is, that the combined sum of those performers playing ‘their own thing’ simultaneously would lead to a valuable musical result. It is therefore unlikely that an individual improvisor would engage this paradigm without the understanding of their fellow performers. Unlike in the previous example, if one improvisor brazenly launches into ‘parallel play’ in a manner that prevents other performers from engaging other paradigms should they so choose, then that would arguably represent a poor – or, at least, inconsiderate – improvisational choice.

4. Connections with the literature

There is much in the preceding model that connects with the existing research literature on improvisation; however, much of this tends to focus on a particular approach, often matching with one of the paradigms we have proposed here.

For example, MacDonald and Wilson (2020: 77) propose a set of ‘key choices’ for group improvisors, stemming from some solid practice-research work with free improvisors: maintaining, initiating, adopting, augmenting and contrasting. While it may be possible to interpret some of these terms as falling under the ‘social communication’ paradigm, it becomes clear through the descriptions and examples offered that in fact these all fall under paradigm 1, ‘sound composition’. Another example of paradigm 1 is Roger Dean’s Creative Improvisation, which offers a ‘method’ for improvisors, some of which fits well with the ‘sound composition’ approach to free improvisation – especially the chapter ‘Timbres and Textures for Improvising’ (Dean 1989: 33–9).

Paradigm 2, ‘social communication’, has received a great deal more research attention in recent decades, though some of this is in more advanced realms of theory than the relatively straightforward approach intended here. Ingrid Monson’s Saying Something: Jazz Improvisation and Interaction (Monson 1996) provides a good example of an approach focused primarily on improvisation as a network of interactions between performers. There is also a range of choice expressions, often found in interviews with performers, that fit well within this paradigm, such as vocalist Maggie Nicols describing improvisation skill as being a question of ‘social virtuosity’ (Smith 2004: 236), or Eddie Prévost’s discussion of free improvisation as ‘psycho-drama’ (Prévost 2004: 11). Borgo (2022: xvii) discusses ‘social cognition’ in improvisation, while Born (Born, Lewis and Straw 2017: 9–10 and 41) use the term ‘social aesthetics’ (although Born is focused on much broader questions of the political implications of performance). A number of writers point to Bourriaud’s ‘relational aesthetics’ (Bourriaud 2002) as a key touchstone here – for example, Marino (2021: 7), Born (2017 et al.: 33–9) and Marcel Cobussen, who describes relational aesthetics as ‘an aesthetic of the inter-human, of encounters, of proximity … the realm of human interactions and its social context; it is an art form where the substrate is formed by intersubjectivity, taking being-together as its central theme’ (Cobussen 2017: 96). However, while this would seem to apply very well to the ‘social communication’ aspect of free improvisation, Bourriaud is in fact more concerned with a broader question of ‘participation’ – for example, an audience’s engagement with an artwork – than with the detailed interactions between performers.

One of the closest terms to what is described here as paradigm 3’s ‘parallel play’ is Eddie Prévost’s description of ‘non-dialogical interactions’ (Prévost 2004: 92). Elsewhere, in describing the work of foundational free improvisation ensemble AMM (of which he was a founding member), Prévost says that ‘[p]art of AMM’s philosophy, its ethos if you like, is the idea of concurrent commentary: separate voices speaking at the same time, interweaving and interleaving. But each voice is not atomised or individuated’ (Bailey 1992: 129). This strikes me as a very good description of the ‘parallel play’ paradigm – fittingly perhaps, as we will use a performance by AMM as an example of this paradigm in the next section.

There has been increasing interest in recent years in the idea of a ‘flow state’ in improvisation and elsewhere (Csikszentmihalyi 1997); this is central to the final of our paradigms, ‘One beast with many heads’ – which, it could be argued, is one of the clearest examples of the ‘flow state’ in action. However, this paradigm relies on an extension of ‘flow’, to what Sawyer has termed ‘group flow’: ‘the magical moment when it all comes together, when the group is in sync and the performers seem to be thinking with one mind. … This is when audience members think [the performers are] reading from the same script – even though there’s no script’ (Sawyer 2007: 51). However, Csikszentmihalyi and Sawyer intend these as a broader category, which describe a fundamental prerequisite for ‘good’ improvisation in any paradigm – with performers fully focused and engaged, and so on. The ‘one beast with many heads’ paradigm goes further and is perhaps at the extreme end of the ‘flow’ or ‘group flow’ state. For this we need to look, for example, to theories of the ‘extended consciousness’ that stems from ‘inter-brain neural synchronization’ between collaborators working extremely closely and creatively together (Valencia and Froese 2020); however, this is beyond the scope of the discussion here. Pierrepont describes something similar, but with a slightly different, dual model, in which improvisors ‘abandon themselves to themselves, to a stream of consciousness … while at the same time according the most rapt attention to what their colleagues are playing, without ever abandoning them definitively, without ever following them definitively … placing themselves in a state of unconsciousness and hyperconsciousness at the same time: a state, that is, of double consciousness’ (Pierrepont 2013: 211). A final touchstone for paradigm 4 is the idea of improvisation as a way of accessing ‘spiritual, ecstatic, or trance-like performance states’ (Borgo 2022: 28; see also Borgo 2003), which is most often discussed in relation to free jazz (e.g., the work of Albert Ayler) but can clearly be applicable to free improvisation as well.

Some sources reinforce several paradigms at once. For example, MacDonald and Wilson report on having conducted a trio study, in which the performers described the same passage from a recorded improvisation thusly: ‘one person in a trio saw a texture being created through the concerted efforts of all three members; another member considered only herself to be responding to someone else’s choices; the third saw all players’ input as unrelated, with each having a different idea about the music: “three different thoughts going on at once”’ (MacDonald and Wilson 2020: 101). This, rather elegantly, demonstrates paradigms one (sound composition), two (social communication) and three (‘parallel play’), in perfect order, while also serving as an excellent example of how the same performance can be experienced or interpreted from different paradigm perspectives by different performers or audience members. ⁷

5. Analysis

We will now attempt to apply and demonstrate the above paradigms through analysis of recorded improvisations. As has already been stated, most free improvisation is likely to access or suggest some degree of several of these paradigms, or at least of the first two, simultaneously; however, representative examples have been selected that highlight or prioritise a specific paradigm, and that are therefore usefully illustrative. ⁸

5.1. Paradigm 1, sound composition: Fergus Kelly, Max Eastley and Mark Wastell – improvised music at St Mark’s Church, Myddelton Square, London

As an example of the ‘sound composition’ paradigm, we will consider an improvised performance by the trio of Fergus Kelly, Max Eastley and Mark Wastell, which took place on 28 March 2009 at St Mark’s Church, Myddelton Square, London, accessed via a video made available online (Kelly, Eastley and Wastell 2009). ⁹ In this performance, all three performers play a variety of percussion instruments and objects, both pitched (chimes, singing bowls, xylophone bar, tam-tam) and unpitched (including Eastley’s performance, which is largely focused on several flat stones), together with a number of sustained electronic drones and feedback. The video is 10 minutes long; but, as the performers are already in action when the video begins and are still in action when the video ends, the impression is that this is an excerpt from a longer performance. However, for the purposes of our consideration here, we will take the video as though it were the complete work.

The performance is slow, studied, and careful, suggesting very close listening by the performers. The primary focus is on timbre and texture. The choice of sound materials – both the sustained materials and the shorter percussive materials – as well as the location of the performance suggest that space is also a substantial element of the performance; however, this is not properly available to us via the recording, or at best only partially. There is a primary division between an ‘environmental’ layer, made up of long sustained sounds – both pitched (percussion, tones, feedback) and more granular (e.g., shuffling sounds, scraping sounds) – and an ‘event’ layer of short individual interjections (e.g., woodblock, stones scraping). One could argue that there is also an intermediary layer, in which what starts as an ‘event’ triggers a sustained sound that becomes part of the ‘environment’ layer (e.g., Wastell striking the xylophone bar or the singing bowls). For the most part, the dynamic is quite low, but there is nevertheless a trajectory of the dynamics across the performance, which allows us to propose a form for the piece:

• 0:00–3:00: section 1 – introduces the primary materials and layers; low dynamic, medium density.

• 3:00–5:00: section 2a – sustained materials drop out; focus is on short individual gestures and events, with lots of air and space around them; while there is variety in the dynamics of individual events, the overall dynamic remains low, with very low density.

• 5:00–6:40: section 2b – similar to section 2a, but a low quiet drone gives this section a slightly different character.

• 6:40–10:00: section 3 – this section builds to a climax, both in terms of dynamic and density; sustained pitched material becomes louder and stronger; gestural events become louder and more active; climax shortly before 8:30 with louder percussive sounds, followed by a slight decrease in dynamic, density and tension.

Thus, the performance can be broken into three roughly equal sections: 1 (3 minutes), 2a/2b (3 minutes 40 seconds), 3 (3 minutes 20 seconds).

As the description thus far might indicate, my focus as a listener is almost entirely on the performance as real-time composition: my attention is drawn to the qualities of the sound materials, to the evolution of the form and so on. In addition, in the video there is no indication of any extra-musical involvement between the performers: there is no eye contact; no one is watching to see who is performing what, or what they might be planning; and so on. Perhaps more importantly, the nature of the sound materials and the manner in which they are deployed – either sustained or sparse, with no indications of clear or immediate sonic or musical interactions or direct responses – discourages any attempt to interpret the performance primarily through the lens of the ‘social communication’ paradigm, instead strongly suggesting that the performers are primarily concerned with the performance as an evolving composition and that the listener might usefully do the same. The performers’ apparent close listening discredits ‘parallel play’ as a likely candidate here; paradigm 4, ‘one beast with many heads’, remains possible but is not necessarily identifiable to the outside observer.

5.2. Paradigm 2, social communication: Evan Parker Electro-Acoustic Nonet – Festival Météo, 2015

To demonstrate paradigm 2, ‘social communication’, we will take as our example a live concert video by the Evan Parker Electro-Acoustic Nonet, recorded on 28 August 2015 at the Festival Météo/Mulhouse Music Festival, France (Parker 2015). While the full performance is an hour long, we will examine only the first six minutes, which is enough to get a general sense of the approach taken.

For this performance, the ensemble consists of:

• Evan Parker, saxophone

• Barry Guy, double bass

• Paul Lytton, percussion

• Peter Evans, trumpet

• Okkyung Lee, cello

• Sten Sandell, piano and keyboard

• Sam Pluta, computer

• Richard Barrett and Paul Obermayer, sampling keyboards.

The performance begins (1:55) with a strong solo by Sam Pluta, using controllers to perform a granular software patch, that mounts quite quickly in dynamics and density, adding in short, pitched gestures as it builds. This prompts the entry of Barrett and Obermayer (shortly before 3:00), contributing similar electronic gestures with some contrast in their characteristics. Paul Lytton enters simultaneously with some quiet percussion gestures and/or textures; however, these are quite subtle and suggest that Lytton is perhaps more interested in indicating his collaborative participation than in making a significant impact on the overall sound. As Pluta’s solo reaches a climax and begins to dissolve, Barrett and Obermayer provide increasingly strong, clear gestures that mark (or trigger?) the end of the first section.

At 3:55 Parker introduces a clicking/bubbling semi-pitched texture on the saxophone, which effectively serves to transition from a similar material in the electronics at the end of section 1 to a new section focused around the other instruments. Lytton enters with more active gestural/textural work on the percussion, to mark the new section, while Barrett and Obermayer continue with similar gestures to those at the end of the first section (again aiding in the transition between sections), building into a clear duet between the saxophone and the electronics, which peaks and then builds back down.

At 5:00 Barry Guy marks the start of a new section with a clear three-note gesture on the double bass, which is immediately answered by a short gesture from Okkyung Lee on cello. After a brief gestural dialogue between Guy, Lee and Barrett, the new section fills out around a more sustained, dronelike texture, introduced first by Parker, joined next by Peter Evans on trumpet and then Lee on cello. At 5:29 Sandell interjects sharply on piano, before joining the drone with a low rumble on the lower strings, then interjecting again at 5:48, to which Guy responds. This constructs a conversational layer on top of the drone layer, with performers moving somewhat flexibly between the two. There are clear call-and-response interactions throughout, for example starting at 5:52 that moves from piano, to bass, to saxophone, to cello, back to piano and to bass; or this might alternately be interpreted as a counterpoint of two dialogues – one between piano and saxophone, the other between cello and bass. Barrett enters at 6:15 to dialogue with the piano, using piano samples as his sound material. The music then moves towards a new section, anchored initially around the cello, bass and piano; however, we have covered enough ground for our purposes, so we will stop our analysis here (as an analysis of the complete hour-long performance is beyond the space available).

Compared to the previous example, there is much in the preceding description that focuses on the roles taken by individual performers, and how these interact and change over time. To summarise:

• 1:55: Pluta (computer) begins – solo

• 2:55: Barrett and Obermayer (sampling keyboards) join – dialogue

• 3:55: Parker (saxophone) enters – transition to new section

• 5:00: Guy (double bass) launches new section

• Guy (bass), Lee (cello), Barrett (sampling keyboard): dialogue

• 5:08: Parker (saxophone) launches drone; joined by Evans (trumpet) and Lee (cello)

• 5:29: Sandell (piano) interjects, then joins drone

• 5:48: Sandell (piano) interjects again; Guy (bass) responds

• Call-and-response: piano/bass/saxophone/cello (or, counterpoint: piano/saxophone dialogue + bass/cello dialogue)

• 6:15: Dialogue – Sandell (piano) and Barrett (piano samples).

So, we have solos, dialogues and conversations; we have people joining each other, or interrupting each other; we have multiple conversations happening simultaneously; and so on. These are the hallmarks of paradigm 2: the experience or interpretation of an improvisation primarily as a map of interactions between performers. Lytton’s percussion entry early in the performance provides a useful, and common, case in point: joining the performance with a very quiet contribution, possibly too quiet to be heard (and thereby ineffective in terms of paradigm 1), but which demonstrates the performer’s willing participation and collaboration (extremely effective in terms of paradigm 2).

However, note also that the preceding description does not focus solely on social factors; in fact, it moves regularly between a description of the roles and interactions of the various performers, and the musical consequences of those choices and decisions. In fact, we could rewrite our summary thusly:

• 1:55: Intro/section 1 – granular materials, plus gestural materials

• 3:55: Transition – bubbling material

• 4:20: Section 2: counterpoint – longer phrases and gestures

• 5:08: Section 3a: drone

• 5:50: Section 3b: counterpoint.

In other words, this performance also provides an excellent example of a point that has been stressed many times already: the differences between paradigms are not black and white, but rather are more a question either of interpretation or of the fluid mixing of and moving between paradigms. As listeners, this performance by the Evan Parker Electro-Acoustic Nonet is quite satisfying whether taken on the terms of paradigm 1, or paradigm 2; and, we can assume with some confidence that the performers are themselves moving quite fluidly between these two positions.

5.3. Paradigm 3, ‘parallel play’: AMM – ‘Like a Cloud Hanging in the Sky’, from The Crypt – 12th June 1968

To demonstrate the ‘parallel play’ paradigm, we turn to a classic performance from an ensemble credited as one of the founders of free improvisation practice: AMM, and the opening track from their live recording of a performance at The Crypt, London, on 12 June 1968 (AMM 1992). Unlike the more contemplative music for which AMM later became known, the sound here is huge and very dense – although admittedly, as a late 1960s live club recording, the sound quality is not fully up to the task, so we must to some extent use some imagination.

The performers for this recording are credited as Eddie Prévost and Chris Hobbs, percussion; Keith Rowe, electric guitar; Lou Gare, saxophone; and Cornelius Cardew, piano. However, this does not prepare the listener for the screeching wall of sound that results, especially during this opening 45-minute track. ¹⁰ The most prominent sound materials appear to be scraped metal sheets and feedback, with some softer textural work on drums and cymbals providing a background. Occasional bursts of electric guitar appear, as well as periodic piano notes, bowed cello or violin, and what sounds occasionally like bits of radio (although this could be imagined). It is only very occasionally that a squealing saxophone can be identified, and then only briefly. Around the 30-minute mark this titanic sound mass collapses, to reveal sustained high-pitched drones and somewhat mysterious rattlings and rumblings in the background. The texture continues to thin, and the sounds become increasingly delicate, leading the listener into something like a state of hypnotic rapture after the assault of the opening 30 minutes.

For the most part, this performance is a clear contender for the ‘parallel play’ category, for both practical and musical reasons. Practical, because it seems unlikely the sheer mass and volume of the soundworld during the bulk of the piece would allow for anything else; musical, because this is a canvas of enormous sonic blocks, often very chaotic and sometimes without clear differentiation for minutes at a time, again making the close listening and detailed responses usually involved with the first two paradigms unlikely. Instead, performers are busily and effectively making their contributions to the overall mass, likely shifting according to the building and crashing of the waves of sound, following the broader shape of the evolving performance and trusting that their own intensity will help sustain the momentum of the overall musical monolith. Consider, for example, the section from 20:00 to 30:00: no room here for the detailed interaction of paradigm 2, nor for the subtler compositional touches of paradigm 1.

This changes very substantially, however, in the final third of the piece, in which both the dynamics and the density reduce dramatically. The move away from ‘parallel play’ is extremely striking, and we very quickly find ourselves in the subtlest, most responsive realms of the ‘sound composition’ of paradigm 1: the engagement between the performers becomes exceedingly clear, as sustained tones float and align together, intrusions and interventions are sparse and carefully considered, and this beautiful final section becomes increasingly delicate. In fact, it could be argued that it is this dramatic shift from paradigm 3 to paradigm 1 that creates the main drama of the piece, and it is very moving indeed. ¹¹

5.4. Paradigm 4, ‘one beast with many heads’: VCA – ‘Umlaut’, from Ceres

Unlike the first three paradigms, it is rather more challenging to confidently offer a definitive example of paradigm 4, ‘one beast with many heads’. There are several reasons for this:

• It is the rarest of the four paradigms.

• It is the only one that cannot be recognised from the outside, by the audience or listener, but can only be reported by the performer(s), after the fact.

• It is the only one that cannot simply be ‘chosen’ – that is, a performer cannot simply ‘choose’ to engage this paradigm, but, as described earlier, can only open the door and hope for the best.

• Since this paradigm is rare, it is even more rare that it is captured and documented when it does arise (and, even if one has been lucky enough to document such a performance, rarer still that one has the presence of mind to note down for posterity that the ‘beast with many heads’ had made an appearance for this or that portion of a concert or set, on this particular date, for future analysis and examination).

Thus, it is very difficult to examine something that, generally speaking, cannot be confidently identified from the outside and cannot deliberately be invoked by the performers – indeed, any pressure placed on improvisors to do so is almost guaranteed to completely prevent its appearance. Nevertheless, we will consider a track titled ‘Umlaut’ (VCA 2017; Sound Example 1), from the album Ceres, performed by a trio of Sergio Castrillón on cello, Marc Vilanova on saxophone, and myself on percussion and electronics, as our example of paradigm 4, with a caveat immediately following.

The performance unfolds in layers of texture and timbre that come and go throughout the piece, with long builds and swells in dynamics, sudden cuts and constant micro-variations. Layers range from low drones to very high-pitched electronics; textures include extended techniques on the saxophone, providing a gentle ‘bubbling’ sound around the reed and mouthpiece, and both mechanical and synthesised textures from the percussion and electronics. Energy across the piece is carefully controlled, from its gentle opening to more manic passages and rougher, more frenetic playing. The piece ends by ‘peeling away’ layers of texture and timbre one by one, until only a low, quiet pitch on the cello is heard, which then stops.

If we hear the ‘beast with many heads’ in this example, then this is perhaps through the sense of multiple performers all working together to achieve a single compositional aim. For much of the performance it is difficult to single out any individual performer’s contribution; instead, the piece works through its shifting landscape of textural layers, with all performers working hand-in-hand to establish these layers and to shape a counterpoint between them as the piece unfolds. It is this which gives the sense, one hopes, of multiple performers working together ‘as one’.

However, I will immediately hold up my hands and confess to having misled the reader; in fact, the ‘beast’ did not make an appearance during this performance. Nearly, perhaps; this was a very relaxed and comfortable recording session, with excellent colleagues and good communication between the performers, and generally an inviting atmosphere for the ‘beast’ to be welcomed in – but, alas, no trance state was enacted; egos did not merge and disappear; and, despite a strong performance, we remained three distinct performers, regardless of how seamlessly we might have played together that day. So, owing to the challenges described earlier in attempting to capture and record the ‘beast’, we must be satisfied with my contention that, when the ‘beast’ does appear, it might sound something like this.

6. Shortcomings and future directions

Despite the usefulness of the four-paradigm model presented here, it is not without its shortcomings in its current form; these primarily revolve around the need for a more rigorous examination and analysis of these paradigms and their performance outcomes. For example, a future project hopes to chase the ‘beast with many heads’ paradigm and attempt to collect enough documented examples for a substantial examination to potentially yield interesting results; however, to do this properly would require a research project of substantial scale, in terms of duration, documentation and so on. Another shortcoming is that, with the exception of the VCA performance, all the examples included here were presented only from the listener’s perspective; since the proposed paradigms address both performer experience as well as audience interpretation – or, if anything, prioritise the former – the performer’s viewpoint on their engagement with these paradigms in their own performances would be important. However, this can be tricky, as this is obviously very subjective, open to bias and agenda, and can be reconfigured in the memory – or, indeed, simply be forgotten. As a result, in order to collect meaningful data on performer interpretation regarding these paradigms, one has to document live performances and collect feedback from both performers and audience immediately after the performances, ¹² which can then be compared and contrasted. This work has begun and is currently underway, at least in a ‘proof of concept’ form; but I look forward to the opportunity to tackle this at the scale it requires and deserves and to reflect further upon the model presented here at that point.