2.1. Bottom-up empiricism and top-down rationality

The idea that the design and manufacture of acoustic instruments can be characterised as being craft-orientated, led by ‘bottom-up exploration’ whereas digital instrument design is necessarily tied to the top-down application of rational theoretical principles, is on the face of it a persuasive generalisation (Magnusson Reference Magnusson2009: 173–4). Nonetheless, it is a generalisation, as Magnusson (Reference Magnusson2009: 173) acknowledges. The basis for explaining differences between technological paradigms shifts significantly if we cease to treat these design orientations as mutually exclusive. There are two components to this. The first is to consider the question in terms of the specific type of rationality that Magnusson implies when he talks of ‘a mathematical understanding of sound’ (Magnusson Reference Magnusson2009: 174), insofar as this represents a particular corpus of conceptual scaffolds. The second is to consider the nature of conceptual apparatus in making more generally.

On the first front, scientific knowledge can be seen to have been applied to the design of acoustic instruments alongside more exploratory design strategies. This is particularly evident with various innovations to the design of woodwind instruments during the nineteenth century (Campbell, Greated and Myers Reference Campbell, Greated and Myers2004). Conversely, we can observe that not all approaches to digital instrument design are predicated on the application of scientific knowledge, such as with George Lewis’ account of developing Voyager:

Secondly, considering the relationship between making and the world of ideas more generally, the situation appears a little more complex. Magnusson notes that in order to design digital musical instruments, extra-technical ideas about music itself are filtered through the top-down, symbolic process of modelling:

The need to model, Magnusson argues, is inevitably reductive because such numerical models are partial; complex and dynamic ideas such as ‘music’ are irreducible to a finite set of symbolic parameters. However, the interaction between tool design and conceptual ideas becomes more complicated when we consider that this process of reduction is not the only way in which designs and extra-technical ideas interact. Andrew Feenberg notes that reductions are accompanied or compensated for by further interventions from both designers and users in a process he calls ‘mediation’ (Feenberg Reference Feenberg1999: 206). These interventions are unconnected to the technical logic of the device but have to do with its suitability for particular social contexts, such as decisions about ‘look and feel’ or the range of functionality that is exposed. As such, the concept applies equally to the ways in our equipment is packaged and marketed, and subsequent ways in which users may customise or modify equipment, ranging from the cosmetic, such as taping over the Apple logo, to the more drastic and invasive, such as circumventing copy protection.

Feenberg argues that these mediations establish feedback loops with the processes of reduction in design, as certain interventions are liable to become enshrined over time as design principles, either explicitly through legislation and standardisation, or implicitly through convention. As such, the top-down design process takes on a less linear and more interconnected character as something embroiled in a complex set of historically and socially situated interactions, not only with ideas about the thing being modelled but also with ideas about the modelling of the thing. Furthermore, this process takes on a continuous and dynamic quality when interventions by users are considered. Designs are only ever contingently complete as they are always subject to modification in order to adjust their fit to sets of ideas outside of their technical logic.

2.2. The boundaries of the instrument

Mediating interventions on technologies also spill out into the world, over the boundaries of any particular device. Consider, for instance, the report by John Bowers that

Bowers’ design is not simply a matter of modelling musical preferences and devising more or less optimal ergonomics, but is also inflected by preferences about his social relationships in performance, which in turn affect the way in which resources are interconnected and located in space. As the use of physical space becomes part of the design, the conceptual boundary between instrument and environment is rendered more ambiguous.

This ambiguity has explanatory potential when considered in relation to two ideas from Feenberg and from Clark. Firstly, that just as the technological objects we encounter are very often systems of other technologies that have been interconnected and coupled, so usage also very often consists of assembling systems of interconnections and couplings. Feenberg argues for a degree of equivalence for these two forms of what he calls ‘systemisation’ (Feenberg Reference Feenberg1999: 205). This is particularly evident in computers, which are highly complex coalitions of interconnected and interdependent systems, but is also familiar to any user of audio technologies where much activity, both physical and virtual, consists of making assemblages of interconnected components. Secondly, systemising activity is more radically positioned by Clark as being integral to ways in which our cognitive processes can be seen to take place in continuous interaction between our brains and bodies and the wider environment. Clark argues that cognition involves continuous acts of ‘ecological assembly’ (Clark Reference Clark2008: 13), where external resources are enrolled into ad hoc networks in ways that distribute the cognitive burden of some task out into the world. Significantly, how technological or even how material a resource is does not appear to have much bearing on how likely it is to be involved in such a network, as Clark identifies a cognitive tendency to use whatever is at hand, resulting in complex mixtures of material resources, uses of space and organisation, and the deployment of symbols as ways of thinking through the world.

So, just as we can regard a computer as a system, it is also part of a larger ecosystem of potential interconnections some of which may be physical, some of which may be mediated by the ‘virtual’ activities of our brains. Furthermore, if Clark's account is true, then there is a case for regarding the systems we make and encounter around acoustic instruments as contingent unities in their own right, so that, under some conditions, systems such as instrument-score (and more complex extensions) can be considered as constituting the boundaries of the instrument. Thus the instrument becomes a term for describing the coalition of resources being used at a particular moment, as Franziska Schroeder and Pedro Rebelo observe:

Redrawing the boundaries of the instrument in this way means that the presence of symbolic mediations in our musical systems can be seen not to be an exclusive property of the digital, but is conspicuous throughout our environments and acts of assembly. With this in mind, we can examine the idea that symbolic forms can disrupt embodied relationships with the world.

2.3. Embodiment, symbols and computers

Magnusson suggests that ‘to work with symbolic tools means that one has to continually switch modes from focusing on the world to focusing on the tool with regular intervals and to a more pronounced degree than in acoustic instruments’ (Magnusson Reference Magnusson2009: 173). This continuous switching implies, for Magnusson, an interruption of our embodied inhabitation of the world that yields to a hermeneutic orientation, based on intellectual understanding. Consequently, he argues that the basis of skilled action developed around tacit knowledge that applies to acoustic instruments gives way to skill development through intellectual understanding for digital instruments. I will examine this in two parts, as it is a complex claim. In this section I will argue that with digital systems our interactions remain embodied. I then consider the implications of ‘understanding’ and skill formation in the following section.

If we follow Magnusson in accepting the role that tacit knowledge, gained through active exploration in the world, plays in skill formation more generally, and accept my claim that as a result of ecological assembly symbolic mediations in music are not exclusive to digital systems, then it remains to be considered whether there is something peculiar about the nature of computers or about the specific types of symbolic mediation employed that disturbs embodied engagement. Armstrong (Reference Armstrong2006) contributes a useful discussion on this point that identifies the switching of modes, described by Magnusson above, as being a continual feature of human cognition rather than an extraordinary symptom of computer usage.Footnote 2 In this sense, then, the switching of modes by itself does not appear to account sufficiently for a detachment of our engagement.

What Magnusson describes when he distinguishes between focus on the world and focus on the tool is encapsulated in Heidegger's concept of the ‘breakdown’ that occurs when performative flow is disrupted and the Otherness of the instrument becomes suddenly apparent (Armstrong Reference Armstrong2006: 64). Such an experience is obviously not limited to digital instruments, but are there grounds for treating it as an inherent property of the digital?

Initially, we can perhaps place different computer-orientated musical activities along a nominal continuum of embodiedness. We could, for example, consider performing EQ adjustments in real-time, selecting and refining parameters for offline processing, and non-live (undead?) coding as being progressively more distant from direct, bodily engagement with the world. EQ adjustment is uncontroversially embodied: the character of my adjustments is continually guided by what I hear. In the case of the more temporally interrupted process of applying, auditioning, refining and reapplying parameters, there would seem to be a case for this being characterised by more textual engagement. However, assessing what the various symbolic options on offer mean in context still happens out in the world, beyond the boundaries of the symbolic system. The possible interactions between the parameters of, say, a phase vocoder, and any particular sonic material are sufficiently complex that it is simply not possible to determine in advance without a lot of practice quite what the result might be.

Even when writing code, which would seem on the surface to be paradigmatically disembodied, it is possible to flow such that the separateness between myself and the machine seems to dissipate. When coding most fluently, I find that rather than simply implementing some structure that I have imagined in my head, what happens is between myself, the computer and the environment. I will play with space in the code as a way of helping me think more clearly about a particular bit; I will make use of doodles, sticky notes, changes of posture and musings out loud, as my work overspills the boundaries of the coding environment.

These very simple examples suggest that, as embodiment can happen with digital interfaces, the reasons for why breakdown occurs are reducible neither to the presence of symbolic mediations in the instrument, nor to an inherent property of digital interfaces. It is in more complex settings such as performance, however, that disruptions of flow are commonly experienced. Performing brings to bear networks of inter-related skills, many of which depend on being able to maintain focus on the world at large. Yet it is not reducible simply to an aggregation of skills, because the qualitative aspects of what we try to achieve when performing are inflected by social and cultural aspects of the environments in which we perform. The meaning of the symbolic aspects of an interface are in turn inflected by the wider context at hand. I argue in the following section that by situating our understanding of symbols in context they are subject to continual and contingent re-evaluation, and that the propensity for breakdown lies at the intersection of socially informed intentions and the technical affordances of the instrument.

2.4. Do you read me? Understanding

Discussions around the significance of particular symbolic forms – scores and texts – are well-established musicological concerns. Significantly, for the purposes of this discussion, the nature of such texts as authoritative, autonomous, self-contained carriers of meaning has been called into question in the musicology of recent decades (Cook and Everist Reference Cook and Everist1999), just as such notions have been rendered problematic for texts and symbolic forms more generally by currents like poststructuralism. Is there a reason to suppose that the symbolic forms that we encounter when we use digital tools are particularly different in this respect? I propose that there is not, and that this is made clearer when we adopt an analytic perspective that situates player and instrument into a rich social context. Two points arise from this. First, it becomes possible to view the process of enskilment with digital systems as involving the same degree of tacit knowledge and situated understanding as Magnusson rightly identifies with acoustic systems (Magnusson Reference Magnusson2009: 170). Second, we can reposition the affordances of systems (or their lack) as arising in the particularities of the social, material and historical context at hand. This helps move towards answering Magnusson's question at the start of this section in a way that sees such a ‘double aspect’ (Feenberg Reference Feenberg2002: 74–5) of computers as a property of any technology in situ, rather than of digital tools in particular.

One persuasive reason for regarding the symbolic forms used in digital interfaces as different and possibly less ambiguous than other texts is that they refer to technical and quantitive concepts that are simply there to be understood so that ‘ergonomically, the interaction happens primarily through a symbolic channel, which gradually teaches the user to operate with technical terms (such as “low-pass”)’ (Magnusson Reference Magnusson2009: 174). The meaning of a concept such as ‘low-pass’ is, of course, easy to describe in the abstract as a filtering operation on a signal that admits content below a certain frequency threshold whilst rejecting content above it. However, such a definition does not encompass the range of possible things that a person might understand the term to mean in practice. We can perhaps map this ambiguity on two broad fronts.

The first is that people bring with them historically contingent interpretations of concepts such as ‘low-pass’ that are shaded by lived experiences. That is, one's understanding of what a filtering operation is will be shaped by previous encounters with this and related concepts. We may understand filtering operations more or less solely in terms of our sonic experience of them, as memories and projections of their interactions with different types of sonic material, or we may understand them in more technical, conceptual terms such as specific types of circuit or difference equation. These are evidently schematic depictions, but sufficient to make the point that the epistemology of a concept such as ‘low-pass’ is at least partially accounted for by the histories of people's prior encounters.

Second, it follows that what we as individuals understand by such concepts changes with time and context. Most practitioners will understand the concept of ‘low-pass’ from multiple perspectives, both sonic and technical, and this understanding is subject to constant revision. Moreover, particular understandings are informed by context: what we are doing, and who we are doing it with. Consider, for instance, the range of particular meanings that the numeric properties of an equaliser may have. For a somebody new to such things, there is little or no experiential basis for associating these numbers – frequency, gain, resonance – with any particular aspect of what they are hearing. As such, the numbers demand attention as they learn, for example, how frequency as an aural phenomenon maps to frequency as a numeric range, helped or hindered to some extent by the clarity of their listening situation.

Such learning is necessarily an active, empirical affair. When we search around with an equaliser trying to locate some specific aspect of a sound, this can be likened to what Tim Ingold has called ‘wayfinding’, as distinct from what he calls ‘navigation’. The latter, Ingold argues, is characterised by knowing in advance one's destination, the former by knowing one's destination when one gets there (Ingold Reference Ingold2000: 235–7). This description of wayfinding characterises working with sound well, and applies particularly to skilled, expert use; one knows when the right point has been arrived at without necessarily being able to articulate how one knows. Skilled practitioners will home in more quickly as they have learned to interpret the response in the sound as a guide of where to go next, developing skills that are akin to Sudnow's characterisation of ‘grabbing’ for the right chords on the piano (Sudnow Reference Sudnow2001).

The distinction between wayfinding and navigation yields two additional, related points concerning the development of skill and the possibility of breakdown (see previous section). Ingold proposes a further distinction between mapping as an activity, and maps as artefacts. The former, he argues, can be considered as a ‘narrative re-enactment’ of our wayfinding (Ingold Reference Ingold2000: 234) – that is, as a learning process that arises through doing. The latter, on the other hand, is a guide to navigation: a set of instructions for getting from A to B.

Concerning enskilment, we can suggest that numerical parameters might initially serve as a map for novices, who are given out to seeking out advice in the form of particular parameter recipes to accomplish a particular task.Footnote 3 Experts, on the other hand, will be able to proceed intuitively, finding their way through the interaction of material and processor on the basis of learned, tacit skills (Dreyfus and Dreyfus Reference Dreyfus and Dreyfus1986). The parameters come to mean something quite different in each case. For novices, the numbers define the territory; the audible spectrum, for instance, becomes divided into pre-established zones where particular instrumental characteristics are taken to live (vocal presence at 5 kHz, for example). For experts, the relationship is reversed, as they have learned the skill of translating back from what is heard to a numeric representation.

On the second point, concerning breakdown, Ingold's distinction resonates with mapping in the sense of determining which external controls affect what parts of a digital algorithm. In particular, there is scope for contextualising critiques of overly simple and direct mapping strategies as yielding unsatisfactory experiences in terms of the degree to which such schemes afford only navigation rather than wayfinding, and thus militate against skilled development but also give rise to breakdown more frequently. By this measure, these problems are not intrinsic to the digital nature of our interfaces, but are related to particular choices by interface designers about what they expect their interfaces to mean to users. Direct parameter mappings of a granulator, for instance, are not especially musically suggestive, and are probably also quite opaque to someone unfamiliar with granulators. Fortunately, however, more wayfinding-orientated approaches to mapping are available (Bowers Reference Bowers2003; Van Nort and Wanderley Reference Van Nort and Wanderley2006).

Finally, our interpretation of symbolic channels is also inflected by context. Nominally technical concepts and their symbolic representations are enlarged by connotations specific to social and material circumstances. The meaning of numerical parameters is tied up with the complex interaction between a process and the signal it operates upon. For instance, what counts as a ‘broad’ resonance setting for a filter depends on the sonic material, the particularities of the filter being used, the job at hand (techno filter sweep or mastering?) and the propensity of the loudspeaker-room-listener system to pick up on those changes. Moreover, in discourse quite what these concepts are taken to mean is sensitive to what is being talked about, and the relationship between interlocutors, as indicated by Thomas Porcello's investigation of the different uses of technical and metaphorical language between experts and novices in recording studios (Porcello Reference Porcello2004).

The process of developing expertise with digital systems is not merely a case of coming to an understanding of a technical system qua technical system but of developing complexes of skills that loop through and around the instrument (some of which will take the form of tacit knowledge), because the use of symbolically mediated systems still hinges on bodies of tacit knowledge extrinsic to the device itself. This, as Clark points out, applies equally to symbolically mediated activities on computers (such as playing Tetris) and to playing piano (Clark Reference Clark2008: 75). Skills arise within ‘the whole system of relations constituted by the presence of the artisan in his or her environment’ (Ingold Reference Ingold2000: 291; Suchman Reference Suchman2007: 262), a system of relations that includes interaction with other agencies and with various traces of histories and cultures. Furthermore, the meaning of technical concepts encountered in technology arise, irrespective of their digitality, in this same system of relationships, as Feenberg says: ‘the social character of technology lies not in the logic of its inner workings, but in the relation of that logic to a social context’ (Feenberg Reference Feenberg2002: 79).

2.5. Active technologies

This brings us, finally, to the idea that digital instruments manifest some kind of active quality or agency where other technological paradigms do not, an argument made by both Magnusson (Reference Magnusson2009) and Bown, Eldridge and McCormack (Reference Bown, Eldridge and McCormack2009). Within the context of contemporary performance practices this is clearly a notion that needs accounting for, as various artists’ work involves systems that exhibit some degree of autonomy of operation, such as George Lewis’ Voyager (Lewis Reference Lewis1999, Reference Lewis2000), Arne Eigenfeldt's agent-based systems (Eigenfeldt Reference Eigenfeldt2008) and Agostino Di Scipio's Audible Ecosystemic Interface (Di Scipio Reference Di Scipio2003). However, for all that our digital instruments, as highly complex systemisations of technologies (see Section 2.2), may make this active nature more apparent, there are grounds for not regarding it solely as an attribute of digital technologies.

Considering first the condensed ecosystem of player–instrument, or agent–tool, relations, Ingold helps us appreciate that the material resources of un-digital practice are not altogether passive with a detailed account of wood-cutting in which the character of action can be seen to be in response to the very particularities of the wood and the consequences of prior acting (Ingold Reference Ingold2006). Perception and action, as Magnusson describes, ‘co-emerge through the agent's active involvement with the world’ (Magnusson Reference Magnusson2009: 169), rather than the process of making being one of mechanically executing an intention preformed in all its particulars (Ingold Reference Ingold2000: 294–311). This resonates with Bennett Hogg's depiction of ‘the violin as lie detector’ (Waters Reference Waters2007: 5), highlighting that the instrument is far from being passive, but is a participant in an ongoing negotiation.

We might, however, regard digital technologies as being especially active because, as Magnusson notes, they seem to embody knowledge in such an explicit manner. Computers are, by this account, tools for automation, things to which we delegate tasks in formalised terms. Consequentially, they exhibit agency by apparently making skilled human action peripheral where once it was central; hence the concern expressed by Vaughan that the transition from analogue tape-orientated studios to digital-workstation-orientated studios could herald a process of deskilling (Vaughan Reference Vaughan1994). However, Ingold provides an additional insight when he notes that ‘as fast as machines have been contrived to do what had previously been done by skilled hands, different skills have sprung up to cope with the machines themselves’ (Ingold Reference Ingold2000: 332). Again, this can be seen to apply equally to digital and nondigital technologies. Just as delegating the role of co-player to a computer, as Lewis does with Voyager, will entail human players developing skills to cope with the computer-player, so the delegation of remembering certain features of a musical work to scores or recordings (Frith Reference Frith1996: 226–7) opened up new realms of skilled action.

Finally, we may question whether the propensity of digital systems to ‘script’ action really is a special property of a given device or class of device when we consider the instrument as an assemblage of material and immaterial resources rather than as a single artefact. ‘Agencies – and associated accountabilities – reside neither in us nor in our artefacts but in our intra-actions’, writes Lucy Suchman (Reference Suchman2007: 285), so that when technologies seem to script our action, this scripting is as much a social matter as a technical one. The apparent imperatives of given technologies exist as matters of convention and consent in the same way that cultural codes of behaviour, dress, language and music arise through the co-dwelling of humans in environments shaped and filled by artefacts. It follows that to withdraw consent, to disrupt apparent imperatives is also a socially inflected matter rather than a purely individual action.

When it seems that our digital systems are imposing upon us restrictive ‘cultural specifications’ (Magnusson Reference Magnusson2009: 175) then this can be understood more fully as arising in a particular complex of social, historical and material circumstances than as an inherent property of digitality. It allows us to acknowledge, for instance, that those things we are trying to achieve that the computer appears to be ‘making’ difficult are socially constituted aims, whether that be virtuosity (Jordà Reference Jordà2004), intimacy (Cook Reference Cook2004; Plans Casal Reference Plans Casal2008) or togetherness (Schroeder, Renaud, Rebelo and Gualdas Reference Schroeder, Renaud, Rebelo and Gualdas2007). Cultural specifications and, it follows, affordances arise in the fold between our socially and historically ensconced priorities and the social-technical histories of our machines, themselves reflective of particular priorities. Affordances, therefore, are partly constituted by convention and by consent. Consent can be withdrawn, of course, and new affordances are found by disrupting what is taken to be correct usage or by trying to achieve something apparently perverse – such as when Phil Archer deliberately overloads his CPU to create variations of current on a USB bus that is controlling a slide guitar made from eviscerated CD players (Archer Reference Archer2006: 24–5). This noncompliance is socially inflected in the same manner; apparent perversity is relative to socially situated expectation, and affordances for disruption emerge more readily in a social milieu that sanctions such things.

2.6. Summary

The epistemic features that seem to distinguish the digital from the acoustic when viewed from the close-up vantage point of player-instrument dissipate somewhat when we zoom out and consider players and instruments as situated in an ecosystem of connections and histories.

1. 1. ‘Top-down’ and ‘bottom-up’ design practices are less readily distinguishable, as a complex web of different types of conceptual framework that relate and interact in various ways can be seen to feature both in both digital and acoustic making.

2. 2. Artefacts are systems within systems, and can readily incorporate disparate mixtures of the high- and low-tech, and material and immaterial, resources, such as symbols and ideas. The practical boundaries of instruments, in such a view, become contingent and dynamic, prompting the suggestion that such configurations as violin-plus-score may be considered as an operational unity that presents symbolic mediations, like a digital instrument.

3. 3. Symbolic mediations are not unique to digital systems, and do not, on their own, account for breakdowns of embodied flow.

4. 4. The symbolic components of performance ecosystems are not stable channels of fixed or unambiguous meaning, but are subject to contingent interpretation in practice. Digital instruments, like acoustic instruments, are interacted with on the basis of mixtures of tacit and formal knowledge.

5. 5. The propensity for agency of digital instruments is also present in the acoustic realm, both through the active nature of working with raw materials and acoustic instruments and through the way that both opportunities and foreclosures of creative potential emerge in the context of particular social, material, historical and political systems.

It is worth emphasising again that arguing against the necessity of the difference between acoustic and digital instruments is not an attempt to negate or otherwise devalue the experiences of such difference. Rather it is an attempt to relocate the explanation for these experiences to a wider context with, I believe, more explanatory potential, and to propose that where digital instruments might seem particularly potent in their ‘cultural specification’, this is a difference of degree that arises in particular configurations of people, materials, symbols and histories.

That our relationships are spread out to such an extent across the worlds we inhabit serves to remind us that our practice is similarly diffused, something in Ingold's terms in which we dwell (Ingold Reference Ingold2000). When we come to perform and design digital instruments we bring to bear a range of cultural, social and bodily competencies beyond the technical ability to communicate with the machine on ‘its’ terms. Where there do seem to be significant differences in the musical experience between digital and acoustic practices then, in addition to Magnusson's call for designers to take seriously their potential in establishing cultural horizons, it is also a matter for practitioners to communicate about, both between themselves and more widely. Specifically, there may be fruitful discussion to be had on the kinds of conjunctions of skills, priorities and preferences we find ourselves bringing to bear, and the kinds of coping strategies we develop.