Improvisation

All about learning - conscious? unconscious? What's the difference? Does it matter?

Previously I quoted from Nick Ellis's (2011) chapter: Implicit and Explicit SLA and Their Interface, which is found in Sanz, C., & Leow, R. P. (Eds), Implicit and explicit language learning: conditions, processes, and knowledge in SLA and bilingualism. Washington, DC: Georgetown University Press.

Ellis is one of the few writers who not only distinguishes between the two types of learning process (implicit and explicit) but also tackles the issue of what role each process plays in learning. I began by quoting his distinctions:

• Implicit learning is acquisition of knowledge about the underlying structure of a complex stimulus environment by a process that takes place naturally, simply, and without conscious operations (p.38)
• Explicit learning is a more conscious operation, where the individual attends to particular aspects of the stimulus array and volunteers and tests hypotheses in a search for structure (p.38)

But then proposed my own assertion that implicit (unconscious) acquisition of the rules necessary for improvisation was not sufficient for improvisation. Why not? If such processes are sufficient for learning the complexities of our mother tongue (L1), why not for music? Such a comparison is, after all, valid if we believe that the ideal improviser is one who can 'speak' music in the same way that one speaks one's birth language. Would implicit (unconscious) acquisition of the rules then be sufficient if we learnt to improvise in the same way as we are taught to speak a birth language? To reconstruct these conditions would be an interesting experiment! It would mean:

• thousands of hours of one-to-one instruction.
• the use of music to communicate one's basic emotional and physical needs.
• the acquisition of skill during the years of greatest neurological plasticity and cognitive growth, i.e., one would learn to express oneself in music in parallel to the processes of adapting to and understanding the surrounding social and physical environment.

It's relevant I think to note that L1 acquisition is not as implicit as it is often implied. Explicit instruction need not be restricted to the teaching of abstract rules and theory, but can also apply to forms of guidance in which the individual receives 'explicit' (i.e., clear) feedback, guidance and correction. Successful learning could hardly occur through meer passivity, as Ellis remarks:

Whatever the domain, learning the patterns, regularities or underlying concepts of a complex problem space or stimulus environment with explicit instruction, direction, and advances clues, heuristics, or organisers is always better than learning without any cues at all. Ellis (2011, p.38)

It's my personal conviction that musical improvisation could be learnt in the same fashion as L1 under the optimal conditions outlined above. Quite possibly, improvisers who learn the skill of improvisation also benefit from some of these conditions - in particular the learning of improvisatory skills alongside other musical skill formations, and the development of these skills during periods of general cognitive and personal development. In a sense, young improvisers in common with L1 learners are equipped with a mental white sheet, whereas classically-trained performers have to adapt their existing systems of musical perception to the new ones of improvisation, in common with L2 learners. In this, second case, the contexts of learning (of both music and language) are usually not comparable, calling for a different approach to learning, as Ellis (2011) describes: 

In contrast to the newborn infant, the L2 learner’s neocortex has already been tuned to the L1, incremental learning has slowly committed it to a particular configuration, and it has reached a point of entrenchment where the L2 is perceived through mechanisms optimised for the L1. Thus the L1 implicit representations conspire in a “learned attention” to language and in the automatised processing of the L2 in nonoptimal, L1-tuned ways. (Ellis, 2011, p.40)

Let's look more closely at the contrasting nature of these two types of knowledge could be applied to learning situations, using the insights of Ellis (2011), Hiebert (1986), Carpenter (1986) and others.

EXPLICIT LEARNING.

• Ellis (2011, p.38). When the material to be learned is simple, ... and the critical features are salient, then learners gain from being told to adopt an explicit mode of learning where hypotheses are to be explicitly generated and tested and the model of the system updated accordingly. As a result they are also able to verbalise this knowledge and transfer to novel situations. 

• Hiebert & Lefevre (1986, p.3-4). Conceptual knowledge is characterised most clearly as knowledge that is rich in relationships. It can be thought of as a connected web of knowledge, a network, in which the linking relationships are as prominent as the discrete pieces of information. Relationships pervade the individual facts and propositions so that all pieces of information are linked to some network. In fact, a unit of conceptual knowledge cannot be an isolated piece of information; by definition it is a part of conceptual knowledge only if the holder recognises its relationship to other pieces of information.

• Ellis (2011, p.41). ... conscious processing is spread wide over the brain and unifies otherwise disparate areas in a synchronised focus of activity. 

• Ellis (2011, p.41). Conscious activity affords much more scope for focused long-range association and influence than does implicit learning. It brings about a whole new level of potential associations.

• Ellis (2011, p.42). Compared with the vast number of unconscious neural processes happening in any given moment, conscious capacity evidences a very narrow bottleneck. But the narrow limits of consciousness have a compensating advantage: consciousness seems to act as a gateway, creating access to essentially any part of the nervous system. ... It is a facility for accessing, disseminating, and exchanging information and for exercising global coordination and control: consciousness is the interface.

• Ellis (2011, p.43) ...where L1 experience has tuned the learner’s attention elsewhere. ... Instruction is thus targeted at increasing the salience of commonly ignored features by first pointing them out and explaining their structure and, second, by providing meaningful input that contains many instances of the same grammatical meaning-form relationship. 

• Ellis (2011, p.45).Consciousness creates access: its contents are broadcast throughout the brain to the vast array of our unconscious sources of knowledge, and by these means, consciousness is the interface.

IMPLICIT LEARNING.

• Ellis (2011, p.35): The acquisition of L1 grammar is implicit and is extracted from experience of usage rather than from explicit rules - simple exposure to normal linguistic input suffices, and no explicit instruction is needed.

• Ellis (2011, p.38): When the material to be learned is more randomly structured with a large number of variables and when the important relationships are not obvious, then explicit instructions only interfere and implicit mode of learning is more effective. This learning is instance-based but, with sufficient exemplars, an implicit understanding of the structure will be achieved.

• Ellis (2011, p.39): Implicit learning collates the evidence of language, and the results of this tallying provide an optimal solution to the problem space of form-function mappings and their contextualised use, with representational systems modularising over thousands of hours on task.

• Ellis, 2011, p.41). ... implicit learning occurs largely within modality and involves the priming, or chunking, of representations or routines within a module; it is the means of tuning our zombie agents, the menagerie of sepcialised sensorimotor processors ... that carry out routine operations in the absence of direct conscious sensation or control.
• Ellis (2011, p.41). ... our implicit systems automatically process the input, allowing our conscious selves to concentrate on the meaning rather than the form. Implicit, habitual processes are highly adaptive in predictable situations. But the more novelty we encounter, the more the involvement of consciousness is needed for successful learning and problem solving

THE ROLE OF CONSCIOUS KNOWLEDGE IN LEARNING. One thing that emerges from these useful definitions is that the role of conscious knowledge (declarative knowledge, conscious awareness or conscious control) becomes more circumscribed. It's clear that, although Hiebert & Lefevre (1986) describe conscious knowledge as 'rich in relationships' and 'a connected web of knowledge' these connections are conceptually different to the kind of multitudinous connections calculated by implicit knowledge, which, over time - Ellis mentions 'thousands of hours on task' - furnishes us with instincts for what is appropriate or fitting in form and gesture. There is so much of skilled learning and performance which cannot be ascribed to consciousness: most importantly (i) I cannot consciously control or train my muscles, (ii) I cannot consciously control, generate or even fully understand the functioning of the music I improvise in real time, (iii) I cannot consciously understand the processes of automatisation which are so essential to fluency, (iv) the very nature of automatisation is that it reduces demands on consciousness and allows the performance of parallel rather than serial processes.

Yet, Ellis (2011) claims that implicit learning is itself inadequate for the learning of a new language, due to effects of interference arising from already existing skills in the native language. In what way do the existing language skills interfere with the acquisition of new skills? Can these effects be comparable to learning improvisation? 

Problems seem to arise when existing language skills prevent the learner from noticing what is important (salient) in the new material: '...where L1 experience has tuned the learner’s attention elsewhere' (Ellis, 2011, p.43). Why should this occur? Presumably because individuals try to understand or process the new language in similar ways to that in which they habitually process or understand their native tongue. This feature alone I think is directly replicable in the learning of improvisation (= the acquisition of novel generative skills) by musicians trained in performance (= possession of interpretative skills). The fact that music is not learned implicitly in an L1 fashion should not disrupt the comparison because, for trained musicians, the initial declarative methods of learning music will have been long transformed into automatic procedures. (In other words, the trained and fluent performer possesses an intact and specialised knowledge base sufficient for the task of performance, and it is this specialised and intact knowledge base which is brought to the task of learning improvisation.) Of course, when we have great familiarity with a task we don't need to analyse the techniques we use to perform it, not do we think of the constituent parts of the task, or in what way these parts are combined to make greater structures. The step-by-step learning of our native tongue, so embedded in communicative needs and direct feedback, is long since transformed into unconscious performance; likewise, the hard-won technical mastery of a musical instrument and the skills of meticulous score-reading and interpretation are second nature to the trained performer. How then should a musician with such a training look anew at a score and understand it as a model, say, for improvisation? As Ellis suggests, ',,, our zombie agents, the menagerie of specialised sensorimotor processors' which characterise implicit learning are inadequate for these initial stages of learning, powerful instruments though they are. The processes of unconscious learning must be steered in a certain direction by consciousness, given specific goals to head for, and have these goals reiterated through a clear framework of instruction: 'Instruction is thus targeted at increasing the salience of commonly ignored features by first pointing them out and explaining their structure and, second, by providing meaningful input that contains many instances of the same grammatical meaning-form relationship (Ellis, 2011, p.43). 

In this way, consciousness, though restricted as a tool for the actual performance of improvisation, and inadequate in terms of perceiving intrinsic matters of form in the material (whether music or language) through experience, can be seen to play an essential, leading role in the early stages of adult learning. How does it do this? Ellis suggests that there is no part of the nervous system to which consciousness does not have 'access'. Through this right of access then, consciousness can act as an interface between disparate, automatic systems, as a force which coordinates, sychronises and ultimately exercises 'global coordination and control' (p.42).

PROCESS PURE: Are the two types of knowledge really distinct? Much of the research giving rise to these distinctions comes from tests on amnesiac patients (Green & Shanks, 1993, p.304): 'Although amnesics are severely impaired on tasks of explicit memory, their performance is often within the normal range on implicit memory tasks. Normal controls, as well as amnesics, exhibit this independence of performance on implicit and explicit memory tasks.' The same authors also warn that, with regard to implicit memory (recall of unconsciously learnt material) '... individual memory tasks are not process-pure. In other words, a single task may involve the contribution of more than one type of memory process. A given task, for example, may involve any combination of conscious and automatic processes, such as those that are conceptually and data driven (Green & Shanks, 1993, p.314).

Hiebert & Lefevre (1986), also warn that '... the distinction provides a classification scheme into which all knowledge can or should be stored. Not all knowledge can be usefully described as either conceptual or procedural. Some knowledge, seems to be a little of both, and some knowledge seems to be neither' (p.3).