100% found this document useful (1 vote)

169 views58 pages

Pressing - Improvisation - Methods and Models

This chapter provides a monumental survey of concepts from a number of disciplines that are relevant to an understanding of improvisation, and outlines a theory of improvisation that concentrates on the psychological bases for moment-to-moment choices within an improvisational structure. The theory is associationistic in nature, and proposes that improvisation proceeds by means of choices of elements which either continue or interrupt some aspect of the immediately preceding context.

Uploaded by

Daniel De León González

We take content rights seriously. If you suspect this is your content, claim it here.

Available Formats

Download as PDF, TXT or read online on Scribd

100% found this document useful (1 vote)

169 views58 pages

Pressing - Improvisation - Methods and Models

Uploaded by

Daniel De León González

We take content rights seriously. If you suspect this is your content, claim it here.

Available Formats

Download as PDF, TXT or read online on Scribd

You are on page 1/ 58

Improvisation: methods and models

University Press Scholarship Online

Oxford Scholarship Online

Generative Processes in Music: The Psychology of

Performance, Improvisation, and Composition
John Sloboda

Print publication date: 2001

Print ISBN-13: 9780198508465
Published to Oxford Scholarship Online: March 2012
DOI: 10.1093/acprof:oso/9780198508465.001.0001

Improvisation: methods and models

Jeff Pressing

DOI:10.1093/acprof:oso/9780198508465.003.0007

Abstract and Keywords

This chapter provides a monumental survey of concepts from a number of
disciplines that are relevant to an understanding of improvisation, and outlines a
theory of improvisation that concentrates on the psychological bases for
moment-to-moment choices within an improvisational structure. The theory is
associationistic in nature, and proposes that improvisation proceeds by means of
choices of elements which either continue or interrupt some aspect of the
immediately preceding context. It deals with the note-by-note structure of an
improvisation.

Keywords: improvisation, psychology, neuropsychology, creativity

Introduction
How do people improvise? How is improvisational skill learned and taught?
These questions are the subject of this chapter. They are difficult questions, for
behind them are long-standing philosophical quandaries such as the origins of
novelty and the nature of expertise, which trouble psychologists and artificial
intelligence workers today almost as much as they did Plato and Socrates in the
fourth and fifth centuries BC.

Page 1 of 58

PRINTED FROM OXFORD SCHOLARSHIP ONLINE (www.oxfordscholarship.com). (c) Copyright Oxford University Press, 2019. All
Rights Reserved. Under the terms of the licence agreement, an individual user may print out a PDF of a single chapter of a
monograph in OSO for personal use (for details see www.oxfordscholarship.com/page/privacy-policy). Subscriber:
UNIVERSIDAD NACIONAL AUTONOMA DE MEXICO; date: 13 April 2019
Improvisation: methods and models

In a previous article (Pressing 1984a) I summarized a number of general

properties of the improvisation process on the basis of the diverse historical
writings of artists, teachers, and musicologists. This material was integrated
with precepts from cognitive psychology to sketch out the beginnings of a
general theory of improvisation.

In this article a much more explicit cognitive formulation is presented, the first
proper (though by no means necessarily correct) theory of improvised behaviour
in music. The building of this theory has required input from many disparate
fields with which the general musical reader may not be familiar. For this reason
I begin with the survey of appropriate background research and its relation to
improvisation. Some of these areas may initially seem distant from the topic at
hand.

A survey of pertinent research

Some physiology and neuropsychology
Although our state of knowledge in these areas is far too meagre to have any
definite repercussions for improvisation, there are a few facts which are at least
strongly suggestive.

To begin with, improvisation (or any type of music performance) includes the
folowing components, roughly in the following order:

(p.130) (1) Complex electrochemical signals are passed between parts

of the nervous system and on to endocrine and muscle systems;
(2) muscles, bones, and connective tissues execute a complex sequence of
actions;
(3) rapid visual, tactile, and proprioceptive monitoring of actions takes
place;
(4) music is produced by the instrument or voice;
(5) self-produced sounds, and other auditory input, are sensed;
(6) sensed sounds are set into cognitive representations and evaluated as
music;
(7) further cognitive processing in the central nervous system generates
the design of the next action sequence and triggers it.

– return to step (1) and repeat –

It seems apparent that the most starkly drawn distinctions between

improvisation and fixed performance lie in steps (6) and (7), with possibly
important differences in step (3). This chapter therefore inevitably focuses on
these aspects.

Page 2 of 58

The given steps are often collapsed into a three-component information-

processing model of human behaviour which has ready physiological analogies:
input (sense organs), processing and decision-making (central nervous system,
abbreviated CNS), and motor output (muscle systems and glands).

Control of movement by the CNS is complex: the cerebral cortex sends signals to
both the cerebellum and the basal ganglia, which process the information and
send a new set of signals back to the motor cortex. The brainstem nuclei are also
involved in details of motor co-ordination. It has been suggested that the basal
ganglia and cerebellum have complementary roles, with the basal ganglia
initiating and controlling slow movements while the cerebellum is active in the
co-ordination of fast, ballistic movements (Sage 1977).

Motor signals from the cortex pass to the spinal cord and motor nuclei of the
cranial nerves via two separate channels: the pyramidal and extrapyramidal
systems. These two nerve tracts illustrate the simultaneously hierarchical and
parallel-processing aspects of CNS control, for they run in parallel but
interconnect at all main levels: cortex, brainstem, and spinal cord. Hence while
each tract has some separate functions there is a redundancy that can be used
to facilitate error correction and motor refinement. Similar redundancy and
parallel processing is found at lower levels of motor control. Alpha-gamma
coactivation, for example, describes the partial redundancy of neural
information sent to two distinct types of motoneurons, alpha and gamma, whose
axons and collaterals terminate on the main skeletal muscles and the intrafusal
muscle fibres, respectively.

The organization of behaviour has often been linked with the existence (p.131)
of motor action units (or equivalent concepts), and their aggregation into long
chains to develop more complex movements. The validity of the concept of motor
action units can be seen mirrored physiologically in the existence of command
neurons, single nerve cells in invertebrates whose activation alone suffices to
elicit a recognizable fragment of behaviour. The effect is achieved by excitation
and/or inhibition of a constellation of motoneurons (Bentley and Konishi 1978;
Shepherd 1983). While there are no known single cells that fully trigger complex
behaviour in mammals, populations of neurons in the brains of higher animals
are strongly suspected of serving a similar function (Beatty 1975). It is therefore
possible to speculate that skilled improvisers would, through practice, develop
general patterns of neural connections specific to improvisational motor control.

Page 3 of 58

Finally, it is of interest that neurological correlates have recently been

discovered for a division of knowledge and memory into two separate categories:
declarative and procedural. A degree of independence of these two types of
memory (for facts or procedures) has been reported among amnesic and post-
encephalitic patients for some time (for example Milner 1962; Brooks and
Baddeley 1976). Typically, patients can not remember new facts, but are able to
learn new motor skills over a period of time, yet without any awareness on
successive days of having performed the tasks before. In recent studies, Cohen
(1981) and Cohen and Squire (Cohen (1981) have shown that declarative
learning is linked to specific diencephalic and bitemporal brain structures.
Unaware of this work, I drew a related distinction in a recent paper (Pressing
1984a) between object and process memory, based on the rehearsal strategies of
improvising musicians. As Squire (1982) has pointed out, there are parallel
distinctions in earlier writings; artificial intelligence (Winograd 1975), knowing
how and knowing that (Ryle 1949), habit memory and pure memory (Bergson
1910), and memory with or without record (Bruner 1969). What is suggestive
about these correlations is that physiological locations for some specific
cognitive skills used in improvisation might very well exist.

Motor control and skilled performance

This area traditionally has centered around industrial skills, sport, typing,
handwriting, specially designed laboratory tasks like tracking, and to a lesser
degree music. It is a complex field of considerable relevance to improvisation,
even though improvisation per se is scarcely mentioned. Therefore I first review
general theories of motor control, and then delve into a number of special issues
in skilled performance and skill development that are relevant here.

(p.132) Theories of motor control and skill

The starting point for nearly all the existing theories is the three-stage
information-processing model mentioned earlier, based on sensory input,
cognitive processing, and motor output. To this must be added the notion of
feedback (auditory, visual, tactile, or proprioceptive). Traditional ‘open-loop’
theories include no feedback, and hence no mechanisms for error correction. In
its starkest form this theory is clearly inappropriate for improvisation; however,
there is persistent evidence, dating back to the medical work of Lashley (1917),
and including studies of insect behaviour and de-afferentiation techniques in
monkeys that points to the existence of motor programmes that can run off
actions in open-loop fashion.

Page 4 of 58

In contrast stand ‘closed-loop’ theories, which contain feedback, and hence

allow for the intuitively natural possibilities of error detection and correction.
The closed-loop negative feedback (CLNF) model is one of the oldest. In this
model the feedback (primarily auditory in the case of musical improvisation) is
sent back to an earlier stage in the control system which compares actual output
with intended output, producing a correction based on the difference between
the two (see for example Bernstein 1967). Such closed-loop models have their
historical roots in engineering models of servomechanisms, control theory, and
cybernetics.

A wide variety of closed-loop formulations has been given. Gel’fand and Tsetlin
(1962, 1971) used a mathematical minimization procedure to model the
cognitive search for appropriate motor behaviour. Pew and Baron (1978)
sketched out a theory of skilled performance based on optimum control (see also
Kleinman et al. 1971). Powers (1973) proposed a hierarchy of motor control
systems whereby the correction procedures of higher-order control systems
constitute reference signals for lower-order systems. Another hierarchical model
was given by Pew (1974), in which specific single movements are combined into
sequences, and ultimately into various subroutines that make up goal-directed
action. Actions are then organized and initiated by an executive programme
(Fitts 1964). As is apparent, many such hierarchy theories are based on the
application of computer programming principles (see Miller et al. 1960).

These ideas offer a more sophisticated understanding of motor behaviour, but

they have serious limitations. They model motor learning either poorly or not at
all, and are not based on empirical findings about human actions (Adams 1961).
A closed-loop theory of motor learning was proposed by Adams (1971, 1976) in
an attempt to rectify some of these problems. In this theory there are ‘memory
traces’ which select and initiate movements and ‘perceptual traces’ which are
representations of the intended movements, and are used as templates for error
correction. A perceptual trace is gradually built up by repeated practice from
feedback, knowledge of (p.133) results (often abbreviated KR), and error
correction. Eventually the perceptual trace can function as an internalized goal,
diminishing dependence on the externally based knowledge of results (Namikas
1983). Hence open-loop control characteristics are not completely excluded.

By the late 1970s the consensus was that both open- and closed-loop control
must occur in skilled performance (Keele and Summers 1976; Delcomyn 1980;
Paillard 1980; see Summers1981 for a review). That is, movements are both
centrally stored as motor programmes, and susceptible to tuning (adjustment)
on the basis of feedback. Coupled with the well-established concept of flexibility
characteristic of skilled (but not rote) performance (Welford 1976), this
promoted approaches based on more abstract programming notions that
brought the field closer to artificial intelligence (and made it more germane to
improvisation).

Page 5 of 58

In this spirit Schmidt (1975, 1976) proposed a theory of motor schemata that
models both recall and recognition. The schema is considered to contain the
general characteristics of a movement which must be organized in any given
situation to satisfy environmental requirements and the goals of the performer.
Context then guides the production of a series of motor commands that
ultimately generate a spatiotemporal pattern of muscle actions. Feedback is
based on a template-comparison idea.

Because schemata are not specific movement instructions but are ‘generalized’
motor programmes, this theory is capable of modelling novelty (at least in a very
general way), which the others above could not (except Pew 1974, which also
uses a schema notion). The possibility of novelty is also catered to by Schmidt’s
inclusion of degree of variability of practice conditions as one determiner of
schema ‘strength’. At its core, the ‘novelty problem’ is very close to that of
improvisation.

Similar to schemata is the notion of action plan. Miller et al (1960) gave a

general description of plans, while Clark and Clark (1977) described plans for
language discourse, and Sloboda (1982) and Shaffer (1980, 1981, 1984)
specified plans for playing music. As discussed by Shaffer (1980), a plan is an
abstract homomorphism representing the essential structure of the performance
and allowing finer details to be generated or located as they are needed during
execution.

Other related theories include Allport’s proposal of a system of condition-action

units which are links between sensory calling patterns and categories of action
(Allport 1980). Also related are adjustable control or description structures for
artificial intelligence such as frames and scripts-(see below).

Page 6 of 58

This convergence of theory is useful in constructing a model of improvisation

(see below). However, it remains rather unspecific, and has run far ahead of
experiment. But as of this writing there seems only one alternative in the area of
motor behaviour. This is the organizational invariant approach of Turvey, Kugler,
Kelso, and others (Turvey 1977; (p.134) Kugler et al. 1980; see Kelso 1982 for
further references). This approach draws on two sources: the ecological
perspective of Gibson (1966, 1979) and the dissipative structure model of non-
equitibrum thermodynamics (Prigogine 1967; Prigogine and Nicolis 1971).
Essentially the theory de-emphasizes notions of cognitive process and control,
replacing them with, in so far as is possible, ‘organization invariants’. These
organization invariants are characteristic constraint structures that allow the
emergenceof specific spatial relationships and dynamic processes in the
behaviour of non-linear systems when the parameters controlling these systems
fall in certain critical ranges. Thus if the human motor action apparatus is
considered to be (as it certainly is) a non-linear system, characteristic properties
of muscle groups and patterns of human limb co-ordination will naturally
emerge from the constraints imposed by a given task situation (Kelso et al. (1981;
Saltzman and Kelso 1983). The proposals are exciting, but their ultimate fate
remains unclear. The theory is still being formulated, and comparable ideas from
non-linear mathematics have infiltrated many fields in the last 10 years, with
uneven results.

Organizational invariant theory seems also likely to apply primarily to the

dynamics of motor programme execution, and not to the formulation of
intentions and high-level decision-making (Wilberg 1983). Since these functions
are vital elements in improvisation in any but an extreme mechanistic approach,
the theory as it stands is not particularly suitable for improvisation modelling.
Nevertheless, these ideas may be used in an understanding of the sources of
behavioural novelty, and are discussed further below.

Some special issues relevant to improvisation

Skill classification

Various dimensions of skill classification have been proposed and improvisation

can be placed within these. Two possible categories are ‘open’ skills, which
require extensive interaction with external stimuli, and ‘closed’ skills, which may
be run off without reference to the environment (Poulton 1957). Solo
improvisation is basically a closed skill, as it relies only on self-produced stimuli,
whereas ensemble improvisation is more open. Other dimensions of skill
classification are gross-fine, discrete-serial-continuous, complex-simple, and
perceptual-motor (Holding 1981). Improvisation is a fine, complex skill, with
both perceptual and motor components; continuous actions predominate,
although there are also discrete and serial motor aspects. This last point varies
somewhat with the nature of the instrument played.

Page 7 of 58

It is important to also emphasize the contrast between unskilled and highly

skilled performance. A vast majority of reported skill studies treat simple motor
tasks like tracking, under an implicity reductionistic scientific methodology. It is
increasingly acknowledged, however, that highly developed skills have
distinctive emergent properties missed in these (p.135) earlier short-term
studies, properties such as adaptability, efficiency, fluency, flexibility, and
expressiveness (Welford 1976; Shaffer 1980; Sparrow 1983). These are vital
components of improvisatory skill.

Feedback and error correction

Feedback is a vital component in improvisation for it enables error correction

and adaptation—a narrowing of the gap between intended and actual motor and
musical effects. But feedback is also important for its motivational (Gibbs and
Brown 1956) and attention-focusing effects (Pressing 1984a).

Feedback redundancy is an important concept for music. Aural, visual,

proprioceptive, and touch feedback reinforce each other for the instrumental
improviser, whereas the vocalist has only hearing and proprioception available
(Pressing 1984a). Likewise the design of some instruments allows more precise
visual feedback and more categorical kinaesthetic feedback than others. This is
almost certainly why sophisticated improvisation using advanced pitch materials
is more difficult on the violin than the piano, and extremely challenging for the
vocalist. For every first-rate scat-singer in the world there must be 500 talented
jazz saxophonists.

Feedback can also be considered to operate over different time scales. Thus
short-term feedback guides ongoing movements, while longer term feedback is
used in decision-making and response selection. Still longer term feedback
exists in the form of knowledge of results (KR) for skills where external
evaluation is present or result perception is not sufficiently precise or
immediate. The importance of this for improvisation has been demonstrated by
Partchey (1974), who compared the effects of feedback, models, and repetition
on students’ ability to improvise melodies. Feedback, in the form of playbacks of
recordings of the students’ own improvisations, was clearly superior to listening
to pre-composed model melodies, or repetition, as an improvisation learning
technique. In group improvisation, feedback loops would also operate between
performers (Pressing 1980).

Page 8 of 58

In view of the inter-connectedness of the parts of the central nervous system, it

is also clear that there exist internal feedback (and feedforward) loops not based
on sensory processing (Brooks 1978). That is, if higher cognitive levels set the
design of motor programmes while movement fine structure is specified in
closed-loop fashion by lower levels of the CNS, notably the spinal cord, then
copies of these lower level motor instructions are almost certainly sent directly
back up to higher centres. In other words, there is some kind of central
monitoring of efference. This would serve to increase overall processing speed
and accuracy.

The role of errors in improvisation has been discussed previously (Pressing

1984a). It will simply be pointed out here that errors may accrue at all stages of
the human information processing system: perception, movement/musical
gesture selection and design, and execution. Minor (p.136) errors typically
demand no compensation in following actions, whereas major errors typically do.

Anticipation, pres-election, and feedforward

These three concepts have to do with preparation for acton. Physiological

recording of the Bereitschaft potential (BP) and contingent negative variation
(CNV) (see Brunia 1980) now provides explicit support for the long-standing idea
that higher cognitive control centres bias lower ones towards anticipated
movements. This is therefore a type of feedforward, and has been described
from various perspectives: spinal ‘tuning’ (Turvey 1977; Easton 1978), corollary
discharge or efference copy (von Holst 1954), and pre-selection (see Kelso and
Wallace 1978 for discussion).

The idea of preparation is very important for improvisation, where real-time

cognitive processing is often pushed up near its attentional limits. It can be
formally proved, for example, that only a control system with a model of
disturbances and predictive power can become error free (Kickert et al. (1978).
For improvised performance that aims at artistic presentation, where
discrepancies between intention and result must be kept within strict bounds,
practice must attempt to explore the full range of possible motor actions and
musical effects, to enable both finer control and the internal modeling of
discrepancies and correction procedures, including feedforward.

Hierarchy vs. heterarchy

Page 9 of 58

Because of influences of the physical sciences and control theory, an

overwhelming majority of models for motor behaviour have used a hierarchical
control system. However, the inter-connectedness between difference locations
in the CMS and the many documented types of feedack and feedforward
processes mentioned above argue that this perspective is probably too narrow.
Furthermore, explicit parallel-processing possibilities exist due to the separate
pyramidal and extrapyramidal neural tracts, alpha-gamma coactivation, etc., as
mentioned above. Hence other types of organization, referred to as heterarchical
or coalition, have been proposed (McCulloch 1945; von Foerster 1960; Greene
1972; Turvey 1977). In this perspective, executive control of the system may be
transferred between different ‘levels’ depending on the needs of the situation
(Miller et al. 1960).

Time scales for the control of movement

This is a subject with an enormous and complex literature. For background

purposes in modelling improvisation a few points only seem sufficient.

Actual neural transmission times are on the order of tens of milliseconds.

According to Davis (1957; see also Sage 1977), auditory stimulus activity
reaches the cerebral cortex 8–9 ms after stimulation while visual stimulation
involves a longer latency of 20–40 ms. Since the two neural pathways are of
comparable length, this difference points to a greater (p.137) transmission
speed for audition than vision. It should, however, be noted that the auditory
system contains both ipsilateral and contralateral pathways, while the pathways
of the visual system are exclusively crossed. The cortical response time for a
movement stimulus appears to be on the order of 10–20 ms (Adams 1976).

Reaction time is the time taken for a sense stimulus to travel to the CNS and
return to initiate and execute a largely pre-programmed motor response. Simple
reaction time (RT) with only one chosen motor response typically fall in the
range 100–250 ms, depending on conditions and sensory modality (Summers
1981). Auditory, kinaesthetic, and tactile reaction times have typically been
found to fall in the range 100–160 ms (Chernikoff and Taylor 1952; Higgins and
Angel 1970; Glencross 1977; Sage 1977), while visual reaction times have been
considered longer, typically reported as at least 190 ms (Keele and Posner 1968).
Reaction times for other sensory modalities seem to be in the range above 200
ms, while RTs involving choice of response are in general longer and are
reasonably modelled by Hick’s Law (Hick 1952). Kinaesthetic and tactile choice
reactions seem also to be faster than visual (Leonard 1959; Glencross and
Koreman 1979). Data on auditory choice RTs do not seem to be readily available.

Page 10 of 58

Error correction (EC) times vary with sensory modality and context. EC times
are important for improvisation because it may reasonably be argued that they
reflect minimum times for decision-making that is expressive or compositional.
Visual error correction is usually reported to be about 200 ms, whereas
kinaesthetic EC can occur over intervals as short as 50–60 ms (Kerr 1982), as
seen in reports on tracking tasks (Gibbs 1965; Higgins and Angel 1970).
However, other recent work in the case of vision has found some instances of
visual EC times down in the range near 100 ms as well (Smith and Brown 1980;
Zelaznik et al. 1983). It seems likely that the time taken for error correction
would be a function of the degree of invoked processing involvement; that is,
motor programme construction would take more time than selection, while
exacting criteria of discrimination or motor accuracy or a wide range of
response choice would naturally increase EC time. Rabbitt and Vyas (1970) and
Welford (1974) have enunciated this view, one which is well supported by the
introspective reports of improvisers going back for many centuries (Ferand
1961).

Explicit information on auditory error correction times does not seem to be

available, but it is possible to point out a general tendency in the above data.
Namely, processing speed seems to be greatest for audition and touch/
kinaesthesia, of all the possible sensory systems. These are precisely the
elements involved in musical improvisation and provide a vivid psychological
interpretation for the historical fact that music, of all art and sport forms, has
developed improvisation to by far the greatest degree. Under this interpretation,
human beings, as creative agents, have as a (p.138) matter of course drawn on
the sensory systems most adapted to quick decision-making: in other words, a
predilection for improvised sound manipulation might be genetically
programmed. Of course, such an interpretation remains highly speculative.

Finally it should be noted that unexpected sensory changes requiring significant

voluntary compensations require a minimum time of about 400–500 ms (Welford
1976). This is therefore the time scale over which improvising players in
ensembles can react to each others’ introduced novelties (about twice a second).
Nuances in continuous improvised performance based on self-monitoring are
probably limited by error correction times of about 100 ms (Welford 1976), so
that speeds of approximately 10 actions per second and higher involve virtually
exclusively pre-programmed actions (Pressing 1984a). An informal analysis of
jazz solos over a variety of tempos supports this ball-park estimate of the time
limits for improvisational novelty.

Timing and movement invariants

Page 11 of 58

Up to this point very little has been said about the timing of skilled performance,
yet it is obviously a vital point. Considerable experimental work in the domains
of fluent speech (Huggins 1978), typing (Shaffer 1978; Terzuolo and Viviani 1979),
handwriting (Denier van der Gon and Thuring 1965; Viviani and Terzuolo 1980;
Hollerbach 1981), generalized arm trajectories (Morasso 1983), and piano
performance (Shaffer 1980, 1984) has established that invariant timing and
spatial sequences, strongly suggestive of schemata, underlie skilled actions.
Such performance rhythms, or ‘hometetic’ behaviour, as some have termed it,
shows great tuneability: over wide variations in distance and overall time
constraints, invariance of phasing and accelerations (equivalently, forces) can be
observed (Schmidt 1983). By phasing is meant the relative timings of component
parts of the entire movement sequence. But it is also true that the relative
timings of movement components can be changed intentionally, at least to a
considerable degree. Hence the improviser has access to generalized action
programmes (in both motor and music representation), which allow overall
parametric control (time, space, force) and subprogram tuneability. This may
well be responsible for the flexibility of conception characteristic of experienced
improvisation.

Motor memory

It has often been suggested that a distinct form of memory for action, called
motor memory, exists. The subjective impression of improvisers (and other
performers) is certainly that potentially separate yet often interconnected motor,
symbolic, and aural forms of memory do exist. For a review of this extensive
topic and its relationship to verbal memory the reader may wish to consult Laabs
and Simmons (1981).

(p.139) Skill development

All skill learning seems to share certain common features. In the early stages, a
basic movement vocabulary is being assembled and fundamental perceptual
distinctions needed for the use of feedback are drawn. In intermediate stages,
larger action units are assembled, based on stringing together the existing
movement vocabulary in accordance with the developing cognitive framework.
These action units begin to enable predictive open-loop response. The ability to
perceive distinctions is refined considerably, and internal models of action and
error correction are developed. Expressive fluency begins to appear,
characterized by a feeling of mindful ‘letting go’ (Schneider and Fisk 1983;
Pressing 1984a). By the time advanced or expert stages have been reached, the
performer has become highly attuned to subtle perceptual information and has
available a vast array of finely timed and tuneable motor programmes. This
results in the qualities of efficiency, fluency, flexibility, and expressiveness. All
motor organization functions can be handled automatically (without conscious
attention) and the performer attends almost exclusively to a higher level of
emergent expressive control parameters.

Page 12 of 58

In the case of improvised music these emergent control parameters are notions
such as form, timbre, texture, articulation, gesture, activity level, pitch
relationships, motoric ‘feel’, expressive design, emotion, note placement, and
dynamics. There must also be a developed priority given to auditory monitoring
over kinaesthetic and especially visual monitoring. This idea is supported by
research on typists (West 1967), which showed that the dominant visual control
used for optimal results in early stages of learning to type gave way later to
reliance on tactile and kinaesthetic cues. It also seems likely that sensory
discrimination and motor control functions make increasing use of higher order
space-time relationships (velocity, acceleration) as skill learning progresses
(Marteniuk and Romanow 1983).

The change from controlled processing to automatic motor processing as a result

of extensive skill rehearsal is an idea of long standing (James 1890; Shiffrin and
Schneider 1977), and it undoubtedly improves movement quality and integration
(Eccles 1972). The accompanying feeling of automaticity, about which much
metaphysical speculation exists in the improvisation literature, can be simply
viewed as a natural result of considerable practice, a stage at which it has
become possible to completely dispense with conscious monitoring of motor
programmes, so that the hands appear to have a life of their own, driven by the
musical constraints of the situation (Bartlett 1947; Welford 1976; Pressing
1984a). In a sense, the performer is played by the music. The same thing
happens with common actions like walking and eating. As Welford (1976) has
cogently pointed out, automaticity is therefore especially likely when the actions
involved are always, or virtually always, accurate to within the (p.140)
requirements of the task. Hence automaticity in improvisation can be frequent in
both free and highly structured contexts, since task requirements are often self-
chosen, but is more likely to be successful in musical terms for the less
experienced player towards the free end of the specturm.

Schneider and Fisk (1983) have proposed an interesting corollary to the above,
based upon a classification of tasks into those requiring consistent or varied
processing: ‘Practice leads to apparently resource free automatic productions
for consistent processing but does not reduce (attentional) resources needed for
a varied processing task.’ (p, 129) This idea is appealing and perhaps widely
valid, but is too simple to encompass the full complexity of improvisation. For
part of the result of extensive practice of improvisation is an abstraction to
greater and greater generality of motor and musical controls to the point where
highly variable, often novel, specific results can be produced based on the
automatic use of general, highly flexible and tuneable motor programmes. More
irrevocable constraints causing attentional loading seem to be timing and
interhand co-ordination (Pressing 1984a).

Page 13 of 58

Another relevant area is the optimum distribution and nature of practice.

Generalizations here are particularly hazardous (Newell 1981) and I will confine
my comments specifically to improvisation.

The extremes of massed and distributed practice typically have complementary

functions for the improviser. Distributed practice develops immediacy, and
consistency of results under variable conditions, whereas massed rehearsal, by
perhaps bringing to the player’s awareness otherwise unperceived repetitive
aspects of his or her music, enables the transcendence or improvement of stale
musical design. One is reminded of the opinion of master trumpeter Miles Davis
that his sidemen only really got loose in the last set of the night, after they had
used up all their well-learned tricks (Carr 1982).

Variability of practice conditions is vital for improvisation, for obvious reasons,

and this seems to be true of nearly all skilled behaviour (Schmidt 1983). Mental
practice away from the instrument can be important for performers of fixed
music, based on internal hearing of scores, but there seems very little record of
its use in improvisation. This is presumably due to the intrinsically vital motoric
link between performer and instrument for improvisation.

Techniques used by musicians to teach improvisation will be described below.

However, some general principles of skill teaching are pertinent here. The
successful yet contrasting approaches of the ‘discovery’ method and structural
prescription (the use of instructions or demonstrations) may be mentioned. The
basic trial-and-error idea of the discovery method probably requires little
explanation; it has often been used as an industrial training procedure, where
learning sessions are arranged so that trainees must make active choices which
are normally correct, and which therefore (p.141) do not lead to ingrained
errors (Welford 1976). Less formalized self-discovery techniques are certainly
characteristic of much learning in the arts. But structural prescription is also a
vital part of skill learning. For all but very simple skills, instructions seem
particularly effective when kept simple, and when focusing on goals and general
action principles rather than kinematic details (Hendrickson and Schroeder
1941; Holding 1965; Newell 1981). This certainly holds for improvisation.
Probably too much intellectual detail both interferes with the fluid organization
of action sequences, as mentioned earlier, and strains attentional resources.

Studies and theories of musical improvsation

Page 14 of 58

A cognitive overview of much of this literature has been given earlier (Pressing
1984a, which includes references to dance and theatre), and will not be
repeated here. Historical surveys of improvisation in Western music may be
found in Ferand (1938, 1961), The new Grove dictionary of music (1983), and
Pressing (1984b,c). These deal primarily with the period to 1900. Discussion of
avant-garde improvisation since 1950 is included in Cope (1984). Non-Western
musical improvisation is described by Reck (1983), Datta and Lath (1967), Wade
(1973), Jairazbhoy (1971), and Lipiczky (1985) for Indian music; by Nettl and
Riddle (1974), Nettl and Foltin (1972), Zonis (1973), Signell (1974, 1977), and
Touma (1971) for various Middle Eastern traditions; by Béhague(1980) for Latin
American musics; by Hood (1971, 1975), Sumarsam (1981) for gamelans and
other stratified ensembles in Southeast Asia, and by Jones (1959) and Locke
(1979) for Ewe music of Ghana. Park (1985) has described the improvisation
techniques of Korean shamans, Avery (1984) structure and strategy in Azorean–
Canadian folkloric song duelling, and Erlmann (1985) variational procedures in
Ful’be praise song. Nettl (1974) has provided thoughtful general insights from
the perspective of the ethnomusicologist.

In the twentieth century, prescriptive teaching texts on Western music

improvisation are legion. Few, however, have the sorts of cognitive insights
useful in model building, and almost all are concerned with the specifics of jazz
(a small related number with blues and rock) or keyboard (particularly French-
tradition organ) improvisation. The jazz texts are too numerous to survey fully
here and are in any case mostly quite repetitious. Important perspectives are
however given by Coker (1964, 1975), Schuller (1968), Baker (1969), Owens
(1974), Liebman et al. (1978), Dobbins (1978), Howard (1978), Murphy (1982),
and Radano (1985). Among the better organ and piano texts may be mentioned
the works of Dupré (1925/37), Schouten (no date given), Gehring (1963),
Berkowitz (1975) and Weidner (1984). Analytical and prescriptive texts which
stand apart from the typical stylistic conventions above are the works of Bailey
(1980), (p.142) Bresgen (1960), Sperber (1974), Stumme (1972), and Whitmer
(1934). Except for Bailey, all of these take tonal music as their primary area of
discourse. Discussions which emphasize free improvisation often take a more
cognitive approach, but their usefulness is sometimes compromised by
vagueness or subjectivity. Valuable readings in this area include Silverman
(1962), Jost (1974), Parsons (1978), Bailey (1980), and special issues of
Perspectives of new music (Fall-Winter 1982/Spring-Summer 1983, 26–111), the
Music educator’s journal (1980, 66, (5), 36–147), Keyboard (1984, 10(10)), and
The British Journal of Music Education (1985, 2(2)). Other works of interest are
those on choir improvisation (Ehmann 1950, Ueltzen 1986), silent-film
accompaniment (Hanlon 1975), dulcimer improvisation (Schickhaus 1978), and
percussion gestures (Goldstein 1983).

Page 15 of 58

Musical improvisation has also been considered as a vehicle for consciousness

expansion and the tapping of deep intuitions, A full history of this ‘transpersonal’
approach would go back thousands of years to the sacred texts of many
religions. Here I only survey recent Western opinion. Hamel (1979) has
intelligently chronicled music of the avant-garde (for example Riley,
Stockhausen) from this perspective. Laneri (1975) has developed a philosophy of
improvisation based on different states of consciousness, featuring the concepts
of synchronicity and introversion. The resultant music is primarily vocal, since
the voice is considered the primal instrument. A powerful system of sonic
meditation most applicable to local improvisation groups has been developed by
Oliveros (1971). ‘Sensing’ compositions have been published by Gaburo (1968).
An attempt to connect music, altered states of consciousness, and research in
parapsychology has been given by Pressing (1980), while Galás (1981/82) has
created a primal vocal music based on obsession, excessive behaviour, and
trance states of severe concentration.

The approaches in the literature to the teaching of improvisation may be broadly

grouped as follows. First, there is the perspective overwhelmingly found in
historical Western texts, that improvisation is real-time composition and that no
fundamental distinction need be drawn between the two. This philosophy was
dominant in pre-Baroque times but had become rare by the eighteenth century.
In practice this results in a nuts-and-bolts approach with few implications for the
modeling of improvisation beyond basic ideas of variation, embellishment, and
other traditional processes of musical development. A second approach, which
historically took over as the first one waned, sets out patterns, models, and
procedures specific to the improvisational situation, which, if followed by those
possessing a solid enough level of musicianship, will produce stylistically
appropriate music. In this category fall the many figured bass and melodic
embellishment texts of the seventeenth and eighteenth centuries (for example
Mersenne 1635; Quantz 1752/1966; Bach 1778/1949; Arnold 1965), as well as
the riff (p.143) compendia and how-to-do-it books in the field of jazz (such as
Coker, et al. 1970; Slonimsky 1975; Nelson 1966).

A third technique is the setting of a spectrum of improvisational problems or

constraints. The philosophy behind this technique shows a clear contrast with
the second approach above, as described by Doerschuk (1984), referring to the
Dalcroze system.

The art of improvisation rests on…a developed awareness of one’s

expressive individuality. This knowledge grows through interactive
exercises with a teacher, whose function is not to present models for
imitation, but to pose problems intended to provoke personal responses,
(p.52)

Page 16 of 58

Jaques-Dalcroze (1921) seems to have pioneered this approach in our century

with a revealing series of improvisation exercises for piano. These include
composition-like problems in rhythm, melody, expressive nuance, and harmony;
muscular exercises; imitation of a teacher; exercises in hand independence; the
notation of improvisation just after performing it; and what may be termed an
‘interrupt’ technique. In this last technique the word ‘hopp’ is recited by the
teacher, as a cue for the student to perform pre-set operations such as
transposition or change of tempo during the performance. This technique is
reminiscent of a much later suggestion by Roads (1979) that musical grammars
used in improvisation might be ‘interrupt-driven’. This idea is developed in the
model below.

Parsons (1978) has made effective use of this third technique in a collection of
short pieces by many different composers defined largely by improvisational
instruction sets; he also presents a taxonomy of psycho-improvisational faults
and recommended exercises for correcting them. A shorter multi-author
collection of improvisational exercises is found in Armbruster (1984). Jazz fake
books like the Real book (no listed authors or dates) or The world’s greatest fake
book (Sher 1983) may also be considered to act along the lines of this technique.

A fourth approach is the presentation of multiple versions of important musical

entities (most commonly motives) by the teacher, leaving the student to infer
completely on his or her own the ways in which improvisation or variation may
occur by an appreciation of the intrinsic ‘fuzziness’ of the musical concept. This
imitative self-discovery approach is found in the Persian radif, which is a
repository of musical material learned in a series of increasingly complex
versions by the aspiring performer (Nettl and Foltin 1972), and in Ghanaian
traditions (K. Ladzekpo, personal communication), for example. A related
procedure made possible by the use of recording technology in the twentieth
century is for the student to directly copy a number of improvised solos by
repeated listening to recordings, and from this extract common elements and
variation procedures. Song-form based improvisations, in which solos consist of
a number of choruses which repeat the same underlying chord (p.144)
progression, are particularly suitable. This method has been widely used in jazz
and blues since the end of the First World War.

A fifth approach is allied to the self-realization ideas of humanistic psychology. It

is based on concepts of creativity and expressive individuality which go back in
music explicitly at least to Coleman(1922), implicitly certainly to Czerny (1829/
1983), and probably in a general sense at least to the Enlightenment. Important
educational applications of this idea are found in the works of Carl Orff, Zoltán
Kodály, Suzuki (see Mills and Murthy 1973), and particularly Jaques-Dalcroze
(1976, 1930) and Shafer (1969). In the words of Jaques-Dalcroze.

Page 17 of 58

Improvisation is the study of direct relations between cerebral commands

and muscular interpretations in order to express one’s own musical
feelings…Performance is propelled by developing the students’ powers of
sensation, imagination, and memory.

(In Abramson 1980, p.64.)

Little actual research on optimal techniques for teaching improvisation has been
carried out. The important study by Partchey (1973) which showed the value of
models and particularly of subsequent aural feedback in learning to improvise
has already been mentioned above. Work by Hores (1977) has shown that visual
and aural approaches to the teaching of jazz improvisation can be equally
effective. Burnsed (1978) looked at the efficacy of design of an introductory jazz
improvisation sequence for band students. Seuhs (1979) developed and assessed
(by adjudication) a course of study in Baroque improvisation techniques. Bash
(1983) compared the effectiveness of three different instructional methods in
learning to improvise jazz. Method I was a standard technical procedure based
on scales and chords. Method II supplemented this technical dimension with
aural perception techniques which included rote vocal responses to blues
patterns, blues vocalizations, and instrumental echo response patterns based on
rote or procedures of generalization. Method III supplemented the same
technical procedures of Method I with a historical-analytical treatment. All three
methods gave improved results over that of a control group, and methods II and
III, though no significant difference was found between them, were both
superior to method I. The results show the value of specific theoretical and
technical instruction, and also of its supplementation by relevant aural training
or analyses of performance strategies used by virtuoso improvisers.

One final comment on improvisation teaching seems apposite. This is the fact
that the optimally effective teacher is able to direct evaluative comments on
several different levels. One is the technical—‘Your notes don’t St the chord’,
‘The piano is lagging behind the bass’, etc. Another is the compositional—‘Try to
develop that motive more before discarding it’, ‘Use more rhythmic variety in
pacing your solo’, ‘Musical quotations seem (p.145) inappropriate in this free a
context’, etc. Yet another level is the use of organizing metaphor, a vital part of
the tradition of jazz teaching—‘Use more space’, ‘Dig in’, ‘Go for it’, ‘Play more
laid-back’, ‘Don’t force it-follow the flow’, etc. Simple comments of this kind can
be remarkably effective at removing improvisational blocks, when delivered at a
proper time.

Page 18 of 58

Pike (1974) has presented a brief but insightful phenomenology of jazz. His
approach considers the projection of ‘tonal imagery’ to be the fundamental
process in jazz improvisation. Tonal imagery is either ‘reproductive’ (memory-
based) or ‘productive’ (creative). The improviser operates in a ‘perceptual field’
which acts as a framework in which the improviser’s imagery appears and
originates. This field includes not only the perception of external tonal events,
but the perception of internal images, as well as the states of consciousness
evoked by these images. Images in this field are combined, associated,
contrasted, and otherwise organized. The phenomenological operations
describing this are processes such as repetition, contrast, continuity, completion,
closure, and deviation. Other aspects of improvisation defined by Pike include
‘intuitive cognition’, an immediate penetration into the singular and expressive
nature of an image, and ‘prevision’, a glimpse into the developmental horizons of
an embryonic jazz idea.

Although some of Pike’s claims are open to question, for example his uncritical
acceptance of concepts like Hodeir’s ‘vital drive’ (Hodeir 1956), his short paper
remains an important introspective analysis of the experience of improvisation.
The only other extensive phenomenological treatment of improvisation seems to
be Mathieu’s (1984) study of musician/dancer duo performances. Other
perspectives on the experiences of the improviser have been given by Milano
(1984), in an interview with jazz pianist/psychiatrist Denny Zeitlin, and Sudnow
(1978), who has produced a basic ethnomethodological description of learning to
play jazz on the piano. Related philosophical issues have been raised by Alperson
(1984) and Kleeman (1985/86).

Finally it may be proper to note that the computer age has spawned new hybrids
of composition and improvisation. Fry (1980, 1982/83) has described music and
dance improvisation set-ups using computer sensing and control devices.
Chadabe (1984) has described a method of ‘Interactive composition’ whereby
movements of the hands in space near two proximity-sensitive antennas trigger
and exert partial control over real-time computer sound generation. Interactive
computer-based performance systems have also been used by trombonist
George Lewis and a host of ‘performance artists’, including this writer. And
recently available commercial software, such as the Macintosh-based M and Jam
Factory, has an interactive improvisational component that seems rich with
promise.

(p.146) Oral traditions and folklore

The idea that traditional folk-tales from many cultures have underlying unities,
which may be interpreted as narrative grammars, is a fairly well-established one
(Propp 1927; Thompson 1946; Nagler 1974). Explanations of this fact have
tended towards one or the other of two viewpoints.

Page 19 of 58

A common (particularly European) perspective in the study of oral tradition and

folklore has been a focus on their repetitive and imitative aspects, with the
frequent assumption of an Urtext which has undergone historical and
geographic transformation. A powerful opposing view, and one which seems
increasingly relevant as a description of referent-based improvisation, is found
in the ‘formulaic composition’ proposals of Milman Parry and Albert Lord (Parry
1930, 1932; Lord 1964, 1965).

Formulaic composition was derived from Milman’s intense study of the Homeric
epics, particularly the Odyssey, and given further support by research on
Yugoslav folk-epic poetry conducted by Milman and Lord. It is also considered to
be applicable to other oral epics such as Beowulf and the Chanson de Roland,
and has been used to analyse Latvian folk-song texts (VĪkis-Freibergs 1984). In
this view epic oral poetry is created anew at each performance by the singer
from a store of formulas, a store of themes, and a technique of composition.
There is no ‘original’ version; instead the tradition is multiform. A ‘formula’ is a
group of words regularly employed under the same metrical conditions to
express a given essential idea; it has melodic, metric, syntactic, and acoustic
dimensions. By choosing from, a repertoire of roughly synonymous formulas of
different lengths and expanding or deleting subthemes according to the needs of
the performance situation, the experienced performer is able to formulaically
compose (in real-time, hence improvise) a detailed and freshly compelling
version of a known song epic. As a result of the composition system, instances of
pleonasm and parataxis are common.

The formulas considered as a group reveal further patterns. In the words of Lord
(1964): ‘the really significant element in the process is…the setting up of various
patterns that make adjustment of phrase and creation of phrases by analogy
possible’ (p.37). In addition, the permutation of events and formulas may occur,
as well as the substitution of one theme for another.

Yet the traditional singer does not seek originality with this technique, but
heightened expression. Lord speculates that formulas originally grew out of a
need for intensification of meaning or evocation. ‘The poet was sorcerer and
seer before he became artist’ (Lord 1964, p.67).

Page 20 of 58

The relevance of formulaic composition to specific types of musical

improvisation has recently been discussed by several writers. Treitler (1974) has
argued that Gregorian chant was composed and transmitted in an analogous
process to that used in the oral epics. Smith (1983) has used (p.147) the
process to describe the constraints imposed on the song-based jazz performer,
and has gone on to analyse piano improvisations by Bill Evans, Kernfeld (1983)
has examined how far formulas may be used to describe the music of
saxophonist John Coltrane. Reck (1983) has produced the evocative idea of a
musician’s ‘tool-kit’, in a mammoth study of five performances by South Indian
musician Thirugokarnam Ramachandra Iyer, The tool-kit is considered to be
piece-specific and to contain both individually chosen and culturally determined
formulas, musical habits, models of improvisational and compositional forms,
aesthetic values, and social attitudes.

The application of Parry-Lord theory to musical improvisation is thus a clear

contemporary trend. The limits of its validity and usefulness are still open
questions, and are probably linked to whether a satisfactory agreement can be
reached on the principles to be used to define musical ‘formulas’.

Intuition and creativity

These are two related concepts, each with a vast literature. Their connection
with improvisation is undeniable, yet explicit mention of it in either field is rare.
On the other hand, ‘free’ musicians and many music educators commonly use
the two terms, but often without a very clear notion of just what is being
discussed. This section attempts to bridge that gap.

The concept of intuition is much older than creativity, and it has separate
philosophical and psychological traditions. Westcott (1968) has provided an
excellent general survey, enumerating three historical approaches to
philosophies of intuition. First comes Classical Intuition (for example Spinoza,
Croce, Bergson), which views intuition as a special kind of contact with a prime
reality, a glimpse of ultimate truth unclouded by the machinations of reason or
the compulsions of instinct. Knowledge gained through this kind of intuition is
unique, immediate, personal, unverifiable. The second approach, called by
Westcott Contemporary Intuitionism (for example Stocks 1939; Ewing 1941;
Bahm 1960), takes the more restricted view that intuition is the immediate
apprehension of certain basic truths (of deduction, mathematical axioms,
causality, etc.). This immediate knowing stands outside logic or reason and yet is
the only foundation upon, which they can be built. Knowledge gained through
intuition constitutes a set of ‘justifiable beliefs’, which are nevertheless subject
to the possibility of error. A third approach is positivistic (for example Bunge
1962) in that it rejects as illusory both the notions of immediacy and ultimate
truth found in some earlier views. Rather, an intuition is simply a rapid inference
which produces a hypothesis.

Page 21 of 58

Of all these views, it is perhaps that of French philosopher Henri (p.148)

Bergson (1859–1941) which shows the greatest affinities with the common
metaphors of improvisation. Bergson saw intuition as a way to attain direct
contact with a prime reality ordinarily masked from human knowledge. This
prime reality is an ongoing movement, an evolving dynamic flux which proceeds
along a definite but unpredictable course.

The prime reality is referred to as ‘the perpetual happening’ or ‘duration’.

The mind of man, according to Bergson, is shielded from the perpetual
happening by the intellect, which imposes ‘patterned immobility’ on prime
reality, distorting, immobilizing, and separating it into discrete objects,
events and processes, In the perpetual happening itself, all events, objects,
and processes are unified’

(In Westcott 1968, p.8).

In Bergson’s view, the intellect can freely interact with the fruits of intuition
(special knowledge and experience), to develop an enriched personal
perspective.

The notion of tapping a prime reality is very similar to the improviser’s aesthetic
of tapping the flow of the music, as mentioned above. The same apparent
process has been eloquently described with regard to the origins of folk-tales
from many cultures by English writer Richard Adams:

I have a vision of—the world as the astronauts saw it—a shining globe,
poised in space and rotating on its polar axis. Round it, enveloping it
entirely, as one Chinese carved ivory ball encloses another within it, is a
second…gossamer-like sphere…rotating freely and independently of the
rotation of the earth.

Within this outer web we live. It soaks up, transmutes and is charged with
human experience, exuded from the world within like steam or an aroma
from cooking food. The story-teller is he who reaches up, grasps that part
of the web which happens to be above his head at the moment and draws it
down—it is, of course, elastic and unbreakable—to touch the earth. When
he has told his story—its story—he releases it and it springs back and
continues in rotation. The web moves continually above us, so that in time
every point on its interior surface passes directly above every point on the
surface of the world. This is why the same stories are found all over the
world, among different people who can have had little or no
communication with each other.

(Adams 1980, p.12.)

Page 22 of 58

There is a clear convergence of imagery in this and other descriptions that

points to a likely transpersonal component to improvisation.

The psychological perspectives on intuition are many and varied, but only two
seem relevant here. The first is the widely occurring idea that intuition is a
special case of inference which draws on cues and associations not ordinarily
used (Westcott 1968). A similarity with certain theories of skill learning
mentioned above is apparent. A second and wide-ranging approach is found in
the recent work by Bastick (1982), which includes a search of over 2.5 million
sources for common properties underlying intuition. After the identification and
detailed analysis of some 20 of these (p.149) properties, Bastick ends up
describing intuition as a combinatorial process operating over pre-existing
connections among elements of different ‘emotional sets’. These emotional sets
apparently contain encodings, often redundant, of many different life events
(intellectual activities, movement, emotion, etc.). By giving strong emphasis to
the role of dynamics, bodily experience, and the maximizing of redundancy in
encoding, and by a series of suggestive diagrams of intuitive processing, Bastick
seems to be on an important track parallel to emerging ideas of improvisation.

Research in creativity is probably more extensive than that in intuition, for

intuition is most commonly considered a subcategory of creativity. Creativity
research in music education has been recently surveyed by Richardson (1983).
The only clear relations to improvisation she found were in specialized
educational methods and a growing tendency to use improvisation tests in
assessing musical creativity. Vaughan (1971), Gorder (1976), and Webster (1977)
have designed and implemented such tests, but results show uneven patterns of
correlation between general intelligence, creativity, musicality, composition, and
improvisation, and seem to have no clear consequences for improvisation
modelling.

General studies of creativity abound, and follow many divergent paths. Two
alone seem, relevant here. Guilford’s Structure-of-Intellect (SI) model proposed
a taxonomy of factors of intelligence (Guilford and Hoepfner 1971 (and earlier
references mentioned therein); Guilford 1977). These intelligence factors, which
number 120, are classified along three dimensions: thought content: visual,
auditory figural, semantic, symbolic, and behavioural information; kinds of
operation performed on the content: cognition, memory, convergent production,
divergent production, evaluation; products (the results of applying operations to
content): units, classes, relations, systems, transformations, and implications.
These classifications are related to improvisation in a general way, but despite
their intuitive appeal, they have so far been fairly resistant to empirical
verification.

Page 23 of 58

Guilford further defined a set of six aptitudes for creative thinking: fluency,
flexibility, originality, elaboration, redefinition, and sensitivity to problems.
Torrance (1966) used this same set in designing a more open-ended approach to
the testing and definition of creativity. Some of these six aptitudes are identical
to the ones found in skilled performance above; they are considered here to be
further guidelines for testing the plausibility of improvisational modelling.

Finally, Guilford and Hoepfner classified techniques of evaluation (in problem-

solving), which they held to be due to appeals to logical consistency, past
experiences, feeling of Tightness, or aesthetic principles. Such a classification
also has implications for improvisation (see model below).

(p.150) Artificial intelligence

This field is concerned with programming computers to be intelligent problem
solvers. The framework of action is usually formulated in terms of a problem
space which must be searched for correct solutions. Since interesting problem
spaces are nearly always too large to be investigated completely, a major focus
of the field is the design of better heuristic search techniques. Coupled naturally
with this are many methods and frameworks for the representation of
knowledge.

There is traditionally no explicit mention of improvisation in the field. In making

such a link, it seems clear that the successful application of artificial intelligence
concepts to improvisation rests to a large degree on the appropriateness of
considering improvisation to be a kind of problem solving. There is little doubt
that such an analogy can be fruitful, particularly for referent-guided
improvisation. For example, the process of improvisation may be divided up into
a number of time points, and viewed as a succession of small problems, each of
which is the production of an appropriate chunk of musical action at the current
time point, where the constraints on action are the referent, goals, and musical
actions at earlier time points. Alternatively, the time-scale may be drawn much
coarser, and each complete improvisation may be considered a solution to a
much more generally stated problem: for instances, improvise a chorus on ‘I Got
Rhythm’ changes, within the constraints of be-bop style.

Before surveying the fruits of this approach it may be wise to spell out its
limitations. Experientially, improvisation can seem to be far removed from
problem solving. This is particularly so where the goals of the music making are
exploration and process, rather than the presentation of artistic product. It is
also very difficult to imagine how one could ever specify the ‘problems’ in freer
types of improvisation with sufficient detail to allow specific artificial
intelligence techniques to be used in modelling. Such problem formulations,
even if possible, would be very personal, open ended, and sometimes
contradictory.

Page 24 of 58

With these provisos, we examine how various artificial intelligence problem-

solving techniques might apply to improvisation. Search techniques come in
several variants, including depth-first, breadth-first, and best-first. All use a
generate-and-test procedure to find solutions to a problem. Clearly there are
possible connections with improvisation. Generate-and-test could be applied to
learning to improvise, where generation is sound production and testing is
listening to generated music; or, it could describe internal cognitive selection
processes, where testing is based on internal hearing of generated possibilities,
before one is chosen as the actual musical output at a given time. Unfortunately
with regard to this second interpretation there is a serious limitation: the
inevitable use of back-tracking in the search processes cannot be very
significant in improvisation (p.151) due to the cognitive limitations of real-time
processing. The need of the improviser is for a good solution, not the best, for
there is probably no single ‘best’ solution, and even if there were, it would take
too long to find it. Therefore, the number of solution paths compared at any one
step is probably very strongly limited, perhaps to two or three.

Another problem-solving technique is problem reduction: that is, reducing a

problem to a set of subproblems. This is a common way to look at the teaching of
improvisation, but seems less likely to apply to doing it, where integration of
action is required. Of course there is no proof of this; we know far too little
about the workings of the brain. Constraint satisfaction, on the other hand, is a
technique whose principles seem to apply to improvisation. The constraints are
the referent, goals of the performer, stylistic norms, etc. Finally, means-ends
analysis is a technique that is based on comparing current and goal states.
Because it involves considerable back-tracking, it is unlikely to apply to the
improvisation process. Yet like other methods above, it seems relevant to the
process of learning improvisational skill. In general, then, learning to improvise
(that is, to structure musical impulses within aesthetic guidelines) is more like
problem solving than is improvising itself.

Another main branch of artificial intelligence is knowledge representation. The

relevance to improvisation seems clear, for any particular mode of knowledge
representation makes it efficient to do certain things and inefficient to do others.
And efficiency is what the improviser needs above all.

Page 25 of 58

Knowledge representation in artificial intelligence is based on many ideas,

including indexing, conceptual dependency, hierarchies, semantic nets, multiple
representation, blackboards (actually a type of interprocess communication),
frames, scripts, stereotypes, and rule models (Rich 1983; Lenat 1984). With
respect to improvisation, many of these are more suggestive than readily
applicable. Indexing, for example, is too artificial, whereas conceptual
dependency, in which information is represented by certain conceptual
primitives, is too strongly linked with natural language structure. Hierarchies
have been discussed previously. Semantic nets are perhaps more promising:
information is represented as a network of nodes connected to each other by
labelled arcs, each node representing an object, event, or concept, and each arc
a relation between nodes. Such a graph could be drawn for musical objects and
events, but parametrically tuneable processes are not easy to represent, and this
is a serious drawback.

Multiple representation, however, is an important idea, and one which is implicit

in parallel-processing ideas mentioned earlier. The increased flexibility and
efficiency possible with multiple representation argue very strongly for its
inclusion in any model of improvisation. Gelernter (1963) successfully applied
the idea to problems in plane geometry by using simultaneous axiomatic and
diagrammatic representations. Another interesting application is the notion of
the ‘blackboard’, an organization of the (p.152) problem space into multiple
levels of representation, typically along a dimension indicating level of
abstractness. Thus a spoken sentence may be processed at levels of acoustic
wave form, phonemes, syllables, words, word sequences, phrases, etc. Each part
of the blackboard is triggered automatically as relevant information comes in.
Multiple representation also strengthens the possibilities for analogy, and
promotes synergy, by which is meant the co-operative action of parts of a
complex system (Lenat 1984).

Page 26 of 58

The last four ideas mentioned above, frames, scripts, stereotypes, and rule
models, are considered to be various types of schemata (Rich 1983). The use of
the word here is slightly different from that in the area of motor behaviour (see
Adams 1976 for a survey). Frames are used to describe collections of attributes
of an object. A frame consists of slots filled with attributes and associated
values. Like most slot-and-filler structures, frames facilitate the drawing of
analogies. Ideas equivalent to the frame are found in the improvisation model
below. Scripts are simply normative event sequences and in so far as they apply
to improvisation have much in common with the generalized motor schemata
described above. Stereotypes have their usual meaning and are parts of the
norms of musical style, but are often avoided by the best improvisers. Rule
models describe the common features shared by a set of rules which form the
basis for a ‘production system’. If the improvising musician is the production
system, the important rules will be largely heuristic and the rules about rules
may be termed metaheuristics. Some of these will be culturally and historically
based, while others presumably reflect intrinsic properties of the cognitive
apparatus. Serafine (1983) has presented an insightful discussion of this
distinction from the standpoint of the cognitive psychologist.

In principle it should be possible to integrate appropriate artificial intelligence

techniques to construct an expert system which improvises. One of the very few
such attempts is the unpublished work of Levitt (1981), which dealt with jazz
improvisation. The idea awaits further development.

A model of improvisation
Any theory of improvisation must explain three things: how people improvise;
how people learn improvisational skill; and the origin of novel behaviour. It must
also be consistent with the numerous recurring themes reviewed above. The
model given here seems to satisfy these conditions.

How peopic improvise

The first part of this model describes the process of improvisation. It begins with
the observation that any improvisation may be partitioned into a (p.153)
sequence of non-overlapping sections. By non-overlapping it is simply meant that
sounds are assigned to only one section, not that the sounds themselves do not
overlap. Let each of these sections contain a number of musical events and be
called an event cluster Ei. Then the improvisation I is simply an ordered union of
all these event clusters. Formally,

(1)

Page 27 of 58

From a naïve analytical perspective there is a large number of ways such a

partitioning could be made. Our first major assumption is that every
improvisation is actually generated by triggers at specific time points t 1 ,t 2 …tn
that instigate the movement patterns appropriate to effect intended musical
actions. Each time point is thus the point at which decided action begins to be
executed. Note that it is schemata for action that are triggered, not precise
movement details, and subsequent motor fine tuning based on feedback
processes goes on after each time point. Often time points will have clear
musical correlates, with adjacent event clusters being set off from each other by
local musical boundary criteria; pauses, phrase junctures, cadences, grouping by
sequence etc.; but this need not always be the case.

With this interpretation, equation (1) is a unique specification of the timing of

central decision making made by the improviser. The improvisation may then be
viewed as a series of ‘situations’, where the (i + 1)th situation is confined
primarily to the time interval (ti, ti +1) and entails the generation of the cluster
Ei+1 on the basis of the previous events {E 1, E2,…Ei} ≡ {E}i, the referent R (if
one exists), a set of current goals g, and long-term memory M, The referent R is
an underlying piece-specific guide or scheme used by the musician to facilitate
the generation of improvised behaviour (Pressing 1984a). The process of event-
cluster generation may then be written

(2)
Decision-making in the (i+1)th situation may in principle extend well back
before time ti, depending on the degree of pre-selection used by the performer,
and will also extend slightly into the future, in that fine details of motor control
will be left to lower control centres and hence may occur after t i+1.

Equation (2) applies strictly only to solo improvisation. The only changes with
group improvisation are that, first, all performers will have their own distinct
time-point sequences (even though they would often be partially correlated),
and, second, players will normally interact. Equation (2) can be readily extended
to apply to all K members of an improvisation ensemble by writing

(3)
(p.154) where subscripts refer to the kth performer, and C stands for
performer k’s cognitive representation of all previous event clusters produced by
the other performers and any expectations of their likely future actions. For
simplicity, we use the formalism of equation (2) and speak primarily in terms of
solo improvisation in what follows, adding in the effects of other performers in a
straightforward manner as needed at certain points.

Page 28 of 58

Any given event cluster E has a number of simultaneously valid and partially
redundant ‘aspects’. Each aspect is a representation of E from a certain
perspective. Most important are the acoustic aspect (produced and sensed
sound), the musical aspect (cognitive representation of the sounds in terms of
music-technical and expressive dimensions), and the movement aspect
(including timing of muscular actions, proprioception, touch, spatial perception,
and central monitoring of efference). Visual and emotional aspects normally also
play a role, and in principle there may be others. Furthermore each aspect exists
in two forms, intended and actual. Each intended form is specified at a specific
time point: the corresponding actual form is constructed from subsequent
sensory feedback. The gap between these two forms is reduced by sound
training in musicianship and improvisation practice, but it never dwindles
completely to zero. Hence in equation (2) or (3) the variable {E}i represents
intended and actual forms of all aspects of event clusters E 1 to Ei –1, the
intended form of Ei, plus, over the course of the time interval (ti, tj+1),
increasing feedback on the actual form of Ei. By ti +1, when central commands
for Ei+l are transmitted, the ongoing nature of improvisation probably demands
that integration of the intended and actual forms of Ei be virtually complete.

The details of the proposed model of what occurs in the (i+1)th situation, that is,
the selection of Ei+t, are as follows:

(A) Ei is triggered and executed (it may spill on briefly to times t≫ti +1).
(B) Each aspect of Ei may be decomposed into three types of analytical
representation: objects, features, and processes. An ‘object’ is a unified
cognitive or perceptual entity. It may, for example, correspond to a chord,
a sound, or a certain finger motion. ‘Features’ are parameters that
describe shared properties of objects, and ‘processes’ are descriptions of
changes of objects or features over time. At t¡ this decomposition is based
only on intended information (efference); by ti +1 much of the actual form
of Ei, received through the senses and internal feedback, has been used
to refine the cognitive representation of Ei. This may continue after ti +1.
Let this decomposition into objects, features, and processes (for each
aspect) be represented by three variable-dimension arrays O, F, and P,
and assume that they represent all information about Ei needed by the
improviser in decision making.

Page 29 of 58

(C) The structures of the three types of arrays are as follows. The object
array is a 2×N array where row 1 components label the objects present
and row 2 gives their associated cognitive strengths sk (explained below).
The (p.155) feature and process arrays are typically non-rectangular.
Their first rows consist of object and process labels respectively, and each
column below that row is built up of a number of pairs of elements which
give the values vjk of associated features or process parameters and their
corresponding cognitive strengths sjk. The arrays are non-rectangular
because different objects may possess different numbers of significant
features or process parameters. The feature and parameter process
values vjk vary over ranges appropriate to their nature, whereas cognitive
strengths sjk are normalized to vary between 0 and 1. Cognitive strength
is essentially an indicator of attentional loading, that is, the importance
that the given factor has in the performer’s internal representation. Thus
even though certain features may be objectively present, as analysed by
others, if the player does not use them in his or her cognitive
representation, their s values would be zero. Sample object, feature, and
process arrays for the following event cluster (a short trombone motive)
are given by way of example (Fig. 7.1), for the musical aspect only.
Considerable redundancy of representation has been set out in the
process array.

Page 30 of 58

(D) Production of Ei+l occurs primarily on the basis of long-term factors

(R, g, stylistic norms, and ongoing processes), and by evaluation of the
effects and possibilities of E¡, There seem to be only two methods of
continuation used; associative or interrupt generation. In associative
generation the improviser desires to effect continuity between Ei and Ei+1
and picks new arrays O i+I, F i+1, P i+1, whose set of strong cognitive
components includes all or nearly all of the strong cognitive components
of O i, F i, and Pi, with the parameter values of these shared components
being directly related (as described in (E) below). In other words the Ei
components with high s values carry their information on in some way to
Ei+l. These new arrays act as a set of constraints which determine, in
conjunction with various generation processes, the musical actions
generated for Ei+1, The relative importance of different constraints in the
generation process is indicated by their respective cognitive strengths sk
and sjk. Note that the Ei+l arrays may contain new strong components
(constraints) that were previously weak or completely absent. In
particular, it is possible to add a new independent musical process to a
continuing one to produce an associative continuation which has a clear
sense of novelty (e.g. the introduction of a new part in polyphonic music).
In the case of interrupt generation the improviser has had enough of the
event train ending with Ei (for whatever reasons) and breaks off into a
different musical direction by resetting a significant number of strong
components of O i+I, F i+1, P i+1 without any relations to E¡ except
possibly those chosen to be normative with regard to style in the piece, or
intrinsic to the referent (if present) or goals. Clearly, the more strong
components that are reset, the greater the sense of interruption.

Page 31 of 58

(E) Associative (p.156) Fig. 7.1. Possible object,

generation is based on feature, and process arrays
either similarity or corresponding to a short trombone
contrast. In tie case of motive.
similarity all or nearly all
important (important as
determined from the vantage point of the improviser) array components
stay approximately the same. In other words, for those components vjk
with sjk’s signicantly above zero at time t = t i , ( v j k ) t i ≈ ( v j k ) t i + 1.
Significant object array components behave analogously. In the case of
constrast-type associative generation, at least one strong component of
either the feature or process arrays must either move from near one end
of its possible range of values to near the opposite end, or cross some
perceptually significant boundary. Meanwhile, all other strong
components change either very (p.157) little or not at all. Examples are
when a group of high notes is followed by a group of low notes, or an
accelerando changes of decelerando, or bright timbres are replaced by
dull timbres. The idea behind this classification is that the most powerful
and general types of improvisational control are those that are cued to
features and processes. The objects, though a crucial part of the entire
procedure, are at the same time often merely the very familiar musical
clothing of cognitive action space.
(F) Interrupt generation is based on the resetting of all or a significant
number of the strong array components without regard to their values in
the current event cluster E¡. A decision to interrupt brings to an end a
sequence of related event clusters, say K = {E i–r, E i–r + 1,…Ei}, where the
number of event clusters in this ‘event-cluster class’ is r+1. Hence
interrupt decisions partition the entire improvisation into A discontinuous
event-cluster classes Kα, so that the formal design of the piece becomes
(4)
Each event-cluster class Kα contains at least one event cluster, and may
be defined in terms of the strong components of the object, feature, and
process arrays shared by all the member event clusters. If these special
components are represented as O s α , F s α , and P s α, then Kα is defined
by (O S, F s P s,)α. One of the sets F s α and P s α must be non-empty. If (O
s, F s P s)α = or ≍ (O s, Fs, Ps)β, for some β not immediately following a, we
have recursion in formal design of the improvisation. Under these
assumptions the process of improvisation may be sketched
diagrammatically as in Fig. 7.2.

Page 32 of 58

(G) The choice between association and interrupt generation may be

formally modelled by a time-dependent tolerance level for repetition, L(t).
An interrupt tester, whose inputs are presumably the time since the onset
of the Kα event cluster class, (t–ti–r), and the size and nature of Kα.
computes the degree of current repetition, Z(t), and if Z(t)≧L(t), institutes
an interrupt generation, so that Z(t) jumps to a low value. Otherwise
associative generation continues. Diagrammatically this is shown in Fig.
7.3 for the same improvisation as in Fig. 7.2.
(H) Once O i+1, F i+1, and P i+I are selected for all relevant aspects,
tuneable cognitive and motor subprogrammes are set in motion that
generate, on the basis of these higher constraints and current motor
positions, a specific action design. At this point we have reached ti+ 1 and
this loop of the process (Ei↠Ei+ 1) is complete. By iteration, then, the
entire improvisation is built up. The starting point E 1 may be considered
a situation of interrupt generation (where E 0 is silence) and the final
event cluster En is simply a second case of interrupt generation where
En+ 1 = silence, after which the improvisation process is turned off.

These, then, are the salient features of the model in outline. They are
diagramatically displayed in Fig. 7.4.

Next we look more deeply at certain critical stages of the improvisation (p.158)
model. To begin with, it is characterized throughout by extensive redundancy.
There is first of all redundancy between the aspects of each event cluster. The
performer knows, for example, that certain motor actions involved in striking a
kettle drum (motor aspect) will correspond to a particular sound (acoustic
aspect), with associated musical implications (musical aspect). Furthermore,
each aspect is decomposed into extensive object, feature, and process
representations which contain considerable redundancy. For example, the
musical motive of Example 7.1 may be pitch encoded as the objects D2F2A2B2, or
as the object BΦ diminished 7 chord in first inversion, or as a diatonic sweep to
the leading tone in the key of C major, or as a ii Φ diminished 7 chord in a minor,
or as an ascending contour, and so forth. Its features include melodic motion by
seconds or (p.159) thirds, diatonic note choice, the degree and speed of
crescendo, rhythmic regularity of attack, certain values of finger force and
velocity used by the performer, and so forth. Many processes could be implicated
to generate the given motive: arpeggiate a BΦ diminished 7 chord, pick notes
consistent with a triplet feel in C major, move the fingers 4321 of the left hand in
such a fashion as to depress keys on the piano, and so forth. If the nature of
improvisation entails the seeking out of a satisfactory trajectory in musical
action space, such redundancy of description and generation allows maximal
flexibility of path selection, so that whatever creative impulse presents itself as
an intention, and whatever attentional loadings (p.160) (p.161) may be set
up, some means of cognitive organization and corresponding motor realization
will be available within the limiting constraints of real-time processing.
Page 33 of 58

Fig. 7.2. An improvisation in musical

action space, showing four event-cluster
classes, and form ABA’C.

Fig. 7.3. Interrupt generation via the

repetition functions L and Z.

Page 34 of 58

Such extensive redundancy I

take here to mean that control
of event production is
heterarchical, and may
potentially shift rapidly from
one cognitive control area to Example 7.1
another. Indeed this must be
considered the most effective
strategy for improvisation,
Experientially it very probably
corresponds to ‘letting go’, or
‘going with the low’ as
described earlier, whereby
central hierarchical control,
identified here with conscious
monitoring of decision making, Fig. 7.4 The improvisation model in
yields to heterarchical control diagrammatic form. Only the process
(and corresponding unconscious Ei↠Ei+1 (intended) is detailed. Each event
allocation of attention). cluster E¡ is present in a number of
partially redundant aspects, and each of
these is decomposed into object, feature,
and process arrays. Largely on the basis
of musical representation a decision
about type of continuation is made by an
interrupt tester. In accordance with this
decision an intended array decomposition
is generated, with input from Ei arrays,
referent, goals, and memory. This
decomposition acts as a set of constraints
in the generation of musical action, and
production of Ei+1 is subsequently begun
by a movement trigger at ti +1. The
diagram detail shows what happens in
the time interval (ti,ti +l), so that the
indicated decomposition of Ei is
integrated (that is, intended plus actual
forms of E¡ are combined), whereas the
indicated decomposition of Ei+1 is
intended (no feedback has been received
yet). Hence O, F, and P at time point ti +1
do not have indicated outputs.

Page 35 of 58

Next we look further at the object, feature, and process arrays that are critical
in the representation and generation of event clusters. First of all it may well be
asked how such arrays are formed. The answer given here is based on an
ecological perspective, which considers that the capacity to extract or create
such arrays is neurologically innate, but that they are only brought into being by
interaction with the environment. More specifically, cognitive objects are
inferred to exist on the basis of perceived invariance in sensory input over time,
and boundedness in a space (whether physical, musical, or abstract). Features
are tuneable parameters and come to be abstracted on the basis of perceived
similarity or contrast in sensory input. Processes come about from perceived
change in an object or along a feature dimension with time.

Thus over the course of one’s life new arrays and array components are
constantly being created by new perceptions and new perceptual groupings.
During any given improvisation at most very few new features or processes will
be created, and only a limited number of new objects. In general, though, this is
one source of novel behaviour: the evolution of movement control structures for
newly discovered objects, features, and processes. However, there seems to be
another, probably more common source of behavioural novelty: the motor
enactment of novel combinations of values of array components. This second
possibility is shown for example by considering a child musician who has learned
motor actions corresponding to the distinctions loud/soft and fast/slow
separately, but without encountering soft and fast simultaneously. By combining
these two dimensions an action novel to the child’s experience can result.
Furthermore, the results of such novel parametric combinations need not be so
predictable. If we recall that the human performance system is non-linear, then,
as mentioned above in the paragraphs on organizational invariant theory, novel,
strikingly different behaviour may follow when controlling system parameters
assume certain novel combinations of ranges. It can further be shown
mathematically that behaviour described as ‘chaotic’ may occur under such
conditions (Li and Yorke 1975; May 1976), even for simple systems. This
perspective has led to a biomathematical analysis, for example, of many so-
called (p.162) ‘dynamical’ diseases, including schizophrenia, AV heart block,
epilepsy, and some haematological disorders (see Guevara et al. 1983 for a
survey). The point with regard to improvisation is that the same sort of smooth
parametric tuning can be used to generate abrupt intentional novelties in
movement and musical expression. The integration of the results of novel ranges
of array components is presumed to be handled by control structures of the CNS
responsible for timing and smoothness of action.

Page 36 of 58

During any given improvisation, when possible object, feature, and process array
types are basically fixed, novel sensory input will be analysed and assigned to
existing categories, or, if the it is too poor, into existing categories plus
deviations. In this model such a description is also considered to apply to the
generation of action. That is, novel actions are built primarily by distorting
aspects of existing ones. This sheds light on the organizing power of the
metaphor, mentioned earlier, since it may be considered to be a global link
across categories, one that facilitates movement integration. In other words, the
image or metaphor enables the co-ordinated modification and resetting of whole
classes of array components in a fashion ensuring spatial and temporal
coherence.

The central core of the model is the generation of a new set of array components
for Ei+l from those preceding it. To make this process clearer, we now look at
two examples.

(1) Let Ei be played by the right hand at the piano.

Above are a number of possible

improvisational continuations,
based on attentional emphasis
(that is, cognitive strength)
given to the mentioned array
components (see Fig. 7.5). Example 7.2
Emphasis given to a particular
component means that it will
guide the generation of subsequent events. The type of arrays emphasized are
also indicated; note that this is not uniquely determined, since the model makes
a feature of redundancy. Continuations 1–8 exemplify associative continuation,
with numbers 7 and 8 more abstract than the others, while number 9 is interrupt
based.

(p.163) (p.164)

Continuation Emphasized components used for continuation Type of

arrays

1 key of A major; quaver durations O,F,P

2 perfect fourth interval F

3 notes E, A, D; rhythmic displacement O,P

4 melodic contour O,F

5 motor generation with right-hand fingers 1, 2, O,F

and 4

Page 37 of 58

Continuation Emphasized components used for continuation Type of

arrays

6 gesture (note use of contrast), perfect fourth F,P

interval

7 phrase design (antecedent/consequent), F

interval class 2

8 notes E, A, D: chromatic decoration O,P

9 interrupt generation: new motive O,F,P

If the same line had been

played on flute, the
continuations might all have
been very similar except for
continuation number 5, which
has as its constraint focus the
actual movement patterns for
manipulating the instrument.

(2) Here we consider an event

cluster less clearly tied to
structural-historical processes.
Let Et be a segment of sounds
produced by a single slow
tilting and rotation of a
tambourine one-quarter filled
with a single layer of small lead
shot. E i is a coloured noise
sound which subjectively is
reminiscent of distant ocean
waves or rain. Some possible Fig. 7.5. Examples of continuation of an
continuations are then as event cluster under the emphasis of
follows: selected array components.

Continuation Description Type of

arrays

1 continue tilt but speed up rotation of F,P

tambourine

2 shake tambourine from side to side F,P

3 stop motion of tambourine F,P

Page 38 of 58

Continuation Description Type of

arrays

4 toss lead shot in air and catch it F,P

5 perform a drum roll on the bottom of the O,F,P

tambourine skin with fingers of the right hand

Continuations 1 and 2 are of associative type, whereas 3, 4, and 5 are interrupt

type. Notice that here description emphasizes the motor aspect, since there is no
extensive tradition of music theory which applies to such a sound source.

If these examples succeed in illustrating how continuations may be constructed,

they are mute on the details of how one continuation comes to be chosen over all
other possible ones. What has been said so far is only that, in associative
generation, a set of constraints is produced associatively, while in interrupt
generation the set of strong constraints on action includes uncorrelated
resetting. Obviously, event generation is informed by a vast panorama of
culturally and cognitively based musical processes and stylistic preferences
(motivic development, phrase design, historical forms, transposition, rhythmic
design, etc.). But a considerable degree of residual decision-making remains, as
for example the choice of array components that will be singled out to act as
strong constraints or to be reset. How are such residual decisions made?

It does not seem possible to give a final answer to this question, for it has (p.
165) at its ultimate root the question of volition and hence the mind-body
problem, about which there is no general philosophical agreement in our culture
or even among scientists. There is also no conclusive empirical evidence to
support one view or another, despite the opposing claims of some positivists and
phenomenologists. It seems useful therefore to characterize a number of
strategies of explanation for the residual decision-making mentioned above, and
subsequently explore what possibilities exist for experimentally decided among
them.

It is first of all possible to take the intuitive perspective, that the individual acts
best when he or she merely taps a certain powerful source that dictates the
course of musical action in a naturally correct fashion, one that may not be
analysable or predictable in physical or musical terms. Although this perspective
is usually transpersonal and may seem romantic to some, this does not imply
that it is untestable and therefore unscientific.

Page 39 of 58

A second perspective is to assume that this residual decision making actually

reflects the effects of individual free will. In other words, the improviser is a
unique conscious entity, and residual decision making rests to some degree on
internal variables not predictable even in principle from a fully detailed
knowledge of the physical-state variables of the improviser and his or her
environment.

A third perspective is the physicalist one. Here complex decision making is seen
to be an emergent property of the fantastically complex physical system known
as a human being, in interaction with a series of environments. Free will in this
perspective is either illusory, or simply a somewhat misleading metaphor for
certain complex characteristics of the system. There are a number of models
possible within this perspective for residual decision making: interactive control
with lower CNS centres, network statistical voting models, distributed memory-
type models, decision making based on fuzzy logic, etc.

Fourth and last is the perspective of randomness. Here the unconstrained

residual decision making is simply modelled by use of random generators. As the
improviser becomes more and more expert through practice and more and more
control procedures are built up, random processes need to be invoked less and
less frequently and overall error levels decrease, perhaps approaching a
minimum threshold.

To experimentally distinguish between these points of view a high resolution

improvisation transcription system has been built here at La Trobe University.
Co-worker Greg Troup and myself, as well as technical staff of the departments
of psychology and music, have designed and set up the apparatus. It is a
synthesizer-based system, using modified MIDI format, and enables detailed
recording, to millisecond resolution, of musical actions at a keyboard. It is also
possible to input sound from other (non-keyboard) instruments. Simultaneously
as music is recorded a videotape of the performance can be made. The results of
this investigation (p.166) are not yet complete and will be reported elsewhere,
but it seems likely that the limits of validity of intuitionist and random
perspectives will be determinable. There seems to be, however, no obvious
experimental design that will decide between the physicalist and free-will
perspectives. Hence the two may be considered co-existing formulations. The
problem in deciding between the two rests with setting up the repeatable
conditions which should theoretically lead to the same Improvised’ result in the
hardcore physicalist model and to a different improvised result under conditions
of free will. But since each event potentially affects all those that follow, all
initial conditions are intrinsically unrepeatable.

The development of improvisational skill

Page 40 of 58

The modelling of this process remains in a less developed state and only a brief
discussion is included here. Its starting point is the emergent results of practice
found in all types of skill, as mentioned earlier: improved efficiency, fluency,
flexibility, capacity for error correction, and, less universally, expressiveness. But
there are at least two additional components of improvisational skill:
inventiveness and the achievement of coherence. In more fixed skills these are
less important, since inventiveness provides few tangible advantages, and
coherence is built in by the rigidity of the task demands.*

The specific cognitive changes that allow these properties to develop in

improvised musical behaviour are considered to be:

(1) an increase in the memory store of objects, features, and processes—

in musical, acoustic, motor (and other) aspects;
(2) an increase in accessibility of this memory store due to the build-up of
redundant relationships between its constituents and the aggregation of
these constituents into larger cognitive assemblies;
(3) an increasingly reined attunement to subtle and contextually relevant
perceptual information.

The build-up and improved access to memory of points (1) and (2) is presumably
central to any learning process. In the language of the model of this chapter this
involves the use of extensive redundancy, and also the aggregation of memory
constituents (objects, features, processes) into new cognitive assembles which
may be accessed autonomously. Because such a procedure can presumably be
nested to arbitrary depth, very complicated interconnected knowledge
structures may develop.

Page 41 of 58

This last idea is not new here. It has a considerable history and has been most
clearly outlined for the purposes of this paper by Hayes-Roth (1977), who
generalized an earlier model of Hebb (1949). The central feature of (p.167)
aggregation of memory elements Hayes-Roth termed unitisation, and her
knowledge-assembly theory was built up around the presence of elemental
‘cognitive units’. In the terminology of this chapter these are object, feature, and
process array components. In knowledge-assembly theory such cognitive units
are associatively activated and may combine to form assemblies, whose
‘strengths’ are increasing functions of recency and frequency of activation, and
decreasing functions of their own complexity. From these strengths are derived
probabilities and speeds of activation. There is a level of redundancy appropriate
to improvisation in this model, since for example a cognitive unit may be
activated individually or as part of a larger assembly of cognitive units. In her
paper, Hayes-Roth shows that knowledge-assembly learning theory is consistent
with a large body of experimental results. It is also consistent with the
introspective reports of improvisers and the review given above of improvisation
teaching methods. But a decision on the superior applicability of this theory to
improvisation over those of other related formulations must rest upon
experimental work as yet undone. For this reason I give no further speculation.

The third point mentioned above may be elaborated as follows. The refinement
of improvisational skill must depend partly on increasing the efficiency of
perceptual processing to allow the inclusion of more and better-selected
information in the improviser’s decision-making procedures. The need for this
efficiency is imposed by every performer’s more or less limited individual
capacity, per unit time, to process novel sensory input. It seems likely that
practice leads to the increasingly efficient use of information in two ways: by
reducing the effective amount of information by the recognition of patterns of
redundancy in the sensory input, and by focusing attention increasingly on the
information that is most relevant for producing a successful improvisation. The
increased use of such subtle and ‘higher-order’ information leads to the higher-
order skill characteristics mentioned earlier. The main differences in this
process between fixed and improvised actions may be said to reside in the
nature of the attention focus used in the two situations. The fixed-skill situation
evolves towards a minimal size attention set, whereas the unpredictability of
improvisation demands that the attention focus remain wide. To go beyond such
insufficiently specific observations experimental work is clearly required.

Conclusions

Page 42 of 58

This chapter has attempted to iluminate the process of musical improvisation by

first examining the modelling tools available from a number of different
disciplines. Based on this examination, a cognitive model has then been
presented for the process itself, followed by a brief discussion of its relation (p.
168) to improvisational skill acquisition. The central features of the model are
as follows. It is reductionist, in that cognitive structures of processing and
control are considered to be broken down into aspects (acoustic, musical,
movement, etc.), each of these into types of analytical representation (objects,
features, processes), and each of these into characterizing elements (array
components). At the same time the model is synergistic and capable of
behavioural novelty, due to the extensive redundancy of the cognitive
representations and the distributed and non-linear character of the outlined
control processes. The extensive presence of feedback and feedforward
contributes to this. The fundamental nature of the improvisation process is
considered to be the stringing together of a series of ‘event clusters’ during each
of which a continuation is chosen, based upon either the continuing of some
existing stream of musical development (called here an event-cluster class) by
association of array entries, or the interruption of that stream by the choosing of
a new set of array entries that act as constraints in the generation of a new
stream (new event-cluster class).

The model seems to be specific enough to allow its use as a basis for the design
of ‘improvising’ computer programs. Work in this direction is in progress. At the
same time some fundamental philosophical questions remain about the origin of
certain kinds of decision making in any such model, and four types of answers to
these have been outlined; intuition, free will, physical causation, and
randomness. Some of these alternatives should be distinguishable on the basis of
experimental work currently in progress at our laboratories, which also has as
its aim the testing of the basic assumptions of the model. This will be described
in subsequent publications.

Acknowledgement
I am indebted to John Sloboda, Margot Prior, Geoff Cumming, Geoff Webb, Denis
Glencross, and Glynda Kinsella for helpful criticism.

References

Bibliography references:

Abramson, R. M. (1980). Dalcroze-based improvisation. Music Educator’s Journal

66, (5), 62–8.

Adams, J. A. (1961). Human tracking behaviour. Psychological Bulletin 58, 55–79.

Adams, J. A. (1971). A closed-loop theory of motor learning. Journal of Motor

Behaviour 3, 111–49.

Page 43 of 58

Adams, J. A. (1976), Issues for a closed-loop theory of motor learning. In Motor

control; issues and trends (ed. G. E. Stelmach). Academic Press, New York.

Adams, R. (1980). The iron wolf, Penguin, Reading.

(p.169) Allport, D. A. (1980), Attention and performance. In Cognitive

psychology; new directions, (ed. G. Claxton). Routledge and Kegan Paul, London.

Alperson, P. (1984), On musical improvisation. Journal of Aesthetics and Art

Criticism 43, 17–29

Armbruster, G. (1984). First steps in improvisation. Keyboard 10 (Oct.), 37–44.

Arnold, F. T. (1965). The art of accompaniment from a thoroughbass as practised

in the XVIIth and XVIIIth centuries, Dover, New York.

Austin, L., Oliveros, P., et al. (1982/83). Forum: improvisation. Perspectives of

New Music, Fall-Winter 1982/Spring-Summer 1983, 26–111.

Avery, T. L. (1984) Structure and strategy in Azorean-Canadian song duels.

Unpublished Ph.D. thesis, Indiana University.

Bach, C. P. E. (1778/1949). Essay on the true art of playing keyboard

instruments. Norton, New York.

Bahm, A. (1960). Types of intuition. University of New Mexico publications in

social sciences and philosophy, No. 3.

Bailey, D. (1980). Improvisation: its nature and practice in music. Moorland.

Baker, D. (1969). Jazz improvisation. Maher, Chicago.

Bartlett, F.C. (1947). The measurement of human skill. British Medical Journal 1,
835, 877.

Bash, L. (1983). The effectiveness of three instructional methods on the

acquisition of jazz improvisation skills. Unpublished Ph.D. thesis. State
University of New York at Buffalo.

Bastick, T. (1982). Intuition: how we think and act. Wiley, Chichester.

Beatty, J. (1975). Introduction to physiological psychology. Brooks/Cole,

Monterey, Calif.

Béhague, G. (1980), Improvisation in Latin American musics. Music Educator’s

Journal 66, (5), 118–25.

Bentley, D. and Konishi, M. (1978). Neural control of behaviour. Annual Review

of Neurosciences 1, 35–59.

Page 44 of 58

Bergson, H. L. (1910). Matter and memory (authorized trans. N. M. Paul and W.

S. Palmer). Allen, London.

Berkowitz, S. (1975). Improvisation through keyboard harmony. Prentice-Hall,

Englewood Cliffs, NJ.

Bernstein, N. (1967). The co-ordination and regulation of movements. Pergamon,

London.

Bresgen, C. (1960). Die improvisation. Quelle and Meyer, Heidelberg.

Brooks, D. N. and Baddeley, A. (1976), What can amnesic patients learn?

Neuropsychotogia 14, 111–22.

Brooks, V. B. (1978). Motor programs revisited. In Posture and movement:

perspectives for integrating sensory and motor research on the mammalian
nervous sytem. Raven, New York.

Bruner, J. S. (1969). Modalities of memory. In The pathology of memory (ed. G.

Tallard, and N. C. Waugh), pp. 253–9. Academic Press, New York.

Brunia, C. H. M. (1980). Motor preparation, recorded on the cortical and spinal

level. In Tutorials in motor behaviour (ed. G. E. Stelmach and J. Requin). North-
Holland, Amsterdam.

Bunge, M. (1962). Intuition and science. Prentice-Hall, Englewood Cliffs, NJ.

(p.170) Burnsed, C. V. (1978). The development and evaluation of an

introductory jazz improvisation sequence for intermediate band students.
Dissertation Abstracts International, 41A, 1214A.

Carr, I. (1982). Miles Davis, Paladin, London.

Chadabe, J. (1984). Interactive composing: an overview. Computer Music Journal

8, (1), 22–7.

Chernikoff, R. and Taylor, F. V. (1952). Reaction time to kinesthetic stimulation

resulting from sudden arm displacement. Journal of Experimental Psychology 43,
1–8.

Clark, H. H. and Clark, E. V. (1977). Psychology and language. Harcourt Brace

Jovanovitch, New York.

Cohen, N. J. (1981). Neuropsychological evidence for a distinction between

procedural and declarative knowledge in human memory and amnesia.
Unpublished Ph.D. thesis. University of California at San Diego.

Page 45 of 58

Cohen, N. J. and Squire, L. R. (1980), Preserved learning and retention of

pattern analysing skill in amnesia: dissociation of knowing how and knowing
that. Science 210, 207–9.

Coker, J. (1964). Improvising jazz. Prentice-Hall, Englewood Cliffs, NJ.

Coker, J. (1975). The jazz idiom, Prentice-Hall, Englewood Cliffs, NJ.

Coker J., Casale, J., Campbell, G., and Greene, J. (1970). Patterns for jazz. Studio
Productions, Lebanon, Ind.

Coleman, S. N. (1922). Creative music for children, G. P. Putnam’s Sons, New

York.

Cope, D. H. (1984), New directions in music. William Brown, Dubuque, Iowa.

Czerny, C. (1829/1983). Systematic introduction to improvisation on the piano

(Trans. A. L. Mitchell). Longman, New York.

Datta, V. and Lath, M. (1967), Improvisation in Indian music. World of Music, 9,

(1), 27–34.

Davis, R. (1957). The human operator as a single channel information system.

Quarterly Journal of Experimental Psychology 9, 119–29.

Delcomyn, F. (1980). Neural basis of rhythmic behaviour in animals. Science

210, 492–8.

Denier van der Gon, J. J. and Thuring, J. Ph. (1965). The guiding of human
writing movements, Kybernetik 2, 145–8.

Dobbins, B. (1978), The contemporary jazz pianist, 4 vols. GAMT Music Press,
Jamestown, RI.

Doerschuk, B. (1984). The literature of improvisation. Keyboard 10 (Oct.), 48–52.

Dupré, M. (1925/37). Cours complet d’improvisation a’ l’orgue, 2 vols. A. Leduc,

Paris.

Easton, T. A. (1978), Coordinative structure—the basis for a motor program. In

Psychology of’motor behaviour and sport (ed. D. M. Landers and R. W.
Christina). Human Kinetics, Champaign, Ill.

Eccles, J. C. (1972). The understanding of the brain, McGraw-Hill, New York.

Ehmann, Wilhelm (1950). Chorische Improvisation in der Kantorei. Kirchenchor

10 (Sept.–Oct.), 65–71.

Page 46 of 58

Erlmann, V. (1985). Model, variation and performance, Ful’be praise song in

Northern, Cameroon. Yearbook for traditional music 17, 88–112.

(p.171) Ewing, A. (1941). Reason and intuition. Proceedings of the British

academy 27, 67–107.

Ferand, Ernst (1938). Die Improvisation in der Musik. Rhein-Veriag, Zurich.

Ferand, Ernst (1961). Improvisation in nine centuries of western music, Arno

Volk Verlag, Hans Gerig KG, Cologne.

Fitts, P. M. (1964). Perceptual-motor skill learning. In Categories of human

learning (ed. A. W. Melton). Academic Press, New York.

Fry, C. (1980), Computer improvisation. Computer Music Journal 4 (3), 48 ff.

Fry, C. (1982/83). Dancing musicians. Perspectivies of New Music, Fall-Winter

1982/Spring-Summer 1983, 585–9.

Gaburo, K. (1968). Twenty sensing compositions. Lingua Press, La Jolla, Calif.

Galás, D. (1981/82). Intravenal song. Perspectives of New Music, Fall-Winter

1981/Spring-Summer 1982, 59–62.

Gehring, P. K. (1963). Improvisation in contemporary organ playing. Unpublished

Ph.D. thesis Syracuse University.

Gelernter, H. (1963). Realisation of a geometry-theorem-proving machine. In

Computers and thought, (ed. E. A. Feigenbaum and J. Feldman). McGraw-Hill,
New York.

Gel’fand, I. M. and Tsetlin, M.L. (1962). Some methods of control for complex
systems. Russian Mathematical Surveys 17, 95–116.

Gel’fand, I. M. and Tsetlin, M. L. (1971). Mathematical modelling of mechanisms

of the central nervous system. In Models of the structural-functional
organisation of certain biological systems (eds. I. M. Gel’fand, V. S. Gurfinkel, S.
V. Fomin, and M. T. Tsetlin). MIT Press, Cambridge, Mass.

Gibbs, C. B. (1965). Probability learning in step-input tracking. British Journal of

Psychology 56, 233–42.

Gibbs, C. B. and Brown, I. C. (1956). Increased production from information

incentives in an uninteresting repetitive task. Manager 24, 374–9.

Gibson, J. J. (1966). The senses considered as perceptual systems. Houghton

Mifflin, Boston.

Page 47 of 58

Gibson, J. J. (1979). The ecological approach to visual perception. Houghton

Mifflin, Boston.

Glencross, D. J. (1977). Control of skilled movements. Psychological Bulletin 84,

14–29.

Glencross, D. J. and Koreman, M. M. (1979). The processing of proprioceptive

signals. Neuropsychologia 17, 683–7.

Goldstein, M. (1983). The gesture of improvisation: some thoughts, reflections

and questions regarding percussion music. Percussionist 21, (3), 18–24.

Gorder, W. D. (1976) An investigation of divergent production abilities as

constructs of musical creativity. Unpublished Ed. D. thesis. University of Illinois
at Urbana-Champaign.

Greene, P. M. (1972). Problems of organisation of motor systems. In Progress in

theoretical biology, Vol. 2 (eds. R. Rosen and F. M. Snell). Academic Press, New
York.

Guevara, M., Glass, L., Mackey, M., and Shrier, A. (1983). Chaos in neurobiology.
IEEE Transactions on systems, Man, and Cybernetics SMC-13, (5), 790–8.

(p.172) Guildford, J. P. (1977), Way beyond the I.Q. Creative Education

Foundation, New York.

Guildford, J. P. and Hoepfner, R. (1971). The analysis of intelligence, McGraw-

Hill, New York.

Hamel, P. M. (1979). Through music to the self. Shambhala, Boulder, Col.

Hanlon, E. S. (1975). Improvisation: theory and application for theatrical music

and silent film. Unpublished Ph.D. thesis. University of Cincinnati.

Hayes-Roth, B. (1977). Evolution of cognitive structures and processes.

Psychological Review 84, (3), 260–78.

Hebb, D. D. (1949). Organisation of behaviour. Wiley, New York.

Hendrickson, G. and Schroeder, W. H. (1941). Transfer of training in learning to

hit a submerged target. Journal of Educational Psychology 32, 205–13.

Hick, W. E. (1952). On the rate of gain of information. Quarterly Journal of

Experimental Psychology 4, 11–26.

Higgins, J. R. and Angel, R. W. (1970). Correction of tracking errors without

sensory feedback. Journal of Experimental Psychology 84, 412–16.

Page 48 of 58

Hodeir, Andre (1956). Jazz: its evolution and essence (trans. David Noakes).
Grove, New York.

Holding, D. H. (1965). Principles of training, Pergamon, London.

Holding, D. (1981). Skills research. In Human skills (ed. D. Holding). Wiley,

Chichester.

Hollerbach, J. M. (1981). An oscillation theory of handwriting. Biological

Cybernetics 39, 139–56.

Hood, M. (1971). Aspects of group improvisation in the Javanese gamelan. In

Musics of Asia (ed. José Maceda), pp. 17–21. Manila.

Hood, M. (1975). Improvisation in the stratified ensembles of Southeast Asia.

Selected Reports in Ethnamusicology (UCLA) 2, (2), 25–33.

Hores, R. G. (1977). A comparative study of visual- and aural-orientated

approaches to Jazz improvisation with implications for instruction. Unpublished
Ed. D. thesis. Indiana University.

Howard, J. (1978). Improvisational techniques of Art Tatum. Unpublished Ph.D.

thesis. Case Western Reserve University.

Huggins, A. W. F. (1978). Speech timing and intelligibility. In Attention and

performance VII (ed. J. Requin). Lawrence Erlbaum, Hillsdale, NJ.

Jairazbhoy, N. A. (1971). The rags of North Indian music. Faber and Faber,
London.

Jaques-Dalcroze, E. (1921/1976). Rhythm, music and education. B. Blom, New

York.

Jaques-Dalcroze, E. (1930). Eurythmics, art, and education. Ayer, New York.

James, W. (1890). Principles of psychology, Vol. 1. Holt, New York.

Jones, A. M. (1959). Studies in African Music. Oxford, London.

Jost, E. (1974). Free jazz. Universal, Graz.

Keele, S. W. and Posner, M. I. (1968). Processing of feedback in rapid

movements. Journal of Experimental Psychology 77, 353–63.

Keele, S. W. and Summers, J. J. (1976). The structure of motor programs. In

Motor control; issues and trends (ed. G. E. Stelmach). Academic Press, New
York.

Page 49 of 58

Kelso, J. A. S. (1982). Two strategies for investigating action. In Human motor

(p.173) behaviour (ed. J. A. S. Kelso). Lawrence Erlbaum, Hillsdale, NJ.

Kelso, J. A. S. and Wallace, S. A. (1978). Conscious mechanisms in movement. In

Information processing in motor control and learning (ed. G. E. Stelmach),
Academic Press, New York.

Kelso, J. A., Holt, J. G., Rubin, P., and Kugler, P. N. (1981). Patterns of human
interlimb co-ordination emerge from the properties of non-linear, limit cycle
oscillatory processes: theory and data. Journal of Motor Behaviour 13, 226–61.

Kernfeld, B. (1983). Two Coltranes. Annual Review of Jazz Studies 2, 7–66.

Kerr, R. (1982). Psychomotor learning. Saunders, Philadelphia.

Kickert, W. G., Bertrand, J. W., and Praagman, J. (1978). Some comments on

cybernetics and control. IEEE transactions on systems, man and cybernetics
SMC-8, 805–9.

Kleeman, J. E. (1985/86). The parameters of musical transmission. The Journal of

Musicology 4, 1–22.

Kleinman, D. L., Baron, S., and Levison, W. H. (1971). A control theoretic

approach to manned-vehicle systems analysis. IEEE Transactions on Automatic
Control AC-16, 824–32.

Kugler, P. N., Kelso, J. A. S., and Turvey, M. T. (1980). On the concept of co-
ordinative structures as dissipative structures; I. theoretical lines of
convergence. In Tutorials in motor behaviour (ed. G. E. Stelmach and J. Requin).
North-Holland, Amsterdam.

Laabs, G. J. and Simmons, R. W. (1981). Motor memory. In Human skills (ed. D.

Holding). Wiley, New York.

Laneri, Roberto (1975). Prima materia: an opus in progress. The natural

dimension of music. Unpublished Ph.D. thesis. University of California at San
Diego.

Lenat, D. B. (1984). Computer software for intelligent systems, Scientific

American 251, (3), 152–60.

Leonard, J. A. (1959). Tactual choice reactions. Quarterly Journal of

Experimental Psychology 11, 76–83.

Levitt, D. A. (1981). A melody description system for jazz improvisation.

Unpublished M.S. thesis. MIT.

Page 50 of 58

Li, T. Y. and Yorke, J. (1975). Period three implies chaos. American Mathematical
Monthly 82, 985–92.

Liebman, D., Beirach, R., Tusa, F., Williams, J., and Roy, B. (1978). Lookoutfarm,
Almo Publications, Hollywood.

Lipiczky, T. (1985). Tihai formulas and the ‘composition’ and improvisation in

North Indian music. The Musical Quarterly 71, 157–71.

Locke, D. (1979). The music of atsiabeko. Unpublished Ph.D. thesis. Wesleyan

University.

Lord, A. B. (1964). The singer of tales. Harvard University Press, Cambridge,

Mass.

Lord, A. B. (1965). Yugoslav epic folk poetry. In The study in folklore (ed. A.
Dundes). Prentice-Hall, Englewood Cliffs, NJ.

McCulloch, W. S. (1945). A heterarchy of values determined by the topology of

nervous nets. Bulletin of Mathematical Biophysics 7, 89–93.

Marteniuk, R. G. and Romanow, S. K. E. (1983). Human movement organisation

and learning as revealed by variability of movement, use of kinematic (p.174)
information, and Fourier analysis. In Memory and control of action (ed. R. A.
Magill), North-Holland, Amsterdam.

Mathieu, L. (1984). A phenomenological investigation of improvisation in music

and dance. Unpublished Ph.D. thesis, New York University.

May, R. M. (1976). Simple mathematical models with very complicated

dynamics. Nature 261, 459–67.

Mersenne, M. (1635). Harmonic Universelle.

Milano, D. (1984). The psychology of improvisation. Keyboard 10 (Oct.), 25, 30–

Miller, G. A., Galanter, E., and Pribram, K. (1960). Plans and the structure of
behaviour. Holt, Rinehart and Winston, New York.

Milner, B. (1962). Les troubles de la memoire accompagnant des lesions

hippocampiques bilaterales. In Phsyiologie de l’Hippocampe. Cent. Natl.
Recherche Scientifique, Paris.

Morasso, P. (1983). Three dimensional arm trajectories. Biological Cybernetics

48, 187–94.

Page 51 of 58

Murphy, F. (1982). The cornet style of Leon Bix Beiderbecke (1903–1931).

Unpublished Ph.D. thesis. La Trobe University, Melbourne.

Nagler, M. N. (1974). Spontaneity and tradition: a study in the oral art of Homer.
University of California Press, Berkeley.

Namikas, G. (1983), Vertical process and motor performance. In Memory and

control of action (ed. R. A. Magill). North-Holland, Amsterdam.

Nelson, O. (1966). Patterns for improvisation. Nelson Music, Hollywood.

Nettl, B. (1974). Thoughts on improvisation. Musical Quarterly 60, 1–19.

Nettl, B. and Foltin, B. Jr. (1972). Daramad of Chahargah. Information Co-

ordinators, Detroit.

Nettl, B. and Riddle, R. (1974). Taqsim Nahawand: a study of 16 performances

by Jihad Racy. Yearbook of the International Folk Music Council 5, 11–50.

Newell, K. M. (1981). Skill learning. In Human skills (ed. D. Holding). Wiley,

Chichester.

Oliveros, P. (1971). Sonic meditations. Smith Publications/Sonic Art Editions,

Baltimore, Md.

Owens, T. (1974). Charlie Parker: techniques of improvisation, 2 Vols.

Unpublished Ph.D. thesis. University of California at Los Angeles.

Paillard, J. (1980). The multichanneling of visual cues and the organisation of

visually guided response. In Tutorials in motor behaviour (eds. G. E. Stelmach
and J. Requin). North-Holland, Amsterdam.

Park, M. (1985). Music and shamanism in Korea: a study of selected ssikkŭm-gut

rituals for the dead. Unpublished Ph.D. thesis, University of California at Los
Angeles.

Parry, M. (1930). Studies in the epic technique of oral versemaking: I. Homer

and Homeric style. Harvard Studies in Classical Philology 41, 73–147.

Parry, M. (1932). Studies in the epic technique of oral versemaking: II. the
Homeric language as the language of poetry. Harvard Studies in Classical
Philology 43, 1–50.

Parsons, W. (1978). Music for citizen’s band. W. Parsons, La Jolla, Calif.

Partchey, K. C. (1973). The effects of feedback, models, and repetition on the (p.
175) ability to improvise melodies. Unpublished D.Ed. thesis. Pennsylvania
State University.

Page 52 of 58

Pew, R. W. (1974). Human perceptual-motor performance. In Human information

processing: tutorials in performance and cognition (ed. B. H. Kantowitz).
Erlbaum, NJ.

Pew, R. W. and Baron, S. (1978). The components of an information processing

theory of skilled performance based on an optimal control perspective. In
Information processing motor control and learning (ed. G. E. Stelmach).
Academic Press, New York.

Pike, A. (1974). A Phenomenology of jazz. Journal of Jazz Studies 2, (1), 88–94.

Poulton, E. C. (1957). On the stimulus and response in pursuit tracking. Journal

of Experimental Psychology 53, 57–65.

Powers, W. T. (1973). Behaviour: the control of perception. Aldine, Chicago.

Pressing, J. (1980). Music, altered states of consciousness, and psi. In

Proceedings of the Psychic Orientation Conference (ed. A. Gabay). La Trobe
University, Melbourne.

Pressing, J. (1984a). Cognitive processes in improvisation. In Cognitive

processes in the perception of art (ed. W. R. Crozier and A. J. Chapman), pp. 345–
63. North-Holland, Amsterdam.

Pressing, J. (1984b). A history of musical improvisation to 1600. Keyboard 10

(11), 64–8.

Pressing J. (1984c). A history of musical improvisation: 1600–1900. Keyboard 10,

(12), 59–67.

Prigogine, I. (1967). Introduction to thermodynamics of irreversible processes.

Interscience, New York.

Prigogine, I. and Nicholis, G. (1971). Biological order, structure and instabilities.

Quarterly Review of Biophysics 4, 107–48.

Propp, V. I. (1927/1968). Morphology of the folk tale (trans. L. Scott). University

of Texas Press, Austin.

Quantz, J. J. (1752/1966). On playing the flute (trans. E. Reilley). Free Press, New
York.

Rabbitt, P. M. A. and Vyas, S. M. (1970). An elementary preliminary taxonomy for

some errors in laboratory choice RT tasks. Acta Psychologica 33, 56–76.

Radano, R. M. (1985). Anthony Braxton and his two musical traditions, the
meeting of concert music and jazz (2 vols). Unpublished Ph.D. thesis, University
of Michigan.

Page 53 of 58

Reck, D. B. (1983). A musician’s tool-kit: a study of five performances by

Thirugokarnam Ramachandra Iyer, 2 Vols. Unpublished Ph.D. thesis. Wesleyan
University.

Rich, E. (1983). Artificial intelligence. McGraw-Hill, New York.

Richardson, C. P. (1983). Creativity research in music education: a review.

Council for Research in Music Education 74, (Spring 1983), 1–21.

Roads, C. (1979). Grammars as representations for music. Computer Music

Journal 3, 48–55.

Ryle, G. (1949). The concept of mind. Hutchinson, London.

Sadie, S. (ed.) (1980). The new Grove dictionary of music and musicians.
Macmillan, London. Listings under improvisation, aleatory, cadenza, continuo,
division. prélude non mesuré.

(p.176) Sage, G. S. (1977). Introduction to motor behaviour: a

neuropsychological approach, Addison-Wesley, Reading, Mass.

Saltzman, E. L. and Kelso, J. A. S. (1983). Toward a dynamical account of motor

memory and control. In Memory and control of action (ed. R. A. Magill). North-
Holland, Amsterdam.

Schickhaus, K-H. (1978). Neues Schulwerk für Hackbrett. Josef Preissler,

Munich.

Schmidt, R. A. (1983). On the underlying response structure of well-learned

motor responses: a discussion of Namikas and Schneider and Fisk. In Memory
and control of action (ed. R. A. Magill). North-Holland, Amsterdam.

Schneider, W. and Fisk, A. D. (1983). Attention theory and mechanisms for

skilled performance. In Memory and control of action (ed. R. A. Magill). North-
Holland, Amsterdam.

Schouten, H. (no date given). Improvisation on the organ (trans. by J. L. Warren).

W. Paxton, London.

Schuller, G. (1968). Early jazz: its roots and musical development, Oxford
University Press, New York.

Serafine, M. L. (1983), Cognition in music. Cognition 14, 119–83.

Shafer, R. M. (1969). Ear cleaning. Universal, London.

Shaffer, L. H. (1978). Timing in the motor programming of typing. Quarterly

Journal of Experimental Psychology 30, 333–45.

Page 54 of 58

Shaffer, L. H. (1980). Analysing piano performance: a study of concert pianists.

In Tutorials in motor behaviour (ed. G. E. Stelmach and J. Requin). North-
Holland, Amsterdam.

Shaffer, L. H. (1981). Performances of Chopin, Bach and Bartok: studies in motor

programming. Cognitive psychology 13, 327–76.

Shaffer, L. H. (1984). Timing in solo and duet piano performances. Quarterly

Journal of Experimental Psychology A36, 577–95.

Shepherd, G. M. (1983). Neurobiology. Oxford University Press, New York.

Sher, C. (1983). The world’s greatest fake book. Sher Music, San Francisco.

Shiffrin, R. M. and Schneider, W. (1977). Controlled and automatic human

information processing: II. perceptual learning, automatic attending, and a
general theory. Psychological Review 84, 127–90.

Signell, K. (1974). Esthetics of improvisation in Turkish art music. Asian Music 5,

(2), 45–9.

Signell, K. (1977). Makam: modal practice in Turkish art music, Asian music
Publications, Seattle.

Silverman, M. L. (1962) Ensemble improvisation as a creative technique in the

secondary instrumental program. Unpublished Ed.D. thesis. Stanford University.

Sloboda, J. A. (1982). Music performance. In Psychology of music (ed. D.

Deutsch). Academic Press, New York.

Slonimsky, N. (1975). Thesaurus of scales and melodic patterns. Duckworth,

London.

Smith, G. E. (1983). Homer, Gregory, and Bill Evans? the theory of formulaic
composition in the context of jazz piano improvisation. Unpublished Ph.D. thesis,
Harvard University.

Smith, W. M. and Bowen, K. F. (1980). The effects of delayed and displaced

visual feedback on motor control. Journal of Motor Behaviour 12, 91–101.

(p.177) Sparrow, W. A. (1983), The efficiency of skilled performance. Journal of

Motor Behaviour 15, 237–61.

Sperber, M. (1974). Improvisation in the performing arts: music, dance and

theatre. Unpublished Ed.D. thesis, Columbia University.

Squire, L. R. (1982). The neuropsychology of human memory. Annual Review of

Neuroscience 5, 241–73.

Page 55 of 58

Stocks, J. (1939). Reason and intuition (ed. D. M. Emmett). Oxford University

Press, New York.

Stumme, W. (1972). über improvisation, B. Schott, Mainz.

Sudnow, D. (1978). Ways of the hand. Harvard University Press, Cambridge,

Mass.

Suehs, H. C. (1979). The development, implementation, and assessment of a

course of study for instruction in certain improvisational techniques in the
performance of Baroque music from 1679 to 1741. Unpublished Ph.D. thesis,
Catholic University of America at Washington.

Sumarsam (1981). The musical practice of the gamelan sekaten. Asian Music 12,
(2), 54–73.

Summers, J. J. (1981). Motor programs. In Human skills (ed. D. Holding), Wiley,

Chichester.

Suzuki, S., Mills, E. and Murphy, T. (1973). The Suzuki concept: an introduction
to a successful method for early music education. Diablo Press, Berkeley, Calif.

Terzuolo, C. and Viviani, P. (1979). About the central representation of learned

motor patterns. In Posture and movement (eds. R. Talbot and D. R. Humphrey).
Raven, New York.

Thompson, S. (1946). The folktale. Holt, Rinehart and Winston, New York.

Torrance, E. P. (1966) Guiding creative talent. Prentice-Hall, Englewood Cliffs,

NJ.

Touma, H. H. (1971). The maqam phenomenon: an improvisation technique in

the music of the middle east. Ethnomusicology 15, 38–48.

Treitler, L. (1974). Homer and Gregory: the transmission of epic poetry and
plainchant. Musical Quarterly 9, 333–72.

Turvey, M. T. (1977). Preliminaries to a theory of action with reference to vision.

In Perceiving, acting and knowing; toward an ecological psychology (eds. R.
Shaw and J. Bransford). Lawrence Erlbaum, Hillsdale, NJ.

Ueltzen, D. R. (1986). Improvisation in Kirchengesang und mit dem Kirchenlied

und am Gesangbuch für das Jahr 2000. Gottesdienst und Kirchenmusik 3, 71–7.

Vaughan, M. M. (1971). Music as model and metaphor in the cultivation and

measurement of creative behaviour in children. Unpublished Ed.D. thesis.
University of Georgia.

Page 56 of 58

VĪkis-Fribergs, V. (1984). Creativity and tradition in oral folklore or the balance

of innovation and repetition in the oral poet’s art. In Cognitive processes in the
perception of art (eds. W. R. Crazier and A. J. Chapman). North-Holland,
Amsterdam.

Viviani, P. and Terzuolo, C. (1980), Space-time invariance in learned motor skills.

In Tutorials in motor behaviour (eds. G. E. Stelmach and J. Requin). North-
Holland, Amsterdam.

von Foerster, H. (1960). On self-organising systems and their environments. In

(p.178) Self-organising systems (ed. M. C. Yorits and S. Cameron). Pergamon,
New York.

Von Holst, E. (1954). Relations between the central nervous system and the
peripheral organs. British Journal of Animal Behaviour 2, 89–94.

Wade, Bonnie (1973). Chiz in Khyal: the traditional composition in the

improvised performance. Ethnomusicology 17, 443–59.

Webster, R. P. (1977). A factor of intellect approach to creative thinking in music.

Unpublished Ph.D. thesis. University of Rochester, Eastman School of Music.

Weidner, R. F. (1984). The improvisation techniques of Charles Tournemire as

extracted from his five reconstructed organ improvisations. Unpublished Ph.D.
thesis, Michigan State University.

Welford, A. T. (1974). On the sequencing of action. Brain Research 71, 381–92.

Welford, A. T. (1976). Skilled performance, Scott, Foresman, Glenview, Ill.

West, L. J. (1967). Vision and kinesthesis in the acquisition of typewriting skill.

Journal of Applied Psychology 51 (2), 161–6.

Westcott, M. R. (1968). Towards a contemporary psychology of intuition, Holt,

Reinhart and Winston, New York.

Whitmer, T. (1934). The art of improvisation. M. Witmark and Sons, New York.

Wilberg, R. B. (1983). Memory for movement: discussion of Adams and Saltzman

and Kelso. In Memory and control of action (ed. R. A. Magill). North-Holland,
Amsterdam.

Winograd, T. (1975). Frame representations and the declarative-procedural

controversy. In Representation and understanding (ed. D. Bobrow and A.
Collins). Academic Press, New York.

Zelaznik, H. N., Hawkins, B., and Kisselburgh, L, (1983). Rapid visual feedback
processing in single-arming movements. Journal of Motor Behaviour 15, 217–36.

Page 57 of 58

Zonis, Ella (1973). Classical Persian music: an introduction. Harvard University

Press, Cambridge, Mass.

Notes:
(*) It is interesting to note that these two skills push in opposite directions, for
inventiveness comes from the commitment to avoid repetition as much as
possible, while coherence is only achieved by some degree of structural unity,
which is only possible with repetition.

Access brought to you by:

Page 58 of 58

Jeff Pressing 1988 Improv-Methods
No ratings yet
Jeff Pressing 1988 Improv-Methods
50 pages
(Cambridge Companions to Music) Martin Iddon (Editor) - The Cambridge Companion to Serialism-Cambridge University Press (2023)
No ratings yet
(Cambridge Companions to Music) Martin Iddon (Editor) - The Cambridge Companion to Serialism-Cambridge University Press (2023)
462 pages
Thoegersen Peter Alexander Polytempic Polymicrotonal Music
No ratings yet
Thoegersen Peter Alexander Polytempic Polymicrotonal Music
188 pages
Short Note Biology Form 5-Chapter 1 Transport
83% (6)
Short Note Biology Form 5-Chapter 1 Transport
7 pages
Preparation Strategies in Percussion For The Music of Iannis Xenakis and Other Composers
No ratings yet
Preparation Strategies in Percussion For The Music of Iannis Xenakis and Other Composers
64 pages
Boulez, Pierre, Alea PDF
No ratings yet
Boulez, Pierre, Alea PDF
13 pages
Schoenberg's Farben - An Analysis of Op. 16, No. 3
100% (1)
Schoenberg's Farben - An Analysis of Op. 16, No. 3
33 pages
Gulu University Dipoma Entry Scheme Government Admission List 2024 - 2025
No ratings yet
Gulu University Dipoma Entry Scheme Government Admission List 2024 - 2025
3 pages
Lindberg Related Rocks-2 PDF
No ratings yet
Lindberg Related Rocks-2 PDF
106 pages
Luciano Berio - Sequenza XIII (Chanson) PDF
No ratings yet
Luciano Berio - Sequenza XIII (Chanson) PDF
7 pages
Giacinto Scelsi
50% (2)
Giacinto Scelsi
6 pages
Story Wheel Lesson Plan
No ratings yet
Story Wheel Lesson Plan
4 pages
Funda Quiz Duma
No ratings yet
Funda Quiz Duma
15 pages
Samir Amin
No ratings yet
Samir Amin
5 pages
Stockhausen Punkte
No ratings yet
Stockhausen Punkte
60 pages
LLaMA Open and Efficient Foundation Language Models
No ratings yet
LLaMA Open and Efficient Foundation Language Models
27 pages
Grammatical Names and Roles
100% (1)
Grammatical Names and Roles
10 pages
Spain National Football Team
No ratings yet
Spain National Football Team
24 pages
Kampala Campus Attendance Firm Lists Ay 2024-2025 RM
No ratings yet
Kampala Campus Attendance Firm Lists Ay 2024-2025 RM
54 pages
ITR-25
No ratings yet
ITR-25
1 page
Writing For Saxophones
100% (1)
Writing For Saxophones
2 pages
Aleatory
No ratings yet
Aleatory
7 pages
ALL SEMESTER MARKSHEET (ALL IN ONE) - Compressed - Rem675757oved
100% (1)
ALL SEMESTER MARKSHEET (ALL IN ONE) - Compressed - Rem675757oved
1 page
(Solution manual) Elementary Surveying An Introduction to Geomatics 15th Edition download
100% (1)
(Solution manual) Elementary Surveying An Introduction to Geomatics 15th Edition download
38 pages
An Interview With Olivier Messiaen
100% (1)
An Interview With Olivier Messiaen
8 pages
Robin Holloway's 'Second Concerto For Orchestra': A Masterpiece Bridging Romanticism and Modernism
No ratings yet
Robin Holloway's 'Second Concerto For Orchestra': A Masterpiece Bridging Romanticism and Modernism
29 pages
Download Complete (Original PDF) Meteorology Today 11th Edition by C. Donald Ahrens PDF for All Chapters
100% (1)
Download Complete (Original PDF) Meteorology Today 11th Edition by C. Donald Ahrens PDF for All Chapters
30 pages
Notas de Necessitá D'interrogare Il Cielo
No ratings yet
Notas de Necessitá D'interrogare Il Cielo
22 pages
PMI Training Brochure
No ratings yet
PMI Training Brochure
11 pages
Companies That Hire Felons List
No ratings yet
Companies That Hire Felons List
4 pages
Ism Class Xi Marketing (812) (Fe) Qp & Ms (22-23) Set A
No ratings yet
Ism Class Xi Marketing (812) (Fe) Qp & Ms (22-23) Set A
13 pages
Evolution of Tonal Organiztion PDF
100% (1)
Evolution of Tonal Organiztion PDF
36 pages
Steven Stucky, Contemporary Music
No ratings yet
Steven Stucky, Contemporary Music
6 pages
CAD Viewers
No ratings yet
CAD Viewers
24 pages
VPLEX - How To Reboot The VPLEX Management Server. - Dell Singapore
No ratings yet
VPLEX - How To Reboot The VPLEX Management Server. - Dell Singapore
3 pages
Motive and Spatialization in Thomas Tallis Spem in Alium
100% (1)
Motive and Spatialization in Thomas Tallis Spem in Alium
55 pages
Ligeti - African Polyrhythm
100% (1)
Ligeti - African Polyrhythm
13 pages
1res I STA NCE 4 S WI TCH 6 D I ODE 9 7 AMP 3mult IME T E R 9volt 5 Cond E Nser 8battery
No ratings yet
1res I STA NCE 4 S WI TCH 6 D I ODE 9 7 AMP 3mult IME T E R 9volt 5 Cond E Nser 8battery
3 pages
Musical Styles Since 1945
No ratings yet
Musical Styles Since 1945
7 pages
A THEORETICAL AND PERFORMANCE ANALYSIS OF TWO WORKS FOR SAXOPHONE QUARTET: Xenakis and Donatoni
No ratings yet
A THEORETICAL AND PERFORMANCE ANALYSIS OF TWO WORKS FOR SAXOPHONE QUARTET: Xenakis and Donatoni
70 pages
17 The Book of Secrets
100% (1)
17 The Book of Secrets
3 pages
The Music of The Environment: For Making Mu Performances Orga A N
No ratings yet
The Music of The Environment: For Making Mu Performances Orga A N
6 pages
Deutz Fahr Agrotron K 90 100 110 120 Profiline Workshop Manual
98% (66)
Deutz Fahr Agrotron K 90 100 110 120 Profiline Workshop Manual
20 pages
Lei Liang - Harp Concerto
No ratings yet
Lei Liang - Harp Concerto
79 pages
Class-Vi Final Term Datesheet & Curriculum
No ratings yet
Class-Vi Final Term Datesheet & Curriculum
4 pages
VOID CONTRACT SEC 517 (2a) Clause 1,2,3 &4 517 (2b) 517 (2c) by Jeevan Regmi
No ratings yet
VOID CONTRACT SEC 517 (2a) Clause 1,2,3 &4 517 (2b) 517 (2c) by Jeevan Regmi
2 pages
J Msea 2016 06 014
No ratings yet
J Msea 2016 06 014
6 pages
APznzaYy7PKfy EppSwN1yjagnEJ3p8aZTvkccuJxRGMREtLQLv 2a9vTyybwMjJAL08gDzWxFRxrJb1HqRhiYMN1U6wwy49i5c4 RUl4GF16ByPTBRPHNdlDjcDG7oeg0Vx5Hdrb4wM06KcjMZhn1nUA5 FZlivv38GjahGbjt7koRTYVnJEV6H0HUZTrx8r6ReOfc0yB1XJQ1CqXFiCMW
No ratings yet
APznzaYy7PKfy EppSwN1yjagnEJ3p8aZTvkccuJxRGMREtLQLv 2a9vTyybwMjJAL08gDzWxFRxrJb1HqRhiYMN1U6wwy49i5c4 RUl4GF16ByPTBRPHNdlDjcDG7oeg0Vx5Hdrb4wM06KcjMZhn1nUA5 FZlivv38GjahGbjt7koRTYVnJEV6H0HUZTrx8r6ReOfc0yB1XJQ1CqXFiCMW
40 pages
Other Notes On John Zorn's Cobra
No ratings yet
Other Notes On John Zorn's Cobra
3 pages
13.G.R. No. L-24772 May 27, 1968
No ratings yet
13.G.R. No. L-24772 May 27, 1968
2 pages
Tenney - About Changes
100% (1)
Tenney - About Changes
25 pages
Template For Lesson Plan MCT S
No ratings yet
Template For Lesson Plan MCT S
3 pages
Cartography Fact Sheet 4.0 PDF
No ratings yet
Cartography Fact Sheet 4.0 PDF
2 pages
Notation Perform A 00 Will
100% (1)
Notation Perform A 00 Will
190 pages
Young, La Monte & Marian Zazeela (1969) - Selected Writings 1959-1969.
No ratings yet
Young, La Monte & Marian Zazeela (1969) - Selected Writings 1959-1969.
78 pages
Words As Music Notation, Nielsen
100% (1)
Words As Music Notation, Nielsen
18 pages
Extended Techniques For Bass Guitar - Summary Sheet: by Philip Langshaw (z5316375)
No ratings yet
Extended Techniques For Bass Guitar - Summary Sheet: by Philip Langshaw (z5316375)
1 page
Bok, Wendel - Techniques of Bass Clarinet Playing (New Techniques For The Bass Clarinet) - Henri Bok
No ratings yet
Bok, Wendel - Techniques of Bass Clarinet Playing (New Techniques For The Bass Clarinet) - Henri Bok
91 pages
Solved Paris Participates in Her Employer S Nonqualified Deferred Compensation Plan For
No ratings yet
Solved Paris Participates in Her Employer S Nonqualified Deferred Compensation Plan For
1 page
"When They Are Dying, Men Sing... ": Nono's Diario Polacco N. 2
No ratings yet
"When They Are Dying, Men Sing... ": Nono's Diario Polacco N. 2
7 pages
Constitutive Silences
100% (1)
Constitutive Silences
60 pages
Sebutkan 10 Nama Hewan Beserta Ciri-cirinya(Minimal 3 Ciri-ciri) Dalam Bhs Inggris - Brainly.co.Id
No ratings yet
Sebutkan 10 Nama Hewan Beserta Ciri-cirinya(Minimal 3 Ciri-ciri) Dalam Bhs Inggris - Brainly.co.Id
1 page
Kahn 1995
100% (1)
Kahn 1995
4 pages
ATS2122 Jazz Tech 2019
No ratings yet
ATS2122 Jazz Tech 2019
6 pages
Music of John Cage
No ratings yet
Music of John Cage
16 pages
Subharmonics - A Revolutionary Technique For The Violin
No ratings yet
Subharmonics - A Revolutionary Technique For The Violin
5 pages
Pitch and Register Edgard Varese
No ratings yet
Pitch and Register Edgard Varese
26 pages
Appropriation, Culture and Meaning in
0% (1)
Appropriation, Culture and Meaning in
8 pages
Picclo Fingering 3
No ratings yet
Picclo Fingering 3
3 pages
01 Cook N 1803-2 PDF
100% (1)
01 Cook N 1803-2 PDF
26 pages
An Examination of Minimalist Tendencies in Two Early Work
100% (1)
An Examination of Minimalist Tendencies in Two Early Work
24 pages
Scratch Music - Readings
No ratings yet
Scratch Music - Readings
10 pages
Anthony Braxton Notes
No ratings yet
Anthony Braxton Notes
7 pages
Analytical Studies in World Music, Introduction
No ratings yet
Analytical Studies in World Music, Introduction
40 pages
Composing Sound Structures With Rules
No ratings yet
Composing Sound Structures With Rules
14 pages
DEAR ABBY (Score) - Nina Shekhar
No ratings yet
DEAR ABBY (Score) - Nina Shekhar
44 pages
Oggetti I: Jason Thorpe Buchanan
No ratings yet
Oggetti I: Jason Thorpe Buchanan
14 pages
Azathoth - J.O'Connor
No ratings yet
Azathoth - J.O'Connor
19 pages
Bryars: "1, 2, 1-2-3-4" Notes
No ratings yet
Bryars: "1, 2, 1-2-3-4" Notes
3 pages
Winslow. The Psycology of Musical Memory PDF
100% (1)
Winslow. The Psycology of Musical Memory PDF
3 pages
Berio para Estudiar
No ratings yet
Berio para Estudiar
6 pages
Roig Francoli - Hexachordal Combinatoriality
No ratings yet
Roig Francoli - Hexachordal Combinatoriality
11 pages
Lesson Plan Graphic Notation
No ratings yet
Lesson Plan Graphic Notation
8 pages
Artificial Life 2007: (George Lewis 2007) Page One
No ratings yet
Artificial Life 2007: (George Lewis 2007) Page One
1 page
Lutoslawski Intrada
No ratings yet
Lutoslawski Intrada
3 pages
Self-Organizing Systems, 1963
From Everand
Self-Organizing Systems, 1963
James Emmett Garvey
No ratings yet
Kinematics of the Brain Activities
From Everand
Kinematics of the Brain Activities
Mostafa M. Dini
No ratings yet
Brain Functioning and Regeneration: Kinematics of the Brain Activities Volume Iv
From Everand
Brain Functioning and Regeneration: Kinematics of the Brain Activities Volume Iv
Mostafa M. Dini
No ratings yet
Outside and Inside: Race and Identity in White Jazz Autobiography
From Everand
Outside and Inside: Race and Identity in White Jazz Autobiography
Reva Marin
No ratings yet
Neuropsychology: The basics of the matter
From Everand
Neuropsychology: The basics of the matter
Stefano Calicchio
No ratings yet

Note: This service is not intended for secure transactions such as banking, social media, email, or purchasing. Use at your own risk. We assume no liability whatsoever for broken pages.