Consciousness, Literature and the Arts
Archive
Volume 1 Number 1, April 2000
_______________________________________________________________
Creativity,
Structure and Randomness
by
The
study of creative improvisation necessarily involves a variety of disciplines
and a range of analytic levels. A
paradigm known as dynamic systems theory (DST)
is currently in vogue as a means of describing motor learning and motor control.
It is suggested that elements of this approach may be relevant to the
task of organizing the wealth of research related to improvisation.
Particularly useful is the concept of arranging processes in terms of
their position on a spectrum of structure,
ranging from determinism to anarchy. A
further assertion is that most ‘interesting’ behaviour lies at an
intermediate position, the so-called edge-of-chaos.
Ideas such as these are in harmony with much of the creativity studies
literature; the image of a chaotic
edge serves as a reminder that both deterministic and random elements are
essential to the process of innovation.
Interdisciplinary
studies are dangerous opportunities. The
danger resides in the ongoing potential for miscommunication.
Or worse, conceptual mismatch hidden beneath a layer of linguistic
similarity. The pitfalls of
homography are familiar to language learners;
words like sensible are easily confused in translating between English and
French. They are referred to in the
ESL jargon as ‘false friends’.
The
opportunity resides in the unifying power of analogy.
Chaos was initially considered an obscure property of the mathematics of
topology. However, its utility as
an organizing framework was slowly recognized during the 1960’s and 70’s.
Formal aspects of the definition (e.g. sensitive dependence on initial
conditions) were adapted to the description of dynamical systems ranging from
ecology to the weather. Chaos was
to become a key player in a conceptual revolution, involving the coupling of
determinism with chance (Stewart, 1989).
A
similar mix of opportunity and danger is currently arising in the field of motor
learning and motor control. Dynamic
systems theory (DST) is a new approach to the description of self-organization
in nature. DST-related papers are
appearing in a variety of contexts - everything from kinematics to computer simulations of the origin of life.
Central to these projects is an attempt to understand the simultaneous
interplay of structure and randomness in organic action.
Biological (and perhaps creative) processes typically occupy a position
intermediate between the extremes of determinism
and anarchy. Heart rate, for
example, reflects an ongoing negotiation between physiological set-points and
environment stimuli. Somehow, the
rhythm remains in an intermediate state of dynamic tension - neither overly
fixed nor overly flexible. The tendency towards real-time balance has been given
a variety of labels including anti chaos, edge-of-chaos, and self-organized criticality (Bak,
1997).
Dynamics
of this type are not unlike those encountered in the study of expressive
movement training. Physical improvisation is a disciplined reinterpretation of
external contingencies. The texture
of performance is based on an interplay of forces, operating over a range of
time-scales. Serendipity is sifted
and filtered within the continuity of performance training.
The result is sometimes described in the third person, as if the fusion
can only be appreciated fully in retrospect:
…Creating
in this way is a delicate negotiation between inner and outer reality – self
and others – to shape a whole. It
is very exciting when such moments occur. I
don’t feel I create but rather recognize them. (Lacy, 1990)
Homography
in mind, it would be prudent to avoid too large a stretch.
While theatre might colloquially be referred to as a complex system, the
intended meaning is rather different from the equivalent technical definition.
Nonetheless, I suggest that the imagery of DST
is potentially useful as a means of better understanding the creative
act. The chaotic edge is a potent
metaphor, irrespective of its ultimate mathematical relevance to the study of
motor organization. As imagery it
encapsulates the dual nature of improvised movement - the integration of both
structure and randomness within the moment of innovation.
For actor and audience, performance is
about flow, edges, limits and an ongoing tension between boredom and anarchy.
Performance
research could certainly benefit from the use of reconciling imagery.
Ironically, the richness of the topic has itself led to a stratification
by vocabulary. This is particularly
true of the study of expressive movement. Analysis
can occur at a variety of levels, in contexts ranging from sociology to
biophysics. One school has crystallized around the application of Laban’s
Effort/Shape analysis at the
rudimentary level of impetus to action. The
Bartenieff FundamentalsTM are an acknowledgement of the impact of basic symmetry
relationships on motor sequencing. Studies
focus on the integrating aspect of the breath, and the importance of various
intrinsic somatic divisions. These
involve the proximal distal (trunk/limb), vertical orientation and the spinal
response, laterality and independent motion as prerequisites for phrasing and
3-dimensional presence. More
conventional physiological perspectives introduce their own (independent)
lexicon and emphasis. Fitt
discusses the impact of kinesiology on dance training.
Interest here is focussed on how the shape and capabilities of the body
affect the range of motion. Efforts
are aimed at increasing the efficiency of training programs and the prevention
of injuries (Fitt, 1996). Biomechanical
literature is more likely to chart the position, velocity, and acceleration of
the line; papers range from the improvement of efficiency in skilled action to
the role of feedback in performance.
The
diversity extends to larger and smaller scales.
Neuroscience provides another perspective on improvisation.
Here, quality of line is related to brain structures including the
relative input of the primary motor area, the basal ganglia and the cerebellum.
At the level of cell assemblies, central pattern generators have been
studied in animals, and their influence invoked in relation to the
microstructure of skilled motor action (Kelso, 1997).
A shift from biology to psychology leads to a different set of
conventions. Discussions range from
artistic motivation (Rank, 1932) to the use of improvisation in a mental health
setting. Larger contexts lead to
speculation on the intrinsic meaning of motor phrases, the political implication
of gesture, as well as the aesthetics of form involved in theories of
choreography for groups.
Chapple
and Davis, in their stimulating overview of the field, employ the image of a
tree. The roots draw from the
physics of motion and the cycles of biology.
These are integrated at higher and higher levels, until the genres of
performance emerge as branches (1988).
Complexity provides a complementary, process-oriented framework. “Interesting” behavior is approached by describing the
relative contributions of long-term (stabilizing) and short-term (energizing)
influences. The players in the dynamic are a function of the level of analysis.
From the perspective of creativity research, training and self-reflection
provide the performer with the continuity of a supporting structure.
Preparation and incubation meet task and environmental constraints in the
generative moment of insight.
The
dialogue of anarchy and constraint continues at the level of the movement trace.
Spontaneity of line is moderated by the definition of task and the
stability of skills and temperament.
Patterns are shaped by the
inevitability of gravity and the unpredictability of the moment.
At finer scales, we find a subtle balance in the relative contributions
of long- and short-term tempos in the micro-variations of the execution.
Physiological research suggests that ‘structured randomness’ may be
integral to the performance of well-learned motor behaviors’ (Hausdorff
et.al., 1995). There is also
evidence that intermediate complexity in rhythm texture may play a role in
aesthetic judgements (Swickard).
Reciprocity
and Innovation
A
concept similar to the edge-of-chaos metaphor appears early in the creativity
literature. Innovation has long
been seen as occupying an analogous position of dynamic tension, poised between
competing rational and irrational forces.
The edge has been defined implicitly by the simultaneous emphasis in
various schools on structural and anarchic influences.
Platonic advocates have stressed the need for (paradigm-breaking) liminal
states prior to creative leaps. Frequently
the unconscious is seen as a kind of stochastic engine - a churning producer of
spontaneity. In contrast, the
rule-based component has been pursued
by Gestalt psychologists such as Arnheim. The
ultimate extension of this line of thought is a demystification of the creative
act, in which creativity research becomes a sub-discipline of the psychology of
problem-solving.
The
link between structure and surprise is clarified when improvisation is seen as
an aspect of creativity research. Serendipity and continuity are fused in the
folk definition of creativity: that which is at once novel
and appropriate. Historically, however, the emphasis tends to favor one
extreme or the other. Houseman
traces the rationalist perspective (artist-as-craftsman) back at least as far as
Aristotle; Plato is credited with notion of divine inspiration (1975).
The fulcrum of the balance shifts with time and fashion.
19th century writers typically highlight the irrational
component, presumably a reflection of the concerns of the Romantic period.
Coleridge’s account of the writing of Kubla Khan is a good example. The
accident of circumstance involves the reading of a travelogue while mildly
sedated. The result is a miracle of
the esemplastic mind:
The
Author continued for about three hours in a profound sleep, at least of the
external senses, during which time he has the most vivid confidence, that he
could not have composed less than from two to three hundred lines; if that
indeed can be called composition in which all the images rose up before him as things,
with a parallel production of the correspondent expressions, without any
sensation or consciousness of effort. (Coleridge in Ghiselin,
19xx)
20th
century research has tended to focus on those aspects of the creative process
which yield to empirical analysis. Lowes’s
retake on the creation of Kubla Khan is written in this vein (1927).
Starting from the entries in Coleridge’s notebooks, the poet’s
acquisition of imagery is traced through a labyrinth of books and references.
The singularity of the opium-dream is lost, the import of the drama now weighed
against a foundation of extended preparatory research.
Serendipity is downplayed in favor of the continuity of procedural
elements (i.e. knowledge acquisition and an extended period of reflection).
Fluidity
and structure are also thematic in the study of imagery and the creative
process. Again there is a tendency
to emphasize the extremes. Singer
describes the imagination as an engine for novelty.
Creativity results from shifting attention to the spontaneity of the
inner flow (1976). Liminal states
at the sleep/wake boarder are considered particularly fertile grounds for the
appearance of unusual material. Roset
sees the source of the artistic process in the autistic imagery of dreams,
drug-states and mental illness (1984). This
idea will be developed into a theory of remote associations by Baggott
(1997).
The study follows changes in response quality with administration of
cylocybin; a rise in distant analogies is likened to that seen in schizophrenia.
Rothenberg however, notes that verbosity in clinical settings is not
typically matched by a rise in coherence.
A
complementary emphasis on imagery and structure was initially championed by the
Gestalt school of psychology. Kohler
describes the creative process as conditioned by fundamental rules of visual
organization. A number of
such factors have been identified including the primacy of figure/ground
distinctions, and the role of similarity, proximity and closure in the stability
of perception (1925). Arnheim
proposes a connection between sudden shifts in the perception of ambiguous
figures and the apparent discontinuity of insight.
Finke et al. have extended
the study of the Gestalt laws to include the conscious manipulation of imagery. Their model of creativity involves a give-and-take between
generative and evaluative phases. New
ideas are brainstormed by combining visualization with various transforms (such
as rotation, fusion, and changes of context). Emergent properties are then screened for utility, ease of
recall and simplicity of form (1992).
Imagery
is increasingly recognized as an integrating factor – a means of channeling
thought and performance. Conceptual
breakthroughs in science are closely linked to changes in prevalent analogy.
Kepler’s studies of the rules of planetary motion were initially
stifled by over-reliance on a geometry of symmetry.
Subsequent progress involved the adoption of a facilitating image (a
comparison of the solar system to the Trinity, with gravity in the role of the
Holy Ghost) (Koestler, 1964). Body
image and motor performance are similarly open to conditioning through imagery.
Franklin presents a program of dance training based on the cultivation of
specific visual metaphors. Each is
designed to effect a particular quality of the dynamics of movement.
His work draws on the use of active imagination for the promotion of
change, a tradition pioneered by the analytic school (1996).
Stochastic
and structural forces are irreducible aspects of the creative process. They remain despite changes in the approach or methodology.
Recently, computational metaphors have become the focus of attention.
These models employ the ideas and vocabulary of artificial intelligence.
A typical example is the concept of search space.
Weisberg reformulates the creative process as an exercise in problem
solving. The set of all possible
responses is described as a two-dimensional grid, with elevation above the graph
representing the fitness of the solution. Creativity becomes a journey through a
fog-covered landscape. The goal
becomes the arrival in a reasonable amount of time at high ground (Weisberg,
1993). A key topic in these models
is the links between productivity and effective search strategies.
Interestingly, one of the most popular suggestions involves the
application of controlled randomness. Initial forays are left haphazard, as a
means of sampling widely. Later, as experience is gained, the degree of randomness is
gradually lowered, leading to a fine-tuning of the search.
Judicious use of chance allows for a degree of performance superior to
that of deterministic selection rules.
Hofstadter
integrates the limits of artificial intelligence with the practical work of the
artist. His Copycat project is an attempt to engineer a machine capable of
independently designing aesthetically pleasing type fonts (1995). Random
elements are explicitly incorporated in the software.
Procedures are invoked probabilistically rather than in predetermined
sequence. Within structural
limitations there is no a priori means for the programmer to anticipate the
outcome of a given run. The
resulting mix of surprise and structure is a powerful metaphor for the artistic
process.
Fluid
Structure and the Psychology of Movement
A
real-time dialogue between structure and randomness has long been described in
somatic research. In this
literature, improvisation is seen as contingent on the interplay of the general
with the specific - i.e. the
interplay of biological tendencies with the uniqueness of individuation. The
deterministic aspect is emphasized in the discussion of movement archetypes -
primal loci of organization which serve to channel undifferentiated action.
Stochastic influences involve the move from predisposition towards the
specificity of life history. The
manifestation of archetype against the actor’s unique background is the
essence of creative improvisation.
The
complexities of improvisation have long been the subject of psychological
interest. Chodorow presents an
ambitious attempt to discern inner structure beneath the particularity of the
moment. Her terminology is borrowed
from Jungian analysis, although she works as a Dance Movement Therapist.
In this model, creative action emerges from the interplay of a series of
internal influences. Details of the inner architecture reflect both evolutionary
and personal development. The
deepest layers provide a fundamental energizing influence. Their effect is undifferentiated, but serves as a stimulus to
action. There is perhaps a close analogy here to brainstem function.
The reticular activating system in particular plays a role in the
modulation of attention, and the maintenance of arousal (Chodorow,
1991).
Superseding
the primordial unconscious are the emotional archetypes.
At this level the stimulus to action is canalized into a series of primal
affects. Following Stewart,
Chodorow outlines a system of four responses: fear, anger, sadness and
contempt/shame. These in turn are
grounded on the more basic aptitudes of centering (startle) and orientation
(joy/interest). Physiological
correlates are also possible here. Structures
of the mid brain are directly involved in the creation of emotional response.
Animal studies have highlighted the role of the amygdala in priming the
motor system for fight or flight. More
controversial is the identification of each archetype with a prototypic icon and
sensory channel.
In
the analytic model, primal activity is subject to further differentiation at
higher levels. The cultural unconscious integrates the capacity for
self-reflection with the power of emotion, leading to productive imagery.
This is the foundation for myth and of the sensation of the inner
journey. It also represents an
explicit opening-outward to the social world.
Symbolism results when affect-tendencies meet with the specifics of
locale and era. Motion necessarily
reflects this balance.
The
personal unconscious adds the history and conflicts of the individual to the
layering of action. The persona is
a constellation of movements representing the idealized projection of the self. The shadow, in contrast, is the reservoir for unacknowledged
or dissociated psychic energy. In
analytic thought, tension between these forces is said to result in
idiosyncratic action. Fragments
recur in daily life (and presumably in movement training) until the underlying
complex is resolved. The focus of
Mindell’s process oriented psychology is on the alleviation of this state.
His procedure involves the cultivation of focussed awareness – a type
of movement meditation. When re-search is combined with presence, gesture is given an
opportunity for elaboration and integration.
The action border then shifts towards deeper issues.
Ego
consciousness is primarily responsible for volitional action.
It is dominant in the setting of tasks and in phases such as the warm-up.
Phylogenetic association is with the neocortex.
The canalization of energy associated with deeper layers reappears here
in terms of personality patterns. Choice
in movement is conditioned by broad tendencies towards introversion and
extroversion. There are also preferences with respect to sensing and
intuiting, or thinking and feeling. Higher
on the Self-ego axis are factors representing individuation and the life-arc.
They appear as organizing principles of motion involving the use of
symmetry and closure symbols such as the circle.
Analytic
psychologists reject both determinism and anarchy as meaningful approaches to
the understanding of creative movement. Action
tendencies are broadly constrained by physiological principles and genetic
memory. However the influence of
culture, personal baggage, time, place and conscious choice are no less present
in the reality of improvisation. Traditionally,
the study of dynamical systems obeys a similar logic.
Description in terms of the specifics of position is contrasted with
explanation in terms of general movement principles (velocity, acceleration).
When deeper rules are combined with a knowledge of history (initial
conditions) understanding of the system is said to be complete (Elman et.al.,
1996). Chodorow’s approach is schematically similar, with a
universality of structure expressed through the prism of individual experience.
Hawkins
presents an alternative framework for the analysis of creative movement. Her perspective is more experiential, and less tied to
Jungian terminology. Creative
action is presented as the external correlate of a process of internal
re-organization. Psychic dynamics
are conditioned by innate tendencies such as sensing, feeling and the power of
the image. However, maturity of the
performance is contingent on training and the development of raw potential (Hawkins,
1991).
Each
of the organizing principles is open to growth and elaboration.
Sense memory and awareness are heightened by observational exercises.
These include attention to structural detail, as well as to aspects of
tempo-rhythm. Feeling is related to the depth of involvement – i.e. the sense
of presence. Motivation is
encouraged by providing a safe physical and emotional environment.
Movement tasks are assigned which tap basic somatic issues such as
approach/withdraw, conceal/reveal and tighten/loosen. With time, fragments of authentic action are woven together
into extended improvisations.
Imagery
is central to the fusion of sense and feeling as coherent action. Active imagination is encouraged firstly through guided
relaxation. Practice is given in
the ability to externalize abstractions as concrete dynamics.
Students are advised to experiment with concepts open to both physical
and emotional interpretation. Symbolism
is also employed as a means of binding sensation to social reality.
Exercises include the exploration of entanglement, attachment, search and
suspicion.
Transformation
is deepened with the acquisition of choreographic skills.
Awareness of energy flow is guided by studies of birds and animals.
Spatial design and relationships are also addressed through imagery.
Rhythmic sensitivity is heightened by attention to word usage and
phrasing. There is also concern for
fostering the reintegration of separate skills as organic movement. Latent organization is nurtured through performance
re-search. Hawkins contends that
development is marked by a shift from personal to social, from fragmentation to
elaboration and from constraint to communication.
In
comparison with the analytic model, Hawkins is more aware of the synergy between
improvisation and discipline. While
both frameworks acknowledge the influence of inner architecture, the emphasis
here is on professional development. Latent
tendencies are seen as in perpetual flux, gaining or losing efficacy in relation
to training and experience. While
action arises from a general instinct to growth, content and quality of line are
open to revision.
Structure,
Randomness and the Position Trace
The
metaphor of an edge is no less relevant at the level of kinematics. Laban was one of the first theorists to see within the
apparent anarchy of real-life action the regularity of
recurrent dynamic ‘themes’. These
he described in terms of clusters on the axes of weight, space, time and flow.
His effort actions are roughly analogous to the somatic notion of motion
archetypes. They represent the
background presence of structure, organizing principles in ongoing tension with
the spontaneous element of performance. Further,
his clusters are also similar to the DST notion of an attractor - regularities
in the transitions of activity which a system manifests during its evolution.
Detailed
analysis of movement paths is constrained by inherent methodological
difficulties. One issue is the
fidelity of recording instrumentation. Novak
discusses the merits and drawbacks of applying codified systems in the context
of contact improvisation (1988). Layers
of distortion are necessarily introduced. The
precision of Labanalysis, for example, is no doubt qualified by both the skill
of the analyst and a degree of inherent cultural bias.
Aspects of the dynamic schema in particular have proven to be
problematic. Fitt suggests the
replacement of the flow axis with a more value-neutral estimate of force. She
also describes attempts to demonstrate the validity of the quality scales with
reference to changes in muscle activity traces (EMG) (1996).
Despite
the difficulties, Effort/Shape coding, an aspect of the Laban system, provides
an intriguing method for documenting the complexities of improvised motion.
The emphasis is on action quality rather than position/trajectory.
The result is an abstraction of performance,
often revealing latent structure beneath the particularity of the moment.
In his own research, Laban identifies a series of dynamic archetypes
(Effort Actions). Each represents a
distinct triplet, drawn from all possible combinations of four axes of notation.
Floating for instance, is the confluence of a light touch, an indirect
use of space and a leisurely tempo-rhythm.
The psychological correlate is indulgence. In contrast, thrusting denotes the opposite – an attitude of
resistance to time, space and weight. Eight such primary centers are contrasted
with six indeterminate regions known as incomplete actions.
These are behavioral transition points (axis crossings) in which only two
of the four polarities are distinct (Newlace, 1993; and Laban,
1971).
Laban
presents Effort/Shape as a means of describing both the objective manifestation
and the motivation to action. His movement constellations are at once pure
observation and an inference regarding the inner world of the performer. The argument is tailored to the constraints of conventional
theatre. From this perspective,
movement is meaning; dynamics become the medium through which the audience
reaches an understanding of intent. The
impact of effort actions in a nonrepresentational setting is left unexplored;
however, a universality of influence is implied.
Movement
archetypes represent a subjective awareness of long-term regularities. Laban
performs what in statistical parlance would be referred to as cluster analysis. Traces accumulate with extended observation, until patterns
begin to emerge. Prototypes (such
as effort actions) are then identified, and provided with labels.
A question for future research is the degree to which the action centers
(punching, pushing, slashing) generalize beyond the details of Labanalysis.
Does gliding represent a regularity of expression or of the notation?
Are other dynamics inherently unstable, or are they less often noted
because of a lack of contrast? One
response is to retain the essence of the recording process, but to adapt an
eclectic approach to content. Novak’s photo essay is a case in point;
methodology is grounded in the Effort/Shape tradition, but attention is geared
to the specifics of performance style.
Technology
provides a means of reducing some of the bias associated with motion study.
Mulder describes the range of equipment available in the field of human
movement tracking. Sensors
take advantage of acoustic, optical and magnetic designs.
Precision is increased when systems are worn on the body (the inside-out
strategy). The disadvantage in the
performance context is the clash between mechanical encumbrance and artistic
expression. Outside-in systems such
as video cameras produce a courser trace, but are less intrusive.
There
are other more subtle difficulties with the recording process.
Improvements in reliability are won by a narrowing of focus; a loss of
perspective is always possible. The
subjective impression of Performer A’s use of space is replaced by the
objective trace of her center of gravity. Or
her elbow. The element of social
interaction is similarly obscured.
However,
the use of such equipment does create exciting new research opportunities.
This is particularly true when the precision of the recording is combined
with contemporary approaches to path analysis.
The field of complex systems has been particularly active in the
development of tools for the understanding of subtle dynamics.
Their methods have evolved as a means of coping with ambiguous structure,
such as the pattern of eddies in flowing water.
One of the icons of this interdisciplinary science is a graphic analysis
of the weather by Lorenz. His
3-dimensional model reveals flickering chaos subsumed within a larger, less
obvious aesthetic. Building on
these results, a more general procedure (the Method of Delay) was introduced for
the study of nonlinear change. The
approach starts with the spatial trace, and argues backwards to arrive at a
statement of general movement quality. Action profiles emerge from the
superposition of many short-term events. The
result is a statistical fingerprint in which the details of the improvisation
are sacrificed for knowledge of long-term behavior. The objectivity of the image is one answer to the issue of
bias in conventional scales of notation.
Contemporary
analysis also benefits from advances in computer technology.
Connectionism is an increasingly popular alternative to conventional
software design. Based loosely on
biological models, the parallel architecture is well-suited to the detection of
subtle regularities. Usage is
typically split into two distinct phases. In
the training period, data samples are paired with a category label. Sets of
movement traces might, for example, represent the work of a single performer, or
samples from a particular genre (mime, contact, etc.).
The task of the computer becomes the discovery of invariant patterns –
common elements which transcend the specifics of each recording.
Analysis becomes the basis for the creation of an internal prototype.
The
results are useful as an objective assessment of recurrent themes in the
dynamics of action. They may also
be used for comparing or contrasting new work with previous samples.
The utility of the approach is well-demonstrated in the field of speaker
verification. Here the technology
is applied to the problem of determining the identity of an individual given a
sample of her speech. Patterns of
vocal quality are abstracted, and compared with a library of templates.
Differences are to be expected from day to day as people catch colds, and
respond under various levels of stress. However,
there are thematic elements beneath changes in surface details.
Self-Organization
and The Body in Motion
Metaphor
comes closest to mathematics in current attempts to understand motor learning
and motor control. Theorists such
as Thelen and Kelso describe a universe of action in which the movement trace is
a negotiation between the elusiveness of intention and the physics of the
actor/environment matrix. Structure
is modulated through the necessary
presence of randomness. The role of
‘noise’ in the motor system is seen as that of a lubricant, allowing the
actor the flexibility to escape from fixed action patterns. Fluid movement is described as contingent on the real-time
relationship between stochastic and structured forces - yet again a phenomenon
contingent on dynamic balance.
The
application of these new tools has meant a step forward in the sophistication of
movement studies. Thelen presents
an overview of the paradigm shift occurring in the analysis of skill acquisition
(1995). Her research into infant
development focuses on the process of self-organization.
The ‘degrees-of-freedom’ dilemma is particularly intriguing – given
the incredible plasticity of the human body, how is it that most children
autonomously arrive at the same motor milestones? Her work implies that skilled movement results not only from
neural development. Action is also shaped in real-time by the physical
properties of the limb. In this
view, the potential for walking is distributed across the interface between the
child and her environment. Change
any aspect of the system, and the nascent form is affected. Thelen notes the emergence of new motor styles during the
Apollo space program. A jumping
pattern automatically replaced conventional walking under the lighter conditions
on the moon’s surface.
Another
example of the blending of form and function comes from the investigation of the
stepping reflex. Newborns are
capable, if supported, of demonstrating a walking pattern.
The behavior is lost within a few weeks, only to reappear in a year’s
time. Conventional thought was that
the motor schedule was conditioned by changes at the level of the central
nervous system. However, Thelen offers a more pragmatic solution; the gain and
loss of potential reflects the very physical ratio of limb size to strength.
Reduce the pull of gravity (through suspension in warm water, for
example), and the phantom skill returns. Walking
becomes not a ‘thing the child does’ but rather a negotiation between the
will to move and the context. In
Thelen’s words:
Nonetheless,
some patterns are preferred under certain circumstances: They act as attractors
in that the system ‘wants’ to perform them.
Other patterns are possible but [are] performed with more difficulty and
are more easily disrupted. (1995, 84)
Kelso
is similarly interested in the link between physiology, environment and emergent
action. His work documents the relative contributions of task, symmetry and
coordination in the organization of skilled movement.
Often his experiments focus on the nature of spontaneous shifts
(transition points) in behavior. In
a typical example, subjects are asked to simulate patterns of locomotion
familiar from the animal kingdom. These
include pacing, trotting, galloping and jumping.
Tracking equipment is used to monitor the ‘cross talk’ between arms
and legs as the tempo is varied. At
macro levels, action patterns are cleanly separated; there is an either/or
quality to limb synchrony. Preferred
directions (bound to jump, not the other way round) further contribute to the
sense of coherence.
However,
a complimentary picture emerges at smaller scales.
Microanalysis reveals a graded instability (fluctuation enhancement),
prior to a switch (Kelso, 1997). Kelso
suggests that this is cause rather than effect; noise may be a necessary element
in pattern shifts. Using a physical model, the argument is that endogenous
excitation or randomness may serve to facilitate phase transitions by weakening
the hold of established action. The
idea is exciting because it dovetails with other studies on the role of
‘background noise’ in the central nervous system. Freeman and colleagues at the University of California have
spent 30 years studying the physiology of perceptual shifts (1991).
Most of this work has been with the sense of smell in animals.
A major finding has been the demonstration that distinct firing patterns
of brain cells correlate with the recognition of different odors.
Following on this however, has been research into the ‘reset’
mechanism. To use a more familiar
analogy, how does the wine taster clear her palate between glasses?
Freeman contends that the brain naturally returns to a state of
controlled randomness (chaos). The
uniformity of noise serves to neutralize the impact of previous impressions, and
to facilitate accommodation to new ones. It
is the dynamic at the center of fresh perception, and possibly fresh action.
Is this the physiological correlate to Tufnell’s ‘Gap’ in
improvisation – the felt sensation of a facilitating emptiness?
My
work at the University of Malta involves the application of trace recording to
the study of expressive movement. A basic question I have is whether motion
archetypes (clusters in the dynamics) can be used to document individual
differences in performance. Personality
has been described as endogenous, enduring patterns of behavior. Typically this
is interpreted linguistically in terms of responses on pencil and paper tests.
Is it possible also to demonstrate the nonverbal equivalent in terms of
movement consistency?
In
the current project, ultrasonic equipment is being used to record improvised
hand movements. The participants
are undergraduate and post-graduate actors.
Their experience with movement training ranges from zero to nine years.
Experimental sessions last for thirty minutes.
The recording ‘studio’ consists of a table 1.5 m in length, situated
at waist height. Performers are
asked to create an air trace by moving a target towards and away from the eye of
the recorder (Fig. 1a). The context
of the dynamics is varied in 5-minute intervals; the requested quality of line
shifts between the four combinations of staccato/legato at fast/slow speeds. Each actor is recorded on two occasions, separated by at
least 24 hours. The basic question
is whether individual differences recur consistently in the line traces.
Interesting
patterns emerge from the comparison of records. The quality of line is
apparently conditioned by both acquired skills and by performance
interpretation. Precision of action
correlates with years of experience. The
beginners are able to capture the fast/slow dichotomy, but the flow categories
are lost, particularly at high speeds. The
effect is presumably related to the development of performance skills.
Movement pedagogy is another factor. One performer showed an unusual
emphasis on the transition between stops; in
discussion with him the phenomenon was attributed to a background in mime.
The
tools of nonlinear analysis provide additional levels of insight. When movement
is processed with the Method of Delay, details of the partitura are suppressed
in favor of general action signatures. In
this format, similarities of performance are heightened (Fig.
1b). Knowledge of an actor’s work on one day suggests qualities
of line that will evolve in later sessions.
Patterns are further clarified by extending the results with new computer
technology (networks). It becomes
possible to identify individual differences in the range of preferred dynamics (Fig.
1c). In effect, the software
is asked to identify an actor based on a current motion profile and a library of
previous samples. Movement
recognition is in principle no different than analysis by voice.
I have had some success in this area, although the difficulty increases
exponentially with the number of performers.
Complexity
and Tempo
At the
level of tempo-rhythms, complex
organic processes demonstrate a DST-related concept known as self-similarity.
It appears that under certain conditions, systems are able to
spontaneously self-organize, creating a subtle organization of short-term
fluctuations within a framework of long-term influences.
Cardiac variability is one example - the heart maintains a steady rhythm
yet fluctuates rapidly in response to environmental pressures.
Musical composition is analogous in the successful reconciliation of
monotony and surprise. It is
expected that a similar pattern would appear in the physical analysis of
improvised movement patterns. However,
this assertion has yet to be rigorously tested.
The
‘fluid-structure’ dynamic is equally apparent when the emphasis shifts from
space to time. Hausdorff et al.
report on the intricate layering of tempo-rhythms in natural walking patterns (1995). Participants are fitted
with a device that measures the interval between steps of the same foot. Analysis is focused not on pace continuity, but on the
variations in speed from the average rate.
The nature of these fluctuations is intriguing; the profile of residuals
is consistently intermediate between randomness and over-determination.
Technically, the results belong to a category known as scale-free
processes; while slow changes predominate, there are contributions to the
movement occurring at all frequencies. Quite
literally, there is ‘something to see’ regardless of the interval of
attention. Scale-free behavior is
not an artifact of the recording. It
is lost when the data is reanalyzed after shuffling.
It is also sensitive to the presence of self-organization.
When the subjects are asked to match their steps to the sound of a
metronome, the signature disappears.
The
origin of the pattern is currently a matter of speculation.
Simple models (e.g. adding noise to a base rhythm)
fail to recapture the distribution of tempos.
Presumably there is a connection with the layering of influences present
in all organic processes. It is
perhaps not surprising that the authors of this study are known for their
investigation of heart rate intervals. In
this field, micro-variations have been a focus of attention for more than thirty
years. Here too, the phenomenon
represents a confluence of forces (cardiac, respiratory, neuronal) operating
over a range of time scales. Details
of the transition from cacophony to symphony remain poorly understood. What is clear is the link between biological processes and
health. Variability is suppressed
in the presence of stress hormones. It
is also reduced by diseases such as diabetes.
This fact has been put to good use; scaling in inter-beat intervals is
now recognized as a marker for recovery following heart transplantation (Stanley
et.al., 1996).
Human
activity is colored by a superposition of tempos.
Further, the balance of frequencies reflects both internal organization
and accommodation to external factors. Does
the eye use this information in evaluating the authenticity of action? This is currently an open question. Certainly the ability to distinguish organic movement
persists even under difficult viewing conditions.
Kelso reports on a study in which subjects are recorded while performing
various tasks. Motion is rendered
abstract by storing only the light from markers placed on key points of the
body. Resulting patterns are then
embedded in larger cloud of dots. Viewers
have difficulty identifying the outline of a person when the images are static.
However, the recognition is instant once the sequence is run (Stanley
et.al., 1996).
Voss
is responsible for investigating the specific relationship between perception
and scale-free processes. His
pioneering work focuses on the acoustics of music.
One study involves the analysis of frequency relationships in extended
recordings of radio programs. The
emphasis is on global features of organization in the distribution of audio
power. The results show a pattern
analogous to that described by Hausdorff et al.
Music apparently represents the same combination of multilevel activity
and moderate complexity (Voss, 1975). Early
studies demonstrated invariance across a spectrum of pieces, ranging from
classical to jazz. The results have
since been confirmed in the spectral profiles of nonwestern compositions.
Schroeder speculates that perception is biased towards phenomena intermediate
between monotony and anarchy. It
seems that music captures our attention by matching our range of sharpest focus.
Voss
extends these findings by exploring the relationship between scaling and
perceived musicality. In one study,
frequency distributions are used as the starting point for the creation of
original scores. The synthesis
occurs in several steps; the pattern of scaling is tapped first for the sequence
of pitch, and then again for the duration of the notes.
Three compositions are prepared, representing the default condition, as
well as shifts in the balance towards randomness and predictability.
Listeners consistently judge the signal with the intermediate degree of
complexity to be the most pleasingx.
I
have been looking at ways to apply these results in the study of artistic
expression. Classical musicians differ in their interpretation of a score;
enthusiasts are quickly able to identify a pianist’s ‘signature’ despite
changes of instrument or musical selection.
Is it possible to find an empirical correlate to this invariance?
As in the
Figure
1a
top: hand movement trace – displacement versus time
1b
middle: motion signature, derived from Fig. 1a using Method of Delay
1c
bottom: separation of 3 trace
samples by performer, based on tempo and velocity
analysis
of movement profiles, a machine learning approach is adopted.
Connectionist technology is well-suited to the analysis of slippery
categories. Caudell & Butler
for example describe the development of a filter capable of distinguishing
between the sonar echoes of natural and man-made objects (1991).
The beauty of the procedure is that the salient qualities of the signal
do not need to be specified beforehand. Rather
the software is presented with the raw data, paired to a series of desired
targets (‘artificial’ versus ‘natural’ in the sonar study).
The computer is asked to assimilate the information.
This is accomplished by searching for a means of summarizing the data
such that the labels provided are preserved.
The performance summary can then be tested against new, unlabelled
material. Furthermore, it is
possible to work backwards from the solution in order to gain insight into
exactly which features of the data were relevant to the distinction.
In
a project on performer identity, I attempt to compare the interpretation of two
pianists. The material consists of
excerpts from Bach’s Goldberg Variations.
Recordings are digitized and summary profiles are constructed from the
distribution of frequencies. These
are paired with the target categories (“Performer A” and “Performer B”)
and presented to the software. The
computer is indeed able to formalize the distinction.
Further, the recognition by performer extrapolates to new recordings.
The
most interesting aspect of the exercise is in the ‘debriefing’ of the
classifier following its success. Resolution
turns out not based on isolated features of performance.
Rather, the separation by style draws on information distributed
uniformly across the entire range of frequencies. The smoothness of this pattern is only disturbed when the
data is corrupted. In all
likelihood there is no literal relationship between the machine’s logic and
the details of subjective perception. However,
these results do suggest that the performer identity is a holistic construct,
informed by input across a range of scales.
Conclusions:
Towards a Deeper Understanding of Fluid Structure
Results
to date represent little more than a sketch for further work.
The icons of complexity (scaling, self-organization, reciprocity) are
useful; they stress similarities of form across diverse knowledge domains. The
implications for research, however, need to be tested against the reality of
performance.
A
key issue in the field of creativity studies is the development of the
distinction between online and off-line processes.
Most of the attention over the past century has been focused on the lives
of writers or scientists. Paradigm
shifts and literary works have in common the luxury of pauses for reflection.
Wallas’s four-stage model notes the importance of incubation prior to
insight. Kohler’s studies treat
innovation as the outcome of a protracted exercise in problem solving.
Creative movement, on the other hand, represents continuity of flow.
There is no opportunity for a time-out in performance.
This suggests that artistic improvisation involves a qualitatively
different set of production rules.
Continuity
and innovation are linked in music as well as theatre.
Sternberg studies the nature of online composition in jazz.
He believes that successful musicians base their sessions on the use of a
limited set of guiding principles, known as production rules (1988).
The complexity of these heuristics is ‘unpacked’ by applying them
cyclically as the performance unfolds. A
similar dynamic is commonly encountered in the field of complex systems.
The creation of highly evocative fractal images is based solely on the
recursive use of very simple instructions.
Organic textures are particularly easy to capture with this method.
Trees and ferns emerge through the actions of translation, rotation and
scaling. Is there a comparable
dynamic at work in the artistic use of the motor system?
Complexities
of structure are similarly hinted at by research based on analytic and
humanistic psychology. Insight into the role of the unconscious in physical
action will presumably deepen as theoretical frameworks are evaluated by
performers. Certainly the writings
of dance movement therapists can be adapted to research in creative
improvisation. A number of
hypotheses are implicit in Chodorow’s system; it would be interesting to match
them against felt experience in the work of the actor.
Does improvisation reflect the ebb and flow of autonomous centers of
influence? Is this interaction
relevant to the evolution of structure in training sessions? How important is imagery to the process of organization?
What is the link between repetition and awareness?
To what degree are fragments of movement a result of gesture and culture?
Laban’s
work offers another perspective on the psychology of movement.
Some of the arguments appear dated, particularly the notion of dynamics
as a window to the inner world of value. However,
potential links between Effort/Shape analysis and cognitive science await
development. Laban describes what
would now be referred to as altered states of consciousness in terms of patterns
of movement. Further, he offers
clear predictions regarding the correlation between action quality and inner
experience. It would be fascinating
to combine the potential of current biotechnology with the heritage of dance
notation.
Thelen’s
work with dynamical systems raises interesting questions regarding locus of
control. Where is the stability of
the dynamic? Is it imposed by the
central nervous system, or does it arise as a negotiation with space and
gravity? Current thought on the
physics of motion is intimately connected with the decline of Cartesian dualism.
From the viewpoint of complex systems, improvisation becomes a tangible
expression of the field of performance, rather than the will of the performer.
There are implications here for the work of the artist. How does quality of movement relate to the spotlight of
attention? What if the center of
perception expands to include the stage, or the stage-and-audience?
Perhaps soliloquy is inevitably dialogue.
Kelso
extends the ambiguity to include the act of volition.
Change is conditioned by the dissolution of attractors, mediated by an
enhancement of endogenous ‘noise’. The
similarity to training in improvisation is striking;
discipline becomes a means of clearing
away of old patterns in expectation of new form.
The relationship between physiology and subjective experience is an open
question. Is the uncertainty of spontaneity exciting, or does excitement
facilitate spontaneity? How sensitive is the process to the balance of tension and
relaxation?
Technological
progress provides a different perspective on the complexity of performance.
Sessions of improvised movement exist within a larger context of
training. Software modeled on the
process of speech recognition is able to express this continuity in objective
terms. A fundamental issue (for
actors and their empirical audience) is the question of novelty:
to what degree is new work inevitably conditioned by past experience and
motor personality?
The
most interesting arena for interdisciplinary research is the relationship
between tempo-rhythm, performance and the aesthetic of organic action. The process through which multilevel contributions to
movement are blended into a coherent whole is still poorly understood.
Bak argues that nature is often biased towards textured (scale-free)
behaviors. However, the assertion is still controversial.
There is as yet no agreement on a mathematical basis for the assertion.
There is also debate regarding the relevance of the theory under
real-world conditions. Improvised
movement is a case in point. To
what level does the term ‘self-organization’ imply?
Is it meant with respect to the physical trace, the interface of motor
programs with muscle, or the will to form?
Or is it a summary of the multilayer influences?
Psychophysics
will eventually answer whether the phenomenon of listening bias is relevant to
the visual appreciation of movement. Is
the sense of texture contingent on a balanced superposition of tempos? Is
complexity of line related to truth in action?
Or is scaling of relevance only to the appreciation of music?
For the actor, research linking biological rhythms to external influences
is equally provocative. It is an
invitation to re-examine the impact of stress, physical condition and
psychological locus of control on the mercurial quality of performance.
Anon., ‘False Cognates in
Swedish and Dutch’ http://users.bart.nl/~sante/engfalsk.htm
Chapple, E., & Davis M.,
1988, ‘Expressive movement and performance: Toward a unifying theory’.
The Drama Review, 53-79.
http://www.cs.sfu.ca/~amulder/personal/vmi/HMTT.pub.html