neuroaesthetics of dance patrick haggard institute of cognitive neuroscience university college...
TRANSCRIPT
Neuroaesthetics of dance
Patrick HaggardInstitute of Cognitive Neuroscience
University College London
Funding: Leverhulme Trust, Royal Society, EU, Wellcome Trust, ESRCColleagues: Beatriz Calvo, Daniel Glaser, Corinne Jola, Dick Passingham, Deborah
Bull, Emma Maguire, Tom Sapsford, Shantel Ehrenberg, Elena Daprati, Marco Iosa, Paolo D’Oto
Dance requires:
• Performer
• Observer
• (Possible other stuff)
Disclaimer 1: Observer Approach
Dance requires:
• Performer
• Observer I will focus here!
• (Possible other stuff)
Disclaimer 1: Observer Approach
Disclaimer 2: Scientific Approach
Advantages
• Simplify• Measure• Explain• Predict
Disadvantages
• More than sum of parts• Misses the subjective
element, what “I” feel• Culture-bound
Observe, clarify, hypothesise, but also appreciateReduce, but also respect
Study of something’s:
• Artistic or potential artistic value:
• Beauty
• Ability to induce a range of mental states, including pleasure, various emotions, awe etc.
Aesthetics: Definition
Study of something’s:
• Artistic or potential artistic value: Definition by practice
• BeautyDefinition by objective properties
• Ability to induce a range of mental states, including pleasure, various emotions, awe etc.Definition by mental effects
Aesthetics: Definition
1. Aesthetic preference
2. Brain basis of aesthetic judgement
3. Aesthetic drivers of artistic change
Roadmap
1. Aesthetic preference
2. Brain basis of aesthetic judgement
3. Aesthetic drivers of artistic change
Roadmap
• Objectivist/Platonist aesthetics– Intrinsic aesthetic value– Aesthetics is a property of things
• Subjectivist/relativist aesthetics– No intrinsic aesthetic value– Only personal liking– Aesthetics is not a property of things, but of how
we see things
1. Aesthetic preference
• Aesthetic preference: Which do you like most?
• Universal application, high face validity
• Low explanatory value and efficiency:does not say why you like it?
• Subjectivists: express personal aesthetic taste
• Objectivists: find universal preferences
1. Aesthetic preference
1. Aesthetic preference
• Preference judgements give experimental control
• Preference judgements give experimental control
1. Aesthetic preference
Golden section:1.618 : 1
• “Balance”
• Aesthetic value depends on configurations, relations between parts and wholes
Aesthetic preference: balance
Balance(McManus et al., 1985)
GroupA
GroupB
Balance
GroupA
GroupB
Balance
GroupA
GroupB
Balance
Balance
Some body configurations may be especially pleasing
Dance aesthetics may reflect a structured visual pattern
Dance may exploit basic brain mechanisms of pattern perception
• Familiar things are preferred to unfamiliar
• Basis of aesthetic subjectivity:– Different previous experience, different evaluation– Basis of cultural relativism aestheticism
• Exploration vs. exploitation
• Challenge for artists
Aesthetic Preference: “Mere exposure”
Ballet Capoeira
Non-expert
Non-expertSubjects
Ballet dancers
Capoeira dancers
Expert
Expert
Non-expert controls Non-expert Non-expert
Calvo-Merino et al., Cerebral Cortex, 2005
Ballet Capoeira
Aesthetic preference for dance movesCalvo-Merino et al., Consciousness and Cognition, in review
Non-expert
Non-expertSubjects
Ballet dancers
Capoeira dancers
Expert
Expert
Non-expert controls Non-expert Non-expert
• 6 naïve subjects, retested after brain scanning• 24 ballet and capoiera moves• “How much do you like this move?” (1-5)• Which brain areas correlate with average liking?
Neural correlates of liking
a b• Bilateral Visual• Right premotor
• “embodied”aesthetics?
“Neurotargeting”- Project neuroaesthetic activations back into stimulus space
• People disliked this, right premotor cortex inactive
• People liked this, right premotor cortex active
“Neurotargeting”- Project neuroaesthetic activations back into stimulus space
• People disliked this, right premotor cortex inactive
• People liked this, right premotor cortex active
“Neurotargeting”- Project neuroaesthetic activations back into stimulus space
• People disliked this, right premotor cortex inactive
• People liked this, right premotor cortex active
• Neural correlates of average preferences
• Key parameters for liking:– Overall amount of movement– Jumping
• Future work:– More people, more brain areas– Other aesthetic dimensions beyond liking– Neural correlates of liking in experts?
Aesthetic preferences: conclusion
1. Aesthetic preference
2. Brain basis of aesthetic judgement
3. Aesthetic drivers of artistic change
Roadmap
Brain basis of aesthetic judgement
• Pure dance movement stimuli
• Point-light display
• Sequence of two dance moves
• Aesthetic preference task
• “Which move do you prefer?
Brain basis of aesthetic judgement
Vis
ual P
erce
ptio
nA
esth
etic
E
valu
atio
n
“Same or different?”
“Which one do you prefer?”
Brain basis of aesthetic judgement
Vis
ual P
erce
ptio
nA
esth
etic
E
valu
atio
n
“Same or different?”
“Which one do you prefer?”
• What brain circuits underlie aesthetic evaluation?
Brain basis of aesthetic judgement
Vis
ual P
erce
ptio
nA
esth
etic
E
valu
atio
n
“Same or different?”
“Which one do you prefer?”
• What brain circuits underlie aesthetic evaluation?• Are these circuits influenced by experience?
Naïve subjects Expertdancers
Results
Aesthetic preference -Visual discrimination
BExpert dancers
Naïve subjects
A
L&R parietal:Attention?Mirror system?Sensory cortex?
R temporal:Memory?
Results
Aesthetic preference -Visual discrimination
BExpert dancers
Naïve subjects
A
L&R parietal:Attention?Mirror system?Sensory cortex?
R temporal:Memory?
• Aesthetic network:– attention and embodied feeling (SI)
• Clear expertise effects– dancers’ aesthetic judgements are
based on reference memory– internal stored representation of a
perfect pirouette?
Conclusions
• Aesthetic appreciation can be learned, and taught
• Brain basis for ‘connoisseurship’
• Explains effects of culture and familiarity
Implications
1. Aesthetic preference
2. Brain basis of aesthetic judgement
3. Aesthetic drivers of artistic change
Roadmap
Aesthetic geometry
• Balance and pattern central to aesthetic preference
• Body ‘line’ in dance may be an aesthetic pattern
• Historical evolution of body line– Royal Ballet archive 1942-2004
• Fixed position in choreography– Developpe, arabesque etc – Rose Adagio, Sleeping beauty
ANATOMICAL REFERENCES AND DERIVED STIMULI
B. Schematic Figures
Right finger tip
Right Toe
Left finger tip
Left toe
A. Location of markers and segments
+
finger tip
wrist
elbowshoulder
sternum
navel
forehead
nose
hip
knee
ankle
tip of the toes
ANATOMICAL REFERENCES AND DERIVED STIMULI
B. Schematic Figures
Right finger tip
Right Toe
Left finger tip
Left toe
A. Location of markers and segments
+
finger tip
wrist
elbowshoulder
sternum
navel
forehead
nose
hip
knee
ankle
tip of the toes
ARABESQUE PENCHEE
120
130
140
150
160
170
180
190
1945 1955 1965 1975 1985 1995 2005
Years
Le
g E
lev
ati
on
(d
eg
)
r=0.8306
1962 1979 1996 2003
DEVELOPPE A LA SECONDE
r=0.7793
90
110
130
150
170
190
1945 1955 1965 1975 1985 1995 2005
Years
Leg
Ele
vati
on
(d
eg)
r=0.7877
1962 1979 1996 2003
B
ARABESQUE PENCHEE
120
130
140
150
160
170
180
190
1945 1955 1965 1975 1985 1995 2005
Years
Le
g E
lev
ati
on
(d
eg
)
r=0.8306
1962 1979 1996 2003
DEVELOPPE A LA SECONDE
r=0.7793
90
110
130
150
170
190
1945 1955 1965 1975 1985 1995 2005
Years
Leg
Ele
vati
on
(d
eg)
r=0.7877
1962 1979 1996 2003
B
• Body line becoming more vertical
• Why?
Results
• Body line becoming more vertical
• Why?
1. Dancer fitness?– But, both easy/supported and hard/unsupported
show similar trend towards vertical
Results
• Body line becoming more vertical
• Why?
1. Dancer fitness?– But, both easy/supported and hard/unsupported
show similar trend towards vertical2. Evolution of aesthetic value?
– Standardise stick figures– Aesthetic preference testing, 12 naïve subjects
Results
Schematic Figures
Schematic Figures
Right finger tip
Right Toe
Left finger tip
Left toe
Right Toe
Right finger tip
Left finger tip
Left Toe
Outline polygon
Stick figuresp=0.004, r=0.400
Polygonsp=0.006, r=0.387
Stick figuresp=0.004, r=0.400
Polygonsp=0.006, r=0.387
Conclusions
• Gradual aesthetic evolution within an art form• Socio-aesthetic feedback loop
Artist(Choreographer,
Dancer)
ObserverEvaluative Feedback
1. Aesthetic experience can be studied scientifically
2. Aesthetic experience has neural correlates
3. Brain network for aesthetic evaluation
4. Aesthetic appreciation can be learned/taught
5. Aesthetics is part of rich socio-cultural exchange
Overall Conclusions
End
1. Seeing body postures
1. Seeing body configuration
EBA V1 SPL vPMcInvers
ion e
ffect
(% a
ccura
cy d
iffere
nce
)
• Dance postures are configurations of the body• The human brain sees dance as ‘visual wholes’
• The premotor cortex underlies this way of seeing• Configural vision may be learned
1. Seeing body configuration
-30
-20
-10
0
10
20
30
40
50
Naïve subjects Expert Dancers
Groups
Bo
dy
inv
ers
ion
eff
ec
t:
up
rig
ht-
inv
ert
ed
(m
s)
Ballet postures
Asian dance postures
1. Seeing body postures
2. Liking body movements
3. Evolving aesthetics?
Roadmap
Ballet Capoeira
2. Liking body movements
Non-expert
Non-expertSubjects
Ballet dancers
Capoeira dancers
Expert
Expert
Non-expert controls Non-expert Non-expert
Expert - non-expert
• Expert dancers see dance differently from non-experts• Because they have stronger sensorimotor responses to
watching dance
Experts show more sensorimotor brain activity than non-experts
Mirror neuron areas
Ballet Capoeira
2. Sensorimotor Liking
Non-expert
Non-expertSubjects
Ballet dancers
Capoeira dancers
Expert
Expert
Non-expert controls Non-expert Non-expert
• 6 naïve subjects, retested after brain scanning• 24 ballet and capoiera moves• “How much do you like this move?” (1-5)• Find brain areas that correlate with liking?
2. Sensorimotor Liking
a b• Visual areas• Right premotor
“Neurotargeting”
• People disliked this, right premotor cortex inactive
• People liked this, right premotor cortex active
• Aesthetic evaluation of dance has measurable neural correlates
• Dance may be aesthetically successful because of specific effects on the brain:
‘All visual art must obey the laws of the visual system’ (Zeki and Lamb, 1994)
• But, note the problems– Average of 6 subjects, not universal– ‘Liking’ is not aesthetically rich: other aesthetics?
2. Sensorimotor Liking
1. Seeing body postures
2. Liking body movements
3. Evolving aesthetics?
Roadmap
• Does dance change over time? How? Why?• Same choreography
– Different dancers– Different aesthetics– Different communicative messages
• Rose Adagio, Sleeping Beauty• 1946-2003. Royal Opera House Archive
3. Evolving aesthetics
ARABESQUE PENCHEE
120
130
140
150
160
170
180
190
1945 1955 1965 1975 1985 1995 2005
Years
Le
g E
lev
ati
on
(d
eg
)
r=0.8348
DEVELOPPE A LA SECONDE
r=0.779390
110
130
150
170
190
1945 1955 1965 1975 1985 1995 2005
Years
Leg
Ele
vati
on
(d
eg)
r=0.7793
1962 1979 1996 2003
A
B
1962 1979 1996 2003
• Physical fitness or Aesthetic evolution?
3. Evolving aesthetics
• Aesthetic judgement12 naïve observers50 stick figures
• 50 boundary shapes
60
70
80
90
100
110
120
1945 1955 1965 1975 1985 1995 2005
YearsL
eg
Ele
va
tio
n (
de
g)
r=0.6473
ARABESQUE
1962 1979 1996 2003
C
• Does dance change over time? How? Why
• Interpretations of a single choreographic moment have changed
– Dancer fitness– Artistic interpretation– Aesthetic effect
3. Evolving aesthetics
• Performing art is looking at bodies in motion(plus many other aspects not yet studied)
• Dance uses special ‘ways of seeing’ in the brain
• Identified visual and motor brain systems for dance
• These systems contribute to aesthetic experience
Overall conclusion
• End
1. Seeing body postures
1. Seeing body postures
• Are dance postures also seen as configurations?– Or as individual body parts?
• Which brain areas underlie configural vision?
1. Seeing body postures
V1
Configural System
Body part System
Transcranial Magnetic Stimulation (TMS)
X
Reduced body inversion effect
V1
Configural System
Local System
V1
Configural System
Local System
V1
Configural System
Local System
V1
Configural System
Local System
V1
Configural System
Local System
V1
Configural System
Local System
Inversion Effect
Upright > Inverted
NoInversion Effect
Upright = Inverted
IncreasedInversion Effect
Upright >> Inverted
X
X
X
Upright Inverted Predicted effect
Scientific Approach
Advantages
• Simplify• Measure• Explain• Predict
Disadvantages
• More than sum of parts• Misses the subjective
element, what “I” feel• Culture-bound
Observe, clarify, hypothesise, but appreciateReduce, but also respect
Proprioception:
Sensory information about the state of the body
Receptor neurons in:MusclesSkinJoints
Sensations from muscles
• Key role in maintaining posture
• Can be trained, superior in dancers
• Gives a coherent sense of body– Primary level: muscle length/joint angle– Secondary level: spatial configuration of
the whole body (body scheme)
Proprioception
Coherent sense of bodyDe Vignemont, Ehrsson & Haggard (2005)
ControlBiceps Triceps
x
y
Where is my hand?Jola, Davis & Haggard (in prep.)
Where is my hand?Jola, Davis & Haggard (in prep.)
Proprioceptive:Proprioceptive
Matching(PP)
Visual:Proprioceptive
Matching(VP)
VisuoProprioceptive:Proprioceptive
Matching(VPP)
PP VP VPP
PP VP VPP PP VP VPP
-3
-2
-1
0
1
2
3
-10 0 10 20 30 40 50 60
-3
-2
-1
0
1
2
3
-10 0 10 20 30 40 50 60
-3
-2
-1
0
1
2
3
-10 0 10 20 30 40 50
-3
-2
-1
0
1
2
3
-10 0 10 20 30 40 50
-3
-2
-1
0
1
2
3
-10 0 10 20 30 40 50
-3
-2
-1
0
1
2
3
-10 0 10 20 30 40 50
PP VP VPP
Exp
ert
Dan
cers
Nor
mal
su
bjec
ts
Conclusions
• Proprioception is superior in dancers– Lower matching error
• Coherent body image: Secondary Proprioception– Dancers have a less fragmented body image
– Common egocentre, “core stability”
• Future directions…– Proprioceptive representation of leg position?
– Stability of proprioceptive representation over time?
Doing dance:internal sensation1. Proprioception(2. Motor prediction)
Watching dance:external sensation3. Motor simulation4. Neuro-aesthetics
Premotor cortex: “Mirror” neurons
• Does a similar brain system exist in man?• We understand the actions of others by simulation
Active when a monkey performs a specific action
Or watches another person doing the action
Non-expert
Non-expert
Sensorimotor responses to dance:Watching actions you can do or can’t do…
Subjects
Ballet dancers
Capoeira dancers
Dance videosBallet Capoeira
Expert
Expert
Expert - non-expert
• Expert dancers see dance differently from non-experts
• Because they are better able to simulate what they see
• Strong sensorimotor responses to watching dance
Experts show more sensorimotor brain activity than non-experts
Mirror neuron areas
Visual vs. motor experience in acquired skills
Set of gender-specific ballet moves
Subjects
Female dancers
Female moves
Male dancers
Male moves
Visual familiarity
Visual familiarity
Visual familiarity
Visual familiarity
Motor familiarity
Motor familiarity
Fem
ale
Female
Mal
e
Male
Visual familiarity
Visual familiarity
Visual familiarity
Visual familiarity
Motor familiarity
Motor familiarity
Su
bje
cts
Gender specific actions
=
Pure motor expertise effect
Inte
ract
ion:
Motor expertise+ gender congruence
gender congruence
MaleFemale
Gender-common actions
Visual familiarity
Visual familiarity
Visual familiarity
Visual familiarity
Motor familiarity
Motor familiarity
Motor familiarity
Motor familiarity
Pure motor expertise effect: mirror system
SPM t: interactions: gender-specific – gender-common
1. Left IPS: -42, 57, 48
2. Left dPMC: -48, 6, 45
Conclusions
• A strong motor component to watching dance
• Watching dance activates the same brain structures as performing the corresponding dance moves
• Potential application in dance training/rehabilitation?
• Depressing for the rest of us?
Doing dance:internal sensation1. Proprioception(2. Motor prediction)
Watching dance:external sensation3. Motor simulation4. Neuro-aesthetics
• 3 aspects of aesthetic response
Why does dance have aesthetic appeal?
• Dance is universal in human culture• Aesthetic appeal of dance may exploit natural
patterns of brain activity
• Visual• Emotional
• Sensorimotor
simple-complex
dull-interesting
dislike
-like
weak-powerful
lowest rated highest rated
Visual form
Sensorimotor entrainment
7 naïve subjects
Consensus ratings
Aesthetic dimensions
(Berlyne, 1970)
Strongly Liked
a2.
a1.
b2.
b1.
Strongly Disliked
Consensus ratings
Neural correlates of aesthetic judgement
a. b.
Best
Worst
Best
Worst
Vis
ual
Mot
or
Premotor cortex tuning
Best
Worstb.
a.
Visual cortex tuning
Best
Worstb.
a.
Conclusions
• Not so depressing for the rest of us…
• Premotor cortex activity correlates with aesthetic evaluation
• Degree of movement has strong influence on aesthetic evaluation
• Strong motor component to dance aesthetics
Aesthetics as visual balance(McManus et al., 1985)
GroupA
GroupB
GroupA
GroupB
GroupA
GroupB
Aesthetic balance
Some body configurationsare especially pleasing
Aesthetics may reflect abalanced visual whole
Do dance postures exploitbasic visual brain mechanisms?
“Line”
What brain areas correlatewith aesthetic evaluations?
Aesthetic geometry
• Body ‘line’ as aesthetic pattern
• Historical evolution of line
• Fixed position in choreography– Developpe from Rose Adagio, Sleeping beauty– Thanks to ROH archive
• How and why has this position changed in 75 years?
Aesthetic geometry
Angle L defined by the two legs, fit line based
[the larger the angle, the higher goes the dynamic leg]
90
105
120
135
150
165
180
1956-1979 1986-1996 2002-2004
deg
judge1 judge2 judge3
Aesthetic geometry
Conclusions
• Body geometry changes, even for fixed choreography– Dancer fitness?
– Evolution towards more aesthetic position?
• ‘Line’ may be getting more vertical
• Future directions:– Are some body lines aesthetically preferred?
– Do these preferences reflect tuning of specific brain areas?
• Thank you
Viewing wholes
Different!(EASY)
Er, Different!(Much Harder)
Inversion Effect:Reaction Time Difference
Viewing wholes
Different!(EASY)
Er, Different?(Much Harder)Inversion Effect:Reaction Time DifferenceLarger for bodies/faces thanNon-biological stimuli
0 500 600 1100 Response:“Same” or“Different”
ms
Visual form: Inversion Effect
0 500 600 1100ms
Visual form: Inversion Effect
Response:“Same” or“Different”
Inverted bodies:- judgement is harder, slower- body processed as an overall visual form
Dance postures & control stimuli
Next step: making “Pseudopostures”
Ballet posture
Bharata Natyam posture
Top BNBottom Ballet
Questions:
Has choreography evolved to prefer ‘global’ body positions?
Does looking globally make things look aesthetically better?
3. The paradox of vision
• Vision is the key sense for the audience• Vision raises several computational problems for
the dancer’s brain:
1. Too slow to contribute to movement control 2. No clear relation to motor commands3. Over-rides proprioception
• Helpful for reinforcement, not for online control• How should dancers use mirrors?
2. Prediction
When we move fast, the brain can’t wait for the body to detect and respond to an error
The problem of feedback delays
Sensory delay: 25 ms
Motor delay: 25 ms
The brain’s solution:Predict the consequences of our movements as we make them- don’t wait for sensory feedback
• Brain sends command to leg muscles
• Cerebellum receives copy of command
• … predicts how the body will move
• … asks for a corrected command
• Total time < 20 ms !
Prediction is important for:
• Speed of movement• Accuracy and coordination
• Learning depends on changing connections in the cerebellum
• Genetic factors limit predictive learning ability?
Doing ballet1. Proprioception2. Motor prediction
Watching ballet3. Motor simulation4. Neuro-aesthetics
Background
• Cortical facilitation during action observation– Fadiga (1995), Strafella & Paus, Aziz-Zadeh– Action understanding vs low-level resonance?
• All studies assume derivative simulation– Predict effects for viewing others’ actions are a reduced version of effects for viewing one’s own actions– (Patuzzo et al., Maeda et al.)
• Self/other comparison generally involves several visual confounds