10: Perception of Action

Imitation and Mirror Neurons

crossmodal transfer

the transfer of abilities across different sensory modalities

perception-action mapping

the ability to seamlessly map visual representations of actions onto our motor systems to produce a copy of the action

developmental evidence for perception action mapping

infants can imitate caregiver’s facial expressions, hand and mouth movements, head turns, etc.— babies build up a representation of the visual image of the caregiver’s face/mouth and map this onto their own motor representation of the movement

age that babies can allegedly imitate specific facial expressions (meltzoff and moore, 1977)

12-21 days

what specific acts do infants imitate (meltzoff and moore, 1977)

lip and tongue protrusions, open mouth

evidence against neonate imitation

when does true imitation occur (oostenbroek et al., 2016)

6-9 months

active intermodal matching (AIM; meltzoff & moore, 1997)

neonates recognise equivalences between body transformations they see and those of their own body that they “feel” themselves— their emotional expressions induce adults to produce similar expressions, which provides the infant with a visual input to match his motor output

AIM involves:

• perception and action having independent coding/representation

• a “specialist” module for imitation

aim vs other models (e.g. IM and ASL)

others posit common coding for perception and action, and an imitation part of “generalist” processes for motor control and learning

ideomotor (IM) theory

close link between motor movements and perception as they are very similar

associative sequence learning (ASL)

emphasises learning through experience (e.g. see consequence of own hand action)

dual route of imitation (rumiati & tessari, 2002)

• semantic: meaningful actions, stored in repetoire

• visuomotor/direct: meaningless actions— mirror neurons

semantic route of imitation

stored representation of doing things

visuomotor/direct route of imitation

carefully observe an action and then map it onto one’s own motor system

mirror neurons

bimodal, visuo-motor neurons (respond to both visual and motor stimuli) that discharge when an individual performs an action and when they observe the same action performed by another individual

mirror neurons and action understanding (umilta et al., 2001)

mirror neurons active during observation of partially hidden actions predicts action outcome even in absence of complete visual information (no response in the absence of an object)

mirror neurons and action understanding (kohler et al., 2002)

audio-visual mirror neurons respond to the sound typically produced by the action (though no activity in audio-only stimulus, as macaques are not aware of what is about to be done until it is done/do not imitate)

mirror neuron properties

• somatotopically organised

• responds only to goal-directed actions

• also canonical visuomotor neurons (also called “object observation-related” neurons)

mirror neurons in monkey

found in area F5 of premotor cortex and inferior parietal lobe

mirror neurons in humans (location)

human homologue in

• broca’s area (B144)

• ventral inferior frontal gyrus (BA6)

• posterior parietal lobe

• superior temporal lobe

indirect evidence of human mirror neurons

• close link between perception and action

• behavioural

• brain imaging (fMRI)

• transcranial magnetic stimulation (TMS)

direct evidence of human mirror neurons

recording from neurons (difficult to obtain as it unlikely to be ethically cleared due to risk)

behavioural human mirror neurons

faster responses observed when compatibility between observed and executed movements, known as “automatic imitation” (brass et al., 2000; 2001)

automatic imitation

lab analogue of mimicry

human mirror neurons and brain imaging

somatotopic activation of pre-motor and parietal cortex, areas correspond to observations of different body parts (buccino et al., 2001)

perception-action overlap (hardwick et al., 2018)

overlap in brain activity between imagined, observed and executed movements

motor imagery

imagined movement without action

TMS and human mirror neurons (fadiga et al., 2005)

motor evoked potentials to show that observing an action produces increased motor excitability

direct recording of human mirror neurons (mukamel et al., 2010)

action observation-related neurons found in medial frontal lobe (SMA) and medial temporal lobe (hippocampus)

mukamel et al. (2010) study

recorded from 1177 neurons in 21 patients undergoing surgery for intractable epilepsy

examples of mirror neurons beyond movement (bonini et al., 2022)

• evolution of language

• empathy

• social cognition

• “broken mirror” theories of autism and schizophrenia

intersubjectivity

imitation, empathy and intention reading allowing us to predict the behaviours of others

empathy and direct mapping

viewing pain in others and feeling pain yourself has an overlap in anterior cingulate cortex (BA24b; morrison et al., 2004)

limitations in primate data (dinstein et al., 2008)

• small number of examples

• often qualitative rather than quantitative

• need more studies to ask how well cells can distinguish pairs of movements (e.g. keysers et al., 2003)

• need evidence of mirror neurons firing in spontaneous social interaction

limitations with human data

• many areas outside of mirror neuron areas are activated during action observation

• adaptation protocols concerning what neurons are firing

• TMS effects could be produced by areas outside of mirror neurons

use of adaption

can be used to explore whether the same area is involved in different tasks

problems for mirror neuron theory

• motor theories of perception are not new

• over-emphasis on “action-understanding” function

• do mirror neurons go beyond other sensory-motor neurons?

evaluation of mirror neurons supporting action understanding in monkeys

• inactivation of F5 disrupted grasping but not perception

• could just be association or working memory

• no measurement of understanding

evidence of action understanding existing without mirror neurons

• F5 also responds to objects, but it is not argued to underpin understanding of objects

• superior temporal sulcus may be more critical for action understanding

human-monkey differences

• higher cognitive functions attributed to mirror neurons are not seen in monkeys

• assumed that MNs in humans have developed to include both action understanding and imitation

• cannot assume that conclusions from monkey MNs apply in humans (few human imaging studies examine overlap between observation and own action