The role of motor system in action understanding Flashcards
Rizzolatti et al. (1996)
studying Macaque monkeys’ ventral premotor cortex, recording neurons in these regions for different types of grips. Neurons also shown to discharge during the observation of action - by chance, fired when scientists made action.
Rizzolatti et al. (1996) Area F5
Both observation and execution of action trigger the discharge, neuron not interested in properties, just the action. Pliers didn’t cause activation when they were naive, but once monkeys learned how to use them they did.
Gallese et al. (1996) F5
Recorded >500 neurons, ~90 classified as mirror neurons and then ~50 defined as congruent to specific actions, broad neurons discharge to less specific action, have less specificity.
Neurons found bilaterally, no marked evidence for hemispheric effects.
Gallese et al. (1996) area PF
Directly connects to F5. After initial movement to reach and grasp, monkey has options (move to mouth, move to place). Planning to do one changes discharge in area PF.
Mirror neurons here seem to focus on overall goal.
The mirror system
As well as F5 (premotor cortex) and PF there is STS (Superior Temporal Sulcus), with particular discharge for bio motion.
Useful identifying bio moving entities in the world. Has no motor response, just visual response.
What evidence to date of mirror neurons in humans?`
Only indirect because can’t invasively record
Fried et al.
Single-neuron recordings show neurons discharging at observation and execution of hand grasping actions and facial emotional expressions.
However, activating areas related to memory so be activating memories to do with action.
Fadiga et al. (1995)
First evidence of motor system activation.
Used Transcranial Magnetic Stimulation (TMS). Motor Evoked Potentials (MEPs) from hand muscles discharge of neurons.
TMS while participants (1) observed experimenter grasping object, (2) looked at the same object, (3) observed experimenter tracing geometrical figure in air with arm.
If population of neurons active during action execution then applied pulse should increase/excite and cause a twitch.
MEPs significantly increased in conditions where participants observed . Pattern reflected pattern of muscle activity recorded when participants executed observed actions.
Seems to resemble monkey system - activation matches between observation and action.
Hari et al.
Stimulated nerve causes increase in beta oscillation in motor system.
What have fMRI/PET studies shown about mirror neurons?
Meta analysis has shown activity in F%/STS when observing acts.
Main criticisms of fMRI/PET data?
Poorly defined Region of Interest (ROI).
Very few studies compare execution to observation, probably because it’s difficult to do movements in scanners.
Analyse specific areas in which lots of neurons are likely to exist. Overlap of activity doesn’t mean that the same neurons are active, could be different populations.
Dinstein et al. (2007)
Ps played rock-paper-scissors against videotaped opponent.
Sometimes would repeatedly observe or execute the same movement on subsequent trials.
If mirror neurons exist then response suppression should occur for cross modal instances.
Only found in six small areas of brain than previously thought large areas.
Chong et al. (2008)
Found mirrorlike activity in the parietal lobule (area pf) but not the inferior temporal gyrus (area f5 in monkeys).
Kilner et al. (2009)
Used simple gripping task, real life actions instead of videos and more precise scanning and demonstrated mirrorlike activity in ventral premotor cortex/inferior frontal gyrus.
What do mirror neurons do?
Mirror neurons seem to underlie a long list from action understanding to speech perception, to empathy.
Seems to focus on action understanding.
How do mirror neurons mediate understanding of actions?
Rizzolatti and Craighero (2001): Whenever someone sees someone else make an action, neurons representing the action are activated in the premotor cortex. This representation corresponds to that which is spontaneously generated during active action and whose outcome is known to the acting individual. This transforms visual information into knowledge.
Kilner et al. (2007) visual input activates STS/parietal MN/frontal MN, using this activation we work backwards to establish intentions
What problems are there with a recognition model account of how mirror neurons mediate understanding of actions?
Patients with lesions to MN areas are still able to infer the intentions of others.
Also not always 1:1 mapping of input and cause.
What does predictive coding consist of?
Generative model and recognition model.
Generative model makes predictions of visual consequences of someone else’s action (backwards connection). If wrong prediction then prediction error is account for and changed accordingly (forward connection).
What is the generative model?
Based on internal input decisions the generative model produces an output.
What are the three steps in action selection in action execution?
1) Visual processing of object (recognise as cup)
2) Retrieve from semantic memory all actions associated with object
3) Choose most appropriate action (semantic selection)
