Interaction in social context : Action semantics

Question 1

Outline

Answer 1

I. Evocation of motor information during object recognition

II. Functional role of motor information evocation

III. Various types of motor information, different status

a) Distinction between structural and functional gestures
b) Distinction between HOW (gestures) and WHY (intention)

IV. Automaticity and plasticity of motor involvment

a) Contextual modulations
b) Impact of novel interactions

Behavioral, neuropsychological, neuroimaging, developmental evidence

Question 2

Traditional model

Answer 2

Riddoch & Humphrey, 2001
From vision to semantics :
- Processing basic components
- Features binding
- Structural description system
- Semantic system- Naming

Dorsal stream > Visual route for action
Ventral stream > Visual route for recognition

Traditional approach
Object visual recognition :
- involves retrieving object visual and semantic representations
- Relies on the occiptio-temporal cortex (ventral stream)
- Does not involve retreiving motor information
- Does not rely on the parietal cortex
- Is not influenced by how we interact with objects

Question 3

Def : recognition ?

Answer 3

Accessing some knowledge from an object.
Naming, categorizing (high or low level), reading a word

Question 4

Embodied cogntion approaches

Answer 4

Embodied cogntion approaches

Cognitive processes are simulations of previously-acquired sensory and motor experiences

Object visual recognition :

• Involve retreiving perceptual and motor information
• relies on distributed neural network including both the temporal and parietal cortices
• Is influenced by how we interact with objects

Question 5

I. Evocation of motor information during object recognition

Answer 5

I. Evocation of motor information during object recognition

> Some motor information is evoked during object recognition
Stimulus-response compatibility effects : fraise+pinch =++
Action-object priming : main avant objet compatible = ++
Object-Object priming : objet avec meme action = ++

> Object recogntion activates fronto-parietal regions of the visuomotor system
Chao - fMRI
Pulvermuller : noun reading

> Motor information may not participate in object information : it could be an epiphenomenom (Hickock 2010)

Question 6

I. Optic ataxia / Visual agnosia / Apraxia

Answer 6

Optic ataxia (dorsal stream)
Typical deficit of the dorsal stream; impaired visually)guided actions towards objects but preserved object recogntion (cf patient AT Jeannerod)

Visual Agnosia (ventro stream)
typical deficit of the ventral stream : impaired object recognition but preserved reaching and grasping (cf patient DF Milner & Goodale)

Apraxia (ventro-dorsal stream)
inability to perform particular purposive actions, as a result of brain damage

Question 7

II. Functional role of motor information evocation

Answer 7

II. Functional role of motor information evocation

Object recognition is less efficient when motor information is not accessible :
> motor interference (crush ball, 3 step mvt) - lesion to brain regions associated to the visuo-motor system (Buxbaum 2010 : VLSM)

Critical regions for action and object recognition overlap in the posterior temporo-parietal cortex
→ Some motor information has a functional role in object recognition

But it does not always work …
→ Need to characterize what type of information is integrated in object concepts

Question 8

III. Various types of motor information, different status

Answer 8

III. Various types of motor information, different status

A single object can be associated with many different gestures. In particular, distinct structural (move) and functional (use) gestures.

→ Do those gestures representations have the same role in object recognition ?

a) Distinction between structural and functional gestures

Functional and structural gestures rely on different cognitive processes and are neuroanatomically distinct.

Functional gestures are more likely than structural getures to be a semantic feature of objects and to participate in object recognition.
[Campanella & Shallice 2011 - Incremental Semantic Interference Phenomena ISIP]

→ Are gesture representations the only level of action representations involved in object recognition ?

b) Distinction between HOW (gestures) and WHY (intention)
Actions may represented at a dictinct hierarchical levels Gestures VS Intention

Verbal action intentions can influence object recognition

Many remaining questions about the role of intention-level action representations in object recognition

→ Are intention representations a semantic feature? A context ? Are they represented in a different way as object functional features ?

Question 9

IV. Automaticity and plasticity of motor involvement

Answer 9

IV. Automaticity and plasticity of motor involvement
Motor information :
- “Some” is critical
- “Some” = functional gestures (+ action intentions?)
- Is “some” always critical

a) Contextual modulations
b) Impact of novel interactions

Action evocation can have a functionlal role in object recogntion, but it is not automatic

The involvement of motor information in object visual recognition depends on the context, and may be modified by sensorimotor experience

Question 10

Conclusion

Answer 10

“Action semantics: A unifying conceptual framework for the selective use of multimodal and modalityspecific object knowledge”

Van Elk, Van Schie, & Bekkering, Physics of Life Reviews, 2014