Multiple semantic memory systems or a unitary semantic store?

Question 1

4 main neurological disorders associated with impaired semantic knowledge

Answer 1

• Stroke: extensive lesions in the left hemisphere often lead to aphasia, which can include poor performance on semantic tests
• Herpes simplex virus encephalitis – a viral infection which can affect medial and lateral temporal lobes, often associated with poor semantic memory but also impaired episodic memory and other functions
• Alzheimer’s disease – most common form of neurodegenerative disease, which impacts on medial and later lateral temporal lobes and is associated with profound amnesia, and later semantic and other deficits.
• Semantic dementia – relatively rare neurodegenerative disease, associated with progressive atrophy of anterior temporal lobe(s) and selective impairment of semantic memory, later spreading to other regions and associated functions
  
  • First reported by Warrington (1975). Associated with profound degradation of semantic memory with relative sparing of most other cognitive dfunctions.
  • Hodges et al. 1992: the disorder follows progressive atrophy of the temporal lobes, usually bilaterally but often affecting the left temporal lobe more than the right. Atrophy begins in the temporal pole, spreading posteriorly and medially as the disease progresses.

Question 2

Evidence of modality-specific semantics: stroke patiesnt who exhibited impairments in semantic knowledge only when tested through one modality

• Patient TOB: deficit only apparent when tested using words, performed well with pictures

Answer 2

McCarthy and Warrington 1988

Question 3

Evidence of modality-specific semantics: stroke patiesnt who exhibited impairments in semantic knowledge only when tested through one modality

• Patient DRS
• Very impaired at picture naming but accurate at naming from verbal description.

Answer 3

Warrington & McCarthy 1994

Question 4

Pattern of deficits in semantic dementia doesn’t fit with idea of multiple semantic memory systems:

Longitudinal study of naming performance in patient JL, who demonstrated progressive deterioration of semantic knowledge, with the pattern of his naming error suggesting gradual ‘pruning back’ of the semantic tree. Progressively worsening deficits are seen on tasks that tap semantic knowledge through all modalities of input and output, with consistency or errors across all tasks.

Answer 4

Hodges et al 1995

Question 5

Proposed a unitary semantic store:

Argued that apparently modality-specific effects observed in some stroke patients reflect disruption to one of the access routes to the unitary semantic store.

Answer 5

Caramazza et al 1990

Question 6

Examined activity in healthy volunteers during performance of semantic memory tasks that used either words or pictures.

Found evidence for a common semantic system for words and pictures

PET study. Contrasted activity during two semantic tasks (probing knowledge of associations between concepts and knowledge of the visual attributes of these concepts) and a baseline task (discrimination of physical stimulus size) performed with either words or pictures.

• Modality specific activations unrelated to semantic processing occures in the left inferior parietal lobule for words, and the right middle occipital gyrus for pictures.
• Semantic network common to both words and pictures extended from left superior occipital gyrus through teh middle and inferior temporal cortex to the inferior frontal gyrus.
• Picture-specific activation related to semantic tasks: left posterior inferior temporal sulcus
• Word-specifica semantic activation: left superior temporal sulcus, ledt anterior middle temporal gyrus, and left inferior frontal sulcus.

Thus: semantic tasks activate a distributed semantic processing system shared by both words and pictures, with a few specific areas differentially active for either words or pictures.

Answer 6

Vandenberghe et al 1996 Nature

Question 7

Category specific semantics? Some studies of stroke patients suggest that different categories of knowledge may be represented in different brain regions

• Reported patient VER who was selectively impaired at naming pictures of nonliving things, performing well at naming even low frequency examples of living things.

Answer 7

Warrington and McCarthy 1983

Question 8

Reported 4 patients who were selectively impaired at tasks tapping semantic knowledge about living things

Proposed that category specific semantic disorders could be explained by a sensory-functional distinction. Living things are defined more by their sensory properties (colour, shape, taste, etc., and non-living things are defined more by functional properties (what they are used for).

BUT closer analysis suggests that this view may be too simple

• Patient JBR, one of those reported by W+S to show a selective impairment of knowledge of living things, performed well on naming body parts (an example of a living thing) but was poor at some nonliving things e.g. precious stones, musical instruments.

Answer 8

Warrington & Shallice 1984

Question 9

Some studies reporting category specific semantic deficits have been criticised because the different categories used differed on factors like frequency, familiarity, age of acquisition, etc,

BUT this study: controlled for all these factors and still found category specific deficits in patient EW.

Answer 9

Caramazza and Shelton 1998

Question 10

Evidence that the view of category specific semantid disorders explained by a sensory-functional distinction may be too simplistic.

Reviewed 79 cases of category-specific semantic impairment in the literature, finding strong evidence for other category distinctions within the living and nonliving domains.

Review, aimed to answer two questions:

• What are the categories of category-specific deficits?
• Is there an interaction between impairment for a type of knowledge and impairment for a given category of objects?

Concluded:

• The categories of category-specific semantic deficits are animate objects, inanimate biological objects, and artefacts.
• Domain of inanimate biological objects fractionates into two independent semantic categories: animals and fruit/vegetables.
• The types of category-specific deficits are not associated with specific types of conceptual knowledge deficits.

Answer 10

Capitani et al 2003

Question 11

Functional neuroimaging study.

Observed considerable variability in activity associated wiht naming pictures of people, animals and tools

Answer 11

Damasio et al 1996

Question 12

Reviewed a number of studies indicating that object categories are represented in distributed networks that parallel the organisation of perceptual, motor and language processing systems in the brain.

• E.g. knowledge about tools activates regions associated with motion perception and action planning (Martin et al., 1996)
• Knowledge about characteristic object colour activates areas associated with colour perception (Martin et al., 1995).

Answer 12

Martin and Chao 2001

Question 13

Proposed a semantic hub view:

• Various category-specific regions connect to a general semantic hub in the anterior temporal lobe, where associations between different object attributes are processed to allow generalisation, inference and discrimination between overlapping concepts.
• This view can explain the data from patients with strokes, those with semantic dementia, and the neuroimaging data.
• Also consistent with the semantic hub view, Patterson et al. reviewed 40 functional neuroimaging studies of semantic processing, finding consistent activation in anterior temporal cortex irrespective of the kind of task, type of stimuli involved, etc.

Answer 13

Patterson et al 2007