#10626 - Semantic Dementia - Neuropsychology of Memory

Semantic Dementia (1)

Semantic Memory general knowledge about the world – memory for facts, ideas, and concepts – associated with a feeling of “knowing” gives meaning to our sensory experience - it is culturally shared , not temporally specific + is to a large extent acquired early in life

Tulving’s taxonomy of memory -> LTM divided into declarative + non-declarative declarative – divided between – episodic + semantic

MODELS OF SEMANTIC MEMORY how are concepts organised in semantic memory ?

Spreading Activation Model Collins & Loftus (1975)

• Notion of logically organised hierarchies is too inflexible

• Preferable to assume that semantic memory is organised on the basis of semantic relatedness or semantic distance

• Semantic relatedness can be measured by asking people to decide how closely related pairs of words are/list as many members of a particular category

• Those which are produced most often are regarded as most closely related to that category

• Whenever a person sees, hears or thinks about a concept the appropriate node in semantic memory is activated

• This activation spreads most strongly to other concepts that are closely related semantically

Parallel Distributed Processing Model McClelland & Rumelhart (1985)

Model assumes that info processing takes place through interactions of large numbers of simple processing elements (units) – neurons, which send excitatory or inhibitory signals to other units

• Processing takes place by the spread of activation among simple, neuron like processing units

• Semantic info is not stored as such, but is reconstructed in response to probes , in a process called pattern completion

• Info not input in step by step manner info is distributed to all parts of the memory system at once

• Filling in occurs through the spread of activation among units + their connections

• The outcome depends on the strengths (weights) of the connections

• These are shaped by experience

• E.g. “A canary is...” – this is filling in

• The model has a feed-forward structure activation flows only in one direction

• From units which represent items (e.g. canary) + relations (e.g. is a) through ‘hidden’ layers to an output layer containing units corresponding to possible completions to 3 constituent propositions e.g. fill a canary is a…

• Hidden units combine concept + relation info

Schemas & Scripts

Semantic memory doesn’t just store info about objects or animals – but also about more complex knowledge bases

Schema well integrated chunk of knowledge about the world, events, people or actions

Scripts type of schema relating to typical sequences of event – e.g going to a restaurant

Frames knowledge structures referring to aspects of the world containing fixed structural info

Schemata is useful for 4 reasons:

1. Forms expectations e.g. in restaurant expect to be shown table + given menu

2. Enables us to fill in the gaps when reading or listening schema allows us to fill in the gaps in what we see/hear – enhances understanding

3. Assists in the perception of visual scenes activation of schematic knowledge facilitates visual perception – make inferences about what we expect to see if a given scene

4. Efficiency Anderson & Schooler (1991) don’t need to remember all specific details of all experiences – cognitive efficiency

Can amnesics learn new semantic information?

Kitchner et al (1998)

• Case study of RS - severely amnesic

• He can recognise some recent famous faces for the time e.g. John Major

• He can define the meaning of some recently acquired vocan e.g. interent

• Has some knowledge of recent famous events

• BUT

• He didn’t acquire semantic information normally – was slow + laborious – involving many repetitions

• HM also used to use 1950’s vocab – he could not define new words – e.g. biodegradable

• Attempts to teach him 8 new words failed (Gabrielli et al, 1983)

• Patient R.S’s ability suggests that some semantic knowledge can be acquired in the absence of EM

• Seems to support the dichotomy between semantic + EM

• However many amnesics – HM have impairment in acquiring both new episodic info + semantic info

• Semantic problems are not v. noticeable in adult amnesics – little semantic info is learned in adulthood

• Amnesia is seen as a syndrome – not a pure deficit – some have problems in recall + familiarity + therefore both in EM + SM

Developmental Amnesia

• Vargha-Khadem et al (1997)

• 3 patients with amnesia resulting from early events

• Profound amnesia for episodic materials

• Still progressing through school – normal IQ + academic achievement

• Able to comprehend passages + could acquire knowledge about the world

• Seems semantic memory can develop in the absence of EM

• SM may develop normally in developmental amnesia – may be an adaptive thing a– but adults have more difficulties in SM due to EM impairments – more reliant on EM to form SM whereas children adopt coping strategies

Visual Object Agnosia

• Def => failure in visual object recognition which is not due to general intellectual impairment, sensory impairment or language disorder

• Word “recognition” here is used to mean ‘comprehending the meaning’ of the object (e.g. its name or function)

• Rare neurological disorder

• Lissauer (1890) divided agnosics into 2 subcategories

• Apperceptive agnosics => fail to recognise visually presented objects + cannot copy and match - no adequate percept – visual problem (not related to neuropsychology)

• Associative agnosics => fail to recognise visually presented objects yet can copy + match adequately – percept is adequate but no association with stored info this could look like a disorder of semantic memory

Warrington (1975) first report of an impairment interpreted as an impairment of SM

Described 3 patients with cerebral atrophy presenting with progressive anomia + impaired word comprehension

• 3 patients selected for having visual object agnosia – each had a progressive dementia – visual perception intact

• Visual + perceptual tests : shape discrimination, matching faces, matching objects v/similar results to controls shown NOT apperceptive agnosics

• Visual recognition by semantic probe test presented with pictures of animals + object - asked “Is it an animal” – questions become progressively more detailed – asking about the semantic properties of the objects/animals

• Patients did know some details about the objects/animals – but as more semantic info required the more they struggled to correctly answer

• Was not a limited problem in knowing the name of the object – issue in knowing the function or properties of the object

Anomia a specific deficit in naming (Kay & Ellis, 1987)

• Patient EST – presented him with pictures of animals + objects + asked to name them

• He had a specific deficit in naming – yet could provide semantic details about the object

• E.g. snowman “its cold, it’s a man, cold, frozen”

• Verbal comprehension OK + articulation OK

• This is not a problem restricted to the inability to pronounce the words – as when told the name – they could repeat it back

• Found a frequency effect words which are used more frequently were more likely to be named than those which were not frequently used

• Used objects ordered from high frq to low freq = pain, accordion, igloo, mita

• This is a deficit in speech output lexicon (accessing words or names) after having identified the object

• When you have identified the semantic properties of an object but cannot name – “tip of the tongue phenomenon”

N.B –Warrington (1975) – check that not anomia = auditory recognition by semantic probe test

- presented names in auditory

- they had a similar problem – could understand the name of the object – BUT – could not give any semantic info about the objects

cannot be a problem of anomia

SEMANTIC DEMENTIA

• Selective deficit in semantic knowledge first documented by Warrington (1975)

• Snowden et al (1989) first used the term “semantic dementia” – to describe patients with fluent progressive aphasia , deficits in word comprehension + in knowledge about objects + people

Neary et al (1998) criteria for diagnosing semantic dementia

• Core diagnostic features

1. Insidious onset + gradual progression

2. Language disorder characterised by:

• Progressive, fluent, empty spontaneous speech

• Speech production is effortless, without hestiancies, + patient doesn’t search for words – HOWEVER – little info is conveyed = reduced use of nominal terms + increased use of broad generic terms e.g. thing empty nature of speech output

• Loss of word meaning - manifested in impaired naming + comprehension

• Semantic paraphasias – using non-descript words to get around breaks in speech e.g. “thing, that woman” to refer to their wife

3. Perceptual disorder characterised by

• Prosopagnosia – impaired recognition of identity of familiar faces and/or

• Associative agnosia – impaired recognition of object identity

4. Preserved perceptual matching + drawing reproduction shows they have seen it properly

5. Preserved single word repetition indicates they have heard it properly

6. Preserved ability to read aloud + write

Testing Semantic Dementia

Consistently poor performance on this range of tests - the absence of other contributory deficits - is diagnostic of a semantic memory impairment (Hodges & Patterson, 1997)

1. Category fluency tell me names of as many animals in one minute

2. Confrontation naming what is the name of this object/picture?

3. Naming to description what do we call the large African animal with a trunk?

4. Picture sorting sort these pics into living + man made items

5. Generation of verbal definitions how would you describe an elephant to someone who had...