Acquired Neurological Language Disorders HCSNet SummerFest 2006 Prof Helen Chenery

1
Acquired Neurological Language Disorders HCSNet
SummerFest 2006 Prof Helen Chenery
Rules of Engagement1. Sit next to a
stranger!2. Introduce yourself.
2
Learning Objectives

• What has a knowledge of psycholinguistics
• contributed to our understanding of ANLD?

2. What have advances in our understanding of
brain/behaviour relationships brought to ANLD?
3. What has new research contributed to our
understanding and management of people with
ANLD?
4. What have the knowledge and skills from other
disciplines contributed to our understanding of
ANLD?
3
(No Transcript)
4
Course Structure

• I will address each of these collaborations with
  reference to a person who has sustained
  neurological damage.
• A person with nonfluent aphasia
• A person with Parkinsons disease
• A person with schizophrenia
• A person with dementia

5
A person with nonfluent aphasia

• Meet Gerry, a 75 year old man who suffered a left
  hemisphere stroke 16 months ago and has aphasia.

conversation
Group Task With your partner, answer the
following questions?
6
Structured Observation in Aphasia

• Is Gerrys language production fluent or
  nonfluent?
• What severity level is his aphasia?
• Are there any paraphasic errors (words that are
  substituted for others)? If so, give examples.
• What is Gerrys articulation like?
• Can Gerry understand everything that is said to
  him?
• Does Gerry have word finding difficulties?
• Are memory problems causing Gerry to be unable to
  speak?

7
The big questions!

• What has gone wrong as a result of the stroke and
  how can we help Gerry recover his language?
• Firstly, what does psycholinguistics contribute?

Group Task How can we describe Gerrys language?

Picture description
8
(No Transcript)
9
(No Transcript)
10
Agrammatism

• Entails the dropping out of articles, connective
  words, auxiliaries and inflections so that
  grammar may, in extreme cases, be reduced to
  rudimentary form - the use of one or two words
  sentences.
• Yes ah Monday er Dad and Peter H (patients
  name), and Dad er hospital and ah Wednesday
  Wednesday, nine oclock and oh Thursday ten
  oclock, ah doctors two an doctors and er
  teethyah
• (Goodglass Geschwind, 1976)
• Can a theory of normal sentence production help
  us to understand Gerrys language deficits?

11
Garretts sentence production model
12
(No Transcript)
13
Thematic role assignment and Predicate-argument
structure

• Verb Kiss
• Implies that someone is doing the kissing
  (agent) and someone is being kissed
  (recipient/theme)
• Verb Deliver
• Implies that someone (agent) is doing the
  delivering (agent), that some kind of item
  (theme) is being delivered, and that someone is
  receiving the delivery (recipient)

14

• Verb Delivered
• Implies
• agent theme recipient
• John delivered the guns to the military base
• The choice of a particular verb commits a speaker
  to specify the persons or objects related to the
  thematic roles associated with that verb.

15

• Where do these errors come from?
• She offered him some cake.
• He offered her some biscuits.
• He offered her some dake.
• He threw some cake at her.

16
Disturbances affecting the realisation of
Thematic Roles in Agrammatism

• disturbance has been said to affect the patients'
  ability to use the basic word order of English to
  convey propositional features such as thematic
  roles.
• Saffran, Schwartz and Marin (1980) analysed 5
  agrammatic aphasics describing simple pictures of
  actions.Used actions and locatives of like and
  contrasting animacy, e.g.,

17
(No Transcript)
18
Aphasic responses

• Target
• a key is in a suitcase
• a chair is on a car
• a boy is pulling a baby
• a man is washing a baby

• Aphasic response
• the luggage .. The key is the luggage. The key
  is in the luggage
• Automobile is The chair isthe chair was
  under the car.
• The boy and girl was falling down.
• Washing bath me.. boy .. boy.bath

19
Results

• Found that AA could not with any consistency
  produce NP-V-NP orderings that reflected
  underlying semantic roles.
• Most errors occurred for both animate actions and
  both inanimate locations.

20

• noted a strong effect of animacy upon the
  position of nouns around the verbs
• suggested that thematic roles were not mapped
  onto the canonical noun-verb-noun word order of
  English, and that animacy determined the position
  of nouns around verbs in these patients.
• Concluded that AA have either lost the basic
  linguistic notions of thematic roles (gent, theme
  etc) or else cannot use even basic word order of
  the language to express this sentential semantic
  feature.
• Possible that some very severe patients lose
  these concepts or do not attempt to convey them.

21
Also linked to difficulty in producing verbs

• Many AA have particular difficulties with the
  production of verbs.
• Not only ties to trouble in producing the correct
  inflectional and derivational forms of a verb in
  a given context, also affect ability to produce
  verbs themselves, resulting in omissions,
  paraphasias, and nominalisations of verbs

22
Examples of errors affecting the production of
verbs

• Patient attempts to describe the picture of a
  girl presenting flowers to a teacher
• Omission of main verbs
• The young the girl the little girl is the
  flower.
• The girls is .. Going to flowers.
• Nominalizations used instead of verbs
• The girl is flower the woman.The girl is is
  roses. The girl is rosin.
• Picture of a man taking a photograph of a girl
• The man kodaks and the girlkodaks the girl.
• from Badecker and Caramazza (1985)

23
(No Transcript)
24
Verb retrieval - Gordon Dell (2003)
INPUT LAYER Semantic Syntactic

Determiner Noun
Verb Motion
Air

OUTPUT LAYER Lexical items The boy go The bird
fly
25
Verb retrieval - Gordon Dell (2003)
INPUT LAYER Semantic Syntactic

Determiner Noun
Verb Motion
Air

OUTPUT LAYER Lexical items The boy go The bird
fly
Weaker input
Lesion of syntactic input
26
Verb retrieval - Gordon Dell (2003)
INPUT LAYER Semantic Syntactic

Determiner Noun
Verb Motion
Air

OUTPUT LAYER Lexical items The boy go The bird
fly
Weaker input
Lesion of semantic input
27
Verb retrieval in nonfluent aphasia

• Barde et al. (2006)
• Examined verb retrieval in agrammatic and
  non-agrammatic aphasics
• Worse retrieval of simple verbs relative to
  complex verbs was only evident in the group of
  agrammatic aphasic patients.
• Results confirm weakened syntactic input for verb
  retrieval in non-fluent aphasia

28
Approaches to therapy?
29
(No Transcript)
30
Mapping Therapy Byng et al. (Stage 1)

• Present two pictures
• Stan washing the dishes/Mary washing the dishes
• Patient selects a picture
• Four cards with phrases on are presented
• Each phrase has a coloured line below it denoting
  its syntactic class (e.g., red for noun phrases
  and green for verbs)

31
Mapping therapy (Stage 1)

• Three lines, one red, one green and another red
  are also drawn in that order on a blank card.
• The patient has to identify the roles played by
  the entities involved in the event and then
  select the phrases which describe those entities.
• Start with non-reversible events.
• Repeat the process for the second picture. Ask
  patient to identify which part of the sentence
  has been changed.
• Then move on to contrasts between change of
  agent, action, and theme/patient (use different
  stimuli)
• Repeat each set of stimuli with, and then
  without, the colour cues.

32
Mapping Therapy (Byng et al., 1994) Stage 2

• Use stimulus cards from previous therapy and ask
  patient to describe the picture within a
  structured format.
• Give the patient a cue consisting of three
  horizontal lines representing the three
  constituent phrases of a possible sentence, and
  ask patient to describe the components within
  this sentence frame. Having produced any lexical
  item, patient is asked to identify where in the
  sentence frame that item should go.

33
Computer based delivery

• Beveridge et al. (2002) delivered mapping
  therapy to patients via computer program.
• 8 X 1 hour sessions
• comprising picture building or sentence building
  tasks
• minimal input from clinician
• Written sentence comprehension improved
• Treatment effects generalised to untreated
  sentences
• Some generalisation to spoken sentence
  comprehension

34
Verb production therapy

• Training of argument structure assignments
• (Schneider et al., 2003)
• 1) Picture shown to patient
• 2) Patient told This picture shows
• catch. It shows a man catching a ball.
• The man is the person doing the
• catching, the ball is the thing
• being caught
• 3) Process repeated for other pictures.
• 4) Patient shown all pictures again and asked to
  describe them.

35
Functional Neuroimaging
-
time (seconds)
Deoxy
Oxy

• Functional magnetic resonance imaging

36
Neuroimaging sentence production

• Haller et al. (2005)
• Inferior frontal gyrus was the site of most
  activity during sentence generation
• Confirms the critical role of this area in
  syntactic encoding

37
Neuroimaging mechanisms of language recovery

• Neuroimaging studies have revealed various
  mechanisms of language recovery
• Reactivation of functionally disrupted regions
• Up-regulation of other language networks
• Recruitment of homologue areas and/or surrounding
  tissue

38
Mapping therapy fMRI (Wierenga et al., 2006)
Patient 1 72 year old woman Left middle
cerebral artery CVA
Pre-post therapy change Increased recruitment of
Brocas area and adjacent inferior frontal cortex.
39
Mapping therapy fMRI (Wierenga et al., 2006)
Patient 2 58 year old man Left middle cerebral
artery CVA
Pre-post therapy change Reduction of activity in
inferior frontal sulcus middle temporal gyrus.
40
Neuroimaging Wierenga et al. (2006)

• Data suggest that the neural mechanisms
  underlying changes to language function
  post-surgery can differ between participants.
• Further exploration of the neural mechanisms
  underlying recovery is required
• Which may lead to conceptually driven
  rehabilitation strategies and improved treatment
  efficacy

41
What about sentence comprehension?
42
What must a model of sentence comprehension be
able to account for?
(or an Absolutely Fabulous view of sentence
comprehension)

• The model must be able to account for the fact
  that we can detect an acceptable versus an
  unacceptable sentence,
• e.g., Eddie and Patsy went to
• town.
• Eddie and Patsy put to town.

43

• The model must be able to account for the fact
  that we know implicitly about breaking up
  sentences into their parts or pieces
• E.g., Eddie and Patsy went to town
• Vs Eddie and Patsy went to town?

?
44
and

• The model must account for the fact that we can
  make judgements about reference
• e.g., Patsy hit herself.
• Patsy hit her.
• And to recognise sameness of meaning
• e.g., Patsy hit Eddie.
• Eddie was hit by Patsy.
• And to recognise structural ambiguities
• e.g., Patsy drank the champagne
• in the car?

45
and

• The model must be able to account for the fact
  that we can understand spoken sentences at an
  astonishingly rapid rate.
• That most of these characteristics are innate or
  automatic and are performed by us implicitly,
  without conscious awareness.
• And that we can understand quite complex spoken
  sentences if we are required to.
• The man that the child hugged laughed.
• The man kissed the girl who hugged the woman.

46
Friederici, 2001 overview
Processes of reanalysis repair
Syntactic integration
Integration of semantic and morphosyntactic info.
Semantic relations
Identification of lemma and morphological info.
Thematic role assignment
Identification of word category
Syntactic structure building
Lexical-semantic access
Identification of word forms
Identification of phonemes
Phonological segmentation and sequencing
Primary acoustic analysis
47
Lexical-semantic networks
lion
fur
tiger
animal
stripes
48
Investigation of language processing using
semantic priming

• Refers to a reaction time advantage in
  identifying a particular word when it is preceded
  (or primed) by a related word.
• Can be investigated subconsciously or when other
  cognitive systems are active.
• Assessed via computerized tasks.

49
Semantic Priming

Prime
bird
350msec
Prime
?
night

Target
OWL
250 1000 msec
SOA
50
A little aside
I need a volunteer from the audience.
51
Multiple priming task results 250 SOA
52
1000 ms SOA
Delayed lexical access contributes to the
sentence comprehension difficulties associated
with nonfluent aphasia
53
Summary

• Damage to inferior frontal cortex
• A disruption to the integration of semantic
  syntactic information during sentence
  comprehension and production
• Delays in lexical-semantic activation

54
Parkinsons Disease

• Meet actor Richard Moir

55
Language changes in PD

• Impaired naming and definitional abilities
• Reduced verbal fluency abilities
• Difficulties in interpreting ambiguity and
  figurative language
• Reduced performance when comprehending complex
  sentences (e.g., object-relative sentences the
  man that the woman hugged was happy)

56
The controversy

• Are the language changes in PD an isolable
  deficit or are they the result of associated
  deficits in other systems (e.g., executive
  resources, working memory, etc).
• Is language represented and processed
  subcortically?

57
(No Transcript)
58
What causes PD

• Most pronounced neuropathological feature is a
  loss of dopamine neurons in the subcortex (the
  substantia nigra), which leads to a reduction in
  dopamine levels in the other subcortical regions,
  e.g., the striatum.

59
The role of the subcortex in human language
processing

60
The action of dopamine

• Has two functions
• specific effects as a neurotransmitter (phasic DA
  release)
• as a non-specific neuromodulator (tonic DA
  release)

61
Finely balanced system. DA allows the system to
process salient information (the phasic
controlled) while acting to filter out less
salient information (the tonic controlled). DA
modulates the signal-to-noise ratio by both
facilitatory and inhibitory mechanisms
62
In balance (S/N 1)
If remove filter (or make the gaps wider),
parallels situation where tonic DA is
reduced. Leads to increased noise, (S/N
and the DA system would fail to screen
irrelevant information (equivalent to
unfocussed activation).
If put finer filter, amount of information screene
d as irrelevant would increase, and lead to
focussed activation. Situation where tonic DA is
increased, lead to decreased noise, (S/N -
1).
63
Investigation of indeterminacies of meaning using
semantic priming

• Refers to a reaction time advantage in
  identifying a particular word when it is preceded
  (or primed) by a related word.
• Can be investigated subconsciously or when other
  cognitive systems are active.
• Assessed via computerized tasks

64
Rationale

• Proposed that additional dopamine increases the
  signal to noise ratio within semantic networks by
  reducing spreading activation or dampening
  weaker background signals.
• Biased lexical ambiguities provide an ideal
  avenue for investigating this hypothesis.

65
Processing of semantic ambiguities
Bank
River
Money
Prime
Target
Target
Lexical
66
Aims

• To investigate lexical ambiguity processing in
  healthy participants both on and off levodopa,
  and in participants with Parkinsons disease both
  on an off levodopa.

67
Stimulus set contained

• Two meanings frequencies
• dominant
• subordinate
• Two stimulus conditions
• related e.g, ball round (dominant)
• ball dance (subordinate)
• unrelated e.g., left round
• left dance
• and was presented at two stimulus-onset-asynchroni
  es (SOAs)
• 100 msec
• 1000 msec

68
?
Prime
Target

calf
money
350msec
500msec
100 1000msec
SOA
69
The influence of hyperdopaminergic states

• Students given Madopar? which contains 100 mg
  levodopa and 25 mg benserazide or an identical
  capsule of placebo.
• Testing began 60 minutes after ingestion of the
  capsule
• Neither subject nor examiner was informed about
  the capsules content.

70
Results


PE Dom
PE Sub
PE Dom
PE Sub
100 msec
1000 msec
71
Discussion

• Reduced subordinate meaning activation at the
  short SOA in the levodopa group suggests that
  dopamine acts to focus activation by dampening
  weaker signals within the semantic network.
• This process is magnified at the longer SOA with
  both dominant and subordinate meaning activation
  lost.

72
The influence of hypodopaminergic states
PE Dom
PE Sub
PE Dom
PE Sub
100 msec
1000 msec
73
Discussion

• Reduced dopamine enhances signal leading to
  unfocussed activation of semantic information.

74
The effect of increased hypodopaminergic states
PE Dom
PE Sub
PE Dom
PE Sub
100 msec
1000 msec
75
Sentence processing in PD Electrophysiological
investigations
76
Semantic ERP components

• N400 Reflects lexical-semantic processing

Correct The cat drank the milk
Incorrect The cat drank the ball
From www.sissa.it/cns/lcd/calacei/calacei/Site/CA
LACEI-2006_files/10.Friederici.CALACEI.May06.ppt
77
Syntactic ERP components

• Early left anterior negativity (ELAN)
• Reflects early automatic phrase structure
  building processes
• P600
• Reflects late syntactic integration processes

Correct The shirt was ironed
Incorrect The shirt was on ironed
From www.sissa.it/cns/lcd/calacei/calacei/Site/CA
LACEI-2006_files/10.Friederici.CALACEI.May06.ppt
78
Sentence processing ERPs in PD

• Friederici et al. (2003)
• N400
• ELAN
• P600
• Suggests basal ganglia contribute to late
  syntactic integration processes
• May explain sentence comprehension deficits in PD

?
79
Lexical-semantic activation

• Evidence to suggest that significantly slowed
  lexical-semantic activation may only be evident
  in those PD patients with comprehension
  difficulties (Angwin et al., 2005 Grossman et
  al., 2002)
• Slowed lexical-semantic access may contribute to
  sentence processing difficulties in PD

• Dopamine dependent frontal-striatal system may
  influence the speed of lexical-semantic
  activation
• Further evidence to support this theory
• Slower lexical-semantic activation when patients
  tested off dopaminergic medication (Angwin et
  al., 2005)
• Poorer sentence comprehension when off medication
  (Grossman et al., 2001)

80
Summary

• Dopamine depletion in striatum and associated
  disruption of frontal-striatal pathways
• Reduced signal-to-noise ratio of information
  processing ? subsequent difficulties suppressing
  weak semantic information
• Slowed lexical-semantic activation
• Disruption of syntactic integration processes
  associated with sentence comprehension

81
Schizophrenia

• Meet Leslie

82
Schizophrenia and thought disorder

• A subgroup of people with schizophrenia who also
  have positive symptoms
• Delusions
• Hallucinations
• Disorganised speech (e.g., frequent derailment or
  incoherence)
• Thus, disturbances of associational processes
  considered to be a primary factor in
  schizophrenia, i.e., they play a major role in
  the formation of schizophrenic symptomatology
• "In this malady the associations lose their
  continuity. Of the thousands of associative
  threads which guide our thinking, this disease
  seems to interrupt, quite haphazardly, sometimes
  such single threads, sometimes a whole group, and
  sometimes even large segments of them. In this
  way, thinking becomes illogical and often
  bizarre....." (Bleuler, 1911, p. 14)

83
The language of schizophrenia

• Samples of language from people with
  schizophrenia who are thought disordered
• "The next day when I'd be going out you know, I
  took control, like uh, I put bleach on my hair in
  California".
• "Why do people believe in God?" elicits a
  response like "Because make a twirl in life, my
  box is broken help me blue elephant. Isn't
  lettuce brave? I like electrons, hello."

Meet Ray
84
Locus of language deficits

• Increased spreading activation in semantic memory
• Dysfunctional hemispheric laterality
• Executive deficits and dysfunction of contextual
  processing

85
Direct and indirect priming in schizophrenia
86
Indirect priming
flag
sour
sugar
sweet
wine
87
Study 1

• Aims to examine how people with schizophrenia
  process semantic information across a defined
  time period and the relationship between semantic
  processing difficulties and Thought Disorder.

88
Participants

• 14 PwS (mean age 33 years SD 9.62)
• 11 male and 3 female
• Mean education 12.23 years (SD 2.30)
• Mean length of illness 8.21 years (SD 5.74)
• Mean NART 96.17 (SD 10.29)
• Symptom summary scores also calculated
• Reality Distortion Score, Psychomotor Poverty
  Score and Thought Disorder Summary Score
• 12 matched healthy controls

89
Stimuli

• Related Related target (RRT) soap plug BATH
• Related Unrelated target (RUT) soap gulf BATH
• Unrelated related target (URT) jail plug BATH
• Unrelated Unrelated target (UUT) jail gulf BATH
• Given at two SOAs (250 and 1000ms) and two RPs
  (.75 and .33)

90
The task

• A semantic priming task
• Participants made a lexical decision as to
  whether the final word was a real word or a
  nonsense word.

?
Prime
Prime
soap
Target
plug
BATH
500msec
250 1000msec
SOA
91
Results

• PwS who are also thought disordered hyperprime on
  related word pairs.
• Positive relationship between TDSS and priming
  effects for RRT

(Chenery et al., 2004)
92
Schizophrenia and nicotine

• Schizophrenia and nicotine usage are inextricably
  linked.
• PwS are markedly prone to smoke tobacco
  (prevalence rates estimated at 80-90 compared
  with 20-30 in the general community),
• schizophrenia is associated with higher daily
  cigarette consumption,
• the use of stronger cigarettes,
• the extraction of more nicotine from their
  cigarettes, and
• less success in quitting.

93
A possible explanation for nicotine use in sz

• Nicotine serves as a form of self medication that
  may
• reduce the side effects of antipsychotic
  medications,
• enhance the therapeutic effect of antipsychotics
  and so alleviate negative symptoms, and/or
• ameliorate a number of cognitive deficits
  associated with schizophrenia.
• But investigations of PwS can be confounded by
  many other variables.

94
The schizophrenia spectrum
Schizotypy in the healthy population
Schizophrenia
Relatives of PwS
Schizotypy Personality Disorder
95
Example of items used to rate schizotypy

• Unusual Experiences
• Are the sounds you hear in your daydreams
  usually clear and distinct?
• Have you occasionally felt as though your body
  did not exist?
• Introvertive Anhedonia
• Do you like mixing with people?
• Do you prefer watching television to going out
  with other people?
• Impulsive Nonconformity
• Do you at times have an urge to do something
  harmful or shocking?
• Have you ever felt the urge to injure yourself?

96
Questionnaire items related to Cognitive
Disorganisation

• I sometimes jump quickly from one topic to
  another when speaking.
• I have trouble keeping on the topic.
• My speech seems confused to others.
• A word I was about to say often gets lost when I
  go to speak.
• I find it difficult to understand if someone
  speaks in long sentences.
• Do you ever feel that your speech is difficult to
  understand because the words are all mixed up and
  dont make any sense?
• No matter how hard you try to concentrate, do
  unrelated thoughts always creep into your mind?

97
Aims.

• To investigate whether people with high
  schizotypy ratings (loaded for odd speech items)
  exhibit semantic processing abnormalities.
• Predictions Low schizotypy will show priming for
  strong associates but have reduced or absent
  priming for weak associates. People with high
  schizotypy ratings will show priming for the
  weakly associated word pairs, providing support
  for the hyperactivation hypothesis.
• Nicotine will normalize the high schizotypy
  priming patterns.

98
Stimuli

• 24 strongly related word pairs (e.g., chef food)
• 24 weakly related word pairs (e.g., couch relax)
• 16 unrelated word pairs (e.g., appear golden)
• 16 word nonword pairs (e.g., bible garshar)

99
The N400
The horse galloped in the custard.
100
The EEG Priming Trial Sequence
Prime
Target
sky
?
money
500msec
SOA 1000msec
1000msec
1500msec
101
(No Transcript)
102
High schizotypy with nicotine
103
Low schizotypy with nicotine
Unrel Strong Weak
104
Discussion

• Low schizotypy group
• Our data show a priming effect only for the
  strongly related prime target pairs.
• High schizotypy group
• Priming for both strongly and weakly associated
  words but a nicotine patch normalizes the
  priming.

105
Implications for future management

• If nicotine "normalises" performance, this result
  will have marked consequences for anti-smoking
  and quit campaigns in schizophrenia
• Improved knowledge of the role of
  neurotransmitters such as dopamine in cognitive
  performance might eventually lead to safe and
  effective nicotine agonist therapy thus giving
  people with schizophrenia the cognitive enhancing
  effects of nicotine without the significant
  carcinogenic risk.

106
Dopaminergic modulation of information processing

• Dopamine exerts a neuromodulatory influence on
  prefrontal cortex via meso-cortical pathways
• Prefrontal cortex involved in maintaining
  updating contextual information crucial for
  cognitive function and language processing

From http//mywebpages.comcast.net/epollak/PSY255_
pix/DA_paths.JPG
107
Cohen et al. (1999) Braver et al. (1999)

• Dopaminergic modulation of prefrontal cortex
• Phasic dopamine updating of context
• Tonic dopamine maintenance of context
• Computational modeling increased noise in
  mesocortical dopamine system leads to
• Increased tonic activity ? deficits in context
  maintenance
• Decreased phasic activity ? deficits in updating
  of context

108
Dementia

• Meet Claudia

109
Alzheimers Disease - neuropathology

• hallmark changes associated with the disease are
• neurofibrillary tangles
• senile plaques
• cerebral atrophy
• ventricular dilation

110
(No Transcript)
111
Senile plaques - an extracellular abnormality
112
The cerebral atrophy is clearly shown at autopsy
113
A cross-sectional view
114
Dementia of the Alzheimers type

• He thinks that he is the only impatient man in
  this House to get things done. I will beat him
  50 any day he likes. No doubt he has a hawk-like
  desire for action, without bridle and without
  saddle, across the Atlantic.

from Mahendra, 1987
115
Diary entries
116
Diary entries
To speak now is a great effort, and in the
ground to be covered and the development of the
argument, I get more and more confused.
from Mahendra, 1987
117
(No Transcript)
118
Remember these words
119
Individual task Write down the words just
presented.
120
Semantic impairment

• What is the nature of the semantic impairment in
  DAT?
• Is the deficit one of impaired access to semantic
  knowledge or actual loss of information?
• What implications does this have for implicit
  versus explicit memory?

121
Loss of semantic knowledge
ANIMAL
DOG
Direction of semantic degradation
TAIL
PAWS
FUR
122
Semantic priming in dementia (Giffard et al.,
2001)
Word pair presented during priming
LION
TIGER
Wild Animal
Wild Animal
?
?
STRIPES
MANE
123
We know that

• Semantic strategies can assist in memory
  processing
• Implicit memory is better preserved than explicit
  memory
• Word meanings follow a known pattern of erosion
  in dementia.
• Can we use this knowledge to help PwD make the
  transition from home to a RCF less traumatic both
  for them and their caregivers?

124
Implications for future management

• Begin the process of transition to a RCF while
  still at home using structured semantic therapy
  to assist the laying down of implicit memories of
  the new environment.

125
Conversations
http//www.abc.net.au/4corners/stories/s495625.htm
126
Group Task Whats wrong with the conversation?
127
Conversational Analysis

• Why is it valuable to the assessment of
  disordered interaction?
• Because it uses everyday conversations as data
• because it focuses on conversation as a
  collaborative achievement
• because it emphasises the social role of language
• because it is a client-driven approach
• because it is a minimally obtrusive tool

128
Features of natural conversations

• In conversation, many unspoken rules exist that
  create order
• The aim of CA is to ascertain and describe the
  nature and characteristics of the rules/features
  that exist.
• We can let our conversational partner know if a
  mis-understanding has taken place
• We use trouble-indicating behaviours to highlight
  a point of trouble in understanding the message

129
Features of natural conversations

• Now the trouble source can be altered or
  clarified
• We use repair types to remediate the trouble
  source

130
Definitions of terms

• Trouble indicating behaviours - behaviours that a
  listener can use to highlight a point of trouble
  in understanding the message the speaker is
  intending (e.g., the listener can request a
  repetition or can take a best guess at what was
  meant by the speaker)
• Repair types - a strategy that aims to alter or
  clarify the trouble source (e.g., paraphasing)

131
Conversational Analysis in DAT
Individual Task Circle the trouble in this
conversation
132
WC17 Do you have any kids? DAT18 Huh? WC19 Do
you have any children? DAT20 Children, well my
my fam-family were a heavy family. three three
or four. WC21 Oh, you mean a large family.
DAT22 Yeah, well we might have one of
three. WC23 Have you got one or three
children? DAT24 Um, I I didnt have any one or
three for myself They were my ah mother,
father, sister /w/ when I went to school. WC25
Did you have your own children? DAT26 Ah no I
used to go to school right (unint) about that
size. WC27 What would be that size? DAT28
Cant go to school in the morning cause all
depends on where you are. WC29 Did you grow up
in the country or in the city?
133
Data analysis

• Trouble Indicating Behaviours (TIBs)
• 12 categories in total which included
• 1. Requests for repetition
• 2. Requests for specific information
• 3. Corrections
• 4. Lack of uptake/lack of continuation
• 5. Hypothesis formation
• 6. Metalinguistic comment
• 7. Reprise/minimal dysfluency

134
Frequency of TIBs
Proportion
135
Data analysis

• Repair types
• 7 in total including
• 1. Repetition
• 2. Revision/reformulation
• 3. Addition/specification
• 4. Inappropriate/withdrawal

136
Frequency and effectiveness of repair types.



137
Summary of Results

• Hypothesis formation and Request for specific
  information were used most often by normal
  speakers to indicate trouble.
• Both these TIBs were highly interactive, meaning
  that the well partner was better placed to take a
  greater burden when signalling and repairing
  trouble.

138
Summary of Results

• The people with DAT used non-interactive ways of
  signalling trouble.
• During these breakdowns, the normal partner most
  often repaired the trouble, although some
  misunderstandings were left unresolved.

139
Treatment options

• Naming therapy
• Can improve naming but only for treated items
• Principles of spaced-retrieval and errorless
  learning may facilitate learning of new
  information
• Caregiver training in suitable communication
  strategies is crucial
• Family/home-based caregivers
• Staff in residential care facilities