Speech production

1
Speech production

• What is anomia?
• Types of anomia
• Cognitive models
• Category specific naming deficits?
• Summary

2
Speech production

• We have all experienced having the idea of some
  familiar object but being unable to call up its
  name.
• The moment some other person uses the word it is
  recognised.
• The "Tip of the tongue" phenomenon illustrates
  the normality of pathology.

3
Word form problems (selection and production)

• (1) Tip of the Tongue effects are accompanied by
  a feeling of knowing some knowledge of sound
  structure is preserved and phonemic cueing (i.e.
  starts with) can help word retrieval.
• (2) Malapropisms describe an error that is a real
  word sounding like the target
• Thats a colourful flower derangement
• (3) Neologisms are nonsense words that are
  produced in place of the target
• She cannot see without her nexicles.

4
Anomia

• Patient BP (Weekes Robinson, 1997)
• guitar -gt cello swan -gt duck onion -gt vegetable
  
• An inability to translate ideas into symbols.
• An extreme form of tip-of-the-tongue.
• Types of anomia
• Semantic
• Phonological
• Optic?

5
semantic
phonemic
dog
house
6
A four legged animal with a long mane and a tail
dog
horse
7
HAT
8
Brocas area
Brocas area
Motor cortex
Perisylvian region
Prefrontal cortex
Wernickes area
9
Frontal lobe
Motor strip
Transcranial Magentic Stimulation TMS
Temporal lobe
10
Ashcroft

• "The most powerful realization I had during the
  episode ... was a dissociation between the
  thought and the word or phrase that expresses the
  thought. The subjective experience consisted of
  knowing with complete certainty the idea or
  concept that I was trying to express and being
  completely unable to find and utter the word that
  expressed the idea or concept. ... The experience
  was not one of being unable to articulate a word
  currently held in consciousness. Instead, it was
  one of being fully aware of the target idea yet
  totally unable to accomplish what normally feels
  like a single act of finding-and-saying the word."

11
Symptom or syndrome?

• As a symptom or neurological sign.
• Predominant symptom is word finding difficulty
  and this is found in a majority of patients with
  damage to the brain.
• Confrontation naming is selectively impaired
  whereas other aspects of language and memory can
  be intact.

12
Syndrome

• Difficulties with word finding in spontaneous
  speech and confrontation naming tasks.
• Seen in most aphasic patients and in conditions
  ranging from dementia to Parkinson's disease.
• Associated with left hemisphere damage in the
  temporal lobe (perisylvian region).

13
Optic aphasia

• Patient JF was impaired when naming visually
  presented objects but could define them from
  their spoken names or when given the object to
  touch (Beauvois et al., 1973, 1982, 1985).
• He could mime the use of the objects so this was
  not agnosia (i.e. knew what object was).
• C.f. Tactile aphasia e.g., RG could not name
  objects by touch but could name visually
  presented objects and pantomime their use
  (suggesting that he knew what the object was).

14
Optic aphasia

• Beauvois, et al, (1978) argued for disconnection
  between a separate modality specific action
  memory system and a verbal memory system.
• Dissociable knowledge systems one for actions
  one for objects (also Warrington Shallice, 88).
• Note that an alternative explanation given by
  Riddoch, Humphreys, Coltheart Funnell, 88).

15
Frontal?
Temporal?
Patient RG
Patient JF
Ellis and Young Chapter 2 page 57
16
Semantic anomia

• Naming problems with poor comprehension of object
  names from speech or from print.
• Semantic errors in naming saying "daffodil" for
  tulip (semantic paraphasia) and comprehension
  pointing to a daffodil after hearing the word
  tulip.
• Performance not improved by phonemic cueing i.e.
  the word begins with t.
• There is a general problem with knowing the
  meaning of a word from spoken or written input.

17
Patient JCU

• Patient JCU (Howard Orchard-Lisle, 1984) is a
  severe 'global' aphasic who was extremely poor at
  picture naming (only 3 correct).
• She could be induced to make semantic errors by
  giving her the first sound of an associate e.g.,
  picture of a tiger "l" -gt "lion".
• When shown a picture and asked "Is it an X?
  (e.g., picture of a tiger and "Is it a lion?")
  she would accepted 56 of closely related words
  as correct, but only 2 of unrelated names.

18
Phonological/lexical anomia

• Naming problems in the context of very good
  comprehension of object names (e.g. BP).
• Semantic errors in naming only
• Naming saying cap" for hat (paraphasia).
• Comprehension will point to a picture of a hat
  after hearing/reading the word hat.
• Phonological errors may also occur as patient
  searches for the correct name (e.g., nexicles).
• Patient immediately recognises the target name
  upon hearing the tester produce it.

19
Patient EST

• Patient EST (Kay Ellis, 1987) could sort
  pictures into semantic categories and match
  pictures with spoken words (comprehension OK).
• When given pictures to name he sometimes gave a
  similar sounding (incorrect) response.
• Performance was improved by phonemic cueing and
  he knew the sound structure of word.
• General problem in selecting or producing the
  appropriate sound.

20
A cognitive framework for naming
articulation of phonemes
horse
21
Points of breakdown

• 1. Object perception / identification Optic.
• 2. Object 'comprehension' Semantic anomia.
• 3. Word retrieval Phonological/lexical anomia.
• 4. Articulation of phonemes cause phonological
  errors e.g., tulip -gt "dulop . . .tupil
  dyspraxia or dysarthria.

22
Exclusionary testing for anomia

• Confrontation naming tasks
• e.g., Boston naming test Graded naming test
  (McKenna Warrington, 1983).
• Semantic knowledge
• e.g., Spoken word-picture matching (PALPA)
  Pyramids and Palm Trees Test (Howard and
  Patterson, 1996).
• Verbal fluency (FAS test) semantic category
  fluency.
• Effects of psycholinguistic variables
• word frequency (spoken and written)
• age-of-acquisition
• imageability
• number of letters/phonemes/syllables.

23
A cognitive framework for naming
dog
articulation of phonemes
24
Agnosia?

• It is important to consider the possibility of
  pre-semantic (visual) deficits when assessing a
  patients naming difficulties using pictures.
• If the deficit is pre-semantic then patient
  should be able to perform well when tested using
  other modalities (written words, spoken words,
  etc.).
• The patient should also have normal speech
  production other than for naming objects e.g.,
  normal spontaneous speech naming to definition
  verbal and category fluency.

25
A cognitive framework for naming
A four legged animal with a long mane and a tail
dog
articulation of phonemes
26
Tell me everything that is happening in this
picture
27
hat
28
Semantic impairment

• If the semantic system is impaired then both
  comprehension and production of single words
  should be compromised.
• Examples of patients with this form of anomia
• Category-specific patients (e.g., JBR Warrington
  Shallice, 1984).
• Some cerebro-vascular accident (CVA) patients
  (Chertkow, 1997 Howard Orchard Lisle, 1984).
• Patients with dementia (Parris Weekes, 2001).

29
Thompson-Schill et al (1999) Neuropsychologia

• Living Nonliving Example cases
• Animate Inanimate
• Animal Fruit Artefacts
• x x v (Warrington Shallice,1984).
• v v x (Sheridan Humphreys, 1993).
• x v v (Hart Gordon, 1992).
• v x x (Hillis Caramazza, 1991).
• v x v (Hart, Berndt, Caramazza 1985)
• x v x Not reported

30
Picture-picture matching task
31
Severe damage to (verbal) semantic system

• Gianotti, et al (1986) reported patients who had
  no knowledge of the meaning of a word.
• In speech production, errors were semantic in
  nature e.g., calling a hat a cap.
• Warrington Shallice (1984) patient JBR had
  problems comprehending pictures and names of
  living things and problems producing names
  spontaneouslycategory specificity.

32
Milder damage to (verbal) semantic system

• Patient JCU had a deficit in comprehension of
  spoken words but some semantic information was
  clearly preserved because she could often provide
  the correct answer ("tiger to a picture of a
  tiger) if given a phonemic cue (/t/).
• Even if given a wrong phonemic cue (e.g., the
  sound /l/ to a picture of a tiger) she produced a
  semantically plausible answer (e.g., lion) rather
  than an inappropriate answer (e.g., lemon)
  indicating some semantic processing.

33
A cognitive framework for naming
A four legged animal with a long mane and a tail
dog
articulation of phonemes
34
Phonological lexicon impairment

• Comprehension should be normal.
• If make semantic errors but should be able to
  reject them on a forced choice task.
• Phonemic cueing should improve naming.
• Phoneme level should be normal so should see good
  repetition and reading aloud.
• Example cases Patient EE (Howard, 1995)
  Patients JS and GM (Lambon Ralph et al., 2000)
  Patient MOS (Lambon Ralph, 1999).

35
Selection or production?

• Ellis and Young (1996) do not distinguish between
  selection and production but other authors do
  (McCarthy Warrington, 1990).
• They distinguish between deficits in lexical
  selection and deficits in lexical production.
• Best evidence for two separate stages is AF,
  (Hier Mohr 1977) who could write words he
  couldnt name therefore selection of the word is
  preserved but spoken production is not.

36
Category specific anomia

• Nouns versus verbs
• Naming actions (verbs) intact (Robinson et al
  1999)
• Naming objects (nouns) intact (Cappa et al 1998).
• Caramazza (1991) has argued that nouns and verbs
  are represented as discrete forms in output
  systems.
• Proper names versus common names
• Names of faces impaired but names of objects
  intact.
• Geographical terms and proper names are
  selectively preserved in some patients (McNeil et
  al, 1994).
• Why? The names are constrained and refer to a
  unit of meaning whereas common nouns refer to
  types.

37
A cognitive framework for naming
verbs
nouns
articulation of phonemes
38
Dissociation between nouns and verbs

• Selective deficits in producing verbs relative to
  nouns in speech are associated with
  left-hemisphere pre-frontal lesions (BA 45 again)
  whereas difficulty in producing nouns relative to
  verbs is associated with left temporal damage.
• Issue Do these impairments arise because of
  differences in the way grammatical categories of
  words are organized in the brain or because of
  differences in the neural representation of
  actions and objects (i.e. semantic properties)?

39
Syntax or semantics?

• Caramazza argues that lexical-grammatical
  knowledge has a neuroanatomical and also a
  functional basis that is separate from the
  semantic features of a words representation.
• Morphological transformations of verbs (e.g., he
  sings -gt they sing) and nouns (e.g. the songs -gt
  the song) cannot result from damage to semantics
  as they are found in the production of
  pseudo-words that function as verbs (e.g. he
  wugs-gtthey wug) or nouns (the wugs-gtthe wug).

40
Parris and Weekes (2001)

• Report an anomic patient RS with dementia
  affecting his access to knowledge about objects
  from visual, verbal and tactile input.
• RS displays preserved knowledge and naming of
  actions but an effect of instrumentality on
  action naming i.e. actions that depict an actor
  using a tool are worse than actions depicting an
  actor performing an action without a tool.
• Shows that the relationship between actions and
  objects affects action naming in dementia.

41
dog
Instrumental x
With objects v
Non-instrumental v
42
A cognitive framework for naming
Action Knowledge v
Object Knowledge x
Verbs v
Nouns x
articulation of phonemes
43
verbs
nouns
tools
44
Dynamic aphasia

• Patient ANG had a malignant meningioma in BA45.
• Patient ANG has normal repetition, naming, and
  reading but spontaneous speech is impaired.
• BA45 is involved in the selection of competing
  alternatives for response output (Shallice).
• Dogs have a good sense of _______v
• Helen reached down to dust the _______x
• Haylings Test
• Robinson et al (1998) argued ANG had damage to a
  module for selection of competing verbal
  responses.

45
PET study of normal word generation

• Wise et al, (1991) reported an early PET study
• 1. Rest.
• 2. Perception of non words.
• 3. Noun/noun semantic category comparisons
  (e.g., fruit -apple..furniture-shirt).
• 4. Verb/noun semantic category comparisons
  (e.g., eat apple.knit-spectacles).
• 5. Verb generation similar to verbal fluency
  test Ss presented with a noun (e.g. flower) and
  had to generate as many associated verbs as
  possible.

46
Subtractive methodology

• 2 minus 1 prelexical phonetic processing.
• 3 minus 2 post phonetic semantic processing of
  nouns.
• 4 minus 2 post phonetic semantic processing of
  verbs.
• 5 minus 4 regions involved in the retrieval of
  verbs from memory.
• The subjects were silent in all conditions in
  order to avoid activation due to vocalisation.

47
Results

• In the verb generation condition there was
  activity in the posterior left inferior frontal
  gyrus (Brocas area), posterior left middle
  frontal gyrus and the supplementary motor area
  (SMA).
• If left SMA is lesioned, speech output is lost
  though comprehension is preserved in patients.
• The posterior left middle frontal gyrus (BA 45)
  is associated with dynamic aphasia (c.f. Robinson
  et al 1998) and is also involved when several
  verbs must be generated for each noun rather like
  the requirements for spontaneous speech.

48
Brain and behaviour

• These results show that in the normal brain,
  there is a close correspondence between the brain
  regions involved in the deliberate, voluntary
  generation of words, and the brain regions
  involved in Brocas aphasia and transcortical
  motor (dynamic) aphasia.
• The most likely organisation of the lexical
  network is an interactive arrangement of brain
  regions serving spoken word production.
• This would allow for a modular pattern.

49
Summary

• Anomia is referred to as a neurological sign in
  neuropsychology because this deficit is often
  observed in patients with brain damage.
• Studies of anomia have enabled us to develop
  cognitive models of naming and to identify the
  necessary cognitive processes.
• Studies of category specific anomia have
  contributed to the identification of brain
  regions used for noun and verb processing using
  neuroimaging techniques.

50
Primary reading sources

• Parkin Chapter 7.
• Chapter 5 (2 and 9) - Ellis, A. Young, A.
  (1996) Human cognitive neuropsychology. Hove
  Psychology Press.
• McCarthy, R. Warrington, E.K. (1990) Cognitive
  Neuropsychology A Clinical Introduction. London
  Academic Press. Chapter 6 "Auditory Word
  Comprehension Chapter 7 "Word Retrieval".

51
References

• Cappa S et al (1998). Object and action naming in
  Alzheimer's disease and fronto-temporal dementia.
  Neurology, 50(2) 351-355.
• Caramazza A Hillis A. (1991). Lexical
  organization of nouns and verbs in the brain.
  Nature (346), 269.
• Lambon Ralph M. et al (2000). Classical anomia A
  neuropsychological perspective on speech
  production. Neuropsychologia, (38), 186-202.
• Parris, B Weekes, B. (in press). Action naming
  in dementia. Neurocase.

52
Patient BP
53

• RED

• RED

54
A cognitive framework for naming
lemma
Written word output lexicon
dog
articulation of phonemes
HORSE
55
Levelts model of naming
Language based
Not language based
ANG
ToM, Actions, Knowledge of world WM, Discourse
model
Pre-verbal message
Conceptual preparation
Grammatical encoding
lemmas mental lexicon
Surface structure
Morpho-phonological encoding
syllabary
phonetic encoding
articulation
overt speech
56
verbs
nouns
tools
57
Phoneme level

• Comprehension should be normal.
• Errors are mainly phonological but individual
  phonemes can be pronounced correctly.
• Phonological errors similar to spoonerisms
• He received a blushing crow
• Phoneme level is used for repetition, naming,
  speech production and reading aloud -gt so should
  be impaired for all language tasks.
• Example cases RD (Ellis, et al., 1983) LT
  (Shallice, Rumiati, Zadini, 2000).

58
Post phoneme level problems

• Pure articulation problem (speech musculature).
• Comprehension, speech output lexicon and phoneme
  level intact but cannot programme speech into an
  articulate output (dyspraxic).
• Affects all forms of speech output.
• Errors halting speech and poorly articulated
  phonemes (for all phonemes and word types).
• Specific to language apraxia for speech.
• General oral apraxia.

59
(No Transcript)
60
(No Transcript)
61
verbs
Nouns
62
Phenomenology

• Aschcroft - an academic interested in aphasia.
• A personal case history of transient anomia
  Brain and Language, 1991, 44, 47-57.
• 45-minute anomic episode.
• Caused by disrupted boodflow (later treated by
  operation).

63
Category specific agnosia

• Warrington and Shallice (1984) reported a patient
  called JBR who following an acute lesion to the
  left temporal lobe (as a result of herpes
  encephalitis) had a selective deficit when asked
  to name pictures from just one semantic category
  living things.

64
Artefacts

• By contrast JBR was able to name non-living
  objects very well including those with low
  frequency names such as accordion that were
  matched for the number of letters in the name and
  the visual complexity of the object.

65
Double dissociation

• Patient YOT (Warrington McCarthy, 1987) was
  worse at naming nonliving things but was very
  good at naming large, nonmanipulable objects.
• Suggests that the organisation of semantic memory
  might break down into taxonomic categories or
  reflect different properties of living
  (visual-sensory) and nonliving (functional)
  items.
• LA (Gainotti Silveri, 1996) was worse at living
  than nonliving things and was worse at giving
  visual information compared with functional
  information about both living and nonliving
  things.

66
Categories of knowledge about objects?

• Multiple semantic systems (Warrington/Shallice)
• Patients who have no access to object knowledge
  from verbal input but can name pictures
  successfully.
• Modality specific (visual and verbal) meaning
  systems.
• Verbal semantic and visual semantic systems?
• Unitary semantic system (Caramazza/Coltheart)
• Different types of object depend on different
  types of encoding e.g., living things require
  featural encoding.
• Knowledge only appears to fractionate because of
  differences in task difficulty (verbal visual
  tasks).

67
Confounding factor accounts

• Confounding factors for verbal/visual knowledge
• Pictures afford more information than words.
• Agnosia causes apperceptive visual problems.
• Modality of input impairments cannot be excluded.
• Confounding factors for living/nonliving things
• Visual familiarity, similarity, complexity
  (Funnell, 2000)
• Animals are more structurally similar to each
  other than artefacts (Humphreys,et al, 1995
  Funnell, 2000).
• Monkeys discriminate nonliving things better than
  living things (Gaffan Heywood, 1993).

68
Differential-weighting hypothesis

• Category specific effects on recognition result
  from a correlated factor such as the ratio of
  visual versus functional features of an object
• living more visual and nonliving more functional.
• Farah McClelland (1991) report a dictionary
  study showing the ratio of visual to functional
  features for living things and nonliving things
• living things was 7.71 and nonliving was 1.41.

69
Farah and McClelland (1991)

• A single system with functional and visual
  features.
• Model was trained to associate functional and
  visual features differently for living and
  non-living things.
• Two different inputs (verbal and visual).
• Model lesioned to either visual or functional
  units
• A) Visual unitsgtliving things Functional
  unitsgtnonliving things.
• B) Visual unitsgtimpaired functional knowledge of
  living things.
• Loss of both visual and functional knowledge
  about a living thing can occur in a unitary
  single system that only distinguishes between
  visual and functional attributes.

70
Thompson-Schill et al., (1999)

• Interactive modality hypothesis predicts that
  questions about visual and functional semantic
  knowledge for living things will engage visual
  semantic processing.
• Questions about the visual semantic attributes of
  nonliving things will also engage visual semantic
  processing.
• Scaned the fusiform gyrus while asking questions
  about visual and functional properties of living
  and nonliving objects.

71
Thompson-Schill et al., (1999)

• There was increased activity in the ventral
  (What) pathway - middle temporal gyrus - for all
  conditions suggesting a common system.
• But questions about the visual semantic
  attributes of living and nonliving things engaged
  the fusiform gyrus differentially an
  interaction.
• Visual semantic retrieval may depend on both
  modality of input and the category of retrieval.
• See Caramazza (2000) for a reply.

72
Funnell (2000)

• Reported a patient with dementia (NA) who showed
  a category specific naming deficit
• not due to any of the confounding factors (visual
  complexity, similarity, familiarity).
• not due to semantic problems.
• resulted from an early visual problem possibly
  due to the access of stored structural
  descriptions.
• conclusion is that category specific deficits can
  be a consequence of visual impairments in agnosia
  and do not necessarily reflect categories of
  knowledge.