ling411-08 - Rice University

Download Report

Transcript ling411-08 - Rice University

Ling 411 – 08
Disorders of Syntax and Morphology
Goodglass 1993: Chapter 6
Major Language Areas
Supramarginal gyrus
(Goldstein’s area) Angular gyrus
Wernicke’s area
(Geschwinds’s area)
Exner’s
area
Broca’s
area
Agrammatism
 Usually associated with Broca’s aphasia
 Generally present in Broca’s aphasia
 But other aphasics also have grammatical
dysfunctions
• Paragrammatism – common in Wernicke aphasia
 A lot of variation among different patients
Agrammatism vs. Paragrammatism
 Paragrammatism – too much speech
• Normal or excessive fluency
• Use of inappropriate words
• Neologisms
• No lack of function words and inflections
•
 But not always used appropriately
Common in Wernicke’s aphasia
 Agrammatism – not enough speech
• Lack of fluency
• Omission (NOT deletion!) of function words and
•
inflections
Common in Broca’s aphasia
Omission vs. Deletion
 Goodglass (106):
• Sentences with a deleted main verb (“Joan




and I . . . Coffee”) may continue to appear.
• . . . misuse or deletions of morphology . . .
Is he talking about deletion or omission?
Deletion implies that it was first there, and
then removed
Omission – it wasn’t put in at all
Goodglass is following a practice that was
common among linguists at the time he wrote
the book
Broca’s Aphasia
 Damage to frontal lobe
• Mainly, inferior frontal gyrus
 Largely intact comprehension
 Nonfluent, agrammatic speech
 “Telegraphic speech” –
• Abundance of content words (e.g., nouns)
• Lack of function words (e.g. prepositions)
 Impaired verb processing
• Bates, Chen, Tzeng, Li & Opie, 1991; Damasio &
Tranel, 1993; Daniele, Giustolisi, Silveri,
Colosimo & Gainotti, 1994; Lamb & Zhang, 2010;
Shapiro & Caramazza, 2003
Verbal short-term memory deficit
(in Broca aphasia)
 Patients can readily point to individual
objects or body parts named by the
examiner
 But when asked to point to the same items
in a specific sequence they often fail at
the level of only two or three items
Benson & Ardila 124
How to explain?
Subtypes of Broca aphasia
 Type I
• A.k.a. little Broca aphasia
• Milder defects
• Less extensive damage
• Better prognosis
 Type II
• Symptoms worse
• More extensive damage
 These are not distinct, but variations
• Two spans along a scale
Agrammatism: an early observation (1819)
 Deleuze (1819), referring to a Frenchspeaking patient: The patient “used
exclusively the infinitive of verbs and
used no pronouns. … She produced
absolutely no conjugated verb.”
Goodglass 1993: 104
Example of agrammatic speech
Examiner: Can you tell me about why you came back
to the hospital?
Patient: yes … eh … Monday … eh … dad … Peter
Hogan and dad … hospital.
Er … two … er … doctors … and … er … thirty
minutes … and ... er … yes … hospital. And … er …
Wednesday … Wednesday. Nine o’clock. And … er …
Thursday, ten o’clock … doctors … two … two …
doctors… and … er … teeth … fine.
E: Not exactly your teeth … your gP: Gum … gum …
E: What did they do to them?
P: And er … doctor and girl … and er … and er gum …
Goodglass 1993: 107
Some features of agrammatism
 Telegraphic speech
• Short utterances
• Omission of grammatical functors
 Relative abundance of substantives
 Verbs are uncommon, rare in some patients
• When present, uninflected or –ing form
 For French aphasics, infinitive form
 Use of word order is generally spared
 Comprehension is impaired for complex
sentences
Problems in the study of agrammatism
 Must be distinguished from paragrammatism
 Grammatical aberrations – even among Broca
aphasics – vary from patient to patient
 Linguistics has not (yet) provided clear
answers to important basic questions:
• What normal grammatical functions are
• How they operate
Syntax
 First, we need to dispel the notion that
syntax is one capacity, that can be lost (or
spared) as a unit
 Syntax can be understood as a set of
constructions
• Learned by children (and others) one by one
•
 Like vocabulary
Some can be lost, others spared, in aphasia
 It is a label of the grammarian for multiple
things
 Word order is often spared in Broca’s
aphasia while a lot of syntax is lost
Stability of word order in agrammatism
 Agrammatic patients can usually handle
word order in both production and
comprehension
 Evidence (comprehension)
• passive sentences misconstrued
• The horse was kicked by the dog
•
 Broca’s aphasic: horse as kicker
 Passive marker not apprehended
Canonical word order guides the interpretation
 Possibly aided by conceptual knowledge
Reading and writing in agrammatism
 Agrammatic difficulties are also seen in
• Oral reading
• Writing to dictation
• Repetition
 But:
• Some patients are agrammatic in speech but not
in writing (Goodglass 1993: 110)
• Some can repeat correctly
• How to explain?
 Menn & Obler (1990) describe some patients who
are less agrammatic in oral reading than in
spontaneous speech (Goodglass 1993:111)
Variation among agrammatics
(Goodglass 1993:107)
 Syntax and morphology (study of agrammatic
French aphasics)
•
•
Some patients have fairly good syntax but defective
morphology
Some patients have fairly good morphology but
defective syntax
Both types of patients fail to use inflected verb forms
•
•
Some patients use –s plural marker but not articles
Other patients use articles but not –s plural marker
•
 Gleason et al. observations (1975)
Loss of the use of relational markers
in receptive agrammatism (118)
 E.g. father’s sister
• Ex: Is “my father’s sister” a man or a woman?
• Patient answers randomly
 Unable to grasp the relational function of –’s
 Command given in testing:
• Ex: Touch the comb with the pencil
• Patient may touch the pencil with the comb
 Perhaps picks up comb because the word
comb comes first in the instruction
 Locative relations somewhat less fragile
• in back of/in front of, over/under,
before/after
Linguistic structure in the cortex:
What we learn from agrammatism
 Agrammatism is generally associated with
Broca’s aphasia
 Therefore, the grammatical skills lost in
Broca’s aphasia must be supported at least in
part by either
• Broca’s area, or
• Area(s) adjacent to Broca’s area
 In other words: There must be something in
or near Broca’s area that is essential for
correct grammatical production
• And grammatical comprehension –
 Receptive agrammatism
Receptive processing in Broca’s aphasia?
 Problem:
• Broca’s area is in frontal lobe
• Frontal lobe is supposed to be for motor
•
•
•
production
 Motor production is top-down processing
Receptive functions involve bottom-up
processing
Comprehension involves sensory processing
 in frontal lobe?
Bottom-up (receptive) processing in frontal
lobe?
Receptive agrammatism in Broca’s aphasia
Two avenues to explanation
1. The role of short-term memory, and
Broca’s area in short-term memory
2. Maybe the frontal lobe can have receptive
function
•
To explore this possibility we must first
examine the phenomenon of imagery
For perspective,
A related problem: Imagery
 Types of sensory imagery
• Visual
• Auditory
• Somatosensory
 Cf. Motor imagery
Visual Imagery
 Visual images of people, buildings, etc.
• What is a visual image?
 What does it consist of?
• Is it a little picture?
•
» If so, where are the eyes to see it?
» What is it drawn on?
» Where is the visual perception system to
interpret it?
If not, what?
Auditory Imagery
 Auditory images of words, music, etc.
• We can hear things in our heads
• What is an auditory image?
 What does it consist of?
• Sound?
» There is no air inside the head to vibrate
 What hears it?
• Little ears inside the head?
How Imagery Operates
 It’s unlikely that visual imagery uses some
mechanism independent of that for vision
 Therefore, it must use (some of) the same
neural connections used in perception
• For visual imagery, pathways in the occipital lobe
• For auditory imagery, pathways in the temporal lobe
• For tactile imagery, pathways in the parietal lobe
 Imagery is activation of some of the same
neural pathways that get activated upon
receiving input from sense organs
Anatomical consequences
 Consequences of imaging explanation
• Top-down processing in perceptual areas
• Perceptual pathways must have parallel
pathways of opposite direction
 Why are imagined scenes less vivid than
those resulting from input to the eyes?
Bidirectional Processing
 Imagery requires top-down processing
• Using pathways that typically operate bottom-up
 Therefore, perceptual pathways must
generally be bidirectional
 Anatomical evidence supports the hypothesis
• Reciprocal pairs of cortico-cortical axons
Bidirectional Connections
 Most corticocortical connections are
bidirectional
 An established finding from neuroanatomy
 It’s not because the connecting nerve
fibers (axons) are themselves bidirectional
 It’s because we find different but roughly
parallel fibers going in opposite directions
Bidirectional Processing in Frontal Lobe?
 Frontal lobe processing: typically top-down
 But there is a large amount of uniformity in
cortical structure
 Hypothesis: Bottom-up processing also in
frontal lobe
• From perceptual (i.e. posterior) areas to
•
locations in frontal lobe
We already have seen evidence: the arcuate
fasciculus
Bidirectional connections in frontal lobe
 Would explain how Broca’s area is
involved in receptive grammatical
processing
 Would account for the finding that
interpretation of prepositions and
verbs is a frontal lobe function
• Finding from the study of agrammatism
Attempts to explain agrammatism
 Many theories have been proposed
• Cf. Goodglass 1993:111ff
 Some intriguing ideas
• Loss of relational use of words (Jakobson, Luria)
•
•
 Difficulty with markers of such relationships
Impairment of inner speech (Luria)
 Hence, impairment of auditory working
memory
Difficulty with unstressed words (Goodglass,
Kean)
 Substantive words are commonly stressed
 Functors are generally unstressed
Caution in interpreting
 Agrammatism may not be just one
phenomenon
• Syntax is not one structure but several
• All agrammatics and probably all Broca’s
•
aphasics are deficient in use of verbs
Other phenomena of agrammatism show
more variability
 The problem (or part of the problem)
may not be grammar as such:
• Short-term memory – the inner speech loop
• Phonology: stressed vs. unstressed words
Phonological factors
 Function words are (in general) unstressed
 Maybe the difficulty is in production of
unstressed words
 Intriguing finding of Goodglass et al.
• Function words
•
•
 May be produced after a stressed word
 But almost never produced initially
Production starts with stressed word
Even with repetition
 Open the door > Open the door
 The door is open > Door is open
More evidence on relational markers
(119f)
 Grammatical particles that do not mark
relations are exempt from omission
• and
• Japanese clause-final particles
 Emphatic yo
 Question marker ka
 Confirmation-seeking particle ne
 Verbs always have a syntactic implication
• I.e. relationship to one or more nouns
 Menn & Obler: Impairment affecting
grammatical elements that mark
relationships within the sentence
Nouns and Verbs:
Back Brain & Front Brain (?)
 “A Neurolinguistic Universal” –E. Bates
• Verb deficit in Broca’s aphasia
• Noun deficit in Wernicke’s aphasia
 Suggests that
• Verbs are represented in frontal lobe
• Nouns are represented in or near temporal
lobe – angular gyrus and/or supramarginal
gyrus) and/or middle temporal gyrus
 Supports what we derive from the
proximity principle
Proceed with Caution!
 We already know that a noun or a verb has a
complex cortical representation
 Therefore it is not in a single location
 Rather, a functional web
 So what are we talking about?
• The cardinal node of the functional web of a noun
• The cardinal node of the functional web of a verb
A patient with non-fluent aphasia
 Patient ROX (McCarthy & Warrington 1985)
• Impaired production and comprehension of verbs
• Excellent production and comprehension of nouns
• Had no difficulty imitating common actions
• But had difficulty in performing same actions in
•
response to verbal commands
 E.g., confused open and close
In action-naming test, sometimes substituted
nouns for verbs or omitted verbs:
 “chairing” for sitting
 “The man is a sack of potatoes” for
a man carrying a sack of potatoes
Verb deficit and agrammatism: Why?
 Syntactic hypothesis
• Verbs are by their nature syntactically complex
• Nouns are not complex – they can stand alone
 Semantic hypothesis – using proximity
• Verbs represent processes and processes are
•
managed by the frontal lobe
Nouns represent things, and things are known
mainly through perception, which is managed by
the occipital, temporal, and parietal lobes
Noun-Verb vis-à-vis Speech & Writing
(908b)
 Patient S.J.D.
•
•
•
Written naming of verbs defective
But oral naming okay
Nouns okay for both writing and speaking
•
•
•
Oral naming of verbs defective
But written naming okay
Nouns okay for both writing and speaking
 Patient H.W.
 Comparable results independently of
mode of stimulus – picture naming,
reading, writing to dictation
 (More, next slide..)
More on H.W. & S.J.D.
Noun-Verb vis-à-vis Speech-Writing
(Rapp & Caramazza 908-9)
 Inputs: pictures, oral dictation, reading
 Tasks: (1) speaking, (2) writing
 Example:
• There’s a crack in the mirror (crack as n.)
• Don’t crack the mirror (crack as v.)
• S.J.D.
 crack as n. correctly produced, both
modalities
 crack as v. correct only in spoken modality
• H.W. – the opposite modality effect
 Data from Caramazza & Hillis (1991)
Broca’s Area:
A closer look
Subdivisions of Broca’s area
 Broca’s area includes two different (but
adjacent) Brodmann areas
• BA 44 – Pars Opercularis
• BA 45 – Pars Triangularis
 (Some people also include the Pars
orbitalis, just inferior to the pars
triangularis)
Frontal Operculum
 Operculum: little cover
 The part of the frontal lobe that covers
(part of) the Sylvian fissure and anterior
insula
 Adjacent to and inside the anterior portion
of Sylvian fissure
 Opposite it (across Sylvian fissure) in
temporal lobe is the temporal operculum
Subdivisions of Broca’s area
 Another view
Major Language Areas
Supramarginal gyrus
(Goldstein’s area)
Angular gyrus
(Geschwinds’s area)
Wernicke’s area
Broca’s
area
Brodmann area 37
Left hemisphere,
showing
middle
cerebral
artery
Frontal
Operculum
A closer look at Broca’s aphasia
 Broca’s original patient
• Lesion was extensive
• Not just Broca’s area but also
 Adjacent areas
 Subjacent white matter
 A tradition has followed Broca
• Broca’s area held responsible for symptoms of
•
Broca’s aphasia
Confounding factor:
 Broca’s area is only part of the area of
damage with Broca’s aphasia
More recent findings
 Modern imaging methods help
 Some patients have damage restricted
mainly or entirely to just Broca’s area
 They have less severe symptoms that
typically associated with Broca’s aphasia
 Influential paper: Alexander et al. 1990
• Examined Broca patients with different
•
areas of damage
Cast doubts on importance of Broca’s area in
Broca’s aphasia
Broca’s area and Broca aphasia
 Maybe it’s not just Broca’s area damage
that is responsible for some of the
symptoms of “Broca’s aphasia”
 Maybe some of them result instead from
damage to neighboring areas
 Alexander et al. (1990) propose
distinguishing 3 subtypes
Three subtypes in Alexander study
1. Impaired speech initiation
•
•
Symptom traditionally attributed to
transcortical motor aphasia
Area of damage: frontal operculum
•
Area of damage: lower primary motor cortex
2. Disturbed articulatory function
3. The classical Broca’s aphasia syndrome
• More extensive damage
To be continued …
 (Later)
end