Title: Language%20Comprehension%20%20Speech%20Perception%20Naming%20Deficits
1Language Comprehension Speech PerceptionNaming
Deficits
2Triangle Model
Thought / High-level understanding
Semantics meaning
Reading
Speech perception
Orthography text
Phonology speech
Connectionist framework for lexical processing,
adapted from Seidenberg and McClelland (1989) and
Plaut et al (1996).
3Speech Perception
- The first step in comprehending spoken language
is to identify the words being spoken, performed
in multiple stages - 1. Phonemes are detected (/b/, /e/, /t/, /e/,
/r/, ) - 2. Phonemes are combined into syllables (/be/
/ter/) - 3. Syllables are combined into words (better)
- 4. Word meaning retrieved from memory
4Spectrogram I owe you a yo-yo
5Speech perception two problems
- Words are not neatly segmented (e.g., by pauses)
- Difficult to identify phonemes
- Coarticulation consecutive speech sounds blend
into each other due to mechanical constraints on
articulators - Speaker differences pitch affected by age and
sex different dialects, talking speeds etc.
6(No Transcript)
7How Do Listeners Resolve Ambiguity in Acoustic
Input?
- Use of context
- Cross-modal context
- e.g., use of visual cues McGurk effect
- Semantic context
- E.g. phonemic restoration effect
8Effect of Semantic Context
- Pollack Pickett (1964)
- Recorded several conversations.
- Subjects in their experiment had to identify the
words in the conversation. - When words were spliced out of the conversation
and then presented auditorily, subjects
identified the correct word only 47 of the time. - When context was provided, words were identified
with higher accuracy - clarity of speech is an illusion we hear what we
want to hear
9Phonemic restoration
Auditory presentation Perception Legislature
legislatureLegi_lature legi latureLegilature
legislature It was found that the eel was on
the axle. It was found that the eel was on
the shoe. It was found that the eel was on the
orange. It was found that the eel was on the
table.
wheel
heel
peel
meal
Warren, R. M. (1970). Perceptual restorations of
missing speech sounds. Science, 167, 392-393.
10McGurk EffectPerception of auditory event
affected by visual processing
Demo 1 AVI http//psiexp.ss.uci.edu/research/teac
hingP140C/demos/McGurk_large.avi MOV
http//psiexp.ss.uci.edu/research/teachingP140C/de
mos/McGurk_large.mov Demo 2 MOV
http//psiexp.ss.uci.edu/research/teachingP140C/de
mos/McGurk3DFace.mov
Harry McGurk and John MacDonald in "Hearing lips
and seeing voices", Nature 264, 746-748 (1976).
11McGurk Effect
- McGurk effect in video
- lip movements ga
- speech sound ba
- speech perception da (for 98 of adults)
- Demonstrates parallel interactive processing
speech perception is based on multiple sources of
information, e.g. lip movements, auditory
information. - Brain makes reasonable assumption that both
sources are informative and fuses the
information.
12Models of Spoken Word Identification
- The Cohort Model
- Marslen-Wilson Welsh, 1978
- Revised, Marslen-Wilson, 1989
- The TRACE Model
- Similar to the Interactive Activation model
- McClelland Elman, 1986
13Online word recognition the cohort model
14Recognizing Spoken Words The Cohort Model
- All candidates considered in parallel
- Candidates eliminated as more evidence becomes
available in the speech input - Uniqueness point occurs when only one candidate
remains
15Analyzing speech perception with eye tracking
Point to the beaker
Eye tracking device to measure where subjects are
looking
Allopenna, Magnuson Tanenhaus (1998)
16Human Eye Tracking Data
Plot shows the probability of fixating on an
object as a function of time
Allopenna, Magnuson Tanenhaus (1998)
17TRACE a neural network model
- Similar to interactive activation model but
applied to speech recognition - Connections between levels are bi-directional
and excitatory ? top-down effects - Connections within levels are inhibitory
producing competition between alternatives
(McClelland Elman, 1986)
18TRACE Model Predictions
(McClelland Elman, 1986)
19Semantic Representations and Naming Deficits
20Representing Meaning
- Mental representation of meaning as a network of
interconnected features - Evidence comes from patients with
category-specific impairments - more difficulty activating semantic
representation for some categories than for others
21Category Specific Semantic Deficits
Patients who have trouble naming living or
non-living things
22Definitions giving by patient JBR and SBY
Farah and McClelland (1991)
23Explanation
- One possibility is that there are two separate
systems for living and non-living things - More likely explanation
- Different types of objects depend on different
types of encoding - ? perceptual information
- ? functional information
24Representing Meaning
25Sensory-Functional Approach
- 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.
- How do we know that?
- Farah McClelland (1991) report a dictionary
study showing the ratio of visual to functional
features 7.71 for living things and 1.41 for
nonliving things
26A neural network model of category-specific
impairments
A single system with functional and visual
features. Model was trained to associate visual
picture with the name of object using a
distributed internal semantic representation
Farah and McClelland (1991)
27A neural network model of category-specific
impairments
lesions
lesions
Simulate the effect of brain lesions
Farah and McClelland (1991)
28Simulating the Effects of Brain Damage by
lesioning the model
Functional Lesions selective impairment of
non-living things
Visual Lesions selective impairment of living
things
Farah and McClelland (1991)