Transcript
8/17/2019 MA Lecture 4
1/63
Language Behaviour: Psycholinguistics
Psycholinguistics investigates the mental
mechanisms underlying language
processing. (Cognitive)
How to perceive words and store them in the
mind, how to understand a sentence, how tolearn to read, how language and writing
systems influence mental organizations.
8/17/2019 MA Lecture 4
2/63
What is Psycholinguistics
Two Aspects:
Language Comprehension –how weunderstand the meaning of words and
sentences (receptive process)
Language Production –how we speak and
use language (productive process)
8/17/2019 MA Lecture 4
3/63
Example: Visual Word Recognition
When people encounter a printed word,
how do they identify it?
Properties of the word
Visual or auditory (Modality)
Orthographic –whether a word is constructed in a legal
way (e.g. “ math” vs “ mtah” )
Phonological (syllable: /cat/; phoneme: /k/, /e/, /t/,
Meaning
Time course of activation of the 3 types of linguistic info in
visual word identification
8/17/2019 MA Lecture 4
4/63
Example: Word Recognition
In identifying a written word,
Visuo-orthographic analysis occurs.(activate, activation)
Is the meaning of a word accessedautomatically?
Is the phonological information of a wordaccessed automatically?
8/17/2019 MA Lecture 4
5/63
Example: Word Recognition
An Experiment:
Task: Speak out the color name of the stimulus youwill view. (red, blue, green)
Requirement: Complete the task as quickly andaccurately as possible.
(The task was devised by Stroop, 1935 --the “ Stroop task” .)
8/17/2019 MA Lecture 4
6/63
8/17/2019 MA Lecture 4
7/63
8/17/2019 MA Lecture 4
8/63
8/17/2019 MA Lecture 4
9/63
Example: Word Recognition
Expected results:
For color words (red, green, blue), when theirink color is inconsistent with the meaning of
the words,
Naming time is longer (i.e.response latency);
Responses are less accurate.
8/17/2019 MA Lecture 4
10/63
Example: Word Recognition
Why there is such an interference effect?
8/17/2019 MA Lecture 4
11/63
The mental lexicon
red
*
/blu:/
red
blue
8/17/2019 MA Lecture 4
12/63
The above finding suggests that themeaning of words is activatedautomatically –people cannot
control the activation of meaning.
automaticity
Example: Word Recognition
8/17/2019 MA Lecture 4
13/63
Example: Word Recognition
What about the phonological information
of a written word in a silent reading task?
Is a word’s phonology activated in a task
that does not require reading aloud?
8/17/2019 MA Lecture 4
14/63
Example: Word Recognition
Another experiment:
Task: On each trial, you will first see a semanticcategory name (e.g., “flower”). Following thesemantic category, you will see a target wordwhich may be an exemplar of the category(e.g., “rose”). Judge if the target word you willsee is an correct exemplar of the category.
8/17/2019 MA Lecture 4
15/63
Example: Word Recognition
A servant
8/17/2019 MA Lecture 4
16/63
Example: Word Recognition
maid
8/17/2019 MA Lecture 4
17/63
Example: Word Recognition
Type of food
8/17/2019 MA Lecture 4
18/63
Example: Word Recognition
meet
8/17/2019 MA Lecture 4
19/63
Example: Word Recognition
Part of a mountain
8/17/2019 MA Lecture 4
20/63
Example: Word Recognition
peek
8/17/2019 MA Lecture 4
21/63
Example: Word Recognition
A flower
8/17/2019 MA Lecture 4
22/63
Example: Word Recognition
rows
8/17/2019 MA Lecture 4
23/63
Example: Word Recognition
Expected results:
High false “ YES” responses to
homophones of the exemplars.
Suggest that the meaning of words is
accessed via phonology.
8/17/2019 MA Lecture 4
24/63
flower
rose/rouz/
rows
rows
Yes
(false)
8/17/2019 MA Lecture 4
25/63
Why we do this sort of stuff???
Research on normal readers’ word identificationprocesses provide a solid basis for treating
abnormal readers.
13% poor readers
7-15% developmental dyslexics (dyslexia)
(Developmental dyslexia: Individuals with normal to
high intelligence have severe reading problems.)
8/17/2019 MA Lecture 4
26/63
Findings with normal readers indicate that
word meaning and phonology are activated
automatically.
Apply the same tasks to dyslexia. There is a
phonological deficit for dyslexics –they
cannot activate phonological informationobligatorily.
8/17/2019 MA Lecture 4
27/63
Research Methodology inPsycholinguistics
1. Reaction time (RT) approach
It measures people’s RT (response latency) to alanguage stimulus. It includes many on-line
methods of studying people’s language behavior
continuously in a laboratory setting“ RT measures time from onset of a stimulus to
onset of a response.”
Infer the mental activity in terms of RTs.
8/17/2019 MA Lecture 4
28/63
dog
8/17/2019 MA Lecture 4
29/63
dog/dawg/
8/17/2019 MA Lecture 4
30/63
Research Methodology inPsycholinguistics
2. Memory and Classroom Approach
It investigates language behavior inclassroom, a more natural setting.
To study the developmental pattern of
children’s reading ability:
cross-sectional (grades 1 to grade 5)
longitudinal (a 5-year study)
8/17/2019 MA Lecture 4
31/63
Research Methodology inPsycholinguistics
3. Eye movements Approach
Use the eyetracker to measure the positionand duration of people’s eye fixations uponthe text
It is a most reliable psycholinguistictechnique.
8/17/2019 MA Lecture 4
32/63
Research Methodology inPsycholinguistics
How the Eyes Work?
When people read, the eyes do NOT movecontinuously. Instead, the eye will “ stop” on
some word for a brief period (150 to 500
milliseconds). This is called fixation.Between the fixations are periods where the
eye moves rapidly. These rapid eye
movements are called saccades.
8/17/2019 MA Lecture 4
33/63
Research Methodology inPsycholinguistics
To understand a sentence, the eyes
sometimes move back to some word, re-fixating it. These “ backward” saccades aretermed regression.
Words that are fixated are in the fovea; wordssurrounding the fixated words are in theparafovea.
8/17/2019 MA Lecture 4
34/63
Research Methodology inPsycholinguistics
A healthy body may seem reward enough for. . . . . . .
177 196 175 244 302 112 177
most people. However, for all those who. . . . .
266 188 199 216 212
question the payoff, some recent research…. . . .
179 266 245 188
8/17/2019 MA Lecture 4
35/63
Research Methodology inPsycholinguistics
How the Eyes Work?
Fixation Saccades
duration length (characters)
English 231 msec. 7.8
Hebrew 265 msec. 5.5
Chinese 300 msec. 2.0
8/17/2019 MA Lecture 4
36/63
Research Methodology inPsycholinguistics
How the Eyes Work?
Information extracted from the parafovea aids in reading:
Reading rate is about 200 words per minute when only
the fixated word is visible on each fixation, but it is 300words per minute when both the fixated word and the one
immediately to the right are visible on each fixation.
8/17/2019 MA Lecture 4
37/63
Research Methodology inPsycholinguistics
4. Neurolinguistic Approach
It studies the brain mechanism for languagefunctioning –where is language functioning
localized in the brain?
Lesion study –using brain-damage patients
Functional brain imaging study –using the
intact human brain
8/17/2019 MA Lecture 4
38/63
Research Methodology inPsycholinguistics
Lefthemisphere
Righthemisphere
cortex
8/17/2019 MA Lecture 4
39/63
Research Methodology inPsycholinguistics
Basic Neuroanatomy:
The outer layer of the brain –cerebral cortex
Hidden underneath the cortex are
subcortical parts of the brain
The left and right hemispheres are
connected by a band of nerve fibers –
corpus callosum
8/17/2019 MA Lecture 4
40/63
Research Methodology inPsycholinguistics
Basic Neuroanatomy:
An important feature of the human nervoussystem is that each cerebral cortex isconnected to the opposite side of the body.
This is termed contralateral connections.
(Same-side connections are very weak.)
8/17/2019 MA Lecture 4
41/63
Research Methodology inPsycholinguistics
Split-brain patients –those who have a
damaged corpus callosum but an
undamaged brain.
8/17/2019 MA Lecture 4
42/63
8/17/2019 MA Lecture 4
43/63
Research Methodology inPsycholinguistics
8/17/2019 MA Lecture 4
44/63
Research Methodology inPsycholinguistics
Results:
The left hemisphere is a language
hemisphere. (dominant).
8/17/2019 MA Lecture 4
45/63
Research Methodology inPsycholinguistics
Brain Imaging Techniques
Language processing requires energy.
Brain imaging techniques depend on the
fact that working brain tissue calls moreblood its way and consumes more
glucose.
8/17/2019 MA Lecture 4
46/63
Research Methodology inPsycholinguistics
8/17/2019 MA Lecture 4
47/63
• General Comments on LesionStudies:
1. The association between language
processing (e.g. syntactic processing)
and the damaged brain areas is direct.
8/17/2019 MA Lecture 4
48/63
8/17/2019 MA Lecture 4
49/63
2. The primary limitation is that lesions studies
rely on damaged brains. This has severaldrawbacks because
* patients with distinct lesions and specific
functional impairments are rare; in many
case, lesions are large;
* following brain damage, performance may
not reflect normal language processingbecause patients may compensate for their
deficits by using unusual processing
strategies.
8/17/2019 MA Lecture 4
50/63
• Can we reliably investigatenormal people’s intact brain by
using noninvasive techniques??
8/17/2019 MA Lecture 4
51/63
Research Methodology inPsycholinguistics
• Brain Imaging Techniques (PET & fMRI)
Language processing requires energy.
Brain imaging techniques depend on the
fact that working brain tissue calls more
blood its way and consumes moreglucose.
Functional brain imaging:
8/17/2019 MA Lecture 4
52/63
Functional brain imaging:
Directly visualize brain activities of normalpeople.
8/17/2019 MA Lecture 4
53/63
• PET (positron emission tomography)
Subjects are injected with glucose that
has been tagged with a radioactivesubstance; through this, brain imagescan be got indicating which regions of
the brain have the greatest blood floware are using the most energy.
8/17/2019 MA Lecture 4
54/63
8/17/2019 MA Lecture 4
55/63
Washington University at St. Louis:
1986: Fox et al., Nature
1988: Fox et al., Science
1988: Petersen et al., Nature1988: Posner et al. Science
8/17/2019 MA Lecture 4
56/63
•fMRI (functional magnetic resonanceimaging):
When subjects perform a task (e.g., reading), neural
activity in specific areas of the brain increases. This
results in greater need for glucose and oxygen, both of
which are met by increasing blood flow. There is asmall difference in magnetic susceptibility (a property
of molecules) between oxygenated hemoglobin and de-
oxygenated hemoglobin. With increased blood flow toactive brain areas, there is a change in the proportion
of oxygenated to de-oxygenated hemoglobin. This
difference is detected by MRI scanners.
8/17/2019 MA Lecture 4
57/63
Massachusetts General Hospital(Harvard University):
1992: K.K. Kwong et al. (PNAS)
S. Ogawa et al. (PNAS)Peter Bandettini
“ B i I i ” i li th
8/17/2019 MA Lecture 4
58/63
“ Brain Imaging ” -- visualize theactivity of the intact human brain
8/17/2019 MA Lecture 4
59/63
8/17/2019 MA Lecture 4
60/63
8/17/2019 MA Lecture 4
61/63
8/17/2019 MA Lecture 4
62/63
8/17/2019 MA Lecture 4
63/63
• Brain science is (over-)interdisciplinaryand is a team’s work. Psycholinguistics
and Linguistics are parts of it.
top related