Lectures 7 and 8: Word reading

HideShow resource information
What is cognition?
The integration of conscious thoughts (many) and unconscious thoughts (most.)
1 of 31
What are the levels of analysis of word reading?
1. Conscious awareness - read, aware of words 2.Unconscious - representations, modules, processes, architecture. 3.Neural - neurons work out what we do (but we don't know how they do it)
2 of 31
What are the elements of unconscious reading?
1. Mental Dictionary (lexagon) 2. Process- (methods for meaning). 3. Architecture - (mechanisms for task). 4. Control (one task at a time) - remember with MPAC
3 of 31
What is the ultimate aim of a theory?
To sufficiently explicit that we can simulate the computations the mind makes
4 of 31
What is a computational model?
Does simulation, a- get the task done and b- perform like a human doing the task? Just a = artificial intelligence, A and B = computer simulation
5 of 31
What are the component processes in reading for meaning?
Reading is complex, not uniform and has many components.
6 of 31
What is the Lexical Access? What does it consist of?
Identify letters ---> words ---> retrieve ---> sentence structure ---> sentence meaning ---> intention
7 of 31
Define: Word, pronunciation and spelling pattern.
Word = a form - has a function (what is signified). Pronunciation is phonology, meaning is semantic. Spelling pattern is orthography, has a syntactic role.
8 of 31
What is the spoken word form?
Phonemes (consonant and vowel sounds) arranged into stress patterns and syllables. Some cultures are different e.g. chinese have tone pitch pattern. Different languages have different numbers of phonemes e.g. English has 40.
9 of 31
What is the written word form?
Graphemes - written by curves and lines (strokes.) Write left to right and top to bottom.
10 of 31
How do we translate between spelling, pronunciation and meaning?
Relation between form and meaning is arbitrary, so we need a process that identifies the written form (from memory) and retrieves the connected meaning associated.
11 of 31
What is the Word Identification problem?
We know 100,000 words. We recognise 2/3 a second, but takes 200ms to recognise 1 - this does not add up.
12 of 31
How do we solve the Word Identification problem using option 1?
Introspective Reports - can examine what we are consciously aware of as we perform cognitive tasks. But gives little access, is misleading and doesn't tell how word identification is done.
13 of 31
How do we solve the Word Identification problem using option 2?
Option 2 is experimental - artificial lab tasks to capture process of interest and measure response. Online measure (like practical) is more naturalistic.
14 of 31
What is the word superiority effect?
When a word is in the order it should be, or a letter in the correct string they are more likely to be recognised and identified.
15 of 31
Explain the lab phenomena: word superiority effect, frequency effect, and sentence context effects.
WSE = accuracy of letter identification is greater for word than nonword. FE = RT's for lexical decision, categorisation and naming quicker for frequent words. SCE = RT's for lexical decision and naming shorter for word that fits in sentence context
16 of 31
What is the problem with the observed lab phenomena, and what is an alternative?
They don't explain daily reading. Instead we can measure eye fixation (eye tracker.) Eyes jump 8-9 characters, focus on middle words, and ignore function words. We can measure word identification span (currently 7 to right, 3 to left)
17 of 31
What is perceptual span?
The amount of angular span the eye has (horizontal and vertical) - is 120 degrees from whatever the eye is reading.
18 of 31
How do we investigate lexical identification?
Use lab phenomena methods.
19 of 31
What is theory 1 of word identification?
Serial Search Model - Forster 1976 - compare one word at a time, if non-word then not found. Problem= this is too fast for neurons
20 of 31
What is theory 2 of word identification?
Modified Search Model - Forster and Murray, 2004) - mental lexicon of words in 'bins' - categories spelling pattern to select bin. Frequency search is done in individual bin.
21 of 31
What is a parallel-matching word-detector model?
'Feature' unit for each letter feeds up to 'letter' unit for each letter, which feeds up to 'word' layer. Can flow downwards. Transient (one active at a time)
22 of 31
What is the effect of sentence context on word identification?
Words easier to understand in context of a sentence. Frequency and context have effects on recognition, so theories have to facilitate this. Two possible models.
23 of 31
What are the two possible models of frequency effect on words?
Forster Serial Search: lexagon search for frequency, high frequency easier to find. IA-style-parallel matching process: most used detectors = more sensitive, high frequency= easier to find.
24 of 31
What can the brain tell us about reading?
Can look at brain-damaged patients, and use non-invasive measures and manipulations on normal participants to look at identification etc.
25 of 31
What do brain damaged patients tell us about phonological mediation?
It is not necessary - these patients can understand written words but not access sound patterns e.g. chaos - 'chost'
26 of 31
What does deep dyslexia indicate?
nowhere near pronunciation but get meaning of word - see tulip and say daffodil. So phenology not needed to get meaning.
27 of 31
Who suggested that phonological mediation occurs anyway? Why did they think this?
Van Orden - homophone errors e.g. hair/hare - just thought that it was less likely if high frequency words.
28 of 31
What is the dual route model and what does it suggest? (also explains deep dyslexia but no room)
Suggests functionally and anatomically separate pathways for reading, assembly and retrieval. Explains surface dyslexia=OK on words + non-words, but no exception words. Phonological dyslexia=OK on high frequency words, not non/unfamiliar words
29 of 31
What is the main rival of the dual route model? What does it suggest?
The triangle model - says one direct route, single network that learns to translate both regular and exception patterns, provided they are frequent.
30 of 31
How can we manipulate the brain to study brain activation?
fMRI, EEG (especially ERP). Allow us measure activity, fMRI and PET = where, ERP = when - BUT NONE= how. TMS can be used to disrupt function and then measure outcome.
31 of 31

Other cards in this set

Card 2

Front

What are the levels of analysis of word reading?

Back

1. Conscious awareness - read, aware of words 2.Unconscious - representations, modules, processes, architecture. 3.Neural - neurons work out what we do (but we don't know how they do it)

Card 3

Front

What are the elements of unconscious reading?

Back

Preview of the front of card 3

Card 4

Front

What is the ultimate aim of a theory?

Back

Preview of the front of card 4

Card 5

Front

What is a computational model?

Back

Preview of the front of card 5
View more cards

Comments

No comments have yet been made

Similar Psychology resources:

See all Psychology resources »See all Word reading resources »