Language comprehension (understanding) Flashcards
A homograph is a word that is
Spelt the same as another word but doesn’t have the same meaning
Speech is ambiguous because
The speech stream is a continuous stream of sound
If you are unable to hear all the sounds in the word you will
Fill in the gaps by activating mental representations of the words
The mental lexicon is accessed…
via activation of mental representations
Does ‘frequency’ affect the speed with which a mental representation of a word is accessed?
yes
Does ‘neighbourhood density’ affect the speed with which a mental representation of a word is accessed?
yes
Does ‘semantic association’ affect the speed with which a mental representation of a word is accessed?
yes
Does ‘speech onset’ affect the speed with which a mental representation of a word is accessed?
no
If the following sentence is presented ‘she touched the bark of the oa_’ priming should be found for ‘tree’ because it is related to ‘oak’, but will priming also be found for ‘porridge’ because it is related to oat?
yes (/probably)
What is phoneme restoration?
If gap in word = dont hear all then find hard to fill in blank e.g. if someone coughed, but can understand the word as we already know it
= strong evidence for the mental lexicon= access the phonological representation of word
What are homophones?
words that sound the same e.g. muscle and mussel
What are homographs?
Words that are spelt the same e.g. bow (tie) and bow (arrow)
What are homonyms?
Words that sound are are spelt the same e.g. bank £££££££££££££ and bank river
What are homophones, homographs and homonyms an example of?
Word level ambiguity
What is coarticulation?
They way a word changes because of its environment
- depending on what word is coming up next the sound of the preceding word can change e.g. pass me thin book= mouth has to close before b for book= sounds more like an m
but look at the thin carpet= mouth stays open so is clearer
the phoneme /n/ is pronounced differently is an allophone
What can categorical perception help us do?
speak
= can distinguish between two sounds easily because we are involved to do so
- not unique to humans, can train animals to do so e.g. chinchillas
= inherent in us
Does categorical perception help us understand speech streams?
yes, helps unpick whats going on as we can discriminate between particular sounds, and language evolved to use the sounds
What did Werker test with children?
whether they had categorical perception,
could discriminate sounds in children = ba,da in hindi that english speaking adults couldnt
= lose this ability by 10m
Why is there ambiguity in speech streams?
- Because its impossible to know where one word ends and another one starts
- There are no logical breaks
- With some soundwaves when it looks like there is a break there isnt actually a break
How are words recognised in the mental lexicon?
frequency and length of a word will make them quicker/slower to be accessed in lexicon
- low frequency = harder to find = slower
What is neighbourhood density?
Words that sound similar
- when we have a lot= become active in the mental lexicon
when semantic activation, also activates words that are phonetically related e.g. tiger and tie
Luce and pisoni 98 =if lots of neighbours = slow
What basis do we access lexical items?
- Acoustic input
- Lexical characteristics e.g. frequency, neighbourhood density
- Prediction
Lexical items with similar phonology or semantics are activated
Is lexical activation like a dictionary ?
not at all
What is the process of access/activation of lexicons?
1st have to activate word in mental lexicons, then select word and integrate it into the structure to understand what the sentence may be
What is the impact of context?
- Helps us predict
- word monitoring= marslen-wilson and taylor 80
= THE COHORT MODEL
= listen to speech stream and monitor partic word
= when its presented they also have to say the word - clock time when start to speak= speech onset
when sentence made sense = said motorway before hear end of word
when didnt make sense= lot slower
= EVIDENCE in mental lexicons that when hearing speech= you’re working oujt what it means and given the ability to predict what is coming next
= PREDICTS WHATS COMING UP NEXT AND ACTIVATES THEM IN THE MENTAL LEXICON
What did Zwitserlood 1989 study?
lexical access -Cross modal priming task -Present prime word = auditory -Target word= visual e.g. prime =captain - FOUND :despite fact that sentence tells you what will be coming up at the end- priming for both ship and slave THUS CONTEXT NOT AFFECTING PREDICTION AS MUCH AS WOULD EXPECT
Why was the cohort model revised ?
To account for the non-effect of context- to allow for these effects
- Context influences integration of word into sentence, not lexical selection
- Means lexical selection = initial activation of words = so this is not affected by context = shown by Zwitserlood
- INITIAL ACTIVATION IS MOTIVATED BY BOTTOM UP PROCESSES
- Competitor effects leave one word more active than others
Active = in relation to activation of neurons
What methods are used to assess language comprehension?
- word monitoring
- lexical decision task
- cross modal priming
- sentence priming
What is the lexical decision task?
Pts say whether think words on screen are a real word or not
can be combined with a prime e.g. doctor, word is nurse
- prime word= designed to get the lexicon going, should mean the word nurse is already active before its presented
= should react quickly
-when isnt related= need to activate word from scratch = slower
