Lang & Comm 2 Flashcards
three stages in spoken word production
conceptualisation
formulation
articulation
conceptualisation
determining what you want to say (pre-verbal and universal)
formulation
translating pre-verbal message to linguistic form
4 steps involved in formulation (LSPP)
lexicalisation - choosing words
syntactic planning - forming sentence
phonological encoding - turning words to sounds
phonetic planning - pronouncing
articulation
verbal word pronounciation
who extended the process of spoken word production?
Levelt 1989 and Roelofs 1997
Who conceptualised the 3 major steps?
Griffin and Ferreira (2006)
What did Levelt and Roelof extend this process into?
the WEAVER++ model: adds self-monitoring
give the kinds of self-monitoring
internal (planning what to say) and external (you can self-correct when speaking)
what does WEAVER stand for
word form encoding by activation and verification
explain this in 3 parts:
prepare a concept (select lemma)
morphological encoding (grammar)
phonological encoding (select lexeme)
what is your mental lexicon?
store of all the words you know
what is within your mental lexicon?
lemma of words - the meaning of them.
lexeme - basic lexical unit of language of one or several words (play, plays, playing)
what are 3 pieces of evidence for the WEAVER++ model? (SPT)
- speech errors
- picture naming and picture-word interference
- tip of the tongue (ToT) phenomenon
how many speech sounds do we produce per second?
15
how does Levelt characterise speech errors?
they are automatic: “impossible to think in the middle of a word shall I say ‘t’ or ‘d’”
speech production vs. comprehension?
we pay less attention to production
how many speech errors per 1000 words? how many per day?
1-2
7-22
what is the Freudian belief on speech errors?
it’s our repressed thoughts
what is Dell’s belief on speech errors?
it reflects someone’s capacity for using language and its components
what is a major general belief on speech errors?
THEY DO NOT OCCUR AT RANDOM
List the 8 main types of speech errors (SPEDASAB)
shift
persevation
exchange
deletion
addition
substitution
anticipation
blend
word errors vs sound errors?
word errors: happen EARLY, not restricted by distance and always of same word type
sound errors: happen LATER, restricted by closeness and cross word type
list 4 common properties of errors (NECE)
Novel words follow language’s phonology (perple, not peorslpe)
Exchange of phonemes in similar positions
Consonants-consonants & vowels-vowels
Experimentally induced by SLIP technique
explain what this shows about errors:
Garrett’s model of speech production
from highest level to lowest = word errors happen earlier (chosen in lemma) and why sound errors happen later (lexeme errors)
lexicalisation
turning thought into sound
what are the stages of the two-stage retrieval process of lexicalisation
lemma (feline animal noun) and lexeme (/c a t/)
what is evidence for this 2-stage retrieval process?
there are 2 types of substitution errors - ALWAYS separate
what are the 2 types of substitution errors
semantic and phonological
describe ToT
Subjective feeling of knowing lemma but not lexeme
what gets activated when you experience ToT? what are they called?
phonologically similar words - interlopers
how is ToT evidence for the lexicalisation process?
shows that the retrieval of meaning is a separate process to retrieval of sound.
what are the 2 main theories explaining ToT?
blocking hypothesis
transmission deficit hypothesis
who introduced the blocking hypothesis?
Jones and Langford (1987)
what did Jones and Langford (1987) focus on?
blocking hypothesis for ToT - interlopers suppress activation of correct word
who introduced transmission deficit hypothesis?
Burke et al. (1991)
What did Burke et al (1991) focus on?
transmission deficit hypothesis - weak links between grammar and word form = problem with transmission of lemma to lexeme
according to transmission deficit hypothesis, what are the reasons for weak links?
ageing
low frequency words
what 3 pieces evidence favour transmission deficit hypothesis?
blocking is wrong: words with less interlopers = more ToTs
bilinguals = more ToTs
dyslexic children. = more ToTs
Who researched picture naming as evidence of lexicalisation?
Wheeldon and Monsell (1992)
Describe procedure Wheelson and Monsell’s experiment
pps had to name a picture (dog), where one condition did a priming generation task (finish sentence: man’s best friend is a ___) and control had no priming.
Describe findings of Wheelson and Monsell’s experiment
long-term priming occurred, 10-12 minutes after generation task, pps in control were slower at naming picture than experimental condition
does word form priming alone improve picture naming?
no - homophone priming of car boot and footwear boot did not persist in picture recognition
does semantic priming alone improve picture naming?
no - bilinguals finished a sentence in English and picture named in Welsh, and there was no priming seen.
What do picture naming tasks tell us about priming?
need BOTH word form and meaning to activate priming, otherwise there are no noticeable effects.
Explain this study’s procedure:
picture word interference also is evidence of lexicalisation: name picture ignoring distractor words
Explain this study’s findings:
a semantically related distractor word made naming slower than control distractor
a phonologically related distractor made word naming faster than control distractor
How can this study be extended? (Hint: SOA)
Distractor onset can be manipulated = Stimulus Onset Asynchrony
who conducted a picture word interference task study on SOA?
Schriefers et al 1990
Describe procedure of Schriefers et al’s 1990 experiment
name a picture while hearing distractor words:
semantically or phonologically related
150ms before or after picture
Describe findings of Schriefers et al’s 1990 experiment
semantically related distractor before picture = slower
phonologically related distractor just before picture = faster
- means that semantic comes in earlier than phonology
anomic aphasia
milder form of aphasia - speech fluency but unable to generate correct words - mostly verbs and nouns.
what are the 2 types of anomic aphasia
-lexical semantic aphasia - meaning is lost @lemma level
-phonological aphasia - wrong phonology selected @lexeme level
what does anomic aphasia support?
the 2-stage model of lexicalisation
How did Levelt at first describe his WEAVER++ model?
discrete and modular
what made Levelt’s WEAVER++ model discrete and modular?
a single item was produced based on semanticity, then phonological form is retrieved
why could the WEAVER++ model also be cascading?
there are ‘leakages’ between stages
multiple lemmas are activated from semantic to phonological stage.
what do these counteract?
the idea that the WEAVER++ model is discrete and modular
How was the WEAVER++ model tested as a cascading system and by who?
mediated priming (Levelt et al 1991)
whose study is this and what is it exploring?
Levelt et al 1991, effects of mediated priming and whether lexicalisation is a cascading system
Describe Levelt et al’s 1991 procedure
pps named a picture
also wearing headphones hearing a word that they had to identify as real or non-real before they said answer eg. ‘sheep’.
Describe Levelt et al’s 1991 findings
Early inhibitory priming of goat (semantic) @ lemma level
Late priming of sheet (phonological) @ lexeme level
No priming of goal (evidence against cascading systems)
Provide evidence in favour of lexicalisation being a cascading system
Peterson and Savoy 1998 - mediated priming of near synonyms was found (‘couch’ primed ‘soda’ through ‘sofa’)
What major assumption does the feedback system of lexicalisation argue against the discrete one?
errors are not at random
what 2 things prove the feedback lexicalisation model correct for arguing errors not being random
lexical bias (hissed mystery lesson, never missed cystery messon)
similarity effects (mixed substitution errors: comma - colon)
what are lexical biases and similarity effects supporting
that errors are not random as these mistakes result in real words more often than by chance.
what is the alternative to lexicalisation being feedback activated?
output monitor
why are our speech errors usually real words according to output monitoring? (ENG)
External self-monitoring means we suppress non-real words
No taboo slip of tongues (hit shed)
Galvanic skin responses elevated for taboo trials - we catch ourselves
what is subvocalisation
inner speech
what do studies look at to show subvocalisation?
time taken to silently read words
give 3 studies of subvocalisation
Filik and Barber (2011): pps spent longer reading words that didn’t rhyme (bath/kath/garth Northerners vs Southerners)
Ashby and Clifton (2005): people read publication for longer than experiment as it has 2 stresses rather than 1.
Corcoran and Frisson: stutterers took disproportionately longer to silently read words they stuttered on in their head
