PSY 369: Psycholinguistics
Language Production:Models
Uhh…It is a.. You know.. A.. Arggg.I can almost see it, it has two
Syllables, I think it starts with A …..
TOT Meaning access No (little) phonological
access What about syntax?
Tip-of-the-tongue
Semantics Syntax
grammatical category (“part of speech”) e.g. noun, verb, adjective
Gender e.g. le chien, la vache; le camion, la voiture
Number e.g. dog vs. dogs; trousers vs. shirt
Count/mass status e.g. oats vs. flour
Tip-of-the-tongue
Vigliocco et al. (1997) Subjects presented with word definitions
Gender was always arbitrary If unable to retrieve word, they answered
How well do you think you know the word? Guess the gender Guess the number of syllables Guess as many letters and positions as possible Report any word that comes to mind
Then presented with target word Do you know this word? Is this the word you were thinking of?
Tip-of-the-tongue
Vigliocco et al (1997)
Scoring + TOT
Both reported some correct information in questionnaire
And said yes to recognition question - TOT
Otherwise
Vigliocco et al. (1997)
Vigliocco et al (1997)
Results + TOT: 84% correct gender guess - TOT: 53% correct gender guess
chance level Conclusion
Subjects often know grammatical gender information even when they have no phonological information
Supports split between syntax and phonology in production
Vigliocco et al. (1997)
Comparing models Central questions:
Are the stages discrete or cascading? Discrete: must complete before moving on Cascade: can get started as soon as some
information is available Is there feedback?
Top-down only Bottom up too
How many levels are there?
Levelt’s model Four broad stages:
Conceptualisation deciding on the message (= meaning to
express) Formulation
turning the message into linguistic representations
Grammatical encoding (finding words and putting them together)
Phonological encoding (finding sounds and putting them together)
Articulation speaking (or writing or signing)
Monitoring (via the comprehension system)
Message
Lexicon
Grammatical
Form
Articulation
FunctionalProcessing
PositionalProcessing
Network has three strata conceptual stratum
lemma stratum
word-form stratum
Message
Lexicon
Grammatical
Form
Articulation
FunctionalProcessing
PositionalProcessing
Levelt’s model
Tip of tongue state when lemma is retrieved without word-form being retrieved
Message
Lexicon
Grammatical
Form
Articulation
FunctionalProcessing
PositionalProcessing
Levelt’s model
Formulation involves lexical retrieval:
Semantic/syntactic content (lemma)
Phonological content (word-form)
has stripes is dangerous
TIGER (X)
Fem.
Noun countable
tigre
/tigre/
/t/ /I/ /g/
Lexical concepts
Lemmas
Lexemes
Phonemes
Levelt’s model
Lexicon
Conceptual stratum
Conceptual stratum is not decomposed one lexical concept node for
“tiger” instead, conceptual links from
“tiger” to “stripes”, etc.
has stripes is dangerous
TIGER (X)
First, lemma activation occurs This involves activating a
lemma or lemmas corresponding to the concept
thus, concept TIGER activates lemma “tiger”
Lexical selection
Fem.
Noun countable
tiger
TIGER (X)
First, lemma activation occurs This involves activating a lemma
or lemmas corresponding to the concept
thus, concept TIGER activates lemma “tiger”
Lexical selection
tiger But also involves activating other lemmas
TIGER also activates LION (etc.) to some extent
and LION activates lemma “lion”
TIGER (X) LION (X)
lion
Selection is different from activation Only one lemma is selected Probability of selecting the
target lemma (“tiger”) ratio of that lemma’s activation
to the total activation of all lemmas (“tiger”, “lion”, etc.)
Hence competition between semantically related lemmas
Lemma selection
tiger
TIGER (X) LION (X)
lion
Morpho-phonological encoding (and beyond)
The lemma is now converted into a phonological representation
called “word-form” (or “lexeme”)
If “tiger” lemma plus plural (and noun) are activated
Leads to activation of morphemes tigre and s
Other processes too Stress, phonological
segments, phonetics, and finally articulation
/tigre/
/t/ /I/ /g/
Modularity Later processes cannot affect earlier processes
No feedback between the word-form (lexemes) layer and the grammatical (lemmas) layer
Also, only one lemma activates a word form If “tiger” and “lion” lemmas are activated, they
compete to produce a winner at the lemma stratum
Only the “winner” activates a word form The word-forms for the “losers” aren’t accessed
Model’s assumptions
tiger
Picture-word interference task Participants name basic
objects as quickly as possible
Distractor words are embedded in the object
participants are instructed to ignore these words
Experimental tests
Semantically related words can interfere with naming e.g., the word TIGER in
a picture of a LION
Basic findings
tiger
However, form-related words can speed up processing e.g., the word liar in a
picture of a LION
Basic findings
liar
Experiments manipulate timing: picture and word can be presented
simultaneously
time
liar
Experiments manipulate timing: picture and word can be presented
simultaneously
liar
time
liar
or one can slightly precede the other We draw inferences about time-course of processing
Schriefers, Meyer, and Levelt (1990)
SOA (Stimulus onset asynchrony)
manipulation -150 ms (word …150 ms … picture) 0 ms (i.e., synchronous presentation) +150 ms (picture …150ms …word)
Auditory presentation of distractors DOT phonologically related CAT semantically related SHIP unrelated word
Schriefers, Meyer, and Levelt (1990)
Auditory presentation of distractors DOT phonologically related CAT semantically related SHIP unrelated word
500520540560580600620640660680700
-150 0 150
DOTCATSHIP
EarlyOnly Semantic effects
Schriefers, Meyer, and Levelt (1990)
Auditory presentation of distractors DOT phonologically related CAT semantically related SHIP unrelated word
500520540560580600620640660680700
-150 0 150
DOTCATSHIP
LateOnly Phonological effects
Early semantic inhibition Late phonological facilitation Fits with the assumption that semantic processing
precedes phonological processing No overlap
suggests two discrete stages in production an interactive account might find semantic and phonological
effects at the same time
Interpretation
Dell’s interactive account
Dell (1986) presented the best-known interactive account other similar accounts exist
3 levels of representation semantics (decomposed into features) words phonemes (sounds)
Interactive because information flows “upwards” as well as “downwards”
e.g., the semantic features mammal, barks, four-legs activate the word “dog”
this activates the sounds /d/, /o/, /g/ these send activation back to the word level, activating
words containing these sounds (e.g., “log”, “dot”) to some extent
this activation is upwards (phonology to syntax) and wouldn’t occur in Levelt’s account
Mixed errors Both semantic and phonological relationship to target word Target = “cat”
semantic error = “dog” phonological error = “hat” mixed error = “rat”
Occur more often than predicted by modular models if you can go wrong at either stage, it would only be by chance
that an error would be mixed
Evidence for Dell’s model
The semantic features of dog activate lemma “cat” Some features (e.g., animate, mammalian) activate “rat” as
well “cat” then activates the sounds /k/, /ae/, /t/ /ae/ and /t/ activate “rat” by feedback this confluence of activation leads to increased tendency for
“rat” to be uttered Also explains the tendency for phonological errors to
be real words Sounds can only feed back to words (non-words not
represented) so only words can feedback to sound level
Dell’s explanation
Why might interaction occur?
Can’t exist just to produce errors! So what is feedback for?
Perhaps because the same network is used in comprehension
So feedback would be the normal comprehension route Alternatively, it simply serves to increase fluency
in lemma selection advantageous to select a lemma whose phonological
form is easy to find
Schriefers, Meyer, and Levelt (1990)
Also looked for any evidence of a mediated priming effect
hat dog
DOG (X) CAT (X)
cat
/cat/ /hat/
/t//a//k/ /h/
Found no evidence for it
Alternatively, it simply serves to increase fluency in lemma selection advantageous to select a lemma whose
phonological form is easy to find
Evidence for interactivity
A number of recent experimental findings appear to support interaction under some circumstances (or at least cascading models) Peterson & Savoy (JEP:LMC, 1998) Cutting & Ferreira (JEP:LMC, 1999) Griffin & Bock (JML, 1998) Damian & Martin (JEP:LMC, 1999)
Peterson & Savoy found evidence for phonological activation of near synonyms:
Participants slower to say distractor soda than unrelated distractor when naming couch
Soda is related to non-selected sofa
remember that Levelt et al. assume that only one lemma can be selected and hence activate a phonological form
Levelt et al’s explanation: Could be erroneous selection of two lemmas?
Damian and Martin (1999)
Extension of Schriefers et al.’s picture-word interference task remember that semantic inhibition occurred
early, phonological facilitation occurred late (with no overlap)
various methodological changes and developments
focus on Experiment 3
The critical difference from Schriefers et al. is the addition of a “semantic and phonological” condition
Picture of Apple peach (semantically related) apathy (phonologically related) apricot (sem & phono related) couch (unrelated) (also no-word control, always fast)
Results
Relatedness -150 ms 0 ms + 150 ms
Unrelated 670 702 691
Semantic 702 733 697
Phono. 665 683 646
S & P 679 676 654
Summary of findings
early semantic inhibition (- 150 and 0 ms)
late phonological facilitation (0 and + 150 ms) shows overlap, unlike Schriefers et al.
but S & P condition didn’t show early semantic inhibition
This last finding demonstrates that semantic interference is reduced in the simultaneous presence of a phonological relationship (which should facilitate)
Thus the finding appears to contradict the “discrete two-step” account of Levelt et al.
Can the two-stage account be saved?
Evidence for interaction is hard to reconcile with the Levelt account however, most attempts are likely to
revolve around the monitor basically, people sometimes notice a
problem and screen it out Levelt argues that evidence for interaction
really involves “special cases”, not directly related to normal processing
Summary
Levelt et al.’s theory of word production: Strictly modular lexical access Syntactic processing precedes phonological
processing Dell’s interactive account:
Interaction between syntactic and phonological processing
Experimental evidence is equivocal, but increasing evidence that more than one lemma may activate associated wordform
Caramazza’s alternative
Caramazza and colleagues argue against the existence of the lemma node
instead they propose a direct link between semantic level and lexeme
syntactic information is associated with the lexeme Also assumes separate lexemes for written and spoken
production This is really a different issue
Much evidence comes from patient data But also evidence from the
independence of syntactic and phonological information in TOT states see discussion of Vigliocco et al. also Caramazza and Miozzo (Cognition,
1997; see also replies by Roelofs et al.)