1/143
Looks like no tags are added yet.
Phonetics
The study of the physical production, transmission, and perception of speech sounds.
Phonology
The study of how speech sounds are organised and used within a language.
Morphology
The study of the structure of words and the formation of word meanings.
Semantics
The study of the literal meanings of words, phrases, and sentences.
Pragmatics
The study of how context influences the interpretation of meaning.
Lemma
A mental representation of a word containing its meaning and grammatical information.
Lexeme
An abstract unit of vocabulary representing all forms of a single word.
Lexicalisation
The process by which a concept becomes established as a recognised word in a language.
Picture-Word Interference Task
A psycholinguistic experiment in which participants name pictures while ignoring distracting words.
Minimal Pairs
Two words that differ by only one sound and have different meanings.
Allophone
A variant pronunciation of a phoneme that does not change the meaning of a word.
Onomatopoeia
A word whose sound imitates or resembles the sound it describes.
The structure of language
Phonetics - sounds made for speech
Phonology - system of sounds
Morphology - formation of words
Syntax - adds structure to the sentence
Semantics - meaning of sentence
Pragmatics - meaning in context
Phone
phonetic segment occurring in language
Phoneme
sounds in language that distinguish between words e.g. s/v
Thought to speech process (Griffin & Ferreria)
Conceptualisation
Formulation
Articulation
Self monitoring (weaver)
Conceptualisation
what you want to express
Formulation
How you want to express
Articulation
Actually expressing it the word
Self monitoring model - WEAVER ++
Monitor speech internally (what one will say) and externally (what is said during speaking)
WEAVER - Evidence
Speech errors
Tip of tongue
Picture naming
Picture-word interference
Speech errors
Errors not randomly occuring
Automatic
less attention to production than comprehension
Phonemes in similar positions exchanged
constants, vowels exchanged
novel words follow phonological rules
experimentally induced
8 types of speech error
shift
exchange
Anticipation
presereration
addition
deletion
substitution
blend
Shift
A speech element (often a grammatical ending) moves from its correct position to another word. e.g. She decide to hits it
Exchange
Two sounds, words, or phrases swap places. e.g. model renosed (nose remodelled)
Anticipation
A sound from a later word appears too early. (e.g. bake my bike
Perseveration
A sound from an earlier word persists into a later word. (pulled a pantrum)
Addition
An extra sound, syllable, or word is inserted.e.g. explain it clarefully
Deletion
A sound, syllable, or word is omitted. intelligbly
Substitution
One sound or word is replaced by another. e.g. feather is too heavy (when supposed to say light)
Blend
Two competing words are combined into one. e.g. He is cable (calm and stable)
Interpretation of errors
Retrieving word in mental lexicon
Then syntactic frame consturcted where wordas are slotted e..g understand who did what to who
Two main type of errors
Word error
sound error
Word error
not restricted by distance - always same type - happen earlier rather than later
Sound error
happen close together - can cross word types - happen later
Garrets speech-production model
Conceptualisation → Formulation → articulation
Conceptualisation
Thinking about what message you want to convey
Formulation (levels)
Different levels:
Functional - lexical selection
Positional - grammatical encoding
Sound level - sound level
Functional level
lexical selection - The speaker selects the appropriate words (lemmas) and assigns grammatical roles. e.g. horse (actor), kick (action), man (object) - word errors ocurring here
Positional level
Gramatical encoding - The sentence structure is built, grammatical endings are added, and the sounds of words are retrieved. - correct order
Sound level
sound form - A detailed motor plan for speaking is created. - where sound error occur
Lexicalisation - two stage process
Semantic (meaning) - lemma level - The speaker produces a word with a similar meaning. e.g. glass → cup
Phonological (form) - lexeme level - The words sound similar but have different meanings. e.g. historical → hysterical
Semantic processing and phonological processing are separate stages.
double dissassociation evidence
Double Disassociation
Some patients with aphasia can:
Understand meanings but struggle to retrieve sound forms.
Others can:
Produce sounds relatively well but struggle with word meanings.
This provides evidence that:
Lemma retrieval (meaning) and
Lexeme retrieval (form)
are supported by separate cognitive processes
Tip of the tongue (TOT)
Temporary inability to retrieve a word despite knowing its meaning
. -Often involves partial access to the word's form (e.g., first sound, number of syllables, stress pattern).
interlopers
Similar-sounding words that become activated and interfere with retrieval of the target word (e.g., oxygen, oxytocin for oxymoron).
How does ToT support the two-stage model of lexicalisation?
Meaning can be retrieved without retrieving the word's sound, suggesting separate stages for semantic and phonological processing.
Blocking theory (TOT)
Interlopers prevent acitvation of the correct word
Transmission deficit (TOT)
Weak links between meaning and word form → limited actiation of target word
Why does evidence challenge the blocking hypothesis of ToT?
Similar-sounding words reduce rather than increase ToTs, suggesting they aid retrieval.
Why do bilingual speakers experience more ToTs?
Weaker meaning-to-sound connections from dividing language use across two languages.
What does ToT reveal about lemma and lexeme retrieval?
Meaning and grammatical information (lemma) can be retrieved even when word form (lexeme) cannot.
Why do dyslexic children support the transmission-deficit account?
They show normal semantic retrieval but greater phonological retrieval difficulties.
Picture naming (william & monsell)
To investigate whether lexical retrieval occurs in separate stages for meaning (lemma) and word form (lexeme).
What was the main finding of the picture-naming studies?
Prior generation of a word facilitated later picture naming, and the effect persisted across many trials.
Why is long-term facilitation important evidence for lexicalisation?
It suggests that activating a word's lexical representation has lasting effects on later retrieval.
What is homophone priming?
Facilitation caused by a word that sounds the same as the target word.
Why is the lack of long-term homophone priming important?
It suggests that semantic/lexical activation and phonological activation are separate processes.
Why does the lack of cross-language facilitation support a two-stage model?
It suggests facilitation occurs at a language-specific lexical level rather than a shared semantic level.
What is the picture-word interference task?
Participants name a picture while ignoring a distractor word presented at the same time or at a manipulated SOA (stimulus onset asynchrony).
What is the aim of the picture-word interference task?
To investigate the stages of lexicalisation and the effects of semantic and phonological distractors on word production.
What happens when the distractor word is semantically related to the picture?
Picture naming becomes slower (semantic interference).
Example: Picture = CAT, Distractor = DOG
Why do semantically related distractors slow picture naming?
Because related concepts compete during lemma selection.
What happens when the distractor word is phonologically related to the picture?
Picture naming becomes faster (phonological facilitation).
Why do phonologically related distractors speed up picture naming?
Because they activate similar sounds, facilitating lexeme retrieval.
What does semantic interference suggest about speech production?
Competition occurs during lexical/lemma selection before phonological retrieval.
What does phonological facilitation suggest about speech production?
Phonological information is processed after lexical selection and can aid word form retrieval.
How does picture-word interference support the two-stage model of lexicalisation?
Semantic distractors interfere with meaning retrieval (lemma stage), whereas phonological distractors facilitate sound retrieval (lexeme stage).
What is a semantic distractor?
A word related in meaning to the target picture.
What is a phonological distractor?
A word that sounds similar to the target word.
How does the picture-word interference paradigm provide evidence for separate lemma and lexeme stages?
Semantically related distractors cause interference and slower naming, whereas phonologically related distractors facilitate naming. This indicates that meaning-based (lemma) and sound-based (lexeme) processes occur at separate stages of lexical retrieval.
What was the task in Schriefers et al. (1990)?
Participants named a picture while ignoring an auditorily presented distractor word.
Stimulus Onset Asynchrony (SOA)
The time difference between presentation of the distractor word and the picture.
inhibition
Occurs when semantically related distractors (priming) slow picture naming
facilitation
Pccurs when Phonologically related distractors speed up picture naming
How does Schriefers et al. (1990) support a two-stage model of lexicalisation?
Semantic interference occurs first and phonological facilitation occurs later, suggesting separate stages for meaning (lemma) and sound (lexeme).
What evidence from Schriefers et al. (1990) suggests that lemma retrieval precedes lexeme retrieval?
Semantically related distractors caused early interference, whereas phonologically related distractors produced later facilitation, indicating that meaning is processed before word form.
What is the debate between discrete and interactive models of lexicalisation?
Whether meaning and sound are retrieved in separate sequential stages (discrete) or whether activation spreads between stages before selection is complete (interactive).
What does the discrete model propose?
A word's meaning (lemma) is selected first, then its phonological form (lexeme) is retrieved.
How does activation flow in the discrete model?
Semantic → Phonological (one-way, step-by-step).
What is cascading activation?
Activation begins spreading to phonological representations before a single lemma has been fully selected.
What does the cascading model predict?
Multiple candidate words activate their phonological forms simultaneously.
How does activation flow in the cascading model?
semantic → Phonological, but before lemma selection is complete.
What is the feedback model of lexicalisation?
Activation spreads from semantic to phonological levels and then back up again to influence lexical selection.
How does activation flow in the feedback model?
Semantic ↔ Phonological (two-way interaction).
What is the key difference between cascading and feedback models?
Cascading allows activation to spread downwards only; feedback allows activation to spread both downwards and upwards.
What does the feedback model predict about related words?
Activated phonological forms can feed back and strengthen activation of semantically related lemmas.
Which model is most consistent with speech errors involving multiple competing words?
Interactive models (cascading and feedback).
Which model is most consistent with Garrett's original theory?
The discrete model.
Compare discrete and interactive models of lexicalisation.
The discrete model proposes that lemma selection is completed before phonological retrieval begins. Interactive models propose that activation spreads between semantic and phonological levels before selection is complete; in feedback models, activation can also spread back from phonological to semantic levels.
What was the basic task in Levelt et al. (1991)?
Participants named pictures while completing a lexical decision task (LDT) on words presented through headphones.
What is mediated priming?
Activation spreads indirectly from a target concept to phonologically related words through semantic connections.
In the cascaded model, which semantically related word becomes activated when naming "sheep"?
Goat
How does activation spread during cascaded processing?
SHEEP (semantic) → GOAT (semantic neighbour) → /goat/ (phonological form) → related words such as GOAL.
What does the discrete model predict during naming of "sheep"?
Only the phonological form of SHEEP should be activated.
What does the cascaded model predict during naming of "sheep"?
The phonological forms of both SHEEP and semantically related competitors (e.g., GOAT) become activated.
Why is activation of "goal" important evidence for cascaded processing?
Goal" is phonologically related to GOAT rather than SHEEP, suggesting activation spread through multiple levels.
What finding would support a cascaded model over a discrete model?
Priming of words related to semantic competitors (e.g., GOAL activated while naming SHEEP).
How does cascaded processing differ from Garrett's discrete model
Garrett proposed that phonological retrieval begins only after lemma selection, whereas cascaded processing allows phonological activation before selection is complete.
What does mediated priming suggest about lexical retrieval?
Multiple lexical candidates may be partially activated simultaneously.
What did Levelt et al. (1991) find regarding semantic priming?
Early inhibitory priming of a semantic competitor (e.g., goat) at the lemma level.