Use LEFT and RIGHT arrow keys to navigate between flashcards;
Use UP and DOWN arrow keys to flip the card;
H to show hint;
A reads text to speech;
55 Cards in this Set
- Front
- Back
|
While coarticulated signals help the listener identify the phonemes that the speaker intended to produce, coarticulation is a major factor that makes it difficult to formally analyze the acoustic properties of speech. |
No 1:1 correspondence between acoustic signal and phonemes |
|
|
WEAVER ++ (levelt) |
Conceptual preparation -> lexical concept Lexical selection -> lemma Morphological encoding -> morpheme Phonological encoding -> phonological word Phonetic encoding -> phonetic gestural score Articulation -> sound wave |
|
|
Lexical concepts |
Concepts for which your language has a specific word |
|
|
Motor Theory of Speech Perception |
Proposes that gestures, rather than sounds, represent the fundamental unit of mental representation in speech |
|
|
Motor Theory of Speech Perception |
Proposes that gestures, rather than sounds, represent the fundamental unit of mental representation in speech By knowing what the gestures are, you can tell what the set of words was that produced that set of gestures |
|
|
Per second |
3 words 5 syllables 15 phonemes |
|
|
Basic process in speech production |
Conceptualization - thinking about what we want to say Formulation - choosing the right words to say it (the message). Articulation - moving muscles to make a sound wave that a listener can perceive |
|
|
Per second |
3 words 5 syllables 15 phonemes |
|
|
Basic process in speech production |
Conceptualization - thinking about what we want to say Formulation - choosing the right words to say it (the message). Articulation - moving muscles to make a sound wave that a listener can perceive |
|
|
Lexicalization process |
When you (need to) come up with combinations of lexical concepts to express the idea for which your language does not have a single term |
|
|
Lemmas |
A mental representation that incorporates semantic (meaning) and syntactic (combinatorial) informational |
|
|
Per second |
3 words 5 syllables 15 phonemes |
|
|
Basic process in speech production |
Conceptualization - thinking about what we want to say Formulation - choosing the right words to say it (the message). Articulation - moving muscles to make a sound wave that a listener can perceive |
|
|
Lexicalization process |
When you (need to) come up with combinations of lexical concepts to express the idea for which your language does not have a single term |
|
|
Lemmas |
A mental representation that incorporates semantic (meaning) and syntactic (combinatorial) informational |
|
|
Morphemes |
Smallest unit of language that carries meaning |
|
|
Phonological words |
A set of syllables produced as a single unit (not a word) Ess-core-tuss |
|
|
Per second |
3 words 5 syllables 15 phonemes |
|
|
Basic process in speech production |
Conceptualization - thinking about what we want to say Formulation - choosing the right words to say it (the message). Articulation - moving muscles to make a sound wave that a listener can perceive |
|
|
Lexicalization process |
When you (need to) come up with combinations of lexical concepts to express the idea for which your language does not have a single term |
|
|
Lemmas |
A mental representation that incorporates semantic (meaning) and syntactic (combinatorial) informational |
|
|
Morphemes |
Smallest unit of language that carries meaning |
|
|
Phonological words |
A set of syllables produced as a single unit (not a word) Ess-core-tuss |
|
|
Phonological gestural score |
A detailed map of the phonemes and syllables needed for an utterance, as well as metrical Information, such as stress; used to plan specific motor movements (banana vs. panama) |
|
|
Metric structure |
Indicates the relative emphasis or loudness (accent) that each syllable receives |
|
|
Evidence for stages in Levelt’s speech production model |
Speech errors Tip-of-the-tongue experiences Picture-naming studies |
|
|
Evidence for feedback |
Lexical bias effect: errors are more likely than chance to produce real words instead of random gibberish Phonological exchanges happen more often when the result is two real words. Big feet - Fig beet
Errors respect phonotactic constraints rules about how phonemes can be combined, and they create real words more often than They should purely by chance (slip>tlip) Mixed-errors: the word that a person produces by mistake is related in both meaning and sound to the intended word. Interpreted as evidence for phonological-lemma feedback during lemma selection |
|
|
Potential limitations for lemma theories |
Lemma representations include Information about a word’s grammatical features - how the word can combine with other words. Damage to the lemma would mean that speech errors & writing the same. However this is not the case. Semantic substitution error: people have seperate word-form Information source for spoken and written words. Also, patiënts have different difficulties between content (eg. Cat) words & function (eg. Light) words. |
|
|
Self-repair and Self-monitoring. |
Self-repair happens after error Self-monitoring catches errors before error is produced (pre-output monitoring). Wordt aangenomen dat dit bestaat omdat mensen zichzelf vaak al vrijwel direct verbeteren na een fout Pre-output monitoring pays more attention to possible embarassing outputs. Galvanic skin response (a measure of “emotional stress”) higher under time pressure. Error-detection is greatest when planning load is lightest (end of phrases and clauses) |
|
|
Articulation |
Perturbs air flow to create different patterns of sound waves. Articulation consists of contrastive gestures Each contrastive gesture creates a noticable change in the speech signal (pattern of sound waves) |
|
|
Articulation |
Perturbs air flow to create different patterns of sound waves. Consists of contrastive gestures Each contrastive gesture creates a noticeable change in the speech signal (pattern of sound waves) |
|
|
Articulatory phonology theory |
Speech planning creates a gestural score (program) that tells the articulators how to move 1. Move a particular set of articulators 2. Toward a location in the vocaal tract where a constriction occurs 3. With a specific degree of constriction 4. Occuring in a characteristic dynamic manner |
|
|
Articulator movement produces phonemes (basic speech sounds). Phonemes can be classified according to: |
Place of articulation Manner of articulation Voicing |
|
|
Co-articulation |
Gestures for adjacent phonemes overlap in time Gestures are influenced by preceding and following phonemes (pool vs pan) |
|
|
3 important problems |
Segmentation problem: speech is sticky - there are no spaces between words in running speech - and the pauses that are there are not in the right place Invariance problem: speech sounds are not stable - different conditions - different speakers - coarticulation Flow of Information - top-down/bottom-up, interactive, serial or parallel |
|
|
Embedded words |
Words within and across words. (Key in Donkey) |
|
|
Speech error types (errors are not random) |
Semantic substitution errors Sound Exchange errors Word Exchange errors |
|
|
Semantic substitution errors |
A word has been changed in the order or in something else. “Left my keys on the chair instead of the table”. Something goes wrong in conceptual preparation & lexical selection |
|
|
Sound exchange |
Replacing 2 words by 2 other words (one gestural score) Most involve within the same phrase Most involve a single pair of phonemes Errors are more common after phonological primes “fat beer, fun bed, far base” Target: “big feet” Most respect positional constraints. Onsets exchange onsets. Codas onsets exchange codas Because planning unit = syllable; positions within syllabled are marked Spoonerism is an error in speech or deliberate play on words in which corresponding. “Three cheers for our queer old dean!”
BIG FEET Two syllable frames Each marked with order tag Activated all simultaneously Can cause error when close enough |
|
|
Word exchange |
Word substitution but not semantic. “My girlfriend plays the piano” Error in conceptual preparation, lexical selection & phonetic encoding. Obeys the category constraint; nouns Exchange nouns, verbs verbs etc. Rarely otherwise Reflects clause-level planning. |
|
|
Onset, nucleus, codas |
Onset = initial sound of a word Nucleus = vowels Codas = final sound of the word
Onsets have stronger effect than coda. |
|
|
Tip-of-Tongue experiences |
When you have the correct lemma but not the correct phonemes. Error in morphological encoding, phonological encoding/syllabification Become more frequent when getting older Not a conceptual problem. Evidence: picture naming. No matter how frequent a concept is spoken, each concept is recognized as fast as each other. Research: Diary methods Prospecting: asking people questions of words that are hard to find |
|
|
During TOT experience, people: |
- accurately predict whether they will come up with the correct word soon - report the correct number of syllables - accurately report the first phoneme - are more accurate about the beginning and end phonemes than the middle - report words that sound like the target - have more TOT’s for less frequent words - resolve about 40% of TOT’s within a few seconds to a few minutes |
|
|
Picture-naming |
Picture-naming depends on frequency of spoken. Picture-word-interference paradigm: naming a picture with a word on top Identity condition: naming a picture with a word on top. Identify condition is the fastest because word and picture point towards same concept Semantic condition: when have to say Apple when image of lemon -> interference between lexical concepts. Identify condition is slow because word and picture interfere at the concept selection stage Phonological condition: a word refers to an object whose name is similar to the object of the picture. House/mouse. This condition is fast because word and picture don’t interfere at the concept selection stage. |
|
|
Difference WEAVER ++ and Spreading Activation |
Weaver = discrete selection has to take place before activation at the next level starts Spreading = cascading activation spreads throughout the system immediately. |
|
|
Coarticulation produces redundancy |
Individual segments provide clues about preceding and following segments Silent center vowels “b_g perceived as bag” Franken-words (cross-splicing): “Jo + b. Perceived as jog. Onsets are stronger than codas” |
|
|
Coarticulation produces variability |
No 1:1 correspondence between acoustic signal and phoneme |
|
|
Properties of sound waves |
Each episode of compression and rarefraction constitutes a cycle The amount of time during a cycle determines the frequency (Hz; pitch). Amount of energy determines the amplitude (Db; loudness) Speech consists of a complex mixture of sound waves: each component had its own frequency and amplitude |
|
|
Gestures (movement of articulators) represent the fundamental unit of mental representation in speech. Steps: |
Which gestures created a speech signal? What was the gestural plan? This will take you back to the sequence of syllables or words that when into the gestural plan in the first place Now you know what the set of words was that produced that set of gestures |
|
|
McGurk Effect |
Example of multi-modal speech perception. When both auditory and visual, people perceive a blend of these gestures. Ready explained by motor theory of speech perception |
|
|
Relationship between gesture and phoneme closer than between acoustic signal and phoneme |
Speech is Special (dedicated module) Speech production and perception are tightly linked Explains categorical perception Explains the mcgurk effect |
|
|
Problems with Motor Theory of Speech Perception |
Theory makes a strong connection between perceiving and production (can only understand the articulation movements correct if you’re able to produce them yourself) but infants are able to perceive as well. Brain-damaged patients can sometimes produce fluent speech but are not able to perceive it. |
|
|
General auditory approach to speech perception (GA) |
States that speech perception is not special but uses general (learning) mechanisms and states that speech perception and perceiving is a seperate function. This theory therefore not vulnerable to many criticisms that apply to the motor theory. |
|
|
“Top-down” info favors the real word over the non-word (Ganong effect) and restores missing phonemes (phonemic restoriation effect). |
“Top-down” info favors the real word over the non-word (Ganong effect) and restores missing phonemes (phonemic restoriation effect). |
|
|
The Ganong effects (top-down processing) |
the tendency to perceive ambiguous speech stimuli as real words if possible. Your lexical knowledge influences your phoneme perception - does word context affect how phonemes are perceived? -Dash-tash with small increments of VOT - Low VOT: first consonant sounds more like /D/, hight VOT more like /t/ Perception of phonemes is biased towards real words (dash or task) Lexical knowledge influences phoneme perception |
