Task 9: Speech Perception Flashcards
What produces speech sounds?
position/movement of structures in vocal apparatus
Acoustic signal
pressure changes in air
Respiration
To generate an acoustic signal, air is pushed up from lungs past vocal cords and into vocal tract
Phonation
Process through which vocal folds are made to vibrate when air pushes out of lungs. In fact, sound depends on shape of vocal tract as air is pushed through.
Articulation
Shape of vocal tract is altered by moving articulators
Articulators
Tongue, lips, teeth, jaw and soft palate.
Resonance characteristics
Changing size and shape of space through which sound passes increases and decreases energy at different frequencies
How are vowels produced?
- Produced by vibration of vocal cords.
- Specific sounds of each cord are created by changing shape of vocal tract. By changing it, resonant frequency of vocal tract changes, producing peaks of pressure at different frequencies, called formants (F1 formant has highest frequency, F2 has next highest and so on)
How are consonants produced?
- Produced by closing of vocal tract.
- Movement of tongue, lips and other articulators create patterns of energy in acoustic signal. Rapid shifts in frequency before or after formants (vowels) are called formant transitions and are associated with consonants (T1, T2).
Sound spectrogram
Three-dimensional display that plots time on horizontal axis, frequency on vertical axis and amplitude (intensity) on color, with redder showing greater intensity, or gray scale.
- indicates patterns of frequencies and intensities that make up acoustic signal
Phoneme
the shortest segment of speech that, if changed, changes meaning of word.
What kind of relationship is between a phoneme and acoustic signal?
variable
Coarticulation
the articulation of two or more speech sounds together, so that one influences the other.
Perceptual constancy
We perceive sound of phoneme as same even if acoustic signal is changed by coarticulation.
Discuss variability from different speakers x2
- Individual differences – Some voices are high-pitched, other low-pitches, some talk rapidly, others slowly.
- Sloppy pronunciation – When in conversational speech, people sometimes do not articulate each word individually.
Name 3 ways speech perception systems deals with this variability problem in different ways (perceiving phonemes)
- categorical perception
- information provided by face/audiovisual speech perception
- information from our knowledge of language
Categorical perception (perceiving phonemes)
For speech and other complex sounds and images, it occurs when stimuli that exist along continuum are perceived as divided into discrete categories.
Voice onset time (categorical perception)
Time delay between when sound begins and when vocal cords begin vibrating.
Example: /da/ has a short VOT and /ta/ has a long.
Phonetic boundary (categorical perception)
Cross-over point from perception of one phoneme to another, depending on VOT.
Name and explain 2 qualities defining categorical perception
- Sharp labeling function – Small changes to simple acoustic stimuli (pure tones) leads to gradual changes in people’s perception of these stimuli.
- > when VOT goes above 35ms, perception from /da/ changes to /ta/. - Discontinuous discrimination performance – Two stimuli with VOTs on same side of phonetic boundary (25ms) are judged to be same (= perceptual constancy), whereas two stimuli on different sides of phonetic boundary are judged to be different.
- > If we did not have perceptual constancy we would perceive different sounds every time we changed VOT.
Information provided by face/audiovisual speech perception (perceiving phonemes)
Speech perception is multimodal, meaning that our perception of speech can be influenced by information from a number of different senses.
McGurk Effect
Effect illustrating that although auditory information is major source of information for speech perception, visual information can also exert strong influence on what we hear (= audiovisual speech perception).
When someone perceives someone talking i.e. their face what area is also activated
FFA
Information from our knowledge of language
Research showed that it is easier to perceive phonemes that appear in meaningful context.
Bottom up processing of speech perception
nature of acoustic signal
Top down processing of speech perception
context that produces expectations in listener
Phonemic restoration effect
effect in which sounds missing from speech can be restored by the brain and appear to be heard
- more likely to work for higher frequencies -> the mask must have a similar frequency for us to fill in the blank correctly
What does the phonemic restoration effect show?
shows that speech perception can be determined by:
1 bottom up processing
2 top down processing
Discuss perceiving words in sentences
Words are more intelligible when heard in context of grammatical sentence than when presented as items in list of unconnected words
Speech segmentation
Perception of individual words in a conversation
What does speech segmentation prove?
our perception of words is not only based on energy stimulating receptors, but also on knowledge of meaning of words
Transitional probabilities
Chances that one sound will follow another sound
Statistical learning
- when acquired?
Process of learning about transitional probabilities and about other characteristics of language
- acquired as soon as 8 months
Indexical characteristics
carry information about speakers such as age, gender, place of origin, emotional state and whether they are being sarcastic or serious
- listeners take in 2 levels of information about words: its meaning and the characteristics of speaker’s voice
Aphasia
language problems due to damage
Broca’s aphasia
Condition in which people have difficulty speaking but are capable of comprehending what others are saying. It results from damage to Broca’s area in frontal lobe.
Wernicke’s area
Condition in which people can speak fluently but what they say is disorganized and not meaningful, plus they have great difficulty understanding what other people are saying. Results from damage to Wernicke’s area in temporal lobe -> what pathway for recognizing speech.
Word deafness
Extreme case of Wernicke’s aphasia in which people cannot recognize words even if they can still hear pure tones.
Voice area
superior temporal sulcus, activates more for human voices than other sounds
Voice cells
temporal lobe, respond more to “voice” sounds than “non-voice” sounds
Ventral (what) pathway
temporal lobe, recognizing speech
Dorsal (where) pathway
parietal love, responsible for linking acoustic signal to movements used to produce speech
Are the sound waves more like yanny or laural?
laurel
Do younger ears hear laurel or yanny more often?
yanny
The motor theory of speech perception
the hypothesis that people perceive spoken words by identifying the vocal tract gestures with which they are pronounced rather than by identifying the sound patterns that speech generates (activity of motor cortex can influence speech perception) – however theory is highly criticized