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;
47 Cards in this Set
- Front
- Back
Glottal wave
|
like guitar string, vocal cords vibrate, they have a resonance frequency
|
|
Filter
|
As the glottal wave passes through a vocal tract, it forms a resonance chamber and forms resonating frequencies
|
|
Perceptual Property
|
Quality, loudness, pitch
|
|
Acoustic Property
|
Formants, amplitude/intensity, fundamental frequency
|
|
How is a vowel created?
|
Source+filter= vowel
|
|
Formants
|
- Product of resonant frequencies plus filters
-different sets of resonant frequencies - Some resonant frequencies have higher energy than others |
|
Spectrogram
|
- Visual Represenation of frequencies over time
|
|
Difference btween F1 and F2 in spectrogram
|
- It indicates backness
- a large gaps means a front vowel - a small gap means a back vowel |
|
What does F1 tell you
|
there is an inverse relationship between height of vowel and height of the first formant
|
|
How can you tell lip rounding in spectrograms
|
It lowers the F2, F3, and higher
|
|
How can you tell dipthongs
|
They have a changing formant structure
|
|
Fundamental frequency is a formant: true or false?
|
false
|
|
How to tell if there is voicing in a spectrogram
|
- there is a voice bar
|
|
How to tell if there is aspiration in a spectrogram
|
- there is a period of "noise" after burst
|
|
How to tell Alveolars in a spectrogram [t] and [d]
|
Tail of F2 will point to 1700-1800 hertz
|
|
How to tell bilabial [b], [p]
|
Tails of f1 and F2 go down (they rise before the sound and then fall after the sound)
|
|
Accent of Korean or Japanese
|
[upside down r] and [l] are often confused
|
|
Co-articulation
|
The overlapping of adjacent articulations, usually in the form of the vowel overpowering the articulation of the consonant
- Ohman's discovery- coarticulation of a vowel |
|
The three challenges of speech perception
|
1. Speaker variability
2. Segment variability 3. Segmenting the speech stream into words |
|
Speaker variability
|
1. Variability between speakers
- fundamental frequencies vary between speakers - speaking rate - accents between speakers 2. Variability within speakers - Emotional state - speaking rate - Possition of word within a sentence |
|
Segment Variability
|
1. Stress
2. Co-articulation |
|
Non-variation (things that dont change)
|
Vowels: Relative formant levels
Consonants: importance of formant transitions |
|
Word Segmentation
|
How we segenment speech streams into words
1. Largely a guessing games 2. Segmentation strategies |
|
Word Segmentation errors
|
Mondegreens- word segmentation
|
|
Segmentaiton strategies
|
1. Starting at the end of utterances
2. Using words you already know 3. Looking for stressed syllables 4. Use phonotactics 5. Use experience (the last thing) |
|
How we learn segmenation
|
1. patterns (linkind to the end of utterances)
2. Statical patterns 3. Stress |
|
Preferencial listening experiment (can babies tell words from non-words)
|
- The baby uses statical patterns to tell if a word a non-sense word or a real word
- Dependent variable- how long the baby look at a light - "novelity prerence for new words" |
|
Forced Choice identification
|
we simply ask the participant to distinguish different sounds which are different VOT times, and “what did you hear”… a very simple concept… it doesn’t work on infants obviously
|
|
Categorical perception over time
|
1. As soon as we are 1 month old we are able to detect VOT
|
|
Three categorical perception tests in babies
|
1. High Amplitude sucking (dihabuation of the same sound)
2. Preferencial listening in visual domain (we loose sensitivity to visually recognize different languages after 8 months) 3. Conditioned Head turn (conditioned a kid to turn his head when they turned to Ba, shows that babies can understand categorical perception) |
|
Categorical perception in Animals
|
- Chinchillas detect +30 boundary
- Crickets use chirp frequencies (high frequencies indicate threat) |
|
Losing Categories
|
- Infants are able to distinguish nearly every phonemic contrast they have been tested
- They loose this ability after 12 months |
|
Language in British Columbia Thompson Salish (Nthlakampx)
|
[k'] ‐ air compressed between tongue/velum and glottis, [q'] ‐ air compressed between tongue/uvula and glottis
|
|
Position of Vowel phonemes is dictated by
|
1. Maximize perceptibility
2.. Articulatory dispersion correlates with acoustic dispersion 3. Confounded by ease of articulation |
|
Margi
|
- Two vowel phomenes
- They are not at the extremes, they central vowels because they are distinguishable but easy to pronounce |
|
Hyper-articulation
|
In babies led speech we hyper-articulate speech /i/, /a/, /u/, in order to assist in the aquision of vowel phonemes
|
|
Great Vowel shift
|
A shifted first, then all vowels seemed to move upward
|
|
Middle english
|
1100-1500
|
|
Modern English
|
1500 on
|
|
Broca's area
|
Frontal lobe
- controls the movement of the mouth - fragmented speech - Aware of deficit |
|
Wernickies area
|
temporal lobe
- controls language comprehension - fluent but empty speech, poor comprehension |
|
Phonetics focuses on
|
spelling to phonology to meaning
|
|
Whole word/whole language
|
(skips phonology)
- goes from spelling to meaning |
|
Evidence in phonology in reading
|
Asked if it is a flower (rose vs rows) sometimes people said rows meant flowers… based on how the word sounds
|
|
Common mistakes by spanish influenced English
|
- Pronouncing vowel phonemes as if in spanosh
- Dipthongs to mono-thongs - No TH sounds, no v sound, no /v/ sound - Voiced fricative to stop shift - Fricatives to stops - Palatal consonant shift |
|
English Influenced Spanish
|
- aspiration of voiceless stops
- vowel reduction in unstressed syllables - diphthongization of mid vowel |
|
How to tell consonant stops in a spectrogram
|
- Constriction: look at formant transitions for F1 – it rises out of a constriction and falls into one
- Alveolars- [t],[d] – tail of f2 will point to 1700-1800 hertz - bilabial- [b],[p] – the tails of f1 and f2 go down (rise before the sound, go down after) - Velar stop consonant (the velar pinch) [k] |