Exam 3 Flashcards
acoustic correlates of vowels
- lower F1, higher jaw height
- higher F2, more fronted vowel
- F1 correlates with jaw height–>inverse relationship
- F2 correlates with tongue frontness/backness
- lip rounding lowers formant frequencies
vowel space chart
- F1 is on ordinate (y-axis)
- F2 is on abscissa (x-axis)
- both scales in reverse order to represent the tongue positions more intuitively (scale is high-to-low)
the “schwa trick”
- to determine formant frequency ranges for any given speaker, map that person’s mid-central vowel and use it as a frame of reference for the rest of the words
- an avg adults schwa will have F1=500Hz, F2=1500Hz, and F3=2500 Hz
wideband spectrogram
provides info about articulation
narrowband spectrogram
gives info on phonation
acoustic correlates of consonants
oral stops, fricatives, affricates, nasal stops, approximates
manner of voiced stops
- evidence of onset closure: abrupt drop in amplitude on waveform; sudden loss of sound tracings in frequencies above F0 on spectrogram
- presence of complex periodic sound during stop gap on waveform and of voicing bar in spectrogram
- abrupt release of closure with soft-to-moderate burst (may be visible in waveform and spectrogram)
place of voiced stops
- formant transitions (especially F2 & F3) provide clues for place of articulation
- F2 dips down from vowel for labial closure [b]
- F2 is level for tongue-front closures [d]
- F2 rises and F3 dips down (velar pinch) for dorsal closure [g]
manner of unvoiced, aspirated stops
- evidence of onset closure: abrupt drop in amplitude on waveform; sudden loss of all sound tracings on spectrogram
- no sound during stop gap on waveform and spectrogram
- abrupt release of closure with moderate-to-loud burst (usually visible in waveform and spectrogram)
- presence of aspiration noise on waveform and spectrogram following burst (+VOT)
place of unvoiced, aspirated stops
- formant transitions (F2 and F3) provide clues
- notice that formant transitions vary depending on which vowels and consonants are adjacent
- aspiration may obscure formant transitions at onset of vowel
manner of voiced & unvoiced fricatives
- more amplitude than stops
- waveform, if expanded, will show complex aperiodic sound; spectrogram will show scratchy noise tracings
- voiced fricatives will show periodicity on waveform, and both voicing bar and glottal pulses on spectrogram
- durations may vary considerably; voiced fricatives tend to be longer
place of voiced and unvoiced fricatives
- most of the energy in fricatives is in the higher freq ranges; without a sufficiently large frequency scale, the place differences are hard to interpret
- the alveolar [s z] and alveopalatal [ʃ ʒ] fricatives tend to have louder frication, they are called sibilant
sibilant
the alveolar [s z] and alveopalatal [ʃ ʒ] fricatives that tend to have louder frication
voiced and unvoiced affricates
- stop gap (with or without voicing) like an alveolar stop
- alveopalatal fricative offset (with or without voicing)
- looks similar to [dʰ] [tʰ] except that the frication tends to be longer and is almost always more sibilant in affricates [dʒ] [tʃ]
nasal stop
- abrupt onset and offset like oral stops, but without a burst
- voicing bar + first nasal formant (N1) = nasal murmur
- more formants than oral sounds–often low-intensity and not visible
- antiformants: space between formants
- place clues from adjacent vowel transitions
approximates (liquids & glides)
- formant structure like vowels but with lower amplitude; greater amplitude than other consonants
- larger transitions than vowels
- [w] has F1 & F2 similar to [ʊ]
- [r] has significant drop in F3 (can dip below 2000 Hz)
- [l] may show F1 “step” transition
- [j] has F1 & F2 similar to [ɪ]
spectrogram reading steps
- segmentation
- acoustic description
- articulatory description
- transcription
- stress analysis
segmentation
look for boundaries between sounds
acoustic description
describe the acoustic feature of each segment (voicing bar, glottal pulses, acoustic energy)
articulatory description
what can you infer about articulation from the acoustic characteristics? (vowel, non-high, high, non-front)
transcription
for each segment, can you narrow the transcription symbols down to a few possible candidates?
stress analysis
pitch contour (look for prosodic clues using the pitch and energy contours to determine the stress patterns)
phonetics
the study of speech sounds, their production, perception, and physical properties
phonology
the study of the distinctive sounds and characteristic patterns of a spoken language
types of transcription
- broad vs narrow
- impressionistic vs systemic
- phonetic vs phonemic
broad transcription
doesn’t show a great detail of phonetic detail–usually just main symbols
ex: phonemic transcription
narrow transcription
captures pronunciation in great detail using diacritics
impressionistic transcription
transcription of speech that is unknown to you
you don’t know the language or phonological rules
systemic transcription
transcription that knowingly represents the regularities of a language’s unique phonology
ex: dogs –> [dogz]
phonetic transcription
transcription of the actual pronunciation of a word
may be broad or narrow, systemic or impressionistic
indicated by [ ]
phonemic transcription
a broad transcription of the underlying phonemes of a word (as it might be said in exaggerated citation-form speech, with no syllable carrying more stress than any other
indicated by / /
minimal pairs
two words, with different meanings, that sound identical except for one sound
ex: [kæt] vs [pæt], [k] & [p] are phonemes
words (minimal sets) often help determine the phonemes of a language)
phonological patterns at segmental level
- VOT
- glottal substitution: [garʔn] instead of [garden]
- nasal & lateral position: [sænd] vs [sæ.dn̩]
- flapping: neutralization of [d]/[t] (quarɾer]
- velarization of nasals
- nasalization of vowels in nasal contexts: [mæ̃n]
- rhotacization of neutral vowels: [ɚ]
- vowel lengthening: lid vs lit
- vowel reduction
nasal & lateral plosion (closure)
nasals and laterals following a homorganic stop may become a plosive
ex: [m̩ n̩ ŋ̩ l̩]
velarization of nasals
nasals that precede (come before) velars become nasals
ex: [ræn] [ræn̬] [ræn̬k]
velar pinch on spectrogram
homorganic sound
sounds made in the same place of articulation
[p] [b] [m]
[s] [t] [r] [l] [n] [d]
rhoicization of neutral vowels
when “r-coloring” occurs with a neutral vowel, the entire vowel becomes rhoticiized
ex: [bɪrd] (beard) – [bɚd] (bird)
r-coloring
occurs when the /r/ is pronounced after vowels in the same syllable
ar, er, ir, or sounds in words
vowel reduction
vowels in unstressed syllables immediately adjacent to stressed syllables may be reduced
ex: demonstrate
suprasegmental patterns
- intonation contours
- stress pattern
- tonic syllables
intonation contours
english has a rising intonation pattern on yes/no interrogative questions and a falling intonation on most other phrases
stress patterns
english marks the stressed syllable lengthening, syllable loudness, and/or rising pitch (F0)
tonic syllables
when stress is marked with rising pitch, especially at the level of the phrase–>that syllable is called the “tonic syllable”
- english is a non-tonal language
- variations of “ma” in mandarin
