Stop Place Contrasts before Liquids

1
Stop Place Contrasts before Liquids

• Edward Flemming
• MIT

2
Does phonology have a biological grounding?

• Phonology is shaped by the nature of speech
  perception and speech production.
• Speech production and speech perception depend on
  biological endowments.
• A case study adopting this approach to the
  analysis of a phonological pattern Restrictions
  on stop place contrasts before laterals.

3
Restrictions on coronal-lateral clusters

• Many languages allow initial pl, kl/bl, gl
  clusters, but exclude tl, dl (Kawasaki 1982).
• E.g. English, German, Norwegian, Thai, etc
• English
• b-g, p-k contrast before l, blue-glue,
  plan-clan
• initial dl-, tl- are not possible.

4
Restrictions on coronal-lateral clusters

• This is a result of a dispreference for
  coronal-dorsal stop contrasts before laterals
  (Flemming 1995).
• Some languages reverse the English pattern,
• tl-, dl- OK, but no kl-, gl-.
• e.g. Haroi and other Chamic languages (Mudhenk
  Goschnick 1977), Katu dialects (Wallace 1969)

• Some languages have free variation between
  coronal and velar stops before lateral (but
  contrast elsewhere), e.g. Bolton English
  (Shorrocks 1998), Mong Njua (Lyman 1974)

5
Restrictions on coronal-dorsal contrasts before
l

• Kawasaki (1982) hypothesizes that this
  dispreference is due to perceptual similarity of
  dl-gl, tl-kl.
• General hypothesis There is a preference for
  perceptually distinct contrasts.
• Before l, contrasts between coronal dorsal
  stops are not very distinct.
• Previous evidence
• Kawasaki (1982) Evidence from 1 speaker that
  formant transitions are very similar in dl-,
  gl-.
• But bursts can be sufficient to distinguish stops.

6
Restrictions on coronal-dorsal contrasts before
l

• Further investigate Kawasakis hypothesis through
  acoustic analysis of American English (and
  Hebrew).
• Hypothesis English dl-gl and tl-kl contrasts
  would be less distinct than legal stop place
  contrasts (e.g. bl-gl, before vowels).
• How do we infer the expected realization of dl-,
  tl-?
• Other stop-liquid clusters
• Medial -dl-, -tl- clusters
• tl-, dl- in languages that allow these clusters
  (e.g. Hebrew, Russian).
• But NB languages that allow these contrasts may
  realize stop-liquid clusters in a different
  fashion, e.g. less gestural overlap.
• Focus on voiced clusters (Am. English medial -tl-
  often realized as -?l-).

7
Materials

• 6 near-minimal pairs for bl-gl, e.g. blow, glow
• 9 triplets for b, d, g, each preceding the same
  set of nine vowels.

• Sentence frame Say X to me
• Presented twice in random order
• 5 native speakers of American English, 4 female,
  1 male.
• 6 near-minimal pairs for medial -dl-gl-, e.g.
  Ridley, wrigley
• Same frame, presented twice in random order
• 3 native speakers of American English, all female

8
Cues to stop place contrasts

• Prevocalic stops (e.g. Dorman et al 1977)
• Release burst - transient frication
• Formant transitions
• Examine similar cues in stops preceding l.

formant transitions
burst
9
Measurements

• Formants
• Measured F2 and F3 as soon as possible after the
  end of the burst.
• Burst
• Burst duration - from stop release to onset of
  first formant.
• Shape of spectrum of first 6 ms of the burst.

10
Quantifying burst shape

• Measured from smoothed spectra (Hanson Stevens
  2003)
• Calculate a series of seven DFTs on 3 ms windows
  at 1 ms intervals.
• Average these spectra.

11
Quantifying burst shape

• Burst peak frequency of the highest amplitude in
  the spectrum
• Amid-Ahi (cf. Suchato et al 2005)
• Amid average amplitude from 1.25 kHz - 3 kHz
• Ahi average amplitude from 3.5 kHz - 8 kHz

• bl bursts

mid
high
gl bursts
12
Stop-l clusters - formants

• English stops are not well differentiated by
  formant transitions before l.

• Medial -dl-, -gl-
• Small differences in F2 onsets
• -dl- seems to allow more coarticulatory influence
  of a preceding front vowel.

• Initial bl-, gl-
• No significant differences in formant onsets.

13
Stop-l clusters - formants

• Stops are not well differentiated by formant
  transitions before l.
• Similar results from a preliminary study of
  Hebrew initial bl-, dl-, gl-, pl-, tl-, kl-
  clusters (3 speakers)
• Stops appear to be overlapped with following
  lateral, so formants at onset are largely
  determined by l.
• Effect is particularly striking for English
  -dl-, since prevocalic d is usually
  characterized by relatively high F2 at release
  (2100 Hz)
• Possibly lateral release of d facilitates the
  low F2 onset (1311 Hz)

14
Stop-l clusters - burst

• Initial bl-, gl- clusters are distinguished by
  burst quality and duration.

• gl burst has a higher peak than bl burst.

• gl is more compact than bl.

• gl has a longer burst than bl.
• p

15
Stop-l clusters - burst

• Medial -dl-, -gl- are also differentiated by
  their bursts.
• But the properties of -dl- bursts deviate
  substantially from prevocalic d bursts in the
  direction of g/gl bursts.

16
(No Transcript)
17
Lateral release

• The properties of the dl burst are expected
  consequences of laterally releasing d.
• Centrally released d burst has significant high
  frequency energy because it is filtered by the
  short cavity in front of the alveolar closure.
• But with lateral release, the front cavity
  includes the side passages opened up by lowering
  the sides of the tongue, and thus is
  significantly longer.
• Hence lower frequency peak, more compact burst
  shape.

18
Dispreference for d-g contrast before l

• Stop-l clusters are distinguished by burst not
  formant transitions.
• The bl-gl contrast is plausibly more distinct
  than dl-gl because the burst of laterally
  released d is compact like a velar burst.
• Not clear what leads to a preference for the
  labial-coronal contrast in Chamic etc.
• preference for homorganic clusters?

bl-
-gl-
-dl-
19
Lateral release

• Similarity between dl and gl appears to be a
  result of lateral release of d.
• Lateral release may play a role in unusually low
  F2 onset in dl.
• Lateral release results in non-canonical d
  burst.
• So if d were not laterally released, dl-gl
  contrast should be more distinct.

20

• Thanks to Patrick Jackson Jones

21
Stop-l clusters
(Ba)gley
(ba)dly
22
br vs. dr vs. gr - Formants
p

Formant transitions distinguish br, dr, gr.

23
br vs. dr vs. gr - burst

• burst distinguishes br, dr, gr
• p