Vwl Systems

Josef Fruehwald

October 24, 2014

Intro

Questions and topics

  1. What happens to /uw/, /ow/ and /aw/ before /l/?
  2. How does this interact with these vowels fronting?
  3. What does this tell us about phonology and phonetic changes?
  4. Assorted loose ends.

What happens to /uw/, /ow/ and /aw/ before /l/?

“What’s really happening to short-a before L in Philadelphia?” (Dinkin, 2013)

  • There is a close phonetic similarity of /awl/, /æl/ and /aw/ (Powel, pal, pow)
  • /awl/ → [æ:l] (Tucker 1944)
  • Apparent tensing of /æl/ is due to it merger to either /awl/ or /aw/

Making it more general

Both /owl/ and /uwl/ are blocked from fronting in Philadelphia

  • Argument 1:
    This is because they both also undergo monophthongization
  • Argument 2:
    This is one phonological rule
  • Argument 3:
    The phonological features selecting /aw/, /ow/ and /uw/ to undergo monopthongization before /l/ are the same that select them to all front

The data

plot of chunk unnamed-chunk-3

Generalizing

plot of chunk unnamed-chunk-4

Monophtongization

Vowel Trajectories

Monophthongization

Proposal:There is a generalized monophthongization process in the grammar.

  • Vwl → V:l
figures/vwl_notext.svg

Phonetic Parallism

Parallelism

plot of chunk unnamed-chunk-5

Parallism

  • Linguistic or social parallism? (Watt 2000)

Parallelism

Random Effects

plot of chunk unnamed-chunk-6

Linguistc & Social Factors

plot of chunk unnamed-chunk-7

Linguistic & Social Factors

plot of chunk unnamed-chunk-10

Linguistic & Social Factors

plot of chunk unnamed-chunk-11

Tying them together

“Features”

Set defining labels

figures/features.svg
  • Phonology: Pick out members to undergo a phonological process
    • Here: Monophthongization
  • Phonetics: Defines a target phonetic realization
    • Here, the diachronic fronting

Monophthongization and Fronting

vw_front
vwl_block

Nucleus movement

plot of chunk unnamed-chunk-13

Nucleus Movement

plot of chunk unnamed-chunk-14

Non-complications and Complications

/uw/, /ow/ and /aw/ don’t always all front

Seems like there’s an implicational scale:

  • if aw
    • then ow and uw
  • if ow
    • then uw
  • uw can front alone.

Phonological Natural Classes

figures/cross_back.svg
  • I would predict it would be impossible for just /ow/ to front, or only /ow/ not to front.

A real problem

A dialect where only /aw/ fronts, or only /aw/ and /ow/ front, but not /uw/ should be possible, but not attested?

  • But maybe Philadelphian men fit the bill.

/uw/ fronts before /l/ in Manchester

  1. Maybe /l/ doesn’t have the same phonological effect on Manchester /uw/
  2. The phonology → phonetics mapping that’s changing in Manchester might be different from Philadelphia

No gull/goal or pull/pool merger

  1. Al~owl is pretty near-mergery anyway.
  2. It seems like /awl/ → [æ:l] potentiated merger of /æl/ to [æ:l]

Conclusions

Conclusions

  • It’s good to take a systematizing eye to phonetic changes.
  • Phonetic change isn’t commonly taken as evidence for how the phonology → mapping operates, but it’s a rich vein of evidence.