PHONOLOGICAL TEAMWORK IN LAAL ROUNDING...

PHONOLOGICAL TEAMWORK IN LAAL ROUNDING HARMONY: AN ABC ANALYSIS ABC Conference, UC Berkeley, 18-19 May 2014

Florian Lionnet UC Berkeley

UC Berkeley Phonology Lab Annual Report: ABC↔Conference

223

Introduction

  “phonological teamwork”  i.e. processes involving more than one necessary trigger

  Specifically: multiple-trigger assimilation involving subphonemic/subfeatural cumulative effects

2


224

Introduction

  Similar to Cantonese inter-coronal fronting (Flemming 1997, 2001) *TuT > TyT (morpheme structure constraint)  kʰyt ‘decide’ kʰut ‘bracket’   tʰuk ’bald head’   tʰyt ‘take off’ *tʰut

3


225

Introduction

  Flemming’s (1997: 11)

“cases of doubly-conditioned assimilation (…) provide evidence that coarticulation is relevant to uncontroversially phonological processes, and therefore must be represented in the phonology.” (cf. also Steriade’s (2009) P-map)

4


226

Introduction

IN THIS TALK:   CASE STUDY:

 Rare doubly triggered rounding harmony in Laal  Laal: isolate, Chad (ca. 750 speakers)

  ANALYSIS:  Agreement by Correspondence (ABC, modified)

  I WILL SHOW THAT:  There is phonology below the phonological feature  ABC can handle this subfeatural phonology  ABC can handle local effects of assimilation as well as long-distance assimilation

5


227

1. Laal doubly triggered rounding harmony


228


  Words are maximally disyllabic:

C V1 (C) . C V2 (C) hi i ü ɨ u i -- ɨ u

mid e üo ə o e -- ə olo ia

(<ɛ)üa a ua

(<ɔ)-- -- a --

+Front -Front +Front -Front

V1 V2

7


229


Double trigger[labial] [round]or or

C1 V1 (C2) C3 V2

[α height, -front]

⇒ V1>[rd]

p, b, ɓmb, m, w

oooooooooooooooooooooor]rrrrrrrrrrrrrrrrrrrrrr ooooorrrrrrrr

1 2

eighhhhhhttttt

3

ffront]tttt

r

hi i ü u i -- ɨmid e üo o e -- ə

lo ia (<ɛ)

üa a ua (<ɔ)

-- -- a --

+Front -Front +Front -Front

V1 V2

[α he[α he

ü u

üooo o

üa aaaaaaaaa u

ɨ ə ə

a

uo

DD 8


230


1) Lab, V2[rd], Height, -Front > Rounding:

a.  /ɓᵼr+-ú/ > ɓùr-ú ‘hook-pl’

b.  /wəər+-ó/ > wòòr-ó ‘mongoose-pl’

c.  /təb+-ó/ > tòb-ó ‘fish(sp.)-pl’

d.  /cᵼrm-+-ú/ > cúrm-ú ‘tree(sp.)-pl’

e.  /pəb+-ó/ > pób-ó ‘cobra-pl’

9


231


2) No Rounding:

a.  /gōbər/ > gōbər ‘cloud’ (*gōbōr)

b.  /məəm-ər/> məəm-ər ‘my gd-mother’ *V2[rd]

c.  /gᵼn+-ù/ > gᵼn-ù ‘net-pl’ *Lab

d.  /ɓər+-ú/ > ɓər-ú ‘plant.sp-pl’ *Height

e.  /bìrú / > bìrú ‘burn’ *-Front

10


232


  Note: the doubly triggered rounding harmony is only attested in stratum 1:

 Root-internal (MSC)

 Root + number marking suffixes (mainly nouns + a few verbs)

11


233


C1 V1 (C2) C3 V2

[round]

12

Spread [round] from V2 to V1: •  if V1 and V2 = [α height, -front] •  if C1, C2, and/or C3 = [labial]

Mere descriptive generalization, not explanatory.


234

2. Subphonemic similarity threshold


235


[labial] [round]or or

C1 V1 (C2) C3 V2

[α height, -front]

⇒ V1>[rd]oooooooooooooooooooooor

]rrrrrrrrrrrrrrrrrrrrrr ooooorrrrrrrr

1 2

[α heighhhhhhhtttttt

3

ffront]ttttt

14

V1 needs the coarticulatory rounding of the neighboring consonant to become a target of Rounding Harmony


236


Similarity: V1 V2

First threshold: Amount of rounding coarticulation

/ɓ ᵼ r ú/ ɓùruSubphonemic: V2 = [round] V1 = non-round (subphonemically rounded)

[αheight, -front]Second threshold: Rounding similarity between V1 and V2

15


237


Similarity: V1 V2

ɓùru /ɓ ᵼ r ú/ Subphonemic: V1 = non-round (coarticulatory

rounding) V2 = [round]

[αheight, -front]Second threshold: V1 and V2 must be similar enough in

rounding

gggggggghhhhhhhhtttttttt, -fro[[[[[[[[αheig

16


238


Similarity: V1 V2

Threshold: V1 must be rounded enough (cumulative

coarticulation)

/ɓ ᵼ r ú/ ɓùruSubphonemic: V1 = non-round (subphonemically

rounded) V2 = [round]

[αheight, -front]Second threshold: V1 and V2 must be similar enough in

rounding

17


239


CV coarticulation: / / + C[+labial]

I = ɨ°I = ɨ + Lab

Measurements =center of vowel, far from transitions

18


240


  Second threshold: rounding similarity (qualitative)

/ ɓ ᵼ r ú / [ɓùrú]

[αheight, -front]

19


241


  ‘ostriches’: two alternate forms:

a. /bəəg-ú/ bəəgú b. /bəəg-ó/ bóógó

20


242

  Rounding enhancement and rounding similarity

“worst” round V


“best” round V

(cf. Kaun 2004, Terbeek 1977, Linker 1982)

21


243

  Subphonemic rounding similarity

“worst” round V


“best” round V

22


244

  Rounding similarity scale

2.3 Rounding Similarity Scale

Degree ofsimilarity Property Similar to [u]

Phonemic level 5 [rd] u

Subphonemiclevel

4 rd/[-front]/height ɨᵒ3 rd/height ɨᵒ, iᵒ2 rd/[-front] ɨᵒ, əᵒ, aᵒ1 rd V°

23


245

  Rounding similarity scale (threshold = level 4)

2.3 Rounding Similarity Scale

Degree ofsimilarity Property Similar to [u]

Phonemic level 5 [rd] u

Subphonemiclevel

4 rd/[-front]/height ɨᵒ3 rd/height ɨᵒ, iᵒ2 rd/[-front] ɨᵒ, əᵒ, aᵒ1 rd V°

24


246

  Level 5 acts like a magnet

Rounding Similarity Scale

Similarity a. /gᵼn-ù/ b. /məəg-u/ c./bìru/ d. /ɓᵼr-ú/scale g ɨ n u m əə g u b i r u ɓ u r u5 [round] u4 rd +[-fr]/height ɨᵒ3 rd +height iᵒ2 rd +[-fr] əəᵒ1 rd

<1 rd ɨ

25


247

3. Agreement by Correspondence


248

Main insights: agreement is driven by a similarity threshold effect

  Similarity: Harmony = agreement between segments in a correspondence relation based on similarity.   Threshold: unstable correspondence (Inkelas and Shih 2013b):

 some segments are similar enough to interact,  but too uncomfortably similar to co-exist;  two possible repairs: harmony and disharmony

3 Agreement by Correspondence 27


249

ABC is used for:

  Long-distance C agreement (Hansson 2001; Rose & Walker 2004)

  V Harmony (Sasa 2009; Walker 2009; Rhodes 2012)

  Long-distance C Dissimilation (Bennett 2013)

  Local effects of assimilation/dissimilation

(Inkelas & Shih 2013a,b; Shih 2013)

3 Agreement by Correspondence 28


250

3.1CORR-XX and IDENT-XX redefined

  CORR-XX: establishes correspondence between phonologically similar segments (i.e. shared phonological features).

  IDENT-XX: enforces identity in a particular feature between two segments in correspondence

29


251

IDENT-XX[F]Segments in an output string agree in the phonological feature [F] 1) if they are in the correspondence relation defined as CORR-XX(P≥n), 2) and if [F] ⊂ P.

CORR-XX (P≥n)Segments in an output string are in correspondence if their

similarity in the ponetic property P is at least n on the P-similarity scale.

[F] ⊂ P


re inP is ae.

is

Tree in the ph

30


252


  Laal Rounding Correspondence Hierarchy.

Similarity scale Correspondence hierarchy Degree 4: rd/[-front]/height CORR-VV(rd≥4) Degree 3: rd/height ⟩⟩ CORR-VV(rd≥3) Degree 2: rd/[-front] ⟩⟩ CORR-VV(rd≥2) Degree 1: rd ⟩⟩ CORR-VV(rd≥1)

31


253


  CORR-XX and IDENT-XX are co-dependent

IDENTα-VV[RD] CORRα-VV(RD≥4)

CORR-VV(RD≥3)

CORR-VV(RD≥2)

CORR-VV(RD≥1)

32


254

3.2 ABC ANALYSIS

  Analysis: /ɓᵼr-ú/ > ɓùrú ‘fish hooks’

/ɓᵼr-ú/ IDENT-IO[RD]

CORRα-VV(RD≥4)

IDENTα-VV[RD]

IDENT-OI[RD]

a. ɓᵼᵒr-ú *! b. ɓùr-u *! * c. ɓᵼᵒαr-úα *!

d. ɓuαr-úα * e. ɓᵼᵒr-ᵼ *!

33


255

3.2 ABC ANALYSIS

  Analysis: /mèn-ù/ > mènù ‘hoes’

/mèn-u / IDENT-IO[RD]

CORRα-VV(RD≥4)

IDENTα-VV[RD]

IDENT-OI[RD]

a. mèᵒn-ù b. müòn-u *! c. mèᵒn-ᵼ *!

34


256

4. A purely phonological alternation


257

  Is this alternation:  purely phonetic?

 purely phonological?

 both: phonology has access to phonetic information?

4. A purely phonological alternation 36


258

  Argument 1:

Slow speech rate / pause between 1 and 2 does not undo the harmony

it is not the actual coarticulatory effect and phonetic realization that are driving the harmony



259

  Argument 1:

Slow speech rate / pause between 1 and 2 does not undo the harmony

it is not the actual coarticulatory effect and phonetic realization that are driving the harmony



260

Argument 2: Opacity of intervening /w/ :

a.  /wəər+-ó/ > wòòr-ó ‘mongoose-pl’

b.  /səw+-ò/ > səw-ò ‘warthogs’ (*sówò)c.  /məw+-ó/ > məw-ó ‘scorpions’ (*mówó)

NB: Also a general MSC: *Uw

(no exception in 2200 word lexicon)



261

  Undominated *Uw constraint:

Interesting case where the system “chooses” to keep the unstable correspondence, the “uncomfortable” level of similarity unchanged.


/ɓᵼr-ú/ *Uw IDENT-IO[RD]

CORRα-VV(RD≥4)

IDENTα-VV[RD]

IDENT-OI[RD]

a. məᵒw-ó *! b. mów-o *! * * c. məᵒw-əᵒ *! *

d. məᵒαw-óα * e. moαw-óα *! *


262

  Argument 3: stratum-dependent In stratum 2, rounding harmony is systematic and unconditional

E.g. verb + object suffix:

a)  /tɨr+-ùn/ > túr-ùn ‘put her across’ *Lab

b)  /dəg+-òn/ > dòg-òn ‘drag her’ *Lab

c)  /dəg+-nǔ/ > dòg-nǔ ‘drag us (ex.)’ *Lab, *Height

d)  /léér+-nǔ/ > lüóór-nǔ ‘wrap us’ *Lab, *Height, *-Front

Coarticulation and phonetic realization are expected to be identical in both strata!



263

5. An alternative (and very tentative) ABC analysis


264

5. Alternative analysis

/ɓᵼr-ú/ → ɓùrú  Hypothesis:

 V1 [ɨᵘ] does not assimilate to V2 [u] … but to /u/ in the inventory:

  i.e. [ɨᵘ] and [u] are not perceptually contrastive enough, and [ɨᵘ] is reinterpreted as [u].s

43


265


  Contrast as a systemic property (Flemming 2001, 2004; Hayes & Steriade 2004:24-25)   “Constraints favoring distinct contrasts are constraints on the differences between forms rather than on the individual forms themselves” (Flemming 2004:232) “paradigmatic constraints”, which evaluate whole sets of possible forms, rather than single output forms.

44


266


  i.e. Comparison between each segment in the output form and the full phonological inventory. E.g. for vowels:

45

i ü ɨ ue üo ə oia üa a ua

ɓ ᵼᵒ r ú


267


  i.e. Comparison between each segment in the output form and the full phonological inventory. E.g. for vowels:

46

ɓ ᵼ r à ɓ ᵼᵒ r ú

i ü ɨ ɨ ɨᵒ u

e üo ə o

ia üa a ua


268


  PROPOSAL: Output-to-Inventory Correspondence

CORRα-VOUTVINV(RD≥4)

47

ɓᵼᵒαru ↔ { uα }


269

4. Alternative analysis 48

/ɓᵼr-ú/ IDENT-IO[RD]

CORRα-VOUTVINV (RD≥4)

IDENTα-VV[RD]

IDENT-OI[RD]

a. ɓᵼᵒr-ú ↔ {u} *! b. ɓùr-u ↔ {u} *! * c. ɓᵼᵒαr-ú ↔ {uα} *!

d. ɓuαr-ú ↔ {uα} * f. ɓᵼr-ᵼ ↔ {…} *!


270

CONCLUSION


271

Conclusion

  Laal doubly triggered rounding harmony is an argument in favor of the inclusion of subphonemic / subfeatural information in phonology

50


272

Conclusion

  Enrichment of ABC: ABC accounts for subphonemic/sub-featural similarity effects Both local and long-distance effects of assimilation can be insightfully captured by one and the same model Potentially interesting account of paradigmatic similarity effects (à la Flemming)

If ABC is a theory of similarity driven processes, it should be able to account for processes driven by perceptual distinctiveness and contrast effects.

51


273

Conclusion

  Enrichment of ABC:  ABC accounts for subphonemic/sub-featural similarity effects

Both local and long-distance effects of assimilation can be insightfully captured by one and the same model Potentially interesting account of paradigmatic similarity effects (à la Flemming)


52


274

Conclusion

  Enrichment of ABC:  ABC accounts for subphonemic/sub-featural similarity effects  Both local and long-distance effects of assimilation can be insightfully captured by one and the same model

Potentially interesting account of paradigmatic similarity effects (à la Flemming)


53


275

Conclusion

  Enrichment of ABC:  ABC accounts for subphonemic/sub-featural similarity effects  Both local and long-distance effects of assimilation can be insightfully captured by one and the same model  Potentially interesting account of paradigmatic similarity effects (à la Flemming) If ABC is a theory of similarity driven processes, it should be able to account for processes driven by perceptual distinctiveness and contrast effects.

54


276

Conclusion

  Enrichment of ABC:  ABC accounts for subphonemic/sub-featural similarity effects  Both local and long-distance effects of assimilation can be insightfully captured by one and the same model  Potentially interesting account of paradigmatic similarity effects (à la Flemming)   If ABC is a theory of similarity driven processes, it should be able to account for processes driven by perceptual distinctiveness and contrast effects.

55


277

THANK YOU!

56


278

References 1/1

  Bennett, William G. 2013. Dissimilation, Consonant Harmony, and Surface Correspondence. Ph.D. dissertation, NJCU, Rutgers.

  Flemming, Edward S. 1997. Phonetic Detail in Phonology: Towards a unified account of assimilation and coarticulation. In Suzuki, K. and D. Elzinga (eds.), Southwest Workshop on Optimality Theory: Features in OT (SWOT), Coyote Papers.

  Flemming, Edward S. 2002. Auditory Representations in Phonology. Routledge: New York & London.

  Flemming, Edward S. 2004. Contrast and perceptual distinctiveness. In B. Hayes, R. Kirchner and D. Steriade (Eds.), 2004, Phonetically Based Phonology. Cambridge Unviersity Press.

  Hansson, Gunnar Ólafur. 2001. Theoretical and Typological Issues in Consonant Harmony. Unpublished doctoral dissertation. University of California, Berkeley.

  Inkelas, Sharon and Stephanie Shih. 2013. ABC+Q: Contour segments and tones in (sub)segmental Agreement by Correspondence. Paper presented at Phonology 2013, UMass, 8-10 Nov. 2013.

  Inkelas, Sharon and Stephanie Shih. Unstable surface correspondence as the source of local conspiracies. Paper to be presented at NELS 44, University of Connecticut, Oct. 18-20, 2013.

  Kaun, Abigail. 2004. The typology of rounding harmony. In B. Hayes et al. (eds.), Phonetically Based Phonology. 2004. Cambridge: CUP

57


279

References 2/2

  Linker, Wendy. 1982. Articulatory and acoustic correlates of labial activity in vowels: A cross-linguistic study. Doctoral dissertation, UCLA. Published in UCLA Working Papers in Phonetics 56.

  Rhodes, Russell. 2012. Vowel Harmony as Agreement by Correspondence. Annual Report of the UC Berkeley Phonology Lab (2012), pp.138-168.

  Rose, Sharon and Rachel Walker. 2004. A typology of consonant agreement as correspondence. Language 80:475-31.

  Sasa, Tomomasa. 2009. Treatment of Vowel Harmony in Optimality Theory. Ph.D. dissertation, University of Iowa.

  Shih, Stephanie. 2013. Consonant-tone interaction as Agreement by Correspondence.

  Terbeek, Dale. 1977. A cross-language multidimensional scaling study of vowel perception. PhD dissertation, UCLA. Published in UCLA Working Papers in Phonetics 37.

  Walker, Rachel. 2009. Similarity-sensitive blocking and transparency in Menominee. Paper presented at the 83rd Annual Meeting of the Linguistic Society of America. San Francisco, California. 9 January 2009. http://www-bcf.usc.edu/~rwalker/MenomineeLSAHdt.pdf

58


280

Acknowledgments:

  Many thanks to my Laal consultants and the Laal-speaking community in Chad.

  Larry Hyman, Sharon Inkelas, Stephanie Shih, Donca Steriade, John Sylak-Glassman for helpful comments

  Audiences at UC Berkeley, Stanford University, UMass, LSA 2014 and OCP11.

  This work is supported by the Volkswagen Foundation DOBES program (grant #85538)

59


281

Appendix: Lowering of the threshold

  One exception with a front vowel:  pıır-u p ǘǘru ‘compounds’

60


282


  Different co-indexation:

61

CORR-VV(RD≥4)

IDENTα-VV[RD] CORRα-VV(RD≥3)

CORR-VV(RD≥2)

CORR-VV(RD≥1)


283


  + re-ranking CORRα-VV(RD≥3) :

a. CORR-VV(RD≥4)

IDENT-OI[RD]

CORRα-VV(RD≥3)

62


284


  + re-ranking CORRα-VV(RD≥3) :

a. CORR-VV(RD≥4) b. CORR-VV(RD≥4),

IDENT-OI[RD]

CORRα-VV(RD≥3) CORRα-VV(RD≥3)

IDENT-OI[RD]

63


285

a. /bìru / bìru ‘burn’


/bìr-u/ IDENT-IO[RD]

CORRα-VV(RD≥4)

IDENTα-VV[RD]

IDENT-OI[RD]

a. bìr-u b. bür-u *! c. bìr-ı *!

64


286

a. /bìru / bìru ‘burn’

b. /pıır-u/ p ǘǘru ‘compounds’



CORR-VV(RD≥4) IDENT-VV[RD] IDENT-

OI[RD]a. bìr-u

b. bür-u *! c. bìr-ı *!


CORR-VV(RD≥4)

CORRα-VV(RD≥3)

IDENTα-VV[RD]

IDENT-OI[RD]

a. pııᵒr-ú *! b. p ǘǘr-ú *! * c. pııᵒr-ᵼ *! d. pııᵒαr-úα *!

e. p ǘǘαruα *

65


287

PHONOLOGICAL TEAMWORK IN LAAL ROUNDING...

Documents

Transcript of PHONOLOGICAL TEAMWORK IN LAAL ROUNDING...