Philippe Ségéral Tobias Scheer

Université Paris 7 Université de Nice

32th Pozna½ Linguistic Meeting 30 April – 2nd May 1999

Full Interpretability: evidence from Virtual Geminates (1) Full interpretability (Harris 1994,1996, Harris&Lindsey 1990,1995) a. phonological primes possess a phonetic interpretation in isolation. b. hence, all phonological representations at any derivational stage are phonetically interpretable. c. there is no component of systematic phonetic interpretation because all phonetic information is defined in phonological representations. d. phonology is mapped onto phonetics one-to-one: for a given phonological representation, one and only one phonetic realisations, and vice-versa. (2) Virtual Geminates there is converging evidence for the existence of consonants that are underlyingly double, but always simplex on the surface. a. Somali (Coushitic) b. Cologne German c. German

Somali (3) conjugation of an unmarked verb: keen "bring" habitual present stem pers. tense stem 1sg keen ø aa keen 2sg keen t aa keen 3m.sg keen ø aa keen 3f.sg keen t aa keen 1pl keen n aa keen 2 pl keen t aa n keen 3 pl keen ø aa n keen imper keen ø ø

past pers. ø t ø t n t ø

tense ay ay ay ay ay ee n ee n

-2-

(4) Somali vowel-zero alternations follow the usual pattern a. cross-linguistic pattern1 zero C__C-V

vowel C__C-ø

vowel C__C-CV

Moroccan Arabic

ki-tøb-u

køti-b-ø

ki-tti-b-ø

German (optional elision) Tangale (Chadic) Turkish Slavic (e.g. Czech) Hungarian

innør-e

inner-ø

inner-lich

dobø-go devør-i lokøt-e majøm-on

dobe devir-ø loket-ø majom-ø

dobu-n-go devir-den loket-ní majom-ra

nouns

plural nirøg-o

sg. indefinite nirig-ø

gloss write perf.act.3pl, 3sg, 3sg causative inner+infl, inner, internal called, call, called me transfer ACC, NOM, ABL elbow GEN, NOM, adj. monkey Superessive, NOM, Sublative

Somali

verbs: CViCViC stems, 1sg first conjugation gudb-aa tifq-aa boqr-aa Íerg-aa bar"-aa verbs: 3rd conjugation suffix –(a)t√CVC √CVCC

sg. definite nirig-ta

young female camel

2sg gudub-taa tifiq-daa * boqor-taa Íereg-taa bara"-daa *

1sg

3sg masc.

3pl

qab-t-aa tuur-t-aa Ía"l-ad-aa kabb-ad-aa joogs-ad-aa

qab-t-aa tuur-t-aa Ía"l-ad-aa kabb-ad-aa joogs-ad-aa

qab-t-aan tuur-t-aan Ía"l-ad-aan kabb-ad-aan joogs-ad-aan

inverse s'égoutter proclaim king se rassasier split by water

hold plie up inherit sip stop oneself

in alternating stems √CV1CV2C-, V2 is always a copy of V1 * t-->d / Guttural__

(5) generalisations: alternation-sites show a vowel if a. a CC-cluster occurs on their righthand side: 1st conjugation. √CViC__C-C b. a CC-cluster occurs to the lefhand side: 3rd conjugation. √CVCC-__t-V *CCC, *CC# in Somali. (6) general process in Somali: voiceless consonants voice in intervocalic position e.g. t-->d / V__V bar bar-ta tache, la tache mindi mindi-da knife, the knife

1

See Scheer (1997) for a general survey of vowel-zero alternations. Data regarding the languages mentioned can be found in Scheer (1996) for Czech, Nikiema (1989) for Tangale, Kaye (1990) for Moroccan Arabic, Barillot (1997) for Somali, Törkenczy (1992) for Hungarian, and Wiese (1995), Noske (1993) for German.

-3-

(7) 3rd conjugation paradigm imperative qab-o tuur-o 1sg 2sg 3sg masc. 3sg fem 1pl 2pl 3pl

Ía"l-o

kabb-o

joog-s-o

hold

pile up

inherit

sip

stop oneself

qabtaa qabataa qabtaa qabataa qabannaa qabataan qabtaan

tuurtaa tuurataa tuurtaa tuurataa tuuranna tuurataan tuurtaan

Ía"ladaa Ía"lataa Ía"ladaa Ía"lataa Ía"lannaa Ía"lataan Ía"ladaan

kabbadaa kabbataa kabbadaa kabbataa kabannaa kabbataan kabbadaan

joogsadaa joogsataa joogsadaa joogsataa joogsanna joogsataan joogsadan

(8) a.

2sg, 3sg fem, 2pl: qabataa(n) according to generalisation (5), either [b] or [t] must be a CC-cluster. b. overt geminates in Somali: [bb, dd, ÍÍ, gg, mm, nn, rr, ll] if [b] were a geminate, it would appear as such on the surface. ==> [t] is a geminate

(9) 3 arguments in favour of [t] being /tt/: a. vocalisation of the alternation-site. b. resistance of [t] to intervocalic voicing, cf. (6): [t] does not voice because it is a geminate. c. [t] is morphologically twofold: 2sg /qab-(a)t-taa/ [qabataa] vs. 1sg /qab-(a)t-aa/ [qabtaa] /Ía"l-(a)t-taa/ [Ía"lataa] vs. /Ía"l-(a)t-aa/ [Ía"ladaa] 3sg masc. /qab-(a)t-taa/ [qabataa] /Ía"l-(a)t-taa/ [Ía"lataa] 2pl /qab-(a)t-taan/ [qabataan]

Kölsch (10)

standard Ger. Cologne German a. n - N b“awn b“uN N b. t – k hçjt´ hyk d – g Snajd´n Snig´ c. nt - Nk bunt buN Nk nd - N bind´n biN N´

braun heute schneiden bunt binden

(11) diachronic velarisation: OHG - Old High German, about 850-1050 A.D. MHG - Middle-High German, about 1050-1350 A.D. NHG – Modern Standard German MHG [t,d,n] > Col [k,g,N] / {ii,uu,yy}__ MHG [nt,nd] > Col [Nk,Ng] / {i,u,y}__

"brown" "today" "cut" "colorful" "tie"

-4-

n

t,d

MHG NHG Col NHG - Col uu__ aw__ u__ braun - bruN Zaun - ZuN ii__ aj__ i__ fein - fiN Rhein - RhiN yy__ çj__ y__ neun – nüN uu__ aw__ u__ Braut – Bruk Haut – Huk ii__ aj__ i__ Seide – Sik Streit – Strik

nt,nd i__

i__

u__

u__

y__

y__

MHG NHG Col NHG – Col ou__ au__ aw__ — ei__

aj__

aj__

Bein – Bein Stein – Stein

ou__ au__

au__

Flaute – Flaute

ei__

aj__

neiden – neide breit - breit

aj__

i__

blind - bliNk Pinte - PiNk u__ bunt - buNk Hund - huNk y__ Bündel - PüNel

(12) generalisations a. simplex dentals velarize only if preceded by a long vowel. b. when velarization of simplex dentals takes place, the preceding long vowel shortens. c. dental clusters [nt,nd] velarize after short vowels. (13) velarisation may be construed as a a unique process only if the result is geminate (T=dentals, K=velars): VNT > VNK VVT > VKK (14) a. velarisation of simplex consonants takes place only after long vowels because only long vowels may provide an additional skeletal slot. b. velarisation of dental clusters [nt,nd] is triggered by short vowels because they do not need to acquire an additional skeletal position in order to be geminate. c. long vowels preceding velarized consonants are shortened because their second skeletal position is occupied by the velar. (15) MHG hût, zît, niun

>

Kölsch Huck, Zick, nüng

O N O N O N | | h u t z i t n y n

>

O | h z n

(16) possible diachronic analysis: a. MHG b. velarisation and gemination c. degemination Kölsch

[VVT] *[VKK] [VK]

N O N O N | u k i k y N

-5-

(17) However, /Vg/ cannot be the synchronic identity of [Vg] < VVg: synchronic palatalisation of /g/. ANY German [g] appears as Kölsch [j]: German günistig [gYnstik] ungeheuer [ung´hçjå] groß [g“oos] glauben [glawb´n]

Kölsch [jYnstiS] [unj´hçjå] [j“oos] [jløøv´]

egal balgen Gig

German [Egaal] [balg´n] [gig]

Kölsch [Ejaal] [balj´] [jik]

(18) Hence, there should be no [g] in Kölsch at all. The only surface [g] in Kölsch are velarized dentals: German krauten lauten leiden

Kölsch krug´ lug´ lig´

German reiten schneiden schreiten

Kölsch rig´ S“ig´ S“ig´

German streiten Beutel deuten

Kölsch StXig´ byg´l dyg´

(19) [g] < [d] does not undergo synchronic palatalisation because only simplex /g/ palatalise. ==> [g] < [t,d] is /gg/.

Standard German (20) instability of #sC- clusters Romance prothesis (Spanish, Italian, French,…) accent circonflexe "^" indicates vowel-length in Old French, cf. Bourciez (1926:207). word-internally: Compensatory Lengthening Lat VsC > Old Fr VVC Lat French musca mouche *piscare pêcher testa tête *foreste forêt castellus château as(i)nu âne i(n)s(u)la île

word-initially: prothetic vowel

[#eC-] étudiant épagneul été réponse

Lat #sC- > [EsC-] estudiantin Espagne estival responsable

conclusion: initial +ATR vowels are long. They are the result of Compensatory Lengthening. where does the initial [e] compensate to?

-6-

(21) existence of an "initial CV", cf. Lowenstamm in press, Ségéral&Scheer (ms) "#" = CV. Romance strategy when faced with the instability of #sC: identification of the initial CV. Latin student C V - C V C V C V C V C V | | | | | | | s t u d e n t French estudiantin [estydijãtE)]: #VsC -ATR

French étudiant [etydijã]: #VVC +ATR

C V - C V C V C V C V | | | | | | E s t u d i…

C V - C V C V C V C V | | | | | e s t u d i …

(22) instability of #sC: German

st str sp spl spr sl sn sm sv

sC > SC / #__

word-initial MHG [#sC] NHG [#SC] stein Stein strît Streit sparn sparen splittern splittern sprechen sprechen slâfen schlafen snël schnell smal schmal swîn Schwein

word-internal MHG NHG wüeste Wüste Astrid wespe Wespe

Breslau

(23) [S] is the phonetic manifestation of the gemination of [s]. The strategy is the same as in Romance: identification of the initial CV. MHG snël [snEl]

NHG schnell [SnEl]

C V - C V C V C V | | | | s n E l

C V - C V C V C V | | | | s n E l

(24) *VVCC in monomorphemic sequences (Mond < Monat, fahnden < Low German, etc.) kalt, Wald, Wild, krank etc. *[VVS] except loans: Nische, Dusche

-7-

(25) what about #sk ? cf. Braune & Eggers (1886:136ff.) OHG sk > Sk > S / #__ OHG scado > M/NHG schade [Saad´] OHG sk > S / …__… OHG lescan > M/NHG löschen [løS´n] (26) loss of [k] causes Compensatory Lengthening of the following vowel in #__ OHG M/NHG scaban schaben Saab´n Lat scatula Schädel SQQd´l scado Schaden Saad´n scala Schale (Hülse) Saal´ scama Scham Saam scara Schar Saar scelah scheel Seel scamil Schemel Seem´l sceran scheren (Wolle) Seer´n scerôn scheren (sich fortmachen) Seer´n scivaro Schiefer Siifå scilihen schielen Siil´n scina Schiene Siin´ scer(i)linc Schierling SiirliN scuren schüren Syyr´n (27) OHG scado > Skade > Saade OHG scado [skado] C V - C V C V C V | | | | | s k a d o

gemination of #s ==> [Skado] C V - C V C V C V | | | | | s k a d o

loss of [k] ==> M/NHG schade [Saad´] C V - C V C V C V | | | | s k a d ´

(28) Compensatory Lengthening does not occur in closed syllables, cf. (24) *VVCC, exhaustive data cf. Appendix OHG M/NHG scarf scharf Sarf scalc Schalk Salk scenken schenken SENk´n scilt Schild Silt scinten schinden Sind´n sculd(a) Schuld Sult

-8-

(29) NHG: words #S V words #S VV

226 91 317

exhaustive data from Duden

a. √__CxCy-, cf. (28) = 54 b. √__CxCx- graphic geminates = 62 OHG scaffan scella scerran scutten got skatts scuzz(i)o scuZZila

M/NHG schaffen Schelle scharren schütten Schatz Schütze Schüssel

Saf´n SEl´ Sar´n Syt´n SatÉs SytÉs´ Sys´l

c. loans = 71 < French < Arabic < Slavic < Hebrew < English < Latin < Greek < Italian < Chinese < Iranian < Dutch < Sanskrit < Yiddish < Turkish

Schablone, Schalotte, Scharande, Schärpe, Schecke, Schikane Scharia, Scherif, Schirokko Schapka, Schelsky, Schaschlik, Schostakowitsch, Scharwenzel Schacher, schalom, schächten, Schibboleth Scheck, Schelf, Schock, Schirting Scharlach, Scharteke, Schatulle, Schilf Schimäre, Schisma, schizogen Schamotte, Scharmützel, Scherzo Schanghai, Schansi, Schantung, Schenjang Schador, Schillum Schafott, Schellack Schakal, Schamane Schamott, Schickse Schabracke

d. dialectal = 39 < Swabian scherbeln < Austrian Schachen < landschaftlich Schalander

(30) conclusion: √__CxCx are closed syllables. (31) a. why does [s] become [S] word-internally in [sk], but not in [sp,st]? Because there is no free position it can spread to word-internally. b. why does the loss of [k] cause Compensatory Lengthening of the following vowel word-initially, but of [s] word-internally? Because [s] is already geminated word-initially. c. why is there Compensatory Lengthening of the vowel after [sk] word-initially, but not word-internally? Because the vowel to the lefthandside of […sk…] is schwa.

-9-

OHG lescan [leskan]

M/NHG [løS´n]

C V - C V C V C V C V | | | | | | l e s k a n

C V - C V C V C V C V | | | | | l ø s k ´ n

(32) virtual geminates in NHG a. [S] b. graphic geminates

References Barillot, Xavier 1997. Alternances voyelle-zéro et gémination consonantique en somali. DEA thesis, Université Paris 7. Bourciez, Edouard 1926. Précis Historique de Phonétique Française. 6th edition Paris: Klincksieck. Braune, Wilhelm and Hans Eggers 1886: Althochdeutsche Grammatik. 14th edition Tübingen: Niemeyer 1987. Harris, John 1994. English sound structure. Oxford: Blackwell. Harris, J. (1996). Full interpretability in phonology. Current trends in Phonology: Models and Methods, edited by J. Durand & B. Laks, 305-332. Salford: ESRI. Harris, J. & G. Lindsey (1990). Phonetic interpretation in generative grammar. UCL Working Papers in Linguistics 2, 355–369. Harris, J. & G. Lindsey (1995). The elements of phonological representation. Frontiers of Phonology, edited by J. Durand & F. Katamba, 34–79. London: Longman. Kaye, Jonathan 1990. Government in Phonology: the case of Moroccan Arabic. The Linguistic Review 6, 131-159. Lowenstamm, Jean in press. The beginning of the word. In: Syllables?!, edited by John Rennison. Amsterdam: Holland Academic Graphics. Nikiema, Emmanuel 1989. Gouvernement Propre et licenciement en phonologie: le cas du Tangale. Langues Orientales Anciennes, Philologie et Linguistique 2, 225-251. Noske, Roland 1993. A theory of syllabification and segmental alternations. With studies on the phonology of French, German, Tonkawa and Yawelmani. Tübingen : Niemeyer. Scheer, Tobias 1996. Une théorie de l'interaction directe entre consonnes. Doctoral Dissertation, Université Paris 7. Scheer, T. (1997). Vowel-zero alternations and their support for a theory of consonantal interaction. Certamen Phonologicum III, edited by P.M.Bertinetto, L.Gaeta, G.Jetchev and D.Michaels, 67-88. Torino: Rosenberg & Sellier. Ségéral, Philippe & Tobias Scheer ms. The Coda Mirror. Törkenczy, Miklos 1992. Vowel-zero Alternations in Hungarian: a Government Approach. In: Approaches to Hungarian 4, edited by Isv