14
 A CVCV Analysis of Syllabic Consonants in Coptic Jean-Marc B ELTZUNG  1 & Cédri c PATIN  1 ,2 1 LPP, Paris 2 ZAS, Berlin [email protected] [email protected] 24 May 2007 1 In tr oduc ti on (1) sylla bic c onson ants hav e rec eiv ed d iffe rent treat ments in ph onolo gical litera - ture in CVCV phonology, a radical extension of the Government Phonology theory: syllabic consonant = a segment associated both to a C slot and a V slot   right-branching hypothesis  (Blaho 2004, among others)   = left-branching hypothesis (Scheer 2004, among others) (2) this pap er dea ls wi th syl lab ic con sonants in Cop tic , a now e xti nct ed Afro- Asiatic language (3) it will be sh own that copti c ph enome na s uppor t the left-b ranch ing h ypoth esis 2 Sch was and syll abi c consonants (4) no pla ce her e for an exten ded dis cus sio n we will assume a basic knowledge of the nature of syllabic consonants  howev er, it is important for our purpose to focus on the close relationship between  syllabic consonants  and  sequences of a schwa and a consonant (5) three diffe rent evi denc es supp orting the relat ion exis ting betwe en  syllabic con- sonants  and  schwa + consonant  sequences:  free variation  +  interdialectal vari- ation +  complementary distribution. 2.1 Evi den ce 1: fr ee variation (6) rst evidence in fav or of a direct link between C  and  C: they are in free varia- tion in many languages Many thanks to Gwendoline Fox for helpful corrections 1

2007 Beltzung&Patin Coptic Syllabic Consonants

Embed Size (px)

DESCRIPTION

Study about syllabic consonants in coptic language

Citation preview

  • A CVCV Analysis of Syllabic Consonants inCoptic

    Jean-Marc BELTZUNG 1 & Cdric PATIN 1,21 LPP, Paris

    2 ZAS, [email protected]

    [email protected]

    24 May 2007

    1 Introduction(1) syllabic consonants have received different treatments in phonological litera-

    ture in CVCV phonology, a radical extension of the Government Phonologytheory: syllabic consonant = a segment associated both to a C slot and a V slot right-branching hypothesis (Blaho 2004, among others) 6= left-branchinghypothesis (Scheer 2004, among others)

    (2) this paper deals with syllabic consonants in Coptic, a now extincted Afro-Asiatic language

    (3) it will be shown that coptic phenomena support the left-branching hypothesis

    2 Schwas and syllabic consonants(4) no place here for an extended discussion we will assume a basic knowledge

    of the nature of syllabic consonants however, it is important for our purposeto focus on the close relationship between syllabic consonants and sequencesof a schwa and a consonant

    (5) three different evidences supporting the relation existing between syllabic con-sonants and schwa + consonant sequences: free variation + interdialectal vari-ation + complementary distribution.

    2.1 Evidence 1: free variation(6) first evidence in favor of a direct link between C

    "

    and C: they are in free varia-tion in many languages

    Many thanks to Gwendoline Fox for helpful corrections

    1

  • (7) the presence or absence of the schwa is most of the time related to the styleof speech (e.g. Halh Mongolian (Karlsson Mukhanova 2006, 2007): alternationbetween a N sequence (in formal speech) and an N

    "

    (in casual speech))(8) the most discussed cases of C

    "

    / C free alternation are probably in Germaniclanguagesa. English (Clark & Yallop 1995: 67)

    sudden [s2dn] [s2dn"

    ]medal [mEdl] [mEdl

    "

    ]b. German (Siebs 1961, cited in Clark & Yallop 1995: 68)

    haben [ha:bn] [ha:bm"

    ] to havegeben [ge:bn] [ge:bm

    "

    ] to givesagen [za:gn] [za:gN

    "

    ] to saydanken [daNkn] [daNkN

    "

    ] to thankc. NE Dutch (Toft 2002: 112)

    kiken [kikn] [kikN"

    ] to look

    (9) note that the presence of a schwa is supported in 8-b and 8-c by the absence ofhomorganicity, while its absence is related to homorganicity

    2.2 Evidence 2: interdialectal variation(10) the direct link between C

    "

    and C is also supported by the fact that a C se-quence in a dialect may correspond to a syllabic consonant in another one(e.g. syllabic nasals in Hellendoorn Dutch but [n] in other variants of Dutch(van Oostendorp 2001, 2004))

    (11) Berber any consonant, even a voiceless stop, may be syllabic in BerberChleuh:Berber (Chleuh) (Ridouane 2003: 269-70 and personnal communication)

    kks

    "

    take offfq

    "

    .qs

    "

    irritatetk

    "

    .Sf

    "

    it driedtf

    "

    .ss

    "

    she is quiettk

    "

    .ks

    "

    t you took offtq

    "

    s.sf

    "

    it shrunktf

    "

    .tk

    "

    .ts

    "

    tt you sprained it (fem.)(12) Ridouane (2003) has demonstrated that there is no schwa in the varieties of

    Haha or Anti-Atlas, but that a schwa is present in the variety of Agadir:Berber (Chleuh) (Ridouane 2003)

    Agadir Hahafk f

    "

    k giveks ks

    "

    feed onsXf pas

    "

    Xf fade away

    2.3 Evidence 3: complementary distribution(13) last evidence : complementary distribution

    2

  • (14) Britannic English (Toft 2002) the final /l/ is almost always syllabic (e.g.people, vocal, etc.), while final /n/ shows two different behaviors: i. it tendsto be syllabic after /t/ (beaten) ii. there is a schwa between /p,k/ and a final/n/ (deepen, beacon)

    (15) German (Noske 1996) "there is postlexically a free alternation betweensyllabic sonorant cononants in German and a sequence of schwa + sonorant[...] except for R, which is always vocalised under these circumstances" (p.172-5).

    (16) Klaaifrysk Frisian (van Oostendorp 2004) "syllabic sonorants and schwaare in a form of complementary distribution: the former occur in exactlythose environments where the latter do not occur." (p. 28)

    (17) Piro Arawakan, Peru (Lin 1997,2005) "the two forms, C"

    and Cv, areconsidered allophones of the corresponding underlying C (Matteson 1965:33), and [that] the choice between C

    "

    and Cv is rule-governed." (Lin 1997:406)

    3 Syllabic consonants in CVCV

    3.1 The CVCV framework(18) CVCV is a radical extension of Government Phonology (Lowenstamm 1996,

    Scheer 2004): Prosodic structure is universally composed of strictly alternat-ing CV units : CVCV. . . CnVn. Two lateral relations structure the string:a. Governmentb. Licensing

    (19) two major predictions are made by CVCV:a. No coda (constituent)b. No branching constituents

    (20) according to CVCV:a. a coda is a consonant followed by an empty nucleus ()b. a geminate is a segment associated with two Onset positionsc. a long vowel is a segment associated with two Nuclei positionsd. Sgral & Scheer 1999, p. 16

    closed syllable geminate long vowel [...C#]O N O N

    C V C

    O N O N

    C V

    O N O N

    C V

    O N

    ... C #

    (21) the Empty Category Principle (Kaye et al. 1990) states that a nucleus may re-main unexpressed if, and only if, it is:a. subject to Proper Government (PG)b. enclosed within a domain of Infrasegmental Government (IG)c. domain-final (parametrized)

    (22) lateral relations structuring the string:

    3

  • a. Proper Government (PG) is a dependency relation that holds between twonuclei. One is the head (governor), the other its complement (governee).Only plain vowels, IG enclosed vowels and domain-final vowels maygovern another one.

    b. Infra-segmental Government (IG) is a dependency relation that holds be-tween the Place primitives of two different Onsets. Sonorants are pre-dicted to be governors and Obstruents governees.

    c. Government Licensing: a consonant may govern another one (throughIG) if licensed by its own nucleus.

    (23) Government inhibits the segmental expression of its target + Licensing strength-ens the segmental expression of its target

    (24) Proper Governmenta. an ungoverned vowel may governs an empty nucleus

    O1 N1 O2 N2

    C C Vb. a governed vowel cannot govern a preceding empty nucleus

    O1 N1 O2 N2 O3 N3

    C C C Vc. N1 must be phonetically expressed in this case

    O1 N1 O2 N2 O3 N3

    C C C V

    (25) Infra-segmental Governmenta. a sonorant R may govern an obstruent T, but a T cannot govern a R

    a. O N O N

    T R V

    b. O N O N

    R T V8b. a consonant must be licensed by its own nucleus to be governer

    O N O N O N

    T R V C V

    c. if not, the consonant cannot govern another one

    O N O N O N

    R 8 T C V

    (26) Lowenstamm (1999) claimed that the diacritic # used at the beginning of thewords is an empty initial CV (C0V0)

    (27) accounts for the asymmetry of attested inital clusters in languages that have:a. #TR only (Indo-European languages)

    4

  • b. both #TR and #RT (some Semitic languages)c. implication: #RT #TR

    (28) initial site is a site of clitization Clitization can take place iff the site islicensed

    (29) a. an infra-segmental domain allows the government of V0C0 V0 C V C V

    # T R A

    b. If IG fails to appear, V0 is not governed and may be pronounced

    C0 V0 C V C V

    # R T A9

    (30) under CVCV, a syllabic consonant is a segment associated both to a C slot anda V slot explains why syllabic consonant patterns with both vowels andconsonants (eg. tones, stress, homorgany)

    (31) almost all authors agree on the dual nature of syllabic consonants. However,most of the discussion deals with Germanic or Slavic languages. A centralquestion regarding their representation: Are syllabic consonants:a. right-branching (Blaho 2004) orb. left-branching (Scheer 2004) ?

    a. Right-branching b. Left-branchingC V

    V C

    3.2 Left- or right-branching ?(32) Right-branching: advocated by different authors Rennison 1993, Afuta 2002,

    Blaho 2004, among others) + discussed in detail in Blaho (2004)(33) Blaho (2004) cites several arguments to support her analysis:

    a. consonantal phonotactics: in Slovak, syllabic liquids are only sensitive toconsonants preceding them.

    b. the minimal word constraint : she claimed that there is no word only com-posed of a syllabic consonant

    (34) weakness: arguments are builded upon a small set of languages (Germanic andSlavic). There are counter-arguments in other languages:a. consonantal phonotactics: syllabic nasals homorgonize with the follow-

    ing consonants in Yao (Ngunga 2000) and Dagaare (Bodomo 1997)b. the minimal word constraint : syllabic consonants may appear as a word

    of its own in Berber (Ridouane, pc): imperative [f"

    ] give and in ShaoxingChinese (Zhang 2006): [m

    "

    ] yes, [N"

    ] five

    5

  • (35) Left-branching: advocated by Harris (1994), Toft (2002) and Scheer (2004)among others + discussed in details in Scheer (2004)

    (36) arguments:a. alternation of C and C

    "

    in Germanic languagesb. diachronical emergence of syllabic consonants through the loss of a pre-

    ceding vowel in Slavic (Scheer 2004)(37) weakness: arguments are builded upon a small set of languages (Germanic and

    Slavic) e.g. counter-argument regarding syllabic consonants emergence inother languages: in African languages, most of the syllabic nasals emergedfrom the loss of following vowel (see for instance Creissels (1999) on Tswana) synchronic alternations (Saghala our own data , Ganda: Kawaha 2001)

    4 Syllabic consonants in Coptic

    4.1 The Coptic language(38) Coptic is an extinct Afro-asiatic language.

    a. last stage of Ancient Egyptian.b. attested between the 3rd and 14th century CE.c. not a uniform language but a set of five main dialects (from north to

    south): Bohairic (B), Fayyumic (F), Sahidic (S), Lycopolitan (L) andAkhmimic (A)

    (39) written with Greek alphabet augmented by seven Demotic letters which dontexist in Greek e.g. r

    p p

    m-

    p p

    n-khme Egyptian ("man of Egypt")

    4.2 Syllabic consonants: graphical expression(40) in Coptic SAL, a non-etymological diacritic called supralinear stroke

    may appear over all consonants:p p

    m,

    p p

    n. . .

    (41) different interpretations in the litterature:a. Till (1932) : denotes a vowel since it alternates with schwab. Worrell (1934): indicates syllabic division of a word syllabic conso-

    nant.c. Polotsky (1933): denotes a syllabic consonant

    (42) nowadays, supralinear stroke as syllabic marker has been commonly acceptedamong authors (see Polotsky 1933, Nagel 1966, Plumley 1948, Vergote 1973, Depuydt1985, Vicychl 1990, Peust 1999, Layton 2000 among others)

    4.3 Syllabic consonants as left-branching: some indices(43) if supralinear stroke denotes syllabic consonants in SAL,

    a. all consonants can have a syllabic function in unstressed syllables butb. only syllabic sonorants can occur in stressed syllable (Worrell 1934, Nagel

    1966, Vicychl 1990 among others)

    6

  • (44) our goal is to show that syllabic consonants are left-branching in Coptic. Wewill focuse merely on syllabic sonorants and four arguments will be advanced:1. inter-dialectal complementary distribution of C

    "

    and C and intra-dialectalfree variation of C

    "

    and C2. emergence of stressed syllabic sonorants3. reduplication process4. coronal nasal homorgany before labial obstruents

    4.3.1 Complementary distribution and free variation

    (45) Inter-dialectal: when a syllabic consonant appears in S, A and L (signaled bysupralinear stroke), B and F generally have a sequence of schwa (signaled bye) and the non-syllabic counterpart:

    C"

    (SAL) C (BF)a. S n

    "

    to S nto robe (of linen)hr

    "

    man hrman pomegrenatehm

    "

    s xms ear

    l

    "

    hem lhEm to shouthotB

    "

    xotB to killb. SOlms

    "

    SOlms gnat, mosquitosolp

    "

    solp to break

    (46) C appears in BF in the following cases:a. when the syllabic consonant in SAL is a sonorant orb. when the syllabic consonant in SAL is an obstruent preceded by a sono-

    rant1

    (47) Intra-dialectal. Free variation of C"

    and C is sometimes observed in SC

    "

    Ca. B

    "

    SE BSE forgetfulnessl

    "

    hem lhem roar

    wm

    "

    - wm- eat (construct form)OBn

    "

    OBn alumm

    "

    pr

    "

    - m

    "

    pr- do notb. m

    "

    pk

    "

    - m

    "

    pk- negative past tensesolp

    "

    solp to breakhoBs

    "

    hoBs cover

    onh

    "

    onh livelahmf

    "

    lahmf meaning unknow

    (48) this first argument leads us to represent Coptic syllabic consonant as left-branching structure in a strict CV framework. Thus, we assume the followingrepresentations for the form S. [Sn

    "

    to], B. [Snto] "robe (of linen)":

    a. Sahidic C1 V1 C2 V2 C3 V3

    S n t o

    b. Bohairic C1 V1 C2 V2 C3 V3

    S n t o

    1This case may explain why examples with syllabic obstruents alternations are scarce cross-dialectally.

    7

  • 4.3.2 Stressed syllabic sonorants

    (49) the 2nd argument comes from stressed syllabic sonorants. Stressed syllabicsonorants emerge diachronically through the loss of a stressed vowel preced-ing a non-syllabic sonorant. The lost vowel is always the outcome of Egyptianstressed */i/, that is /a/ in closed syllables in S and B (but /E/ in other dialects)and /i/ or /e/ in open syllables in all dialects (Vycichl 1990:74)

    (50) stress in Egyptian languages appears on either the ultimate or the penultimatesyllable of a word (trochaic system): */"rama/ > /"rom/ man vs */"rin/ > /"ran/name

    (51) a sequence *CiRT (where R=sonorant and T=obstruent) > (CaRT) > CR"

    T inS and A but CRT in B and F, as shown by the following examples:

    Egyptian S and A B and Ftbt tB

    "

    t tBt fishm hm

    "

    tS hmtS vinegarfn-w fn

    "

    t fnt worms

    Slf Sl

    "

    f Slf fearwrt wr

    "

    t wrt rose

    (52) this fact is very common in S and A and explain why *i > a and e is rare inthis environment. In B and F, a schwa occurs instead (written as e)

    (53) stressed syllabic consonants in S and A can be represented as follows (whereboxed positions represent the stress):

    a. Sahidic b. Bohairic

    C1 V1 C2 V2 C3 V3

    t a B t

    C1 V1 C2 V2 C3 V3

    t B t

    C1 V1 C2 V2 C3 V3

    t B t

    (54) if the left-branching structure is correct, we must follow Scheer (1998:281)and admit that Coptic final empty nuclei may govern the previous nucleus ifthis nucleus is not lexically filled.

    (55) vowel loss is also synchronically reflected in the S and A verbal morphol-ogy where the vowel /i/ (i) of some infinitive verbs alterns with /a/ (a) in thepronominal form when the following consonant is not a sonorant (56a.). Whena sonorant appears after the infinitive vowel /i/, the pronominal form has a syl-labic consonant instead (56b.):

    (56) Infinitive Pronominala. mis mast- to give birth

    S ik Sakt- to dighit hat- move around, convulsesih saht- be removed

    b. in n"

    t- (*ant-) to findS iB SB

    "

    t- (*SaBt-) to changeS in Sn

    "

    t- (*Sant-) to seek, asksmin smn

    "

    t- (*smant-) to establish(57) B and F generally have a schwa instead of syllabic consonants in (56b.). Right-

    8

  • branching representation fails to predict the forms in (56a.) but correctly pre-dicts the forms in (56b.)2:Right-branching representationa. mast-f

    "

    "to give him birth" b. Sn"

    t-f

    "

    to ask him

    C1 V1 C2 V2 C3 V3-C4 V4

    m a s t f

    C1 V1 C2 V2 C3 V3-C4 V4

    S a n t f

    (58) in (57), the well formedness of the representation involves the absence of avowel in V1. If V1 is present, C2 cannot spread on V2 but V2 must be ex-pressed since it cannot be governed by V3, already governed by V4. A left-branching representation correctly predicts both forms in (56a.) and (56b.):Left-branching representationa. mast-f

    "

    "to give him birth" b. Sn"

    t-f

    "

    to ask him

    C1 V1 C2 V2 C3 V3-C4 V4

    m a s t f

    C1 V1 C2 V2 C3 V3-C4 V4

    S a n t f

    4.3.3 Reduplication

    (59) Coptic reduplication is formed by reduplication of the second and third con-sonants at the right of the base. In the infinitive form of verbs, the reduplicantappears with no vowel but with a syllabic consonant wheras the root appearswith a full vowel in Sahidic (the dot denotes syllable boundary):

    hBOr.Br

    "

    to become loosenedSOr.Sr

    "

    to knock downStOr.tr

    "

    to disturbBas.Bs

    "

    sort of duckkas.ks

    "

    to dansesOl.sl

    "

    to comfortkrOm.rm

    "

    to become darkmOn.mn

    "

    to move, shake (earth)(60) if syllabic consonants are right-branching in Coptic, the representation of redu-

    plicated forms such as [mOnmn"

    ] to move, shake (earth) is ill-formed:Right-branching structure

    C1 V1 C2 V2 C3 V3 C4 V4

    m O n m n

    (61) if syllabic consonants are Left-branching in Coptic, the representation of thereduplicated forms are well-formed:Left-branching structure

    C1 V1 C2 V2 C3 V3 C4 V4

    m O n m n

    2/-f/ is the third person masculin singular suffix

    9

  • (62) some reduplicated verbs appear without vowels neither in the base nor in thereduplicant. This fact can be explained by vowel loss in S and A. Rememberthat /a/ < */i/ in stressed closed syllables was lost before sonorants. In theseforms, the sonorant is always syllabic both in the base and in the reduplicant:Reduplication in S (Plumley 1948:72, Vergote 1973:48)

    Br

    "

    .Br

    "

    to boiltm

    "

    .tm

    "

    to become heavykm

    "

    .km

    "

    to strike a musical instrumentkrm

    "

    .rm

    "

    to murmursn

    "

    .sn

    "

    to resoundtl

    "

    .tl

    "

    to driphm

    "

    .hm

    "

    to roar

    (63) if syllabic consonants are rigth-branching in Coptic, the representation of redu-plicated forms such as [krm

    "

    rm

    "

    ] to murmur is ill formed:

    C1 V1 C2 V2 C3 V3 C4 V4 C5 V5

    k 8 r m r m

    (64) if syllabic consonants are left-branching in Coptic, the representation of thereduplicated form is well-formed:

    C1 V1 C2 V2 C3 V3 C4 V4 C5 V5

    k r m r m

    4.3.4 Homorgany

    (65) in Coptic, a consonant homorgany occurs in initial coronal nasal + obstru-ents sequences, where the nasal consonant is a proclitic element (/n-/ definiteplural article, genitive marker):

    10

  • a. m"

    -mE:w the thoughtsm

    "

    -mayn the signsm

    "

    -mes the interestsm

    "

    -Ba:mp the goatsm

    "

    -p ew the heavensm

    "

    -p-nut of Godb. n

    "

    -wot (*m"

    -wot) of onlyn

    "

    -wOyn (*m"

    -wOyn) of lightc. n

    "

    -rom the menn

    "

    -nOyt of flourn

    "

    -tSOyt of olivesn

    "

    -Sn

    "

    s of linenn

    "

    -swO of wheatd. n

    "

    -kOw (*N"

    -kOw ?) the othersn

    "

    -kem (*N"

    -kem ?) of Egypt(66) in (65a.), coronal syllabic nasals are assimilated to a following labial obstru-

    ent3, leading to a labial syllabic nasal. If the first consonant of the root is alabial glide [w], as in (65b.), assimilation doesnt take place. In (65c.), coronalsyllabic nasal keep coronal articulation before coronal obstruents.

    (67) possible representations of Coptic homorgany stated in (65):a. Left-branching b. Right-branching

    C V C0 V0 C1 V1- C2 V2

    m m e s

    C0 V0 C1 V1 C2 V2

    m m e s

    (68) the left-branching structure under (67a.) involves the clitic CV site. The right-branching structure under (67b.) doesnt involve this site as C1 spreads onhis right (the clitic site withered away, as stated by Lowenstamm (1999)). Theright-branching structure fails to predict assimilation since it violates the Prin-ciple of Free Occurence (Kaye 1990), which states that there are no restrictionson the melodic content of neighbouring onsets and nuclei.

    (69) an additional clue can be invoked throught the fact that there are no assimila-tions before /m/ filling the syllable nucleus (Till 1951:6567, Peust 1999:161 data from S):

    a. m"

    -mes the interestsm

    "

    -Ba:mp the goatsm

    "

    -p ew the heavensb. n-m

    "

    lO:h (*m-m"

    lO:h) the fightsn-m

    "

    sah (*m-m"

    sah) of crocodilen-m

    "

    to (*m-m"

    to) the deapth (of a sea)(70) the forms standing in (69b.) show that no assimilation takes place when the

    initial consonant of a stem is a syllabic consonant. As for examples in (67), ifwe represent syllabic consonants as right-branching, assimilation is predictedto occur in the forms given in (69b.):

    3However, assimilation sometimes fails to appear before the labial fricative /B/ in S: [n"

    -Bal] the eyes(*m

    "

    -bal), [n"

    -BaSOr] the foxes, etc.

    11

  • C0 V0 C1 V1 C2 V2 C3 V3

    m m s a h

    (71) if syllabic consonants are left-branching we can account for the fact that noassimilation takes place in (69b.) because of the syllabic /m/:

    C V C0 V0- C1 V1 C2 V2 C3 V3

    n m s a h

    *m

    (72) since V1 is filled by the following consonant, the Free Occurrence Principle isviolate and no assimilation occurs in this case. According to Till (1951), thefact that no assimilation take place can only be due to an intervening vowel(graphically rendered by supralinear stroke)4.

    (73) our analysis is supported by Scheer (2004: 707-41)s analysis of homorganic-ity according to him, a nasal in coda position (i.e. before an empty nucleus)is weak and will therefore assimilates and or spread5 in order "to consolidateits existence" (p. 710). This is what happens in (69)

    (74) Right-branching hypothesis predicts that in (69) the first nasal will stand in acoda position too, and will then be unstable Left-branching hypothesis pre-dicts that the nasal in (69) will stand in a strong position, since the followingnucleus is filled by the syllabic nasal, and then will not need to assimilate

    5 References

    Blaho, S. 2004. Syllabic consonants in strict CV. Masters thesis, Pzmany Pter Uni-versity at Piliscsaba.

    Bodomo, A. B. 1997. The Structure of Dagaare. Stanford Monographs in African Lan-guages. CSLI Publications

    Clark, J. & C. Yallop. 1995. An introduction to phonetics and phonology (2nd edition).Oxford: Blackwell

    Creissels, D. 1999. Remarks on the sound correspondences between Proto-Bantu andTswana (S.31), with particular attention to problems involving *j (or *y) and *i andsequences *NC. Bantu historical linguistics: Theoretical and empirical perspectives,Hombert, J.-M. & L. Hyman (eds.). Stanford: CSLI Publications. pp. 297-334.Karlsson Mukhanova, A. 2006. Rhythm and Intonation in Halh Mongolian. Lund: Stu-dentlitteratur.

    Kaye, J. 1990. Scoda licensing. Phonology 7. pp. 301 U330Kaye, J., Lowenstamm, J., and J.-R. Vergnaud. 1990. Constituent structure and govern-ment in phonology. Phonology 7. pp. 193-231

    Layton, B. 2000. A Coptic Grammar with Chrestomathy and Glossary : Sahidic Di-4Remember that Till (1951) claim that supralinear stroke denotes a vowel, not a syllabic consonant.5see Scheers analysis of German (Scheer 2004: 719-34

    12

  • alect. Porta LinguarumOrientalium 20. Harrassowitz, Wiesbaden

    Lin, Y.-H. 2005. Piro affricates: Phonological edge effects and phonetic antiedge ef-fects? In The Internal Organization of Phonological Segments, van Oostendorp M. &J.M. van de Weijer (eds.). Berlin : Mouton de Gruyter.Lin, Y.-H. 1997. Syllabic and moraic structures in Piro. Phonology 14. pp. 403-436

    Lowenstamm, J. 1999. The beginning of the word. Phonologica 1996, John Rennisonand Klaus Khnhammer. The Hague: Holland Academic Graphics. pp. 153-166

    Lowenstamm, J. 1996. CV as the only syllable type. In Current trends in Phonology.Models and Methods, Jacques Durand & Bernard Laks (eds.). Salford, Manchester:ESRI. pp. 419-441

    Nagel, P. 1966. Zur problem der konsonantischen silbentrger im koptischen. Zeitschriftfr gyptische Sprache und Altertumskunde, 92. pp. 76 U8Ngunga, A. 2000. Phonology and morphology of the Ciyao verb. Stanford: C.S.L.I.

    Nijen Twilhaar, J. & M. van Oostendorp. 2000. Feature Domains in Hellendoorn Dutch.ms.

    Noske, R. 1996. On the Relevance of Underlying Forms: or the genesis of schwa as anattributive adjective marker in the history of German. In Dam Phonology, Nespor, M.& N.S.H. Smith (eds.). HIL Phonology Papers 2. La Haye: Holland Academic Graph-ics. pp. 165-190

    Oostendorp, M. van. 2004. Variation in phonology: Theory and Evidence. ms.NationalResearch Training Seminar in Troms, Norway

    Oostendorp, M. van. 2001. Nasal Consonants in Variants of Dutch and Some RelatedSystems. Neerlandistiek.nl 01.08.

    Peust, C. 1999. Egyptian Phonology: An Introduction to the Phonology of a Dead Lan-guage. Peust & Gutschmidt: Gttingen

    Plumley, J. M. 1948. An introductory Coptic grammar (Sahidic dialect). Home & vanThal: London.

    Polotsky, H. J. 1933. Zur koptischen lautlehre ii. Zeitschrift fr gyptische Spracheund Altertumskunde 69. pp. 125 U9

    Rennison, J. R. 1993. Empty Nuclei in Koromfe: A First Look. Linguistique Africaine11, 35-65

    Ridouane, R. 2003. Suites de consonnes en berbre, phontique et phonologie. Thsede Doctorat, Universit Paris 3

    Scheer, T. in press a. On the Status of Word-Initial Clusters in Slavic (And Elsewhere).To appear in the proceedings FASL 15, the Toronto MeetingScheer, T. in press b. Syllabic and trapped consonants in (Western) Slavic: different butstill the same. Investigations into Formal Slavic Linguistics, edited by Gerhild Zybatow& Luka Szucsich. Frankfurt am Main: Lang

    Scheer, T. 2004. A lateral theory of phonology. Vol 1: What is CVCV, and why shouldit be ?. Mouton de Gruyter: Berlin.

    13

  • Sgral, P & Scheer, T. 1999. The Coda Mirror. ms. version 3.1

    Toft, Z. 2002. The phonetics and phonology of some syllabic consonants in SouthernBritish English. ZAS Papers in Linguistics 28. pp. 111-144

    Vergote, J. 1973. Grammaire copte, volume 1a,b. Peeters: Leuven

    Vycichl, W. 1990. La Vocalisation de la Langue gyptienne. BdE 16. IFAO, Le Caire

    Worrell, W. H. 1934. Coptic sounds. University of Michigan studies, humanistic series26. Ann Arbor, Chicago

    Zhang, J. 2006. The Phonology of Shaoxing Chinese. LOT Dissertation Series 123.Utrecht Institute of Linguistics.

    14