Three-consonant sequences67emphasised here that consonants flanking a trapped sonorant alwaysagree in voicing, e.g. [drv]al ‘wood-cutter’, [krt]añ ‘larynx’, etc. The latterobservation was one of the main reasons why Polish trapped sonorantscaught the attention of phonologists in the past. Many analyseshave been offered for the behaviour of trapped sonorants with respect tovoicing (Bethin 1984, Rubach and Booij 1987, 1990a, b, R u -bach 1996, 1997a, b, Gussmann 1992). Very briefly, trapped sonorantswere claimed to be transparent to voice assimilation or voice neutralisation.In the latter situation a voiced obstruent which occurs beforea word-final trapped sonorant undergoes word-final devoicing, althoughit is not final at all (24).(24)a.bó[pr] — bo[br]a¿u[pr] — ¿u[br]aka[tr] — ka[dr]ub.mieli[sn] — mieli[zn]aboja[‚] — boja[ó]imechani[sm] — mechani[zm]y‘beaver, nom.sg./gen.sg.’‘bison, nom.sg./gen.sg.’‘frame, nom.sg./gen.sg.’‘shallows, gen.pl./nom.sg.’‘fear, nom.sg./gen.sg.’‘mechanism, sg./pl.’It was observed that although an obstruent preceding a trapped consonantis underlyingly voiced, it undergoes neutralisation when the casemarker is zero. Moreover, despite the fact that in Polish the rule of voiceneutralisation applies to obstruents only, trapped sonorants were observedto undergo devoicing too, and the familiar example bóbr shouldactually be transcribed [bup °r], with the whole final cluster devoiced.Unsurprisingly, the same ‘transparency’ was observed in the word-internal/initialposition. The general conclusion was that obstruents flankinga sonorant always agree in voicing. In Polish, unlike in Czech, for instance,TrD clusters (where ‘D’ is a voiced obstruent) are not possible.However, the reason why Polish was such an eagerly studied languagewas that in this language voice assimilation across the sonorant can beboth regressive and progressive, e.g. [kref] — [k °rfi] — [krevné] ‘blood,nom.sg./gen.sg./relative’, [bref] — [brvi] ‘eyebrow, nom.sg./nom.pl.’,[jent °rka] — [jendrek] (a name, gen.sg./nom.sg.). In the case of [kref] thefinal fricative must be voiced phonologically, vis. /v/, because of the form[krevné]. Since /v/ in /krev/ is reduced to [f] due to the general rule ofword-final neutralisation, the one in [k °rfi] must arise due to a progressiveassimilation. In other words, the devoicing of /v/ must be ascribed tothe presence of the initial [k] that precedes the trapped [r]. On the other5*
68 The phonological nature of the beginning of the wordhand, the form [jent °rka] demonstrates regressive assimilation, thus, in[jendrek] both the dental plosive [d] and [r] are voiced. However, thecluster [dr] undergoes devoicing when it occurs in a direct contact withthe following voiceless [k]. The voice value of [k] imposes its voicelesscharacteristic on the entire cluster. It must be clarified here that theidentical voicing specification of obstruents separated by the trapped [r]does not constitute any problem for models like GP or the Strict CV framework,where the sonorants lack the laryngeal element. Moreover, in[bup °r] the bilabial plosive /b/ is devoiced due to the following empty nucleus,which cannot license the laryngeal node in Polish. Finally, the explanationof the devoiced character of /r/ both in the former and the lattercontext, rather than being a pure phonological phenomenon, falls underthe scope of phonetics. 18Additionally, it should be noted here that the forms [jent °rka] – [jendrek]are peculiar in another respect. Specifically, in the previous analysestrapped sonorants were claimed to be extrasyllabic (Rubach andBooij 1990a, Rubach 1996, 1997a, b). However, as pointed out byScheer (2004), in all studied cases extrasyllabicity and extrametricallityoccur only at word-edges. This observation follows from the fact thatword-margins behave in a peculiar, often unpredictable way. If trappedsonorants are extrasyllabic, Polish is the only language which allows forword-internal extrasyllabicity, e.g. Jêdrka [jent °rka] (a name), piosnka[pjos °nka] ‘song’, czosnku [t¡Sos °nku] ‘garlic, gen.sg.’ It follows that previousaccounts relying on the syllabification algorithm left word-medialtrapped sonorants such as [n] in czosnku unsyllabified. Note that [n] canneither form a branching onset with the following [k], nor qualify asa second member of a branching coda [sn]. Both options violate the SonoritySequencing Principle. Needless to say, the same argumentation appliesto word-final TR clusters such as [p °r] in bóbr. The latter clustercannot be interpreted as a branching onset as such constituents do notoccur word-finally. The problem was solved by assigning the [b] to thecoda, which left the final [r] unsyllabified. This analysis was then extendedto cover word-initial trapped sonorants. The representation of[jent °rka] and [bup °r] is given below in (25).In (25a) the extrasyllabic character assigned to the word-internal [r]is the reason why the flanking obstruents agree in voicing. Generallyspeaking, extrasyllabicity explains the transparent character of the wordinternaltrapped sonorants. Both obstruents, i.e. /d/ and /k/, are adjacentat the early stage of syllabification, hence voice assimilation occurs. Thenthe unsyllabified segment is linked to the higher prosodic constituent,18I owe this note to Eugeniusz Cyran
- Page 2 and 3:
Polish and EnglishConsonantal Clust
- Page 4 and 5:
Artur KijakPolish and EnglishConson
- Page 6 and 7:
ContentsPreface . . . . . . . . . .
- Page 8 and 9:
PrefaceThe phonotactic peculiaritie
- Page 10 and 11:
Preface92000), Ploch (1999), van de
- Page 12:
List of abbreviationsBrODIdim.FODge
- Page 15 and 16:
14 The frameworkemploying the simpl
- Page 17 and 18: 16 The frameworksion in section 3 b
- Page 19 and 20: 18 The frameworkmodel is able to ha
- Page 21 and 22: 20 The frameworkhanan 1986). Thus,
- Page 23 and 24: 22 The frameworkare not derived at
- Page 25 and 26: 24 The frameworkWhat is interesting
- Page 27 and 28: 26 The frameworklateral relations,
- Page 29 and 30: 28 The frameworkIn general, we can
- Page 31 and 32: 30 The frameworkobstruents followed
- Page 33 and 34: 32 The frameworkLet us look more de
- Page 35 and 36: 34 The framework(7) PGO N O N O N O
- Page 37 and 38: 36 The frameworkby all sounds. Thus
- Page 39 and 40: 38 The frameworkexist. What is a Br
- Page 41 and 42: 40The frameworkLowenstamm’s (1999
- Page 43 and 44: 42 The frameworksky and Halle’s (
- Page 45 and 46: 44 The phonological nature of the b
- Page 47 and 48: 46 The phonological nature of the b
- Page 49 and 50: 48 The phonological nature of the b
- Page 51 and 52: 50 The phonological nature of the b
- Page 53 and 54: 52 The phonological nature of the b
- Page 55 and 56: 54 The phonological nature of the b
- Page 57 and 58: 56 The phonological nature of the b
- Page 59 and 60: 58 The phonological nature of the b
- Page 61 and 62: 60 The phonological nature of the b
- Page 63 and 64: 62 The phonological nature of the b
- Page 65 and 66: 64 The phonological nature of the b
- Page 67: 66 The phonological nature of the b
- Page 71 and 72: 70 The phonological nature of the b
- Page 73 and 74: 72 The phonological nature of the b
- Page 75 and 76: 74 The phonological nature of the b
- Page 77 and 78: 76 The phonological nature of the b
- Page 79 and 80: 78 The phonological nature of the b
- Page 81 and 82: 80 The phonological nature of the b
- Page 83 and 84: 82 The phonological nature of the b
- Page 85 and 86: 84 The phonological nature of the b
- Page 87 and 88: 86 The phonological nature of the b
- Page 89 and 90: 88 The phonological nature of the b
- Page 91 and 92: 90 The phonological nature of the b
- Page 93 and 94: 92 The phonological nature of the b
- Page 95 and 96: 94 The phonological nature of the b
- Page 97 and 98: 96 The phonological nature of the b
- Page 99 and 100: 98 The phonological nature of the b
- Page 101 and 102: III. Bogus clusters, syllabic conso
- Page 103 and 104: 102 Bogus clusters, syllabic conson
- Page 105 and 106: 104 Bogus clusters, syllabic conson
- Page 107 and 108: 106 Bogus clusters, syllabic conson
- Page 109 and 110: 108 Bogus clusters, syllabic conson
- Page 111 and 112: 110 Bogus clusters, syllabic conson
- Page 113 and 114: 112 Bogus clusters, syllabic conson
- Page 115 and 116: 114 Bogus clusters, syllabic conson
- Page 117 and 118: 116 Bogus clusters, syllabic conson
- Page 119 and 120:
118 Bogus clusters, syllabic conson
- Page 121 and 122:
120 Bogus clusters, syllabic conson
- Page 123 and 124:
122 Bogus clusters, syllabic conson
- Page 125 and 126:
124 Bogus clusters, syllabic conson
- Page 127 and 128:
126 Bogus clusters, syllabic conson
- Page 129 and 130:
128 Bogus clusters, syllabic conson
- Page 131 and 132:
130 Bogus clusters, syllabic conson
- Page 133 and 134:
132 Bogus clusters, syllabic conson
- Page 135 and 136:
134 Bogus clusters, syllabic conson
- Page 137 and 138:
136 Bogus clusters, syllabic conson
- Page 139 and 140:
138 Bogus clusters, syllabic conson
- Page 141 and 142:
140 Bogus clusters, syllabic conson
- Page 143 and 144:
142 Bogus clusters, syllabic conson
- Page 145 and 146:
144 Bogus clusters, syllabic conson
- Page 147 and 148:
146 Bogus clusters, syllabic conson
- Page 149 and 150:
148 Bogus clusters, syllabic conson
- Page 151 and 152:
150 Bogus clusters, syllabic conson
- Page 153 and 154:
152 Conclusionnisms available in th
- Page 155 and 156:
154 ReferencesBotma, B. (2004) Phon
- Page 157 and 158:
156 ReferencesGussmann, E. (1998) D
- Page 159 and 160:
158 ReferencesPawelec, P. (1989) Cy
- Page 161 and 162:
160 ReferencesScheer, T. (1997) Vow
- Page 164 and 165:
Author indexAbercrombie, David 103A
- Page 166 and 167:
Artur KijakGrupy spó³g³oskowe w
- Page 168 and 169:
Zusammenfassung167für alle anderen