Abstract
One of the remarkable features of the phonology of Imdlawn Tashlhiyt is its syllable structure, which allows any segment to be a syllable nucleus. This feature is not without precedent in the phonological literature, but it is sufficiently rare to require a detailed justification. Among the languages which have been claimed to have syllabic obstruents, those which have received most attention from a theoretical point of view are languages spoken on the Northwest Pacific coast of North America, e.g. Bella Coola, and certain Mon-Khmer languages spoken in Laos and the Malay peninsula, see Bagemihl (1991) and Shaw (1993) and references therein.1 In this chapter and the next we determine the basic inventory of syllable types in Tashlhiyt and we discuss the role of sonority in assigning syllable structure to sequences of segments. Syllabic consonants will be the focus of Chapter 6.
Access this chapter
Tax calculation will be finalised at checkout
Purchases are for personal use only
Preview
Unable to display preview. Download preview PDF.
Notes
Closely related issues are raised by certain members of the Northwest Caucasian and Yuman families, see Anderson (1978).
See Elmedlaoui (1985) and DE (1985, 1988, 1996a, b).
This claim has already been made about other dialects, see Applegate (1958: 13) on Ifni Tashlhiyt: ‘all consonants in certain environments have syllabic allophones’, and Mitc]hell (1957: 198) about the dialect spoken in Zuara (Libya): ‘If the statement is at all meaningful at the phonetic level, any “consonant” may be syllabic in Berber’.
For each example we give both the broad phonetic transcription used throughout this book and an IPA transcription.
The prefix /I-/ is the reflex of the definite article of Arabic, v. § 2.5.3.1.
Here is the meaning and the morphemic analysis for each expression in (6): (a) 3ms-collide ‘he collided’, (b) 3fs-stingy=even’ she is even stingy’, (c) f-u-gazelle-fs ‘female gazelle’, (d) 2-surround-2mp ‘you (mp) surrounded’, (e) pick=do3ms 2s-eat:aor-2s=do3ms ‘pick it and eat it’.
At the end of (6)c the underlying sequence /d-t/ surfaces as the geminate tt, in which the first skeletal position is a nucleus.
E.g. [ö] in (6)d, which is a fronted variant of /u/. On u fronting, v. § 3.8.
Here are the meanings of the verbs in column I: (a) ‘jail’, (b) ‘curse’, (c) ‘keep the Sabbath’, (d) ‘become cheap’, (e) ‘grant’ (the subject is usually God), (f) ‘be strong’. The Imdlawn Tashlhiyt words on the right-hand side have meanings identical with those of their MA counterparts.
In lines a, c and e, the medial consonant of the Berber words optionally devoices, see § 6.4.2 on regressive devoicing in Imdlawn Tashlhiyt.
In free variation with Ižhd.
The verbs are in the 2s pf imperative.
V. § 2.5.3.1.
The mishandling of schwas occurs prominently in the skits of Abdeljabbar Luzir and Ahmad Belqas, a comic duo well-known during the Sixties and the Seventies who had made a specialty of imitating the broken MA of Ashlhiy shopkeepers.
The words in (10) are 2s imperative verbs, i.e. bare aorist stems. Here are the meanings of the verbs in (10): (a) ‘be dried out’, (a’) ‘hunt’, (b) ‘add’, (b’) ‘gather firewood’, (c) ‘plough’, (c’) ‘hide’, (d) ‘choke’, (d’) ‘drown’.
Armenian is a language in which vowel epenthesis is motivated by sonority-driven syllabification, see Vaux (1998).
Louali and Puech (1999) and Puech and Louali (1999) report on an experiment in which Tashlhiyt speakers were presented with pairs of Tashlhiyt words and asked to judge whether the words were rhythmically alike or different. The interpretation of their results is unclear.
This fact was already pointed out in DE (1988).
During a conversation a well-known line may be quoted on account of its content, as one quotes proverbs.
V. also the description of a-ħwaš in Basset (1952/1987: 87). On aħwas, a rural party with dances, see Schuyler (1979: 49-52).
From underlying /l-lrays/. The plural is !rrways and the feminine singular !tarrayst. On the Irrways, their music and the social and economic background of their activity, see Schuyler’s (1979) outstanding work.
See Stroomer (1992) and Boogert (1997).
See the appendices at the end of this book for musical scores of Ashlhiy tunes and words sung to these tunes.
The music and words of the song can be found in Davenson (1955: 581).
Actually this is a simplification. It would be more accurate to say that each number represents a point in time, i.e. the onset of a note (v. Cornulier 1995: 116-120, 280) or a musical beat (v. Hayes and Kaun 1996). Text-to-tune alignments are effected primarily by pairing up the metrical structure of the text with the rhythmical structure of the tune.
The implicit conventions which govern text-to-tune alignment in French nursery rhymes favor a ‘feminine schwa’ on the sixth note of tune (13). montre may be pronounced with such a vowel in its final syllable, whereas stylo cannot.
Malone (1996: 124) uses the term ‘orthometric’ to ‘denote the set of systematic euphonic patterns deployed by a given language-cum-tradition in the verbal arts.’
On the versification of melħun, see Chapter 8.
These are the first lines of a poem by Mohmmad Andmsir in Amarir (1975: 132ff). ‘A sick man needed a little honey / He could not gather any (from a hive) nor reach any / All in tears he went to see the owner of the hive’. Our translations do not seek elegance and they are sometimes rather approximative. We give them only to allow those who know Berber to identify morphemes and syntactic structures.
V. below on the phonological differences between the forms of language used in speaking and in singing.
Actually this is a simplification. The music on which ME sang this particular piece in order to retranscribe it requires lines to be grouped in couplets. The tune for the first line of a couplet is different from that for the second line, but both tunes have the same rhythmical structure. Each column in (19) thus corresponds to a given point in that rhythmical structure. The score of the tune in question is given at the end of this book, see Tune 1 in Appendix IV.
These lines are respectively line 69 in the preceding piece and lines 33 and 46 in that of Appendix II. Here is the text of the first line: i-ga zund 1-fqqih i-!ttfar ukan 1-ħaqq ‘like the man of learning, he is indeed entitled’.
Nowhere does one find an answer to that question in Jouad’s work. In his various publications, Jouad tries to answer the following questions, which are different: (i) Given a metrical pattern and a line of verse which matc]hes that pattern, how does one parse the line into orthometric syllables? (ii) Given a sequence of lines which all matc]h the same metrical pattern, how does one discover what that metrical pattern is? Answering our question also answers Jouad’s, but the converse is not true.
See Selkirk (1981).
See Kaye and Lowenstamm (1984). For more recent references see e.g. Harris and Kaisse (1999).
Appendix III, syllable 4 in line 64.
Appendix III, syllable 8 in line 78.
Appendix II, syllable 7 in line 49.
In the first columns of tables (28) and (47) all geminates are indicated by tildes. Elsewhere only heterosyllabic geminates are indicated by tildes.
’stu tutut / Stu tutut / Salt is expensive / Making ends meet is difficult’ (literally ‘the hearth is difficult’).
(a) i³=as ur gum-nt t-fraw-in=a t-zdm i-³al-n, ‘if these twigs are not enough, let her chop armfuls’; (b) !allahukbar d-duni-t=inu t-!štn ul=inu, ‘Allahoakbar! Life here below troubles my heart’.
This meter requires final syllables with ‘compound rimes’. On syllables with compound rimes, see the end of this section.
(a) i³ ur i-!rsi 1-hna t-fk-t a-!dar i=!lutan, ‘if peace does not settle, begin your journey’ (literally: ‘give the foot to the lands’); (b) ixar inn ³=ak i-ban s-!sfa t-ssu-t a-kal, ‘where serenity comes to you, there you should take the ground as your carpet’; (c) t-a-!nqqzbrud-t ad=ax=as t-s³a lalla=tnx, ‘it is a small poncho which our mistress has bought him’.
Lines 19, 20 and 69 of the poem by Hmad Biyzmawn in Amarir (1975: 139-143). (a) Ah! Which mountain did I not comb, which plains!; (b) I have badgered the scholars and the marabouts; (c) I’ve crossed the seas and I’ve been all over the skies and the lands.
V. Galand-Pernet (1969).
In lines 12 and 56 of the poem in Appendix III (but not in line 59) the first half of the initial geminate is ‘left out’ of the parse. ME finds such violations worse when the geminate involved is a sonorant than when it is an obstruent.
For other examples, see for instance the piece recorded in Jouad (1995: 193).
Goldsmith and Larson (1992), Prince and Smolensky (1993), Scobbie (1993), Zec (1995), Shaw (1996), Clements (1997), Frampton (1999).
Implicit in table (28).
On sonority and its role in syllabification, see Clements (1990) for a careful discussion and references.
On the differences between the sonority scale in (35) and that used in our earlier work, see below the end of § 4.9.1.
V. text under (23) in § 4.6.
n is not a sonority peak either since it is adjacent to i. The fact that i does not belong to the same syllable as n is irrelevant, as will be explained below.
SonPeak is Clement’s Sonority Peak Principle (p. 303), slightly reworded.
(40)a,b,c,d are respectively lines 1, 18, 20 and 21 in the song by El-hajj Belaid in Mestaoui (1996: 38ss). Here are translations. ‘He who hunts till nightfall without catc]hing anything / I am ill and physicians have lead me astray / Having agreed on the symptoms, they declare / Unfortunate! Your illness is that of love’.
Every line in this song ends with a stop-gap vowel i, a common occurrence in Ashlhiy singing. The final i has been left out from our transcriptions.
From /yan w-!attan/.
From /n=l-ħubba/. /a#a/ reduces to a short a, see the text below (17).
From /ad=fllak/, see the text under (17).
This line is ill-formed: its sixth syllable should be heavy.
The same situation obtains later on in the same line, when /ibnžla/ is parsed as i.bnž.la rather than as *ib.nž.la.
In Dell and Tangi (1993) the same constraint is posited to prevent M from being turned into a in certain contexts.
It is not obvious that it should be so. For instance if closed syllables with an obstruent nucleus were universally excluded, as proposed by Shaw (1996), and if SonPeak were inviolable, such a sequence as nakzdma could only be parsed with d left unsyllabified.
‘Owner of the threshed grain, let God bless the threshing floor’.
/ad/ underlyingly. On the consonant in /ad/ and /rad/ see DE (1989: 188-189).
For observations on French songs which were made using the same method, see Dell (1989).
In Jouad (1995) one finds mu.nnt (line 4 p. 183 and line 4 p. 196) and li.mmk (line 8 p. 220). These parses do not contradict our assertion since in the sequences in question ‘nn’ et ‘mm’ are not geminates, see later in this section. On the first form v. DE (1985: 128 note 46).
This constraint was already discussed in our 1985 work and in Dell and Tangi (1992: 132-133).
Line 151 in a song composed by ME, to appear. ‘Have you already forgotten the one-stringed violin and the tuning fork?’.
Lines 217 and 218 in the song cited in the preceding note. ‘Patience is a fine virtue as long as it does not go together with humility / Patience and faith are the foundations on which houses and hearths stand’.
Underlyingly /I-!sbr/. The total assimilation of /l-/ to the following coronal gives rise to a geminate, v. § 2.5.3.1.
/ad=f/ in the underlying representation.
The metrical pattern required by the tune allows the seventh syllable of a line to be either L or H.
‘By and large’: see Chapter 7 on the violations of SonPeak and NoRR which are due to underlying glides.
‘Whoever carries flowers and is followed by flies’. This line was coined by ME. From now, on lines for which no reference is given are lines invented by ME.
‘He has always led herds and great riches’.
‘Alas! Brahim has exiled himself here, he has fled here all in tears’. Note that in syllable 8 the glide w violates SonPeak. See Chapter 7 about such violations.
‘The unfortunate whose tooth gives him a rough time will never find relief’.
NoPICOR: No Plateaux In Complex Obstruent Rimes.
See Elmedlaoui (1985), DE (1985, 1988).
‘But the owner was a callous man; he did not deign look at him.’
See also syllable 10 in (33)b.
(a) a y-ukr 1-kas ³=t-!uzzum-t nm a t-a-ħanu-t, ‘he stole the glass from the back of the shop’; (b) bihi y-ukr 1-kas ur=as=gisn t-a-yafu-t, ‘Bihi stole the glass; it is of no use to him’.
‘Ouch! the one whose tooth aches and who, on top of that, is in trouble’.
This corrects DE (1997a: 46), who stated incorrectly that al.ms.ki is unacceptable.
a ymma 1-mskin ur=akwkw y-ufi t-a-glla-t ‘Ah! while the poor man does not even find a turnip’.
Author information
Authors and Affiliations
Rights and permissions
Copyright information
© 2002 Springer Science+Business Media Dordrecht
About this chapter
Cite this chapter
Dell, F., Elmedlaoui, M. (2002). Tashlhiyt Syllables I. In: Syllables in Tashlhiyt Berber and in Moroccan Arabic. Kluwer International Handbooks of Linguistics, vol 2. Springer, Dordrecht. https://doi.org/10.1007/978-94-010-0279-0_4
Download citation
DOI: https://doi.org/10.1007/978-94-010-0279-0_4
Publisher Name: Springer, Dordrecht
Print ISBN: 978-1-4020-1077-4
Online ISBN: 978-94-010-0279-0
eBook Packages: Springer Book Archive