the phonetics and phonology of emphatics in mehri janet c.e. watson & alex bellem workshop on...
TRANSCRIPT
The phonetics and phonology of emphatics in Mehri
Janet C.E. Watson & Alex Bellem
Workshop on Pharyngeals & PharyngealisationNewcastle University – 26 March 2009
Overview
BackgroundModern South Arabian (MSAL)Emphatics across SemiticEmphatics in MSAL?
DataMehri evidence
Previous descriptive accountsFieldwork
Native speaker viewsAcoustic analyses
Prepausal glottalisation in Ṣan’āni ArabicPhonological patterning
Summary, conclusion, further work
Modern South Arabian (MSAL)
Semitic languages
Based on the work of Hetzron (1972, 1974). See Faber (1997), Bennett (1998), Appleyard (2003) for overviews of Semitic classification. See Corriente (1996) for a recent alternative classification.
Emphatics in Semitic…
Proto-Semitic is generally reconstructed as having a series of consonantal triads
Proto-Semitic
Emphatics in Semitic…
These consonantal triads have a voiced member and a voiceless member, with the third member being ‘emphatic’
The ‘emphatic’ member has been the subject of controversy ‘backed’ (Arabic-style)
ejective
It is now increasingly hypothesised that the early Semitic emphatics (inherited from Afroasiatic) were ejectives
Our work (on Arabic, not just Mehri) supports this
Under such a system, ‘emphatic’ is a laryngeal (phonation) contrast in obstruents
Proto-Semitic
Emphatics across modern Semitic…
Ethiosemitic Emphatics are ejectives, and thus ‘emphatic’ is a third laryngeal contrast
(voiced–voiceless–emphatic) Neo-Aramaic
Dialects vary – ‘trajectory’ of emphatic development can be traced through different dialects*
Most often ‘emphatic’ seems to be realised in some kind of ‘backing’ feature alongside non-aspiration (voiceless non-emphatics are aspirated)
Arabic Generally, most salient correlate is ‘backing’ (uvularisation /
pharyngealisation) Certain dialects (/dialect types) show remnant of ‘laryngeal’ function, so
there is dialectal variation in terms of the development of emphatics (2-way vs 3-way phonation systems)**
MSAL Where do they fit into this typology?
* Dolgopolsky (1977)** See Watson & Bellem (in press), Bellem (2007); also see Heselwood (1996)
Emphatics in MSAL: the literature Works based on fieldwork in the first half of the 20th C generally describe MSAL emphatics
as similar to those of Arabic (but less salient)
Viennese expedition in the early 20th C (e.g. Jahn 1902, Müller 1909, Bittner 1909)
Bertram Thomas’ fieldwork (Thomas, 1937)
Wolf Leslau’s work (1947), based on Thomas (1937)
Leslau also notes, in the discussion following Johnstone 1975, his view that ‘[MSAL] glottalization did not sound to me to be of the same type as that in Ethiopic’
T.M. Johnstone’s work in the 1970s breaks with this tradition – he describes MSAL emphatics as (post-)glottalised
For Harsusi, he observes that glottalisation is ‘energetically articulated in initial and final position, but in other positions…rather weak’ (1977, see also 1975, 1987)
Marie-Claude Simeone-Senelle & Antoine Lonnet, 1983–
Simeone-Senelle (1997) describes the ‘prevailing’ articulation of MSAL emphatics as post-glottalised, but ‘The degree of this glottalization varies, depending on the position of the consonant in the word and on the dialects concerned…’
Lonnet (2009) states that emphatics in some MSAL dialects tend to be pharyngealised–uvularised, and sees this in terms of a gradual sound change
Russian researchers in Soqotra
Naumkin & Porkhomovskij (1981) say that glottalisation in Soqotri is restricted to the emphatic stops, with the emphatic fricatives (and occasionally stops) pharyngealised – they suggest that there is an ongoing transition from glottalised to pharyngealised
Question: What are the MSAL emphatics?
Until 1973, Ethiosemitic was believed to be the only Semitic language sub-family in which
emphatics were realised as ejectives
From the 1970s we find varying reports
Without assuming that emphatics are uniform across MSAL varieties, we can still say that
overall there is no clear consensus, and it is not clear from the literature where the MSAL emphatics fit into the emphatic typology
Question: What are the MSAL emphatics?
This paper takes a closer look at one variety of an MSAL – the Mahriyōt dialect
of Mehri, spoken in eastern Yemen
We aim to show why there has been such a lack of consistency in previous
descriptions
Mahriyōt: Data
111 texts recorded by Alexander Sima, to be published by Harrassowitz 2009
Oral material recorded and transcribed by Janet Watson in al-Ghaydha Jan–March 2008
5 oral narratives
Oral descriptions and examples of emphatics and laterals in Mahriyōt
Mahriyōt consonants
lab dent alv pal-alv pal vel uvu phar glott
PLOSVE
vcd b d j
vceless t k ʔ
emph ṭ ḳ
FRICTVE
vcd ð z ġ~q ʕ
vceless f θ s š x ћ h
emph θ$ ṣ č̣$
LATERAL
vcd l
vceless ś
emph ź
sonorants m n r y w
Mahriyōt emphatics
emphatic non-emphatic
ṣ s z
č̣ š
ṭ t d
ź ś
θ θ ð
ḳ k
Mahriyōt evidence
Inconsistency of much previous descriptive work
E.g. transcriptions vary considerably and are inconsistent even within one work
Hein, ed. Müller (1909)
/ḳ/ as g and k: ġalgōt ‘she saw’ for ġalḳōt, occasionally as k, as in tekefôd ‘she goes down’, but ugofôd ‘and he went down’
/ṣ/ as z, ṣ and s: zóṭer ‘basket’ for ṣōṭar, zayd ‘fish’ for ṣayd, but also as ṣ and s, particular in the word for ‘morning’, as in kṣôbaḥ and hesôbaḥ ‘in the morning’
The inconsistencies most often relate to laryngeal category
Mahriyōt evidence: fieldwork
Native-speaker descriptions:
Five sounds not attested in Arabic: /ḳ/, /ṣ/, /č̣ S/, /ś/, /ź/
/ḳ/ = /k/ + ʕayn
/č̣ S/ = ‘heavy Egyptian jīm’ + ʕayn
Mehri /ṣ/ considerably tenser than Arabic /ṣ/, sometimes partially voiced
/ṭ/ and /θ S/ not grouped by informant with ‘five sounds not found in Arabic’
/ṭ/ = ‘same’ or ‘like’ Arabic /ṭ/
/θ S/ = varies between voiced and voiceless
/ṭ, ṣ, č̣ S/
ṣā’ / ṣā’ / wa-ṭā’ wa-č̣ ā / yaʕnī / anṭughā bi-šakl at-tālī /
ṣā’ ṣā’ ṣā’ ṣā’ / wa-ṭ-ṭā’ hūwa nafs aṭ-ṭā’ fi l-... bi-l-
ʕarabīyah / ṭā’ / allī hū ʕalayh al-ʕūd hass-mā ygūlū / e:h
ṭa’ / ṭā’ ṭā’ ṭā’ ṭā’ ṭā’ / wa-ṣā’ ṣā’ ṣā’ / wa-ḥarf č̣ āʕ / e:h č̣ āʕ
č̣ āʕ č̣ āʕ / allī hēh kama l-jīm bi-l-ʕarabī wa-taḥthā θalāθa
nugaṭ tarmīz hāðā kama ttafagnā ʕalayh fī ’almāniyā
iθnā tadrīsnā fī almāniyā maʕ al-jānib al-almāni / wa-hī
tunṭug miθl al-jīm al-maṣrīyah aθ-θagīlah wa-l-ʕayn / č̣ āʕ
/ č̣ āʕ č̣ āʕ / yaʕnī law bayn-axað maθāl maθalan al-ḥarf
ṣād / maθalan / ana xað amθilah bi-xtiṣār / ṣā’ / ṣā’
maθalan ṣayd / ṣā’ ṣayd / ṣā’ ṣayd ṣayd / maθalan ṣā’
bi-ḥarf ṣā kīnaḥ maθalan / ṣift / ṣift / ṣift / maθal .../
Native-speaker judgements:
Rejection by native speakers of ejective tokens of /ṣ/, /č̣ S/, /ṭ/, /ź/, /θ S/
…except in pre-pausal position
Let’s look more closely at the Mahriyōt emphatics and consider the acoustic evidence…
Mahriyōt evidence: fieldwork
w-tarnīk ‘and tarnīk [type of fish]’
w-wīḳad ‘and wīḳad [type of fish]’
ḳannatt ‘small’
w-ō-ð-alhōḳ ‘and I am chasing’
śīwōṭ ‘fire’
ћaṭṭōt ‘a bean / grain’
Mahriyōt evidence: conclusion I
Of the ‘emphatic’ stops:
only ḳ turns out to be ejective in all positions
ṭ is ejective only in final and prepausal position
otherwise, ṭ is similar to (the local) Arabic ṭ
non-prepausally, ṭ is ‘backed’ and unaspirated
This ‘backing’ is also the main correlate of ‘emphatic’ in the other (fricative) emphatics…
/ś/ and /ź/
Time (s)0 0.84517
0
5000
Time (s)0 0.466301
0
5000
F2
F2
śātū ‘winter’
źābal ‘cold’
bā nwās ‘Abu Nuwas’
F2
xalāṣ ‘that’s it!’
F2
marwōź ‘sick [m.pl.]’
krōṣ ‘fleas’
č̣ aʕrīr ‘back of the neck’
yā šadd ‘oh what trouble!’
Importance of environment!Pre-pausal glottalisation
Voiced and emphatic obstruents:
ġayj > ġayč̣’# ‘man’
yanhōč̣ > yanhōč̣’# ‘he shouts to s.o.’
ṭād > ṭāt’# ‘one’
śīwōṭ > śīwōt’# ‘fire’
ṣwārāb > ṣwārāp’# ‘harvest period [dim.]’
mōnaġ > mōnax’# [place name]
Liquids (after long vowel):
syōr > syōr’ # ‘he went’, b-ḥāwēl > b-ḥāwēl’ # ‘firstly’
…but:
šīt > šītʰ # ‘penis’ and yaṣkūk > yaṣkūkʰ # ‘he closes’
ṣwārāb ‘harvest period [diminutive]’
b-ħāwāl’ ‘at first’
Pre-pausal glottalisation
With the exception of ḳ, the emphatics are only clearly glottalised prepausally, otherwise they are ‘backed’
This glottalisation seems to be part of a wider process of prepausal glottalisation, which affects certain segment types:
voiced obstruents
emphatics
liquids in the environment VVL#
Pre-pausal glottalisation
Areal feature
Feature of many Yemeni Arabic dialects
In prepausal position in Ṣan’āni Arabic:
all voiced consonants are devoiced
all voiced stops and emphatics are realised as ejectives
all (non-nasal) sonorants are devoiced and glottalised
Yemen
Ṣan’āni dajāj ‘chicken’
Glottal release
Ṣan’āni dagīg ‘flour’
Glottal release
Ṣan’āni mubargaṭ ‘lumpy’
Glottal release
Ṣan’āni nār ‘flame’
Mehri emphatics: phonological patterning
Mehri emphatics pattern with pharyngeals and uvulars when it comes to vowel allophones Mahriyōt ay and aw may occur to the exclusion of ī and ū
following an emphatic, uvular or pharyngeal:
ba-ḥḥays ‘with energy’
ḳayṯ̣ ‘hot/pre-monsoon period’
ʕayd ‘sardines’
ʕaylūj ‘camel calf’
ḥayḏān ‘ear’
mṣawġat ‘jewellery shop’
śaṭrayr ‘cloth’
(strict adjacency not necessary)
Mehri emphatics: phonological patterning
Allomorphy
feminine nominal, adjectival and numeral ending:
normally -īt, but -ayt in certain words:
ṣarʕayt ‘smell under the armpits’ (cf. šabdīt ‘liver’)
bīźayt ‘egg’ (cf. rēśīt ‘snake’)
habʕayt ‘seven’ (cf. ṯ̣amnīt ‘eight’)
ṣalḥayt ‘fat f.s.’ (cf. xaṯ̣mīt ‘thin f.s.’)
Mehri emphatics: phonological patterning
Allomorphy
nominal feminine suffix:
-āt after an emphatic, uvular or pharyngeal, but -ēt after other segments (except nasals):
ḳaṣṣāt ‘story’ barzēt ‘small hole in boat to let water out’
mṭarḳāt ‘hammer’ raḥbēt ‘village; town’
ṣafḥāt ‘hinge’ mbaxrēt ‘iron frame for incensing clothes’
xabzēt ‘piece of bread’
ḳaśrēt ‘naughtiness’
Summary I
Emphatics:
/ḳ/ = ejective
All other emphatics = non-ejective
There is a process of prepausal glottalisation that affects (among other segments) the emphatics
Why the inconsistency in the reporting of Mehri (possibly MSAL) emphatics?
The perception of ejective emphatics in one position (pre-pausally)
Presence of one ejective – /ḳ/ – in all positions
Assumption that ‘emphasis’ had one main phonetic correlate
= assumption that emphatics as a class were ejectives
Summary II
With the exception of ḳ, Mahriyōt emphatics are phonologically (‘underlyingly’) of the (local) Arabic type:
‘backed’
unaspirated (in the case of voiceless emphatics)
(…although the Mahriyōt emphatics are less ‘backed’ than Arabic emphatics tend to be)
Implications and further work I
What about other MSAL?
Our data is from Mahriyōt, but we believe that the situation may be similar in other MSAL
First impressions are: that other MSAL varieties may also have some degree of
prepausal glottalisation (not just either ‘ejective emphatics’ or freely varying emphatics)
the degree of any prepausal glottalisation may vary across the individual varieties, and that aside from the possibly varying prepausal phenomena, the phonetic correlates of /ṭ/ may also vary
Implications and further work II
It seems that there is a move across MSAL from ejective emphatics > pharyngealised emphatics
For Soqotri, two types of ‘emphasis’ are described:
ejective stops
pharyngealised fricatives and occasionally stops
(Naumkin & Porkhomovskij 1981: 12–13)
Conclusion III: postscript
Our view of the emphatics, in at least this dialect of Mehri, is thus the flip-side of Johnstone et al’s coin
We claim that they are lexically of the ‘backed’ type, but glottalised prepausally (as part of a wider process of prepausal glottalisation)
This does not negate the claim that MSAL emphatics were historically ejectives – our claim relates to the synchronic status of the emphatics, and thus their phonological identity
Selected References
Johnstone, T.M. 1975. The Modern South Arabian languages. Afroasiatic Linguistics 1/5: 93–121.
Ladefoged, P. & I. Maddieson. 1996. The Sounds of the World’s Languages. Oxford: Blackwell. Leslau, W. 1947. Four Modern South Arabian languages. Word 3: 180–203. Lonnet, A. 2009. South Arabian, Modern. In K. Versteegh (ed.) Encyclopedia of Arabic
Language and Linguistics. Vol. IV. Maddieson, I. 2008. Glottalized Consonants. In: M. Haspelmath, M.S. Dryer, D. Gil & B. Comrie
(eds.) The World Atlas of Language Structures Online. Munich: Max Planck Digital Library. Naumkin, V.V. & V.Y. Porkhomovskij 1981. Ocherki po etnolingvistike Sokotry. Moskva. Sima, A. (in press) (edited, introduced and annotated by J.C.E. Watson & W. Arnold). Mehri-
Texte im Dialekt der jemenitischen Šarqiyyah.Wiesbaden. Simeone-Senelle, M-Cl. 1997. The Modern South Arabian languages. In R. Hetztron (ed.), The
Semitic Languages. London: Routledge. 378–423. Watson, J.C.E. & Y. Asiri. 2007. Pre-pausal devoicing and glottalisation in varieties of the south-
western Arabian peninsula. ICPhS. Watson, J.C.E. & A. Bellem (in press). Glottalisation and neutralisation in Yemeni Arabic and
Mehri. In B. Heselwood & Z. Hassan (eds) Instrumental Studies in Arabic Phonetics Amsterdam: John Benjamins.
For a more complete list of references, see Watson & Bellem (in press).