Cadeirydd: Helen Imhoff
The category of aspect is a popular research topic whose usefulness has been explored with regard to various languages. A classical approach to this grammatical category is normally the assumption that aspect should be seen as the depiction of an act(ion) as a whole (perfective) or as a duration (imperfective). This seems to be prototypically encoded in Slavic, most apparently in Russian, predominantly the by prefixation of verbs, in which the perfective is normally seen as marked, but most clearly by suffixation. Such suffixation triggers pairs of verbs, as for instance in the following:
On otkryval okno, no ne otkryl.
He opened_ip window, but not opened_p.
'He tried to open the window, but did not manage.'
When looking at Welsh, it seems difficult to identify the concept of aspect within a verb-based framework. When, however, opting for the concept of aspectuality, a more universally applicable system takes shape. If we define aspectuality as a cognitive domain referring to a time structure of situations, we may include aspect, lexical aspect (Aktionsart) and verb semantics (Verbalcharakter) as subconcepts which may be encoded in paradigm-like form. This is seen in the following Welsh examples:
Yr wyf yn canu ‘I am singing’.
Yr wyf wedi canu ‘I have sung’.
Yr wyf ar fin canu ‘I am about to sing’.
Yr wyf newydd ganu ‘I have just sung (finished singing)’.
Yr wyf heb ganu ‘I have not sung/I did not sing (did not even start singing)’.
Yr wyf wedi hen ganu ‘I have long sung’, etc.
On the basis of field research, this system is further developed and explained in detail, and parallels are shown with other languages.
A number of languages of northern Europe display a suprasegmental contrast that is realised by means of either tone, glottalisation or overlength. For example, contrastive tones occur in Swedish and Franconian, glottalisation in Danish, and overlength in Estonian. As shown in (1), all three are employed in different dialects of Scottish Gaelic (ScG) to distinguish words belonging to two accent classes, Class 1 and Class 2.
|
(1) |
|
|
Lewis |
Islay |
Applecross |
|
|
|
a. |
Class 1: |
||||
|
|
|
aran |
[ˈáran] |
[ˈaʔran] |
[ˈaran] |
'bread' |
|
|
|
dubhan |
[ˈtúan] |
[ˈtuʔan] |
[ˈtuan] |
'hook' |
|
|
|
bodha |
[ˈpôː] |
[ˈpoʔu] |
[ˈpoː] |
'submerged rock' |
|
|
b. |
Class 2: |
||||
|
|
|
arm |
[ˈarám] |
[ˈarəm] |
[ˈaraˑm] |
'army' |
|
|
|
uan |
[ˈuán] |
[ˈuan] |
[ˈuaˑn] |
'lamb' |
|
|
|
bò |
[ˈpǒː] |
[ˈpou] |
[ˈpoːˑ] |
'cow' |
Overlength in Estonian has long been analysed as reflecting contrastive metrical structure (Prince 1980), and similar analyses have recently been proposed for Swedish (Morén-Duolljá 2013), Danish (Iosad 2016) and Franconian (Köhnlein 2016). Metrical structure has also been invoked for the ScG accent contrast in various ways (Oftedal 1956; Bosch & de Jong 1998; Ladefoged et al. 1998; Smith 1999; Iosad 2015; 2018). Using evidence from phonetics, morphophonology and speaker intuitions I argue for a specific analysis where the ScG accent contrast reflects a difference in the extent of the stressed syllable in the surface representation. Setting this apart from existing analyses of Swedish etc., however, I claim that the accent class of a word in ScG is predictable from its underlying segmental content. ScG therefore represents an intermediate step on the diachronic pathway that led to lexically contrastive metrical structure in those languages.
Scottish Gaelic sonorants have been described as participating in a three-way contrast between palatalised phonemes, velarised phonemes and ‘plain’ phonemes with no secondary articulation (Ladefoged et al. 1998). Here, we report an ultrasound analysis of the tongue shapes used in this contrast, in order to better understanding the phonetic implementation of phonological contrast in Scottish Gaelic (see also Sung et al. 2015).
Data were collected from eight speakers of Lewis Gaelic aged 40-60 who grew up in Gaelic-speaking environments. Participants read a word list containing sonorant phonemes surrounded by both front vowels and back vowels in all word positions. Audio recordings were made using a microphone attached to the ultrasound headset. Ultrasound data were recorded using a high-speed ultrasound system (SonoSpeech) at ~100 Hz and synchronised with the audio signal at each frame of the ultrasound image. Splines were fitted to the tongue contour on the ultrasound image across the entire duration of the vowel+sonorant+vowel interval. We conducted two analyses: (1) a comparison of tongue shapes during the production of each sonorant phoneme; (2) a dynamic analysis of the entire vocalic and sonorant intervals.
Results suggest robust articulatory distinctions between palatalised, velarised and plain laterals and nasals, but no distinct velarisation for the rhotics. Contrasts are realised through tongue body fronting or backing but also tongue root displacement (Bennett et al. 2017). These data contribute to the articulatory description of Gaelic, work towards a better understanding of phonological typology in palatalised sonorants, and provide the basis for future socio-articulatory investigations.