The text (tx) line is the one that holds the transcription of phonological words. Its
purpose is to reflect as closely as possible the spoken data, including truncations, internal
and boundary assimilation (the latter may be interesting for the phonology-syntax
interface), and other phenomena found in spontaneous speech. Major and minor
boundaries (see §3.1 below) are indicated (/ & //), and pauses over 100 to 200 ms appear
in a separate unit (the threshold over which we consider that the pause is significant
depends on genre, language, rhythm).
Transcription is broad: vowels and consonants are written according to their broad
phonetic realization, i.e. in a way that makes the transcription close to the phonetic
realization, but with some phonologization, the idea being to retain phenomena that are
relevant to segmentation (assimilation, dissimilation, ...).
IPA characters are systematically used (except for geminated consonants which are
doubled) as opposed to characters specific to some traditions, unless there is motivation
to do differently. The so-called “emphatics” in Arabic and Berber are an example of such a
treatment: as they are not necessarily pharyngealized, but can be velarized, while
nevertheless forming a homogeneous phonological class, we use a subscript dot; i.e.,
velarized /d
ɣ
/
and pharyngealized /d
ʕ
/ (IPA) are both written /ḍ/ in our Arabic files.
The mot line is mainly an intermediary tier that allows the subsequent segmentation into
‘morphosyntactic words’, the definition of those words being language-dependent. This
tier does not reflect exactly the (phonological-)word segmentation of the tx tier. No
morphemic separators (- =) are used, and the transcription is in its essence phonological
(i.e. 'regularized' as compared to the broad phonetic one). The vowels and consonants are
written according to their phonological value. Phonetic assimilation disappears, but
morphophonological changes (which are not automatic and absolute but depend on the
morphological environment) remain.
The comparison between tx and mot lines allows the retrieval of the phonetic and
phonological phenomena that are characteristic of connected speech.
The mot line is then automatically tokenized into morphemes. Then the resulting mb line
is glossed in ge.
The mb line is segmented into morphemes (one cell per morpheme), allowing for
allomorphs and all such variation desirable for a varied morpheme inventory. Whereas
the mot lines may include allomorphs, the mb line has a single underlying form for each
morpheme. Separator - goes in the cell that contains the affix, while separator = goes in
the cell that contains the clitic.
(1 pages)
Manymanuals.com
Manymanuals.de
Manymanuals.fr
Manymanuals.it
Manymanuals.pl
Manymanuals.cz
Manymanuals.es
Manymanuals-pt.com
Comments to this Manuals