Sangi/Accents and Dialects

=Introducation=

Just like English and any other language with enough speakers, Sangi has different accents and dialects. The accents arose at some point before the planet was colonised and a few dialectal differences had also arisen on the ships. These differences became more noticable after about 5 generations or so, which is the time period covered in this article.

=Accents=

Final consonant classification
The two main features which occur in the final consonant classifications are the treatment of voiced consonants and nasals in which some accents may devoice voiced consonants or allow them to exist with some limiting all final nasals to a single nasal, usually "n" or "m" or even remove all final nasals and nasalising the final vowel. Overall these features have combined to form a set of "stop final" glosses that deviate from the norm. Primarily there are three final stop classes based on the treatment of voiced consonants which are further subdivided based on the treatment of the nasals.

There exist further subclasses of these based on the treatment of final vowels and other final consonants. The subclass representation is written in subscript or in italics after a full stop.

The more diverse an accent the more subclass treatments it will have, but, as with the norm classes, the norm subclass representations are not written. For example, the standard is written as being as Class I, while one of the most diverse final classifications would be IIIc.t2.a1.i1.c1.p1 which would turn final "t" into a glottal stop, "a" into "o", "i" into "e" and both "c" and "p" into "h" while allowing devoicing and nasalisation. Generally, any accent that does not allow devoiced will allow p1 or c1 in through the conservation of a full set of final sounds, but this is not a uniform rule and it does occur rarely.

A good example for the main classes is "mendi" (men) which in Class I is "mendi", "mend" in Class II and "ment" in Class III with subclasses of Class I which form "mende".

Further classifications
Aside from the final consonant classifications of accents there are differing treatments of certain medial and even initial sounds which differ from the norm, although, as usual, the norm is always classed simply as Class I.

Medial classifications
There have been observed many more differences word medially than finally in the various Sangi accents, usually concerning the effect on a consonant or vowel by another consonant or vowel or just a difference in pronunciation. Must vowel differences in vowel come from difference in vowel length or the pronunciation of diphthongs, sometimes merging entire systems and reducing the inventory of sounds significantly. Most consonant differences usually concern the dental and retroflex sounds which some affections of velars as well, but these are rare.

Some accents can technically be both subclasses 1, 2 or more through multiple shifts in diphthong switching based on the following vowel. For example, in some accents "ai" becomes "oi" when followed by an "a", "ei" when followed by an "i" and "a:" when followed by an "o" or "a:" and "ai" when followed by other vowels. These shifts are usually accent specific and have to be explained individually. The overall effect of these processes is an overlapping of diphthongs in some accents and dialects with those in the the standard.

Initial Classifications
There are only a handful of initial differences. With the vowels, they are the same as the medial ones (which I have not written, also occur word finally). The consonants hold even smaller differences with only a small number; th>t, dh>d and the same treatments of retroflexes as medially occur and "g" becoming "gx" or "j" before "i" or "e".

There are subclasses with fewer, or even no, instances of initial consonant mutation, with some uses merging into another mutation and losing it or losing even the uses altogether.

Labeling an accent
When labeling an accent the obvious question is in which order do you place the 3 classifications given above. It is considered standard practice to either write "i", "m" and "f" before which classification type that number represents, e.g. "iI-mI-fI" would be the standard accent, or a simple "initial-medial-final" order is assumed, for example "I-I-IIb" for an almost normal accent with voiced final voiced consonants and nasal finals which merge into "n".

=Dialects=