Modern Tzuman

Tzuman is the lifeblood of the Kanašu Peninsula, the speech of the Kanaši, the most successful group of nation-states on western Rašau. Primed by fierce inter-tribal, inter-city, and eventually inter-country rivalries, their ships have plowed the coasts of both the Emanar Venoršanid (the Neverending Water, the vast ocean to the west of Rašau) and both around and across the Xabran Sea, the massive expanse of water beyond the Straits of Heiyain, and have regular contact eastwards with the Xatsan Penisula.

Tzuman is spoken by many millions of people on the Kanašu Peninsula, several islands to the south, parts of Ḫ enǰan Island in the straits, and a strip of land on the eastern peninsula (under Itece suzerainty). It is the main language of the Macro-Tzuman branch of the Macro-Kanašu language family, and is its most widely spoken language by an gigantic margin. While the language described here is widely used, there are local dialects in each Kanaši state: Eghri, Iši/Išeni, Itece (really a varied dialect continuum, the culmination of the straits' fractuous history with several important varieties, some of which Eghri and Išeni descend from; Teśara, spoken in the Kanaši state of Tešahi, is a separate language, as is the original language of the Civoyé Kingdom). This is a koinè, used as the written medium for many centuries with some changes, but not as much as the spoken varieties.

The language has a long history, nearly three thousand years since they separated from the main Macro-Kanašu stock and settled first in the Seǧákne, the wide plain that comprises what is now Išeneli, and later moving to the southern Itece hinterlands and along the coastline. Being the first people to develop writing in the Northern Hemisphere, their records are the oldest preserved (barring myths of the so-called Masaǰilišxên Hegemony in the northlands).

The Macro-Kanašu language family

Phonology

Consonants

	L	D	A	AP	P	V
Stop	p b		t d		ǧ	k g
Affric		ts	tz	c j
Fric	f		s z	š		h gh
Nas	m		n	ň
Lat			l
Rho			r
Glide	v [ʋ]				y

Vowels

i	u
e	o
ĕ/ē [ɛ]/[ɛ:]	a

diphthongs	ao	au	ey	ēi	ea	ou	uy	ai
stressed version	ó	óu	áy~ái	ái	á	ú	íy	ói

Besides the diphthongs listed here, vowels in hiatus developed a glide, y for i, e, ĕ and ē, v for a, o, u. This is especially prevelent in verb conjugation. Ey and uy, while written as vowel-glide, are diphthongs in speech, preserved in writing due to conservative tendencies in spelling, as well as their patterning with vowel-consonant groups rather than the diphthongs.


Ē developed from the Old Tzuman diphthong -ĕi-, presumably originally formed by continuing to distinguish the reflexive third person ending -ĕi-ĕhn (> -ēihn) from the present/preterit active -ĕhn and the various other active forms -ay-ĕhn -(ň)iy-ĕhn and -(c)uy-ĕhn (all of which > -eyhn), though it has since also occured elsewhere in the language (and in particular has played havoc with the ē-conjugation). Word-finally, it merged with a.

The earliest stages of Old Tzuman were not recorded (phonetically) with any consistency (generally called Ancient Tzuman), being contemporaneous with the deveopment of the Kanaši writing system; its ideographic nature encoding little phonetic information at the time. It quickly progressed from ideo- and logograms to a syllabary based on the 'classical' Old Tzuman (namely the forms marked OTz in this file) words and ideas, giving a much fuller picture (though still clouded by the massive variance of symbols between cities). Sound changes quickly made a syllabary cumbersome, and it developed into the modern abugida system, though some common syllabic symbols survive, and the syllabary itself was propagated widely along the Kanaši trade routes.

Tz [ts] generally originates in PMK initial d, accounting for its somewhat limited distribution; later its occurrence was increased where ð became its modern realisation [ts] (not to be confused with orthographic ts), merging with tz < #d. Ts arose as a positional variant of t before y i and a; most likely originally being a true dental [ts̪], its pronunciation shifted to accommodate alveolar tz, in most of Icetrou and Išeneli becoming [f], less commonly [θ] or even [tθ]/[pf] (all precursor stages to the more widespread [f]). The history of the Tzuman dental affricates and stops is quite complicated, made worse by interdialectal borrowing at various stages.

The language described here is a 'common' (standardized is very much an incorrect word at this stage) variety, primarily based on the speech of Istáril of Icetrou (though with significant input from Cenaša and Kemaš as well); it is used as the language of diplomacy in the Kanašu, west into the Wild Plains (where it influences its distant relatives and other, non-related, languages), along the civilised states of the eastern peninsula, and far up the western coasts, reaching into the Čal lands with the former treaty port of Venóšatz, now the Foreign Quarter of Eīlissan.

Sandhi & Clusters

Sandhi

Tzuman has an extensive sandhi system, both as a result of the sound changes from Ancient Tzuman, and from a seeming desire to truncate verbal endings.

Results of final consonant sandhi before ATz u ~ OTz w ~ MTz v (incorporates word-final devoicing):

p > pv [pʋ̥]	t > sv [sʷ]	c > cv [ʧʋ̥]		k > kv [kʷ]
b > pv [pʋ̥]	d > zv [zʷ]/[sʷ]	j > cv [ʧʋ̥]	ǧ > ǧv [cʋ̥]	g > kv [kʷ]
f > v [ʋ̥]	s > šv [ʃʷ]	š > šv [ʃʷ]		h > kv [kʷ]
z > zv [zʷ]	tz > sv [sʷ]			γ > kv [kʷ]
m > mv [mː]	n > mn [mn]	ň > mň [mɲ]
l > ll [lː]	r > rv [ɹʷ]
v > v [ʋ]	y > iv [iʋ]

Clusters

Tz (ts) c j are treated as unitary sounds (affricates). Other clusters, such as fs fk xt xp sl sk st sn zb gn gm cn jg jd ms tm šk kn (not an exhaustive list) are not. Clusters most frequently resolve to fricative/affricate-stop while stop-nasal (and fricative-nasal) metathesize to homorganic nasal-stop (ň for velars; ts does not occur in this position). Fricative-fricative clusters are varied, with some remaining as is (fs), and some shifting to fricative-stop (*fx > fk).
Stop-stop clusters (including affricates) have a split based on voicing as to how they resolve. Voiceless affricate-stop clusters (with two exceptions) have the affricate resolve to a homorganic fricative before the stop, while voiced stop clusters
A set of tables is useful:

	p	t	c	k			b	d	j	ǧ	g
p	-	ft	šp	k(k)		b	-	bd	jb	ǧb	g(g)
t	ft	-	št	kt		d	bd	-	jd	bǧ	gd
c	šp	št	šc	šk		j	jb	jd	-	jǧ	jg
k	k(k)	kt	šk	-		ǧ	ǧb	dǧ	jǧ	-	ǧg
						g	g(g)	gd	jg	ǧg	-

Stress

Tzuman words are ironclad in stressing words on the penult. This is the easy part; complications arise because that stress has a number of effects on vowels; when stressed they undergo a chain shift, i > e > a > o > u > i (ĕ is strangely left to its own devices, while ē patterns with e). In words with more than 3 syllables (or when unstressed particles/adpositions/etc. are grouped with the word in question), secondary stress may occur, which can affect the result of sound changes from Old Tzuman to modern varieties. Special rules exist for some noun phrases and compounds; normally the primary stress is still attached to the noun, and thus in older phrases the adjective can condense somewhat. However if the adjective is being emphasised, the noun can lose the primary stress by the phrase being treated as a single word, and given that the penult is definitively the stressed syllable, a portion of the adjective will then receive it: ésel viǧílu 'white tower' > Zál véǧil, but it also generates the seperate noun Zlevéǧil 'Sorceror's Tower'. This process is highly productive, and frequently speakers in various settings create their own nominals with it; this can be very confusing to second language speakers and/or the uninitiated, to say nothing of variant dialectal forms.