Virtual Verduria

Proto-Eastern Philology

Introduction   The comparative method   Other types of change
The speakers of proto-Eastern
The proto-Eastern sound system   Phonological constraints
Proto-Eastern sound changes   Cuêzi   Caďinor   Axunašin   Obenzayet   Luxajia
Morphological classes   Verbs   Nouns Adjectives
Morphology      Comparative declension-- Nouns   Adjectives   Pronouns   Comparative conjugation
Syntax    Sample
Proto-Eastern Lexicon
Hints for proto-conlangers

The Eastern family [To Index]

A map of the Eastern languages In recent centuries it has become clear that many of the major languages of Ereláe belong to a single family.

The similarity of Cuêzi and Caďinor was noted in ancient times (indeed, In the land of babblers remarks on it, and rehearses the Cuzeians' thinking on the matter; Beretos' conclusion is that it results from the Caďinorians' attempts to imitate Cuêzi); but the extent of the Eastern family did not become apparent until Verdurians became more familiar with Xurnese, and then Axunašin, some two centuries ago.

Over time it was increasingly recognized that these languages were all related; and Caizu, Obenzayet, Eluye-Makši, and Kešvareni were gradually added to the new family (named Sarise, Eastern, to distinguish it from the languages of the western barbarians).

Here are the major branches and subbranches of the family, with representative members. (+ indicates a dead language.)

Proto-Eastern [To Index]

The reconstruction of proto-Eastern has proceeded according to principles similar to those used to reconstruct Indo-European.

Serious reconstruction has only been attempted in the present century. The first attempt was published by Ružeon in 3442; he and others continued their studies at the University of Verduria (with some assistance from Žésifo and Avéla), culminating in the publication of Dekaši Perëi Řonei (Discovery of the First Language), under the editorship of Ružeon's student Sarileya, in 3473.

The University's effort is highly respectable, but contains serious flaws— most notably an overemphasis on Caďinor and Cuêzi, ignorance of the Čia-Ša languages, inattention to clues offered by minor languages (including Kaini and Ṭeôši), and an occasional overactivity of imagination. This chapter presents my own reconstruction, which given the additional data available and the greater rigor of terrestrial linguistics, must be seen as the most accurate attempt yet to reconstruct the tongue of the Easterners.

The comparative method [To Index]

Reading about protolanguages in the popular press, one may get the impression that to compare two languages, linguists simply choose some plausible-sounding middle ground for each reconstructed word. Nothing could be further from the truth.

Plausible-sounding rules, varying word by word according to the scholar's fancy, allow the misguided to derive any language from any other, and can be seen in schemes deriving all terrestrial languages, from Swahili to Finnish, from Hebrew; or the Avélan scholar Pirosolom's derivation of Caďinor, Kebreni, Eluye-Makši, Western, Skourene, and Qaraumcán (but not Xurnese, which he considered 'devilish') from Cuêzi.

Linguists rely on the comparative method: searching various languages for regular sound correspondances, positing sound changes to account for them, and then reconstructing the historical ancestor of the compared languages. This is the process which has been used to reconstruct proto-Indo-European (and proto-Eastern); the method can be tested (and has been, successfully) by applying it the Romance languages to reconstruct late Vulgar Latin.

If you apply the method to unrelated languages, you get precisely nothing, which is another check on the technique.

The first step, looking for sound correspondances, can be illustrated by listing the following words in several Eastern languages:

Cuêzi Caďinor Axunašin Obenzayet Luxajia
bāsi faucir šagi hägia fàxí
bērede veredes šeirvi häraziz fèrey
ferêde vehend šejiz haʔäd fè(tóʔ)
fuāliu fuelis weli halił fùλu
fuse futes šudi hudiz fúsé
From these words— and many others; it must be emphasized that the method relies on the painstaking comparison of literally hundreds of words, not just a few— there seems to be some kind of correspondance going on:
Cuêzi Caďinor Axunašin Obenzayet Luxajia
b/f f/v š/w h f
It is not enough, however, to state that "an f in Luxajia corresponds to a š or w in Axunašin, a b or an f in Cuêzi, to an f or a v in Caďinor, and to an h in Obenzayet." We must respect the principle—in terrestrial linguistics due to the 19th century German Neo-Grammarians, in Almean to Ružeon and his colleagues— of the regularity of sound change: sound changes are not sporadic, but occur without fail in the proper phonetic environment.

The proper environment in this case may or may not be evident from our small example; but a larger data set (such as the Lexicon) makes it clear: in words which begin with f in Luxajia and an h in Obenzayet, an initial b occurs in Cuêzi before a long a: or ē, and an initial v occurs in Caďinor before a front vowel (i, e). (As for weli, the f was simply lost here.)

We can now reconstruct an initial *f in proto-Eastern for all of these words, which via a regular sound change has changed to b or v in certain environments in Cuêzi and Caďinor. And then we move on to other initials; to the medials; to the vowels... it is a prodigious task.

Special emphasis is laid, in reconstruction work, on comparing paradigms (complete sets of inflected forms), rather than simple words. Correspondences like

lūvāo, lūvēo, lūve (Cuêzi)
liubao, liubeos, liubes... (Caďinor)
ruweu, ruwiu, ruwe.. (Axunašin)
lüvala, lüvalaz, lüviz... (Obenzayet)
are riveting for the historical linguist; almost any word can be borrowed, but the everyday inflections of the language almost never.

From such work we can reconstruct not only the sounds and the lexical items of the protolanguage, but whole inflectional paradigms.

Other types of change [To Index]

Once a sound change has taken place, it often triggers morphological, lexical, and syntactic readjustments. Sounds may merge, confusing similar words or case endings; different words may come to be adopted to disambiguate them, or paradigms may be adjusted. (For instance, the definite past and present came, through sound change, to sound very much alike in Old Verdurian; the definite past was therefore replaced by the remote past.)

Sound changes may obscure earlier morphological rules and make them unproductive, creating irregular paradigms. For instance, the proto-Eastern pattern of short root vowels in the singular and long vowels or diphthongs in the plural, with a uniform set of case endings, has become confused in almost every Eastern language, so that Caďinor, for instance, has ten nominal declensions where proto-Eastern has just three.

Language is tolerant of a certain amount of irregularity; but often a paradigm will be regularized, approximating an existing regular paradigm, or creating a new one. Cuêzi, for instance, has made the declension of pronouns resemble that of the nouns.

The classical form of such readjustment is analogy: one set of forms is made to resemble another, usually more common set. The first adjectival declension in Cuêzi, for instance, over time absorbed a feature of the second and third, namely the identity of masculine and neuter forms.

Writing tends to standardize a language, as speakers of different dialects write in a standard orthography, and a literary standard is consciously imitated. Philologists bless and curse writing: it is their only record of language in former times, but it also obscures linguistic change. The extreme case is Axunášin, which was written using the phonetic glyphs of ancient times even as the language radically changed. (It uses a morphosyllabic script, but this nonetheless encodes much phonetic information.) The result is that we know Old Xurnese only through reconstructions.

Another major source of linguistic change is borrowing. The most familiar cases are lexical borrowing and loan-translation, seen in the hundreds of Caďinor words which derive from Cuêzi. Once borrowed, of course, the word undergoes the linguistic changes of the recipient language. Languages, especially in prolonged bilingual situations, are also known to borrow syntax, morphology, and even phonetics, to the point of obscuring actual paths of derivation; this helps explain why Verdurian scholars have not recognized the affinity of the Čia-Ša languages to their own.

The speakers of proto-Eastern [To Index]

While the provenance of the Indo-Europeans has been a puzzle for decades, there is no similar mystery surrounding the origin of the Easterners (the Sariloi, in Almean scholarship). Their invasion of Eretald and Xengiman around -350 is described in early Cuzeian, Axunašin, and Wede:i sources (only the last being contemporary accounts). It is not hard to place their homeland in Bolon (= northeastern Xurno), an area outside the sphere of early Monkhayic and Wede:i civilization, but well placed to invade both regions when the time came. No archeological confirmation is available, simply because no archeological exploration of pre-urban sites has been done on Almea; but our understanding of the early location and culture of the Easterns has been confirmed in its broad outlines by the ilii.

By the time of the invasion, the Easterners were familiar with the horse, with bronze-working, the bow and arrow, and at least the beginnings of agriculture. None of these technologies were their own invention; the Kagöt had domesticated the horse, while the other technologies were first developed by the Wede:i. Theirs was the first "barbarian invasion" of the plains; and yet it differed from those which the nomads of the steppes unleashed centuries later on their civilized descendants. The Easterners lived in the temperate forests and grasslands of Bolon, not on the steppe; and they had not mastered the use of horses in battle. The Wede:i describe them riding their horses to the very edge of their enemies' cities, then dismounting to make their attack on foot. Their military advantage was speed and the war-readiness of the entire adult male population.

Much of our information on the Easterners derives from the linguistic reconstruction described in this chapter. Not surprisingly, for instance, there are no reconstructed words for reading or writing: the Easterners were illiterate. Nor are there words for 'city', 'temple', 'bridge,' or other tell-tales of urbanization. More interestingly, there exist only a bare minimum of agricultural terms (and no words for 'plow', 'farm', 'crop', or 'harvest'); while words such as *onkou 'herd', *tsērer 'tend', *yagem 'hunt' bespeak a familiarity with animal-herding and hunting. Probably the Easterners, like some American Indian tribes, would plant gardens (*ktats) to supplement what they could gain from herding, hunting, and gathering. Most likely they had no permanent settlements (and indeed there are no reconstructed words for 'city' or 'town').

A number of religious words have been reconstructed, including *nūmiu 'god', *fakus 'shrine', *nēr and *ɣrem 'holy', *fants 'spirit'. And the name of one god can be confidently reconstructed: *Endānor— the source for the Verdurian Enäron, the pre-Endajué Xurnese Inbámu, and even the Cuzeian Eīledan. No other names of gods can be reconstructed, from which some scholars have concluded that the Easterners were monotheists. But only the Cuzeians, among the ancient Eastern peoples, are known to be monotheists; and Cuzeian religion derives from that of the ilii. It seems more likely that each clan had its own gods, and perhaps occasionally abandoned one god for another; so only the chief of the gods was remembered by all the peoples (and among the Axunašin he did not even retain his rank; Inbámu holds a subsidiary position to Meša).

Quite a few animal and plant names can be reconstructed, including *bēts 'cock', *bōna 'cow', *ksūta 'pig', *feɣends 'deer', *mots 'sheep', *tīpal 'horse', *ures 'bear', *ɣaprou 'goat', *abrēna 'oats', *akrens 'maple', *grēlu 'wheat', *xogre 'barley'; not to mention *ɣolps 'fruit' and *xufs 'egg'. *ɣāti 'pot' and *ɣipam 'boil' point to an acquaintance with cooking. (Naturally all these terms refer to Almean plants and animals, for which I have supplied the closest terrestrial equivalent.)

Environmental terms include *ayti 'pond', *luti 'meadow', *bāruma 'mountain', and *gēlere 'river'; the absence of any common term for 'sea' confirms the inland origin of the Easterners. (Caďinor and Cuêzi adapted *sīwai 'lake' to mean 'sea'; Axunašin borrowed a Wede:i word, go:rtu.) The existence of *rūts 'ice' and *nīkte 'snow' is also suitable for Bolon.

Family terms are numerous: *ɣīra 'wife', *mīdor 'mother', *pīdor 'father', *sādor 'sister', *baredū 'brother', *abor 'grandfather', *nepou 'nephew'; note also *nōs 'wedding', *kawma 'hearth'. The Obenzayet reflex of *ɣīra is ʔïra 'slave', which doesn't say much for position of women among the Easterners; that we can reconstruct *sinor 'mother-in-law' but no term for 'father-in-law' may indicate that the wives got their revenge by dominating their children's spouses. The similarity of many of these words may derive from adding a uniform 'kinship suffix' *-dor to babytalk expressions. (For this reason it has been suggested that *baredū derives from an earlier *bādor. There is however no philological evidence for this reconstruction.)

Words such as *anōr 'elder', whose cognates all name positions of respect, as well as *maks 'master', *mēdor 'noble', *londs 'honor', *ōps 'wealth' indicate that Eastern society was far from egalitarian; *ox 'gold' even suggests what the wealth consisted of. In compensation there was the concept, and let us hope the pracice, of *rāfs 'justice'.

The numbers from 1 to 10, and the number 100 (*sikātu) can be reconstructed; but no higher numbers.

The proto-Eastern sound system [To Index]

Sound systems of protolanguages are notoriously variable; the failure of the Indo-Europeanists (dealing with the most well-attested of language families) to determine whether there are three series of obstruents or four, and to decide between aspirated and glottalic stops, is a warning against taking the phonology of a reconstructed language too seriously.

As with all else in proto-Eastern, the phonological system represented below is a reconstruction, based on evidence in derived languages. It is a phonemic reconstruction, meaning that this is the set of sounds recognized as distinctive, and capable of disambiguating words. A phoneme generally has several allophones, different means of articulation which will nonetheless be recognized by the language's speakers as "the same sound." For instance, it is widely believed that proto-Eastern l had both clear (l) and dark (l) variants, as they have sometimes diverged in derived languages.

Consonants

labial alveolar velar
plosives p t k
b d g
fricatives f s x
ɣ
nasals m n
laterals l
approximants r
Vowels
short
  
long
high i u    ī ū
mid e o    ē ō
low a    ā

Diphthongs

aw ay ey oy

Details of articulation are all highly speculative; but it is likely that none of the stops were aspirated, and that the middle articulation was alveolar, as in English, rather than dental.

x is clearly an unvoiced velar fricative. ɣ is more controversial, as can be imagined, given its reflexes in Caďinor /q/, Cuêzi /r/, Axunašin /ɟ/, and Obenzayet /ʔ/. I have taken it as the voiced equivalent of x, /ɣ/— directly attested only in the Čia-Ša languages, but a possible ancestor of all these phonemes (much more so than Ružeon's /ʀ/, or Sarileya's /h/).

(Sarileya adds an ŋ to the nasal series, more for reasons symmetry than for any solid philological reasons. One Verdurian scholar, Nařou, has perfected the symmetry of the system by removing Sarileya's h. But real-world languages are rarely quite this pretty (compare the lack of voiced th in pre-Hellenic Greek, or of voiced gh in German); and the derivation of /q/, /ɟ/, /r/, etc. from /h/ stretches credulity. )

The vowels can be taken as IPA /a e i o u/. The short vowels differed mainly in quantity, but perhaps also in quality; they may have been pronounced as in pot pet pit caught put.

aw /aw/, ay /aj/, ey /ej/, and oy /oj/ are the only true diphthongs. Where w and y occur with other vowels they are consonantal; where other vowels occur together (e.g. *lubeor 'love') they form separate syllables.

(The University's reconstruction differs from mine, in addition to the points already mentioned, in positing a three-way tonal distinction among the vowels (ā â a), and in adding a neutral vowel ə. The former I consider an unjustified echo of Cuêzi, and the latter an ashcan, into which are tossed all the difficult etymologies in Eastern.)

Phonological constraints [To Index]

The reconstruction of roots in proto-Eastern reveals some strong patterns, which have been generalized into phonological constraints.

Typical proto-Eastern roots thus include *fwal-, *ugn-, *lād-, *bāɣor, *ksor-. Examples of prohibited roots: *stōr-, *psuna-, *nāmnōt-, *gɣīp, *xass-.

Proto-Eastern sound changes [To Index]

In what follows changes will be exemplified by Cuêzi, Caďinor, Obenzayet, Axunašin, and Luxajia, representatives of each of the Eastern families, and also the chief languages used in reconstruction of proto-Eastern.

The following abbreviations are used below in the descriptions of sound changes:

F = front vowel (i, ī, e, ē)
B = back vowel (o, ō, u, ū)
H = high vowel (i, u, ī, ū)
S = any oral stop
N = any nasal (n, m)
# = word boundary
 
Cuêzi
 
1. ø → o / [x,Vs,l] _s#
ø → os / x_#
gēs → yēsos → yēzos
ox → oxos
2. ø → a / C_s#
ø → e / C_#
dots → dotas
anōr → nōre; dul → dule
3. t → ø / V_#
s → ø / F_#
xūnit → xūni
rugetes → ruyise
4. l → : / _[stop]
r → u / _[stop, s]
kalkou → cāco
dorsas → douzas
5. f → b / #_[ā,ī,ē,i]
f → v / V_V
b → v / V_V
s → z / V_V
t → s / (V,#)_F
k → s / _i(:)
k → s / V_F
g → y / _F
x → c / _i
fāxes → bāxe
rāfs → rāfas → rāvas
lēbes → lēve
kisir → sizi
tīnama → sīma
kisir → sizi
ɣrukek → rusê
rugites → ruyise
twixis → tici
6. ɣ → x / _[u,w]
n → ø / ɣ_
ɣ → ø / _r
ɣwanca → xuâca
neɣne → nere
ɣrem → rēme
7. r → l / _w
r → l /_...r
bērwes → bēlue
mēruɣs → mēlure; krurs → kluros → cloros
8. k → ^ / _#
m → ^ / _#
r → ø / F_#
pelek → pelê
yagem → yayê
kisir → sizi
9. s → ø / #t_
k → u / #_t
k → u / #_s
tsērer → tēre
ktānem → utāne
ksūta → usūta
10. u → i / _s# manus → manis
11. n → ^ / [stop,s]
n → : / V[velar]_
s → ø / yt_
s → ^ / Vt_
k → ^ / _s
k → ^ / _t
londs → lôdas; sons → sonsas → sôsas
ugne → ūne; rakni → rāci
aytsas → aytas → āetas
xotsex → xôtex
kreksir → crêsi
dekt → dêt
12. ɣ → r ɣay → rāe
13. r → u / [k,g,b]_
p → b / _r
t → d / _r
abrēna → abuēna; xogre → xogue
prenam → brinâ
tragem → drayê
14. e → i / _[nasal]
u → a / [ū,ō]Cn_#
w → ø / #_
prenam → brinâ
klāgūtu → clāgūta
wenka → eîca
15. e → ei, a → ae
in monosyllables
ber → beire
16. aw → āu
ay → āe
ey → ēi
oy → ōu
tawka → tāuca
aytsas → āetas
leyfs → lēivas
nōyns → nōumas
17. u → ø / o_#
ū → u / _#
onkou → ôco
xēkū → xēcu
18. k → c
w → u
i → y / V_V
klāger → clāyē
andwōr → aduōre
meiu → meyu
19. combine double vowels kwunas → cuunas → cūnas
20. eliminate ^'s from
any word with long vowels
mēlatses → mēlâte → mēlate
 
Caďinor
 
1. b → p / #_[ā,ē]
g → c / #_[ā,ē]
d → ď / _[u,w]
t → ť / _[u,w]
p → f / _u
bāɣor → pahor
gēlere → celere
dūna → ďun
xūnitu → ȟuniť
crepū → crefu
2. nɣ → k benɣe → beke
3. ɣ → h / _w
/ V_V
→ k elsewhere
ɣwors → huros
froɣes → frohes
ɣīra → kira
4. u → ø / _# aytsanu → aestan
5. ø → o / C_s#
ø → o / C_th#
xūns → ȟunos
xūnthu → xūnth → ȟunoť
6. f → v / V_V
/ #_F
s → z / _[+long,-low]
brefes → breves
feɣend → vehend
sōl → zol
7. ts → st
gn → ng
ks → sc
ktatsit → ctastit
ugne → unge
ksorū → scoru
8. m → n / _# prenam → prenan
9. k → c
x → i / V_#
x → ȟ elsewhere
kisir → cisir
ktofex → ctovei
rīxam → riȟan
10. Vowel harmony. sikātu → secath; lāboni → labanis; grelū → grilu
A change of vowel was obligatory when root vowels differed by two steps of height (e.g. bāruma → parena); sporadic if they differed by just one (rugetes → rugites). A long vowel did not change; if there was no long vowel, the second was the more likely to change. An unusual example of both vowels changing is waruns → ueronos. There was no requirement for harmony between root and affixes. The rule was not productive in later Caďinor (as witness such words as maȟila, faliles).)
11. w → ø / V_V
aw →u
ow → uo
sāwil → sael
tawka → tuca
towram → tuoran
12. V:C → vCC / _V
C = [n, m, vls stop]
dēns → dēnos → dennos
pītir → pittir
13. ā → au / [-cons] C1_C2
C2 not fricative or m
nāre → naure
(tāsu → tas, rāma → rana)
14. ū → iu / [l, r] _ lūbek → liubec
15. V: → V lēbes → lebes
16. ay → ae aytsas → aestas
17. y → i
w → u
yagem →iagen
waruns → ueronos
 
Axunašin
 
1. r → ø / [u,i]_V
r → ø / V_(C)#
kura → xue
furam → šuem, anōr → emou
u → ø / i_ fwaliu → weli
2. x → š
ks, kt → x
ts, tr → č
ex → eš
buks → bux
aytsas → eče, tragem → čejim
3. t → ø / _# ktult → xul
4. ng, gn, kn → nk unge → unki, rakni → renki
5. f, ɣ → ø / _w
ɣ → j
ɣwex → weš
ɣīra → jire
6. n → m / V_(V,#) ktānem → xamim
b → w / V_V
d → z / V_i
d → v / V_V
g → j / V_V
S → [+vcd] / V_V
(s,f) → [+vcd] / V_V
abor → ewu
kadir → kezi
baredū → berivu
klāger → laji
nepou → nebou, ksūta → xude
kisir → kizi, brefes → beivi
7. Ns → z / _#
S[+vcd]s → z / _#
p,f,r,l → ø / _#
dēns → deiz
kuwids → kuwiz
ōps → us, dors → dus
8. l → r / #_
C → ø / #_l
lēbes → revi
plafes → lavi
f → š / #_V far → šer
rV → VV / C_
Vl → VV / _C
kreksir → kēxi → keixi
kalkou → kakou
9. aw, au → o
ow, ou → ou
ay, ey → ei
oy → ui
maux → moš
towram → tourem
ayka → eige
kalkoyu → kakui
a → e
e → i / (syl)_
o → u
manau → menu
mōles → mouli
ksorū → xuru
ē → ei
ō → ou
: → ø / (a,i,u)_
dēns → deiz
dōns → douz
bēgir → beji
10. i, y → ø / _[č,j,š] aytsas → eiče → eče
11. n → ø / _[t,č,z]
w → ø / C_
fants → šeč
gwents → geč
e → ø / (u,o)_š yāloex → yaluš
 
Obenzayet
 
1. l → ~ / B_C, B_s
S → [+vzd] / _u, _w
n → ng / _[+vel]
w → u / ɣ_
ktult → kšuʔ, pols → palaz
tuli → ʔuli (except after y) ayku → aiku
trankem → zrangkang
ɣwanka→ ʔuangka
2. ɣ → ʔ
s → z / [+vcd]_#
s → z / V_#
wV → u / #_
x → k / #_
x → ʔ / _#
t → ʔ / _#
k → g / _#
m → ng / _#
ɣīra → ʔïra
londs → lädz
mōles → mäliz
wenka → ungka
xupe → ʔuba
barex → baraʔ
let → laʔ
lūbek → lüvag
ktānem → kšänang
3. p → b /V_V
t → d /V_V
k → g /V_V
b → v /V_V
d → z /V_V
g, x → ɣ /V_V
s → h / V_V
ø → a / [p,b,r]_[s,z]#
tīpal → tïbal
protes → vradiz
sikātu → sigäʔu
lūbor → lüvä
lādam → läzang
tragem → zraɣang
kisir → kihia
ōps → äpas
4. S, f → [+vcd+fric] / _r, _l
f → ø / _s
f → h / elsewhere
predi → vrazi
olfs → alas
firax → hiraʔ
5. st, ts → t / _V
kt, ks → kš
[+vel]n → ngg
n → " / _S
mēlastes → mälatiz
ktult → kšuʔ
ugne → ungga
fants → häts
6. y → i / _C
e → i / _z#
aystas → aitaz
bērwes → bäliz
7. e, o → a
ē, ō → ā
: → "
prenam → vranang, mots → mats
mērir → märia, nōs → näz
ksūteyē → kšüdayä
8. w → l / V_V, _#
u → l / (V,y)_#
w → [+vzd]
kš → kši / _#
ʔ → ø / _C, C_
nōwer → nälä
ɣayu → ʔail
tawka → taʔa
gēks→ gäkši
lunɣa→ lungɣa →lunga, ɣuds → ʔudz
9. l → l / [+vel]_, _[+vel]
ʔ → l
r → a / _#
klāger → ɣläɣä, kalkou →
kalʔal
tawra → taʔa→ tala
kisir → kihia
 
Luxajia
 
1. ts → č / #_ *tsērer → čère
2. w → ŋ / #_
l → u / _C
*wenka → ŋiká
*kalkou → káuxò → kàxò
3. [+vcd] → ø / V_V
[+vls] → [+fric] / V_V
*naga → na
*makou → maxò
4. V → [4] / [-vcd]_
[+vcd] → [-vcd] /V
[4]V → [51]
*pūde → púŋi
*bolges → *puŋi
*trogam → *troa → rò
5. [+stop] → ʔ / _#
C → ø / _#
*rogs → roʔ
_#*pītir → pìsí
6. C → ø / #_r
pl → ŋ / #_
*prenam → réna
*plafes → ŋáh
7. u → ŋ / V_V *sīwai → sìŋay
8. [+mid] → [+high] / V(C)_, _[+nas] *benγe → pixi
9. il → λ / _V, V_
n → ø / _C
*melie → meλi
*lunγa → luxa
10. t → č / _[+high]
k → j / _[+high]
*tībri → čìpi
*kura → júra
11. C → C2 / _C2 *wiksam → ŋittá, *sakna → sákka
12. ai → i / _C *ayka → ixá

Morphological classes [To Index]

This section introduces the declension and conjugation classes of proto-Eastern. For each part of speech the classes and their typical reflexes in Cuêzi, Caďinor, Axunašin, and Obenzayet are listed, followed by specific examples. Luxajia is omitted as it has no inflectional morphology.

The examples are given in a two-line format, with English glosses on the second line. For instance:
pE Cuêzi Caď Axun Oben
*lūbek
love
lūve liubec ruwik
desire
lüvag
love

If no English gloss is given for a particular form (as for lūve, liubec above), the gloss to the left should be understood. The proto-Eastern gloss should be taken as a conjecture, a plausible semantic source of the attested forms.

Actual reconstruction, of course, is based on more languages (though those listed are the chief ones used), on very many more words, and on all rather than single forms of those words.

Verbs [To Index]

Proto-Eastern has three verb classes (conjugations), easily distinguished by their final letter in the infinitive— k, m, or r. The latter two conjugations have two subclasses each, with minor differences between them.
pE Cuêzi Caď Axun Oben
-ek -ec -ik -ag
-am -an -em -ang
-em -en -im -ang
-ir -i -ir -i -r
-er -e -er -i -r

Examples of verbs:

pE Cuêzi Caď Axun Oben
*lūbek
love
lūve liubec ruwik
desire
lüvag
love
*lādam
go
lāda laudan ravem läzang
*ktānem
come
utāne ctanen xamim kšänang
*bektir
change
bêti bectir
move
bexi
become
bakšia
trek
*klāger
hit
clāye
beat
clager
flog
laji
hit
ɣlaɣä
kill

Nouns [To Index]

There are three basic declension patterns in proto-Eastern, commonly called masculine, feminine, and neuter, though there is some controversy over whether they are true genders or not. The chief uncertainty is over how adjective agreement worked in proto-Eastern, and the test cases are those where the grammatical gender differs from the semantic— e.g the "masculine" *mīdor 'mother'. In Caďinor and Cuêzi such words have generally been "corrected" to feminine gender; and masculine and feminine forms of nouns can be traced back to masculine and feminine forms in proto-Eastern (e.g. *makseis 'mistress' / *maks 'master'), which lends credence to the gender interpretation.

But in Obenzayet semantic feminines like mïzr 'mother' take feminine adjectives, so that gender and declension class are separated. And Axunašin does not associate its gender classes with sex at all. It is difficult to decide which of these patterns, if any, reflects proto-Eastern usage. Since proto-Eastern adjectives had three declension patterns each, corresponding to the nominal declension classes, it seems certain that adjective agreement obtained. But whether the classes were identified with sex or not, or how grammatical/semantic clashes were resolved, is not known. (None of the Čia-Ša languages are inflected, so they are of no use in resolving the controversy. Some of the Sainor languages have five genders (masculine, feminine, animate, concrete, and abstract), but there has been so much rearrangement of the inflectional system, and such likelihood of influence from Lenani-Littoral (with four-gender pronominal systems) that they are not much help either.)

A nominal form in proto-Eastern breaks down into a root, stem vowel, and case ending. The stem vowel indicates both gender and number, and within each gender there are a few different possibilities for stem vowel. Some case endings are uniform across all nouns; some across a single gender; and a few vary somewhat depending on the stem vowel.

Thus there are several declension subclasses within each gender. The table below shows the subclasses, with the stem vowels for singular and plural, and their usual reflexes in the child languages. Following is a set of examples of each subclass.

pE Cuêzi Caď Axun Oben
Masculine
r i -re -or -(r) -a
- i -as, -os -os -s, -z -z, -s
a ay -as -as -e -az
Neuter
a uy -āu -us -o -al
o oy -o -o -ou -al
u uy -u -u -u -ul
i uy -iu -is -i -il
Feminine
a ey -a -a -e -a
e ey -e -e -i -a
i ya -i -is -i -i

Some examples:

pE Cuêzi Caď Axun Oben
Masculine nouns
*anōr
headman
nōre anor
elder
emou
old man
anäraz
headman
*ktats'
garden
utâtos ctastos xech
farm
kšats
oasis
*xūns
land
xūnas ȟunos šuz
province
ʔünz
land
*aytsas
summer
āetas aestas eche aitaz
*manus
hand
manis manus menu manuz
Neuter nouns
*ksorū
darkness
usolu scoru xuru
black
kšarul
difficulty
*yālo
knee
yālo iaulo yalou yälal
*meiu
water
meyu meis mii mail
Feminine nouns
*bāruma
mountain
bārima parena borme bäruma
*gēlere
river
yēore celere geili
stream
gälara
river
*maksei
mistress
mâsei masceis mexi
lady
makšai
mistress

Adjectives [To Index]

Each proto-Eastern adjective belongs to one of the three declension classes, identified by their citation form as the consonantal, -es, and -is classes. In addition it will have three separate forms for each case, corresponding to the three nominal genders.

The table below shows masculine forms only. (The citation form for Cuêzi adjectives is normally the neuter (utûto 'bad'); but the masculine is given here to facilitate cross-language comparison.)

pE Cuêzi Caď Axun Oben
-r -re -r -(r) -a
-l -le -l -l -l
-n -ne -n -m -n
-t -te -t -
-es -e -es -i -iz
-is -i -is -i -iz
Examples of adjectives:
pE Cuêzi Caď Axun Oben
*uwel
old
uele
used up
uil
old
wel ulal
*ktult
evil
utûte
bad
cťelt
evil
xul kšut
wicked
*lēbes
new
lēve lebe revi läviz
*sidis
thirsty
sidi sidis sizi siziz

Morphology [To Index]

This section gives full comparative declensions for each of the proto-Eastern declension and conjugation classes.

Luxajia is omitted from these lists, as it is not an inflected language.

Comparative declension— Nouns [To Index]

Proto-Eastern had three genders, two numbers, and seven cases. All Eastern languages, except those in the Čia-Ša group, have retained the numbers, and most have at least two genders. Our sample languages have retained all three genders (but modern Verdurian has only two, and Ismaîn and Xurnese have abandoned gender entirely).

As noted above, nominal forms can be divided into root, stem vowel, and case ending. The case endings are very nearly uniform within each gender; Verdurian scholars have worked very hard to eliminate the irregularities, and have had the honesty not to succeed. The regularity cannot be coincidence, however.

When the case endings are displayed by gender, there is clearly revealed an agglutinative system on its way to becoming an inflectional one. The uniform dative ending -nu, for instance, was clearly once an independent morpheme.

case M (s/pl) N F
nominative s/t u -
accusative - m a
genitive ex/ē ex/ē ē
dative nu nu nu
ablative tu tu di
instrumental ko ko ru
locative aw aw w

The greater variation in the nominative, genitive, and accusative, which vary by gender and number, is striking compared to the uniformity within genders (and between masculine and neuter) observed in the oblique cases. No doubt at some earlier stage in the language there were just three cases, and the remaining case endings developed from postpositions.

The number distinctions that exist (e.g. the ex/ē alternation in the genitive) can be plausibly assigned to sound changes; almost certainly the reconstructed system derives from an earlier one in which plurality was indicated exclusively by the stem vowel. For that matter, a listing of the singular and plural stem vowels strongly suggests that the stem vowel was once invariable, and was followed by a plural morpheme -y-.

M
N
F
- i o oy a ey
r i u uy e ey
a ay i uy i ya
u uy

In general, the similarities between masculine and neuter suggest that these were one gender at some early stage of the language. (The masculine nouns in -u have become neuter in Caďinor, which suggests that this confusion survived in some form till after dialectalization.) The variation seen in the ablative and instrumental between masculine and feminine (tu/di, ko/ru) is hard to explain; the first perhaps derives from some sound change or vowel harmony rule; the second seems suppletive.

The nominative/accusative distinction is signalled very differently in the masculine and feminine (presence or absence of a morph) and neuter (different morphs). This has led some to posit that proto-Eastern once had a mixed-case system, nominative-accusative for animate (masculine and feminine) nouns, ablative-ergative for inanimate (neuter) ones.

The relative regularity of the proto-Eastern system has been largely lost in the descendent languages, a casualty of sound changes, reinterpretations, and loss of the more regular cases. Only in Obenzayet can one still synchronically distinguish case ending from stem vowel; in all the other languages shown (and all the more so their modern descendants, such as Verdurian) the stem vowel and case ending have merged into a single inflection which expresses case and number.

Masculine
Obenzayet sinä (← *sinor) has been substituted for anäraz (← *anōr), since the latter has migrated to the next declension (that in *-s).

In Obenzayet, the a-a sequences are actually written ä.

Note on *xūns— If the root ends in -s or -x, the s.nom. is the same as the accusative; if the root ends in -m, the s.nom. ends in -ns.

case pE Cuêzi Caď Axun Oben
s. nom anō-r nōr-e anor emou- sin-a-a
s. gen anō-r-ex nōr-ex anor-ei emou-riš sin-a-aʔ
s. acc anō-r nōr- anor- emou- sin-a-
s. dat anō-r-nu nōr-nu anor-an
s. abl anō-r-tu nōr-tu anor-oť
s. ins anō-r-ko nōr-co sin-a-ka
s. loc anō-r-aw emou-ro sin-a-ral
pl. nom anō-i-t nōi- anor-it emou-i sin-i-ʔ
pl. gen anō-i-ē nōi-ē anor-ie emou-rei sin-i-ä
pl. acc anō-i nōi-i anor-i emou-i sin-i-
pl. dat anō-i-nu nōi-nu anor-in
pl. abl anō-i-tu nōi-tu anor-iť
pl. ins anō-i-ko nōi-co sin-i-ga
pl. loc anō-i-aw sin-i-al
case pE Cuêzi Caď Axun Oben
s. nom xūn-s xūn-as ȟun-os šuz- ʔün-z
s. gen xūn-ex xūn-ex ȟun-ei šum-iš ʔün-aʔ
s. acc xūn- xūn- ȟun- šum- ʔün-
s. dat xūn-nu xūn-nu ȟun-an
s. abl xūn-tu xūn-tu ȟun-oť
s. ins xūn-ko xūn-co ʔüng-ka
s. loc xūn-aw šum-o ʔün-al
pl. nom xūn-i-t xūn-i ȟun-it šum-i ʔün-i-ʔ
pl. gen xūn-i-ē xūn-iē ȟun-ie šum-ei ʔün-i-ä
pl. acc xūn-i- xūn-i ȟun-i šum-i ʔün-i-
pl. dat xūn-i-nu xūn-inu ȟun-in
pl. abl xūn-i-tu xūn-itu ȟun-iť
pl. ins xūn-i-ko xūn-ico ʔün-i-ga
pl. loc xūn-i-aw ʔün-i-al
case pE Cuêzi Caď Axun Oben
s. nom ayts-a-s āet-as aest-as eč-e ait-a-z
s. gen ayts-a-ex āet-ex aest-ai eč-eš ait-a-aʔ
s. acc ayts-a- āet-a aest-a eč-e ait-a-
s. dat ayts-a-nu āet-anu aest-an
s. abl ayts-a-tu āet-atu aest-ať
s. ins ayts-a-ko āet-aco ait-a-ga
s. loc ayts-a-aw eč-eo ait-a-al
pl. nom ayts-ay-t āet-āe aest-ait eč-ei ait-ai-ʔ
pl. gen ayts-ay-ē āet-aē aest-aie eč-iei ait-ay-ä
pl. acc ayts-ay- āet-āe aest-ai eč-ei ait-ai-
pl. dat ayts-ay-nu āet-ānu aest-ain
pl. abl ayts-ay-tu āet-ātu aest-aiť
pl. ins ayts-ay-ko āet-āco ait-ai-ga
pl. loc ayts-ay-aw ait-ay-al
Neuter
In Axunašin weli*fwaliu has been substituted for mii, which has migrated to masculine gender.
case pE Cuêzi Caď Axun Oben
s. nom ksor-ū usol-u scor-u xur-u kšar-u-l
s. gen ksor-u-ex usol-ex scor-ui xur-uš kšar-u-aʔ
s. acc ksor-u-m usol-u scor-um xur-um kšar-u-ng
s. dat ksor-u-nu usol-nu scor-un
s. abl ksor-u-tu usol-tu scor-uť
s. ins ksor-u-ko usol-uco kšar-u-ga
s. loc ksor-u-aw xur-o kšar-u-al
pl. nom ksor-uy-u usol-ū scor-ui xur-ui kšar-ui-l
pl. gen ksor-uy-ē usol-uē scor-uie xur-iei kšar-uy-ä
pl. acc ksor-uy-m usol-ū scor-uim xur-uim kšar-ui-ng
pl. dat ksor-uy-nu usol-ūna scor-uin
pl. abl ksor-uy-tu usol-ūta scor-uiť
pl. ins ksor-uy-ko usol-ūco kšar-ui-ka
pl. loc ksor-uy-aw kšar-uy-al
case pE Cuêzi Caď Axun Oben
s. nom me-i-u me-yu me-is wel-i ma-i-l
s. gen me-i-ex me-yex me-ii wel-iš ma-y-aʔ
s. acc me-i-m me-i me-im wel-im ma-i-ng
s. dat me-i-nu me-inu me-in
s. abl me-i-tu me-itu me-iť
s. ins me-i-ko me-ico ma-i-ga
s. loc me-i-aw wel-o ma-y-al
plural— like ksorū
case pE Cuêzi Caď Axun Oben
s. nom yāl-o-u yāl-o iaul-o yal-ou yäl-a-l
s. gen yāl-o-ex yāl-ex iaul-oi yal-uš yäl-a-aʔ
s. acc yāl-o-m yāl-o iaul-om yal-um yäl-a-ng
s. dat yāl-o-nu yāl-onu iaul-on
s. abl yāl-o-tu yāl-otu iaul-oť
s. ins yāl-o-ko yāl-oco yäl-a-ga
s. loc yāl-o-w yal-o yäl-a-al
pl. nom yāl-oy-u yāl-ō iaul-oi yal-ui yäl-ai-l
pl. gen yāl-oy-ē yāl-oē iaul-oie yal-iei yäl-ay-ä
pl. acc yāl-oy-m yāl-ō iaul-oim yal-uim yäl-ai-ng
pl. dat yāl-oy-nu yāl-ōna iaul-oin
pl. abl yāl-oy-tu yāl-ōta iaul-oiť
pl. ins yāl-oy-ko yāl-ōco yäl-ai-ga
pl. loc yāl-oy-aw yäl-ay-al
Feminine
case pE Cuêzi Caď Axun Oben
s. nom ɣīr-a- rīd-a kir-a jir-e ʔïr-a-
s. gen ɣīr-a-ē rīd-aē kir-ae jir-ei ʔïr-a-aʔ
s. acc ɣīr-a-a rīd-ā kir-aa jir-a ʔïr-a-a
s. dat ɣīr-a-nu rīd-anu kir-an
s. abl ɣīr-a-di rīd-adi kir-ad
s. ins ɣīr-a-ru rīd-alu ʔïr-a-ru
s. loc ɣīr-a-w jir-u ʔïr-a-l
pl. nom ɣīr-ey-t rīd-ē kir-et jir-ei ʔïr-ai-ʔ
pl. gen ɣīr-ey-ē rīd-eē kir-eie jir-iei ʔïr-ay-ä
pl. acc ɣīr-ey-m rīd-ē kir-eim jir-eim ʔïr-ai-ng
pl. dat ɣīr-ey-nu rīd-ēnu kir-ein
pl. abl ɣīr-ey-di rīd-ēdi kir-eid
pl. ins ɣīr-ey-ru rīd-ēlu ʔïr-ai-ru
pl. loc ɣīr-ey-aw ʔïr-ai-l
case pE Cuêzi Caď Axun Oben
s. nom makse-i- mâse-i masce-is mex-i makša-i-
s. gen makse-i-ē mâse-yē masce-ie mex-iei makša-y-aʔ
s. acc makse-i-a mâse-a masce-ia mex-ie makša-y-a
s. dat makse-i-nu mâse-inu masce-in
s. abl makse-i-di mâse-idi masce-id
s. ins makse-i-ru mâse-ilu makša-i-ru
s. loc makse-i-w mex-u makša-i-l
pl. nom makse-ya-t mâse-ā masce-ia mex-ei makša-ya-ʔ
pl. gen makse-ya-ē mâse-aē masce-iae mex-iei makša-y-ä
pl. acc makse-ya-m mâse-ā masce-iam mex-iem makša-ya-ng
pl. dat makse-ya-nu mâse-ānu masce-ian
pl. abl makse-ya-di mâse-ādi masce-iad
pl. ins makse-ya-ru mâse-ālu makša-ya-ru
pl. loc makse-ya-aw makša-ya-l
case pE Cuêzi Caď Axun Oben
s. nom gēler-e yēor-e celer-e geil-i gälar-a-
s. gen gēler-e-ē yēor-eē celer-ei geil-iei gälar-a-aʔ
s. acc gēler-e-a yēor-ê celer-ea geil-ie gälar-a-a
s. dat gēler-e-nu yēor-inu celer-en
s. abl gēler-e-di yēor-edi celer-ed
s. ins gēler-e-ru yēor-elu gälar-a-ru
s. loc gēler-e-w geil-u gälar-a-l
(plural— like ɣīra)

Adjectives [To Index]

Each proto-Eastern adjective is provided with three sets of forms, one each for agreement with no uns of the three genders, and using the case endings for the appropriate gender.

The following table lists, for the three proto-Eastern adjectival classes, the singular and plural stem vowels for each gender; plus the comparative and superlative suffixes.

m
n
f
comp
sup
sg
pl
sg
pl
sg
pl
cons. - i o oy a ey or atses
-es e ey e ey e ey eor ekses
-is i uy i uy i ya ior iksis

The masculine nom.s. -s ending is applied only to the -es and -is declensions, not to the consonantal declension. It should also be noted that the final consonant in the consonantal declension is part of the root: compare nominal *anōr/*anōit, adjectival *sōl/*sōlit; the s.nom. forms of the latter adjective in all three genders are *sōl, *sōlou, *sōla.

I have not given comparative tables of declension here for the adjectives; they would take an inordinate amount of space for very little additional information.

Since such widely separated languages as Caďinor, Obenzayet, and Losainor have comparative and superlative forms, it seems certain that proto-Eastern had them as well; but our reconstructions are more tentative here, since they are not inherited by all languages— Axunašin does not have them, and Cuêzi has only the superlatives (but with a comparative meaning). Moreover there is some puzzling variation in the descendant languages; no one knows, for instance, why the -es adjectives have a comparative -eďes in Caďinor.

All the older attested Eastern languages (Cuêzi, Caďinor, Axunašin) are inflected, with fairly free order; it seems likely that proto-Eastern had the same properties. Adjectives normally follow the noun in Cuêzi, Caďinor, and Obenzayet, and precede it in Axunašin and Luxajia. The usual assumption is that adjectives were postnominal in proto-Eastern as well; but the distribution noted could well be areal (north = postnominal, south = prenominal), leaving the question open for the protolanguage.

Pronouns [To Index]

Six personal pronouns can be reconstructed, including both singular and plural 'you'. A single pronoun served for all uses in the third person singular— *taw meant 'he', 'she', or 'it'.

No locative forms are shown; none of the Eastern languages which still have a locative (e.g. Obenzayet and Axunašin) have one for the pronouns, and a locative use for pronouns is hard to come by anyway.

Unusually for proto-Eastern, the pronominal accusatives are normally formed by suppletion rather than by inflection. Some of the child languages have tossed out the suppletive forms (cf. Obenzayet lala 'thee', formed on the analogy of tala 'him/her'); others have increased the variation (see for instance the declension of Caďinor seo 'I'). Cuêzi has modified a number of forms to make them conform better to the nominal declensions. For all these reasons the reconstruction of the proto-Eastern forms is more speculative than elsewhere. Very likely the forms were less regular than shown here; and very probably the pronouns changed over time, as well as across dialects.

(The Čia-Ša languages, except for Tei, have entirely reconstituted their pronoun systems.)

I pE Cuêzi Caď Axun Oben
nom sewo sēo seo siu sala
gen seway soex eae (ir) salai
acc etu etu eth id aʔu
dat sewnu sēnu seon
abl sewtu sētu ed
ins sewko sēco salʔu
thou pE Cuêzi Caď Axun Oben
nom let led let li laʔ
gen leway loex leae (rir) lalai
acc ēɣ ēr ek ej lala
dat lenu linu len
abl letu letu leth
ins leko leco laga
he/she pE Cuêzi Caď Axun Oben
nom taw tāu tu to tal
gen taway tāuex tuae toiš talai
acc tawa tāua tua toe tala
dat tawnu tāunu tun
abl tawtu tāutu toth
ins tawko tāuco talʔa
we pE Cuêzi Caď Axun Oben
nom tāsu tazū tas taz tähu
gen tayē tazuē taie tiei tahä
acc taym tāe taim teim taing
dat tānu tānu taun
abl tātu tātu tad
ins tayko tāco taika
you pE Cuêzi Caď Axun Oben
nom maux māux muȟ moš mauʔ
gen muyē muē muie miei muyä
acc muym muim muim muing
dat muynu mūna muin
abl muytu mūta muoth
ins muyko mūco muika
they pE Cuêzi Caď Axun Oben
nom kayu cayū cai kei kail
gen kayē cayuē caie kiei kayä
acc kaym caē caim keim kaing
dat kaynu caēnu cain
abl kaytu caētu caith
ins kayko caēco kaika
who pE Cuêzi Caď Axun Oben
s. nom ɣayu rāe kae jei ʔail
s. gen ɣayex rāex kaie jeiš ʔayä
s. acc ɣay kaeth jem ʔaing
s. dat ɣaynu rāenu kaen
s. abl ɣaytu rāetu kaeth
s. ins ɣayko rāeco ʔaika
s. loc ɣayaw ʔayal
pl. nom ɣaɣe radē kahe ʔuyu
pl. gen ɣaɣaē radaē kahie ʔuyä
pl. acc ɣaɣam rade kaham ʔuing
pl. dat ɣaɣanu radanu kahan
pl. abl ɣaɣatu radatu kahath
pl. ins ɣaɣako radaco ʔuika
pl. loc ɣaɣeaw ʔuyal

Comparative conjugation [To Index]

There are three conjugations of proto-Eastern verbs: k, m (divided into -am and -em), and r (divided into -ir and -er).

Verbs are conjugated by person and number. The personal endings are similar in all tenses. In the present tense, for instance, the endings all follow the pattern

1s awV
2s ewVs
3s et
1p Vwmu
2p Vwsi
3p Vntu

The vowel V varies by conjugation. The endings are similar in the other tenses, and it is tempting to derive the endings from a previous morpheme sequence

tense + plural + pronoun
e.g. V   w   mu

Ružeon went so far as to reconstruct the clitic pronouns *wo, *su, *tu, *mu, *si, and *ntu, and claimed that *wo was a worn down form of *sewo, *tu of *taw 'he/she', and *mu of *maux— accidentally shifted to the first person! The speculation here runs well beyond the facts. Ružeon is very free with reconstructed -u's, since Caďinor was known to have lost them. Cuêzi generally kept them; Ružeon explains their absence in the Cuêzi 1s and 2s by the frequency of use of these forms, but this explanation seems ad hoc, and their absence from Obenzayet drives the nails in the coffin. A more careful reconstruction (at the pre-proto-Eastern level) would be *AV, *sV, *tV, *mu, *si, *ntu, where A represents an unknown approximant and V unknown vowels. But even these may well be wrong; sound changes may have rendered the original clitics unrecoverable through internal reconstruction.

Almeologist Nikolai Echternacht has offered the fascinating suggestion that the origin of the verbal endings lies not with the pronouns but with the nominal system. In his view, directional affixes have been reinterpreted as personal:

The third person forms would be a more recent development, related to the 3s pronoun *taw. Note also that the 3s/3p forms have only a single vowel, lacking the persistent *w of the other endings. Thus:
1s/2s *-awV, *ewVs ← **-aw-V
1p *-Vwmu ← **-V-uy-nu
2p *-Vwsi ← **-V-uy-si
3s *-et ← **-a-to
3p *-Vntu ← **-V-nu-to

The tenses remain controversial. It is clear that there were present and past indicative tenses, which can be recovered from almost all the Eastern languages. There were almost certainly many more tenses than this; but there is so much variation in the descendant languages, and so much obvious innovation and reinterpretation, that is is impossible to tell what can be traced back to the protolanguage. Some possible candidates:

In the comparative conjugation tables that follow, the same five verbs, one from each conjugation and subclass— *alirek 'live', *rīxam 'see', *lēlem 'see', *bektir 'move', and *klāger 'beat'— are run through their paces.
Present definite
person pE Cuêzi Caď Axun Oben
I alir-awo alir-āo elir-ao el-eu aril-ala
thou alir-ewos alir-ēo elir-eos el-iu aril-alaz
he/she alir-es alir-e elir-es el-e aril-iz
we alir-owmu alir-ōmo elir-om el-oumu aril-amu
you alir-owsi alir-ōzi elir-ous el-ouzi aril-ahi
they alir-ontu alir-ota elir-ont el-utu aril-äʔu
I rīx-awi rīx-āi riȟ-ai riš-oi rïɣ-ali
thou rīx-ewis rīx-ēi riȟ-eis riš-iw rïɣ-aliz
he/she rīx-et rīx-e riȟ-et riš-i rïɣ-aʔ
we rīx-awmu rīx-āmo riȟ-am riš-omu rïɣ-amu
you rīx-awsi rīx-āzi riȟ-us riš-ozi rïɣ-ahi
they rīx-ontu rīx-ota riȟ-ont riš-itu rïɣ-äʔu
I lēl-awi ler-āi leil-ai reil-oi läl-ali
thou lēl-ewis ler-ēi leil-eis reil-iw läl-aliz
he/she lēl-et ler-e leil-et reil-i läl-aʔ
we lēl-awmu ler-āmo leil-em reil-omu läl-amu
you lēl-awsi ler-āzi leil-es reil-ozi läl-ahi
they lēl-entu ler-itu leil-ent reil-itu läl-äʔu
I bekt-awu bêt-āu bect-u bex-eu bakš-alu
thou bekt-ewus bêt-ēu bect-eus bex-iu bakš-aluz
he/she bekt-it bêt-i bect-it bex-i bakš-iʔ
we bekt-umu bêt-umo bect-um bex-umu bakš-umu
you bekt-usi bêt-uzi bect-us bex-uzi bakš-uhi
they bekt-intu bêt-itu bect-int bex-utu bakš-ïʔu
I klāg-awu clāg-āu clag-u laj-eu ɣläɣ-alu
thou klāg-ewus clāy-ēu clag-eus laj-iu ɣläɣ-aluz
he/she klāg-et clāy-e clag-et laj-i ɣläɣ-aʔ
we klāg-umu clāg-umo clag-um laj-umu ɣläɣ-umu
you klāg-usi clāg-uzi clag-us laj-uzi ɣläɣ-uhi
they klāg-untu clāg-uta clag-unt laj-utu ɣläɣ-üʔu
Past definite
Here the pattern is singular iwV, iwVs, V: or ay, plural V:mu, V:si, iVntu, where the V used is distinct in height and (except in the -ek conjugation) position from that found in the present.
person pE Cuêzi Caď Axun Oben
I alir-iu alir-iu elir-i el-i aril-ilu
thou alir-ius alir-iu elir-ius el-i aril-iuz
he/she alir-ū alir-ū elir-u el-u aril-ul
we alir-ūmu alir-ūmo elir-um el-umu aril-ümu
you alir-ūsi alir-ūzi elir-us el-uzi aril-ühi
they alir-iuntu alir-ūta elir-iunt el-itu aril-üʔu
I rīx-iwo rīc-io riȟ-io riš-iu rïɣ-ilu
thou rīx-iwos rīc-io riȟ-ios riš-iu rïɣ-iluz
he/she rīx-ay rīx-ā riȟ-ae riš-ei rïɣ-ai
we rīx-owmu rīx-ōmo riȟ-uom riš-oumu rïɣ-ümu
you rīx-owsi rīx-ōzi riȟ-uos riš-ouzi rïɣ-ühi
they rīx-iontu rīc-ītu riȟ-iont riš-ietu rïɣ-üʔu
I lēl-iwo ler-io leil-io reil-iu läl-ilu
thou lēl-iwos ler-io leil-ios reil-iu läl-iluz
he/she lēl-ay ler-ā leil-ae reil-ei läl-ai
we lēl-owmu ler-ōmo leil-uom reil-oumu läl-ümu
you lēl-owsi ler-ōzi leil-ues reil-ouzi läl-ühi
they lēl-iontu ler-ītu leil-iont reil-ietu läl-üʔu
I bekt-ie bêt-ie bect-ie bex-ie ɣläɣ-ia
thou bekt-ies bêt-ie bect-ies bex-ie ɣläɣ-iaz
he/she bekt-ay bêt-ē bect-ae bex-ei ɣläɣ-ai
we bekt-ēmu bêt-ēmo bect-em bex-eimu ɣläɣ-ämu
you bekt-ēsi bêt-ēzi bect-es bex-eizi ɣläɣ-ähi
they bekt-ientu bêt-ītu bect-ient bex-ietu bakš-äʔu
I klāg-ie clāy-ie clag-ie laj-ie ɣläɣ-ia
thou klāg-ies clāy-ie clag-ies laj-ie ɣläɣ-iaz
he/she klāg-ē clāy-ē clag-e laj-ei ɣläɣ-ai
we klāg-ēmu clāy-ēmo clag-em laj-eimu ɣläɣ-ämu
you klāg-ēsi clāy-ēzi clag-es laj-eizi ɣläɣ-ähi
they klāg-ientu clāy-ītu clag-ient laj-ietu ɣläɣ-äʔu
Infixed tenses
Quite a few Eastern tenses are formed by adding an infix to the verb root. Each of our inflected descendent languages has such tenses (so that this morphological method must have existed in proto-Eastern), but no one infix is attested in all four languages (indicating perhaps that the proto-Eastern system had not stabilized before the breakup into dialects). Here are two sets of forms attested in three languages, which are fairly likely to have existed in the protolanguage.

The first is a present remote or subjunctive tense:

person pE Cuêzi Caď Axun Oben
I rīx-em-awi rīx-ināi riȟ-emai riš-imoi
thou rīx-em-ewis rīx-inēi riȟ-emes riš-imiw
he/she rīx-em-et rīx-ine riȟ-emet riš-imi
we rīx-em-awmu rīx-ināmo riȟ-emam riš-imomu
you rīx-em-awsi rīx-ināzi riȟ-emus riš-imozi
they rīx-em-ontu rīx-inota riȟ-emont riš-imitu

The second is a definite past anterior tense (past conditional in Obenzayet):

person pE Cuêzi Caď Axun Oben
I rīx-er-iwo rīx-erio riȟ-erio rïɣ-ar-ilu
thou rīx-er-iwos rīx-erio riȟ-erios rïɣ-ar-iluz
he/she rīx-er-ay rīx-erā riȟ-erae rïɣ-ar-ai
we rīx-er-owmu rīx-erōmo riȟ-erom rïɣ-ar-ümu
you rīx-er-owsi rīx-erōzi riȟ-eros rïɣ-ar-ühi
they rīx-er-iontu rīx-erītu riȟ-eriont rïɣ-ar-üʔu
Causative present
These forms are only attested in the Central and Karazi families.
person pE Cuêzi Caď Axun Oben
I rīx-uwi rīx-ū riȟ-ui
thou rīx-uwis rīx-ū riȟ-uis
he/she rīx-ut rīx-u riȟ-ut
we rīx-īmu rīc-īmo riȟ-im
you rīx-īsi rīc-īzi riȟ-is
they rīx-intu rīc-īzu riȟ-int
Imperfect
These forms are only attested in the Central and Naviu families.
person pE Cuêzi Caď Axun Oben
I rīx-ior rīc-ior rïɣ-iä
thou rīx-iors rīc-ior rïɣ-araz
he/she rīx-ayr rīx-āre rïɣ-aya
we rīx-omur rīx-ōmor rïɣ-ama
you rīx-osir rīx-ōzir rïɣ-ahia
they rīx-intur rīc-ītur rïɣ-ïʔa
'To be'
The cognates of *esam 'to be' are irregular in most Eastern languages; and it is difficult to reconstruct a wholly regular paradigm in proto-Eastern. Yet a regular conjugation seems tantalizingly close. The present tense conjugation shown below is more or less regular for a first conjugation verb, for instance, except for the alternation between initial *s and *es. Yet it cannot simply be assumed that the initial *e was lost; the Cuêzi and Caďinor 1s and 2s forms should have been *zāi, *zēi / *esai, *esei if the verb began with *e in all forms. And it is likely that the 3s form became *es (rather than *eses or *eset) at an early date.
person pE Cuêzi Caď Axun Oben
I sawi sāi sai zoi sali
thou sewis sēi seis zewi saliz
he/she eses ê es zi sahiz
we esawmu zāmo esam izomu sahamu
you esawsi zāzi esos izozi sahai
they esontu zota sont izutu sahäʔu

In the other tenses the general problem is that *esam can't seem to make up its mind what conjugation it belongs to. These forms seem to belong to the first conjugation; but the rest of the Cuêzi conjugation is divided between the second and fourth conjugations, and in Caďinor between the first and second; the Obenzayet non-present forms have all been normalized to the fourth conjugation. Perhaps the most compelling explanation is that *esam belongs to a sixth, lost conjugation, which the descendant languages have fit into the other five as best they can.

Worse yet, in Cuêzi, Caďinor, and Axunašin, *esam seems to have got mixed up with another verb *fuam. In Caďinor this verb contributes the past tenses; in Cuêzi, the imperfect ones; in Axunašin, the subjunctive. The original distinction of meaning is impossible to reconstruct with confidence; but judging from Cuêzi and Luxajia (in the latter fu means 'habit/habitual'), *fuam probably had the sense of 'be permanently or by nature', *esam was 'be currently or momentarily'.

Simple forms
There are patterns to Eastern irregular verbs that allow a somewhat daring recontruction of the so-called simple forms. We find reflexes of these forms in (e.g.) Caďinor and Obenzayet irregular forms, in the abbreviated Axunašin imperative, in certain fossilized expressions in Cuêzi.

Sarileya reconstructed these forms as *-Qx for first person, *-Qs for second, and *-Qt for third, where Q is an element that deletes the previous consonant. Thus for *ktānem she forms *ktānQx, *ktānQs, *ktànQt, and thence *ktāx, *ktās, *ktāt. With an unusual but consistent vowel raising in the 2nd and 3rd persons, this gives Caďinor ctai, ctes, ctet. There are no plural forms.

I believe the -x and -t are interpolations from Caďinor; note that the I.s irregular verbs there have the regular final vowel (liubec → liuo, ctanen → ctai, faucir → fau), and the one irregular III.s form for a verb in -ec is liubec → lius, with the -s that is normal for this conjugation. These consonants also have no reflexes in Obenzayet. The -Q- is also a description rather than an explanation. I therefore reconstruct -V, -s, -0 (that is, no ending at all in the III.s); with the Caďinor forms modified by analogy.

The first of these are reminiscent of the speculative postposed pronouns from pre-Proto-Eastern (*AV, *sV). These forms may therefore be a glimpse into an earlier, simpler morphology, in which only person (and not tense or number) was marked on the verb... and perhaps an even earlier VSO stage of the language, with the pronouns as independent words.

Syntax [To Index]

Syntax can be reconstructed by the same methods as words— but with a good deal less confidence, since there are many fewer data points. (For a given sound in a given position, we often have a few dozen words with which to reconstruct the proto-Eastern form. For basic sentence order, by contrast, there is just one datum per language.)

The following table summarizes (with a good deal of simplication, of course) the basic syntactic features of the daughter languages:

Cuêzi Caď Axun Oben Luxajia
S order SOV SOV SOV VSO SVO
Adj / Noun NA NA AN NA AN
Gen / Noun NG NG GN NG GN
Clause / Noun NC NC CN NC CN
Prep / Obj PO PO PO PO OP
Verb/ Aux VAux VAux VAux AuxV AuxV
Case nom / acc nom / acc relative status nom / acc none
Pronouns pro-drop pro-drop pro-drop pro-drop required
Negatives mai (V) bu-(V) verb form (V) čin (V)
Questions (V)-me remote tense jiti + subj. (S) sahü (S)

On sentence order, we note that all the daughters agree on SO (but in Ereláe, at least, this is almost a universal). The position of the verb is not quite as obvious; but it should be emphasized that the Čia-Ša languages are not at all uniform on SVO; and the general Naviu VSO order is the same as (and thus possibly adopted from) Western— except for Seia, adjoining Xurno, which uses SOV. Thus we may reconstruct SOV for proto-Eastern.

In terrestrial languages, we tend to see the correlations

VO / PO / NG / NA or AN (modified-modifier)
OV / OP / GN / AN (modifier-modified)

Obenzayet matches the first pattern; Axunašin matches the second except that it has prepositions rather than postpositions. The others have the 'wrong' S order (that is, the other three indicators pattern together).

Except for S order, the Eastern languages seem to pattern geographically: the northern ones (Cuêzi, Caďinor, Obenzayet) are head-first, the southern ones (Axunašin, Luxajia) are head-last. Again, Seia patterns with Axunašin. It is tempting, then, to view the head-first order as an innovation, and suppose that proto-Eastern was consistently head-last: AN, GN, OP.

On case it is fairly clear that Axunašin has reinterpreted the nominative-accusative system of proto-Eastern (it retains it for pronouns)— and the Čia-Ša languages of course have lost it. Similarly, we can assume that proto-Eastern was pro-drop— but pointing, as we have seen, to an even earlier system with (required) postposed subject pronouns.

About all we can say about negatives and questions is that there is no uniformity: each language has gone its own way, and there is no way to tell if any of the daughter languages has retained the methods of the parent language. The Central bu- and Cuêzi bi- prefixes may point to a morpheme *bux, which it's tempting to link to Obenzayet . (Axunašin jiti is a transparent compound of interrogative and demonstrative pronouns from proto-Eastern.)

Sample [To Index]

Here, in the spirit of August Schleicher, is a short text in proto-Eastern. Schleicher got a lot of flak for his fable, but what's linguistics for if you can't have some fun with it?

Giws rīxet ɣēda taway pīdorex bōna bwin ɣetwet.
boy-N watch-3s when he-G father-G cow-N calf-A bear-3s
A boy watches as his father's cow gives birth to a calf.
Bwins net toɣne mīdor lēbenu ditnu nuret. Giws ridit.
calf-N born-3s and mother new-D baby-D give suck-3s boy-N smile-3s
The calf is born, and the mother gives suck to the new baby. The boy smiles.
Bwins kreksit toɣne giws tawtu gomu furet toɣne tawa lūbes.
calf-N grow-3s and boy-N he-Ab with play-3s and he-A love-3s
The calf grows, and the boy plays with it, and loves it.
Taw tawa lawres esam mērit.
he-N he-A beautiful to be consider-3s
He considers it beautiful.
Taway pīdor ktodko gones. "Let sewnu kerans sewis!
he-G father anger-In burn-3s you-N me-D shame be-2s
His father burns in anger. "You are a shame to me!
īl bwins ānuko sōnko aliremes! Tātu gomu tawa kregem ktānemewis?
that-N calf-N 1-in year-In live-3s-rem us-Ab with he-A eat come-2s-rem
This calf will live just one year! Will you come to eat it with us?
ɣiksis fants genaw predi fētet."
weak-N spirit-N clan-Lo before stink-3s
A weak spirit stinks before the clan."
Pīdorex misit giwex ɣōbaw bux soklintu.
father-G word-pl-N boy-G head-Lo not penetrate-3p
The words of the father do not penetrate the boy's head.
Sōns fākit toɣne pīdor bwinex gorgia trankem lādet.
year-N leave-3s and father calf-G throat-A cut go-3s
A year goes by, and the father goes to cut the calf's throat.
Giws fewitet: "Bwin gaksie! Tot laniwo, taway ɣrof towram ktāniwos,
boy speak-3s calf-A hide-1s-past that-A think-1s-past he-G blood-A pour come-2s-past
The boy speaks: "I hid the calf! I thought that you were coming to pour its blood,
toɣne twixiw diwew tawa gaksie."
and quiet-Lo place-Lo he-A hide-1s-past
and I hid it in a quiet place."
Pīdor tawnu kadit: "Meds, let lādemewis toɣne yagemewis.
father-N he-D order-3s son-N you-N go-2s-rem and hunt-2s-rem
The father orders him: "Son, you will go, and you will hunt.
Ures wiks, feɣend wiks; toɣne tawa dōmnu tragemewis.
bear seek-simple deer seek-simple and he-A home-D bring-2s-rem
Look for a bear or a deer and bring it home.
ɣay Endānor det tot tayē tīnama eses."
who-A Endānor-N give-3s that-A we-G meal-N be-3s
What Endānor gives will be our meal."
Giws yaget toɣne feɣend sākit.
boy hunt-3s and deer-A seize-3s
The boy hunts, and seizes a deer.
Bēws giwex kuwidaw eses: feɣend bux lūbū.
peace-N boy-G heart-Lo be-3s deer-A not love-3s-past
There is peace in the boy's heart, because he did not love the deer.
Wenkaē lūbor taway kawmaw fuet.
person-G love-N he-G hearth-Lo persist-3s
A person's love is in his hearth.

Some of the syntactical contrivances here shouldn't be taken too seriously. I used the remote for both future and interrogative meanings, and the simple forms as an imperative. These are plausible uses, but may be incorrect, especially since Proto-Eastern probably had more verb forms than we are aware of.

The negative *bux and the conjunction *toɣne are supported by only two families each (and in each case the families are geographical neighbors). However, it's hardly possible to write a passage without reconstructing something for these!

Proto-Eastern Lexicon [To Index]

Glosses: For space reasons glosses are provided for the proto-Eastern root only. The word may have very different meanings in one or more child languages; see the section on each language for meanings.

Naturally we cannot be certain of the meaning of a reconstructed root; the plausibility of the meaning given ranges from near certainty (in cases like *lēbes, where the descendent languages agree on a meaning) to flimsy conjecture (e.g. *sōl, each of whose reflexes seems to have a different meaning). Names of concrete things (e.g. *tīpal 'horse') have fairly certain reconstructed meanings.

The comparative method inevitably leads to a level of semantic abstraction which almost certainly was not present in the actual language. In many cases the root may have had a more specific meaning than the one shown— perhaps identical to that found in one of the child languages— but our methods cannot say what it is.

Citation forms: Nouns and adjectives are given in the nominative singular (for adjectives, the masculine form), verbs in the infinitive, in all six languages. The declension or conjugation class is recoverable from the proto-Eastern form.

An apostrophe appended to a proto-Eastern form (e.g. *ktats', *ures') indicates that the final s is part of the root (cf. gen.sg. *ktatsex, *uresex), and not simply the nominative ending (cf. *dits, gen. *ditex). A superscript m indicates that the final consonant of the root is m not n (cf. *dōns, *dēns, gen. *dōmex, *dēnex).

The Verdurian reconstruction. The reader is referred to the Dekaši Perëi Řonei for the University of Verduria's reconstruction. The first thing which will be noticed is the smaller size of our lexicon: about 350 words, as compared with nearly three times as many in the Verdurian version. Unfortunately the University's count is inflated by scholarly errors. There is much solid scholarly work; but also much overactive imagination. The principle of the regularity of sound change has been grasped by the Verdurians, but not always followed (however, early efforts were worse). The Verdurians also always reconstruct a proto-Eastern root if Cuêzi and Caďinor seem to agree on a word; but many of these correspondences are due simply to borrowing from Cuêzi, or from the Monkhayic substratum in the Plain. Such sloppiness has prevented the Verdurians from noting the phonological restraints of proto-Eastern (e.g. the prohibition on *pt and *ps).

proto-Eastern Cuêzi Caďinor Axunašin Obenzayet Luxajia
*abor grandpa avore abro ewu avä aw
*abrēna oats abuēna abrenna ebeime avräna apina
*adawr blue adāure adures evo avał aw(li)
*agisam stab ayizâ agasan ejizem aγihaŋ ihá
*akrens maple acuinas acernos ekeiz akén
*alirek live alirê elirec elik arilag
*andwōr powerful aduōre andeor ädüa oču
*anōr elder nōre anor emou anäraz onu
*ānu one āna an ame änu awnu
*arebs tree arevas arbos eriz arabaz ariʔ
*asītses great zîte esistes eziči hìšé
*ax against ex
*awels earth āuelos ales ewis ałaz oŋiy
*ayka thorn āeca aeca eige aiga ixá
*ayti small lake āeti aetanis eidi aidi isí
*aytsas summer āetas aestas eče aitaz išá
*bans route banas banos bez
*baredū brother bardu baraďu berivu vrad paŋu
*barex arm brexos bareȟ beriš baraʔ pari
*bāruma mountain bāruma parena borme bäruma ruma
*bases low baze bases bezi bahiz pahé
*bats’ stick bâtas bastos paʔ
*bayda bowl bāeda baedia beide pay
*bāγor four bāor pahor baju bäʔua paw
*begne blessing bēge beng(ir) benki peŋe
*bektir move bêti bectir bexi bakšia petčí
*benγe clothes uvere beke pixi
*ber mist beire ber ber pe
*bēgir throw bēyi pegir beji bäγia pey
*bērek fear bēre perir bärag peyre
*bērwes first bēlue perues beiri bäłiz
*bēts cock bêtas pettos beč bäts peyʔ
*bēws peace bēu peos
*bolges big bōye bolges buxiz puŋi
*bōna cow bōna bouna boume büna powna
*bōns bull bounos bouz bünz pow
*brefes short breves beivi vrihiz rehé
*brigam fight būga brigan bijem vriγaŋ riy
*buks’ mouth bûsas buscos bux bukši puʔ
*bux no, not bi- bu-
*bwins calf buinas biz puyn
*bwunem distribute vûne bunn- bum- punaŋ puni
*dam give dan dem daŋ ta
*dān flat dāne dannos dam dän ta
*dawr hard dāure dur dor dał taw
*dekt ten dêt dect dex dakš
*demorir exist dêroi demerir demuri
*dēns day dennos deiz dänz teyn
*dikema gums disima dicena digama
*dits baby ditas ditos dič dits čiʔ
*dīm three dīma din dime dïŋ či
*dīmer third dîere tmeres dimi dïmä
*diwe place -due -die diwi diła čiŋi
*doliu hollow dore dolis duli toλu
*dors’ back douzas dorsos dus daraz tor
*doti finger dosi dotis nudi dadi tosí
*dōns m house dōnas domos douz dünz
*dul smooth dule ďul tul čuλ
*duna two duna ďun vume tuna čua
*dwāns wool duānas ďannos daz tänz čuwn
*dwer door ďer der tä
*elūra wrist elūra ilura lure alüra lura
*ēn first eine eimi än ey
*Endānor a god (Eīl)edan Endauron Inbámu Ädänä
*endi wood êdi endis ädi iči
*eris my eris ir ariz
*esam be esc esan izem sahä ehá
*ewel middle el ewil ałal iŋiλ
*fakus shrine facis vacus šegu haγuz fáxú
*fākir leave bāsi faucir šagi hägia fàxí
*fants spirit fâtas fantos šeč häts fáʔ
*far do fi far še
*fāxes skillful bāxe fauȟes häγiz fàxé
*fegi plant fēi vegis šeji haγi
*felais uncle’s velais šelei félay
*fēter stink bēse vetter šeidi hädä fèsé
*fetam travel veťuran šedem hadaŋ fésá
*feγends deer ferêde vehend šejiz haʔädz fè(tóʔ)
*fēredes green berede veredes šeirvi häraziz fèrey
*fewitem speak veiten šuidim
*firax enemy birax viraȟ čiaš hiraʔ fíra
*fōxek blow fōxe foȟec šouš- fòxé
*fōr loud fōre fortes šour häriz
*fronou grass forno runou vranał rónu
*froγes cold feroe frohes ruji vraʔiz
*fuam be fu- fu- šu-
*fuli fern fuli filis fúli
*fūlka color fūca šuke ka
*furam play fūra furan šuem huraŋ fúra
*futes full fuse futes šudi hudiz fúsé
*fwaks yoke fuacas fuacos wax fùʔ
*fwaliu leaf fuāliu fuelis weli halił fùλu
*gaksir hide gâsi kascir gexi gakšia katčí
*gāli stem gāli caulis gäli kali
*gans’ goose gâsas gansos gez kon
*gāps beech, elm gāpas capos gas kaʔ
*gars wit garas garos ges
*gāx lizard gāex cauȟos gaš gäʔ ka
*gēbes pure yēve cep- geivi gäviz
*gēks’ weight yēsas cescos geix gäkši keyʔ
*gelinis long yilini glinis gelmi kelin
*gēlere river yeōre celere geili gälara keri
*gēna beak yēna cenna geime gäna keyna
*gens clan yinos genos ganz jin
*gēs’ idol yēzos ges key
*gina girl yina gina gime jina
*giws boy yine gios gi giłaz jiy
*glayek shine moon xlayê gla- leik γłayag lay
*gomu with go cum mu ju
*gonek burn ogonê ganag juni
*gorgi throat gori gorgis nugi gargi
*grelū wheat goêlu grilu geilu relu
*gutus rat guť gudu xuduz jusú
*gūre lion gūre gurie xüra juri
*gwents metal guîtas guentos geč xäts juyʔ
*īl that īle illu
*kadir order cadi cadir kezi kazia
*kaiu they cayū cai keï kaił
*kalkou heel cāco calco kakou kałk kàxò
*kalpir revere cāpi calpir kapi käpia kàfí
*kawma hearth cāuma cuma kome kama kàma
*kawrec acquire cāurê currec kałag kàri
*kayr shelter coros caer keir kaya
*kerans m shame cranas ceramos kerez karanz kéron
*kisir slip sizi cisir kizi kihia jíhí
*kīra body sīra cira kie kïra jìra
*kiwal sky suale cilel kiul kiłal jiŋáλ
*klāger beat clāye clager laji γłaγä čày
*kōlir gather cōli cullir kouli kälia kòli
*kors’ side corsos kus karaz kór
*kregem eat creyê cregen kejim γraγaŋ
*kreksir grow crêsi crescir keixi γrakšia rétčí
*krepu twilight cuepu crefu keibu γrabu
*krurs leg cloros cruros kus γłuraz rúr
*ksagi vagina usayi scagis xeji kšaγi ittà
*kseγi every seri scehis kšaʔiz ittè
*ksālem breathe usāle scalean xalim kšälaŋ
*ksegne fowl usēge scenge xenki kšaŋga
*kseya neck useya sceia xie kšaya ittèy
*ksorū darkness usolu scoru xuru kšaruł ittóru
*ksutses dead scustes xuči kšutiz ičé
*ksūta pig usūta scutua xude kšüda
*ktats’ garden utâtos ctastos xeč kšats ittáʔ
*ktānem come utāne ctanen xamim kšänaŋ ittàni
*ktods anger utodas ctodos ittóʔ
*ktofs roof utovas ctovos xus kšas ittó
*ktult evil utûte cťelt xul kšut
*kuluda elbow culuda culda xulve kuda júlua
*kura hen cura cura xue júra
*kuriam argue curi curesan kuim kuraŋ
*kuwids core cūidas cuedos kuwiz júŋiʔ
*kwets beetle cuetas cuetos keč jùʔ
*kwuns m wonder cūnas cuomos kuz kunz jù(li)
*lāboni tongue lāvoni labanis rauni lävani luni
*lādam go lāda laudan ravem läzaŋ la
*laksir rise lâsi rexi lakšia latti
*lanem think lanê renim lanaŋ loni
*lāns m flax lāmas lannos raz
*lawns circle lāunas ron-
*lawres beautiful lāure lures rori lałiz luri
*lēbes new lēve lebes revi läviz ley
*lēlem see lerê leilen reilim lälaŋ λe
*lēma milk lemma reme läma leyma
*lēns line lēnas lennos reiz
*leris your leris rir lariz
*let thou led let ri laʔ
*letam fly letâ letan redem lesá
*leyfs wolf lēivas leivos reis lais (ney)ley
*likuns a hard nut licunas licunos riguz lixún
*līxes high līxe lihal- riši liγiz λixé
*londs m honor lôdas londos lädz lun
*lōns m apple lōnas lomos rouz länz
*lunγa bend lûra luka runje luŋa luxa
*luti meadow lusi lutis rudi ludi
*lūbek love lūve liubec ruwik lüvag luŋi
*lūbor love lūvore liubor ruwo lüvä
*mafor black mavore mevu mahar mahó
*makou paste maco megou magał maxò
*maks’ master mâsio mascion mex makš maʔ
*maksei mistress mâsei masceis mexi makšai mattè
*manus hand manis manus menu manuz monu
*maux you māux muȟ moš mauʔ
*meiu water meyu meis mii maił
*mek have mek me
*melie bee melie melie malya meλi
*mēdor noble mēdore medro meivu mäzaraz mey
*meds son medas meďos mez madz meʔ
*mēlatses better mēlate melastes mälatiz
*mēruγs brain mēlure merukos meiruj märus meyru
*mērir consider mēri merir märia meyri
*mēte field mēse mestis meidi
*mētū task mētu meťis d meysù
*mīdor mother mīdore midra mivu mïzä miy
*mis’ word mizos mi misaz
*mōles soft mōle molles mouli mäliz mowli
*mots sheep motas motos muč mats moʔ
*mukses many mûse musces muxi mukšiz mutté
*mūra wonder mūra miura müra
*musa insect muza musa muze muha muhá
*naga foot nega naga neje naγa na
*nape over na nevi naba nafé
*nāre place nāre naure nari nära
*naγi moon nari nahis neji naʔi nay
*nebri nine nebu nebri nebi navri
*negnes same nēge nenges neŋe
*neka daughter neca nege naga nexá
*nem be born onê nen nem naaŋ ne
*nepou nephew nepōre nepo nebou nabał nefò
*nēr holy nēre nier neir ney
*neγne craft nere nege neji nene
*nikts smoke nîtas nictos nïkši niʔ
*nīkte snow nîi neicte nixi nïkša nitté
*nins’ nut ninsos niz ninz nin
*nūmiu god nūmiu niumis numi
*niwōr beauty niōre niwo niłä niŋil
*nōktu night nōtu noctu noxu näkšu
*nōs’ wedding nosos nou näz
*nōwer rain (v) nōue noer noui näłä nowŋi
*nōyns m name nōumas nomos nuz nünz nown
*nurer give suck nure nurir nui nurä nuri
*olfs nose olfas olvos ous alas ow
*onkou herd ôco onco unkou aŋkał ukò
*ox gold oxos o
*oγir hear ori ohir uji aʔia oy
*ōps wealth opos us äpas ow(li)
*pantu five pâtu panť penk tu páčú
*pawsa flea pāuza pusa poze pałsa pàhá
*pawγam push pāurâ puhan pojem páŋa
*pelek press pelê pilik palag péli
*pentam sing pêtâ pentan petim pätaŋ pétá
*peγem stand pejim
*pīdor father pīdore pidor pivu pïzä pìy
*pītir drink pīsi pittir pidi pïdia pìsí
*pitwes little pitue puťies pitiz písù
*plafes yellow flaves lavi vlahiz ŋáhé
*pols floor polos pol pus pałaz
*pons man pomas ponos panz
*potes deep pose potes pudi padiz pósé
*predi before bri pred peivi vrazi
*prenam take brinâ prenam peimem vranaŋ réna
*prosam walk brozâ prosan pouzem vrahaŋ róhá
*protes second brose ptores poudi vradiz rósé
*pūde stone pūd- pied- puvi púŋi
*puts’ stomach pûtas puč púʔ
*rakni thigh rāci racnis renki raŋki rakki
*rāfs justice rāvas ravos ras räs ra
*rāma frog rāma rana rame rawma
*rewmem count rēnê ruemen ramaŋ
*reyir run, flow rēi reii rei
*ridir smile ridi ridir rizi
*ris’ seed rizas risos ri risaz ri
*riγes clear rire riji riʔiz ri
*rīxam look rīxa riȟan rišem rïγaŋ
*rogs horn rogas rogos ruz ragz roʔ
*rōn wide orone roum rän row
*rugetes red ruyise rugites rujidi ruγadiz ruysé
*rūts ice rūtas riotos ruč rüts ruʔ
*sakna pine sāca sacna senke saŋka sákka
*saleter jump salese salter selidi zladä sàsé
*sādor sister sādore saudara savu säzä sàw
*sākir seize sāsi sagi sägia sàxí
*sāren east sārine sar säran
*sāwil salty sāule sael sol sàŋiλ
*selirek roll selirē silirec selik salirag séri
*sewo I sēo seo siu sała
*sidis thirsty sidi sidis sizi siziz
*sikātu hundred sicātu secať sigadu sigädu sísú
*siks willow sicas sicos six síʔ
*siler shine sun sile seler sili síli
*sinor ma-in-law sinore sinera sinu sinä sína
*sīwai lake sīe zieis siweï sïłai sìŋay
*sīxe berry sīxe ziȟe siši sïγa sìxé
*soklir penetrate solci soclir suki sókí
*sonam dream somâ sonan sunem sanaŋ sóna
*sons’ soil sôsas sonsos suz sanz són
*sōl strong sōle zol säl sòλ
*sōns year sōnas zonnos souz sänz sòn
*sules young sules suli suliz súli
*sūris late sūri zurris suri süriz sùri
*sweks wasp suecas suecos sùʔ
*swetsa six sêta suest seče sata
*taw he/she tāu tu to tał
*tawka spot tāuca tuca toge taka tàxá
*tawra fire turan tore tała tàra
*tāsu we tazū tas taz tähu
*teknis narrow sēni tecnis tenki taŋkiz tékki
*tets point setos teťos teč tats té(li)
*teyla rib sēila teila teile taila téλa
*tēxo trunk sēxo teiȟo tešu täγa tèxó
*tībri feel sībui tibris tibi tïvri čìpi
*tīnama meal sīma tienna time tïnama čoma
*tīpal horse tīble tipel tibel tïbał
*toksem enough tôsê toscen toxim takšaŋ tótté
*tops mole topas topos tus tapas
*tos’ dew tozas tosos tu
*tots that one totas totos tuč tats tóʔ
*towram pour tōura tuoran tuwam tałaŋ tòra
*toγne and ton taŋa tóŋe
*tragem bring drayê tragen čejim zraγaŋ
*trankem cut drâcê trancen čenkim zraŋkaŋ ráké
*trogam touch drogâ trogan čujim zraγaŋ
*tsāfs ash tāvas stavos čas čà
*tsasa cedar tasa stassa čeze čáhá
*tsērer tend tēre sterer čeiri tärä čère
*tuli breeze ťulis tuli tuli čúli
*twixis quiet tici ťiȟis tiši tiγiz čùxí
*tyetus palm tieťus tidu taduz čìsú
*tyetenes fourth tietnes tidin tadiniz
*ugne fingernail ūge unge unki uŋga uŋe
*ulis bright uli ilis uli uliz uli
*urākni spider urāci aracnis wanki uräŋki irakki
*ures’ bear (n) urezos ursos uriz (ney)uri
*uwel old uele uil wel ułal uŋiλ
*wardus bird ardis uradus werdu urduz ŋarču
*waruns eagle araunas ueronos waruz urunz ŋarun
*wenka person eîca wenke uŋka ŋiká
*wiksam look for îcâ iscan wixem ukšaŋ ŋittá
*wusa louse uza uosa ŋuhá
*xayps seven xāeps ȟaep šeis kaipas
*xēkū kind xēcu šeigu k xèxù
*xipatou liver cipato ȟepato šidou kidał xísò
*xogre barley xogue ȟorge šugei kaγra
*xots’ bone xôtas ȟostos šuč kats xóʔ
*xufs egg xuvas ȟuvos šus kus
*xupe under xu ȟupe šuvi kuba xúfé
*xurn language ȟron šun kuran xúr
*xūns land xūnas ȟunos šuz künz xùn
*xupwek vomit xupuê šubik xúfù
*yagem hunt yayê iagen yaji yaγaŋ iλi
*yālou knee yālo iaulo yalou yälał iylu
*yektū feather yêtu iectu yexu yäkšuł itčù
*yoki eight yosi ioci yugi yagi
*yorta flower yoreta iorta yute yarta iyttá
*yuli path yuli iulis iyli
*γads buttocks radas kados jaz ʔadz xaʔ
*γaprou goat rabro kapro japou ʔavrał
*γayu who rāe kae jei ʔaił xay
*γāti pot rāsi kattis jadi ʔädi
*γēda when rēda keda xey
*γefer ashes revere kever xehé
*γeksem finish rêsê kescen jexim ʔakšaŋ xetté
*γet which one rete ket ji- xeʔ
*γetwem bear (v) retê keťen jidim ʔat xesù
*γiksis weak risi kiscis jixi ʔikšiz xitčí
*γilema hill rilima kilima jilme ʔilama xiλna
*γipam boil ripâ kipan jibem ʔibaŋ xifá
*γīra wife rīda kira jire ʔïra
*γolps fruit rōpas kolpos jolas ʔapas xowʔ
*γōbs head rōvas kobos jouz ʔäbaz xowʔ
*γorū goodness juru xó(li)
*γrēba spine rēva kreba jeiwe rava rey
*γrem holy rēme krem jeim raŋ re
*γrofs blood cuovas kerovos jous ro
*γrukek control rusê krucec jugik rugag
*γrits’ kidney rîtas kritos jič rits riʔ
*γuds hole xudas kudos juz ʔudz xuʔ
*γutsam dig xûtâ kusan jučem xušá
*γwanka hip xuâca hanca wanke ʔuaŋka xuwká
*γwe eye xue hie we ʔua xuy
*γweksir remove xuêsi hescir wexi ʔuakšï xuytčí
*γwex tail xuecos kueȟos weš ʔuaʔ (xaʔ)xuy
*γwibū owl hibu wiu ʔuivuł xuyŋu
*γwors clay xuras huros wus ʔuaraz xur
*γwōns ball xuōnas hionnos wouz ʔuänz xuwn

Tips for proto-conlangers [To Index]

So how do you do something like this?

First, you must learn some historical linguistics. (In fact, I used the proto-Eastern project as motivation for learning this subject.) This document has already introduced many of the basic concepts. Some useful books:


Don't assume that a protolanguage has to look like IE! Proto-Eastern has many features reminiscent of IE, because the Eastern family is supposed to be a sort of parallel-Earth version of IE. But a protolanguage can be any kind of language at all. Review the Language Construction Kit to understand the ways languages can vary.

The easiest way is to work forward. Create the protolanguage; then use the Sound Change Applier to apply a set of sound changes to each daughter language. Repeat ad nauseam.

Don't just work on the lexicon; think about morphology and syntax as well. Some of the most interesting language histories involve changes in typology (e.g. from agglutinating to inflecting, or from SOV to SVO). Lehmann's book is good on this: his version of IE is very different in feel from Latin or Sanskrit, and of course these are quite different from French or Hindi.

If you're like me, you'll find yourself working backwards instead. This is much like the internal reconstruction that linguists do on real languages. You have to notice patterns that hint at sound changes and earlier regularities.

Don't try to keep the daughter language static; half the fun is to introduce some irregularities due to sound change, to use analogy to undo some of the damage sound changes did to the morphology, and to discover that what seemed like simple nouns in the daughter language are really compounds in the parent. Some examples:

Is there an Almean equivalent of Nostratic? Not yet, but it's tempting to provide one. Based on their earliest locations, the Easterners were probably related to the Wede:i, the Fei, and the Lenani.


© 2001 by Mark Rosenfelder
 
Virtual Verduria