Robinson has posted the first examples of Urdreh words other than nouns. But before I get to them, let me look at the noun stems he posted. I say 'stems' because no Urdreh noun can ever stand alone without number/gender/case endings. Here are the new nouns with appropriate endings:

Concrete neuter (inanimate) nouns

mul-: mulyuh 'a fire' (innumerable fires consuming everything in sight would be mulruh); ironically mul- sounds like Korean 물 mul 'water'

olul-: olulyuh 'a drop of water', olulruh 'water' (the innumerable plural -ruh is used for uncountable substances; -yuh, -wuh, -luh refer to one, two, or three-plus drops/pieces of substances: e.g., 'In the desert there was not an olulyuh to drink until I found an oasis of olulruh.')

olul- is disyllabic, though basic stems are monosyllabic. olul- is probably not the original Urdreh stem for 'water' and may be a compound stem (ol-ul- or o-lul-) coined as a replacement for an earlier monosyllabic stem which has become taboo. Perhaps stems in names of deceased Circle Mothers become banned: e.g., if a Circle Mother were named 'Water-yah', that old stem for 'water' can no longer be used.

yel-: yelyuh 'a bit of dirt', yelruh 'dirt': e.g., 'The blade would have been clean if not for the yelluh [three or more spots of dirt] on its edge.'

sah-: sahyuh 'a bit of food'; sahruh 'food': e.g., 'There were still sahwuh [a couple of bits of food] in his bowl.'

dend- 'lizard': denjuh (singular; -d-y- > -j-), dendwuh (dual), dendluh (plural), dendruh (mass plural)

ha- 'hyakl': hayuh (sg.), hawuh (du.), haluh (pl.), haruh (mass pl.)

Adjectives take the same endings as the nouns they precede and modify: e.g.,

su-: suruh olulruh 'blue water'

lem-: lemluh mulluh 'many red fires'

yus-: yuswuh yelwuh 'a couple of bits of yellow earth'

I assume Urdreh has Adj-N order since its compounds have modifier-modified order: e.g., Yar-nayah 'Circle-mother'. Urdreh also has Adj-N compounds. The above phrases could be shortened to




without the redundant endings.

Adj-Adj compounds are also possible:

yus-su- 'green' (lit. 'yellow-blue')

Nouns can be derived from adjectives:

Suyah (a female name; feminine of su- 'blue')

Yussuyeh (a male name; masculine of yus-su- 'green')

lemyih 'the color blue' (with the abstract neuter ending -i-), lemrih 'innumerable shades of blue'


I am attempting to do three things at once:

1. Devise typologically plausible phonological systems for Robinson's languages; i.e., no languages that look like they were slapped together out of random letters - no Mxyzptlks.

2. Devise the systems in 1 out of names for Robinson's existing texts which go back to 1990, long before he or I gave much thought to this issue. The systems must keep these old names basically intact.

3. Devise systems that are as different from each other as possible, so the reader can ideally tell at a glance whether a name is Kasgen (Superior), Dwar, or Urdreh: e.g.,

Kasgen Dwar Urdreh
Initial consonant clusters Only sC-clusters possible: e.g., Old Kasgen Skurn* Consonant + liquid or glide only: Dw- (in the name of the people!), gw-, etc. but no sC-clusters None
Vowels Five** 36 basic vowels and 60 diphthongs plus syllabic l and r; most readers will not realize many vowel-consonant combinations like -an and -ar are really nasalized or retroflex vowels, but they will notice digraphs like oe in Koen [kø̃] generally absent from the other two languages Four (no o)
Final consonants Any initial can be final; clusters in Old Kasgen: -rn, -st None, though the romanization includes Old Dwar stem-final consonant (clusters) that resurface even in New Dwar when vowel-initial suffixes are added: Koen [kø̃] > Koen-a [køna], Aireund [ajrẽw] > Aireund-a [ajrẽwda] 'of Aireund' A limited set of initials can be final; nouns are typically cited in the absolutive case ending in -h
Obligatory number, gender, and case endings No Yes: e.g., -yeh (singular masculine absolutive); these are as inseparable as the masculine nominative singular ending -us of the second declension of Latin

Goal one has been easy because Robinson never fell into the traps of unusual letter combinations and/or aprostrophe mania: e.g., Z'nox, C'cil, B'nee (alien names from X-Men - can you guess where the last two are from?). (Zn- isn't impossible - Russian has it in znat' 'to know', etc. - but I think Denny O'Neil was simply creating a combination that seemed outlandish to him.)

Goals two and three are hard to accomplish together. On the one hand, I am accustomed to Robinson's names from the last two decades and want to keep them more or less unchanged, but on the other hand I do want the languages to be maximally distinct, so I devised the old versions of the languages as a means to explain the relatively similar extant names and give depth to the languages (and a basis from which dialectal variations could develop).

*7.4.1:36: Since no other Kasgen sC-words are known, I started wondering yesterday evening if Skurn is an irregular abbreviation of Kas-Kurn 'the kurn of the Superiors' (?); cf. 's-Gravenhage 'The Hague' < des Graven ha(a)ge 'forest of the Count (of Holland)'. (Dutch words do not normally begin with 's-Gr-.) If Skurn is really 's-Kurn, then there is no need for any initial consonant clusters in Kasgen.

Perhaps kas is not 'Superior' but is an adjective like 'true', 'great', or even 'superior', so Kasgen may literally be 'true/great/superior people' and Gentun could be 'people-city' or 'people-center' - the resemblance between tun and English town is concidental. If -gen is 'people', then Remgen < Rem 'drone' could refer to the Drone class as a whole.

*7.4.1:57: I originally proposed syllabic r to make Kasgen exotic and to enable Scurn (sic) to be reinterprted as Skrn without any final consonant cluster (since r functioned as a vowel). However, this afternoon I realized that syllabic r felt odd to me because I've only seen it in languages that have rich consonant clusters (Czech, Slovak, Serbo-Croatian, Sanskrit, even American English) unlike Kasgen. So now I propose that Old Kasgen r was lost after vowels before a final consonant:

's-Kurn > 's-Kun

Narn > Nan (cf. British English pronounciation of ar) 'third day of a week'

This is consistent with the treatment of Old Kasgen l in the same environment:

Kulm > Kum (cf. the silent l of Eng calm) 'second day of a week'

If Kasgen has vowel suffixes, it's possible that these liquids 'come back' (as in Dwar): e.g.,

's-Kurn-V (syllabified 'sKur-nV)

Narn-V (syllabified Nar-nV)

Kulm-V (syllabified Kul-mV)

Cf. how Korean talg- 'chicken' is pronounced as tak in isolation but as talg- if a vowel-initial suffix follows: talg-ŭl (syllabified tal-gŭl). 'Chicken' is always spelled as 닭 talk with l (ㄹ) in hangul even when the l is not pronounced (the shift of k to g between l and a vowel is automatic, so there's no need for a special letter g). Similarly, Narn could always be spelled with r even when the r is not pronounced. This is similar to how I always write cast with t even though I don't pronounce t in cast but do pronounce the t if a vowel-initial suffix follows: casting (which I never pronounce as cassing). THE LANGUAGES OF HADANUS: DIALECTAL VARIATION IN THE SUPERIOR CALENDAR

Names in English texts about Hadanus are in slightly archaized forms which are analogous to Peking for Beijing*. Although Hadanus has only been settled for 231 Hadanus years (= 165 Dead Earth years), its languages have changed somewhat from the period of initial settlement, though some dialects are more conservative than others.

The standard Superior dialect is based on the speech of Hadan in Gentun. However, Superiors generally have increasingly un-Hadan-like speech the further away they are from Hadan's court in Gentun. Outlier Superiors may preserve Old Superior traits Hadan has given up (see below) but may also have new traits of their own (see below).

Old Superior allowed syllable-final consonant clusters: e.g., in the names of the first four days of the five-day Superior week:

Kast 'first day of a week'

Kulm 'second day of a week'

Narn 'third day of a week'

Lest 'fourth day of a week'

Standard Superior only allows one final consonant per syllable, so such clusters have been simplified in various ways:

1. -VrC sequences have been changed to -rC: Narn > Nrn (roughly 'nern'; r is a syllabic r)

2. -lC sequences have lost -l-: Kulm > Kum (cf. English calm in which -l- is silent)

3. Other -CC sequences have lost their second consonant:

Kast > Kas (cf. my pronunciation of cast and Cass as homophones)

Lest > Les (cf. my pronunciation of lest and less as homophones)

Nonstandard Superior dialects may have retained some or all the original day names without change, or changed them (and words with similar phonetic characteristics) in other ways: e.g.,

- developing retroflex vowels from Vr: Narn > Nn

- developing a syllabic l from Vl: Kulm > Klm

Old Superior is the ancestor of all Superior dialects and can serve as a neutral bridge between them. A reader familiar with Old Superior forms like Kulm from English texts may be able to recognize standard Kum or nonstandard innovative Klm as well as nonstandard conservative Kulm.

*7.3.0:50: The spelling Peking for 北京 'northern capital' was created before ki > ji [tɕi]. The original k sound of 京 'capital' (once king in earlier Mandarin) is still intact in Chinese languages like Cantonese (in which 北京 is Pakking).

Pe for 北 'north' may represent an obsolete Mandarin pronunciation [pə] that has been replaced by bei [pej]. (Note that the P- of Peking and B- of Beijing both represent the same sound: an unaspirated [p] that sounds like a b to English speakers.) THE LANGUAGES OF HADANUS: URDREH (PART 4)

Many Urdreh work as 'Drones' (Rem in the Superior language) and imitate Superior pronunciation (e.g., by replacing w with v and by adding o to their vowel inventory - see Part 3). Such Drones even replace Urdreh vowel + r combinations with Superior syllable r: e.g.,

Urdreh > Rdreh

This mirrors a similar development within Superior: e.g.,

Skurn > Skrn 'name of a fort'

Urdreh allows only

(r or l or nasal) + (any consonant)

after a vowel within a root. (7.4.23:01: Nasals must be at the same point of articulation as the following consonant: -mp-, -mb-, -nt-, -nd-, -ns-; in -nk-, -ng-, -n- is velar [ŋ] like -k- and -g-; [ŋ] can never occur without a following velar in Urdreh)

Since Urdreh doesn't allow initial clusters, this means the maximum native root shape is


Urd- 'Urdreh' takes up three out of the four possible slots.

In Superior-influenced Urdreh, the maximum native root shape may be simplified to


with s-clusters (as in Superior) and r as a possible V along with a, e, i, u plus the un-Urdreh (but Superior) vowel o. Standard Superior doesn't allow -lC clusters, so -l- is dropped before consonants in Superior-influenced Urdreh.

Compound and foreign polysyllabic roots consist of combinations of (C)V(r/l)(C). Superior-influenced Urdreh allows sC-clusters at the beginnings of syllables: e.g., Skearyuh 'Mount Skear' (disyllabic root Ske-ar) instead of Sekearyuh (trisyllabic root Se-ke-ar-).

Word-medial consonant clusters can only contain three consonants if the third consonant belongs to a following root or suffix: e.g., -rdr- in Urdreh consists of -rd of the root Urd- and the collective plural suffix -r-.

Root-final -t -d -s fuse with -y- to become -ch- -j- -sh- (cf. Japanese): e.g.,

Urd-yeh > Urjeh 'one male Urdreh' (but -d remains in Urdreh, etc.)

Nadus-yuh > Nadushuh 'Nadus' (name of a Superior fort)

Superior-influenced Urdreh tends to retain root-final -t -d -s in Superior loanwords even before -y-: e.g.,

Nadus-yuh > Nadusyuh 'Nadus'

but not native words like

Urd-yeh > Urjeh 'one male Urdreh'

There are no word-final consonant clusters in Urdreh. Urdreh only allows a very limited set of final consonants*:

Labials Alveolars Palatal Glottal
Nasals m n
Voiceless fricative h
Nonnasal sonorants w l r y

-h is a real consonant. It is like an initial h- but is pronounced at the end of a word. In katakana, -h would be represented as -hhV/-ffu, with a vowel repeating the previous vowel: e.g.,

ナヤッハ nayah 'mother'

ウルドレッヘ Urdreh 'Urdreh'

ッヒ Jih 'Jeh-hood'

ナドゥシュッフ Nadushuh 'Nadus'

Final -w and -y can be rewritten as -u and -i to avoid confusion with English letter combinations like aw and ay.

*7.2.1:36: The only final consonants that are attested are the absolutive case ending -h and the ergative case ending -r. I don't know of any language that allows only these two consonants in final position. Moreover, I wanted to distinguish Urdreh from Superior (which allows any consonant in final position) and Dwar (which allows no consonants in final solution), so I proposed that Urdreh is halfway between the two, permitting only a limited number of consonants. This list is identical to the bottom half of the list of Urdreh consonants in Part 1, though I have dropped -s which has weakened to -h and excluded sh which can only arise from s + i or y.

The high frequency of -h in attested Urdreh names and words is reminscent of the high frequency of -s in Latin names in English texts. This is because Urdreh names are given in the absolutive singular, just as Latin names are given in the nominative singular. Both languages have more cases than the one case in those citation forms and they also have other parts of speech. Urdreh -m, -n, -l, etc. may appear in other cases, verb endings, etc. that have not yet appeared in any text. THE LANGUAGES OF HADANUS: URDREH (PART 3)

All Urdreh words I have seen have only one of four vowels:

front back
high i u
nonhigh e [æ] a [ɑ]

e is like the a of cat (or a Dwar ae) whereas a is like the au of caught (or a Dwar ao). In some Urdreh dialects, e and a may be central vowels, and e sounds like 'uh':

front central back
high i u
mid e [ə]
low a

(This alternate set of vowels happens to resemble the four vowels that were once proposed for Old Chinese and Proto-Japonic. [ə] is not in Superior or Dwar.)

I have not yet found any Urdreh words with o. O may be in loanwords that have not been Urdreh-ized.

Drones who have not become Superior monolinguals speak a Superior-influenced dialect of Urdreh with o in loanwords. Hence a Drone will say garonyuh instead of garunyuh 'blue clay used by Superiors' (< Superior garon) and guvorruh instead of guwurruh 'gwor forest; Dwar-Cha' (< Superior guvor < Dwar gwor; note the Superior-influenced v instead of w).

Urdreh roots and syllables can only contain a single vowel (a, e, i, u; o only in loans).

Next: What can follow Urdreh vowels? THE LANGUAGES OF HADANUS: URDREH (PART 2)

Here's a list of all consonants that I can find in Urdreh words known to me. Although I don't know of any canonical Urdreh word with w, I proposed a dual suffix -w- in part 1.

Labials Alveolars Palatals Velars and glottals
Voiceless affricate and stops k
Voiced affricate and stops d j g
Nasals m n
Voiceless fricatives s h
Nonnasal sonorants (w) l r y

s is in Yussuyeh, a male Urdreh name from Chapter 1 of The Seeding, and Suyah, a female Drone name from Chapter 6 of The Seeding.

k is in Kalyeh, the name of the Circle Jeh in Chapter 1 of The Seeding. It is also in the stem Kah- 'Superior', but this may be a loanword. (No, kah is a native word [verb stem?] 'hated'.)

The absence of the voiceless stop t is very unusual; if a language has d, it seemingly always has t (though the reverse is not true).

If a language has d, j, g I would expect b and in turn I would also expect p.

I have already proposed that alveolars plus y become palatals: ty, dy, sy > ch, j, sh as in Japanese. (According to this rule, the name Lisyah from an incomplete draft of chapter eleven of The Seeding would be Lishah < Lis-yah except in very archaic, precise speech.)

Hence I propose five previously unattested consonants in bold to fill out the chart:

Labials Alveolars Palatals Velars and glottals
Voiceless affricate and stops p t ch k
Voiced affricate and stops b d j g
Nasals m n
Voiceless fricatives s sh h
Nonnasal sonorants (w) l r y

In Part 1, I proposed initial clusters:

The d- resurfaces preceding other number endings: e.g., Dweh 'two Jeh', Dleh 'three or more Jeh', Dreh 'innumerable Jeh'.

However, I have yet to find any attested Urdreh word with an initial cluster. Here's my view of Jeh now:

- The original stem is di- (perhaps a verbal root 'to lead'?).

- Hence the original singular masculine absolutive was Diyeh.

- This was irregularly contracted to Dyeh which then became Jeh according to a regular rule:

Old Urdreh ty, dy, sy > New Urdreh ch, j, sh (as in Japanese)

- Old Urdreh ti, di, si became New Urdreh chi, ji, shi (as in Japanese).

- Contraction does not take place before number endings other than -y- in slow speech: Jiweh 'two Jeh', Jileh 'three or more Jeh', Jireh 'innumerable Jeh'.

- In rapid speech, -i- may be dropped: Jiweh > Jweh, etc.

No Urdreh root can ever begin with an initial cluster. Urdreh insert vowels in loanwords whose sources had clusters: e.g.,

-i- before -y-: Old Dwar hyakal > Urdreh hiyakal- (sg. masc. abs. hiyakalyeh, etc.)

-u- before -w-: Dwar Dwar > Urdreh Duwar- (sg. masc. abs. Duwaryeh, etc.)

a copy of the following vowel in all other clusters:

Old Superior Skurn > Urdreh Sukurn- (sg. neut. abs. Sukurnyuh, etc.)

Jana Dresh > Urdreh Deresh- (sg. masc. abs. Deresheh < Deresh-yeh, etc.; the sequences ch-y, j-y, sh-y simplify to ch, j, sh)

Urdreh roots tend to be monosyllabic:

(Consonant) + vowel + (r/l) + (consonant):

A name like Yussuyeh probably has a stem consisting of two roots: Yus-su- (rather than Yussu-).


I haven't been posting because I've been sick for the last week. I should have posted this last Sunday but perhaps it's a good thing that I didn't because I came up with a better idea for Urdreh word structure last night.

Robinson deliberately gave all male Urdreh names ending in -yeh and all female Urdreh names ending in -yah. The name Urdreh itself ends in -reh (not -yeh or -yah). And the Urdreh use a biological weapon that they call naryuh ending in -yuh.

All these endings have the structure approximant + vowel + fricative. Extrapolating from that, I came up with the following scheme for Urdreh nouns. Asterisks indicate unattested affixes.

Monosyllabic noun stem Ending
(Consonant) + vowel + (r/l) + (consonant): e.g., Urd- 'Urdreh', Hal-, Gul- (male names), jum- (a kind of weapon), nar- 'parasitic worm' Number Gender: nonhigh vowels nonneuter; high vowels neuter Case
-y-: singular -e-: masculine -h: absolutive
-r: ergative
*zero: vocative
*-w-: dual -a-: feminine *Other sonorants: other cases
*-l-: plural (countable) -u-: concrete neuter
-r-: plural (large collective) *-i-: abstract neuter

Noun stems cannot appear in isolation. They must be followed by number/gender/case suffixes. This distinguishes Urdreh from Superior and Dwar which allow bare nouns.

Noun stems may appear as first elements in compounds: e.g., Yar- 'Circle' (an Urdreh social unit) combined with na- 'parent' in Yarnayah 'Circle Mother' (not Yaryahnayah; compound nouns need only one set of number/gender/case suffixes).

An exceptional noun is Jeh 'an Urdreh male figurehead leader' (also in the compound Yarjeh 'Circle Jeh'). It was originally d-yeh with a noun stem consisting of a single consonant d-. The combinations t-y-, d-y-, s-y- have fused into ch, j, sh in Urdreh (as in Japanese). The d- resurfaces preceding other number endings: e.g., Dweh 'two Jeh', Dleh 'three or more Jeh', Dreh 'innumerable Jeh'. Young Urdreh may say Jweh, Jleh, Jreh by analogy with Jeh, but J- is only correct when it is derived from earlier d-y-.

Urdreh has a four-way distinction between singular, dual (two), countable plural, and collective plural: e.g.,

Urjeh 'a single male Urdreh' (< Urd-yeh)

Urdweh 'two male Urdreh'

Urdleh 'more than two male Urdreh, but countable'

Urdreh 'innumerable Urdreh; name of the Urdreh people'

Like Jeh, the noun Urdreh has incorrect variant forms with -j-: e.g., Urjweh instead of Urdweh.

Urdreh gender is generally biological: e.g.,

nayeh 'father'

nayah 'mother'

However, some nouns for inanimate objects have gender: e.g., jumyah weapons are considered feminine, perhaps because they are associated with (made by?) Urdrah 'innumerable female Urdreh'. Naryuh are considered 'things' and have -u-. Gender-neutral terms may also have -u-:

nayuh 'parent' (either male or female)

Things that cannot be seen like 'conformity' have -i-. -i- can even be added to stems for living things to indicate qualities or states associated with them: e.g.,

Urjih 'Urdreh-ness' (< Urd-yih)

naryih 'naryuh-ness'

nayih 'parenthood'

Jih 'Jeh-ship; duties of a Jeh' (< D-yih)

Urdreh is an ergative language with a special ending -r for the subjects of transitive verbs. Compare Urdreh with English:

Type of verb Urdreh: ah- 'leader' English
Transitive The ahyar sees the ahyah. He sees him.
Intransitive The ahyah lives. He lives.

Notice that English has a special form of he (him) as the object of a transitive verb whereas Urdreh has the same form (ahyah) for both the objects of transitive verbs and the subjects of intransitive verbs.

In Japanese terms:

X-ga + transitive verb > translate as Urdreh X-r

X-ga + intransitive verb or X-wo > translate as Urdreh X-h

Non-Urdreh heard the -r form of ah- 'leader' and borrowed ahyar into their languages instead of the expected ahyah.

Urdreh presumably has other endings corresponding to Japanese case markers like -no, -ni, etc. These endings would all be sonorants: -y, -w, -l, -n, -m, etc. I don't know of any Urdreh words ending in stops (e.g., -k, -g, -t, -d, -p, -b) or fricatives other than -h (e.g., -s -z -f -v). Earlier final stops may have been lost and earlier fricatives may have weakened to -h.

Urdreh has a zero ending for the vocative: e.g., Halye! 'Halyeh!', Urdre! 'O Urdreh people!'

Urdreh stem-suffix combinations are treated as simple stems when borrowed into other languages: e.g., in Dwar,

ahyar 'Urdreh leader'

ahyar-nir 'the Urdreh leader' (it's unknown whether Urdreh has a suffix 'the')

ahyar-a 'of an Urdreh leader' (with -a rather than the Urdreh suffix for 'no')

ahyar-a-nir 'of the Urdreh leader'

The -yar is considered part of the Dwar stem ahyar- whereas from an Urdreh perspective, the stem is ah- and -yar is only one particular number/gender/case combination.

Next: The sound system of Urdreh.

Tangut fonts by Mojikyo.org
Tangut radical font by Andrew West
All other content copyright © 2002-2009 Amritavision