09.6.6.10:30: THE LANGUAGES OF HADANUS: SUPERIOR (PART 5)

I've been thinking that artificial languages like Superior would be designed to be 'perfect' so its consonant system would have no gaps - not even the gaps that do occur in some actual languages:

I decided against v since it was in the sonorant row but wasn't a sonorant.

The following s-clusters might occur. Only sk- is attested.

sh- could be spelled s'h- to remind the reader that it is an s- followed by a h-. Perhaps it is like Dutch sch-s [sx]. Or maybe it would be better not to have sh- at all.

No other clusters are possible.

The following English consonants do not exist in Superior: ch, j, sh, th, v, x [ks], z, zh (the -s- in vision).

Here's how Superiors might pronounce English words in a Superior accent:

fox > foksu or faksu

church > tiyrti (three syllables: ti-yr-ti; for r, see below)

judge > diyadi

shush > siyasi

thin > tin

these, this > dis (Superior has only one kind of i)

vision > bisiyan or wisiyan

zoo > su

There are five vowels (a, e, i, o, u) plus a single syllabic consonant r in Scurn (phonetically Skrn). Serbo-Croatian has the same five vowels plus r system.

Superior syllables have the structure (s)(C)V(C). There are up to 2635 possible syllables. This figure was calculated by multiplying

31 initials (16 simple + 15 s-clusters)

6 vowels (5 + syllabic r)

16 final consonants

and subtracting 155 (= 31 x 5) syllables ending with -iy, -uw, -rw, -rr, -ry. -iy and -uw are hard to distinguish from -i and -u and I don't know of any language that combines syllabic r with -w -r -y.

09.6.5.21:03: THE LANGUAGES OF HADANUS: SUPERIOR (PART 4)

W. Robinson Mason III, the creator of the world of Hadanus, suggested on Wednesday that hyakl may be a Dwaranir loanword into Superior.

I had been assuming that the word was native since it refers to

[l]arger descendants of Dead Earth hyenas, loosely domesticated by Superiors for tracking Dwarnir [= Dwaranir] during the assault on Dwar-Cha.

This implies that the hyakl are not associated with peoples other than the Superiors, though absence of proof is not proof of absence.

However, there are several linguistic reasons why hyakl looks suspicious:

1. It is the only known Superior word with h (6.6.9:42: False. I forgot about Hadan!)

2. It is the only known Superior word with y

3. It is the only known Superior word with a Cy-cluster instead of an sC-cluster

4. It is the only known Superior word ending in a syllabic consonant

5. It is the only known Superior word resembling English words (hyena and jackal)

If hyakl is not a native word, then my sketch of Superior phonology could be changed as follows:

1. There is no h in Superior. If Superior ever lost its p, that sound would have to become f rather than h (cf. the shift of Old Japanese p to Middle Japanese f).

(6.6.9:42: But there is an h in Hadan.)

2. There is no y in Superior, so there can be no Cy-clusters. Earlier Superior y- could have shifted to d- as in Middle Vietnamese or in the Ryukyuan language Yonaguni.

3. There is no syllabic l in Superior. Superior could just have a syllabic r like Serbo-Croatian, but I'd rather not posit it at all since I hypothesized its existence on the basis of syllabic l.

hy- could be borrowed as k-: cf. how Old Japanese (which had no h) borrowed Middle Chinese h as k: e.g,., Middle Chinese 漢 Han 'Chinese' > Japanese Kan.

-kl could be borrowed as -kul (as suggested by WRM) or even -ko (a Hawaiian English pronunciation of -ckal in jackal).

It's also possible that some Superiors simply pronounce the word as is without modifications (but why would they deign to imitate the alien sounds of an 'inferior' tongue?).

The questions I can't answer are: Why would the Superiors borrow the word from the Dwaranir? Did the Dwaranir introduce the hyakl to the Superiors? Are hyakl native to the Dwaranir's forest home of Dwar-Cha?

09.6.4.6:04: THE LANGUAGES OF HADANUS: SUPERIOR (PART 3)

Most if not all syllables in the eleven probable Superior names and words listed in part 1 have the structure CVC (C = consonant, V = vowel). The exceptions are (. = syllable break):

hyakl CyV.Cl (I regard the final -l as syllabic, so this is not a CCVCC monosyllable)

Scurn sCVCC

Skear sCVC.VC (unless this sounds like skier: sCVC)

Superior seems to have at least two types of initial clusters: Cy- and sC-. Two may not seem to be a lot, but filling in the C-slot with the consonants from part 2 can create some exotic clusters from an English perspective:

Potential Cy-clusters:

Note that y is not a vowel, so by is not pronounced like English by, and Superior bye would be 'byeh' and not like English bye.

Like Japanese, Superior may have ch- j- sh- instead of ty- dy- sy-.

Potential sC-clusters:

sb- sd- sg- could be phonetically [zb zd zg] and could be spelled zb- zd- zg-. Note that Superior has no simple z-; z- only occurs in combinations.

sh- is an s- followed by an h- and may be phonetically [sx] (roughly halfway between s-h- and s-k-): cf. Dutch sch- [sx]. This cluster could be spelled as s'h- to avoid confusion with English sh-.

Perhaps Superior could have sCy-clusters:

sty- sdy- might be like s-ch- and s-j- [zj].

All known Superior syllables end in zero or one consonant with the exception of Scurn which seems to end in -rn. David Boxenhorn suggested that -ur- could be a Tangut-style retroflex vowel [uʳ] which would be like an 'oo' and 'r' pronounced at the same time. If Superior had [uʳ], it probably had retroflex versions of all its vowels. These retroflex vowels could be written with final -r which would be a vowel quality and not a consonant:

Skear and korkus could be [ske aʳ] and [koʳ kus].

Retroflex vowels could precede any consonant other than r: e.g, -urp, -urd, -urh (the -h would not be silent). Such vowels could have originated as contractions of -Vr sequences.

There are only four known Superior final consonants: -d, -n, -s, -g.

The existence of -d -g implies -b and the existence of -n implies -m.

Final -h would not be a silent letter: a would be 'ah' but ah would be like 'ah-huh' but without 'uh'.

-r is not a possible final if Superior has retroflex vowels.

One could write a computer program to generate Superior names and words at random:

1. Since most known Superior names and words are disyllabic, the average Superior word is probably disyllabic. A program could select a random number between 1 and 10:

1-2: monosyllabic (e.g., Scurn)

3-9: disyllabic (e.g., Nadus)

10: trisyllabic (e.g., Ladane which I assume is 'lah-dah-neh' and not 'luh-dane')

The probabilities of choosing different sounds should vary: e.g., k- should be much more common than sngy- and a should be much more common than ur.

3. Syllables with improbable and/or undesirable sound combinations would be rejected and regenerated. Examples of such combinations:

- retroflex vowels + -r (unlikely since retroflex vowels originate from earlier -Vr)

-yi and -iy which would be hard to distinguish from -i

(Perhaps -iy and -uw would be pronounced as long vowels [ii uu].)

- syllabic -l followed by -l (syllabic -r folllowed by -r will not be generated since -r is not a possible final)

One could run such a program several times until one generates a name or word that sounds 'right'. I've written such programs in the past though I haven't made any in the last 13 or so years.

To generate a bigger Superior vocabulary rather than just an occasional name or word, one could set the above program to generate monosyllables that could be assigned root meanings. Future names and words could be created by hand from these monosyllabic roots. If Superior had clusters like sngy-, it could have up to 10,800 different nonhomophonous monosyllabic roots*. Roots like sngyurh would insure that Superior doesn't sound like English or Japanese (while still obeying constraints of human languages).

*This figure was calculated by multiplying 60 x 12 x 15:

60 initials

16 basic initials in part 2

15 sC-initials: sp-, sb-, sm- ...

15 Cy-initials: py-, by-, my- ...

14 sCy-initials: spy-, sby-, smy- ...

12 vowels

5 nonretroflex: a e i o u

5 retroflex: ar er ir or ur

2 syllabic consonants: l r

15 finals (the 16 basic initials minus -r)

09.6.2.23:59: THE LANGUAGES OF HADANUS: SUPERIOR (PART 2)

The eleven probable Superior names and words listed in part 1 have the following consonants:

I use 'back' as a cover term for palatal (y), velar (k, g, ng), and glottal (h).

Consonants in parentheses are possible and those in bold are probable.

I assume that the c in Scurn is simply an alternate romanization of [k] before a nonfront vowel. In English, [sk] is written as sk- before e and i but as sc- elsewhere:

sketch, skit

scat, Scot, scuttle

I have never heard of any human language without labial consonants (p, b, m, f, v, w) at all. Their absence in the data may be due to a very small sample.

The distribution of consonants (and vowels) is never entirely random. In real languages I have examined, the presence of one consonant is often correlated with the presence of others.

The presence of the pairs t/d and k/g imply the presence of the pair p/b, though p could be absent if it had weakened to h as in Japanese. (The Japanese consonant pair h/b was originally p/b. Some original p-s did not weaken to h.) For some reason, languages with b can do without p: an even better example is Classical Arabic which has no p at all, because original p shifted to f.

The presence of n implies the presence of m.

The presence of y implies the presence of v or w or both (cf. Hawaiian w which can sound like v or w). I can think of languages with v/w but without y (e.g., Hawaiian, Hanoi Vietnamese) but I can't think of languages with y but without v/w. (The voiced fricative v is not a sonorant but may appear as a descendant or variant of the sonorant w.)

f and ng may or may not exist. I have posited them simply to fill out the chart. When consonant systems are charted, there seems to be a tendency toward full charts rather than charts with lots of holes such as

which is a highly unlikely (and, as far as I know, nonexistent) consonant system.

It's possible but improbable to interpret hyakl as [hyatɬ] with a voiceless alveolar lateral affricate [tɬ]. This sound is absent from English and has been romanized as kl: e.g., Klingon is really [tɬɪŋan] in Klingon. Since Superior has the voiceless/voiced pairs t/d and k/g, it might have the voiced alveolar lateral affricate [dɮ] which could be romanized as gl. However, [dɮ] is a rare sound that only appears in less than 2% of the world's languages in UPSID.

Next: Superior syllable and word structure.

09.6.1.23:59: THE LANGUAGES OF HADANUS: SUPERIOR (PART 1)

My friend W. Robinson Mason III has been working on his Hadanus universe since at least 1990. Several languages are spoken by the post-human species of Hadanus. Tonight I'll be looking at the tongue of the Superiors.

Obviously the name Superior is not a word in Superior. It would be absurd if the Superiors' autonym (name for themselves) just so happened to sound like English superior and even had the same meaning. (I am assuming that the languages of Hadanus are not descended from English or any other 'Dead Earth' language but still operate within the constraints of human languages.)

It's not clear whether the name of the planet Hadanus (< Hadan, name of the Superiors' male leader) is a word in Superior or is a Latinized Superior name like Confucius < Mandarin 孔夫子 Kongfuzi + Latin -us. I presume that Nadus (the name of a Superior fort) is a native Superior name and not a Latinization of Nad. Hence it is possible that the Superiors could pronounce final -us, but that doesn't necessarily mean that there is a Superior word Hadanus.

Other Superior names I have been able to find in currently canonical material at the Hadanus website are

Ladane - female Superior name

Darad - male Superior name

Nadus, Scurn - names of Superior forts

Gentun - city of the Superiors

Skear - mountain near Gentun

Possible Superior words in the Hadanus Glossary of Terms are

garon - clay used by Superiors

hyakl, kustag - animals used by Superiors

korkus - a plant used by Superiors

What can I conclude from a mere seven names and four words? Quite a bit, actually. Superior may have at least five vowels:

None of the above words contains i, but I have never heard of a language with a e o u but without i. The absence of i may simply reflect a limited sampling.

The exact pronunciation of these vowels is unknown, but given the author's fluency in Japanese, I assume they are like the vowels of Japanese with the possible exception of u which may be rounded [u] 'oo' rather than unrounded Japanese u [ɯ]. Languages with five-vowel systems tend to have [a e i o u]; Japanese [a e i o ɯ] may be sui generis.

There may be more vowels which are not clearly distinguished in Anglicized spellings: e.g.,

- the a of garon could be [æ] as in Garrett [gærɛt]

- e could be [ɛ] and/or [ə] as well as [e]

- the u of Nadus could be [ə] as in the English pronunciation of -us which is not [us] 'oos'

The spelling -ea- in Skear could represent

a. a syllable-final [e] plus a syllable-initial [a] if the word is disyllabic (Ske-ar)

b. a diphthong [ea] (or [eə] or [iə]?) if the word is monosyllabic: cf. Khmer phteəh 'house' (one syllable)

c. [ii] if the name rhymes with English shear

Interpretation b implies the presence of other diphthongs in Superior: e.g., if Superior has [ea], I would also expect [oa] with [o], the back counterpart of [e].

The spelling hyakl with final kl implies a syllabic l like the l of Czech mysl 'mind' and vlk 'wolf'. If there were two vowels in hyakl, I would expect a spelling like hyakul. I assume Superior syllabic l behaves like any other vowel and can appear in closed syllables of the type (C)lC (C = consonant).

Languages that I know of with syllabic l (Czech, Slovak, Sanskrit) also have syllabic r: e.g.,

Czech prst 'finger'

Slovak vrch 'hill'

Sanskrit mṛ 'to die' (as in amṛtas 'immortal; Amritas'; the subscript dot signifies syllabicity)

Hence Superior may also have syllables of the type (C)r(C). (6.2.0:24: Scurn may be a monosyllable [skurn] originating from a compound of [sku] and vowelless [rn]; syllabic [r] may lose its syllabicity next to a vowel.)

09.5.31.23:55: 虎於兔 TIGERS IN RABBITS

In "The Capital-Shadow Problem", I wrote,

I'm uncomfortable with all these presyllables [in my Old Chinese reconstruction]. Although they are typologically plausible - even probable - I have not been able to find direct evidence for most of them. But this morning, I realized that I did have potential evidence for my derivation of MC *x- alternating with velars (parallel to my derivation of MC *ʔ- alternating with velars). I'll reveal this evidence on Monday.

I couldn't blog about it any earlier than the 25th because I was away last weekend visiting my friend

sie raʳ

I've written about my four days at her house on her blog. I had meant to pick up where I had left off after coming home, but my discipline faltered and I'll be blogging this week about what I meant to blog about last week. I am in awe of my friend

ʃɨõ gi 'Guard Wife' (analysis of characters here)

who was able to blog every day for two months in a row. I used to be able to blog daily for long periods, maybe even a whole year. What happened?

Anyway, on to what I was supposed to write six days ago. Sagart (1999: 109) proposed that Old Chinese *ʔ-hl- became Middle Chinese *x-. Back on the morning of the 21st, I realized that Middle Chinese

虎 *xoʔ 'tiger'

could be a compression of an Old Chinese disyllabic word like the Chu dialect forms

於兔 *ʔa-hlah

(written phonetically with graphs which would literally mean 'in rabbit')

於檡 *ʔa-hlak

(檡 'a kind of jujube tree' is a purely phonetic symbol like the first graph)

於 *ʔa- could be a prefix and the second syllable could in turn be a compression of an even earlier disyllabic word resembling the dialectal word

茍竇 *koʔ-loh

from south of the Yangtze. Schuessler (2007: 282) pointed out that "comes close to AA [Austroasiatic] forms": e.g., Proto-Monic *klaaʔ (Diffloth 1984), Munda kula. It also resembles non-Sinitic Sino-Tibetan forms like Proto-Lolo-Burmese *k-la² 'tiger' (Matisoff 2003: 599).

Summing up, MC *x- in 'tiger' could be from *ʔ-hl- < *ʔa-hl- whose *hl- is from *kl- < *kV-l-.

I wonder if there are any Old Chinese disyllabic forms beginning with *ʔV- that have compressed counterparts with *ʔ- in Middle Chinese.

6.1.00:15: The *kh- for 虎 'tiger' implied by some Min forms (Schuessler 2007: 282) could be an alternate compression of *ʔ-hl- (cf. Siamese kl- : Old Chinese *ʔ-l- in Sagart 1999: 109) or *kl-.