Gold language

The Gold language (also called Diʕì) was spoken around 1900 AD along the south coast of Rilola and became the diplomatic language of Nama. It is the parent language of Khulls and Bābākiam, the latter of which later evolved into Poswa and Pabappa.

Scratchpad

Politeness etc

12:51, 21 February 2022 (PST)

The Play private verb system that so mystified outsiders, deriving faufuufupusa "the door you opened for me" from fausa "door", originated from something more transparent in Gold. Gold still pronounced /ʕ/ in the coda, at least as a vowel coloring, and therefore such constructions did not stain following consonants. For example, using Gold ŋàsina "door", one could derive ŋasiʕ duhʷuʕus na "the door you opened for me". It is possible that the final /na/ would need to be spelled attached to the previous word, as if it were a clitic, since it is unstressed and cannot carry a tone.

A naive early Moonshine reflection of these words could be such as

ŋàsina ----> ŋàsʲn

ŋasiʕ duhʷuʕus na ----> ŋasʲ rhʷunt (though probably -a stays after clusters). this would make it seem like an interfix.

Verb prefixes turn into suffixes

11:57, 21 February 2022 (PST)

The archaic verbal prefixes ŋa- hə- indicated 1st and 2nd person patient respectively, but instead came to attach to the preceding noun as ergative suffixes, which due to sound changes appeared as they do in Play. Thus /săpu/ "fish" becomes /sapūn/ "fish that (bit) me", etc. This soon becomes mandatory for all sentences with a 3rd person agent and a SAP patient. Play retains this, and then Poswa loses it but restores something superficially similar. The classifier suffix appears after this inflection, not before, because it was once possible to hyperbate the classifier suffix to the end of the sentence. This is important for both Leaper and Moonshine, and Trout languages probably did not make the switch from prefixes to suffixes.

This means that Gold, like Play, has neither pronouns nor verbal person markers, and thus that the Play system was almost identical to that of Gold.

The loss of /ŋa- hə-/ implies the loss of /na- gə-/, the corresponding agentive prefixes, which means that a new prefix /na-/ can appear in the Leaper/Moonshine branch, or even in Gold proper, so long as it was quickly lost along with all other prefixes in Play. Alternatively, it could be that Moonshine alone shows this development, and that Moonshine may have reattached some of the suffixes back to the following word as prefixes, likely passing through a stage in which they were free particles.

Feminine markers

Jun 26, 2020

Inalienable nouns have a distinct form when the possessor is feminine. These are mostly but not entirely body parts. This is formed by adding a consonant to the plain form of the noun, and that consonant is inherited from pre-Tapilula and thus ~4000 years out of date. This survived because it was originally the form for angels. This means that the only nouns with such special forms are those which were used in the context of angels.

The word stems are inalienable because there is no clear morpheme boundary. Some affected nouns are in fact inanimate, but are considered inalienable because of their original context of angels: for example, tagàda and gaḳàta, both meaning "her horse" and referring to different types of horse. The stems of these words are effectively tagàd- and gaḳàt-, even though Gold stems cannot normally end in those consonants. The free forms of the words are tăga and găḳa, and the tone pattern is no coincidence, as these extra consonants only appear on words whose free form follows this pattern.

The concept of angels as primarily female is common to all of the societies with this religion, which by Gold times had likely encircled the whole planet. Even societies led by men considered angels to be feminine, and in the Gold language ths morpheme that originally meant "of an angel" came to merge in some constructions with a morpheme denoting feminine property, and the two were conflated. There was no corresponding male property marker; sometimes the bare form was used, and sometimes an additional morpheme was needed. It is in someways similar to IE etc where one must say "actor/actress" but "nurse/male nurse". The difference is that this is for property rather than the men and women themselves.

Because inanimate nouns inherit their gender from their possessor, nouns in this category are grammatically feminine. The /ga-/ prefix on the second horse word might be dropped, leading to an isolated word ḳàt-, meaning a horse that can only be ridden (or at least owned) by a woman. In Gold, there was no way to attach a male possession marker to this word. PMS and Leaper innovated a longer suffix deriving from Gold -u-_-nni for certain nouns, but even this could only attach to a stem ending in a vowel, and it's not clear if words like ḳat- could contract to just "ḳà" etc and still be distinct; their feminine consonants might become obligatory,.

Verb ideas

Verbs are 1p 2p from reflexive reciprocal

Can be reverse. Na/ga

That is, Poswa's reflexive ending /p/ and reciprocal /s/ are the same morphemes that indicated 1st person (/k/) and 2nd person (/s/) in Gold. Where did the Gold morphemes come from? Were Poswa's meanings the originals, preserved as a secondary use through all this time? Is the serial verb marker /s/ involved at all? Is it simply the genitive case?

Phonology

See Gold phonology.

The consonant inventory was:

Bilabials:             p           m           w   mʷ           
Alveolars:             t       d   n   s   z   l   nʷ      tʷ  dʷ
Postalveolars:                         š   ž   y                                  
Velars:                k   ḳ       ŋ   h   g   gʷ  ŋʷ  hʷ  
Postvelars:                            ħ  (ʕ)          ħʷ

The vowel inventory was

Short vowels:      a  i  u  ə
Long vowels:       ā  ī  ū
Diphthongs:          ai au
                     əi əu

Tones

Tones were not well developed in Gold. Syllables could be high or low, and when a high tone occurred immediately before a low tone of the same vowel, this resulted in a falling tone which was considered a long vowel and is Romanized with a macron. However, there is no long form of the schwa; there are only ā ī ū. Note that high tone is Romanized with a grave accent, as in à, to keep in line with its descendants where this tone develops a final glottal stop.

Although there were only two tones, vowel sequences like àa and aà were becoming more common, and this is what led to the long tones of Khulls and its descendants, which are spelled ā and á respectively. Long tones also existed in Thaoa and Poswa but died out. The àa ~ ā type is much more common than aà ~ á. These could also occur with diphthongs, but only on the ā tone. That is, ài was common but aì was entirely absent, even over morpheme boundaries. Because of this pattern, the grave accent is omitted in the Romanization of falling diphthongs.

Grammar

Gold grammar was complex, as it was in the process of changing from a prefixing structure to one based on infixes and suffixes, where the infixes tended to appear on the last syllable of the word.

Nouns

See Gold nouns.

Nouns were often preceded with classifier prefixes, a trait that existed in Andanese at the same time, but soon died out in the Gold side of the family.

Noun classifiers

Many noun morphemes were very short, often consisting of just a single CV syllable. Many of the shortest roots were highly polysemic. However, Gold still retained the classifier prefixes of its parent language, a trait it shared with its close relative Andanese but which died out fairly early on in all of the descendants of the Gold language. Thus, the Gold language had prefixes, whereas its descendant languages such as Poswa, Khulls, and Pabappa formed their words with suffixes and infixes.

Noun classifiers could not carry stress. Thus, very few nouns were stressed on their first syllable: there was a zero-morpheme classifier which was used for certain very commonly used nouns, but even these nouns were not always stressed on their first syllable.

Because of the classifier prefixes, nouns were often three or more syllables long, although monosyllabic roots were not uncommon either because the many possible readings of a single-syllable noun root could be easily resolved by those same classifier prefixes.

Different noun classifiers could attach to the same roots. For example, katăda meant "tree", and kadăči meant "pear tree". By contrast, by changing the prefix, one can say gităda "fruit" and gidăči "pear". Similarly, kaŋŭta meant "tree trunk" and liŋŭta meant "bone".

Descendants

All of the descendants of the Gold language lost the noun prefixes early on. They hung on, irregularly, in a few words, often in cases where the speakers did not know they were originally noun classifiers. Because of the deletion of the classifiers, all five main branches of the family faced severe problems with homophony early on, and many word roots simply disappeared from the language. Khulls preserved the greatest number of these, as it was the only branch to retain tones, and preserved more distinctions among the consonants than the other branches.

The reason for the loss of the classifier prefixes was a sound change in the history of the Gold language that deleted all unstressed initial vowels. Thus, classifiers that consisted of single vowels were deleted, and the number of words with no classifier prefix was greatly increased. The classifiers remained in the language up to the stage of the classical Gold language, but they were unstable and were lost in all of the descendants except for the irregulars mentioned above.

Marking pluralization

Pluralization was marked consistently by a prefix u- prepended to the noun, before the classifier if there is one. This succumbed to a sound shift deleting all unstressed word-initial vowels, voicing any following consonants, and labializing the consonants if the lost vowel was u-. Because of this sound shift, new classifier prefixes were created that began with voiced labialized consonants. For example, the classifier for fish and other sea life is sa-, but in the plural, it has become wa-. Likewise nu-, the classifier for buildings, becomes nʷu- in the plural.

Marking the accusative case of animate nouns

In a much older stage of the language, the accusative case of animate nouns had been marked by prefixing i- to the affected noun. This prefix was lost due to grammatical reanalysis at much earlier date than the aforementioned deletion of initial vowels, and therefore there are no matched pairs of classifiers beginning with voiced and voiceless consonants. However, the disappearing i- did affect the initial consonants of noun classifiers in other ways, which had become opaque by the time of the classical Gold language but nevertheless persisted even when noun cases came to be marked by mandatory suffixes instead of prefixes (the accusative for most nouns is -ḳ or a derivative of it).

An example of this shift is the prefix for human males, tə-, changing to hə- in the accusative case, alongside the addition of the accusative suffix on the other end of the noun. In this morpheme, the initial h- descends from what was once a palatalized t-. This prefix may further change depending on the gender of the agent; some genders can "affect" males while others cannot.

Verbs

Private verbs

The grammar of the Gold language was the last to preserve the private verbs of its parent language. Private verbs are those whose meaning is dependent on the noun classes of the subject and object precedes it. Noun class in this context includes species and gender. Thus, for example, only humans have special verbs relating to holding objects. Only "fish" had words for swimming. For example, nusan was a type of fish, where nu- is a classifier for fish. The verb for "swim" is bĭ. Other animals have no verbs for swimming; a duck or human would thus need to take a specially modified form of the verb, nubĭ. Thus, all words describing swimming begin with nu-, though this is omitted when describing a fish. Nubĭ could be analyzed as "to move like a fish".

Infixes

All infixes are inherently stressed, with short, high-tone vowels. All infixes are inserted in the final syllable of the word. All exceptions to this pattern are surface manifestations of tonal sandhi and historical sound changes that have become grammaticalized. Some infixes appear to extend out to the end of the word because of sound changes like these. For example, all of the noun case markers were originally infixes, even though many now appear to be suffixes.

Orthography

The orthography of the language was an alphabet derived from the Tapilula syllabary. The letterforms were of the angular, less-ornate western branch, but the letter order was derived from the eastern (Andanese) branch.

Vowels and consonants were considered to belong to two different alphabets and either of the two could be placed first. In the descendant languages, however, the tradition of placing the consonants first came into practice.

The letter order for consonants was:

ʕ l j h ḳ k ŋ p m t w n hʷ g s d ġ b z č ǯ

And for vowels:

a i u ə

These were used as numbers. One pattern was

 ʕ  l  j  h  ḳ  k  ŋ  p  m  t  w  n  hʷ g  s  d  ġ  b  z  č  ǯ
 0  1  2  3  4  5  6  7  8  9 10 11 12 13 14 15 16 17 18 19 20

But this was not the classical alphabet order because the original alphabet was limited to 20 columns, not 21.

Descendant alphabets

Khulls

See Khulls script.

The classical Khulls alphabet was:

p ṗ b m h ʔ ʕ ḷ ṡ ṣ̌ z ŋ̇ ṁ ṅ l x k ḳ ġ ŋ t ṭ d n gʷ xʷ g s r š ž č ǯ kʷ ḳʷ ġʷ pʷ ṗʷ bʷ ʕʷ hʷ
a i o u e

The vowels had six tones in stressed syllables and one for unstressed syllables, but most writers used only six of the seven columns, and later most used only five, as two of the tones had come to be distinguished more by sandhi than by their primary tone.

Proto-Moonshine

The Proto-Moonshine alphabet was:

p b ʔ ʕ l s š z ŋ m n j h k t w hʷ g r ž č ǯ ň kʷ ŋʷ pʷ mʷ gʷ 
a i o u e

There were four tones: three for stressed syllables and one for unstressed syllables. Tone sandhi had nearly disappeared from the language, causing two of the tones to merge with two other tones, and a later unrelated sound change removed both members of one of these tone pairs entirely.

Thaoa

The Thaoa alphabet was:

l j h k kʰ ŋ p pʰ m t tʰ n s x š b ž č ň ʔ
a i u y e o

Thaoa had a six vowel system. Early on, the three secondary vowels y e o were considered to be centralized variants of the three primary vowels a i u, because even though y had been in the language for a long time, it was very rare. Later, however, Thaoa speakers came to see the setup as a six-vowel system with no internal divisions, and y remained the rarest vowel of the six.

Babakiam

Babakiam did not have a stable orthography, as its people were fond of creating ornate artistic scripts which imitated those of the Andanese people who lived amidst them. These artists of the alphabet could be likened to calligraphers, but they worked in a more abstract medium; for example, one form of the alphabet replaced the letters with interlocking shapes resembling a Tangram puzzle.

Letter order was variable, since the alphabet itself was seen as a work of art, and by grouping the letters together in a particular sequence a distinctive picture could be formed.

However, the use of the original script survived, and the most common letter order in use for it was:

p m s b y v k ŋ š ž t n f č
a i u ə ā ī ū

Long vowels were mostly descended from vowels that had been on the ā tone in the Gold language. They were no longer pronounced with a distinctive tone, but their length had remained. The Bābā people considered them to be separate vowels rather than variations of their short counterparts, even though the quality of each vowel was unchanged.

Descendants

Gold divided into two branches, one for Khulls and the other for Babakiam. Babakiam is the ancestor of Poswa and Pabappa.

• Khulls (by far the most widely spoken language, having more descendants and more speakers than all the others combined)

1. Proto-Moonshine language

• Macro-Pabap languages
• Babakiam (and its descendants Poswa and Pabappa)