Frislander wrote: ↑Tue Oct 23, 2018 6:18 pm
Nominal Morphology
Your wish is my command:
Telpahké morphology
Telpahké, compared to its parent language and closest relatives, is a fairly lightly inflected language. Unfortunately, however, this does not mean that the language's morphology is
simple. Pervasive morphophonological alternations including stress-induced apophony and syncope, hinted at above, make the surface instantiations of inflected words less than predictable from their citation forms.
This series of posts will deal solely with
forms- usage will follow later. Rather than attempting to info-dump all the morphophonological details in one go to start with, I'll instead try to point it out and explain it as we encounter them. So, let's look at the morphology of nouns, pronouns, adjectives and verbs in Modern Telpahké.
Nouns
Telpahké nouns distinguish three cases and two numbers. The marking of case is the same regardless of number, so we shall examine that first (also, number marking is what is technically known as a "pain in the arse").
Case
The three cases of Telpahké are the absolutive, the ergative and the accusative. The case inflections marking these three can be represented as -
Ø, -
Vθ and -
V́, respectively. And this is where the
fun complication starts. Essentially, the difficulties arise due to the interplay of a mobile stress and stress-induced vowel alternation.
Telpahké nouns can be divided into two classes, based on whether the stress is
fixed (the "static" class), or whether it is
mobile (the "dynamic" class).
- Static nouns are straightforward: the stress is always fixed on the final syllable of the word, made simpler given that the final syllable is also open. As an example, let us take the noun amá 'mother': the absolutive is amá, the ergative is amáθ and the accusative is amá. Note the syncretism between the absolutive and the accusative here.
- Dynamic nouns are somewhat more complex. They can be subdivided into two classes which, borrowing terminology from Proto-Indo-European studies, I refer to as "hysterodynamic" and "proterodynamic":
- Hysterodynamic nouns are consistent in having the stress towards the end of the noun (hence the name): on the final syllable in the absolutive and accusative cases, and on the penultimate syllable in the ergative. An example here is yíǝl 'chicken': the absolutive is yiǝl, the ergative is yíǝloθ and the accusative is yeló.
- Proterodynamic nouns, however, shift the stress "backwards" in the case of the absolutive case. For example, let's look at sátol 'soldier': absolutive sátol, ergative sɔtóloθ and accusative sɔtoló.
Here's where we run into the first complication: notice the alternations
yiǝl~
yeló and
sátol~
sɔtoló. The astute will remember here the table of vowels
in a previous post:
| stressed | unstressed |
| á | ɔ |
| ɛ́ | a |
| ɛ́ | ɛ |
| é | e |
| ɔ́ | o |
| ó | o |
| íǝ | e |
| í | i |
| úǝ | o |
| ú | u |
As can be seen, the only ambiguity is whether stressed
ɛ́ becomes unstressed
a or
ɛ: this is not synchronically predictable. Aside from this, going from stressed to unstressed is straightforward, but the other way is more problematic. For example, in theory the ergative of
sátol could be any of
sɔtɔ́loθ,
sɔtóloθ or
sɔtúǝloθ: again, the correct outcome is not predictable synchronically. Ironically, this is about the only area where the native script has an advantage over the romanisation.
A further complication is the nature of the vowel of the inflection (the thematic vowel): yet again, this is not synchronically predictable. For "regular" dynamic nouns there are four options:
| 1 | 2 | 3 | 4 |
| ergative | aθ | ɛθ | eθ | oθ |
| accusative | ɛ́ | é | é | ó |
Let's take an entry from my lexicon file: "
ɛ́fal np3 'home'; stem
afɛ́l-". This basically tells me that
ɛ́fal 'home' is a class 3 (3) proterodynamic (p) noun (n), and its ergative stem is
afɛ́l-. Thus, absolutive
ɛ́fal, ergative
afɛ́leθ, accusative
afalé.
To summarise, to decline a regular noun you need to know the following:
- If it's a static noun or not. If it is, bung a -θ on the end for the ergative and call it a day.
- If it's not a static noun, is it hysterodynamic or proterodynamic? If it's hysterodynamic, you will also need to know which class the thematic vowel is.
- Finally, if it's proterodynamic, you will need to know not only the class of the thematic vowel, but also the ergative stem.
As I believe I have said elsewhere, describing a category as "regular" depressingly implies the existence of a category which can be described as "irregular". So it is unfortunately with Telpahké nouns.
The largest class of irregular nouns are historically derived from those nouns which ended in a nasal consonant in the parent language. For example,
pallú 'slave' at first glance appears to be an unobjectionable static noun with a stressed word-final open syllable. It isn't: it's a class 1 hysterodynamic noun which inflects
pallú~
pallúnaθ~
pallunɛ́ - essentially the etymological
n was lost when word-final, but reappears in derived forms.
Similarly, a noun like
harán 'script, writing' is a totally normal class 1 hysterodynamic noun which inflects
harán~
harámaθ~
harɔmɛ́: here etymological
m has become
n when word-final, but reverts to
m when inflected. In the same vein, we have words like
fɛ́n 'fishing net', wherein an etymological
ŋ has become
n word-finally, so
fɛ́n~
fɛ́ŋoθ~
faŋó.
Another significant group of dynamic nouns masquerading as static nouns are those ending in
íǝ or
úǝ. While a most nouns ending in these sounds
are static nouns (such as
fɔríǝ 'lightning'), an annoyingly significant number aren't. The ones that aren't static nouns generally have an underlying semivowel as part of their stem. For example,
moníǝ 'dagger' is
monɛ́yaθ in the ergative and
monayɛ́ in the accusative. This group is particularly important as the common pluralising affixes -
míǝ and -
θíǝ inflect in this way.
The final group of irregular nouns to watch out for are those which end in a final unstressed vowel. All of these are actually proterodynamic nouns that have lost a word-final consonant (most of the time
h) which resurfaces in the inflected forms. For example, the original name of pirate queen whom
we met earlier,
Fúǝnko, falls in this class:
Fúǝnko~
Fonkɔ́heθ~
Fonkohé. Unfortunately, because unstressed vowels before
h are particularly prone to syncope, things like the paradigm of
míǝro 'boulder' exist:
míǝro~
maróhaθ~
merhɛ́. (No,
maróhaθ isn't a typo. Unstressed
e becomes
a before
r followed by a stressed vowel. Sorry.)
Speaking of syncope, there is a class of nouns whose final consonants are
r or
l that appear to undergo syncope in their inflected forms. For example,
súǝmar 'householder' inflects
súǝmar~
súǝmraθ~
somrɛ́. As it happens, diachronically this is not a case of syncope: the ancestral form was
jʰōmra. What has actually happened is a case of anaptyxis in the absolutive case after the loss of word-final unstressed vowels.
Number
Native speakers of Telpahké will tell you happily that number marking on nouns is entirely optional (not neccessarily in that manner, of course) and that
arún can be used as easily for one coconut as it can for a whole pile of them. They are lying.
Number marking on nouns is obligatory in Telpahké, just not in the same contexts as it is in English. If you were to point at a pile of coconuts and exclaim "arún!", natives would gently correct you (as one does to people who get overly excited about coconuts) to
arunmá, which is the correct plural form. However, if you were to chirp "arunmá hin!" on espying only
three coconuts, you would be again corrected to
arún hin. Supposing this psychotic break from reality were to continue and you saw a rat eating the coconuts; duly concerned, you might turn to your companions and say "kóreθ ehkaná in arunmá", informing them that
a rat is eating the coconuts. They would again correct you to
arún, as number is already marked on the verb. By this point, entirely sick of the farcical manner in which number is encoded, you might mutter to yourself "aruné noθíl i ɛsá", stating that you hate coconuts. Again, you would be corrected to
arunmá, because here number is
not marked on the verb as coconuts are no longer the subject. At this point, you would be justified in using one of the coconuts to thump your interlocutor.
Those who are following along at home might think that they now have the entire number thing in Telpahké sussed: the suffix is -
má, and it produces a static noun ("Please, sir! Can I have a house point sir!"). Wrong. ("No, Jenkins, you cannot.")
Unfortunately, the picture is far more complex than this. The marking of number has inherited a number of quirks and oddities from the parent language, which Telpahke has only made more impenetrable. In addition to using a range suffixes which are not entirely predictable synchronically, there are also vestiges of the parent language's inverse number marking, whereby the plural is marked by
removing rather than adding a suffix. Furthermore, a small yet significant number of nouns mark plurality by means of reduplication, which soundchange often renders impenetrable. From an inflectional point of view, plural nouns mark case in the same way as singular nouns; and it is possibly best to consider plural marking as derivational rather than inflectional morphology. Nevertheless, let us look at each of these strategies in turn:
Suffixing plurals
There are five plural suffixes in common use. They are not interchangeable, and the suffix a noun selects is an lexical property of that noun. That is, you can't pluralise
arún as *
aruní just because you fancy it:
arún can only select -
má to become
arunmá.
It should come as no surprise that it is not entirely possible to predict which suffix a noun will take synchronically. However, there is some semantic correlation between the suffixes:
- -k is most commonly used for human referents: sátol 'soldier' > sɔtólok 'soldiers'. It is formed by replacing the θ of the ergative with k, and produces proterodynamic nouns from hysterodynamic nouns and hysterodynamic nouns from static nouns - both of class 2. Thus, for example, static aθá 'father' becomes hysterodynamic aθák 'fathers' (with ergative aθákɛθ and accusative aθɔké); and hysterodynamic sokúr 'friend' becomes proterodynamic sokúrak 'friends' (with ergative sokurɛ́kɛθ and accusative sokuraké).
- -má generally patterns with organic, non-animal objects: fruits, trees, plants and so on. It creates static nouns, as seen with arún 'coconut' > arunmá 'coconuts' above. It is added directly to the absolutive stem.
- -θíǝ and -míǝ are worth treating together as they both denote plurals of inanimate objects, the distinction being that -θíǝ (or -tíǝ after a consonant) is generally reserved for man-made objects. Thus kɛ́ma 'trunk' > kamamíǝ 'trunks' but suǝn 'house' > sontíǝ 'houses'. The resultant forms are declined like moníǝ above.
- -í is a bit of an all-purpose leftover pluralising suffix, but it is particularly common with non-human animals, e.g. pehɛ́r 'cat' > peharí 'cats'. The resultant form is a hysterodynamic class-1 noun which declines like pallú above.
Suffixing singulars
Nouns which remove a suffix to indicate the plural (or add a suffix to indicate the singular, if you prefer) are fortunately not quite as common as they are in Tarì. They do, however, form some of the oldest strata of the language's vocabulary and again, the relationship between the singular and the plural form is often less than transparent. For example, there is
fénar ‘tree’ >
fen 'trees',
fɛ́łtol ‘foreigner’ >
feɫt 'foreigners',
rohł 'pig' >
rok 'pigs'.
Reduplicating plurals
It is worth mentioning that a) nouns which use reduplication to form the plural are among the most common in the language, such as
siǝl 'woman' and
tuǝn 'man' (pl.
sɛhsíǝl and
tahtúǝn, resp.) and b) have often been so obscured by soundchange that it might just be worth thinking of them as being suppletive anyway. It is also worth noting that totally obscure reduplication marking grammatical oppositions will also become an issue when we get to verbs. Hurrah!
(transliterated ce-ṡ-ˈcē-la according to the above scheme) is variously pronounced /sɛhˈsiǝl/, /ɕɛʔˈɕiɛl/, /ʃeʃˈʃejl/, /tsǝˈtsʰajʎ/, /tʃeˈʃtʃeːlǝ/ or /sǝxˈsjaːw/ (the first is the north-eastern variant found on Paríf and neighbouring islands) - all of which derive from an Old Telpahké cecʰēli.
I have to confess that Indonesia has not really been a conscious inspiration here: the language only really gained an Austronesian-style alignment by accident. Partially this is due to a lack of good, accessible resources on Indonesian native cultures/traditional daily life, while there are plenty on India, Japan and (as it happens) the Aztecs, who are my main three sources of inspiration for Imparin culture. Of course, given the similarities between the geography, geology and climate of Indonesia and the Spice Islands, a large degree of similarity is inevitable: these are going to be wet-rice cultivating canoe-using builders in wood who don't wear a great deal, not barley-and-cattle agriculturalists with clinker-built longships living in stone castles and wearing furs.
