New CBB

Kantaranyan is a language which I've been developing in various forms for years now. In fact, the name at least originates in the name of the first conlang I came up with, although the language as it is now shares very little (I think about two words including the name of the language) with it. I'm (hopefully) approaching a version of the language I'd like to settle on, and would consequently like to be able to get some second opinions on it. I'm particularly interested in input on syntax and morphology, because I feel less confident in those areas, but in this post I'll focus on the phonology so that you know what the words are supposed to sound like. This is meant to be a reasonably naturalistic language, so hopefully nothing here is too unlikely.

So, Kantaranyan phonology:

The phoneme inventory:

Plosives: /p t k q/
Nasals: /m n/
Fricatives: /s ʃ h/
Approximants: /β l j ʁ/ (all of these sounds, especially the bilabial and the uvular, may be pronounced with light friction, but they pattern more like sonorants than fricatives)
Vowels: /i a u/

As far as transcription goes, I'll be transcribing /ʃ/, /β/, /j/ and /ʁ/ as <x>, <v>, <y> and <r> respectively, and the others as in the IPA.

Phonotactics:

Closed syllables are allowed, but complex onsets and codas are not. Fricatives are not permitted in codas, but all other types of consonant may occur. Syllables without onsets are permitted, but vowels are not allowed in hiatus within a word unless the first vowel is high and the second is /a/.

If a suffix consisting of a single consonant (such as the plural marker /ʁ/) is added to a base that already ends in a consonant, then the last consonant of the base is deleted (for example /aʁβiat/ (woman) + /ʁ/ -> /aʁβiaʁ/).

Finally, /h/ is not permitted to occur in consonant clusters at all.

Allophonic rules:

-/i/ and /u/ are lowered when adjacent to a uvular consonant (normally resulting in [e] or [o], except where the syllable is closed with /ʁ/, in which case the surface forms are [ɛ] and [ɔ]). This change is blocked if the high vowel is immediately followed by /a/.

-/a/ is fronted to [e] when the following syllable has an underlying /i/, unless the /a/ is adjacent to a uvular.

-If a high vowel is followed by a high vowel in the next syllable, then it is lowered (to [e] or [o]). A high vowel lowered by an adjacent uvular does not cause a preceding high vowel to lower. This can cause vowel alternations within a word, for example /isumi/ (cloud) -> [ˈisomi], but /isumiʁ/ (clouds) -> [ˈesumɛʁ].

-/a/ is fronted to [e] if there is a coronal consonant in the coda of the same syllable, unless the onset is a uvular consonant.

-A syllable final plosive is debuccalised.

-If a nasal precedes an approximant, the nasal is deleted, while the approximant becomes a nasal with the same position of articulation as the underlying approximant. This causes the preceding vowel to be lengthened in compensation.

-A coda nasal is otherwise assimilated to the position of articulation of the following consonant.

-If an approximant precedes another approximant, the first is deleted, and the preceding vowel lengthened in compensation.

-An epenthetic /h/ is inserted between two vowels than cannot appear in hiatus. Thus, for example /a/ + /akam/ > [ˈahakam] (its heart).

-Finally, /a/ is pronounced [ɑ] adjacent to a uvular which surfaces as such (so excluding debuccalised /q/ and elided /ʁ/).

Stress assignment:

Stress falls on the rightmost non-final heavy syllable, and otherwise on the first syllable of a phonetic word if there is no such syllable.

Secondary stresses fall on all non-final syllables separated from a stressed syllable by a single syllable.

I'll finish off with a few examples or words and their surface forms:

Maku /maku/ [ˈma.ku] (house)
Arviat /aʁβiat/ [ˈaː.βi.eʔ] (woman)
Riaxat /ʁiaʃat/ [ˈʁi.a.ʃeʔ] (water)
Milu /milu/ [ˈme.lu] (pot)
Tanra /tanʁa/ [ˈtɑː.ɴɑ] (tongue)
Vahal /βahal/ [ˈβa.hel] (head)
Uruniq /uʁuniq/ [ˈo.ʁo.neʔ] (fish)
A h akam li /aakamli/ [ˌa.ha.ˈkeː.ni] (inside it)
Kamtaranyamarimi /kamtaʁanjamaʁimi/ [ˌkan.tɑ.ˈʁɑː.ɲa.ˌmɑ.ʁe.mi] (Kantaranyan language)

So, that's how the phonology works out at the moment. I'd like to know what you think of it.

------------

The above phonology is rather out of date, as I've rethought various things. I'm currently working on writing up a nice grammar for the language, and I'll link the the current versions of the various parts as I have them written. Look at these for up to date information!

Grammar pdf links:

Phonology
Diachronic Phonology

_________________
Languages!
Native: Proficient: Decent: Have dabbled in:

Ok, morphology and syntax as promised (or what I've actually managed to work out of it). This is kind of a work in progress so it'll be a bit sketchy and won't cover everything. I'd particularly appreciate help/feedback here (syntax especially is not my strong point), especially as regards the verbs.

Nouns:

Nominal morphology as it presently stands is relatively simple. The basic form of a noun is the singular, and (if it's a countable noun, anyway), the plural is formed by the addition of the suffix -r (which as noted in the phonology above deletes the last consonant of the stem if it ends in one).

Some examples of plurals:

Amak (person) > amar (people)
Maku (house) > makur (houses)
Kam (tree) > kar (trees)

There aren't any case endings. I intend for the primary case roles (at least agent and patient) to be marked on the verb rather than the nouns (somehow), while oblique case roles are denoted by postpositions. The postpositions are clitics and form part of the same prosodic word as the noun. (Given that so far I don't have anything that would actually come between the noun and the postposition though, it might actually be better to analyse them as case endings after all. I'll probably revisit that when I've thought out the syntax in more detail.)

The postpositions I've thought of so far as the locative marker li; ka, which denotes the destination of motion; na, which denotes a point of origin, and the instrumental marker i.

More detailed descriptions of location and the like can be given with prepositional phrases, while are usually derived from the names of body parts: for example, maku a h akam li, literally 'in the house's heart', means 'inside the house', or piqta a vahal ka (lit. 'to the mountain's head') means 'to the top of the mountain'.

Possession is denoted by truncated forms of the pronouns, which precede the noun phrase which is possessed, giving forms like niq maku (my house), mi murat (his/her cat) or murat mi takxa (the cat's tail). (The full forms for 'I' and 'he/her' are naniq and ami respectively.)

As regards verbs, I'm still trying to work out exactly how I want them to work. My main intention is to be able to use word order for something other than denoting syntactic roles (such as denoting things like the topic or focus), and would like to use the verbal morphology to signal which argument is the agent, and which is the patient (as I intend Kantaranyan to be predominantly head-marking). On the other hand, I don't want to have too many affixes on the verb.

The options I've thought of so far are either marking the agent on the verb (although I think I might need to add more genders or noun-classes for that to work, as at the moment I'm only distinguishing between animate and inanimate referents in the 3rd person), or otherwise to have some sort of system whereby a transitive verb is marked in some way if it's agent is less animate than the patient. Any suggestions here would be welcome.

As for tense, aspect and mood, I'll probably mostly be using auxiliary verbs, although I think I'd like to mark some tense or aspect distinctions on the verb itself. For example, I'd quite like to use reduplication to form iteratives. I'd like the basic unmarked form of the verb (i.e. without affixes) to denote a basic present tense (possibly present progressive).

Finally, I intend for adjectives to largely operate like intransitive verbs.

_________________
Languages!
Native: Proficient: Decent: Have dabbled in:

Nice language. I didn't see anything unnaturalistic in your language. I’m a big fan of small phonemic inventories like yours. Your language's phonology reminds me of some of the Inuit languages, such as the presence of the two uvular consonants and the lowering of high vowels when adjacent to uvulars. One of my conlangs called Inaqtuk as a very similar phonology to yours. Keep up the good work!

I need to look over this again and better.

_________________
I am not responsible for the accuracy of my sources; they're responsible for their own mistakes, if any, and also responsible for defending their own statements if you disagree with them.

Thanks! Glad you like it. Funnily enough I don't think I ever had Inuit in mind when making this, although I can see the resemblance now you point it out.

The phoneme inventory's got steadily smaller the longer I've worked on this language (although I don't think it's going to get any smaller than it is now). I added /q/ in recently because I noticed that by taking out phonemic /e/ and adding /q/ I could make the vowel system have a more interesting phonology.

_________________
Languages!
Native: Proficient: Decent: Have dabbled in:

Having learnt more about stress and feet earlier this week, I've had a rethink of the stress system in Kantaranyan. The result is as follows:

Kantaranyan has binary, quantity sensitive trochaic feet, which are constructed right-to-left (under the constraint that all heavy syllables constitute feet in their own right). The main stress tends towards the right edge of the word, but will fall on a heavy syllable out of preference.

In practice, this means that primary stress will fall on the rightmost heavy syllable, or if there are no heavy syllables in the word, on the penultimate, with secondary stresses falling on every (other) heavy syllable, and every light syllable separated from a heavy syllable, the primary stress or the right edge of the word by one other light syllable to its left.

To illustrate the system, here are the example words from the first post parsed with the new rules (brackets indicate feet, acute accents primary stresses, and grave accents secondary stresses):

(máku) [house]
(àr)vi(át) [woman]
(rìa)(xát) [water]
(mílu) [pot]
(tán)ra [tongue]
va(hál) [head]
(ùru)(níq) [fish]
(àha)(kám)li [inside it]
(kàm)ta(rán)ya [Kantaranya]
(kàm)ta(rán)(yàma)(rìmi) [Kantaranyan language]
ma(rími) [language]

Unfortunately, I haven't had any inspiration on the morphosyntax front yet. I was wondering if anyone could suggest some languages with interesting ways of marking the agent and patient that I could look at for inspiration. I could do with looking at some actual head-marking languages to to get a better sense of how they do things. Any suggestions?

_________________
Languages!
Native: Proficient: Decent: Have dabbled in:

So, I've decided to go with the direct/inverse alignment in the end. I've refined the basic points of the grammar, so I'll put them up here (this will be a rather long post as a result!).

Parts of speech:

Kantaranyan has two parts of speech: nouns (including pronouns) and verbs.

Nominal morphology:

Nouns are inflected for number, with the singular taking no suffix, and the plural taking the suffix –r, which replaces the last consonant if the root of the noun ends in one.

Nouns also take affixes to mark possession, as well as postpositional suffixes which take the role of basic adpositions.

The possessive affixes combine prefixes and suffixes. The prefixes mark the person of the possessor, as follows:

1st person singular: ni-
1st person plural inclusive: u-
1st person plural exclusive: xi-
2nd person: lu-
3rd person, animate: mi-
3rd person, inanimate: a-

In addition, the noun takes the suffix –mi if the possessor is plural. This is the case even when there are separate plural prefixes, as in the first person.

The postpositional suffixes are affixed after the plural markers. They are as follow:

-li (location)
-i (instrument)
-ka (destination of motion)
-na (point of origin)
-nu (a point passed, or passed through, by a motion or action)
-ti (comitative)

The personal pronouns, in the full forms, are the following:

Niq – 1st person singular
Luat – 2nd person singular
Ami – 3rd person singular animate
Ahi – 3rd person singular inanimate
Umi – 1st person plural inclusive
Xini – 1st person plural exclusive
Rua – 2nd person plural
Mar – 3rd person plural animate
Yir – 3rd person plural inanimate
Kiv – reflexive

Pronouns, like nouns, may take postpositional suffixes.

Demonstratives:
Kantaranyan distinguishes three demonstrative pronouns, based on distance. These are:

Sia (referring to something nearby)
Kia (referring to something far away, but still visible)
Nua (referring to something out of sight)

The demonstratives do not inflect for number, but like other pronouns may take postpositional suffixes. They can be used to modify nouns by removing the final a, and prefixing the demonstrative to the noun root. For example ‘kimaku’ (from ‘maku’) means ‘that house’.

The interrogative pronoun yaku functions in a similar manner. Its prefixed form is ya-. (E.g. yamaku = what house?)

Verbal morphology:

Kantaranyan verbs inflect for aspect, as well as, in the case of transitive verbs, to denote the agent. Tense is unmarked, and must be determined from context, while modality is expressed using postverbal particles or compound verbs.

There are three main aspectual distinctions: the progressive (denoting an action that is in progress, or a state that is in effect, at the time of reference), perfective (denoting a completed action, or an action within a defined limits) and anterior (referring to an action or state that took place before the point in time the clause refers to). In addition, verbs may form an iterative, which denotes that the action takes place multiple times (with the progressive this has a habitual sense, while with the perfective it refers to multiple discreet instances of an action – compare ‘ami milur qaqatu’ (he makes pots) with ‘ami ahun milur qaqatihu’ (he made seven pots)).

The basic aspects are denoted by the following affixes, which are suffixed to verbal stems ending in consonants, and infixed into verbal stems ending in vowels:

Progressive: -0-
Perfective: -i-
Anterior: -uk-

For example, the basic forms for the verbs ‘axa’ (to eat) and ‘tiat’ (to give) are as follow:

Axa, tiat (progressive)
Axia, tiati (perfective)
Axuka, tiatuk (anterior)

The iterative, on the other hand, is formed by reduplication: of the first syllable (excluding its coda) if the verb begins in a consonant, and of the second if it begins with a vowel. The iterative progressive forms of axa and tiat are thus axaxa and titiat.

The inverse form:

In the case of intransitive verbs, there is no problem determining the subject, as the verb has only one core argument. In the case of transitive verbs, the agent is normally assumed to be the noun with the highest animacy, or, if both nouns have the same animacy, the noun with appears first in the sentence. Animacy is determined according to the following hierarchy:

1st person >> 2nd person >> adults, supernatural forces and dangerous natural phenomena >> children >> animals >> inanimate objects.

If, however, the agent is of a lower animacy, or appears second out of two nouns of the same animacy, then the verb appears in the inverse form, marked by using a different set of aspectual affixes, which follow:

Progressive: -u-
Perfective: -in-
Anterior: -am-

Valency modification suffixes:

Suffixes can be added to transitive verbs to mark that one of the arguments is not expressed. These are –ya, which marks that the patient is not expressed, and –ni which marks non-expression of the agent. E.g. Ami pula axa (he is eating a fruit) vs. ami axaya (he is eating) vs. pula axani (the fruit is being eaten).

Postverbal particles:

Postverbal particles are used to indicate things like modality. The main particles are the following:

May (optative)
Lat (marks hypothetic statements, as opposed to statements of fact)
Kun (marks that the speaker is uncertain of the validity of the statement)
Yak (marks that the statement is a question)
Man (marks the statement as negative)

Multiple markers may be used, in which case they are often formed into a single word.

Further nuances of aspect and modality can be denoted by compounding the verb with roots like ‘mahi’ (be able) or ‘kam’ (start to, become). E.g. ami milur qaqatu (he makes pots) vs. ami milur qatumahi (he can make pots); ami kulu (he is ill) vs. ami kulukami (he got ill).


Basic syntax:

Kantaranyan sentences are verb final, so, excepting any post-verbal particles, the main verb will always be the last word in the sentence. The nouns in the sentence, however, may appear in any order, although the first is usually the topic, and the last noun phrase, immediately before the verb, is usually the information given about the topic that the speaker considers to be of greatest importance.

A verb, with or without other arguments, may also be used to modify a noun, in which case it precedes the noun to be modified. If the verb is transitive, then the modified noun will normally be assumed to be the agent. If the modified noun is the patient, then the verb appears in an inverse form. E.g. tariat tavi (the bird is red) vs. tavi tariat (a red bird); milur qaqatu yahi (a man who makes pots); yahi qatinu milu (the pot the man made).

Kantaranyan does not have ditransitive verbs. Instead, verbs like ‘tiat’ (give) take the recipient as their patient, and treat the thing that is being given as the instrument. Thus ‘He gave me a pot’ would be translated ‘Ami niq miluhi tiatin’. Doing something on behalf of someone else is expressed as giving them a nominalised form of the verb, thus, for example, ‘He made me a pot’ in Kantaranyan is ‘Ami niq mimiluqatumi tiatin’.

Possessive constructions:

The equivalent of English’s X’s Y construction is formed in Kantaranyan by placing the possessed noun after the possessor, with the possessed taking possessive affixes to agree with the possessor, thus for example, yahi mimaku = the man’s house; qaxiq alami = the knife’s edge; kar apularmi = the fruits of the trees and so on.

The equivalent of X has Y is expressed with the possessor in the locative, followed by the possessed, which is marked with the appropriate possessive affix, and then the verb ‘ixa’ (exist). E.g. yahili mimilu ixa = the man has a pot.

Postpositional phrases:

Postpositional phrases are used to make finer distinctions in location, destination, to express temporal relationships and so on. They consist of a noun marked with an appropriate postpositional suffix, and can be used to modify a noun by being placed in a possessive relationship with it. Examples include:

Maku ahakamli ‘inside the house’ (literally ‘in the house’s heart’)
Piqta avahalka ‘to the top of the mountain’
Qimali ‘at night’

Conjunction and disjunction:

The forms ‘in’ (and) and ‘suy’ (or) can be used to link together words, phrases or clauses. Both can be repeated before the first element for emphasis, e.g. arviat in mimurat (a woman and her cat) vs. in arviat in mimurat (both the woman and her cat).

X is Y constructions:

In order to say that one noun is another noun, it is necessary to use suffix –xa to convert the noun into a verbal predicate. Thus, for example ‘kia nimakuxa’ = that is my house.

Obviously that's not everything I need for a fully functioning grammar, but it should be enough to be able to construct some sentences, and I hope to try to work out the other stuff I need by trying to express various things in the language. If anyone can stand reading through such a long post, I'd be happy to hear what you think (or if you can pick up on anything that seems unnatural or might get too unwieldy in practice.

_________________
Languages!
Native: Proficient: Decent: Have dabbled in:

One thing I completely forgot in the above is that verbs can also take directional prefixes.

These are used to indicate the direction of motion, and in fact the verb -la- (go) must appear with a directional prefix. The main ones are ka- (towards) and na- (away from), nu- (through) and su- (back again).

These prefixes can also be used metaphorically with verbs that do not describe motion, such as the verb kanuna (volunteer), which is derived from nuna (want).

_________________
Languages!
Native: Proficient: Decent: Have dabbled in:

As I mentionned in the 'random phonology' thread, I've been reading about Government Phonology (GP) lately, and, given one of the things I like to do with conlanging is experiment with linguistic theories, I've been redesigning Kantaranyan phonology within the framework of GP. For the benefit of those who haven't encountered the theory before I'll try to explain the basics as I go.

The following post concerns syllable structure, which in GP is defined by various government relations. To quote one of the papers I've been reading, phonological government "is defined as a binary, asymmetric relation holding between two skeletal points" (Polgárdi, 2006). One segment, the head, is the governor, and licenses the other, which is its dependent. With the exception of two special types of government (which I won't be employing in Kantaranyan, but will probably play around with in other conlangs at some point), government is local (i.e. a head can only govern a segment adjacent two it), and has a set direction. Furthermore, the dependent may not more more complex (I'll talk about complexity in a later post, but more complex segments are generally less sonorant) than the head.

The two main types of government are the following:

Constituent government occurs within syllabic constituents, and is left-headed. There are three such relationships:
-A branching onset (where the first consonant of the cluster is head, and the second dependent)
-A branching nucleus (i.e. to account for long vowels and diphthongs)
-A branching rhyme (where the nucleus is head, and the coda is the dependent)

Interconstituent government occurs between constituents (rhymes/nuclei and onsets), and is right headed. The main such relationship is that an onset governs the preceding coda. It's also worth noting that nuclei licence the preceding onsets, but this is not quite the same as the nucleus will often be less complex (i.e. more sonorous).

Which of the above government relationships are allowed is defined by parameters. In Kantaranyan, the parameters allowing branching onsets and nuclei are off, but the parameter allowing a branching rhyme, and onset-coda government, is on. Kantaranyan additionally does not allow empty nuclei or onsets, so the permitted syllable structures are thus CV and CVC.

Some further comments on codas are necessary:

-Firstly, due to the observations that word final codas often do not display the same restrictions as medial codas, and that medial codas often have a close relationship with the following consonant, GP analyses most final consonants as onsets followed by an empty nucleus (whether such an empty nucleus is licensed or not is defined by another parameter). On the other hand, some languages show a very restricted set of final consonants, similar to those allowed in medial codas, and in such languages final consonants may be analysed as codas. Kantaranyan is a language of the latter type.
-Another restriction is that 'superheavy' rhymes (consisting of a long vowel/diphthong and a coda) are not allowed, as the head of the nucleus cannot license the coda 'through' the second timing slot in the nucleus (as constituent government must be local). This is apparently contradicted by words like 'child' in English, which suggests that the coda is licensed by the nucleus itself rather than the head segment of the nucleus (this, however, is not possible in some languages, and apparently a branching nucleus may have less licencing potential - English coda consonants following long vowels are subject to stricter restrictions than those following short vowels, for example). This isn't an issue in Kantaranyan as branching nuclei do not occur.

Hopefully that made some sense

(In addition to the paper cited above, I also used the following as sources: http://www.lotpublications.nl/publish/articles/000515/bookpart.pdf, http://profilaktyka.kul.lublin.pl/files/30/gfx/CHAPT1.pdf)

_________________
Languages!
Native: Proficient: Decent: Have dabbled in:

This is a cool idea. I'm guessing that the process starts working on the word from right to left?


From the example that I see later on (specifically the words for "woman" and "water"), I'm guessing that you mean "an underlying coronal consonant"?

That was the plan. I'm presently redesigning the phonology a bit though, and I might be taking that process out, as the phonological theory I'm using as a basis doesn't really account for it. I might have a similar alternation, although it will work a bit differently.



Indeed I did, although once again this rule probably won't be staying in the new version (or if it does it will apply to all coda consonants and not just coronals and uvulars).

_________________
Languages!
Native: Proficient: Decent: Have dabbled in:

The new phonology continues with a bit of lexical phonology:

This post will be rather short. One of the principles of GP is the 'Projection Principle', which states that the government relations within the word are set in the lexicon and thus may not be modified by (post lexical) phonological processes. As some of the processes of affixation in Kantaranyan will affect whether a segment is syllabified as a coda or onset, these must be regarded as part of the lexical phonology. The main lexical phonological processes are the following:

-The nominal plural ending -/ə̯/ replaces the final consonant of a nominal root ending in one.
-The verbal aspect affixes are suffixed to a verb root ending in a consonant (in which case this consonant will be syllabified as a coda in the direct progressive, and an onset (with the affix as rhyme) in the other forms), or infixed into a verb root ending in a vowel (before the final vowel, in which case the final consonant of the affixes /in/, /am/ and /uk/ are syllabified as onsets rather than codas).
-A glottal stop is inserted into any empty onsets.

_________________
Languages!
Native: Proficient: Decent: Have dabbled in:

And now for the post-lexical rules (and the phoneme inventory).

Important here are element theory (which is how GP represents the phonological make up of sounds), and projection government.

In GP, sounds are composed of one or more 'elements'. Elements are either there or not there (so no +/- whatever as in feature based analyses of phonology), and each element is supposed to be phonetically interpretable in isolation - for example, the element |U| in isolation is the vowel /u/, but in combination with other elements can produce other rounded vowels or labial consonants, while the element |ʔ| in isolation is a glottal stop, but combined with other elements produces various stop consonants.

In Kantaranyan, I'm using the set of elements proposed here (in the 'segmental structure' chapter). I've chosen to distinguish the palatals and uvulars from coronals and velars respectively by the use of headed place elements (underlined below). Note that other papers propose different analyses of certain segments - especially in the case of the coronals.

Kantaranyan presently has the following phoneme inventory:

Consonants:
|ʔh.U|-/p/, |ʔh.I|-/t/, |ʔh.A|-/k/, |ʔh.A|-/q/, |ʔ|-/ʔ/
|ʔ@.U|-/m/, |ʔ@.I|-/n/, |ʔ@|-/N/
|h.I|-/s/, |h.I|-/ʃ/, |h|-/h/
|h@.U|-/β/, |h@.I|-/l/, |h@.I|-/ʝ/, |h@.A|-/ʁ/, |@|-/ə̯/

Vowels:
|@.U|-/u/, |@.UA|-/o/, |@.A|-/a/, |@.IA|-/e/, |@.I|-/i/

/ʔ/, /N/ and /ə̯/ appear in codas only (although note that [ʔ] does appear in onset position as an epenthetic consonant before an underlyingly vowel-initial syllable), while /s/, /ʃ/ and /h/ are limited to onsets. Consonants from either cannot appear as the final consonant in a consonant-final verb root.

I mentioned in an earlier post the notion of 'complexity'. This is fairly simple - a segment with more elements is more complex than a segment with fewer elements. (Note, however, that the source I linked to in this post suggests a slightly more complex approach whereby different elements have slightly differing properties, but this is not directly relevant to the phonology of Kantaranyan, so I won't go into detail on that.)

I also mentioned 'projection government', which is how GP deals with non-local processes like vowel harmony. Essentially, constituents with something in common (such as nuclei) can be 'projected' to a special phonological tier where the process operates, so that the government processes can be said to be local within the projection. The directionality of projection government is set in a given language by a parameter. In Kantaranyan it operates left to right. (More details on projection government here.)

To prevent this post from getting too long, the actual post-lexical processes I've set up for Kantaranyan are in a separate document:
http://dl.dropbox.com/u/41280522/Postlexical%20rules%20%28Kantaranyan%29.docx
(Note that the suppression of place elements in the coda is motivated by the need for final codas to be simple enough to be governed by the vowel in the nucleus, and that the 'consonant harmony' rule is based on projection government. Feet and stress are still determined as detailed in this post.)

The example words in the document are /qatu/ 'make', /esome/ 'cloud', /ʃaʃu/ 'fifteen', /kat/ 'take' and /taʔsa/ 'tail'.

Finally, to round this post out, here's the romanisation:


/p/ - <p>, /t/ - <t>, /k/ - <k>, /q/ - <q>, /ʔ/ - <h>
/m/ - <m>, /n/ - <n>, /N/ - <n>
/s/ - <s>, /ʃ/ - <x>, /h/ - <h>
/β/ - <v>, /l/ - <l>, /ʝ/ - <y>, /ʁ/ - <r>, /ə̯/ - <r>
/u/ - <u>, /o/ - <o>, /a/ - <a>, /e/ - <e>, /i/ - <i>

Note that epenthetic [ʔ] is not romanised.

Notes on final <r> and <n>:

-In final position in nouns, the graphemes <r> and <n> represent /ə̯/ and /N/ respectively.
-In final position in verbal stems, <r> represents /ʁ/ and <n> represents /n/.

Romanisation is based on phonemes rather than the surface form. The exception to this rule is the Romanisation of morpheme-internal codas, which are romanised according to the surface form as follows:

[p˺] - <p>, [t˺]/[c˺] - <t>, [k˺] - <k>, [q˺] - <q>
[m] - <m>, [n]/[ɲ]/[ŋ]/[ɴ] - <n>
[w] - <v>, [j] - <y>, [ɰ]/[ʁ̞] - <r>

So, that's the new phonology! Comments and feedback are welcome.

_________________
Languages!
Native: Proficient: Decent: Have dabbled in:

Nice
You really make a phonology, not a phoneme inventory.

Was TAVI a bird or red in your phrases?

Does your language have incorporation? You may have said it but I read badly.

Make a proto language for that and call it kantakantaranyan in Finnish.

Thanks! Phonology is the area of linguistics I'm most interested in, and know most about.

Tavi meant red, although I'll probably change it to teve. Bird is tarieh (tariat in the older version).

There isn't incorporation, no. The language is fairly analytic, although there are various inflections. It does make use of infixes though.

I might just make a proto-language actually!
(It would be good to have some neighbouring languages worked out so I can borrow some words from them.)

_________________
Languages!
Native: Proficient: Decent: Have dabbled in:

I'm currently working on various postpositional phrases.

This post will deal with phrases with the locative suffix -ti (which replaces the earlier -li).

-ti denotes static location or fixed points in time, and appears in the following postpositional phrases:
(-ti affixed to a noun on its own may be translated 'in', 'at' or 'on'.)

Akanti (from akan ‘heart’)

-Refers to location inside something. E.g. omakomi aakanti = ‘inside our house’.
-With words referring to large areas, it can mean ‘in the middle of’. E.g. mara aakanti = ‘in the middle of the forest’.

Vaheti (from vahe ‘head’)

-Referring to physical location, vaheti means ‘on’, ‘above’ or ‘on top of’. E.g. petta avaheti = ‘on top of the mountain’.
-Referring to time, vaheti indicates the beginning of a time period. E.g. qema avaheti = ‘in the early hours of night’.

Yotxati (from yotxa ‘foot’)

-Refers to location below or underneath something. E.g. valasopeh ayotxati = ‘under the carpet’.

Lenti (from len ‘side’)

-Refers to approximate location, or location nearby. E.g. mara alenti varra = ‘a town near the forest’.
-It is also used to refer to the subject of speech, songs, stories etc., like English ‘about’. E.g. tarier milemmiti kolemi = ‘a song about birds’.

Tatxati (from tatxa ‘tail’)

-Referring to physical location, tatxati means ‘behind’. E.g. Emi nitatxati sanki = ‘he is standing behind me’.
-Referring to time, tatxati refers to the end of a period of time. E.g. sahin atatxati = ‘at the end of the day’.

Umonti (from umon ‘back’ [typically of an animal rather than a person])

-Refers to the middle of a period of time. E.g. qema aumonti = ‘in the middle of the night’.

Kemiti (from kemi ‘face’)

-Refers to location in front of something. E.g. Emi mako akemiti sanki = ‘he is standing in front of the house’.

_________________
Languages!
Native: Proficient: Decent: Have dabbled in:

What's "Post-lexical" mean in this context?
----
It looks like you've got morphemes' and words' phonology down, but;
What about the morphology?
----
Did you already cover the prosody (stress and rhythm)?

_________________
I am not responsible for the accuracy of my sources; they're responsible for their own mistakes, if any, and also responsible for defending their own statements if you disagree with them.

It's 'post-lexical' to distinguish it from lexical phonology. I'm using it here to refer to the phonological rules which apply after the syllabic structure of the word is determined, as opposed to the processes that can change the syllable structure as the result of morphological processes, which I'm referring to as the lexical phonology.
----
I've covered morphology a bit here, although I need to revise those morphemes a bit. I should probably redo it.
----
Word-level prosody is in this post. The words used as examples there are out of date (I might go back and update them later), but the rules for determining feet and stress are the same. Sentence/phrase-level prosody is still to be sorted out.

_________________
Languages!
Native: Proficient: Decent: Have dabbled in:

I think that's some sort of sandhi.
Word-internal sandhi, I guess. And/or, environmentally-conditioned allophony.
"Processes that can change the syllable structure as the result of morphological processes" are probably allomorphy, or morphophonology, or inter-morphemic sandhi.
(If both of them are some kind of sandhi, they're both word-internal. If you consider them two different types of word-internal sandhi, you might want to call one "conditioned allophony" and call the other "morphophonology".)

The terminology you have is not bad. But I don't think it's standard (is it?).
If it's not standard, you need to explain it early and clearly before you use it; sensibly, just before the first use.
(If it is standard, can you show me a reference, preferably on-line, that says so?)

_________________
I am not responsible for the accuracy of my sources; they're responsible for their own mistakes, if any, and also responsible for defending their own statements if you disagree with them.

I don't think it's necessarily universally employed, but dividing phonological processes into different levels that occur at different points in the derivation of a word has been proposed by some phonologists. See for example this paper: http://www.stanford.edu/~kiparsky/Paper ... LC1982.pdf.

I'm making the distinction here because I've tried to describe the phonology based on the principles of Government Phonology, and one of the principles of Government Phonology states that the syllable structure of the word is set in the lexicon, and subsequent phonological rules cannot change it. Therefore any processes that do modify the syllable structure have to be part of lexical processes (although you're quite right in saying most of the lexical rules I stated could be referred to as morphophonology).

As you say, I probably should have made it more clear what the terms meant when I used them.

_________________
Languages!
Native: Proficient: Decent: Have dabbled in: