Antarctican

Antarctican is spoken in the far future, after catastrophic global warming has rendered most of the world uninhabitable, forcing survivors to flee to the land exposed by the melting ice sheets in Antarctica. It's vocabulary is drawn from a wide variety of existing languages, across many language families.

Basic Phonology

Vowels

The vowel system is complex, with vowels distinguished not only by quality but also by phonation (breathy, modal and tense). Vowels with breathy phonation (marked with a diaresis e.g. ṳ), are also pronounced with a lower than normal pitch. Vowels with a tense phonation (marked with a tilde e.g. æ̰), are pronounced with a higher than normal pitch.

Monophthongs

There are 11 of these. 5 have modal voice, 5 have breathy voice, and one has tense voice.

	Front	Mid	Back
High	i	ɨ̤	ṳ u
			o̤
Mid		ə̤
	ɛ		ʌ̤ ɔ
Low	a̰ a

Diphthongs

Diphthongs are divided into three categories, "general", "opening" and "closing". Opening diphthongs are only ever found in open syllables (syllables that do not end with a consonant), closing diphthongs are only ever found in closed syllables (syllables that end in a consonant), and general diphthongs can be found in both.

Opening Diphthongs

There are five of these: /aɛ/, /aɔ/, /aə̰/, /ʌə̤/, /ʌo̤/. The first two of these (the modally voiced ones) are more common than the last three, especially /ʌə̤/ and /ʌo̤/ (the breathy voiced ones).

Closing Diphthongs

There are only two of these and they both have modal voice: /ɛi/ and /ɔu/

General Diphthongs

Most of the diphthongs in Antarctican fall into this category. There are four with modal voice: /ai/, /au/, /ɔi/ and /ɛu/. Another eleven have tense voice: /oṵ/, /eḭ/, /ɛə̰/, /ɔə̰/, /aḭ/, /aṵ/, /ɔḭ/, /ɛṵ/, /ɛḭ/, and /ɔṵ/. And six more have breathy voice: /ʌɨ̤/, /aṳ/, /oɨ̤/, /əṳ/, /əi̤/, /oṳ/.

Consonants

At most, there are 28 consonants that could be considered phonemic:

	Labial	Alveolar	Palatal	Velar	Glottal / Placeless
Nasal	m̥ m	n̥ n	ɲ̥ ɲ	ŋ̥ ŋ	ɴ
Stop	pʰ p	tʰ t	cʰ c	kʰ k	ʔ
Implosive	ɓ	ɗ
Fricative	f	sʰ s	ʃ	x	h
Approximant	w		j
Flap		ɾ

The implosives can only occur between vowels, and only if the following vowel has tense voice.

11 of these consonants are classed as "aspirated". These include not only the aspirated stops /pʰ/, /tʰ/, /cʰ/ and /kʰ/ and the aspirated fricative /sʰ/, but also the voiceless nasals /m̥/, /n̥/, /ɲ̥/, /ŋ̥/, as well as the fricatives /f/ and /h/. These consonants can only occur before a vowel with modal voice.

/ɴ/ only occurs in syllable codas. It assimilates to the place of articulation of the following consonant. Before /ʔ/, /h/ or at the end of phrases, it simply nasalises the preceding vowel.

Phonotactics

There is only an extremely small number of syllable shapes permitted, limited to syllables of the form C (j /w) V (ɴ). Furthermore, /j/ can only occur after labial consonants. When it occurs after /w/, the consonant cluster /wj/ is pronounced as a labiodental approximant [ʋ], or sometimes a flap [ⱱ]. The consonant cluster /jw/ is pronounced [ɥ].

Chroneme

The chroneme is vary important in Antarctican phonology (see here for an explanation: http://en.wikipedia.org/wiki/Chroneme). It can either attach to a vowel, or to the first consonant in a syllable onset. In both cases the sound is lengthened. Whether or not a word has a chroneme can distinguish words. So while /paxa/ is a person's name, /paxːa/ means a fool or an abnormal person (kind of like Mr. Bean) and /paːxa/ means a tout or a pimp.

Note that the chroneme cannot occur on the first sound of a word, nor the last sound (unless it is a vowel / diphthong with breathy voice or tense voice). So neither /pʰːaxa/ nor /pʰaxaː/ would be possible words. It can also only ever occur a maximum of one time per word. So none of /paːxːa/, /paxːaː/ or /paːxaː/ could be possible words. If the chroneme occurs on a diphthong, the first element of the diphthong is lengthened e.g. /ta̰saːḭ/ - steak, not /ta̰saḭː/.
Also note that "over-long" clusters of /ɴ/ followed by a geminate nasal are not found.

When /ɾ/ is geminated, it is pronounced as a trill [r].

Morphology

Antarctican is an isolating language, and as such there is little in the way of inflectional morphology. However, there are some complex sandhi processes that affect words strung together in a phrase.

Phonation Spreading

When words are put together to make a phonological phrase, it is possible for phonation to spread backwards, from the initial syllable of one word, back to the previous word in that phrase. This never happens with modal voice, only with tense and breathy phonation, and only with a subset of words with these phonations on the first syllable e.g.

/ʃɛḭ/ - to chat

/mɛ/ - 1st person singular accusative particle

[mɛə̰ ʃɛḭ] - to chat with me/mʌ̤/ - to ambush and rob

[mə̤ mʌ̤] - to rob me

However there are many words that do not spread their phonation onto previous words e.g.

/ʃɛḭ/ - to shine

[mɛ ʃɛḭ] - to shine at me

/mʌ̤/ - to apply

[mɛ mʌ̤] - to apply to me

Here we can see that the words for "rob" and "apply" are homophones when spoken by themselves, only being distinguished by whether or not they spread their breathy phonation onto the previous word.
The reason why some words spread their phonation onto previous words and others do not goes back to the proto-language. This had a pitch-register system like Burmese (see here for an explanation http://en.wikipedia.org/wiki/Register_(phonology)). There were five registers: Mid Modal, Mid Tense, Mid Breathy, High Tense and Low Breathy. There was also a tone spreading system where the High Tense and Low Breathy registers would spread onto previous words. However their counterparts with mid pitch would not.
So the words for "chat" and "shine" were originally not homophones even in isolation. The former had High Tense register and the latter had Mid Tense. Likewise, the words for "rob" and "apply" were not homophones even in isolation, with the former having Low Breathy register, and the latter having Mid Breathy register.
However, the two tense voice registers (High and Mid) merged into one, as did the two breathy voice registers (Low and Mid). Except that words that originally had Mid pitch did not spread their phonation, whereas the ones with High (Tense voice) and Low (Breathy voice) did.
From now on, words that spread their phonation will be marked by a minus sign before the word e.g. /-ʃɛḭ/ - to chat vs. /ʃɛḭ/ - to shine, /-mʌ̤/ - to ambush and rob vs. /mʌ̤/ - to apply. Often, when a word has phonation spread onto it, it can change in other ways too e.g./tʰɛ/ - 2nd person singular accusative particle

[tɛə̰ ʃɛḭ] - to chat with you

[tə̤ mʌ̤] - to rob you

Here the change is simply a loss of aspiration in the first consonant (remember that aspirated consonants can only occur before vowels with modal voice). However the change can be more radical, often involving a suffix added onto the word (usually if the word already has breathy or tense phonation) e.g.

/-nɔə̰/ - 1st person plural inclusive accusative particle

[nɔsːoṵ ʃɛḭ] - to chat with us (including you)

[nɔsːṳ mʌ̤] - to rob us (including you)
This process used to be regular, however it is no longer so as a result of phonological change. In the proto-language, the particle discussed above was *[ˈnosu]. However there was a strong tendency to delete short, modally voiced vowels in open syllables, especially if they were at the end of words and/or immediately after a stressed syllable. So now the particle was *[nos]. But since this sound change did not affect vowels with tense or breathy voice, the final syllable was still there when the word was being affected by phonation change. Finally there was a sound change that deleted all consonants that were not before a vowel or semivowel (or reduced nasals to /ɴ/. But when voiceless obstruents were deleted, they put tense voice on the previous vowel. Hence we end up with /-nɔə̰/, /nɔsːoṵ/, /nɔsːṳ/.
These sound changes mean that the variant forms that arise when tense and breathy voice spread onto a word must be memorised along with the basic form when learning the word. Therefore, from now on when vocabulary is given, the normal form will be listed first, followed by the form when tense voice spreads onto it, and finally the form when breathy voice spreads onto it e.g.

/-nɔə̰/, /nɔsːoṵ/, /nɔsːṳ/ - 1st person plural inclusive accusative particle

/tʰɛ/, /-tɛə̰/, /tə̤/ - 2nd person singular accusative particle

/mɛ/, /-mɛə̰/, /-mə̤/ - 2nd person singular accusative particle/-ʃɛḭ/, /-ʃɛḭsːɔə̰/, /ʃɛitːo̤/ - to chat

/ʃɛḭ/, /ʃɛḭsːɔə̰/, /cʰɛitːo̤/ - to shine

/-mʌ̤/, /-ma̰ŋːoṵ/, /-mʌ̤ŋːṳ/ - to ambush and rob

/mʌ̤/, /ma̰ŋːoṵ/, /mʌ̤ŋːṳ/ - to apply

Initial Consonant Mutation

There is one other mutation process that happens when individual words are joined together into a phrase. If some initial consonants clusters end up between vowels, then they change to other consonants. Exactly what they change into depends on the phonation of the following vowel, as well as whether or not it spreads to the previous word. The changes are listed in the table below:

Original Syllable Onset	Modal Voice	Tense Voice (spreading)	Tense Voice (non-spreading)	Breathy Voice (spreading)	Breathy Voice (non-spreading)
ʔ	k	Unchanged	Unchanged	ŋ	Unchanged
k	ŋ	x	ŋ	Unchanged	Unchanged
x	Unchanged	ɗ	Unchanged	Unchanged	Unchanged
t	Unchanged	s	s	Unchanged	Unchanged
s	Unchanged	ɾ	ɾ	Unchanged	Unchanged
n	Unchanged	Unchanged	ɗ	Unchanged	Unchanged
p	Unchanged	ɓ	ɓ	Unchanged	Unchanged

These consonants mutate in exactly the same way whether or not they are clustered with /j/ or /w/ e.g. between vowels and before a vowel with tense voice, /pw/ will become /ɓw/, and likewise /pj/ will become /ɓj/. Note that before vowels with breathy voice, the consonants never mutate.Below are some examples:
/ʔuɾːa/, /-ʔoṵɾːa̰/, /ʔṳɾːʌ̤/ - water

/-ʔṳ/, /kuɗːɔə̰/, /-ʔṳtːo̤/ - good

[ʔṳɾːʌ̤ ŋṳ] - good water

/nweḭɴ/, /nweḭɴkːoṵ/, /twiɴkṳ/ - a drink

[ʔoṵɾːa̰ ɗweḭɴ] - drinking water