Caryatic

"Sanakas" Upā-a pid âdi-ra???

Caryatic was dreamed up by Iustinus in between taking notes for Andrew Sihler's "Comparative Grammar: Indo-European Phonology," in fall of 1997. It was first committed to computer on Dec. 11 of that year. Work continued on and off until late 2003. The language remained largely dormant until 2013, when work resumed, albeit at an absurdly glacial pace.

Inspiration

The original inspiration was to "reverse engineer" the reconstruction of Indo-European from its daughter languages—which felt like an amazing new idea at the time, but which I now know as one of the most common sorts of conlang. I had actually attempted this a couple times before (ðɛ̃ʃwa ɛ̃nɛ̃nõta, "Indo-Tonal), but never with the depth of knowledge I had acquired from my graduate-level Historical Linguistics classes.

Like most of my conlangs, it draws much inspiration from the classical languages, but has broader influence from the rest of the Indo-European family. The three-vowel system was at least partially inspired by Sanskrit's propensity for the phoneme /a/.

Goals

Caryatic is a deliberate break from my previous conlanging work. Seeking to go in the opposite direction, I gave it a small phonetic inventory, few cases, and a name right from the start.

Setting

Caryatic is from a thus-far still unnamed conworld, based loosely on the ancient Mediterranean. The premise is that this world has the same language families as earth, but different daughter languages. The following languages are known to exist:

• Indo-European
  • Samasian
    • • Caryatic (Detailed)
      • Bataic (sketched)
    • • Aduric (roughly sketched)
      • Melavian (roughly sketched)
  • • Elerain (sketched)

• Afro-Asiatic
  • Semitic
    • Safuntic (named only)
  • Timuric (pretty much identical to Ancient Egyptian, since reconstructing the vowels makes it something of a conlang in itself)

In the context of this word, Caryatic roughly takes the place of Greek and Latin, Elerain that of Latin and Germanic, Safuntic that of Phoenician, and Timuric that of Egyptian.

Phonology

Orthography

Caryatic is believed to be written in an alphabetical system, which indicates vowel quantities. Two proposals for this system were submitted by David Salo, but both have been lost. There is some hope they might some day be recovered (perhaps with the proposed maps he drew).

The standard transcription works as follows:

Transcription	IPA
a	[a]
ā, â	[ɑː]
b	[p]
d	[t]
f	[f ~ v]
g	[k]
h	[x ~ ɣ]
i	[i], [j]
ī, î	[iː]
k	[kʰ]
l	[ɫ]
m	[m]
n	[n]
p	[pʰ]
r	[rʲ], [ʀʲ]
s	[s ~ z]
t	[tʰ]
u	[u], [w]
ū, û	[uː]
v, ϕ	[ϕ ~ β]
y	[j]

Consonants

	Bilabial	Labiodental	Alveolar	Palatal	Velar
Stops	p [pʰ] b [p]		t [tʰ] d [t]		k [kʰ] g [k]
Fricatives	v, ϕ [ϕ ~ β]	f [f ~ v]	s [s ~ z]		h [x ~ ɣ]
Nasal	m [m]		n [n]
Glide	u [w]			i, y [j]

Allophonic variation

• Voicing:
  1. Vowels are always voiced. Stops are always voiceless. Liquids default to voiced, and fricatives to voiceless.
  2. Liquids lose their voicing when adjacent to stops.
  3. Fricatives are voiced between voiced sounds.

• S-Assimilation:
  • Alveolars and nasals are dropped before an /s/, usually without compensatory lengthening. Note, however, that when an s is removed [see Sigmatization below] these segments normally reappear.

• Sigmatization:
  • An aspirate stop preceded by an /s/ deaspirates, and the /s/ drops with compensatory lengthening of the preceding vowel. (The transcription sometimes—albeit inconsistently—marks this by using a circumflex instead of a macron on the lengthened vowel. However, more often than not, circumflex is used ubiquitously)
  • While this process does occur across word boundaries, note that if a word begins with /s/ followed immediately by a stop, it is often lexicalized in the asigmatic form, and compensatory lengthening is unlikely to occur. Furthermore, the dropping of the s at the end of a word often allows elements which had dropped [i.e. alveolars and nasals] to reassert themselves.

• Nasal assimilation: [Note that these sound laws are, for the most part, not reflected in the standard orthography]
  1. Nasals drop before homoörganic sounds, with nasalization of the previous vowel.
  2. Nasals assimilate to the place of the following sound.

Vowels

	Front	Central	Back
High	ī, î [iː] i [i]		ū, û [uː] u [u]
Low		ā, â [ɑː] a [a]

Allophonic variation

• Hiatus
  • Two vowels in hiatus (i.e. ones that are adjacent, but do not form a diphthong) are separated by an epenthetic [ɦ]. This is not reflected in the standard orthography.

Stress

The accent is probably pitch-based, rather than stress-based, but this is uncertain. It is assigned as follows:

1. Accent falls on the long vowel nearest the beginning of the word.
2. If there are no long vowels, accent falls on the first syllable.

Example texts

Letter to Andrew Sihler

Yūdīs Masfīus Andrayā Sīlir-ad

Ti taisam Kaimānam Pirâda-an āsmi. Sas vaiks anāï “Hell” nāma-sānt ādi-ga. Au sī pirâdā a “upānā” au kavānā âdi-ga. Ti mama dimba-sa garfu-ga. Tād āh nāma “Kāriātikā” âdi-ga. Igu tua ī-sa, tāssānt ī “Indaurupayā” âdi-ga, garfu. Tū pid hāsi-ra? Tū tī dimba gankasi-ra?

The above text, read by David Salo: File:David Salo reads the Epistle to Sihler.wav