Sidaan Phonology

Here it comes!

As a direct result of not having fun, I decided to have fun with the phonology of Sidaan. This page is the result of said fun. I've also taken liberties with the romanization system. My, my, such liberties...! But, anyway, on! Read on! Be read on!


  Bilabial Labio-Dental Alveolar Palatal Velar Uvular Glottal
Stops p, b   t, d c, ɟ k, g q, ɢ  
Fricatives   f, v s, z ç, ʝ x, ɣ   h
Nasals m   n ɲ ŋ ɴ  
Liquids     l ʎ   ʀ  
Glides       j w    


  Front Back
Palatalizing Non-Palatalizing
High i ɨ u
Mid e ɛ o
Low   ɑ

Consonantal Allophonic Variation

That consonant table up there probably looks a bit crowded, but, I can assure you, there's sound reasoning and sensible planning behind it. By the end of this section, the consonants should seem boring, making the vowels even more boring. Thus, I inflict upon you: Boredom!

  1. Nasal Assimilation: A nasal must assimilate in place to the following consonant. This goes for all nasals, regardless of their phonemic status (note: [m], [n], [ɲ], [ŋ] and [ɴ] are all phonemes). So, for example, if you have a word like [xoɴ], "mouth", and you add a suffix [-s], the result is [xons], not *[xoɴs].

  2. Historical Voicing: At an earlier time in the language, all voiceless obstruents voiced intervocalically and after nasals.. This produced the voiced obstruents that occur in the language today. (So our example above would actually become [xonz].) Additionally, all fricatives voice when occurring next to any voiced non-approximant.

  3. Historical Degemination: All geminates degeminated, producing a medial contrast between voiced and voiceless obstruents. The degemination process had a couple of exceptions, listed immediately below.

  4. Gliding: Geminate liquids became glides: /ll/ > [j] and /ʀʀ/ > [w].

  5. Palatalization: The high front vowels /i/ and /e/ cause alveolar and velar consonants to palatalize (as well as /h/). This results in the following: /t, k/ > [c]; /d, g/ > [ɟ]; /n, ŋ/ > [ɲ]; /s, x, h/ > [ç]; /z, ɣ/ > [ʝ]; and /l/ > [ʎ]. (Note: This, along with one other change, produced all the palatal consonants of Sidaan.)

  6. H-Deletion: The phoneme /h/ was lost word-finally. In coda positions, the /h/ becomes a copy of the next consonant, producing geminates.

  7. Palatal Assimilation: A coda /s/ will assimilate totally to an onset [ç]. It will then degeminate, via the historical degemination rule.

  8. Epenthesis: When the phonotactics require an epenthetic consonant, that epenthetic consonant is /n/.

There's some other consonantal stuff, but that consonantal stuff is more vowely in nature, so it'll be explained in the next section.

Vocalic Allophonic Variation

The vowels do some stuff. That stuff is described below.

  1. Vowel Lowering: An underlying /i/ or /e/ vowel will lower to [ɨ] and [ɛ], respectively, when preceded by a uvular consonant.

  2. Vowel Lengthening: The vowels of monomoraic substantives (nouns, verbs, adjectives and some adverbs) are lengthened when their final representation is monomoraic (so, in other words, if you add a suffix to such a word, the vowel will remain short).

  3. Vowel Loss: Unstressed [+high, -back] vowels (i.e., vowels without primary or secondary stress) are lost in the following circumstances:
    1. In between a fricative and a stop (oral or nasal).
    2. In between a fricative and a liquid.
    3. In between an oral stop and a liquid.

  4. Palatal Harmony: An underlying front vowel (i.e., /e, i, ɛ, or ɨ/) of a given affix will match in its [palatal] feature with the nearest vowel. If that vowel is not a [-back] vowel, the underlying form of the vowel will emerge.

  5. Epenthesis: When the phonotactics require an epenthetic vowel, that epenthetic vowel is /i/.

You can now move on to see how this mess gets romanized.


I really had to reach deep into my Unicode bag of tricks to romanize this beast. The result is below:

  • The following sounds are romanized just as their transcriptions suggest: p, b, t, d, k, g, q, f, v, s, z, x, h, m, n, l, w, u, o and a. (Note: this means that historical geminates will not be preserved in the romanization—just synchronic geminates.)

  • The phoneme /ʀ/ will be spelled r, while the other uvulars will be spelled with a raised dot: /ɢ/ > ġ (capitalized as Ġ) and /ɴ/ > (capitalized as ).

  • Though I'm not at all happy about it, the phonology has forced my hand: /ŋ/ will be spelled just like it looks: ŋ (capitalized, regrettably, as Ŋ). The voiced velar fricative /ɣ/ will be spelled ğ (capitalized as Ğ).

  • The palatal consonants are a bit of a hodgepodge. For my own peace of mind, the main palatal consonants will be spelled just as they are with other languages on this site (e.g., Sheli): /c/ > č; /ɟ/ > j; /ç/ > š; /ʝ/ > ž; /ɲ/ > ñ; and /j/ > y. I've given a lot of thought to the romanization of /ʎ/, and have decided that I hate every option. Literally, every single one; none of them are desirable. Thus, I'm going to use the only letter I've ever used for an alternate l on any of these pages: ł (capitalized as Ł).

  • Following Hungarian's lead, I decided the only reasonable thing to do with the [-back] vowels is this: /ɛ/ > e; /ɨ/ > i; /e/ > é; and /i/ > í. This way the vowels that have [-palatal] associated with them will be a, e, i, o and u.

  • The letter h will be spelled out word-finally, but it will not be spelled out when it's a medial coda. In that case, the sequence will be spelled as a geminate consonant.

That's it for the funky romanization. I kind of like it: dots, carons and tildes all over the place. Now we can move on to graver matters.


Just a few things to mention:

  • The max syllable of Sidaan is CVCC, modulo the rules above (meaning that you could actually have a syllable that's CCVCC, such as slind).

  • Any consonant may begin a word, and any non-palatal consonant may end a word (modulo the rules above).

  • Most clusters are permitted medially, but two oral or nasal stops cannot occur next to each other (unless they're a synchronic geminate).

  • When three consonants come in a row in the middle of a word, a stop (oral or nasal) cannot follow a stop-fricative cluster (so no *mvt, *psq, *tsk, etc.).

  • Word-finally, the following clusters are disallowed (F = fricative; N = nasal; S = stop; L = liquid): *sf, *sx, *rx, *mv, *ŋğ, *px, *tf, *tx, *kf, *qf, *qx, *FL, *FN, *LL, *LN, *NL, *NN, *SL, *SN, *SS.

  • Geminates are never allowed word-finally. When adding a suffix would produce a geminate, an epenthetic vowel is inserted.

  • H is never permitted in an initial or final cluster, and may only appear as a single coda or an onset.

  • If affixation causes stress to shift, dropped vowels may be reintroduced modulo rule 3 of the vocalic allophonic variation section.

I can't promise that this section is finished.


Sidaan has a stress system that is predictable based on the shape of a word. The penultimate syllable always gets first dibs, when it comes to stress. The only thing that can shift main stress away from the penultimate syllable is if the last syllable is heavy and its coda contains a consonant that's velar or backwards (this includes k, g, ŋ, x, ğ, q, ġ, , r and h [even if there is no h sound there]). This means that codas with two or more members that do not contain a consonant that's further back than the hard palate will not cause stress to shift (so a word like toğenz will be stressed on the first syllable). Secondary stress is placed on every other syllable from the main stress (counting either forwards or backwards).


That's it for the phonology section. As this is the first section of Sidaan that I've completed, I intend to have a tiny little celebration. This will involve my kitchen and food obtained therefrom. Let us all rejoice!

Back to Sidaan Main

This page was last modified on Tuesday, March 3, 2009.
This website was last modified on .
This page can be viewed normally, as a milk or dark chocolate bar, in sleek black and white, or in many other ways!
All languages, fonts, pictures, and other materials copyright © 2003- David J. Peterson.

free counters