Half-Consonants and the Evolution of Cosyran

Over the course of the past two weeks, I’ve been revamping my conscript (writing system): Isyaxé Kiryada. While teaching the script to my friends, I realized that a fatal flaw with the script was that even after previous updates, it was unable to express certain consonant clusters, such as in the phrase cosyran ka (note that c is pronounced like ch, and sy is pronounced like sh). But why? Why didn’t I just design the script to support such a basic feature of the language to begin with?

My slave hard at work spreading the gospel.

The four stages of Rikatisyï

Rikatisyï was originally a submission to a ConJam (a competition where people compete to make languages around a certain theme), so I felt pretty obliged to make a proto-language and evolve my main submission from that since that’s what basically everyone did.

We’re going to focus on the main four. Now, let’s look at the maximum syllable structure for each of these languages:

Now, since Rikatisyï was initially made for the ConJam without anticipating the development of Modern Rikatisyï in the future, the script was made with the intention of only supporting the syllable structure of Standard Rikatisyï. Let’s look at our initial example, cosyran:

Now, I couldn’t just slap the letters for sy and r and call it a day — the writing system was an abugida, which means that just writing a consonant presumes the vowel a, instead making the word cosyaran. Of course, this wasn’t an issue for Standard Rikatisyï, where consonant clusters couldn’t occur in the first place, but how about Modern? (also yes, that’s a ligature for syï)

Introducing the half-consonant

I decided to introduce conjunct forms of consonants — you may be familiar with this if you can write Devanagari. For instance, त + व = त्व. Luckily, I didn’t need to introduce conjuncts for all consonants due to the limited types of clusters in the language.

• Final n and h already had diacritics thanks to Standard Rikatisyï.
• Consonant clusters involving r could be simplified to the cases where r came before or after the consonant.
• The remaining valid consonants that could be used in clusters were l, m, f and k.
• Newly introduced final consonants could be simply marked with a final :

An example of a half consonant:

And now, the cosyran issue resolved:

Wait, how did cosyïran even become cosyran?

It’s time for a brief introduction into the way Rikatisyï evolved. If we go back to our previous languages, the word cosyran used to look like this:

And one of the main reasons behind this change is one of my favourite phonological concepts to explain to people: palatalization. Simply put, that’s when a consonant has a y (/j/) attached to it, and that eventually causes the consonant to shift as a whole.

The quick summary for people who can read phonological changes:

H-deletion intervocalically: h > ∅ / V_V [tʰʲɔ.su.ˈhi.ran > tʰʲɔ.sui.ran]

Unrounding of u: u > ɯ [tʰʲɔ.su.ˈhi.ran > tʰʲɔ.ˈsɯi.ran]

Monopthonigzation: ɯi > jɯ > ʲy [tʰʲɔ.ˈsɯi.ran > tʰʲɔ.ˈsʲy.ran]

Palatalization: tʲ, sʲ > tɕ, ɕ [tʰʲɔ.ˈsʲy.ran > t͡ɕʰɔ.ˈɕy.ran]

Stress shifts to the penultimate syllable of a root rather than the inflected form: [t͡ɕʰɔ.ˈɕy.ran > ˈt͡ɕʰɔ.ɕy.ran]

Vowel deletion when unstressed: V [-stress] > V [-voice] > ∅ / [t͡ɕ,d͡ʑ,ɕ,ʑ,s,l]_[r,m,n] [ˈt͡ɕʰɔ.ɕy.ran > ˈt͡ɕʰɔɕ.ran]

Nasalization of vowels: Vn > V [+nasal] // _[+plosive] [ˈt͡ɕʰɔɕ.ran > ˈt͡ɕʰɔɕ.rã]

This exists in English as well! Say the phrase “don’t you” out loud casually. It often turns out as “don’tcha”, doesn’t it? The “ty” part becomes a “ch”. How about “produce”? Do you say it as prodyus or projus? The latter is the effect of palatalization. The same thing happened to our initial syllable tyo in Classical Rikatisyï. Over time, the consonant shifted from being pronounced as ty /tʰʲ/ to simply ch /t͡ɕʰ/.

How about the two-syllable suhi /su.hi/ becoming syï (/ɕy/, pronounced like shu)? This one was a multi-step process, first with the consonant h being lost. Historically, /h/-loss is extremely common across languages. For those of you who speak Spanish, note how words like hoy and hambre begin with h, yet it isn’t pronounced. At one point, those h-s were being pronounced, but h is such a weak consonant (it’s just exhaling) that it got lost over time, with its remnants existing in writing. This exists in English, too, in phrases like should have, where the h merges into the previous word, forming something like should’ve instead.

Now that the h has been lost, we’re left with sui /sui/. At this stage, the diphthong monophthongizes (two vowels merging into one), with ui becoming ï (click here to listen to how it’s pronounced). This vowels also ends up palatalizing the previous consonant, adding a y to it, forming syï. Palatalization continues further, and eventually, sy is pronounced sh (/ɕ/) — which is why /ɕ/ is spelled as sy!

The ï between sy and r is eventually dropped, as it’s an unstressed vowel which became devoiced, then dropped and turned into a consonant cluster, leaving us with the final syr.

Dealing with syr

I managed to solve the issue of the syr cluster in writing, but now I face the issue of pronunciation; I have trouble trilling the r after the sy sound. As such, I’m currently considering reintroducing the -ï verb ending whenever -ran or -rö appears (as opposed to the -e- infix with -sran and -srö). The inflection table for the aspects will then become:

	Perfective	Imperfective	Continuative	Progressive
Kawifduan	cosyet	cosye	cosyesran	cosyesrö
Present	cosyda	cosy	cosyïran	cosyïrö
Past	cosyda	cosyde	cosyden	cosydö

On the other hand, the mood inflection table for the moods for the verb cosy would end up looking like:

Mood	Verb
Active	cosy
Causative	cosyem
Hypothetical/Conditional	cosï
Passive	cosyá
Imperative	cosye

In my next conlang-related post, I’ll go over development of the four verb endings, and why the conditional depalatalizes from cosy- back to cos-. These posts also act as a good way for me to flesh out my conlang’s history, so I’ll probably keep writing these when inspiration strikes me.