Hi all!

As requested, a little thread on the basics of Erháin. As a matter of fact it began as a support language for a novel I'm writing, and at the beginning it was exactly that: nothing but some random notes on word forms and a very rudimentary grammar. But ever since, the time I spent on the language has largely overgrown the effort I'm putting in the novel itself. The language (and the script) are my main interest(s) now.

Erháin has some interesting features. The most important ones are:

• vowel mutation for both grammatical case and number
• infixes in verb conjugations (except for imperatives, gerunds, gerundives and participles)
• postpositions instead of prepositions

Let's go through them here:

1. vowel mutations
1.1 cases
Nominative vowels in Erháin can only be a /a/ , i /i/ , o /ɔ/ , u /u/. The vowel of the stressed syllable (always the penultimate syllable unless otherwise indicated by means of an acute accent) is called the "stemvowel". The three cases - nominative, accusative and genitive - are indicated by mutating this stemvowel.

system:

• Nominative: no mutation
• Accusative: stemvowel is umlauted (a > e /ɛ/ , i > î /ɪ/ , o > ø /œ/ , u > y /y/)
• Genitive: stemvowel is followed by a i /ɪ/

example:
NOM: caron --- /ˈkarɔn/ (sphere, globe)
ACC: ceron --- /ˈkɛron/
GEN: cairon --- /ˈkaɪrɔn/


NOM: silan --- /ˈsilan/ (star)
ACC: sîlan --- /ˈsɪlan/
GEN: siilan --- /ˈsiʔilan/

NOM: oric --- /ˈɔrik/ (nut, nutshell)
ACC: øric --- /ˈœrik/
GEN: oiric --- /ˈɔɪrik/

NOM: durun --- /ˈdurun/ (drum)
ACC: dyrun --- /ˈdyrun/
GEN: duirun --- /ˈduɪrun/


1.2 number

The number of a noun or adjective is again indicated by mutating the stemvowel. In this case many of the possible preceding consonants will mutate under the influence of this mutation as well.

system:

• stemvowels are iotated (a > iá /ja/ , e > ié /jɛ/ , o > io /jɔ/ , ø > iǿ /jœ/ , etc)
• preceding c , s and g are palatalized (/k/ > /tʃ/ , /s/ > /ʃ/ , /g/ > /dʒ/)
• preceding t, d, p, b are fricantized (t /t/ > th /θ/ , d /d/ > dh /ð/ , p /p/ > ph /f/ , b /b/ > bh /v/)
• preceding sonorants m, n, w, j, l, r, h are not touched

examples:

SG NOM: caron --- /ˈkarɔn/ (sphere, globe)
PL NOM: ciáron --- /ˈtʃarɔn/ (spheres, globes)

SG NOM: durun --- /ˈdurun/ (drum)
PL NOM: dhiúrun --- /ˈðjurun/ (drums)

SG NOM: moran --- /ˈmɔran/ (sail)
PL NOM: mióran --- /ˈmjɔran/ (sail)

This process is unchanged in ACC and GEN. You can imagine however that it would - at least in spelling - collide with stemvowel -i-. the plural of, say, silan (star, SG NOM) would become *siilan, rendering it undistinguishable of siilan (star, SG GEN). For that reason the plural of -i- becomes -ei-. Pronunciation however remains the same:

SG NOM: silan --- /ˈsilan/
SG ACC: sîlan --- /ˈsɪlan/
SG GEN: siilan --- /ˈsiʔilan/
PL NOM: seilan --- /ˈʃilan/
PL ACC: seîlan --- /ˈʃɪlɛn/
PL GEN: seiilan --- /ˈʃiʔilan/

topic 2 will follow in the next post >>>>>

What kind of umlauting gets you [ɪ] from [i]? (The only possibility that comes to my mind is a-umlaut, but all the other changes look to be the result of i-umlaut.)

2. Infixes in verb conjugations

Infinitives are easy to recognize: they always end in -iar. Verbs are conjugated by leaving out that suffix and putting an infix in the middle of the verb stem. Which infix depends on the person and tense of the verb.

system:
The verb I used to illustrate this is saniar /'sanjar/ (say). The stem of the vowel is san. The infixes will be added before the last vowel of the verb stam (in this case there's only one). The stress always remains on the syllable that is stressed in the infinitive. Since this means the stress is on another syllable than the penultimate one (which is occupied by the infix), we always put an acute accent on it.

Okay, here goes:


examples:
He drank a cup of tea = óvudan cøron raim (óvaniar = to drink)
I like your hair = menár meínanes (mariar = to love, to like)
we would help you if we could = øs purmác win tumín (paciar = to help, tiniar = can)
you (pl) are wasting your time = ásserfil éradef (ássiliar = to waste)
the eagle spread his wings = arir dísudiv seíraned (disiviar = spread)

As mentioned before though, imperative, gerund, gerundive and participles are created with suffixes:

• imperative: stem + o
• gerund: stem + os
• gerundive: stem + al
• past participle: stem + il
• present participle: stem + ic

3. Postpositions instead of prepositions

That's the least "impressive" one, I guess. The title says it all. Except for one thing: possessive pronouns are suffixes as well. In case both appear together, the possessive suffix comes first.

system:

• 1 sg: -én
• 2 sg: -és
• 3 sg: -éd
• 1 pl: -éng / -ém
• 2 pl: -éf
• 3 pl: -éb

examples:
in the city (in = de, city = tir): tirde
under the bridge (under = hin, bridge = crivac): crívachin
with their friends (with = nun, friend = marlin): miárlinebnun
I was kissed by a beautiful girl (kiss = sasiar, by = nor, beautiful = san, girl = ganid): nun sasil ganid-sannor

There, that's a very short first overview of the language. If you have questions, do not hesitate to ask!

(I converted your square braquets to slashes, because they messed up my text styles. Just mentioning it because I don't want you to think I 'm trying to misquote you or anything )

Actually, you're absolutely right! I admit it is a long shot. I tried looking for another option, but I found nothing usable. For a while I decided just leaving NOM -i- > ACC -i-, but that would be too difficult since word order is very free. For example, there would be no way to determine whether nigar (wizard, wise man) was supposed to be object or subject in the sentence edám nigar barril. This either could mean "he sees the wizard good" (he can see him without difficulty) or "the wizard sees good" (the wizard can see without difficulty).

Other options are just as impossible. E.g. I cannot prolong the -i- to -ì- or something (/i:/, however you write it), since that would collide with -ii-. I cannot change it to its neighbour /y/, since that one is already occupied for the ACC of /u/.

With /a/, /ɛ/, /i/, /ɔ/, /œ/, /u/, /y/ already in use, there's nothing much left to use for the accusitive of -i-. Therefor, and for no other reason, I chose to a-umlaut /i/ while all other sounds were obviously i-umlauted.

BUT! If you have a suggestion, I'll gladly listen!

My suggestion is this: Learn to live with a little ambiguity. It looks like you're striving to build an artlang, not a loglang, and typically artlangs are naturalistic. Ambiguity is a natural feature of any human languages. (Just look at what I wrote above: Is live pronounced /liv/ or /lajv/? OH NOES HOW CAN ANYONE TELL???)

Moreover, sentences which are ambiguous in isolation are rarely so in context. How do we know which interpretation of edám nigar barril is correct? Well, what were the preceding sentences about, the wizard and his capabilities or another person and his perceptions? (And is it really impossible to put in a disambiguating personal pronoun? Other pro-drop languages such as Spanish do this routinely, e.g. Dice que yo estaba muy flaca "She says that I'm very thin" [where estaba could otherwise be interpreted as a third-person form].)

If you're dead set on avoiding such rare cases, a more naturalistic possibility than what you have is for the i-umlaut to affect all vowels in the word, not just the stem vowel. Thus you'd have nom. nigar, acc. niger. The downside is that this could play havoc with your derivational morphology, but it seems a small price to pay. (Moreover, it brings Erháin more in line with the Celtic languages which served as inspiration. The plural of Welsh carreg "stone", for instance, is not *cerreg but cerrig. It's not completely unheard of for i-affection to umlaut only the stressed stem vowel, but it is a less-common development.)

Not to mention that in Spanish the 3rd pl conjugations aren't marked for gender.

It look like you change back vowels into front vowels. Like /a/ to /e/, /o/ to /œ/, /u/ to /y/ and my suggestion is to change /i/ to /ɯ/. Because ɯ is the back version of i.

I think I like the "Learn to live with a little ambiguity". In fact I had to rack my brains to find an example of a sentence with two possible interpretations. Which proves that NOM -i- > ACC -i- is indeed sufficient.

NOM: silan --- /ˈsilan/ (star)
ACC: silan --- /ˈsilan/
GEN: siilan --- /ˈsiʔilan/

it is then. I'm actually relieved Not only is it more logical, it even sounds better!

menár émos seilan /mɛ'nar 'ɛmɔs 'ʃilan/ (I love to watch the stars)... How lovely is that!

Linguoboy, I thank you for convincing me of the obvious.

one small adaptation:

Up til now, I wrote the plurals of a word by preceding the stemvowel (the stressed vowel) by a i. This, in some cases, led to quite some confusion (e.g. the pronunciation of -tia-: should it be /tia/ or /θa/?). In romanized spelling, plurals are now written with an h between the stemvowel and the consonant preceding it. In many cases this means it's closer to English spelling as well - though that wasn't the main purpose.

Examples:

SG NOM: cann --- /kan:/ (lake)
PL NOM: chann --- /tʃan:/ (lakes)

SG NOM: tomar --- /'tɔmar/ (coward)
PL NOM: thomar --- /θɔmar/ (cowards)

SG NOM: nirod --- /'nirɔd/ (table)
PL NOM: nhirod --- /'njirɔd/ (tables)

SG NOM: baroc --- /'barɔk/ (harbour)
PL NOM: bharoc --- /'varɔk/ (harbour)

Sentence:
Daw ghanid unám dílacil lhárcinde, oan esde win ganid udám en shádordur.
/daw 'dʒanid un'am 'dilakil 'ljarkindɛ, ɔwan 'ɛzde win 'ganid ud'am en 'ʃadɔrdur/
I saw two girls playing in the woods, but only one girl saw me between the trees.

Do you think this is an improvement?

What happened to this thread? It just disappeared. Is Caenwyr no longer working on the language, or is there just nothing more to say (as is probably the case)?