A Phonology and Two Questions

[conlangernoob]

Hi guys! I haven't posted on this board for a while (My interests were diverted mainly towards natlangs, mainly Old English), but I have just started work on a yet unnamed Proto-Lang, and I’m wondering what you guys think about the phonology.

CONSONANTS


Nasals: m, n
Stops: p /pʰ/, b /p/, t /tʰ/, d /t/, c /kʰ/, g /k/
Fricatives: f, v, th /θ/, dh /ð/, s, ch /x/, h
Trill: r
Approximate: l, y /j/, w

1. /s/ is realized as /z/ intervocalically.
2. The cluster /hy/ is realized as [ç].
3. The cluster /hw/ is realized as [ʍ].
4. The cluster /ny/ is realized as /ɲ/.
5. The cluster /kw/ is romanized as /qu/.

VOWELS: Basic average five vowel system + length marked with an acute accent + schwa romanized as y.

PHONOTACTICS AND STRESS
1. Stress is on the penultimate syllable, unless there is no coda consonant on that syllable, in which case the stress is on the antepenultimate syllable.
2. The syllable structure is CCVC. The only permissible clusters are /kw/, /gw/, /ty/, /dy/, /ly/, /ny/, /hy/, /hw/.
3. Any vowel may end a word, but the only consonants that may normally end a word are /n r s/.

By the way, rule three of Phonotactics and Stress is attested in Ancient Greek.

Until Anon,

conlangernoob

P.S. Two questions:
a) In what situations can multilingualism exist stably for hundreds of years in one place?
b) Does morphosyntactic alignment necessarily matter in SOV conlangs, since, when there is only one noun marked, it could be considered either the subject or object?

[bradrn]

Seems fine to me. Only comment I could make is that ⟨c⟩ can be a bit confusing in a romanisation.

a) In what situations can multilingualism exist stably for hundreds of years in one place?

Almost any. Throughout most of history, extensive multilingualism has been the default. It still is in most places.

b) Does morphosyntactic alignment necessarily matter in SOV conlangs, since, when there is only one noun marked, it could be considered either the subject or object?

Yes, it matters a lot. Even if there’s just one argument, you can still consider it as aligning with either S or O, on the basis of noun case or verbal agreement — and if you have neither of those, you can still consider its syntactic behaviour in general as being more similar to that of S or O.

[So Haleza Grise]

My only comment is that at first blush the phonology strikes me as a bit European-y. Of course that's fine, maybe that's what you're going for. And of course a proto-language can have descendants that evolve in any direction (PIE is quite phonologically unlike its descendants). But in terrestrial terms, things like separate sets of voiced and unvoiced fricatives are fairly rare. Again, that's completely fine if it's what you want, and there isn't really such a thing as a "standard/average" phonology, plenty of languages have rare or unusual phonological features.

One argument that I have sometimes followed myself is that it can be quite useful to have a phonology that you are "at home" pronouncing, because if you can't be trusted to pronounce it, who can? Sometimes it's good to start somewhere modest/familiar, especially since it helps to know how phonemes behave typically if you are going to apply sound changes to them.

How will you distinguish <y> as a palatal feature from the schwa <y>?

[bradrn]

But in terrestrial terms, things like separate sets of voiced and unvoiced fricatives are fairly rare.

Not so rare, nor particularly European: in WALS’s sample, a third of languages have fricative voicing, and they don’t seem to have a particularly Eurocentric distribution to me, though it does look like Europe has a cluster (alongside West Africa and SE Asia).

[conlangernoob]

What do you guys think about the early drafts of my morphology and syntax sections:

Morphology


Nouns: Nouns are inflected for singular, dual and plural in animate nouns, and just singular and plural in non-human animate nouns, and optional singular and plural marking in inanimate nouns. The singular is unmarked. The dual is marked with the suffix $$$. The plural is marked with the suffix $$$.

Verbs: Verbs are inflected for either the perfective aspect, the imperfective aspect, and the irrealis mood. The perfective is marked as null. The imperfective is marked with the suffix $$$ and the irrealis is marked with the suffix $$$. Verbs can have any of zero to three arguments.

Adjectives: Agrees with verbs.

Syntax


Basic Sentences: The unmarked sentence order of LANGUAGE is subject-object-verb. More precisely:
Subject | Direct object | Indirect object | Verb phrase | Postpositional phrases
However, the verb phrase can be moved to either before the direct object, or before the subject, although the latter is only common in poetry.

Noun Phrases: Nouns precede numerals, adjectives, prepositional phrases and relative clauses, in that order, while demonstratives follow nouns. Genitives are marked with postpositional phrases.


Until anon,

conlangernoob

[conlangernoob]

Question: What is the best way to format a dictionary entry. What I am currently doing is something like this: edhas (eðas) n. 1. black 2. Evil | Middle Ambaran ethas > Archaic Ambaran etas (edhas is not actually in my language, I’m still working on the Proto-Lang).

[bradrn]

Nouns: Nouns are inflected for singular, dual and plural in animate nouns, and just singular and plural in non-human animate nouns, and optional singular and plural marking in inanimate nouns. The singular is unmarked.

Sounds reasonable, though I’m struggling a little to understand the first part — if the dual is used for animate nouns, but not non-human animate nouns, then wouldn’t it be simpler to just say that ‘human nouns can take the singular, dual and plural’?

Verbs: Verbs are inflected for either the perfective aspect, the imperfective aspect, and the irrealis mood. The perfective is marked as null. The imperfective is marked with the suffix $$$ and the irrealis is marked with the suffix $$$. Verbs can have any of zero to three arguments.

Again, sounds reasonable, but the devil is in the details. I’m maybe a bit more concerned with this than other people are, but I always feel that figuring out the semantics of verbal aspect is really important — when you look at it in detail, terms like ‘perfective’ and ‘imperfective’ often mean quite different things when applied to different languages. You might want to think about exactly what kinds of events are categorised as ‘perfective’ or ‘imperfective’ in your language.

Basic Sentences: The unmarked sentence order of LANGUAGE is subject-object-verb. More precisely:
Subject | Direct object | Indirect object | Verb phrase | Postpositional phrases
However, the verb phrase can be moved to either before the direct object, or before the subject, although the latter is only common in poetry.

Something to think about: if the verb can be moved around so much, what grounds are there to consider SOV the basic word order? Why not just

Noun Phrases: Nouns precede numerals, adjectives, prepositional phrases and relative clauses, in that order, while demonstratives follow nouns. Genitives are marked with postpositional phrases.

Sounds reasonable.

Question: What is the best way to format a dictionary entry.

There’s no one answer! Dictionaries are remarkably complicated things. Personally, I like using software for dictionary management: my preference is SIL Toolbox, but there’s also Lexique Pro (which is simpler) and FieldWorks (which is more complex). All three rely on the MDF text format: Coward & Grimes (2000) wrote an overview, which incidentally contains a lot of really useful tips on how you can structure a dictionary and its entries.

[conlangernoob]

Does this system for my verbs work? Did I get everything? Is it naturalistic?

Subjunctive: All deontic and dynamic modality, future events, interrogatives, polite imperatives, jussives, conditionals, presuppositions, wishes, and fears.

Imperfective: All present events, habitual past events, past events expressed without reference to their completion, and imperatives.

Perfective: Past events expressing completed action, and past-in-the-future events.

[alice]

Does this system for my verbs work? Did I get everything? Is it naturalistic?

It looks suspiciously like several of mine

[bradrn]

Does this system for my verbs work? Did I get everything? Is it naturalistic?

Subjunctive: All deontic and dynamic modality, future events, interrogatives, polite imperatives, jussives, conditionals, presuppositions, wishes, and fears.

Imperfective: All present events, habitual past events, past events expressed without reference to their completion, and imperatives.

Perfective: Past events expressing completed action, and past-in-the-future events.

I’ll repeat what I said earlier: listing the names of grammatical categories is fine, and it sounds reasonable from what you’ve given us, but what’s more interesting is to consider the precise semantics of all these categories and how they relate to each other. (‘Subjunctive’, for instance, has gotten used for enough different things that it’s very hard to know if the term is still meaningful or not.)

[bradrn]

Actually… hold on, I think I misunderstood your post. Disregard my last comment please!

[conlangernoob]

Since my conlang, while having a CCVC syllable structure, only allows certain medial consonantal groups, I devised a Sandhi system to deal with word-internal morpheme contact, but I’ve done anything like this before, so I’m wondering what you guys think.

First here’s the phonotactic rules:

The syllable structure is CCVC.
The only permissible onset clusters are /kw/, /gw/, /ty/, /dy/, and /ny/.
The only permissible syllable codas are /m/, /n/, /s/, /r/, or /l/.
The only permissible medial consonantal groups are …
… permissible onset clusters.
… /l/ or /r/ + any phoneme and permissible onset clusters.
… /m/ + any labial plosives and onset clusters beginning with labial plosives.
… /n/ + any non-labial plosives and onset clusters beginning with non-labial plosives.
… /m/ + /n/ or /n/ + /m/.
… /mm/, /nn/, /ss/, /rr/, and /ll/. This only occurs at morpheme boundaries.

And here’s my Sandhi system:

When two word-internal morphemes meet and an illegal medial consonantal group is formed, seven sound changes are applied, in order, until the consonantal group becomes legal.
1. Nasals, before plosives, assimilate to that plosive’s place of articulation.
2. Fricatives, after nasals, undergo fortition to become stops of the same place of articulation and voicing. /s/ and /θ/ both become /t/.
3. Nasals, before fricatives, assimilate to that fricative’s place of articulation.
4. Semivowels, after nasals or /s/, turn into their corresponding short vocalic form.
5. Nasals or /s/, before /r/ or /l/, undergo metathesis and switch places with the next consonant.
6. Any phonemes, after /s/, become deleted and /s/ become geminated.
7. /s/ and /ny/, after nasals, become deleted, and the nasals become geminated.

Until anon,

conlangernoob

[bradrn]

Sounds fine to me! I think some examples might help for understanding the sandhi effects.

4. Semivowels, after nasals or /s/, turn into their corresponding short vocalic form.

What do you mean by ‘short vocalic form’ here? Generally, a semivowel is considered to be just another name for a very short non-syllabic vowel.

[conlangernoob]

Sounds fine to me! I think some examples might help for understanding the sandhi effects.

4. Semivowels, after nasals or /s/, turn into their corresponding short vocalic form.

What do you mean by ‘short vocalic form’ here? Generally, a semivowel is considered to be just another name for a very short non-syllabic vowel.

I have short vowels and long vowels. Should I rephrase it?

[bradrn]

Sounds fine to me! I think some examples might help for understanding the sandhi effects.

4. Semivowels, after nasals or /s/, turn into their corresponding short vocalic form.

What do you mean by ‘short vocalic form’ here? Generally, a semivowel is considered to be just another name for a very short non-syllabic vowel.

I have short vowels and long vowels. Should I rephrase it?

No, I was just missing context — that clarifies things. (It might be simpler to say they ‘turn into the corresponding short vowel’, but in context that means exactly the same thing.)

[Travis B.]

One thing that complicates things, though, is that people speaking some languages can contrast, say, /ji/ with /i/ and /wu/ with /u/ (myself included). If /j/ and /w/ here were simply non-syllabic vowels, this should solely be a length contrast, but as at least how I myself pronounce /ji/ and /wu/ here this is not the case.

[bradrn]

One thing that complicates things, though, is that people speaking some languages can contrast, say, /ji/ with /i/ and /wu/ with /u/ (myself included). If /j/ and /w/ here were simply non-syllabic vowels, this should solely be a length contrast, but as at least how I myself pronounce /ji/ and /wu/ here this is not the case.

Haven’t we argued about this before?

[Travis B.]

One thing that complicates things, though, is that people speaking some languages can contrast, say, /ji/ with /i/ and /wu/ with /u/ (myself included). If /j/ and /w/ here were simply non-syllabic vowels, this should solely be a length contrast, but as at least how I myself pronounce /ji/ and /wu/ here this is not the case.

Haven’t we argued about this before?

I think we have.

[conlangernoob]

What do you guys think of my vowel Sandhi system? By the way, I have a typical five-vowel system plus phonemic length plus the diphthongs /eu/, /au/ and /ai/.

Here it is:

When two word-internal morphemes meet and multiple vowels come in contact, eight sound changes are applied, in order, simplifying the vowel string.

1. When a short vowel and the same short vowel come in contact, they become a long vowel.
2. When a long vowel and a short or long vowel of the same quality come in contact, they become an extra-long vowel. Note that extra-long vowels are phonemically two vowels, but are marked as one vowel with a circumflex in the orthography.
3. All instances of /ai/, /ei/, or /ae/, short or long, become /ai/.
4. All instances of /au/, /ou/, or /ao/, short or long, become /au/.
5. All instances of /eu/, /eo/, or /iu/, short or long, become /eu/.
6. When two long vowels come in contact the second vowel becomes short.
7. When there are any vowels in contact with the beginning or end of a diphthong of the same quality, the vowel is deleted.
8. When a long vowel comes in contact with the beginning or end of a diphthong, the vowel becomes short.

I’m particularly worried about rules 3-5, so comments are appreciated.

Until anon,

conlangernoob

[alice]

What do you guys think of my vowel Sandhi system? By the way, I have a typical five-vowel system plus phonemic length plus the diphthongs /eu/, /au/ and /ai/.

(etc)

I’m particularly worried about rules 3-5, so comments are appreciated.

No need to worry; I do that kind of thing all the time.

It might be helpful to show a table of all the possible combinations before and after your rules apply btw; it'll help you see if you've missed anything out. I've had exactly that problem a few times.