Keenir's scratchpad -

[bradrn]

...and now the reduction can have the vowel increasing in how many pronounciations (for lack of my brain firing at this hour - heater's busted) and also the vowel decreasing its number of pronounciations.

Sorry for being vague. This is quite correct — vowel reduction usually involves, well, a reduction in the number of vowels. (I believe; confirmation will have to wait until I find that article I was thinking about.) I was talking more about the target of vowel reduction; that is, whether vowels reduce to a central vowel or a peripheral vowel.

EDIT: I think I’ve found the article: Crosswhite’s Vowel reduction, in Phonetically Based Phonology. But it turns out to be much less interesting than I recalled — mostly Optimality Theory stuff.

[keenir]

*snip*

Okay, to see if I understand this...

The pronounciation ['aŋ] is the realization.
The surroundings ['äŋ]are the condition.
The reduction is...

EDIT: GAH...I did the next section as you see it, because when I was reading (yes, and re-re-reading) the advice everyone had helpfully provided, my brain switched the words "stressed" and "unstressed" in statements such as reduction to [ɛ~e~i] in (some) unstressed syllables, variation of [æ~a~ɑ(~ɒ)] in stressed syllables...I'll make the corrections below them.

I'll use the number 5 from earlier on my list, and coin a new number: 100...also, to reiterate an earlier mention: a quick explanation: as this started as a one-vowel conlang, I figured it would work as a Pure Abjad, meaning no marking of any vowels; now that allophony has raised its head, the only markings would be for either stressed or unstressed syllables...so I picked the latter, underscoring them.

'äŋ.'äʃ.ma.ja 05 ang.ash.ma.ya
... becomes ɛŋ.əʃ.ma.ja
{though you're about to see that a second marking would be needed, so the next one is represented as underscored and bold}

While for 100, ['äŋ.ʔä.ʔä.jä] <ang.ha.ha.ya> becomes [äŋ.ʔɒ.ʔɒ.jä] <ang.ha.ha.ja> if we say its moving to the periphery...and hard to further to the edge of the map than there (sorry; that just makes me smile).....and that in turn becomes - not sure if its more likely to become äŋ.ʔɒ.з.jä or äŋ.ʔз.ɒ.jä


CORRECTED VERSION:
'äŋ.'äʃ.ma.ja 05 'ang.ash.ma.ya
... becomes ɒŋ.aʃ.mə.jə]

While for 100, ['äŋ.ʔä.ʔä.jä] <ang.ha.ha.ya> becomes - not sure if its more likely to become [ɒŋ.ʔɛ.ɘ.jə] or [ɒŋ.ʔɘ.ɛ.jə] {is that still reducing, or has it become mutation because its expanding into consonants and syllable breaks, or it the difference a whole 'nother lesson?}
(sorry)


Thoughts?

[bradrn]

The pronounciation ['aŋ] is the realization.
The surroundings ['äŋ]are the condition.
The reduction is...

I think you’re maybe getting a bit too focussed on the terminology here. The important thing to understand is that the underlying phonemes are abstract objects which can be pronounced at the surface level as one or more concrete sounds (a.k.a. allophones), with each occurrence of that phoneme being pronounced (a.k.a. realised) as a different allophone depending on its environment where it appears.

I'll use the number 5 from earlier on my list, and coin a new number: 100...also, to reiterate an earlier mention: a quick explanation: as this started as a one-vowel conlang, I figured it would work as a Pure Abjad, meaning no marking of any vowels; now that allophony has raised its head, the only markings would be for either stressed or unstressed syllables...so I picked the latter, underscoring them.

'äŋ.'äʃ.ma.ja 05 ang.ash.ma.ya
... becomes ɛŋ.əʃ.ma.ja
{though you're about to see that a second marking would be needed, so the next one is represented as underscored and bold}

While for 100, ['äŋ.ʔä.ʔä.jä] <ang.ha.ha.ya> becomes [äŋ.ʔɒ.ʔɒ.jä] <ang.ha.ha.ja> if we say its moving to the periphery...and hard to further to the edge of the map than there (sorry; that just makes me smile).....and that in turn becomes - not sure if its more likely to become äŋ.ʔɒ.з.jä or äŋ.ʔз.ɒ.jä


CORRECTED VERSION:
'äŋ.'äʃ.ma.ja 05 'ang.ash.ma.ya
... becomes ɒŋ.aʃ.mə.jə]

While for 100, ['äŋ.ʔä.ʔä.jä] <ang.ha.ha.ya> becomes - not sure if its more likely to become [ɒŋ.ʔɛ.ɘ.jə] or [ɒŋ.ʔɘ.ɛ.jə] {is that still reducing, or has it become mutation because its expanding into consonants and syllable breaks, or it the difference a whole 'nother lesson?}

This really confuses me. What phonological rules describe these processes of vowel reduction? They don’t seem to be consistent in any way here, or at least I can’t find any consistencies. Which makes little sense to me, because vowel reduction is just another kind of phonological process, and like most other kinds of phonological process, it fires in predictable ways and can be described by consistent sound change laws. As with most other parts of linguistics, it’s not a matter of ‘pick a random thing to put here’.

Some other miscellaneous notes:

• In answer to your last point, it’s totally realistic to have, say /'äŋ.ʔä.ʔä.jä/ → [ɒŋ.ʔɛ.ɘ.jə] (note also the slashes for phonemic transcription!), but that has more than just vowel reduction going on. Vowel reduction is usually taken specifically to mean the alteration and merging of vowels in unstressed syllables, which is present, but in this case you also have a process of glottal stop deletion (ʔ→∅/ʔV_, ‘delete a glottal stop after another glottal stop’, is a perfectly fine thing to have), as well as whatever other process is causing unstressed /ä/ to split into [ɛ ɘ ə] in unstressed syllables… which probably has a name too, but who cares what it’s called anyway?
• Taking a broader view of things: if the realisation of vowels is conditioned purely on whether they’re in a stressed syllable or not (which is what people usually mean by ‘vowel reduction’), that only lets you choose between two sets of allophones. To have more allophones for a single vowel phoneme, you should think about where precisely those allophones occur and why they occur there.
• It’s also illuminating to think diachronically, by considering why your language has only one vowel in the first place. As previously mentioned, in natlangs this is usually a consequence of vowel features being transferred onto surrounding consonants; those consonants then become palatalised or labialised, and the underlying vowel inventory is collapsed into a vertical vowel system — usually /ɐ ə/, more rarely /ɐ/, /ɐ ə ɨ/ or /ɐ ɜ ɘ ɨ/.
• By the way, an important thing to remember about these transcriptions is that they’re abstract, like all phonemic transcription. /ɐ ə/ is really more like [e~ɛ~a~ɐ~ɜ~ɑ~ɒ~ɔ~o i~ɪ~ɨ~ə~ɘ~ʊ~u] or something: the symbols are simply mnemonics, which may or may not be similar to how the concrete allophones are notated. For instance, the Marshallese vowel system has been transcribed variously as /ɐ ɜ ɘ ɨ/, /a e ȩ i/ and / /.
• If your goal is to make a realistic language which can unambiguously be represented in a ‘Pure Abjad’, just giving it one vowel won’t be enough: you’ll still need to distinguish, say, /äŋ.ʔä.ʔä.jä/ from /ŋä.ʔä.ʔäj/. On the other hand, if it can be represented unambiguously without any vowels… well then, that means you can remove all the vowels from the underlying representation, without losing any information at all! Which means that what you have is actually a zero-vowel system, not a one-vowel system.
• Of course, you can feel free to disregard all of this and just make the new conlang the way you want. But either way, you should still think about how many vowel phonemes there really are, how they are realised in speech, and what determines those realisations. If you only want to have one surface vowel, or you want to have lots in free (i.e. conditioned) variation, that’s absolutely fine; just make sure you’re consciously aware that that’s what you’re doing.

[keenir]

This really confuses me. What phonological rules describe these processes of vowel reduction? They don’t seem to be consistent in any way here, or at least I can’t find any consistencies.

for right there, I agree i erred: i focused on just trying to show the change, and limited it to only two numbers, rather than show how the change applies over all the whole list of numbers (like i'd done in prior posts with this numberlang)

Some other miscellaneous notes:

many thanks.

[*] It’s also illuminating to think diachronically, by considering why your language has only one vowel in the first place. As previously mentioned, in natlangs this is usually a consequence of vowel features being transferred onto surrounding consonants; those consonants then become palatalised or labialised, and the underlying vowel inventory is collapsed into a vertical vowel system — usually /ɐ ə/, more rarely /ɐ/, /ɐ ə ɨ/ or /ɐ ɜ ɘ ɨ/.

my original plan, which I should have mentioned at some point, was that these one-vowel numbers would be adopted by a civilization new to the region & who would take the number script and improve upon that.

beyond that, it was Just To See If I Could Do It and nothing more.

[*] If your goal is to make a realistic language which can unambiguously be represented in a ‘Pure Abjad’, just giving it one vowel won’t be enough: you’ll still need to distinguish, say, /äŋ.ʔä.ʔä.jä/ from /ŋä.ʔä.ʔäj/.

In my notes (longhand on paper, nowhere on computer), I wrote that all syllables are CV except for the word-initial ones (which can only be VC - and yes, i retained second syllables for a few numbers which are also VC; though on paper those numbers also have properly-behaved VC.CV.CV.(CV) forms) , which only show up as the "number" marker in this wordlist.

On the other hand, if it can be represented unambiguously without any vowels… well then, that means you can remove all the vowels from the underlying representation, without losing any information at all! Which means that what you have is actually a zero-vowel system, not a one-vowel system.

tempting tempting...

One-vowel first, then none.

[bradrn]

[*] If your goal is to make a realistic language which can unambiguously be represented in a ‘Pure Abjad’, just giving it one vowel won’t be enough: you’ll still need to distinguish, say, /äŋ.ʔä.ʔä.jä/ from /ŋä.ʔä.ʔäj/.

In my notes (longhand on paper, nowhere on computer), I wrote that all syllables are CV except for the word-initial ones (which can only be VC - and yes, i retained second syllables for a few numbers which are also VC; though on paper those numbers also have properly-behaved VC.CV.CV.(CV) forms) , which only show up as the "number" marker in this wordlist.

Well, in that case you already have a zero-vowel system! If the language has only one vowel, and its occurrence is predictable in all words, then you might as well take the underlying forms as being just /CCC…C/ and use a vowel-insertion rule to explain the surface distribution of the vowels.

On the other hand, if it can be represented unambiguously without any vowels… well then, that means you can remove all the vowels from the underlying representation, without losing any information at all! Which means that what you have is actually a zero-vowel system, not a one-vowel system.

tempting tempting...

One-vowel first, then none.

If you want to go that route, I’d suggest starting with a two-vowel system, those being more common and much easier to understand. After that, one- and zero-vowel systems are just two-vowel systems where one or both of the vowels are predictable and hence not phonemes.

[keenir]

Well, in that case you already have a zero-vowel system! If the language has only one vowel, and its occurrence is predictable in all words, then you might as well take the underlying forms as being just /CCC…C/ and use a vowel-insertion rule to explain the surface distribution of the vowels.

If you want to go that route, I’d suggest starting with a two-vowel system, those being more common and much easier to understand. After that, one- and zero-vowel systems are just two-vowel systems where one or both of the vowels are predictable and hence not phonemes.

Understood; i shall do so.

ALSO...was thinking over earlier advice, and I have a hunch of which I'm unsure of...

A vowel's realization (and, I'm suspecting, also mutations) is depending upon the surrounding consonants, at least in CV, VC, CVC enviroments such as I'm using...

So, assuming all other details (stress, etc) are equal, an [ä.] would be inclined to mutate to an [e . i] if the enviroment is more labial or dental...while the [ä.] would be inclined to mutate to a [ɒ.] if the enviroment were more glottal...ditto mutating or being realized as [i . i] if the enviroment is nasal or sibilant fricative. (and in granddaughter langs, returning to [a . ä] if the enviroment is lateral or trilling)

Is that a good guideline, or is my hunch off-base?

[Zju]

I don't know how labial environment would shift [a] to [e] or [ i ] - if anything, I'd expect it to shift [a] to a rounded vowel.

ditto mutating or being realized as [i . i] if the enviroment is nasal or sibilant fricative.

I'm not sure of the phonetic justifications of this either, but it could work for historical reasons.

(and in granddaughter langs, returning to [a . ä] if the enviroment is lateral or trilling)

sure, liquids can centralise neighbouring vowels, I think

[keenir]

I don't know how labial environment would shift [a] to [e] or [ i ] - if anything, I'd expect it to shift [a] to a rounded vowel.

I think I was being a bit of hyperbole in terms of making the vowel go all the way to as far leftward on the vowel chart as possible. sorry for any and all confusion i caused.

[bradrn]

So, assuming all other details (stress, etc) are equal, an [ä.] would be inclined to mutate to an [e . i] if the enviroment is more labial or dental...while the [ä.] would be inclined to mutate to a [ɒ.] if the enviroment were more glottal...ditto mutating or being realized as [i . i] if the enviroment is nasal or sibilant fricative. (and in granddaughter langs, returning to [a . ä] if the enviroment is lateral or trilling)

Is that a good guideline, or is my hunch off-base?

It all depends, but this particular set of alternations strikes me as being somewhat articulatorily implausible. Natural phonological rules tend to a be a bit more articulatorily justified… not always, of course, but most of the time there’s a reasonable explanation of why some change might happen. I think Zju’s suggestions would be a good starting point here.

It might also be useful to consider the vowel alternations in Moloko (A grammar of Moloko, Friesen 2017):

/a/ → [ɔ] / with word-level labialisation
/a/ → [ɛ] / with word-level palatalisation
/a/ → [œ] / next to a labiovelar or /j/
/a/ → [a] / otherwise

(ə) → [ʊ] / with word-level labialisation
(ə) → [ɪ] / with word-level palatalisation
(ə) → [i] / next to [j]
(ə) → [u] / next to [w]
(ə) → [ø] / next to a labiovelar or /j/
(ə) → [ə] / otherwise

Or in Abkhaz (Abkhaz, Hewitt 1979):

/a/ → [eː] / before palatalised consonants, unless preceeded by another /a/ or /ħ/
/a/ → [oː] / before /w/, unless preceeded by another /a/ or /ħ/
/a/ → [ɔ] / irregularly after /w/, labialised velars and uvulars
/a/ → [ə] / irregularly after other labialised affricates and fricatives

/ə/ → [u] / next to /w/ and after labialised consonants
/ə/ → [i] / next to /j/ and after palatalised consonants
/ə/ → [ɨ] / irregularly after palatalised consonants or lamino-postalveolars
/ə/ → [œ] / after ħʷ
/ə/ → [ə] / elsewhere

Of course, not all of these correspondences make complete sense at first sight, but for the most part they all have obvious explanations.

[Travis B.]

I'm with bradrn - vowel allophony in languages with small, vertical vowel systems typically relate rather predictably to adjacent consonants' palatalization, labialization, dorsalness, pharyngealization, etc. (and even though I have not specifically read about a language to which this applies, I would expect nasalization to behave much the same way). I.e. palatalization and /j/ typically condition front realizations, labialization and /w/ typically condition rounded realizations (which are typically back ones unless the vowel in question is also adjacent to a palatalized consonant or /j/), dorsals often condition back realizations, pharyngealization very often conditions lowered back realizations, nasal consonants very often nasalize preceding (and less commonly, following) vowels, and so on.

[keenir]

new material on page one...hit the Edit not Reply.


I'm not aiming to make a conlang with trilitral roots...I just needed to light a fire under me, and I had some nice-looking consonant-pairs from helping a friend learn the Ancient Egyptian alphabet on Wondrium. So these are just the starting point that will hopefully grow into better and better statements, and when it grows, they might be kept as Easter Eggs or may be tossed aside.

DS
[d.s]
Meaning: Thus, therefore, it follows (that)...
Origin: "rocks fall, everyone dies".
Produces: [dys] = thus; [dos] = Q;

GWN
[gw.n]
Meaning: I say, "..."
Origin: A bit of Gwen and a bit of Red Green.
Produces: [gwyn] = I say; [gwyn.gwy ... gwyn.gwy] = "..."

HL
[h.l]
Meaning: place, LOC (locative),
Origin: Holy of Holies.
Produces: [hyl] = place; [hol] = corral, space; [hol.ol] = a thing;

Locatives:
hyl can be used as a locative, so long as it is not prefixed by a radical/classifier (at this point, i'm not sure which way I'm going to take them, so my few notes call them by both names)

ma = close by, near; closed, shut
sa = open; can be far away or nearby
ek / -ek = inwards, enter, within

Reduplication:
-ol



.

(but yes, it's a real idiolect that I'm keeping to myself for now...)

.
Dos - hol.ol ma hol, hyl-ek sa hol
Q - thing close corral, place-inward open space
A private thing in a public place?
Gwyn.gwy (dos practice) gwyn.gwy
I say (Q practice) I say.
"Practice?"

[keenir]

...

as yet, there are no numbers - ordinal or otherwise.


Pyllotys myjur taba ganan?
pyl.lo.tys my.jur ta.ba gan.an
Q.LOC creature leg PL{POSS.by}3rdPerson
Where are its legs?
This is a direct reference to a D&D creature which in write-ups, has legs -- but in art, has none. See here: https://youtu.be/baJUAv3EnNI

Pyllotys (myjur taba ganan) ?
{thats as far as I can follow the syntax tree thus far - need to finish the Construction Kit and double-check I understood the parts I think I understand}
{{also need to re-read the glossing rules; i know i'm rusty}}


Pyllotys hast taba mikkin
pyl.lo.tys hast ta.ba mik.kin
Q.LOC human.body leg glove
Where are their socks?

note: pyllotys can also be pronounced [pyl.lo.tyz] as well as [pyl.lo.tys].....in addition, you could use [dos hyl] in its stead, and it'd mean the same, but the query came to my mind with pyllotyz and so thats what I'm using.

[keenir]

...

Tas gwyn.
Yes I.say
I agree.


Dys-holol myr-hololol, sapat gwynamat {mile} emd hast hul emd hast hul emd hast hul
Dys-hol.ol myr-hol.ol-ol, sa.pat gwyn.am.at {mile} emd hast hul emd hast hul emd hast
Brings sun-thing-(?unanalyzed as yet), walk I.say-I.doing {mile} number 1/4 and number 1/4 and number 1/4
Brings today, walk I {mile} counting/number 1/4 & 1/4 & 1/4
Today I walked 3/4 of a mile.


It gets worse: I'm toying with making "is" (in the sense of "are the same as" and the equals sign) be hol.ol.ol.ol.....yes, thats "thing" + (?whats the reduplication do?) + (?another reduplication - does it do the same as the first reduplication in the word?)


Probably nothing will come of it, but it surprised me when i thought it up and jotted it down.

[Travis B.]

It gets worse: I'm toying with making "is" (in the sense of "are the same as" and the equals sign) be hol.ol.ol.ol.....yes, thats "thing" + (?whats the reduplication do?) + (?another reduplication - does it do the same as the first reduplication in the word?)

Even better, find a way to put a -/z/ morpheme at the end, so you can have hol.ol.ol.olz.

[keenir]

At the library, I was looking through Comrie's Atlas and Dalby's Dictionary of Languages...and I thought to try my hand at one of those one-page entries. Partway through, I got distracted by the list of word orders, and how each group would arrange the words
cow & grass & eat

...and I ended up translating them into the following:

Ashshɛn tskɛ: tashsas
Aʃ.ʃɛn tskɛ: taʃ.sas
do.eat mass.count.noun Class:low.plants (not eaten)

gakkag dɛtap tɛggɛt
gak.kag dɛ.tap tɛg.gɛt
grass one Class:fur

ɛkkow
ɛk.kow
wooly.cow (loan word)

Eat. Grass . Cow

=Cows eat grass.

...and from there, I made an abjad:

[keenir]

Its Lexember time!

This month's language: Nalma

December 1st: mlaataln [mla:.taln]. "Unseen", "Unseeable", "None see (verb) (noun)."
ie, "None see Death."
also, if you place a second...well, I'll just demonstrate: "None see hear Death" = "Nobody hears Death"...which, yes, I worked backwards to get to today's word, from a passage in Gilgamesh (Tablet X, vi)



Some of the of the Nalma regions include the following former kingdoms:
* Sadalanuutama - essentially "nuuta is written here" was the name of their kingdom which had been founded by a Nuuta-literate side branch of a royal family off to the north, who brought their writing system with them, spreading it across their domain. Sadalanuutama and Karsema requested Nalma arbitration when the two kingdoms were unable to come to an agreement on how they should reunite after the long separation which had seen each nation obtain their own writing systems. The Nalma did attempt it, as requested...but eventually took the two kingdoms over.

* Nalma - "nalma is written here"...though the the Nalma writing system - the Royal Script - was named for a mythical ancestor of the royal family who had founded the nation. Most historians agree that they were only able to take Karsema and Sadalanuutama into their holdings, because they made it so both nations's people could continue using their own writing systems -- which influenced the birth of the Nalma Royal Script at the time of the kingdom's founding.

* Karsema - "kars is written here". (TBD)


In the next image, there are four rows. These correspond to:

ROW ONE: The Royal Script is an alphabet for the most part. Two letters are <a> and <ml>, the latter of which can only be syllable-initial: <ML- > <A>

ROW TWO: One of the scripts of the provinces is Nuuta, which is a syllabary. This means we have <ma> and <la> on hand to correspond to the above Royal Script signs: <ma la>

ROW THREE: This is the layout of how the Royal Script and the Nuuta Syllabary are written in Nuuta-majority regions of the kingdom. {I tried to remember how Coe arranged Mayan translations; will correct when i look over his books again}
<ML- m(a) l(a) A (l)a>

ROW FOUR: During the short-lived Rebellion Period, one province's governor-turned-king oversaw two attempts to reform the script layout. Early in his kingship, the phonetic signs were all placed following the main vowel: <MLA- m(a) la>

...while shortly before the death of himself and his loyalists, the repeated phonetic sign was removed, leaving the syllable to end with the main vowel: <ML- m(a) la A>

[keenir]

Working offline on a model language thats sufficient for the logography I'm making, which I'm processing through into a syllabary and mayyybe an alphabet or an abjad (like i use it for in one of the fluency threads)...


Definitions found in my copy of Deciphering Aztec Hieroglyphs by Gordon Whittaker.