Scratchpad (non-Sego)

[sasasha]

A scratchpad for conlang stuff of mine that is not related to my conworld.

Index

• Guqpik Samas

[sasasha]

Guqpik Samas
‘The Language of Choice’

A language which permits aggressively pro-drop behaviour and high ambiguity resolved through context, but which presents a variety of typological options to mark syntactic relationships.

Why this redundancy? Natural languages have redundancy, so this language will be naturalistic in that (and likely only that!) sense. The options make for a potential breadth of strategies of expression as a potentially interesting mode of translinguistic communication, as well as, probably, quite a silly mess. Finally, there is philosophical play to be had: how else might we express ourselves, if we are permitted more of the tools in the box?

This language could be seen as an analogical counterpart for typology to the International Phonetic Alphabet. Just as the IPA allows speakers of various languages to understand phonemically and to pronounce each other's speech, so this language might allow speakers of different languages insights into each other's syntactic structures. (Not going to bring any insights to this crowd; but it might be interesting for people just starting out exploring typology.)

I’m not trying to make an IAL here. It’s just a casual engelang experiment that has filled this afternoon, and possibly only this afternoon: how ‘make-your-own’ can a language’s morphosyntax be before it breaks completely? (It’s probably broken out of the box.)


*

Phonology and Orthography

Consonants

	Labial	Coronal	Palatal	Velar	Glottal
Nasal Obstruent	m	n
Fortis Obstruent	p	t		k
Lenis Obstruent	b	d		g
Affricate		c	q
Fricative	f	s	x		h
Lenis Sibilant		z	j
Lateral		l
Rhotic		r
Approximant	w		y

Vowels

i		u
e		o
	a

⟨v⟩ can be used to denote unambiguously a ‘copy vowel’, e.g. domoh ‘house’ may be written domvh.

An apostrophe may be used to show unambiguously in writing a deleted medial consonant (common in compounds), e.g. muk’sdom’h is an orthographic variant of musdoh ‘barn’ (animal-house).

Precise values can vary by speaker. ⟨z j⟩ may be affricates or fricatives, and either voiced or unaspirated, depending on the speaker’s preferred fortis/lenis distinction.

The orthography thus uses all 26 English letters, with no diacritics, for no particular reason.

Syllable structure is (C)V(C). Any consonant may appear in the onset and coda. Any allophony in clusters is not indicated in writing. Words beginning with vowels are lightly glottalised.

‘Strong’ roots take the form CVCvC, as in, the second vowel is a copy of the first. There are thus 20x5x20x20=40,000 separate strong roots available. ‘Weak’ roots (borrowings) may take any other syllabic structure, e.g. radio. (Many weak roots have derived alternatives, e.g. kimguq ‘prong/antennae sound’, from kicim gubuq.)

Stress falls on the first syllable of a word. As there are so many morphosyntactic strategies available, stress may appear, acoustically, to be free. For instance, xoaigubuq and xo ai gubuq (both likely meaning ‘I spoke/made a sound’) will differ in terms of stress. It is assumed that hypothetical speakers may begin to use these distinctions to various pragmatic effects.



*

Morphosyntax

Verbs and nouns may both appear without any explicit person marking. Unmarked verbs are ambiguous as to their arguments, though context may serve to disambiguate; unmarked nouns are ambiguous as to whether or not they are possessed and by whom.

The following options (TBC) are present for grammatical marking of various cross-linguistically commonly marked features. They may be present as prefixes, suffixes, or free morphemes collocated either before or after roots, hence there is no or little obligatory order of morphemes (TBC) except in certain cases:

Person

• i marks first person
• u marks second person
• a marks third person

Animacy

• a- prefix may be applied to indicate animate marking

Number

• -n- may be applied to mark collocation, or -n by itself to imply plurality (see below*)

Reflexivity

• -s- may be applied to mark reflexivity

Gender

• -li- feminine
• -mo- masculine
• -hu- neuter

Case

• -(a)t(a)- may be applied to mark direct object
• -(a)k(a)- may be applied to mark subject
• -al / la- may be applied to mark indirect object
• -ar / ra- may be applied to mark possessor

*The number marking is essentially additive. X-n-Y means X plus Y. X-n means X plus (something). This results in a plausible plural morpheme na 'plus it'.

Relatively complex pronouns, or affixes, may be constructed:

aal - inanimate 3rd person indirect object
ainu - animate, 1st and 2nd person (inclusive 'we')
uus - 2nd person reflexive
aumo - 2nd person masculine

There is plenty of built in ambiguity to this system. For instance, does uus neglect to include an animacy marker because its referent is inanimate? (Perhaps the speaker is addressing an object anthropomorphically.) Or is it simply assumed that the referent is probably animate? And in the absence of explicit case markings, many forms may be taken as having any or multiple potential syntactic role(s); the case markings may, of course, be applied according to various alignments.

Applying these to roots in various ways gives a panoply of potential modes of expression. And confusion.

*

Compounding

This is a derivational rule I’m chucking in dogmatically for fun. Compounds of strong roots can be created by removing the middle consonant and copy vowel of both parts. There are (probably rather unfortunately) no rules for which part of the compound is to be considered the head. Compounds may imply collocation rather than modification/possession, e.g. guqpik ‘language’ [speak+write].

*

A bit of testing

sutul - harp
garah - antelope
filguq - telephone
gubuq - to speak, to sound
fitil - wire, cord; to wind
noroh - to run
pilik - to write
kicim - prong/antenna
mukus - animal
guqpik - language, communication, to communicate
samas - to choose; choice

i sutul gubuq - my harp sounds, I play the harp
sutulai gubuqa - my harp sounds, it plays my harp
noroh garah - the antelope runs
garahna, aannoroh - the antelopes run
gubuqikalit filguq - I speak to her on the phone

*

Temporality

The following morphemes may be affixed to or collocated with any root, nominal or verbal.

• xo - past
• wa - present
• zi - future

garahxo - the former antelope
garah xo - the former antelope
zigubuqius - we will speak to each other
gubuq - we will speak to each other (aggressively pro-drop and context-driven, remember)

*

Word classes

I think the lexicon would do well to specify potential verbal and nominal meanings of roots. I don’t know, given the flexible vibes here, whether I can impose more meaningful boundaries than that.

[sasasha]

ya

Any two morphemes can be separated by the particle ya, which may prove useful in disambiguating this mess.

[/nɒtɛndəduːd/]

this is a really interesting conlang idea!

[sasasha]

this is a really interesting conlang idea!

Bezek!
Thank you!

bezek - (to be) grateful; gratitude

harak - South; lower half (of body)
netey - East, morning
janaw - West, evening
timiq - North; upper half (of body)
buduk - centre, noon
xalar - world
tidim - pole; apex, zenith
seref - night

Tiqxar - the Global North
Hakxar - the Global South

Neyxar - Asia & Oceania
Bukxar - Europe & Africa
Jawxar - the Americas
Sefxar - the Pacific

Timiq Neyxar - Asia
Harak Neyxar - Oceania
Timiq Bukxar - Europe
Harak Bukxar - Africa
Timiq Jawxar - North America
Harak Jawxar - South America
Timiq Sefxar - North Pacific
Harak Sefxar - South Pacific

Timxar - polar lands
Timiq Timxar - the Arctic
Harak Timxar - Antarctica

All these compounds can have 400 different meanings, before you even get into where the head is, if any. Ambiguity, context... I’m thinking of ASL (and probably other sign languages) where you can establish a referent somewhere in physical space and refer to it just by pointing to it. Unless it’s very common, on first mention, you might want to use the long form of a compound, after which you can clip it down.

Possible morpheme:

(v)c(v) - marker of head position (modifier is represented by copy vowel, which matches what is before or after appropriately)

[sasasha]

yutut - equator; waist
xibib - chest, torso; northern tropics, Tropic of Cancer
panar - calf, upper leg; southern tropics, Tropic of Capricorn
karat - head; northern temperate region
hexep - foot, shin; southern temperate region

Yutxar - equatorial lands
Xibxar - lands around the northern tropics
Parxar - lands around the southern tropics
Xibpar Xalar - tropical lands

Karxar - northern temperate lands
Hepxar - southern temperate lands
Karhep Xalar - temperate lands

I’m not going to lay all these out, but this can be useful for certain regional designations:

Xibib Bukxar - North Africa / Mediterranean / ‘Middle East’
Xibib Neyxar - South & South-East Asia
Xibib Jawxar - Central America / Caribbean

[bradrn]

Why this redundancy? Natural languages have redundancy, so this language will be naturalistic in that (and likely only that!) sense. The options make for a potential breadth of strategies of expression as a potentially interesting mode of translinguistic communication, as well as, probably, quite a silly mess. Finally, there is philosophical play to be had: how else might we express ourselves, if we are permitted more of the tools in the box?

This language could be seen as an analogical counterpart for typology to the International Phonetic Alphabet. Just as the IPA allows speakers of various languages to understand phonemically and to pronounce each other's speech, so this language might allow speakers of different languages insights into each other's syntactic structures. (Not going to bring any insights to this crowd; but it might be interesting for people just starting out exploring typology.)

This reminds me very strongly of Lojban (or at least some of the goals of Lojban).

[sasasha]

Why this redundancy? Natural languages have redundancy, so this language will be naturalistic in that (and likely only that!) sense. The options make for a potential breadth of strategies of expression as a potentially interesting mode of translinguistic communication, as well as, probably, quite a silly mess. Finally, there is philosophical play to be had: how else might we express ourselves, if we are permitted more of the tools in the box?

This language could be seen as an analogical counterpart for typology to the International Phonetic Alphabet. Just as the IPA allows speakers of various languages to understand phonemically and to pronounce each other's speech, so this language might allow speakers of different languages insights into each other's syntactic structures. (Not going to bring any insights to this crowd; but it might be interesting for people just starting out exploring typology.)

This reminds me very strongly of Lojban (or at least some of the goals of Lojban).

To my understanding, Lojban is more prescriptive. In fact, emphatically so: Lojban wants to express everything unambiguously. The experiment here is with flexibility, and aggressive ambiguity.

But yes, there’s still a lot shared.

[bradrn]

This reminds me very strongly of Lojban (or at least some of the goals of Lojban).

To my understanding, Lojban is more prescriptive. In fact, emphatically so: Lojban wants to express everything unambiguously. The experiment here is with flexibility, and aggressive ambiguity.

That makes sense.

Another random thought: I think it might be quite difficult to prevent this from becoming a kitchen-sink. Unless, of course, kitchen-sinkery is the whole point…

EDIT: and after writing this I was truly shocked to discover that Wiktionary has an entry for ‘kitchen-sinkery’. But, curiously, not with the meaning I just used it with!

[sasasha]

This reminds me very strongly of Lojban (or at least some of the goals of Lojban).

To my understanding, Lojban is more prescriptive. In fact, emphatically so: Lojban wants to express everything unambiguously. The experiment here is with flexibility, and aggressive ambiguity.

That makes sense.

Another random thought: I think it might be quite difficult to prevent this from becoming a kitchen-sink. Unless, of course, kitchen-sinkery is the whole point…

I think kitchen sinkery is the whole point, and whilst that was fun for a bit, it is getting a bit exhausting and silly. Boundaries save us time and needless mental effort in communication. I have the sense this could be an interesting language to write poetry in, though.

[bradrn]

I think kitchen sinkery is the whole point, and whilst that was fun for a bit, it is getting a bit exhausting and silly. Boundaries save us time and needless mental effort in communication.

Actually, I do feel that it should be possible to do something genuinely interesting with this idea. It would involve some careful thought about topics like: precisely which linguistic constructions are best at maintaining coreference in the presence of ambiguity? Or: what sorts of ambiguity are tolerated best in comprehending an ambiguous sentence?

For a concrete example of how I‘m thinking: my current conlang Eŋes has a switch-reference-esque system for subject marking. In translating a text, I discovered that this was sufficient to eliminate nearly all ambiguity in subject marking, with barely any explicit NPs needed. Taking this further, a carefully-designed combination of — let’s say — switch-reference marking + noun classes + obviation + case concord could completely eliminate the need for explicit NPs after the first mention. Then making each of these categories optional could give a lot of scope for a speaker to play around with precisely delineated amounts of ambiguity…

[sasasha]

I think kitchen sinkery is the whole point, and whilst that was fun for a bit, it is getting a bit exhausting and silly. Boundaries save us time and needless mental effort in communication.

Actually, I do feel that it should be possible to do something genuinely interesting with this idea. It would involve some careful thought about topics like: precisely which linguistic constructions are best at maintaining coreference in the presence of ambiguity? Or: what sorts of ambiguity are tolerated best in comprehending an ambiguous sentence?

For a concrete example of how I‘m thinking: my current conlang Eŋes has a switch-reference-esque system for subject marking. In translating a text, I discovered that this was sufficient to eliminate nearly all ambiguity in subject marking, with barely any explicit NPs needed. Taking this further, a carefully-designed combination of — let’s say — switch-reference marking + noun classes + obviation + case concord could completely eliminate the need for explicit NPs after the first mention. Then making each of these categories optional could give a lot of scope for a speaker to play around with precisely delineated amounts of ambiguity…

That sounds really interesting and is exactly the kind of thing I’m trying to get at. You’re right that it would require a great deal of careful thought to make the whole system flexible-but-robust enough to achieve its aims, and I’m not sure I can give it the time / am the right person to do such a thing.

Still, I think I’ll continue to play around, and try to develop this to a level where I can try using it for personal expression. I think parsing it will be harder than generating it (I find the same is true of Toki Pona).

[bradrn]

I think kitchen sinkery is the whole point, and whilst that was fun for a bit, it is getting a bit exhausting and silly. Boundaries save us time and needless mental effort in communication.

Actually, I do feel that it should be possible to do something genuinely interesting with this idea. It would involve some careful thought about topics like: precisely which linguistic constructions are best at maintaining coreference in the presence of ambiguity? Or: what sorts of ambiguity are tolerated best in comprehending an ambiguous sentence?

For a concrete example of how I‘m thinking: my current conlang Eŋes has a switch-reference-esque system for subject marking. In translating a text, I discovered that this was sufficient to eliminate nearly all ambiguity in subject marking, with barely any explicit NPs needed. Taking this further, a carefully-designed combination of — let’s say — switch-reference marking + noun classes + obviation + case concord could completely eliminate the need for explicit NPs after the first mention. Then making each of these categories optional could give a lot of scope for a speaker to play around with precisely delineated amounts of ambiguity…

That sounds really interesting and is exactly the kind of thing I’m trying to get at. You’re right that it would require a great deal of careful thought to make the whole system flexible-but-robust enough to achieve its aims, and I’m not sure I can give it the time / am the right person to do such a thing.

Well, I'd be happy to help if you want! I know that I've enjoyed playing around with such things in Eŋes. (I have a recent post about it, which right now I'm far too lazy/tired to link…)

[sasasha]

Well, I'd be happy to help if you want! I know that I've enjoyed playing around with such things in Eŋes. (I have a recent post about it, which right now I'm far too lazy/tired to link…)

Thank you, that’s very generous. I haven’t collaborated on a conlang project since Shaja, a collaborative ZBB conlang of circa... 2007? I can’t remember who was part of that. Nort, I think. Ahribar.

I’ll take a closer look at this aspect of Eŋes.

[bradrn]

Well, I'd be happy to help if you want! I know that I've enjoyed playing around with such things in Eŋes. (I have a recent post about it, which right now I'm far too lazy/tired to link…)

Thank you, that’s very generous. I haven’t collaborated on a conlang project since Shaja, a collaborative ZBB conlang of circa... 2007? I can’t remember who was part of that. Nort, I think. Ahribar.

And I’ve never done it!

I’ll take a closer look at this aspect of Eŋes.

I’m less tired now, so I can give the relevant link as https://www.verduria.org/viewtopic.php?p=85891#p85891.

[sasasha]

Well, I'd be happy to help if you want! I know that I've enjoyed playing around with such things in Eŋes. (I have a recent post about it, which right now I'm far too lazy/tired to link…)

Thank you, that’s very generous. I haven’t collaborated on a conlang project since Shaja, a collaborative ZBB conlang of circa... 2007? I can’t remember who was part of that. Nort, I think. Ahribar.

And I’ve never done it!

I’ll take a closer look at this aspect of Eŋes.

I’m less tired now, so I can give the relevant link as https://www.verduria.org/viewtopic.php?p=85891#p85891.

Thanks, I’m enjoying reading about Eŋes, though haven’t taken in enough yet to comment, but will try to.

I’m going to continue to dump ideas here. Feel free to do the same if you like, we can talk about them and sort out any areas of confusion.

[sasasha]

Adpositions

How to deal with positional information?

I have resisted the temptation to allow (or perhaps, to emphasise) triple compounding in derivational processes. So this leaves a role for grammaticalised reduced nouns, which now can modify other nouns to provide positional info (common adjectives may also frequently appear in reduced form). Here are some of them:

teles - crown of head
bufun - sole of foot
purut - face, front
kadaw - back, behind
bogow - loud, beat, wave; left side
migin - quiet, placid, still; right side
secet - locality, nearby area, reach
xajaq - distance, horizon
fadax - internal organ
kuhur - cloak, blanket, garment, wrap

tes - above
bun - below
put - in front of
kaw - behind
bow - to the left of
min - to the right of
set - nearby
xaq - far away from
fax - inside
kur - surrounding, around

hidiw - person

How to resolve a string like garah put hidiw ‒ antelope in front of person, or person in front of antelope? One strategy is to use ya to break the syntactic chain (presented with somewhat random word boundaries):

garahya put hidiw - an antelope, in front of a person

garahput ya hidiw - in front of an antelope, a person

Another observation is that a speaker may settle into a pattern of generally modified-modifier or modifier-modified during an utterance, and may want to signify a switch using the -c- particle.

Some vocab we have already seen:

karat - head
xibib - chest
yutut - waist
panar - upper leg
hexep - lower leg/foot

kat - at head height
yut - at waist height
xib - at chest height
par - a little below waist height
hep - a little above floor height

Many of these could be productively combined, both with each other and other morphemes:

bowparu - to the left of you a little below waist-height

I feel like I can’t go into a lot of detail re syntax here because the point is, put it all together how you like (to whatever level of ambiguity or specificity you need/want).

Compass directions:

timiq - upper half, north
harak - lower half, south
netey - morning, east
janaw - evening, west

neytiq - north-east
jawtiq - north-west
neyhak - south-east
jawhak - south-west

sihis - a crumb, mote, speck, a little

sisney neytiq - a little east of north-east
etc.

A few more useful positionals, which may be combined:

hahaw - to fly, hover
ganam - to stand, to settle, to touch (a surface)

haw - beyond, nearby but not touching
gam - touching, on, directly on

calas - finger, digit
piniw - eye, view
fuhuh - nose, sense of smell

cas - there, where I’m pointing
piw - here, in view
fuh - here, close enough to smell

[bradrn]

I’m enjoying reading about Eŋes, though haven’t taken in enough yet to comment, but will try to.

Thanks!

[…] this leaves a role for grammaticalised reduced nouns, which now can modify other nouns to provide positional info (common adjectives may also frequently appear in reduced form).

More commonly known as ‘relational nouns’. As in your list, they often derive from body parts or geographical terms. (I have one or two survey papers somewhere if you want.)

How to resolve a string like garah put hidiw ‒ antelope in front of person, or person in front of antelope?

Usually languages use possessive constructions, so it’s not an issue — as indeed is shown in your English translation: ‘in front of (the) person’.

[sasasha]

[…] this leaves a role for grammaticalised reduced nouns, which now can modify other nouns to provide positional info (common adjectives may also frequently appear in reduced form).

More commonly known as ‘relational nouns’. As in your list, they often derive from body parts or geographical terms. (I have one or two survey papers somewhere if you want.)

Thank you, I’ll say yes to any papers you want to link. Of course, relational nouns. I used to (a dozen years ago) be pretty good with linguistics terminology, but whereas most of the denizens of this board have got better over the years, I have definitely regressed. Reminders are helpful and appreciated.

How to resolve a string like garah put hidiw ‒ antelope in front of person, or person in front of antelope?

Usually languages use possessive constructions, so it’s not an issue — as indeed is shown in your English translation: ‘in front of (the) person’.

I may be misunderstanding you here, or vice versa. To restate my question: in the context of no fixed noun phrase order, modifier-modified order, and the absence of a requirement to mark case (including genitive), how is one supposed to understand the relationship between two nouns separated by an adposition? It might be a postposition on the first or a preposition on the second. (Sorry if that is how you took my question, and it is me not receiving your answer properly.)

Since I’m trying to avoid setting one strategy for anything, I mentioned two possible strategies ‒ and you suggest one I left out, using a marked possessive construction, e.g. garahar put hidiw (antelope-GEN in.front person) ‘a person in front of an antelope’. Is that what you meant?

[sasasha]

Predicative relationals

Relational nouns may appear in full form, often supporting a predicative reading:

xihim - mountain

Garah teles xihim.
antelope crown mountain
There is an antelope on the mountain.
The antelope is on the mountain.

(Here, context alone is disambiguating from the reading ‘the mountain is upon the antelope’.)

Relational nouns, either full or reduced, like any unbound morphemes, may take personal affixes:

Bowi.
left.side-1
I’m to the left.

Note that in absence of -c- markers or clarifying context, it is technically possible to apply case markers in either direction, leading to ambiguity.

Bowuri.
left.side-2-GEN-1
I’m to your left.
You’re to my left.

This is valid (context may suffice to resolve) or can be resolved by assigning the hanging person affix an unambiguous case marker of its own, following the tendency of modifying morphemes to stay next to (in one direction or another) their referents. Thus in the following example, the nominative affix -k- cannot be taken to modify the 2nd person -u-.

Bowurik.
left.side-2-GEN-1-NOM
I’m to your left.

The case marking may be applied in any order, however. All of the following are, technically, valid expressions of the same sentence:

Bowruki.
Bowruik.
Bowurki.
Urbowik.
Urbowki.
Rubowik.
Rubowki.
Kibowur.
Kibowri.
Ikbowur.
Ikbowri.

All affixes may, optionally, be detached from the noun (affecting stress, and, potentially, pragmatics). This may aid in avoiding unwanted reanalysis:

huruk - sit
kuhur - surround; cloak, wrap

akuhuruk may be analysed as a-ku-huruk ‘you sit (on) it/him/her’, ak-u-huruk ‘it/he/she sits (on) you’, or a-kuhur-uk ‘you surround it/him/her’, ‘you are its/his/her cloak’; a simply way to disambiguate is to introduce word boundaries (audible in changed stress patterns, and because words beginning with vowels are glottalised).

In addition, it is obvious that, when one person affix is separated from the other by the root, there is less call for explicit marking of both cases, making the following (and more) equally valid:

Rubowi.
Kibowu.

It doesn’t really require repeating, but we are in such pro-drop territory that

Bow.

is a valid utterance by itself (though not, of course, unambiguous). This is, of course, equally true of English, and has as much to do with pragmatics as syntax: “Where are we going?” “Left.”


Optional Copula

-m- can function as a copula, though it is not required:

mi bogow ru - I am to the left of you.
ma tes xihim garah - the antelope is on the mountain / there is an antelope on the mountain.