(Conlangs) Q&A Thread - Quick questions go here [2010-2020]

[Reyzadren]

From a conlang-building perspective, what are the advantages of having a head-initial or head-final choice for compound words, other than the obvious "head-initial has the general most important word first" and "head-final words allow jumping into specifics immediately", especially with regards to syntax/morphology?

[Frislander]

From a conlang-building perspective, what are the advantages of having a head-initial or head-final choice for compound words, other than the obvious "head-initial has the general most important word first" and "head-final words allows jumping into specifics immediately", especially with regards to syntax/morphology?

In the grand scheme of things it doesn't really matter all that much, and that's not even considering that when counts as the "head" of a phrase/clause varies from language to language.

[Creyeditor]

I am working on a tonal conlang right now and I am thinking about how to present it. Should I give a systematic, comprehensive overview of the possible surface forms and the morphological alternations OR should I give just a list of underlying processes and how they can be used to derive the forms? I think I might prefer the first one, so that people can make up their own analysis. On th other hand the second one would be shorter, more elegant and less work for me.

[sangi39]

I am working on a tonal conlang right now and I am thinking about how to present it. Should I give a systematic, comprehensive overview of the possible surface forms and the morphological alternations OR should I give just a list of underlying processes and how they can be used to derive the forms? I think I might prefer the first one, so that people can make up their own analysis. On th other hand the second one would be shorter, more elegant and less work for me.

Just me personally, I'd prefer the second presentation alongside maybe two or three examples to show off each process. Presenting the underlying forms and the processes which lead to surface forms can, as you say, make it quicker to show off, but then giving specific examples as well give the reader a better feel for the language.

[qwed117]

I am working on a tonal conlang right now and I am thinking about how to present it. Should I give a systematic, comprehensive overview of the possible surface forms and the morphological alternations OR should I give just a list of underlying processes and how they can be used to derive the forms? I think I might prefer the first one, so that people can make up their own analysis. On th other hand the second one would be shorter, more elegant and less work for me.

Just me personally, I'd prefer the second presentation alongside maybe two or three examples to show off each process. Presenting the underlying forms and the processes which lead to surface forms can, as you say, make it quicker to show off, but then giving specific examples as well give the reader a better feel for the language.

Based on what I can see, what seems to be the M.O. on the CBB is surface-analysis coming first, then introducing mid-analysis that can be easily seen with the surface analysis. This is comparable to say showing Spanish's 3 main conjugations, and then showing the "stem-changing" roots and their morphology over that.
Strong deep-level analysis is not liked here to a large portion. That's probably because it's relatively difficult to think of and create, much less show. This is more comparable to showing that the ambifixing nature of cum in Latin is because cum used to be an adverb. It's not exactly the easiest to make.
I haven't really see much on just surface-level analysis or just mid-level analysis. I guess the former is simpler for CBBers who actually want to learn the language, but those are few in number, and can be equally sated, if not greater by the latter.

[LinguoFranco]

From a conlang-building perspective, what are the advantages of having a head-initial or head-final choice for compound words, other than the obvious "head-initial has the general most important word first" and "head-final words allow jumping into specifics immediately", especially with regards to syntax/morphology?

I usually go with head initial as I like to think of it in terms of you have to know what you are describing before you can describe it. Obviously, head final languages do just fine by putting modifiers before the noun they modify, and people understand those languages just fine, I just tend to go head initial as I see the head as the most important part of the phrase, and should therefore come first. I can't help you here much, but I haven't looked at head final as putting emphasis on the specifics (the modifiers of the noun) as the more important part.


On an unrelated topic, I am working on a testlang that only has a distinction between first and second persons. How do such languages say something like "He is tall," or "He is running?" Would it be like "This man is tall," or "This man is running?"

[Frislander]

On an unrelated topic, I am working on a testlang that only has a distinction between first and second persons. How do such languages say something like "He is tall," or "He is running?" Would it be like "This man is tall," or "This man is running?"

Or just "This is running", or even just zero marking if you have verbal person marking.

[Porphyrogenitos]

On an unrelated topic, I am working on a testlang that only has a distinction between first and second persons. How do such languages say something like "He is tall," or "He is running?" Would it be like "This man is tall," or "This man is running?"

There are some languages, including Latin, that didn't/don't have "real" third-person pronouns - they would just refer to things with demonstratives, or as "this one" or "that one", "the aforementioned", or (for people) "that girl", "that man", "the child", and so on, pretty much as you described. Though Latin still distinguished a third person on verbs, of course. In the case of verbs I would definitely second the suggestion to use zero-marking, since that's extremely common for third person referents anyways.

[clawgrip]

I usually go with head initial as I like to think of it in terms of you have to know what you are describing before you can describe it. Obviously, head final languages do just fine by putting modifiers before the noun they modify, and people understand those languages just fine, I just tend to go head initial as I see the head as the most important part of the phrase, and should therefore come first. I can't help you here much, but I haven't looked at head final as putting emphasis on the specifics (the modifiers of the noun) as the more important part.

I have heard more than one Japanese person say that in English we put the "conclusion" first. An interesting thought from someone who has the exact opposite perspective of yours, that is to say, the head means nothing without context.

[Vlürch]

From a conlang-building perspective, what are the advantages of having a head-initial or head-final choice for compound words, other than the obvious "head-initial has the general most important word first" and "head-final words allow jumping into specifics immediately", especially with regards to syntax/morphology?

I may be wrong, but IIRC I read somewhere that some Iranian language (maybe some variety of Kurdish?) has modifiers both before and after the noun, but I have no idea about the specifics or if I even remember correctly. From a conlanging perspective, that could be one solution to avoid having to choose either, although some logic would have to be invented for how it's decided.

One possibility would be to use modifiers to distinguish whether the thing is the object or subject of a sentence, so for example the noun could come before the adjective if it was the subject and after if it was the object or vice versa. You could have "pink unicorn eat shit" mean "the pink unicorn is eating shit" and "unicorn pink eat shit" meaning "the pink unicorn is being eaten by shit".

Another option would be to distinguish between integral qualities and acquired qualities; for example, a black person would be "black person" but a happy black person would be "black person happy". This would also enable social statements about sexuality or such, if that's something you want to touch, by making the integral quality postnominal and thus implying that it's actually not integral in the speaker's opinion, eg. gay black person would normally be "gay black person" but a person who thinks being gay is a choice would say "black person gay". I'll probably try to make a conlang with this kind of a thing, but won't get far since I never get far with any of my conlangs.

[holbuzvala]

I like consonant clusters, but I can't decide how I might include them in my conlang. You see, the grammatical baseline of the languages rests on biconsonental roots (like Semitic languages, but only 2 consonants in the roots). For instance:

"k-v" = things to do with writing
kava = to write
kolov = writer (the 'l' is not part of the root)
okvo = writing (abstraction)
ikev = a pen

Now, if I want word-initial consonant clusters, I see two choices.

Choice #1
Have the two 'roots' (called 'X' and 'Y') be capable of being lone consonants, or unplittable clusters, but with forms only of XVYV, XVCVY, VXYV, or VXVY, e.g.

"kz-m" = things to do with punching
kzama = to punch
kzolom = a boxer
okzmo = boxing/punching
ikzem = boxing glove

Choice #2
Having only single consonants in the roots, but allowing the forms to be things like XYV, and VXY in addition to XVYV, XVCVY, VXYV, and VXVY:
"b-l" = things to do with eating
bala = to eat
bolol = dinnerguest (someone who eats)
oblo = eating
ibel = plate
bla = food
ubl = tooth

What do you think?

[Chagen]

Is there any correspondence between the inflectional complexity of a language, and whether it tends towards ambitransitive verbs or distinct transitive and intransitive ones? For instance, English usually has ambitransitives: "I turn on the lights", "the lights turned on", whereas, e.g, Japanese is highly inflected and has distinct transitive-intransitive pairs: denki wo tsukeru "(I) turn on the lights", denki ga tsuku "the lights turn on"

You know I just realized that this can also be analyzed as a causative 90% of the time. I could easily envision in Pazmat vejj- "to wake up, turn on" > vījjay- "to awaken, turn (a device) on":
kumāya vījjayī "I turn on the lights"
kumā vejjū "the lights turned on"

kumā "lights" from kum- "to brighten up" is an mass noun and has no indefinite singular (though it does have an indefinite plural). To say "one light" you use the diminutive kūmasā

Because Pazmat has a distinct and very commonly used causative derivation, shouldn't it have intransitives for bare roots and derive transitives with the causative? For instance, cidh- means "to boil". Logically, you'd expect this to be the intransitive use ("the carrots boiled in the water") and the transitive would be the causative cūdhay-:
jriqā caudhivyū "the water boiled"
jriqāya cūdhēyavyī "I boiled the water"

On the other hand, Japanese has a well-used causative, but these transitivity pairs don't appear to use it. The causative of tsuku, for instance, would be tsukaseru/tsukasu, but its transitive pair is tsukeru, which looks exactly like a potential(?!) of all things!

[Ælfwine]

Sort of ambiguous but

What are some common ways to represent schwa in orthography?

[Sumelic]

Variations of <e>, probably. <e ə ë>. Less common things that seem possible: <a u i ı o eo ae ai eu . v q x c>, <’> or <'>, or just <> (nothing).

[Porphyrogenitos]

<y> is also used for schwa in a few orthographies. Notably Welsh (in some positions) and Lojban.

[Frislander]

Kwak'wala uses <a̱>. I've also seen <ă> (Romanian), <ä> (The romanisation of Ehtiopian Semitic languages and Menya: the former confusingly has /ɨ/ <ə> as well) and <ӕ> (Ossetian) used in natlangs.

Also, who's mad enough to use <q x c> as vowels? No one should be inflicted with that! And it's bad enough using <.> as a consonant (I'm looking at you, Tlingit and Lojban), let alone a vowel!

[clawgrip]

Because Pazmat has a distinct and very commonly used causative derivation, shouldn't it have intransitives for bare roots and derive transitives with the causative?

The existence of a commonly-used causative inflection doesn't imply that all base verb stems must therefore be intransitive. Japanese, as you point out, has a commonly-used causative, but both transitive and intransitive verbal roots are common.

On the other hand, Japanese has a well-used causative, but these transitivity pairs don't appear to use it.

(putting aside for now the suffix -e, which swaps transitivity in either direction)

Japanese transitive verbs derived from intransitive verbs do so by means of the suffix -(a)su, which is directly related to the causative suffix -(s)aseru. They both originate from the Old/Middle Japanese verb su, "to do", ancestor of modern suru.

Intransitive verbs derived from transitive verbs do so by means of the suffix -ar(er)u, which is directly related to the passive suffix -(r)areru. They both originate from the Old /Middle Japanese verb ari, "to be (somewhere) (locative copula)" ancestor of modern aru.

The majority of verb pairs will take a suffix on only one verb, while the other one will be the unmodified stem, though some have suffixes on both (and some have a completely unrelated -e that just confuses the issue, e.g.

Code: Select all

STEM TRNS END   STEM TRNS  END    TR      INTR
yak        -u   yak  -e    -ru    yaku    yakeru
yam  -e   -ru   yam         -u    yameru  yamu
at       -eru   at   -ar    -u    ateru   ataru
d    -as   -u   d         -eru    dasu    deru
kow  -as   -u   kow  -ar  -eru    kowasu  kowareru

In short, the causative is much more connected with transitive verbs than your post makes it out to be, and intransitive verbs are related to the passive.

The causative of tsuku, for instance, would be tsukaseru/tsukasu, but its transitive pair is tsukeru, which looks exactly like a potential(?!) of all things!

(first of all, -(s)aseru is not shortened to (s)asu except in sasete/sasete, where it colloquially becomes sashite/sashita; so tsukaseru cannot become tsukasu)

The potential is weird because in one verb class, it is formed with the transitivity-swapping suffix -e I mentioned above, while with the other verb class, it's formed with the passive suffix as-is (though colloquially, there is a potential suffix for this class that is an abbreviation of the passive (e.g. indicative taberu, passive or potential taberareru, colloquial potential tabereru, which allows for the distinction). This doesn't exactly come out of nowhere, since potential verbs are also passive (and thus intransitive) by default, because they generally require the thing that can be done to become the subject.

The passive suffix is also used as a polite suffix, with no other change in syntax, but that's a whole other story.

[eldin raigmore]

... I think nominal aspect, like gender, may usually be something a noun just has, not something about the noun one frequently changes. ...

I think I may have been wrong about that.
I'm looking at a container of corn starch that says "Great for thickening sauces and gravies".
I think "sauce" and "gravy" are usually mass-or-measure nouns when sitting at a meal at a table;
but they seem to be count-nouns in the table-of-contents of a cookbook.
So even in English, some nouns do change aspect frequently-ish.
There doesn't seem to be any morphology specifically to make that happen, though.
Apparently, if you pluralize it, it must be a count-noun; if you use an indefinite-ish quantifier-like determiner such as "some", it must be a mass-noun; otherwise speakers assume addressees can guess, and addressees assume if they guess wrong it doesn't really matter.

[Creyeditor]

There is actually some newer literature on the topic. Changing between mass and count interpretation of nouns is often called 'grinding' and 'packaging'. You might want to do a search on that. The frequency might depend on the kind of noun, fruits for example seem to be easily 'grinded' in English. Some languages have been argued to easily to it with all words, and I think people also have identified certain structures in some languages that do nothing but 'grinding' or 'packaging'.

[clawgrip]

I would say that here, they're switching from mass to count because they're referring to types of sauce, etc., rather than the substance itself. We can say that marinara sauce, for example, is one type of sauce, while Worcestershire sauce is one other type of sauce, both easily counted. When they say that cornstarch is great for thickening sauces, I don't think they mean several containers of an identical sauce; clearly it is types.