Sew'Kyetuh wrote:If there are no verbs to begin with, how do you categorize the MSA?
The understanding I got from some people, and the conclusion I was leaning towards, was that MSA doesn't apply to Etihus because Etihus can have arguments that omit verbs completely, and it can have arguments within a single word. A single word usually cannot be both a noun, and a verb, and an object all-in-one.
To follow through with your analogy of handedness, wouldn't it be more like, "Ok, a person has no hands. Are they left handed, right handed, or ambidextrous?
Ah, but Etihus
definitely has verbs. The fact that it's possible for a statement to not include a verb is very much different than "not having verbs to begin with." In terms of handedness, Etihus is not a person without hands, just a person with hands that sometimes—say, when playing soccer—doesn't use them.
English is classified as a nom-acc language, even though there are plenty of valid statements in English that don't contain a verb. But we don't look at those statements when determining MSA, because those statements are not what MSA is about. Instead we look at transitive clauses and intransitive clauses, of which there are plenty, and compare them. We can do the same for Etihus, as I've shown.
A language in which "there are no verbs to begin with" would be very tricky to pull off. It's an experiment that conlangers attempt every now and then, though slightly less frequently than trying to create a language in which there are no nouns. The reason it's so hard to do is that for it to really work, the language has to behave in such a way that
there is no reasonable analysis of it where verbs are a meaningful concept. But that's not what usually happens.
What usually happens is something like:
A: I have created a language with no verbs!
B:
(Looking at data from the language) Hm, doesn't look like it. These things here, here, and here all behave like verbs.
A: But they're not verbs. They're adjectives that mean "doing X".
B: You say that, but still, they act like verbs, they do what verbs do. There's no reason not to call them verbs, except that you want to call them adjectives.
This is an important thing to understand, and I think it's what cntrational was getting at when he said "you confuse how you
think it works and how it
actually works".
A conlanger's supreme prerogative is to say what sentences are and are not grammatically correct in his or her conlang, in whatever circumstances. But there is a point at which that jurisdiction ends, which is analysis. If someone else can look at a conlang and show, with well-supported observations of the available data, that a different analysis makes more sense, the creator's analysis doesn't trump by virtue of being the creator's. The claims you make about how your language works are true only so far as they are the best account anybody can find for what the language is doing.
Here's an example of how this can play out. For background, a
clitic is "a morpheme that has syntactic characteristics of a word, but depends phonologically on another word or phrase". The possessive -'s in English is a clitic.
= = = = =
A: I've been working on how my language handles location. There is a collection of clitics related to location that can be applied to a noun.
1.
gafanta eosi.
forest=DAT go.1-PAST
I went to the forest.
2.
gafansu eosi.
forest=VIA go.1-PAST
I went through the forest.
B: Those just look like case suffixes. Why do you call them clitics?
A: They're syntactically independent; they go at the end of the whole noun phrase, not on the end of the noun. Like so:
3.
gafan akasta eosi.
forest big=DAT go.1-PAST
I went to the big forest.
4*.
gafanta akas eosi. (<--the * means "this sentence is not grammatical")
forest=DAT big go.1-PAST
B: Ah, that's definitely not a suffix then. But it might still be a free particle rather than a clitic.
A: They're phonologically dependent on the preceding word. You can tell because they undergo the same word-internal phonetic changes that suffixes do. When a suffix starting with a consonant gets added to a word ending in a vowel, the consonant
lenits:
5a.
gafan
forest
5b.
gafanti
forest-1.POSS
my forest
6a.
sata
house
6b.
satathi (<-- -ti becomes -thi because sata ends in a vowel)
house-1.POSS
my house
7.
satatha eosi (<-- -ta becomes -tha)
house=DAT go.1-PAST
I went to the house
8*.
satata eosi
house=DAT go.1-PAST
So you can tell that -ta and -su are syntactically independent, because they get suffixed to whole phrases, but you can tell that they're phonologically dependent because they undergo the same lenition process as suffixes. Therefore, it makes the most sense to call them clitics.
= = = = =
Notice how this kind of discussion is very heavy on examples and comparisons, and doesn't rely at all on interpretations or "But it literally translates as..." arguments. In the end, you can have confidence that what you say about the language is true not just because you say so, but because
it really is the best description of how the language works.
If you're interested, a few years ago I drew up a sketch of a language that was intended to "not have verbs to begin with"; you can see the discussion of it
here. Note how the whole conversation revolved around analysis of examples—I just let the language speak for itself, and eventually another conlanger determined that the best analysis he could find was in fact that the language had no verbs. (Incidentally, I think that even this would not have meant that MSA "didn't apply" to the language; you could probably find close-enough analogues to transitive and intransitive clauses to make a case for a particular alignment.)
In other news, did my explanation of phonemes make sense?