jal wrote: ↑Mon Jul 01, 2024 11:03 am
Isn't that used for cases where it's a passive form? For example "sang" is "sing.PST" because it is the past form of "sing" and it's difficult to clearly point out the specific past tense morpheme since it's formed by ablaut.
That's just semantics. "went" is still "go.PST", even though it's diachronically from a different root, like "is" is still "be.3s".
Yes, but that is a different form of "go", under the lemma "go". It fits into a specific grammatical paradigm and corresponds to other past tense forms of verbs.
What I'm talking about is simply a semantic modification because the glossing language lacks a single word, e.g. if we didn't have the word "pony", such a word in another language could be glossed as "small.horse". In this case, there is no need to reach for a grammatical category label such DIM because the word "small" is quite compact and "horse.DIM" would imply that it somehow corresponds to other diminutive words in some way (whether suppletively or not). In other cases, however, a gloss using regular words can just be too unwieldy. For example a gloss for "flock" as "group.of.birds" may be too long, but a label such as "COLL" fits better. Should that then be "bird.COLL"?
In any case, the Leipzig glossing rules
do in fact allow for both square brackets and parentheses:
https://www.eva.mpg.de/lingua/pdf/Glossing-Rules.pdf wrote:Rule 6: Non-overt elements
If the morpheme-by-morpheme gloss contains an element that does not correspond
to an overt element in the example, it can be enclosed in square brackets. An
obvious alternative is to include an overt "Ø" in the object-language text, which is
separated by a hyphen like an overt element.
Code: Select all
(22) Latin
puer or: puer-Ø
boy[NOM.SG] boy-NOM.SG
‘boy’ ‘boy’
Inherent, non-overt categories such as gender may be indicated in the gloss, but a
special boundary symbol, the round parenthesis, is used. E.g
Code: Select all
(23) Hunzib (van den Berg 1995:46)
oz#-di-g xõxe m-uq'e-r
boy-OBL-AD tree(G4) G4-bend-PRET
'Because of the boy the tree bent.'
(G4 = 4th gender, AD = adessive, PRET = preterite)
jal wrote: ↑Mon Jul 01, 2024 11:03 amIf there's a word meaning "be eaten" and a word meaning "eat", and the speakers of the language both associate it with the same action, they're definitely two sides of the same medal. One can be a suppletive form of another, or there can be two independent roots, but just because "went" is a different word from "go", you shouldn't analyse it as being two totally different words.
When did I say I have two equivalent, independent roots. It works like this.
- tém-ī
tém-SBJ.R5
'it is/gets eaten'
- múl-ā tém-ī
use.mouth-SBJ.R2 tém-SBJ.R5
'I eat it'
The agent in this sentence is the subject of a clause indicating the use of one's mouth. It is not the other side of the
tém medal because it is semantically much more general, also used for drinking, kissing and speech acts. It's functionally basically like an ergative marker except also indicating manner. Both of these words are valid clauses and this is also grammatical, although semantically fairly non-specific:
- múl-ā
use.mouth-SBJ.R2
'I use my mouth' (I eat/drink/say/speak/tell/kiss/...)
If more specificity is needed (specifically "eat") and only the agent of eating is to be mentioned, not the patient, the antipassive/causative infix <úw> can be used.
- t<úw>ém-ā
tém<ANTIP>-SBJ.R2
'I eat'
jal wrote: ↑Mon Jul 01, 2024 11:03 amAnd no language would have two completely different roots for
every verb in the language. That's just not how our mental model works (of course, when it's aliens, you can get away with everything, though I'd still find it highly unlikely).
Again, I never said there is anything like that. There is just a tendency for roots to equate meanings that are expressed passively in English. There aren't huge lists of underived roots with identical but opposite (in voice) meanings. It's more like how Slavic languages generally have an imperfective stem and then a derived perfective form. The imperfective form tends to be the underived form. Another language could tend to have underived perfectives and then derive imperfective forms from them. A lot of actions in Tobarese are simply named most basically with a form that describes what a patient does, akin to a passive in English, and is glossed as such whenever English lacks a corresponding underived verb ... it's not a passive as it is not inflected or derived from an active form. It's just an intransitive verb that describes the action of a semantic patient.
- mágaz-ī
fall.over-SBJ.R5
'it falls over'
- b-ā mágaz-ī
use.hand-SBJ.R2 fall.over-SBJ.R5
'I knock it over (by hand)'
I don't need to translate
mágaz as knock.over.PASS because there is a simple intransitive verb in English. We don't say that "fall over" is a passive verb. I am forced to use passive verbs in the gloss only that's how the glossing language fills these gaps.
Code: Select all
PATIENT ONLY: CAUSATIVE: TOBARESE PATIENT MARKING WORD:
1. fall over knock over mágaz fall.over
2. die kill yóg die
3. ???????? eat tém get.eaten/eat.PASS
4. fall asleep put to sleep l<iy>óuh sleep<INCEP>
5. be visible show šév be.visible
6. shatter shatter gávag shatter(INTR) ??
The necessity of using a passive translation is simply because the only way this meaning is expressed in English is through a passive as there is no simple, underived intransitive word in English meaning "be eaten".
Also, the huge number of labile (ergative) verbs in English means that sometimes the transitivity of the English gloss is necessary and I usually do this in parentheses, as in "shatter" above. I do this because it is modifying the interpretation of the English word, not to indicate that the root
gávag is intransitive. (That would be redundant as all roots are intransitive in Tobarese.) That's all I'm doing with "eat(PASS)". If it should be "eat.PASS" then it should probably also be "shatter.INTR" although, again, that is only necessary to modify the meaning of the English gloss. All Tobarese verbs are intransitive, so the "INTR" label is not about any particular characteristic of
gávag, just that the English gloss word is ambiguous.
Travis B. wrote: ↑Mon Jul 01, 2024 12:38 pmD'oxac'aqa "increasingly" wiib'adaa ha leecamit'a wiib'adaa.
suggest-LOC-AGT.3.S.INAN "increasingly" change-ACTION COMP see-AGT.1.S-NEG change-ACTION
"Increasingly" implies a change while I see no change.
Zuwíyaukī. Hiázzuwiyádziyaukī?
It has changed. Has it stopped changing?
Hiawwíŋiatsáxoqíŋgilišizzuwíyauk?
Or is it other varieties of English that have changed?
