Evidentiality vs modality?

[Maulrus]

So I understand the basic ideas behind both, and I'm told that they are completely distinct. However, in trying to incorporate them into my baby conlang, I''m finding myself confused about the distinction. Doesn't a mood reflect the speaker's attitude towards what is being said (please correct that if it's wrong)? If that is the case, why isn't evidentiality sort of a subclass under modality?

[zompist]

Speaking very generally, I'd say they overlap, but they deal with different questions.

Mood deals with whether the state of affairs is real or less so. There are different types and degrees of 'unreality'-- doubt, futurity, hypotheticality, conditionality, etc.

Evidentiality deals with what the source of one's statement is-- e.g. direct knowledge, hearsay, deduction, etc.

There's an overlap in the area of probability, which can be handled by either category. E.g. we can express it in English using a modal like "may", while in Quehua it's an evidential -cha.

[Maulrus]

Okay, that makes more sense. So if, for example, I was to have one evidential which always applied to storytelling, then it is likely that every single verb in the story would be inflected for that evidentiality, because the whole story comes from the same source. A verb within the story could be indicative or optative as the situation requires; the evidentiality doesn't need to change. I think I've worked out how to handle this now, thanks!

[Aurora Rossa]

So if, for example, I was to have one evidential which always applied to storytelling, then it is likely that every single verb in the story would be inflected for that evidentiality, because the whole story comes from the same source.

Ah, but must it? One reason that I have not experimented much with evidentials is the feeling that having to mark them on every verb would prove highly repetitive.

[Maulrus]

So if, for example, I was to have one evidential which always applied to storytelling, then it is likely that every single verb in the story would be inflected for that evidentiality, because the whole story comes from the same source.

Ah, but must it? One reason that I have not experimented much with evidentials is the feeling that having to mark them on every verb would prove highly repetitive.

Hm, that's true. For the sake of economy I may make evidentiality more optional than I was intending and just force the speakers to rely a bit more on context, but if I may ask, is it at least reasonable (with respect to natlangs of course) to repetitively mark evidentiality on each verb? This language is purely agglutinative, so it seems that it would fit here better than anywhere else, but I can't pretend that I really know.

[zompist]

I just checked a Quechua story which I was creating a full glossed translation for, and yep, pretty much every line contains an evidential. (An exception is direct quotes within the text.)

(It's possible other texts are different, but this was evidence I had at hand.)

[Maulrus]

I just checked a Quechua story which I was creating a full glossed translation for, and yep, pretty much every line contains an evidential. (An exception is direct quotes within the text.)

Yeah, this is the sort of system I was aiming for. I expect I'll keep it this way for the sake of complexity, as my ultimate goal is running sound changes on this and watching my inflectional system completely crumble.

[Vuvuzela]

Okay, that makes more sense. So if, for example, I was to have one evidential which always applied to storytelling, then it is likely that every single verb in the story would be inflected for that evidentiality, because the whole story comes from the same source. A verb within the story could be indicative or optative as the situation requires; the evidentiality doesn't need to change. I think I've worked out how to handle this now, thanks!

I think of the idea that you have to mark every evidential on every verb in a story is not more weird than the fact that in English, we have to mark tense on every verb in a story. It's also interesting to think of how a culture views evidentiality within a story, for instance, a story first person reportative would be incredibly weird, but in third person you'd have a choice. Maybe some novelists in your conculture like to create a very epic feel, as though the story has been passed down through the generations, and so use the third person reportative, while some like to experiment with an unreliable narrator who constantly reminds us they're unreliable by their use of the inferrential.
Of course, if your culture is a tribe of nomadic herders like mine is, you might want to have a standard evidential for storytelling (my people use the third person reportative for stories, in case you're wondering)

[Maulrus]

Okay, that makes more sense. So if, for example, I was to have one evidential which always applied to storytelling, then it is likely that every single verb in the story would be inflected for that evidentiality, because the whole story comes from the same source. A verb within the story could be indicative or optative as the situation requires; the evidentiality doesn't need to change. I think I've worked out how to handle this now, thanks!

I think of the idea that you have to mark every evidential on every verb in a story is not more weird than the fact that in English, we have to mark tense on every verb in a story. It's also interesting to think of how a culture views evidentiality within a story, for instance, a story first person reportative would be incredibly weird, but in third person you'd have a choice. Maybe some novelists in your conculture like to create a very epic feel, as though the story has been passed down through the generations, and so use the third person reportative, while some like to experiment with an unreliable narrator who constantly reminds us they're unreliable by their use of the inferrential.
Of course, if your culture is a tribe of nomadic herders like mine is, you might want to have a standard evidential for storytelling (my people use the third person reportative for stories, in case you're wondering)

I was tentatively planning on dividing it into witness, quotative (which forms a hearsay when used with the dubitative mood), inferential, and a fourth evidential mainly relegated to the telling of traditional stories; is that last one plausible? I'm considering just dropping it and using the quotative with a specific past tense, but if there's any plausible way such for an evidential to develop I'd like to use it.

(By the way, the idea of using evidentials to create an unreliable narrator is fascinating. I don't have much planned out in the way of a conculture yet but that is something I will definitely be thinking about.)

[Vuvuzela]

I was tentatively planning on dividing it into witness, quotative (which forms a hearsay when used with the dubitative mood), inferential, and a fourth evidential mainly relegated to the telling of traditional stories; is that last one plausible? I'm considering just dropping it and using the quotative with a specific past tense, but if there's any plausible way such for an evidential to develop I'd like to use it.

Hold on, let me get my diachronics hat on.Stories are traditionally told with a hearsay evidential, to emphasize how long they've been passed down. But, in a conversation, it's usually considered more polite to use a quotative evidential. Eventually, hearsay drops out of use, in the same way "thee/thou" did in English or "vosotros/vosotras" did in Latin American Spanish, being replaced by the politer form. Literature, oral or written, from before a certain time keeps the normally-obsolete hearsay evidential, as do stories which try to imitate these, and as soon as everybody forgets it was ever a hearsay evidential,you have the system you're proposing.
Does that work? It sounds sort of plausible, at the very least.

[Vardelm]

Ah, but must it? One reason that I have not experimented much with evidentials is the feeling that having to mark them on every verb would prove highly repetitive.

I think of the idea that you have to mark every evidential on every verb in a story is not more weird than the fact that in English, we have to mark tense on every verb in a story.

What Vuv... said.

Eddy/Jabe.../whatever, you're making a mountain out of an anthill. (Anyone surprised?) Think about the number of plurals in English with -s/-es, coupled with -'s for possessives. I've heard people with other L1s say that ends up sounding very repetitive in English. English also uses -ed & -ing a lot, not to mantion is/was/were/am, etc. I would guess every language is going to have lots of repetition. To native speakers, it will simply be "invisible".

[Maulrus]

I was tentatively planning on dividing it into witness, quotative (which forms a hearsay when used with the dubitative mood), inferential, and a fourth evidential mainly relegated to the telling of traditional stories; is that last one plausible? I'm considering just dropping it and using the quotative with a specific past tense, but if there's any plausible way such for an evidential to develop I'd like to use it.

Hold on, let me get my diachronics hat on.Stories are traditionally told with a hearsay evidential, to emphasize how long they've been passed down. But, in a conversation, it's usually considered more polite to use a quotative evidential. Eventually, hearsay drops out of use, in the same way "thee/thou" did in English or "vosotros/vosotras" did in Latin American Spanish, being replaced by the politer form. Literature, oral or written, from before a certain time keeps the normally-obsolete hearsay evidential, as do stories which try to imitate these, and as soon as everybody forgets it was ever a hearsay evidential,you have the system you're proposing.
Does that work? It sounds sort of plausible, at the very least.

Sounds good to me! Thanks for all the suggestions.

[Nortaneous]

So if, for example, I was to have one evidential which always applied to storytelling, then it is likely that every single verb in the story would be inflected for that evidentiality, because the whole story comes from the same source.

Ah, but must it? One reason that I have not experimented much with evidentials is the feeling that having to mark them on every verb would prove highly repetitive.

Not if you fuse evidentiality marking with something else, it doesn't. Or you could have something like the bound(?) forms in some languages of New Guinea, where only the first verb in a sentence takes the full amount of marking, and the rest get a reduced form, so something like:

ass-barf-th-p fuck shit ass, i-piss shitty fuck cock ass dicks, i-fuck dong hell satan
SOME_EVIDENTIALITY-verb-3S-1S nouns nouns nouns, REDUCED_FORM-verb nouns nouns nouns nouns, REDUCED_FORM-verb nouns nouns nouns

edit: what better use of my 8000th post than "assbarfthp fuck shit ass"

[zompist]

I'm reading a book on modality right now, so I wanted to amend my comment above. The author (Palmer) talks about evidentiality as a subtype of modality. One division of modality is propositional, and that in turn can be divided into epistemic (relating to factual status) and evidentials (relating to source of knowledge). Though it's a bit messy, there seems to be a tendency to pick one of these two paths.

[clawgrip]

I'm reading a book on modality right now, so I wanted to amend my comment above. The author (Palmer) talks about evidentiality as a subtype of modality. One division of modality is propositional, and that in turn can be divided into epistemic (relating to factual status) and evidentials (relating to source of knowledge). Though it's a bit messy, there seems to be a tendency to pick one of these two paths.

Thank you for those two terms. I'm pretty sure epistemic/evidential is the primary division between will and be going to, as discussed in the other thread.

[Imralu]

Thank you for those two terms. I'm pretty sure epistemic/evidential is the primary division between will and be going to, as discussed in the other thread.

What other thread? That's IMPOSSIBLE to search for. Link please.

[clawgrip]

Particularly the second page of this thread.

[chris_notts]

I just checked a Quechua story which I was creating a full glossed translation for, and yep, pretty much every line contains an evidential. (An exception is direct quotes within the text.)

(It's possible other texts are different, but this was evidence I had at hand.)

At least some Quechua dialects have clause chaining, don't they? If so, is it just the last verb in the chain that has an evidentiality clitic floating around, or is it every clause? I know that in some languages with clause chaining on the most finite clause (usually final) gets the marking. Presumably if a switch in evidentiality value is highly significant then in these languages you either use some other lexical means or you start a new clause chain.

[zompist]

The story I was translating was in Cuzco dialect, which doesn't have any clause chaining (indeed, it tries to avoid subclauses entirely).