Re: Evidentiality drift (WAS: Pronouns & sexuality)

Sai Emrys wrote:
> On Thu, Feb 26, 2009 at 6:46 PM, Herman Miller <hmiller@...> wrote:
>> Back to the framing idea though: if you've got a language like Tirelat,
>> where you can't express tense without evidentiality, it's pretty much clear
>> from the language what each side would consider to be reliable facts as
>> opposed to opinion or hearsay. If you have a conflict of opinions, no big
>> deal, but if facts are in dispute, you won't be considered credible unless
>> you can back them up. Ultimately, I'm not sure that would be of much help
>> though, because the meanings of the words would still differ.
>
> I wonder: would such use of evidentials survive over time?
>
> Surely politics (and the more subtle interpersonal sorts) would erase
> these distinctions; simply put, it's useful to misrepresent your
> evidence, or to state your opinions as facts. So evidentials would
> drift to become a matter of simple emphasis or stress. (in the same
> way that people now use "x is literally y" to mean "x is figuratively
> but emphatically y").
>
> Do any of you know of how a natlang with evidentials has handled this IRL?
>
> - Sai
Well, there isn't a specific "fact" evidential in Tirelat, but certain
evidentials are more likely to be used in cases where you believe
something to be a fact. It's likely that the range of evidentials in
Tirelat is unusually large. Some examples of the word "bit" in different
contexts with varying degrees of credibility:

A dog bit me. jĕtaċaŕin
A dog bit him/her; I saw it happen. jĕtaċatin
It's obvious that a dog bit him/her. jĕtaċamin
I heard that a dog bit him/her. jĕtaċálin
I think maybe a dog bit him/her. jĕtaċaxan

In each case you've got a potential opening for argument if the fact
that someone was bitten by a dog is in dispute. It's harder to dispute
"a dog bit me" except maybe to ask "are you sure it was a dog?" (unless
you want to suggest that the other person is outright lying). But in
political situations, this sort of personal experience could be
dismissed as anecdotal.

Tirelat is unusually regular; after all it was originally a personal
language, and only later did I adopt it as a fictional language for the
Sangari people. I've been going back to considering it more of a
personal language, or a model for a typical Sangari language without
many of the irregularities of real languages. It's possible that
languages related to Tirelat (or future descendant languages) could have
a simplified version of this scheme. It could also happen that extra
shades of meaning are distinguished in some languages.