In languages that mark evidentiality morphologically, what evidential categories tend to be used for things like creation stories, myths, statements about theology ("God is good"), etc.?

It would seem that "direct experience" evidentials (visual or other) would most likely not be used for such statements, but on the other hand, "inferential" or "hearsay"-type evidentials might not be appropriate either insofar as they imply lesser confidence, at least in cultures where such things are supposed to be known with certainty.

This must be a problem that missionaries come up against all the time: what do SIL linguists do when they have to translate "In the beginning God created the heavens and the earth" into a language that has evidentials?

I realize the answer to this question will be that different languages do different things, and that it will vary based on what evidential categories the language marks -- I'm interested in specific examples.

  • I imagine that this would vary from language to language. Jan 8, 2014 at 4:39
  • 3
    For spoken Tibetan, the main evidential distinction seems to be between things one has personally observed and things that form part of the speaker’s general knowledge. Religious truths would presumably fall into the second category. And Literary and Classical Tibetan don’t make this distinction, so it wouldn’t be an issue in the Buddhist literature extant in Tibetan. (This is based on Tournadre’s textbook, p. 110 ff.)
    – neubau
    Jan 10, 2014 at 3:03
  • 2
    In this answer to an unrelated question, @Draconis says "[My professor] joked that the New Testament in Lingála spoils the twist way too early, because when Jesus dies, that [ultimate tense-aspect] marking is missing!" Obviously irreversible aspect isn't the same thing as evidentiality, but it is an example of how different tense-aspect-voice-etc. systems can affect translation of religion/mythology.
    – abarnert
    Jan 17, 2019 at 23:50

1 Answer 1


So here are some examples from Ecuadorian Quechua.

First of all, Aikhenvald (2004: 43) classes the Quechua language family as having a B1 Evidential system, meaning there are three evidential distinctions: direct (visual), reported, and inferred. Direct, represented by -mi is used to indicate that the speaker was an eyewitness to the event specified. Reported, or indirect -shi is taken to mean hearsay or secondhand information, or that the speaker is removed from the source of the information. The final evidential, the inferential -cha, can invoke a sense of any of the following: probability, doubt, and uncertainty. This is usually translated as “probably” or “perhaps”.

So here: https://www.bible.com/es/bible/98/JHN.1.MTDS are 2 versions of the Bible in Ecuadorian Quechua (Kichwa/Quichua). In John 1 which begins with the "In the beginning there was the word, and the word was with God..."

In all three versions the segment 'the word was with God' the constituent 'with god' is marked with the direct experience evidential -mi:

Rimashca Shutica Taita Dioshuanmi carca

Chai Shimica, Taita Dios-huanmi carca

LIT. "The word father god-with was"

Perhaps the translators chose the direct evidential because they think of the book being written by John, and therefore from John's perspective and since John is thought of by most Christians as a saint or apostle or whatever, he presumably would have 'directly witnessed' such events through visions or some other kind of divine manifestation, and thus could use the "eyewitness" evidential.

However, I think it's important to point out that while most native Quechua speakers today are Christians (Catholic or Protestant), the concepts of Abrahamic religions are still imported concepts. Pre-Colombian South Americans (whose languages developed these evidential categories) didn't have the same ideas about religion and philosophy that have since been imposed on them.

I work with some Quechua dialects in the Amazon in Ecuador, most Quechua speakers there are Christians but they still have a lot of beliefs and ideas about things from pre-colombian times. And the idea of having a consistent "canon" isn't necessary. You'll find 'amazonianized' versions of bible stories like Adam and Eve and Noah's ark and they often conflict with each other in terms of 'canon' but that's not important. Anyway the overall point here is that the system of evidentiality they have may not be the best fit for western ideas about religion, evidence, and canon.

For example, looking a traditional non-Christian myth, we see that the reportative evidential -shi is used quite a lot and the direct evidential -mi only occurs in quotations as speech from specific characters in the narrative. These examples come from a 'beginning times' narrative about a woman who never died but regenerated her body every time she became old by bathing with a certain root. The narrator uses the reportative evidential -shi to talk about the events in the story.

Alli warmishi ag ashkara

LIT. "good woman be-r was being"

"She had been a pretty young woman"

Chasnashi akchata armasha waktarishashi shamushkara shullata anchu-chi-gri-sha

LIT. "like that hair washing hitting she came drops shaking"

"Having washed her hair, she was shaking the water from it as she headed for home."

The site seems to be down at the moment, but you can create a free account at AILLA and access the Janis B. Nuckolls Quechua Collection and download the whole glossed and translated text which is titled: How people came to grow old and die - Como gente vinieron a ponerse viejo y morirse (QVZ003R001I001.pdf). (Content advisory: the text contains incest and rape).

Quechua Collection of Janis B. Nuckolls. The Archive of the Indigenous Languages of Latin America, ailla.utexas.org. Access: public. PID ailla:119502

I'd also recommend Nuckoll's 2014 paper on Quechua evidentials: 'From quotative other to quotative self: Evidential usage in Pastaza Quichua' which breaks down what the more specific functions of the direct evidential -mi is. She departs somewhat from Aikhenvald's descriptions, the direct evidential need not be from an eye witness perspective only, modality and perspective are the more salient elements in defining evidential usage in that language. A speaker may tell someone else’s story and though they were not an eyewitness to the events they may still use -mi in their role as the narrator, thus possessing personal knowledge.

So I think whether the translator intended that John was an eyewitness or not to the beginning time events, it may not be unprecedented to use the direct evidential. Or perhaps it wouldn't be natural to do so at all on the part of the translator.

  • Very interesting, thanks. (Btw, does -mi also function as a focus marker? I seem to have a vague memory that it can, but it's been years since I looked at Quechua.)
    – TKR
    Jan 15, 2019 at 17:59
  • I think it may be more than just that John is a saint or apostle. Also, you suggest visions, but I don't know if that's relevant. The only evidence that John had revelatory visions is in the Book of Revelation—and even most people who subscribe to traditional authorship (i.e., that the author of John and the Johannine Epistles is John the Apostle) believe the author of Revelation was a different man.
    – abarnert
    Jan 16, 2019 at 21:06
  • Another thing worth considering is the idea that all of the books accepted into the Bible were inspired by the Holy Spirit. In some traditions (especially within fundamentalist evangelical Protestantism) this is taken to mean that the Holy Spirit literally wrote the gospel through John. And presumably the Holy Spirit did directly witness those events. (Some of those people even believe that any proper translation of the Bible is inspired by the Holy Spirit—if that's the case, presumably it would be inappropriate of the translator to change the evidentiality.)
    – abarnert
    Jan 16, 2019 at 21:09
  • Meanwhile, I think the idea toward the end of your post (Nuckolls) is interesting. It seems at least plausible, from what you quote, that narration uses the evidentiality of the implied narrative PoV, not of the actual author. And John is written in third-person-omniscient PoV.
    – abarnert
    Jan 16, 2019 at 21:17
  • 1
    @abarnert True, I was speculating, it's hard to know exactly what exactly they subscribe to in terms of John's perspective on such events. Or perhaps they would avoid using a reportative evidential altogether for such religious events. But I don't necessarily think a native speaker would be worried about that. Aikhenvald states that despite the fact that the category is called 'evidential' it doesn't necessarily connote 'evidence' like we might conceive of in western law. Thus using a reportative evidential wouldn't necessarily show skepticisim on the part of the narrator.
    – Wangana
    Jan 17, 2019 at 23:24

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service and acknowledge you have read our privacy policy.

Not the answer you're looking for? Browse other questions tagged or ask your own question.