Talk:Evidentiality

Cleanup
Should the link under Epistemic Modality go to https://en.wikipedia.org/wiki/Epistemic_modality, the page on epistemic modality as the term is used in linguistics, rather than https://en.wikipedia.org/wiki/Epistemology, an important and interesting philosophical topic but not as directly pertinent to the discussion? 128.214.149.151 (talk) 09:24, 14 February 2018 (UTC)

After reading the intro section, I had no idea what evidentiality is. I understood (very well, thank you) that there are several ways that someone could have evidence of an event, but I was still in the dark about how this relates to linguistics. And, for all that, the intro is too long. On a cursory look through the rest of the article, I noticed that it talks about a bewildering array of forms of evidentiality, but doesn't seem to explain how they all relate to each other. --Smack (talk) 05:34, 8 March 2006 (UTC)


 * hi. Thank you for the comments — these are what i wanted.


 * the main point of the article is to demonstrate how the several ways that someone could have evidence of an event may be indicated through language. You could simply say where you got the info (like "Smithy told me that ..."). Some languages use adverbial-type words or phrases to express the evidence. However, more interestingly, some languages have an inflectional category that is used to indicate the evidence. These languages must, for instance, have grammatical suffixes/prefixes that must occur on verbs, just like some languages require grammatical elements that indicate tense, aspect, person, subject/object agreement, gender, mood, and/or number (among other things). So, you must conjugate verbs according to evidentiality in these languages.


 * all of the bewildering array of forms are related in that they all have to do with evidentiality. I thought this would be understood. Do you think not?


 * i lengthened the intro following a suggestion from a fellow editor. Perhaps it should be shortened.


 * does this help make this clear? If so, how can we help to make the article clearer & more accessible using what we understand here?


 * by the way, i read linguistics stuff almost everyday so, i may not be the best judge of how accessible these technical topics are to a general reader.


 * – ishwar  (speak)  02:41, 21 March 2006 (UTC)


 * Sorry if I didn't make my point clear, but I think you missed it. Have a look at the changes I just made.  Also, please avoid using HTML formatting features to turn wiki into a word processor.  If you like to play with cool features, check out How to edit and related pages.  Also, it would be really nice if someone could convert the footnotes to the Footnotes format. --Smack (talk) 04:35, 21 March 2006 (UTC)
 * P.S. I guess those "footnotes" aren't actually footnotes. Maybe you would be better off using the Footnote3 format. --Smack (talk) 04:39, 21 March 2006 (UTC)
 * I readded the table formatting, as it also served to make the table readable in browsers that are not fully Unicode-compliant. &mdash; mark &#9998; 14:57, 21 March 2006 (UTC)

Cleaup removed
I just removed the cleanup tag. As I take it, the article does a good job at explaining and detailing the complex linguistic subject of evidentiality. As any article, it could be improved still, but it is definitely not in the league of articles needing cleanup. &mdash; mark &#9998; 10:18, 15 May 2006 (UTC)

Comments about Turkish
As a native Turkish speaker I must disagree with the article's analysis of evidentiality in Turkish language. I don't think the -di suffix is unmarked. Most Turkish speakers will agree that geldi means "he/she/it came, and I actually saw that". There is no easy way to be vague about evidentiality in Turkish. Cyco130 (talk) 14:20, 4 May 2008 (UTC)


 * Thanks. Would gelmiş then be the unmarked past tense form? Or are you saying neither is unmarked? (Also, could you use geldi if you heard the person come in, but didn't see them, or if you saw their shoes by the door, but didn't see or hear them?) kwami (talk) 17:43, 4 May 2008 (UTC)


 * Hi. I dont know Turkish or the literature on Turkic indirectivity, but I note that what you mention does dispute other analyses. The analysis in the article is that of Lars Johanson (www.turkiclanguages.com/www/LarsJohanson.html). Perhaps, you are not distinguishing between evidentiality and epistemic modality? Take a look this from Johanson's 2003 article (which explicitly disagrees with what you say above):


 * "Turkic languages display basic contrasts between marked indirectives and their unmarked counterparts. Functionally marked terms, expressing the evidential notions explicitly, stand in paradigmatic contrast to non-evidentials. Thus, Turkish exhibits items signalling indirectivity, e.g. gel-miş [come- IPAST ‘has obviously come/obviously came’ and gel-iyor-muş [come- INTRA-IPAST ‘is/was obviously coming, obviously comes’. It has corresponding unmarked items such as gel-di [come- DPAST ] ‘has come/came’, gel-iyor [come- INTRA ] ‘is coming/comes’.


 * "Although the relations between marked and unmarked terms vary across languages, the unmarked ones always exhibit neutral uses in cases where the speaker considers the evidential distinction unessential and thus chooses not to use it. The widespread opinion that unmarked items such as gel-di ‘has come/came’ consistently signal ‘direct experience’ or ‘visual evidence’ is incorrect. Unmarked items simply do not signal that the event is stated in an indirect way, i.e. acknowledged by a recipient by means of report, inference or perception."




 * "Turkic indirectives may also have epistemic connotations in the sense of reservations about the validity of the event as a fact. The indirect way of referring may create uncertainty concerning the realisation of the event and be interpreted as non-testimonial reference. Indirectives, in particular reportive items, can be used to disclaim direct responsibility for the truth of the statement, suggesting that the speaker is not the originator of the information or does not vouch for it. By contrast, unmarked terms may suggest that the speaker is certain of the truth of information and even responsible for it.


 * "‘Supposition’ is sometimes claimed to be the main meaning of Turkic indirectives. The corresponding unmarked items are said to signal that the speaker regards the event as certain. However, indirectives are not presumptives or dubitatives reducing the factuality of the statement. Their task is not to express the speaker's attitude to the truth of the content, to signal doubt or conjecture concerning the information conveyed...."


 * Anyway, I wonder where the "widespread opinion" (which is your disagreement) comes from and why? There are obviously disagreements here, but I havent read the arguments. There seem to be several articles on Turkish so somebody could do the research to figure out where the disagreements stem from. – ishwar  (speak)  19:35, 4 May 2008 (UTC)


 * Hi. To kwami: I'm saying that neither is unmarked. There are cases where -di is used for unseen acts, such as when one is talking about historical facts, but that is really an exceptional case. When you see the shoes, you should say "gelmiş" unless you saw the person come, people will think you actually saw the person if you say "geldi". To ishwar: My problem is not evidentiality vs. epistemic modality. I really disagree with the second quote: "[...] unmarked items such as gel-di ‘has come/came’ consistently signal ‘direct experience’ or ‘visual evidence’ is incorrect". I think it is correct. I will try to find more support for my opinion. My disagreement comes from my scientifically small sample of 4 people (me and 3 roommates, :)). I remember Türkiye Türkçesi Grameri (Turkish Grammar) by Zeynep Korkmaz was saying basically the same thing. I will post the relevant passages as soon as I have access to the book (probably tomorrow). Cyco130 (talk) 03:22, 5 May 2008 (UTC)
 * Hi again, it turns out that I still have the book (ISBN: 975-16-1643-3). Unfortunately there isn't any semantic analysis on these tenses, it's more about morphology. But it does call -di "Görülen Geçmiş Zaman" (Seen Past Tense) and -miş "Duyulan Geçmiş Zaman" (Heard Past Tense). I will need more sources... Cyco130 (talk) 03:37, 5 May 2008 (UTC)


 * So, if there is an exception with historical facts that are not known from direct experience then this is probably why Johanson is saying that they are unmarked. Johanson gives the following examples in a section on "Testimony, involvement, control" (from the same 2003 article):


 * Çok büyü-dü-n
 * much grow-PAST-2.SG
 * "You have become very big"


 * Kemal Paşa, Selânik’te doğdu
 * Kemel.Atatürk pasha, Salonika.LOCATIVE be.born.PAST
 * "Atatürk was born in Salonika"


 * Johanson says these two sentences with the unmarked past can be used when the speaker has direct experience (from participating in the event, witnessing the event, being in control of the event). This is as you suggest (that is, the speaker actually saw the person grow big or actually saw Atatürk being born in Salonika). But, Johanson also says that these sentences can be used for events that the speaker did not witness, did not control, or has just inferred (based on some other evidence). Thus, the sentences with the unmarked past "just lack the two-layered information typical of indirectives, and may thus be used whenever this specific information seems unessential". When the unmarked past is used to signal (or, perhaps, better worded, imply) that the speaker has direct knowledge, the past suffix "does not signal these meanings explicitly".


 * The point you bring up (that direct experience is a common (the most usual?) interpretation of the past) is interesting because Johanson says that in other Turkic languages (like Uyghur, Uzbek, Kazakh, Turkmen) the indirect past is in contrast with a direct past. That is, while Turkish as an indirect-past/unmarked-past opposition, other langs have an indirect/direct opposition. And, Aikhenvald (who is the editor of the book that Johanson's article is in) notes that it is a common development for an unmarked evidential to turn into a direct evidential in languages like these with a two-way opposition between unmarked/indirect. We also see this in Balkan Slavic which has a direct/indirect contrast (their terminology is confirmative/nonconfirmative, Aikhenvald's terminology is firsthand/non-firsthand) and Bulgarian which has an unmarked/indirect (in Aikhenvald's terms everything-else/non-firsthand) contrast. The Balkan Slavic pattern is a new development (through language change).


 * Aikhenvald also notes that how to analyze languages with a neutral term is controversial. She lists the following people who have noted that Turkish -dI tends to suggest a direct experience interpretation:


 * Grunina, E. A. (1976). K istorii semanticheskogo razvitija perfekta -miş. Sovetskaja tjurkologija, 7, 12-16.
 * Aksu-Koç, Ayhan. (2000). Some aspects of the acquisition of evidentials in Turkish. In L. Johanson & B. Utas (Eds.), Evidentials: Turkic, Iranian and neighboring languages (pp. 15-28).
 * Aksu-Koç, Ayhan; & Slobin, Dan I. (1986). A psychological account of the development and use of evidentials in Turkish. In W. Chafe & J. Nichols (Eds.), Evidentiality: The linguistic coding of epistemology (pp. 159-167).


 * Aikhenvald suggests that the change from unmarked to direct "is best viewed as a continuum .... evidentiality-neutral forms ‘drift’ towards acquiring the meaning complementary to their ‘non-firsthand counterparts’ and thus gradually become associated with ‘firsthand’ information. Some may even be interpreted by linguists as a variety of [a direct/indirect type system], as has been suggested for Turkish [by Grunina, Aksu-Koç, & Slobin]". – ishwar  (speak)  20:20, 7 May 2008 (UTC)


 * Sorry for the late reply. Well, it seems that Turkish is experiencing a similar drift. I'm convinced though, if one of the forms is more neutral then it's -di, but the amount of neutrality is not absolute. But I don't completely agree with "Çok büyüdün" example, I would rather say (and expect) "Çok büyümüşsün" if I didn't see the person while he was growing up. Thanks for the valuable information, at least now I know that others have made a similar analysis :). Cyco130 (talk) 08:19, 14 May 2008 (UTC)


 * Yes, I agree. Çok büyüdün could be used only if the process of growing up was witnessed by the speaker: for example, a parent reminiscing about a child's growth between the ages of 10 and 15.  "You grew up quickly [over those years]."  The Turkish -miş tense is inferential rather than evidential (in the narrow senses).  So if you look out of the window in the morning and see that the ground is wet—which is certainly visual evidence—you would nevertheless say dün gece yağmur yağmış  ('last night rain [I gather] has-fallen').  But if you were woken during the night by the sound of rain on the roof you could say the following morning dün gece yağmur yağdı  ('last night rain fell').  The difference is that in the latter case you actually witnessed (or heard) the process. --NigelG (or Ndsg) | Talk 08:37, 19 May 2008 (UTC)

French terms
Are the French terms in the introduction, with accents and all, really necessary? Unless they are actually used in English, they should be removed, as this is the English wikipedia, and this article has no strong ties to French. — Preceding unsigned comment added by 101.166.150.53 (talk) 08:10, 21 November 2012 (UTC)

External links modified
Hello fellow Wikipedians,

I have just modified 6 external links on Evidentiality. Please take a moment to review my edit. If you have any questions, or need the bot to ignore the links, or the page altogether, please visit this simple FaQ for additional information. I made the following changes:
 * Added archive https://web.archive.org/web/20060917025820/http://www.linguistics.ucsb.edu/faculty/cumming/ling50/evidentiality.htm to http://www.linguistics.ucsb.edu/faculty/cumming/ling50/evidentiality.htm
 * Added archive https://web.archive.org/web/20110605015907/http://www.u.arizona.edu/~fdehaan/papers/ to http://www.u.arizona.edu/~fdehaan/papers/
 * Added archive https://web.archive.org/web/20080906163946/http://www.u.arizona.edu/~fdehaan/papers/evidbib.html to http://www.u.arizona.edu/~fdehaan/papers/evidbib.html
 * Added archive https://web.archive.org/web/20070926151704/http://www.u.arizona.edu/~fdehaan/papers/evidsem.jpg to http://www.u.arizona.edu/~fdehaan/papers/evidsem.jpg
 * Added archive https://web.archive.org/web/20050529042033/http://www.faculty.uaf.edu/ffgmh1/docs/den-evid.pdf to http://www.faculty.uaf.edu/ffgmh1/docs/den-evid.pdf
 * Added archive https://archive.is/20130112230102/http://cf.linguistlist.org/cfdocs/new-website/LL-WorkingDirs/pubs/reviews/get-review.cfm?SubID=12044 to http://cf.linguistlist.org/cfdocs/new-website/LL-WorkingDirs/pubs/reviews/get-review.cfm?SubID=12044
 * Added tag to http://cf.linguistlist.org/cfdocs/new-website/LL-WorkingDirs/pubs/reviews/get-review.cfm?SubID=59434

When you have finished reviewing my changes, you may follow the instructions on the template below to fix any issues with the URLs.

Cheers.— InternetArchiveBot  (Report bug) 22:35, 25 September 2017 (UTC)

Subjunctive I in German
Isn't this (Konjunktiv I) an example of this? I can see subjunctive two being a grammatical mood instead (hypothetical, wishful, etc), but using subjunctive one indicates the speaker is referring someone else, using indirect speech. The speaker is not stating his own opinion or expressing his own beliefs, merely referring someone else's. — Preceding unsigned comment added by Ketil (talk • contribs) 09:42, 3 November 2017 (UTC)

Awkward sentence in intro
I have no idea what the first half of this sentence is intended to mean: "Evidential is a meaning of nature and statement that is whether evidence exists for the statement and what kind of evidence exists."


 * Good point. I have removed that sentence. It's just repeating the opening sentence. Bondegezou (talk) 08:50, 29 June 2021 (UTC)