Evidential markers in Yukaghir languages
As can be seen from the map on the left, many languages around the world have such evidential (or evidentiality) markers. The -di and -miş markers in Turkish are part of the tense system in the language, as they also express past tense. In other languages, evidential markers may be attached to the verb without also expressing tense, or can be a stand-alone word (called “particle”). For example, in Lezgian, a Northeast Caucasian language, the marker -lda, which encodes hearsay, attaches to the verb after a tense suffix, whereas in a Tibeto-Burman language Dumi a marker with a similar meaning is a stand-alone particle.
In Siberia evidential markers are quite common. Siberian indigenous languages with evidential markers include Nenets, Yakut, and Evenki. According to a WALS article, evidential markers appear “to be more of an areal feature than a genetic feature”. Indeed, the three Siberian languages with evidential markers mentioned above belong to three different language families: Samoyedic, Turkic, and Tungusic, respectively. All three languages share with Turkish (another Turkic language) the use of evidential markers to mark both directly witnessed and indirectly inferred events. Several other Siberian languages, including Nivkh and Yukaghir languages, use evidential markers only for indirectly inferred events. In other words, they have a rough semantic counterpart of the Turkish -miş but not of -di.
The following discussion focuses on Yukaghir languages, specifically on Kolyma Yukaghir. This language, sometimes known as Southern Yukaghir, is spoken in one of the coldest and least hospitable areas in Siberia: the forest zone near the sources of the Kolyma River, divided between the Sakha Republic and the Magadan Oblast (see map on the left). Although Southern Yukaghir, with only 5 to 50 speakers, is scarcely used for everyday communication, it sheds significant light on evidential markers and more generally on how people conceptualize and categorize events. Kolyma Yukaghir has an evidential marker for indirectly inferred events, pronounced -l’el, but it lacks an evidential marker for direct evidence, such as the Turkish -di. Thus, if a Kolyma Yukaghir speaker finds himself in a murder mystery involving butlers, he might say the literal counterpart of “The butler it did-l’el” or of simply “The butler it did” (in Yukaghir, like in Turkish, the object ‘it’ comes before the verb). The former case means that the speaker did not personally witness the crime, but infers the butler’s guilt from indirect evidence like fingerprints or hearsay. But crucially, the plain “The butler it did” is not neutral in the same way as the English sentence is. Because it lacks an overt expression of indirect inference – the marker -l’el – the speaker who says it is implicitly claiming to have personally witnessed the murder.
The standards of evidence that are involved in determining whether the event was witnessed directly – and hence does not need the marker -l’el – would make Western lawyers happy, as they are quite strict. First, only events that are observed by the speaker unambiguously qualify as being witnessed directly (and therefore, not needing the marker -l’el). Obtaining evidence for an event through hearing a sound may work for some events but not for others. For example, if a Yukaghir speaker hears a noise that seems to be a gunshot, he might say “Something burst”, without the evidential marker, as the bursting sound is heard directly. But the counterpart of “Somebody shot a gun” requires -l’el. After all, a gunshot-like sound may or may not signify that a person fired a gun, as a gun might have discharged by itself, or a champagne cork might have popped. Thus, if only a sound is heard and the actual shooting is not seen, the fact that someone shot a gun can only be inferred.
Second, what needs to be seen is the event itself rather than its result. For example, if I meet an old friend who used to sport a goatee, I might say “You have shaved your beard!”. For a Kolyma Yukaghir speaker this sentence would need the marker -l’e,l since what the speaker is witnessing is the lack of beard that resulted from a shaving event, but not the actual shaving event itself. In another example, cited in Elena Maslova’s paper “Evidentiality in Yukaghir”, a speaker says “Our mother and her companions came, they had a lot of berries gathered-l’el”. The reason that the evidential marker is used here is that the speaker saw the berries, but not how the mother and her companions gathered them.
Finally, the evidence for a directly witnessed event must be obtained at the time of the event itself. For instance, if the speaker learns that a shooting has occurred after the fact, “Somebody shot a gun” still requires -l’el. In one instructive example, a Kolyma speaker said the literal counterpart of “While walking there, I near a bear lair tea drank-l’el”. The evidential marker, was used because the speaker only noticed the lair after drinking the tea.
While the main purpose of the marker -l’el is to indicate that the event was not witnessed directly, it can also be used to express other, subtler shades of meaning. For example, internal personal properties such as being clever, kind, malevolent, and so on can only be inferred on the basis of their external manifestations. However, predicates signifying such properties can occur both with and without the marker -l’el. This contrast is used to signify whether the property is acknowledged by the speaker for the first time, in which case the marker -l’el is used, or has been established previously and is therefore known to be present, in which case the marker is not used. For example, in a narrative about the speaker’s first hunting experience, in which he was supervised by his elder brother, the brother makes two encouraging statements: first, right after the hunting was over, “Best hunter you are-l’el” and a bit later, “Best hunter you are”, without the evidential marker. The -l’el in the first sentence indicates that the elder brother has inferred, on the basis of the boy’s behavior, that he has the qualities required of the best hunter of a tribe. By the time of the second utterance, this fact has already been established, and therefore the marker is not needed. In an even subtler twist, if the speaker expects a certain quality to be present, this quality does not be expressed with the marker -l’el. Therefore, the marker can express surprise on the part of the speaker, as in “Clever person you are-l’el” which can be best rendered in English as ‘You have turned out to be a clever person!’.
In yet another subtlety, the marker -l’el can be used in describing situations that can be considered ‘eyewitness events’ par excellence, namely occurrences in which the speaker himself is a participant. In such cases the marker -l’el encodes inadvertent actions of which the speaker was not fully aware at the time of their happening. One verb that tends to be used with the marker -l’el in this sense is ‘forget’: for example, when one forgets one’s keys at home, one is not aware of forgetting them at the time. Similarly, “I was sitting down and then fell-asleep-l’el” describes the speaker falling asleep without expecting to do so.
Thus, Kolyma Yukaghir grammatically marks evidentiality while English does not. But it does not mean that such meanings cannot be expressed in English. In some cases we use adverbs such as allegedly, surprisingly, or inadvertently. In other cases, we cite the source, as in according to some linguists or as stated in the New York Times. There are even subtle grammatical ways to indicate in English whether the speaker witnessed the event or learned of it from hearsay. Consider, for example, I heard John play the pipes and I heard John plays the pipes. In the first sentence I report witnessing John’s playing the pipes, whereas in the second sentence I report something I know only through hearing other people mention it.