Tokana: active? (was Re: Active case-marking natlangs)
|From:||Jörg Rhiemeier <joerg.rhiemeier@...>|
|Date:||Monday, February 5, 2001, 23:14|
J Matthew Pearson <pearson@...> writes:
> My characterization of Tokana as active is based on the fact that the
> case-marking of the subject (insofar as it is possible to characterise a notion
> "subject" for this language) correlates with things like degree of
> agentivity/animacy and degree of involvement:
I have the impression from the current and earlier discussions that
Marcus Smith's definition of an active language contains the criterium
"Does not mark nouns for case" ;-) At least, he insisted on the
non-existence of case-marking active natlangs in a private-mail
discussion and refuted my Georgian counterexample.
But earlier, he *did* say that Nur-ellen was active even though it marks
nouns for case! In fact, the case marking in Tokana is quite similar to
that in Nur-ellen, as the following comparison shows:
> Absolutive subjects (undergoers):
> Ne Tsion itskane
> the-Abs John-Abs arrived
> "John arrived"
> Ne Tsion tioke
> the-Abs John-Abs died
> "John died"
Nur-ellen uses the objective case here, which corresponds to the
absolutive of Tokana:
The first example is "fluid", it could also be formed with the agentive
to express that John's arrival is a volitional action; but the second
one can hardly be formed so (unless one wants to say that John committed
> Nominative subjects (volitional agents):
> Na Tsion hostane
> the-Nom John-Nom danced
> "John danced"
> Na Tsion tsitspit kopo
> the-Nom John-Nom smashed-the pot-Abs
> "John smashed the pot"
Nur-ellen uses the agentive here, which corresponds to the Tokana
Jan dringent i drom.
John-AGT smash-PAST the pot-OBJ
Only animate agents can be volitional, hence only animate nouns have an
agentive case. Is this the same in Tokana (it seems so; how can
inanimate objects have volition)?
> Ablative subjects (non-volitional animate agents):
> Inaul Tsionu tsitspit kopo
> the-Abl John-Abl smashed-the pot-Abs
> "John accidentally smashed the pot"
Nur-ellen uses the dative for agents acting accidentally, as in the
Na Jan dringent i drom.
DAT John smash-PAST the pot-OBJ
The dative agent must be animate; inanimate nouns cannot be used with
the dative preposition which governs the agentive form of the noun.
To express that someone acts under external force, one would use the
> Instrumental subjects (inanimate agents):
> Itan suhoua tsitspit kopo
> the-Inst wind-Inst smashed-the pot-Abs
> "The wind smashed the pot" (by blowing it off the table)
Same in Nur-ellen:
Ni i sul dringent i drom.
INST the wind smash-PAST the pot-OBJ
> Dative subjects (experiencers):
> Ine' Tsione hilin ikei
> the-Dat John-Dat saw-the dog-Abs
> "John saw the dog"
> Ine' Tsione kesta
> the-Dat John-Dat happy
> "John is happy"
Same again in Nur-ellen:
Na Jan tirent i hu.
DAT John see-PAST the dog-OBJ
Na Jan gelent.
DAT John be-happy-PAST
Use of the agentive instead of the dative implies deliberate
Jan tirent i hu.
John-AGT see-PAST the dog-OBJ
"John watched the dog."
> Consider also the following triplet of sentences, each translated "John cut his
> Na Tsion hane silh
> the-Nom John-Nom cut finger
> Inaul Tsionu hane silh
> the-Abl John-Abl cut finger
> Ine' Tsione hane silh
> the-Dat John-Dat cut finger
> The choice of which sentence to use depends on the degree of volitional
> involvement exhibited by John: The first sentence would be used if John cut his
> finger deliberately (e.g., in preparation for a blood-sibling ritual). The
> second sentence would be used if John performed an action which caused his
> to get cut, but did so without intent (e.g., he was cutting food and the knife
> slipped). The third sentence would be used if someone or something else cut
> John's finger (e.g., he was running along and accidentally grazed his hand
> against something sharp).
In Nur-ellen, the first situation would be expressed by agentive, the
second by dative, the third also by dative if it is an accident; if
someone else cuts John's finger, it would be:
Ristent lebed e Jan`n.
cut-PAST finger-OBJ PART John
"[Someone] cut John's finger."
An example of the impersonal construction Nur-ellen uses instead of a
passive; more on this see below.
> I acknowledge that languages which are usually called "active" don't really work
> like this, but I don't know what else to call the Tokana case-marking pattern.
Well, up to a few minor differences, it works the same way as
Nur-ellen. So either both languages are active, or both are not. And
the last time I checked, there was a consensus that Nur-ellen was
In a second post, Matt addresses the question of passive and
> Tokana does not have passives or antipassives, in the sense that you mean.
Nor does Nur-ellen.
> there are no constructions which serve to promote or demote arguments, changing
> their case in the process. Perhaps the closest thing to a relation-changing
> operation in Tokana is the reflexive, formed by adding the prefix "uma(k)-" to a
> transitive verb. The subjects of reflexive verbs are always in the absolutive
> case, regardless of the case used for the corresponding transitive subject:
> Ma kahtin
> I-Nom hit-Pst-him(Abs)
> "I hit him"
> Me umakahte
> I-Abs Refl-hit-Pst
> "I hit myself"
Nur-ellen uses a reflexive pronoun, which does not change case marking.
> Otherwise, passive/antipassive-type meanings are expressed by simply leaving out
> arguments, as you guessed.
Nur-ellen does the same.
> To express "John was killed", simply omit all mention
> of the agent. There are actually two possible constructions here: If the focus
> is on the event ("John got killed"), the regular past tense form of the verb is
> used; if the focus is on the resultant state ("John has been killed, John is
> dead"), the present tense stative form is used:
> Ne Tsion kaihe
> the-Abs John-Abs kill-Pst
> "John was killed"
> Ne Tsion kaieha
> the-Abs John-Abs kill-Stat
> "John is (in a state of having been) killed"
> Eliminating the patient requires the addition of the prefix "u(k)-" to the verb
> stem. I call this prefix an antipassive marker, but since the addition of this
> prefix does not cause any change in case-marking/grammatical relations, the term
> is not entirely appropriate. Instead, adding "u(k)-" indicates that the
> participant which *would be* expressed with absolutive case has an
> unknown/indeterminate referent:
> Na Tsion kaihin Taniel
> the-Nom John-Nom kill-Pst-the(Abs) Daniel-Abs
> "John killed Daniel"
> Na Tsion ukaihe
> the-Nom John-Nom Antipass-kill-Pst
> "John killed someone"
In Nur-ellen, the patient is simply omitted, without an antipassive
marker or anything like that:
Notice the difference in case to the "John was killed" example. To
> Notice that the agent is still marked with nominative case in the second
> sentence. Eliminating this agent, we get:
> "Someone was killed"
In Nur-ellen, this is:
So is Tokana active or not? I'd say it is.