Preliminary note: If a Lakhota word ending is in uppercase, the Ablaut may undergo a transformation; this is always the Ablaut A (uppercase), which changes to e in sentence-final position, otherwise it is a; e.g. napȟÁ → napȟá / napȟé  'flee'.

Lakhota is a paramount example for a split-intransitive language (or active-stative language) distinguishing between active and stative verbs (cf. Active-stative languages (split-intransitive languages)), where both active and stative verbs can be subdivided in transitive and intransitive verbs. Van Valin (1977: 1) posits that "Lakhota is the best known representative of what are often called 'stative-active languages' ". With regard to the distinction between active and stative verbs he continues:

This is explicitly expressed in the coding of 'subjects' and 'objects': the 'subjects' of stative verbs have the same pronominal forms as the 'objects' of transitive active verbs, while the 'subjects' of intransitive active verbs have the same form as those of transitive active verbs." (Van Valin 1977:2)

Van Valin motivates placing 'subject' and 'object' between quotes (concerning Lakhota) by  the fact that "While the single argument of an intransitive active verb is an actor, that of a stative verb is a patient or an experiencer". The canonical notions of 'subject' and 'object' (which are in fact notions historically established on the base of Indoeuropean languages) are less relevant or adequate for Lakhota than the interpretation as semantic roles, actor, patient, experiencer. This alignment system is both semantically and lexically founded following language-specific criteria.

A remarkable consequence of the active/stative verb distinction is that many adjectives and nouns can also behave like stative verbs in Lakhota (or Osage, and other languages).

active verb, intransitive psíčA  'to jump/leap/hop'
active verb, transitive apȟÁ  'to hit/strike/beat/spank somebody/something'
stative verb, intransitive
ȟópA  'to be strikingly beautiful, very attractive, very pretty, good-looking'
stative verb, transitive waštélakA  'to like/love somebody/something'
noun wiȟópečA  'a stunningly beautiful woman, attractive or fastidious woman'

Esteban (2013) describes this and provides several examples:

"Lakota grammar makes a distinction between stative and active verbs, in terms of the presence or absence of control by an agent [Footnote: Although the distinction between stative and active verbs is semantically motivated, sometimes this differentiation seems to be triggered by lexical criteria: for example, while yaŋkÁ  'sit' and ečhúŋ  'do' are stative verbs, blokáskA  'hiccup' and pšá  'sneeze' are active verbs, contrary to what could be expected.]. Thus, while stative verbs (e.g. ištiŋmA  'sleep', ičhágA  'grow', kakížA  'suffer', yazáŋ  'hurt', etc.) describe states or conditions over which we have no control, active verbs (e.g. mani  'walk', wačhí  'dance', ya  'go', apȟé  'look for', etc.) describe actions that are controlled by an agent. Consequently, a great number of adjectives (e.g. káŋ  'old', wašté  'good', háŋskA  'tall', khúžA  'sick`, etc.) and nouns (e.g. wičhaša  'man', wíŋyaŋ  'woman', waákisniyA  'doctor', itȟáŋčhaŋ  'chief', etc.) can also function as stative verbs in this language:" (Esteban 2013: 2, also examples (1) and (2))

(1)  Ø-    káŋ-pi
     'They are old.'

(2)  ma-     wíŋyaŋ
     'I am a woman.'


The formal distinction (as opposed to semantic) between active/stative verbs is made by two groups/series of pronominal affixes in Lakhota. The type of pronominals employed formally determines the type of the verb. Rood & Taylor (1996) about Lakhota pronouns: "Independent pronouns are rarely used in ordinary Lakhota but are available for emphatic expressions or to serve as the objects of postpositions such as kičhi 'together with'. There are two sets. The first is simply emphatic; the second is used to contrast one referent with others."

stative active
1st person singular -ma- -wa/bl/l- (Class 1/2/3)
2nd person singular -ni- -ya/l/n- (Class 1/2/3)
3rd person singular -Ø- -Ø- In Lakhota there is no gender, therefore according ambiguities occur (he/she/it).
1st person dual inclusive -uŋ(k)- -uŋ(k)- For 1D and 1P a k is added when the next word begins with a
1st person plural exclusive -uŋ(k)-...-pi  -un(k)-...-pi pi is an enclitic in word-final position and not part of the pronominal
2nd person plural -ni-...-pi -ya/l/n-...-pi (Class 1/2/3)
3rd person plural animate
- collective
- distributive
- The plural of inanimate arguments is marked by reduplication
- wičha is also a noun meaning 'person/guy' used only for males. The homonym here is probably grammaticalized to a syntactic clitic (Esteban 2013).

Active verbs are of three classes:
Class 1 (e.g. slolyA 'know', máni 'walk', lowáŋ 'sing')
Class 2 (e.g. yuhá 'have', waŋyáŋkA 'see', 'go')
Class 3 (e.g. yaŋkÁ 'sit', ečhúŋ 'do', úŋ 'use')
(Esteban 2013)

The same predicate can carry two different meanings depending on whether it is considered as stative or active.

(3)  akhé  wígni i-   bláble
     again hunt  INST-2S.ACT.go (redup)
     'I go hunting.'

(4)  waŋ– čhí          -yaŋke háŋtaŋhaŋš ečhél    i-   má-    yaye
     STEM-1S.ACT+2S.STA-see   if         properly INST-1S.STA-go (redup)
    'If I see you, I get well.'

Monotransitive verbs can be coded for two arguments by two pronominal affixes, one stative and one active.

(5)  yuš´íŋye-  ma-    ya-    ye
     'You frightened me.'
(6)  ečhá- wičha- weči-      čuŋ
     STEM- 3P.STA-1S.ACT.BEN-do
     'I did it for them.'

weči in example (6) is a portmanteau form for the morphemes wa and kiči, which stand for first person singular agent and the beneficiary of the action. (Esteban 2013, including the examples).

Active-stative alignment interpreted by [semantic roles,] thematic relations and semantic macroroles

Van Valin (1977) is interpreting the two groups of pronominal affixes by means of semantic roles. The notion of semantic roles was introduced by Fillmore (1968) and a basic set of semantic roles was proposed by Foley (1976). In Role and Reference Grammar (RRG) (Van Valin & LaPolla 1997, Van Valin 2005) with regard to Foley's roles, a different, ampler set, more specialized on verb semantics, is defined as semantic roles, while a reduced set based on Foley's semantic roles figures as thematic relations instead:

  • Agent (AGT) - usually animate entity responsible for the action
  • experiencer (EXP) - animate being experiencing a mental state or event (subbranches: Cognizer, Perceiver, Emoter)
  • Recipient (RCP) - usually animate receiver of the theme
  • Stimulus (STI) - an entity to be experienced sensitively or psychologically
  • Theme (THM) - an entity to be handled or manipulated by the action
  • patient (PAT) - entity affected by the event

I specified my own abbreviations, they may differ elsewhere. RRG employs thematic relations to describe syntactic-semantic relations in the verb argument structure. The theory also defines and employs semantic macroroles, which are a top class grouping semantic roles. There are only two macroroles, actor (A) and undergoer (U). actor covers the semantic roles Agent, experiencer and sometimes Recipient, while undergoer covers Recipient, Stimulus, Theme and patient. The macroroles (can) overlap by Recipient and Stimulus, which language-specifically and by case can be either actor or undergoer.

Van Valin (1977) speaks of semantic roles (after Foley), but in today's view and considering RRG they should rather be considered thematic relations. He follows Foley in sharply distinguishing nuclear semantic roles from syntactic arguments of a verb. He states that "Nuclear roles relate to the participants in an event, activity, action or state specified by a verb, whereas arguments are the obligatory syntactic manifestations of certain nuclear roles." (Van Valin 1977: 8) Following those notions, they can be derived accordingly to accommodate the RRG thematic relations, where then Agent, patient and experiencer would be nuclear thematic relations, and Recipient, Theme and Stimulus would be syntactic arguments.

(7)  wičháša ki  hená  wówapi ki  Ø   -wičhá     -wa   -k'u.
     man     the those book   the INAN-3PL.ANIM.U-1SG.A-give
     'I gave the book to those men.'
(8)  ma-Ø-nú-wičha-wa-ši.
     'I told them to steal it.'

In a footnote regarding example (8), Van Valin specifies:

 "Lakhota is a split-intransitive language, and therefore the bound markers on the verb indicate actor versus undergoer, not subject versus object. 'Subject' in Lakhota is [S, A]: that is, the single argument of an intransitive verb, regardless whether it is actor or undergoer, and the actor of a transitive verb. The 'Ø' glossed INAN indicates that transitive verbs entail a specific undergoer argument, even though inanimate undergoers are not explicitly indicated morphologically on transitive verbs."

Van Valin (2013: 92), including the examples (7), (8).

Note that the undergoer in both examples is wičhá (3PL.ANIM.U) 'male persons', which in (7) is a Recipient and in (8) a patient. The undergoer is not the Theme (wówapi ki  'the book' in (7) and Ø  'it' in (8)). That is because "Lakhota shows secondary-object alignment, and therefore the recipient rather than the theme is the undergoer" (Van Valin 2013: 113).