In terms of syntactic structure, from the below grammar can we conclude that English is accusative language, not ergative.

S --> NP VP

VPtv --> Vtv NP

VPiv --> Viv

By intuition, I believe it should be accusative but I am confused on why, and how?

  • 1
    @MaryAaronson: No, word order has nothing to do with it. There is a tendency to mark noun case with SOV languages, but it can be either an accusative or ergative case system.
    – jlawler
    Commented Jan 25, 2014 at 16:52
  • 1
    Morphosyntactic alignment and word order are orthogonal concepts. Most languages exhibit free word order.
    – Atamiri
    Commented Jan 25, 2014 at 16:53
  • 1
    @MaryAaronson: And, by the way, you can't conclude anything from a grammar fragment. Grammar fragments are artificial constructs, not data.
    – jlawler
    Commented Jan 25, 2014 at 16:54
  • 1
    @jlawler, word order definitely CAN have something to do with it. Word order is one way to mark case. The general definition of ergativity does not restrict itself to morphology.
    – dainichi
    Commented Feb 4, 2014 at 7:36
  • 1
    @jlawler, "case is a morphological matter". If you define it thus, maybe. But analyses that treat particles and other syntactic devices as case are definitely not uncommon. In either case, e.g. Wikipedia's definition of ergativity explicitly states that the ergative patterning can be syntactic OR morphological: "An ergative language maintains a syntactic or morphological equivalence (such as the same word order or grammatical case) for the object of a transitive verb and the single core argument of an intransitive verb"
    – dainichi
    Commented Feb 5, 2014 at 1:19

2 Answers 2


Actually, there is a rather robust generalization called Mahajan's Generalization, which states that ergativity is found in "verb peripheral" languages (VSO and SOV) but not in verb-medial languages (SVO).

So despite the fact that ergativity is a type of argument-alignment, it does appear to be directly related to word order (or, perhaps more accurately, whatever the structural underpinnings of word order are).

Mahajan, Anoop. 1994. The Ergativity Parameter: Have-be Alternation, Word Order and Split Ergativity. NELS 24.
Mahajan, Anoop. 1997. Universal Grammar and the Typology of Ergative Languages. In Alexiadou and Hall (eds.), Studies on Universal Grammar and Typological Variation.

  • To say that "ergativity is found in verb peripheral languages" does not mean that all "verb peripheral languages" have ergativity. There is thus a logical flaw in your second sentence.
    – fdb
    Commented Jan 27, 2014 at 19:27
  • 2
    But it is directly related to word order -- I did not say that word order alone plays a deterministic role (it doesn't). But, (to the extent that the generalization holds) knowing a language's basic word order can be enough to know if it is possibly ergative. (As in the case of English: knowing it is SVO is enough to know it is not possibly ergative)
    – bta
    Commented Jan 27, 2014 at 19:32

English is (still) an accusative language, although the accusative case has only survived with some pronouns (me, him, her, us, them, whom). Thus, we still say “He (subject) saw me (direct object)". If it were an ergative language we would say things like “him (agent) I (subject) was seen”.

  • 3
    But of course calling those pronouns a "case system" is sort of stretching the concept beyond utility. English is syntactically accusative, beyond doubt; the categories of both transitive and intransitive subject are virtually always treated identically by the syntax. To the point, in fact, that the few ergative structures that surface in English (like -er/-ee and the shooting of the hunters) are matters of puzzlement when encountered by native speakers.
    – jlawler
    Commented Jan 25, 2014 at 17:47
  • I do not accept that your examples have anything to do with ergativity. As you are surely aware, the English –ee suffix is borrowed from French –é(e), which continues the Latin perfect participle in -atus, -a, -um. In Latin (as in other ancient IE languages) the perfect participle has an active meaning with intransitive verbs and a passive meaning with transitive verbs. The French and English words in –é / -ee continue (in pre-modern times surely consciously) the situation in Latin.
    – fdb
    Commented Jan 25, 2014 at 18:49
  • @fdb Ancient IE languages had both active and passive participles. The use of passive participles to express active meaning is a later development due to grammaticalization.
    – Atamiri
    Commented Jan 25, 2014 at 22:50
  • Ancient IE languages like Sanskrit and Avestan have active and middle participles formed from the present and perfect stems and perfect participles formed from the zero-grade stem, which are passive if formed from transitive verbs (e.g. śruta- ‘having been heard’) and active if formed from intransitive verbs (e.g. sthita- ‘having stood’). I am talking about real languages, not reconstructions.
    – fdb
    Commented Jan 25, 2014 at 23:19

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.