Consider German, with its four cases and relatively free argument-order.
Now consider the following German sentence, courtesy of Google Translate.

  • Johan schenkte dem Mädchen eine Katze. (Johan gave the girl a cat.)

Couldn't this sentence be re-arranged with either the dative or accusative arguments first, like so?

  • Dem Mädchen schenkte Johan eine Katze. (argument order: Dative-First, Nominative Next, Accusative Last)

  • Dem Mädchen schenkte eine Katze Johan. (argument order: Dative-First, Accusative Next, Nominative Last)

  • Eine Katze schenkte Johan dem Mädchen. (argument order: Accusative-First, Nominative Next, Dative Last)

  • Eine Katze schenkte dem Mädchen Johan. (argument order: Accusative-First, Dative Next, Nominative Last)

Now remember that German also has a passive voice. So, once more courtesy of Google Translate, we get this sentence:

  • Das Mädchen bekam von Johan eine Katze geschenkt.

So, as German proves, free argument order in a language doesn't rule out the existence of that languages' passive voice.

But are there languages that a) have free argument order, and b) lack a passive voice? If not, why not?

  • 6
    @AdamBittlingmayer yes, many! Ergative languages typically lack a passive voice, although some may have antipassive. Commented Aug 9, 2023 at 5:48
  • 4
    And in response to the final question in the OP, there are many Australian languages with extremely free argument order and no passive voice (and plenty that also do not have antipassive) Commented Aug 9, 2023 at 5:49
  • 1
    Many Austronesian languages can be said either not to have a passive, or to have several passives, depending. Word order varies; it's a big family. But "passives" tend to be more associated with relative clause formation than with word order.
    – jlawler
    Commented Aug 9, 2023 at 9:20
  • 1
    Bekam ... geschenkt is not the normal passive, but a special construction paraphrasing an action involving a dative object (bekommen-Passiv). Normal passivization would be: Die Katze wurde dem Mädchen von Johan geschenkt. And that can be rearranged again. Commented Aug 9, 2023 at 14:52
  • 1
    @JamesGrossmann: Probly the clearest case is Keenan and Manorohanta's work on Malagasy, which has around 4 distinct types of "passive", depending on the semantics of the NP that's promoted to subject. In Indonesian languages (of which Malagasy is one), all relative clauses must have the relative word as subject. If it's not a subject, you have to "passivize" it so it becomes subject and can be relativized. The Malay relative clause marker yang is the focus of a great deal of this debate.
    – jlawler
    Commented Aug 9, 2023 at 22:47

1 Answer 1


I'm not sure I fully understand what you are asking but I can offer some examples of argument structure which may provide some context to it.

Hindustani (Hindi/Urdu), which is often analyzed as an “ergative” language as mentioned above, allows free argument order and lacks true passive forms. Closely related Punjabi is also ergative and retains morphological passive forms of verbs, however their use is rather marginal by speakers of most dialects. (In the Jatki dialect they can be heard very frequently, and their retention seems almost like a stylistic choice. These forms tend to lend themselves well to a form of aphorism-laden storytelling tradition popular in Punjab and Sindh.) The reason why Hindustani has shed itself of passive forms is an open question that nobody has provided a satisfactory answer for as far as I know. My hunch is that the answer lies in cultural differences in the native speech community of Hindi compared to those of Indic languages to the immediate west of it.

Take these example sentences:


لڑکے نے کتاب پڑھی

laṛke ne kitāb paṛhī


منڈے نے کتاب پڑھی

muṇḍe ne kitāb paṛhī

English translation for both: Boy book read (boy has read book)


لڑکا کتاب پڑھتا

laṛkā kitāb paṛhtā

منڈا کتاب پڑھدا

muṇḍā kitāb paṛhdā

English: Boy book read (boy reads book).

Now what's going on here is the argument order is exactly the same in both sentences but in the first the verb form (a perfect participle) agrees in gender with the book (feminine), while in the second the form (an imperfect participle) agrees in gender with the boy (masculine). This semantic distinction cannot be made in English but what the ergative postposition ne is doing is allowing a shift in topic/focus to the book instead of the boy. This is formally similar to a passive construction but not passive in meaning. Due to the agreement rules, we can reorder the arguments in the above sentences without changing their meaning. “kitāb muṇḍe ne paṛhī” is still focused on the book rather than the boy. (Generally though, speakers find it more natural to place animate participants in a sentence first, regardless of the syntactic construction being employed.)

As you can see the grammar of these constructions is practically the same between these two languages. Why does Punjabi retain passive verb forms while Hindustani does not? That is a very hard question to answer and does not necessarily have to do with a difference in verbal argument structure between the two languages.

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.