The Wikipedia page for the perfect aspect says the following regarding the origin of the verb 'have' as the English auxiliary for the the aspect:

The have-perfect developed from a construction whereby the verb meaning have denoted possession, and the past participle was an adjective modifying the object, as in I have the work done.[citation needed] This came to be reanalysed, with the object becoming the object of the main verb, and the participle becoming a dependent of the have verb, as in I have done the work. The construction could then be generalised to be used also with intransitive verbs.

This theory sounds vaguely plausible and I cannot find an adequate explanation elsewhere. However, this claim has no Wikipedia citation (as marked on the page).

Is this the reason 'have' is used as the auxiliary?

If so, can you find a reliable source?

  • There are definitely reliable sources to be found for this (though I can’t be arsed to look for one now after a 14-hour work day). It is clearer in older stages of many languages (and even some modern ones) that this is indeed the origin. For example, in French the past participle is still declined to agree with the object, though in modern French only if the object comes first (i.e., “J’ai fermé les fenêtres” ‘I’ve closed the windows’, but “Je les ai fermées” ‘I’ve closed them’). In Mediaeval Latin, they always agreed (i.e., “Fenestrās clausās habeō”). Nov 2, 2016 at 22:57
  • 1
    It's conjectural, because the habban perfect, although absent in Gothic, is already present in the earliest recorded West Germanic (OE and OS) texts from some five or six hundred years later. There's a fascinating dissertation on the subject, embracing both OE and Romance languages, here. Nov 2, 2016 at 23:49
  • 4
    @StoneyB: That reference, along with a short summary of what it says, would make a great answer. Nov 7, 2016 at 13:13
  • 1
    Another dissertation, on the history of the English verb have, is cataloged here. Nov 7, 2016 at 16:30
  • 1
    @JanusBahsJacquet: Your statement about French is highly misleading IMHO; it's not that the past participle is "still" declined to agree with a preceding object, but that it is now declined to agree with a preceding object, due to a rule introduced in the 14th century on the model of Italian. (But the rule in 14th-century Italian may be evidence of your point.)
    – ruakh
    Nov 11, 2016 at 7:08

1 Answer 1


I think there are really two separate questions here:

  • How did have become a perfect auxiliary in the first place?
  • Why is have the main perfect auxiliary?

It's important to make the distinction because they have different answers in different time periods. You seem to be asking the former question, judging from your question body.

Wikipedia seems to be accurate according to this section from Trask's Historical Linguistics:

The English perfect therefore results from the reanalysis of an original stative construction, in this case accompanied by a shift in meaning. A sentence of the form 'I have him bound' originally meant 'I have him in my possession, in a tied-up state', but this could readily be, and was, reinterpreted as meaning 'I have tied him up'. The participle, formerly a modifier of the object NP, with which it necessarily agreed, was reanalysed as part of the verb form and lost its agreement; as part of the same process, the verb have, originally the main verb in the sentence and bearing its usual meaning of ‘possess’, was reanalysed as an auxiliary expressing perfective aspect, the function it still has today.

This change happened in Old English.

The paper Pieces of the be Perfect in German and Older English answers the latter question. It's actually interesting to note that have and be had a long fight, stretching from Middle English to Early Modern English:

have first started showing up with verbs that used to exclusively take be during the ME period, around 1300, but didn’t completely replace be until ca. 1900.

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

Not the answer you're looking for? Browse other questions tagged or ask your own question.