I am not sure of correct terminology, but let me call the medieval perfect tenses like amatum habeo — as opposed to the classical amavi — the "composite active perfect". One would expect the participle in the composite perfect to follow the gender and number of the object: acroasem auditam habeo, librum scriptum habeo, hoc factum habeo, urbes visas habeo…

However, in Italian the composite active perfect is almost always singular masculine, whatever the gender or number of the object. For example, one would say ho mangiato una pizza or ho mangiato due pizze, not ho mangiata una pizza or ho mangiate due pizze. Does the same ever happen in Latin when composite active perfect is used?

I am not very familiar with post-classical Latin, and I do not know how to figure things like this out on my own.

  • 3
    Good question: French has a funny rule for the passé composé where it agrees with the object if it is before avoir, e.g. je les ai vues As a side note, this form is medieval in the sense that it is certainly wasn't classical, but there are many medieval writers who avoid or don't use this form.
    – brianpck
    Commented Sep 30, 2016 at 21:41
  • 2
    @brianpck, it's similar (not identical?) in Italian. "Hai mangiato la pizza?" "Sì, l'ho mangiata." If it happens in several Romance languages, one would expect to see something in Latin. (I call the thing medieval for the lack of a better term. I didn't mean to imply it was universally adopted at any point.)
    – Joonas Ilmavirta
    Commented Sep 30, 2016 at 21:48
  • I'd call it the "Romance perfect" since that's where I most often see it.
    – Draconis
    Commented May 14, 2019 at 18:26
  • @Draconis That'd make some sense, but the same construction is also found in Germanic languages (e.g. English I have read, Swedish jag har läst) and the Romance languages I know have two perfects, so I'd find that name confusing.
    – Joonas Ilmavirta
    Commented May 14, 2019 at 18:31

1 Answer 1


According to Alkire and Rosen (Romance Languages: a Historical Introduction), a perfective construction with habeō wasn't uncommon in Classical Latin:

Caput cinctum habēbant filō.
They had their heads encircled with cord.

(Varro 5.15, first century BCE, translation mine.)

Over time, this got extended to intransitive verbs. In other words, habeō became an auxiliary; the subject didn't necessarily "have" any specific noun.

Dē eā rē supra scriptum habēmus.
We've written about this above.

(Vitruvius 9.1.4, first century BCE, translation mine.)

Here, habēmus has no direct object: it's just supporting scriptum.

This turned out to be a very useful construction! Latin (like all Italo-Celtic languages) had combined its present perfective and past aoristic tense-aspect combinations, but being able to distinguish those two meanings was really nice. So this new habeō-perfect generally took over the present perfective, leaving the original "perfect" to handle the past aoristic.

At first, when habeō had a direct object, the participle would agree with it. But when it didn't, the neuter singular was used by default. And over time, the neuter singular took over everywhere: the agreement didn't really add anything meaningful, and it only appeared half the time anyway. This was a slow and inconsistent process:

Metuō enim nē vōs ibi habeam fatigātōs.
But I'm afraid I might have exhausted you here.

(Augustine Sermones 37.27, fourth century CE, translation mine.)

Haec omnia probātum habēmus.
We've already tried all these things.

(Oribasius 7.48.189f, fourth century CE, translation mine.)

Both of these examples are from educated, formal writings in the same time period: Augustine was a prolific philosopher, and Oribasius was an imperial physician. So it seems both forms were perfectly acceptable at that point.

Over time, the "always neuter singular" form won out (unless the noun comes immediately before habēre, as pointed out in the comments—then it still acts like a direct object). And this is what we see in all the modern Romance languages.

P.S. A bit of searching has turned up people calling this the "composite perfect", "habeō-perfect", "Romance perfect", or "true perfect" (since it actually has a perfective meaning). I think all of those names would be understood.

  • Many thanks! I guess one could argue that scriptum is the object, meaning "writing" or "text", in that Vitruvius passage. But I do agree that reading it as a perfect tense with auxiliary habere makes a lot of sense.
    – Joonas Ilmavirta
    Commented May 14, 2019 at 19:49
  • 1
    Great answer. I do have one quibble, which is that the Augustine quote may not be a perfect at all. Here's the context: "Attendite, rogo vos: iam in fine lectionis sumus. Metuo enim ne ibi vos habeam fatigatos, ubi maxime exigo intentos." Listen up, please, we're now coming to the end of the lesson. I'm afraid I have you all worn out at the precise point where I most require you to be attentive. Is Augustine taking responsibility, "I've worn you out," or not, "I finally have you [at this point in the lesson], and you're worn out"? Maybe parallel texts could clarify? Commented Jun 25, 2019 at 16:18
  • @Kingshorsey That's a very good quibble! I'll find some more context for these and see what I can do.
    – Draconis
    Commented Jun 25, 2019 at 16:27
  • 1
    In general, caution needs to be exercisesed when equating a romance-looking construction in Latin to an actual romance construction. There's nothing indicating that the Vitruvius quote, to a CL speaker, meant something other than "we have it written above" in the literal, material sense by way of conflating the existential/possessive nōbīs suprā scrīptum est with the periphrastic perfect suprā scrīptum est. In fact, I think the absolute majority of instances of this use can be interpreted in this existential/possessive way. Commented Jun 26, 2019 at 1:56

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.