30

When learning some basic French, I was somewhat surprised to learn that phrases of the form "I have found the cat" generally translate almost word-for-word from English (J'ai trouvé le chat). To me, it's not immediately obvious that possession ("I have"/"J'ai") has a correspondence with past tense, although if I think about it a little more I suppose I can kind of see how it makes sense.

This makes me wonder: Is this a common pattern in other languages? Especially ones not closely related to English.

llama
  • 403
  • 4
  • 5
  • 9
    If you noticed this because it was common to both English and French, why do you consider English the unusual one in the title? – Barmar Oct 10 '19 at 17:31
  • 2
    @Barmar I don't understand what point you're trying to make. I don't think my experience implies that either of the mentioned languages is more unusual than the other for having this feature. It seems like an unusual feature in any language, and I'm a native English speaker so I was curious whether this is an unusual feature in my native language or more common and fundamental than that. – llama Oct 10 '19 at 19:35
  • 3
    phrases of the form "I have found the cat" generally translate almost word-for-word from English (J'ai trouvé le chat) This is not quite true. In the English phrase, "have found" is different from "found," and it emphasizes that the action has been completed. In French, the passe compose is simply how the past is expressed in the spoken language. (There is a passe simple, which is only used in formal written French.) Also, note that French has some verbs, e.g., aller, that take etre rather than avoir. –  Oct 11 '19 at 00:03
  • 1
    English, German, and French all use "to have" and "to be" when forming the past tense - "to have" when the verb is a discrete or completed event (like "I have found" or "I have walked") and "to be" when the verb is continuous or ongoing at the time being referred to (such as "I was looking" or "I was walking"). Other related languages probably do similar, but I wouldn't know; I only speak those three. – anaximander Oct 11 '19 at 08:43
  • 5
    @Barmar My inference is that the OP originally saw this feature and just thought it was an idiosyncrasy of English, but then saw it in French, and started wondering whether it's a general feature of languages, or whether it's a feature particular to English that French happens to share. Like if you see one person walking down the street wearing a gas mask, you might dismiss as one person being weird, but if you see two people, you might start to wondering whether something's going on. – Acccumulation Oct 11 '19 at 17:10
  • 1
    Related – Stephen Powell Oct 11 '19 at 18:46
  • Related, possibly duplicate: https://linguistics.stackexchange.com/questions/7184/relationship-between-possession-to-have-and-tenses-i-have-seen – Jetpack Oct 13 '19 at 17:58
  • This feature does not exist in Persian language. For "passe compose" one says:من گربه را پیدا کرده ام – Ali Taghavi Nov 24 '19 at 07:39
  • please see "Perfect simple" https://en.wikipedia.org/wiki/Persian_verbs#Perfect_simple – Ali Taghavi Nov 24 '19 at 07:44
  • Once you have found a cat, you have a found cat! – Carsten S Feb 04 '21 at 11:31

1 Answers1

58

This is what's called a "Sprachbund" feature: it's a trait shared by a bunch of languages in an area, even ones that aren't genetically related. In particular, this one is a feature of the "Standard Average European" Sprachbund, a group of languages centered in Western Europe, and it's one of the features that was originally used to define that Sprachbund in the first place!

The construction seems to have arisen back in Latin. In Classical Latin, past aoristic verbs ("I loved") and present perfective verbs ("I have loved") look exactly the same: both of those would be written amāvī. But it's a pretty useful distinction to be able to make! So in Vulgar Latin, a new construction arose, using the verb habēre ("to have"). It's thought that a phrase like habeō litterās scriptās "I have (written letters)" got reanalyzed into "I (have written) letters", with habēre no longer indicating that you're actually holding anything in your hands, just that an action's been completed in the past.

Vulgar Latin eventually evolved into French, Italian, Spanish, and all the other Romance languages, and brought this construction with it; habēre is the direct ancestor of French avoir, via a series of sound changes. And once various Romance languages and Germanic languages and others were all being spoken in the same area, this feature spread through the Sprachbund: people speaking Germanic languages started to use the same construction. English "have"/German haben/etc isn't actually at all related to Latin habēre, but they looked similar, so it was the obvious choice when adopting the construction into Germanic.

Nowadays, this feature is called the "have-perfective", and it shows up in all sorts of languages within the Standard Average European (SAE) Sprachbund. It's not at all universal, but can be a good way to determine if a language has been influenced by SAE or not!

Draconis
  • 65,972
  • 3
  • 141
  • 215
  • 4
    I can confirm that this construct is also present in Spanish with haber (the local cognate of avoir), for example, "He encontrado el gato" (="I have found the cat"). – Robert Columbia Oct 09 '19 at 23:40
  • 1
    The Romance and Germanic families use auxiliaries meaning to have with the past passive participle, but Slavic (except modern Russian) languages use to be with a special participle. (Modern Russian simply omits the auxiliary.) – Bert Barrois Oct 10 '19 at 11:48
  • 5
    It may be worth noting how Portuguese developed in this, by using the verb ter from Latin tenēre ("to hold"), cognate to French tenir. – Michaelyus Oct 10 '19 at 14:19
  • @Michaelyus Oh, now that's fascinating! I didn't know any Western Romance languages used anything other than (a descendant of) habēre. – Draconis Oct 10 '19 at 14:45
  • 1
    Did the Continental Celtic languages (at least the well-attested ones) participate in this process at all? – gormadoc Oct 10 '19 at 18:41
  • 3
    @gormadoc To the best of my knowledge, attestation of the Continental Celtic languages ends before the habēre-perfective really catches on in Latin. But Celtic is very far from my area of expertise so I'm not sure. – Draconis Oct 10 '19 at 19:34
  • 10
    It was a linguistics "mind blown" moment for me when I learned that habere and haben (and Germanic friends) were NOT cognates. Certainly helped me understand why we need to be very cautious with assuming cognates, and keep remembering that coincidences are much more likely than our brain is led to think... although I guess in this case it's not fully a coincidence, as the superficial similarity of the words probably helped this way of forming the past spread to Germanic languages. But still. – LjL Oct 10 '19 at 21:39
  • The same construct (the verb for have used to make a past tense) is also present in modern Greek. Since you say that this arose in Latin, does that imply it wasn't in ancient Greek and entered modern Greek through Latin? Or, if not, did it arise independently in Ancient Greek as well? It seems, to my linguistically ignorant self, more likely that it would have come from Greek rather than Latin if it existed in both ancient languages. Is that wrong? – terdon Oct 11 '19 at 09:00
  • Why do you say that "have/haben" is unrelated to "habēre"? Didn't the latter have the same meaning, i.e. "to have"? – Dave Oct 11 '19 at 09:10
  • 2
    @Dave Because 'have' and 'haben' ultimately derived from the Proto-Indo-European root kap- 'to hold', and 'habēre' derives from the PIE root ghabh- 'to give; to receive'. When looking for cognates one does not look for similar-looking words, but rather regular correspondences. I recommend reading up on Grimm's Law. – Angelos Oct 11 '19 at 11:24
  • 1
    @terdon Drinka (2003) actually suggests it's a post-Homeric Greek innovation that spread across Europe. – Michaelyus Oct 11 '19 at 12:13
  • @BertBarrois This Sprachbund feature "neo-perfect" (as opposed to the original Slavic perfect that transformed into the compund preterite) does also exist in Slavic languages. At least in Czech "mám uvařeno", "má napsáno". It is not completely grammaticalized and is not considered a tense but it is quite a special use of the verb. – Vladimir F Героям слава Oct 11 '19 at 13:28
  • @Draconis Any non-IE languages that you know of? – Mitch Oct 11 '19 at 21:30
  • @Draconis Answered my own question: from WALS the only 'possessive' perfect is in Europe. – Mitch Oct 11 '19 at 21:33
  • @Mitch I believe Basque shows the same feature (it's one of the few non-IE languages in the Sprachbund) but I don't know the language personally. – Draconis Oct 11 '19 at 21:33
  • @Draconis WALS says that Basque (and Hungarian/Finnish/Estonian) do not use possessive – Mitch Oct 11 '19 at 21:40
  • 1
    How common is the have-perfective outside the European Sprachbund? I think the original question was about worldwide frequency, not just one corner of the world. In Europe it's more common than average, but what's the average? Is it very very rare or is it moderately common? – Jetpack Oct 13 '19 at 17:53
  • @Jetpack Quite rare. It's mostly a European thing. – Draconis Oct 13 '19 at 17:54
  • Thanks, @Draconis. Yeah, it looks like it doesn't occur at all in this collection of 200 or so languages: https://wals.info/feature/68A#2/25.5/148.5. This asker thinks he saw it in Chinese: https://linguistics.stackexchange.com/questions/7184/relationship-between-possession-to-have-and-tenses-i-have-seen. – Jetpack Oct 13 '19 at 17:58