11

Italian, Spanish, Portuguese, and French all have a (compound) perfect tense, which I find curious, given that Latin did not. (You can alternatively perhaps say that it is either united with the perfect and preterite/aorist tenses, rather than it not existing, but in either case, Classical Latin did not have a separate or compound perfect active tense.)

My question is then: where did this compound perfect tense come from in the (Western) Romance languages? And why are there two of them?

The Germanic languages I'm aware of (English and German) both have a compound perfect tense, while German uses both "to have" and "to be" as auxiliaries. The uses of these auxiliaries aligns rather closely with uses in both French and Italian, while of course Spanish only uses the single auxiliary "to have", like modern English, albeit with a slightly different sense.

Noldorin
  • 486
  • 4
  • 13
  • As for why there are two of them, look at transitivity. Historically transitive verbs formed the habeo-perfect and intransitive verbs the sum-perfect which is quite logical because you would tend to say "I have a book written" (habeo) but "I'm the one who came" (sum). But as the construction spread it got grammaticalised. In Macedonian, for example (which was strongly influenced Latin), it's now grammatical to say имам одено (I have gone, habeo) and сум јаден (I've eaten, sum, literally "I'm eaten"). The two auxiliaries just got mixed. – Atamiri Feb 28 '18 at 12:22

3 Answers3

8

What we know is that the have perfect is a Sprachbund feature of Standard Average European. Where it originated is less clear. Because Romance languages are better documented in the late antiquity and early medieval time we have more early precursors of the have perfect in them. However, Southern German had a real need to develop a new perfect or past tense because the loss of the final -e neutralised the difference between past and present tense in the 3rd person singular (Modern High German present/past er sagt/er sagte, Southern dialects: er sagt/er sagt). Southern German dialect have completely lost the synthetic past tense, a phänomenon known as Präteritumsgrenze in German dialectology.

The Modern German periphrastic tenses already developped in Old High German time.

Sir Cornflakes
  • 30,154
  • 3
  • 65
  • 128
  • Auxiliary have or be is also used for past tense in modern Slavic, Greek, Armenian... Do you know about Celtic? – Adam Bittlingmayer Feb 27 '18 at 11:46
  • You are talking about modern Upper German dialects. I don't see any relevance for something that happened in early mediaeval times. Unless you can demonstrate the same features in Old High German. – fdb Feb 27 '18 at 15:47
  • 1
    @A.M.Bittlingmayer At least in modern Welsh, Gaidhlig and so on, present past and future all are formed with an auxiliary. But these languages are not considered part of SAE. – Omar and Lorraine Feb 27 '18 at 16:28
  • The Modern High German Perfect, Plusquamperfect, and Future II are already seen in Old High German. Unfortunately, I don't have data on the Altbairisch at hand to make a more pronounced statement yet. – Sir Cornflakes Feb 27 '18 at 16:32
  • Interesting. I believe the use of auxiliaries must have been present in Proto-Germanic however, since it appears to be present in all modern European languages, and was present even in early Anglo-Saxon, as far I know (which developed rather apart from other Germanic languages, at least until the Norse influence in the late 9th, 10th, and 11th centuries). The Sprachbund theory does sound very reasonable, though I'd be curious as to the ultimate origin nonetheless. – Noldorin Feb 28 '18 at 00:24
  • @Wilson Thank you. Armenian is not usually considered part of SAE either. So it seems this feature goes a bit beyond the SAE sprachbund. – Adam Bittlingmayer Feb 28 '18 at 20:16
  • 2
    Fun fact: The loss of the preteritum in Bavarian goes so far, that there is a twice analytical plusquamperfect Er hat gesagt gehabt He has said had in the spoken language. – Sir Cornflakes Mar 01 '18 at 11:11
  • 2
    I also find it interesting how in Northern and (some) Central Italian, the perfect is the default/near past tense, whereas in Southern Italian dialects, the preterite is used by default even for near-past events. – Noldorin Mar 01 '18 at 22:22
3

It actually came form Late Latin (e.g., probatum habeo). It's a natural process, a similar construction with "to have" has developed, for example, in Northwest Russian which is very interesting because Russian has no corresponding verb so it had to resort to its "у+NP" construction (e.g., у него корова подоено "he has milked a/the cow", lit. "at him cow milked") which clearly shows the origin of this tense is semantic.

Atamiri
  • 2,590
  • 1
  • 14
  • 15
  • Thanks for this bit of info. I didn't know it was present in Latin Latin (I thought only vulgar Latin). I was just searching around for sources, and found https://www.cambridge.org/core/books/early-and-late-latin/on-the-use-of-habeo-and-the-perfect-participle-in-earlier-and-in-later-latin/38069D2622B19FA68B54A1E61B43998A, though unfortunately I can't access it. Maybe it was originally a feature of the vulgar tongue that became accepted in formal Late Latin? – Noldorin Feb 28 '18 at 00:26
  • Also, do you know if the Russian innovation was independent? One could imagine it either springing up separately, or being transmitted from Late/Vulgar Latin via the Germanic languages to the Slavic tongues. (Russian- and German-speaking regions were of course much closer in the past than in the present day.) – Noldorin Feb 28 '18 at 00:27
  • @Noldorin Yes, the Russian habeo-perfect has developed independently, and only in one dialect in the far North. It's also independently developed in Czech and Polish to some extent, there are a few papers on it (speaking of "new tense in the making" since it's still syntactically restricted). A sum-perfect has also developed in some Silesian and Moravian dialects of Czech independently so this too seems to be a general grammaticalisation pathway. – Atamiri Feb 28 '18 at 12:14
  • Are you sure it developed independently in Czech and Polish? That would surprise me, given the geographical closeness to German and even Romance-speaking areas, thus being part of the Sprachbund. – Noldorin Feb 28 '18 at 19:23
  • 1
    @Noldorin Yes, it developed independently centuries after the contact with German and only in a fringe dialect initially. – Atamiri Feb 28 '18 at 20:10
  • 1
    Okay, thanks for the info! You seem to know your stuff well. – Noldorin Mar 01 '18 at 22:21
  • I know that habere can be constructed with the double accusative with the German translation halten für "to consider" for habere, so probatum habeo may be this construction and not yet a periphrastic tense. – Sir Cornflakes Mar 02 '18 at 13:09
  • Being a native Russian speaker I'd say what "у него корова подоенА" means is not "He has milked the cow" but rather "He has got the cow milked", i.e. someone else has milked his cow. Furthermore, exact English translation could as well be something like "The cow IS milked at him". – Aharon M. Vertmont Mar 04 '18 at 20:25
  • 1
    @AharonM.Vertmont In Northwest Russian it’s подоенО. – Atamiri Mar 05 '18 at 08:42
1

Through the process of grammaticalization. It is very common among various languages that the means of expressing perfectivity develops via those of expressing possession. ("Sum" in Latin used to have a possessive function too, say, in constructions with Dative).

As to why there are two auxiliaries, one traditionally discriminates the verb class called unaccusatives which is conjugate with "be". In Middle English there also existed two auxiliaries whose choice was determined in accordance with the verb class, one residue being "go" (e.g. "She is gone").

Mitch
  • 4,455
  • 24
  • 44
Aharon M. Vertmont
  • 511
  • 3
  • 11
  • Are "unaccusative" verbs related to "intransitive" verbs somehow? – Noldorin Mar 02 '18 at 18:39
  • @Noldorin yes, unaccusative and unergative verbs both constitute a class of intransitive verbs (the classification is based on what semantic role the subject of the corresponding sentence has). – Aharon M. Vertmont Mar 02 '18 at 19:23
  • Okay, thanks. I'm guessing unaccusative verbs are considered the classes of verbs that take the "to be" auxiliary in languages like Italian and French too? – Noldorin Mar 03 '18 at 00:49
  • 1
    @Noldorin, yes, that's right. Note that the lists of unaccusative verbs are language specific. – Aharon M. Vertmont Mar 03 '18 at 09:39