4

In Geoff Pullum's recent post Being an Auxiliary on the Lingua Franca blog, he states that the sense of "have" as an auxiliary (forming the perfect tense) evolved from the possession sense, "but the speciation that separated them took place half a millennium ago".

This struck me as much too recent, given that an almost identical perfect construction exists not only in most (all?) other Germanic languages, but also in the Romance languages (with the exception of Portuguese?). But, sure enough, the OED says that this evolution "to some extent parallels developments in other Germanic and Romance languages, but appears at least partly to reflect development within English". (On the other hand, it also says "This development appears to have largely taken place before the written record.")

Assuming that it indeed happened within the last half millennium or so, how can this kind of fundamental language change occur in parallel between a dozen or so languages dispersed over a continent (from Iceland to Romania)? Did the innovation occur in one region and then spread between languages (like a new vocabulary item might), or is it better to think of it as a natural development that came about independently in multiple language communities (like simplification of noun inflection)?

Sir Cornflakes
  • 30,154
  • 3
  • 65
  • 128
Stephen Powell
  • 143
  • 4
  • 2
    I think that what Prof. Pullum is saying is not that the perfect arose half a millenium ago but that the modern strict distinction between English auxiliaries and lexicals arose half a millennium ago. And I think it's true that many of the characteristic NICE properties which distinguish auxiliaries from lexicals are products of the transition from ME to ModE. – StoneyB on hiatus Jul 08 '16 at 10:48
  • 1
    See also this question and its answers: http://linguistics.stackexchange.com/questions/17462/do-germanic-words-have-romance-qualities-and-vice-versa – Sir Cornflakes Jul 09 '16 at 21:54

2 Answers2

7

Areal features develop when languages from different groups or branches are in contact with each other. There are a few main mechanisms - common substrate, common superstrate, parallel development.

About the feature in question:

The use of have, haben, avoir, avere etc as an auxiliary for the past tense is an example of an areal feature. The areas where languages use this auxiliary correspond roughly to former Celtic-speaking areas and to those where language was heavily influenced by Vulgar Latin, including deep in the Balkans. I will not speculate on the ultimate source and exact mechanism here.

Danish, Dutch, French, German and Italian represent a transitional zone, where to be is the auxiliary for a specific subgroup of verbs (primarily verbs that indicate motion or change of state when used intransitively – in French and Italian also all reflexive verbs, and in Italian also unaccusative verbs in general), and to have for all others. Old Spanish and Middle English had this distinction too, which is to say this feature's area was still evolving in recent centuries - across language families.

It is not common unbroken Indo-European legacy - Latin habere and Germanic haben are not actually cognates, and classical Latin and Greek made little use of auxiliaries anyway. Many Indo-European languages outside of Western Europe - Armenian, Iranian languages, Slavic languages beside Macedonian - use the copula as an auxiliary, but most do not use any translation or cognate of habere or haben as an auxiliary for the past tense.

enter image description here

[Pardon the seeming equation of languages with current political entities]

Janus Bahs Jacquet
  • 4,673
  • 1
  • 22
  • 28
Adam Bittlingmayer
  • 7,664
  • 25
  • 40
  • The use of the verb "to be" as an auxiliary to form transitive and intransitive verbs is wide-spread in Iranian and Indo-Aryan languages. Also in Aramaic (Semitic). – fdb Jul 08 '16 at 18:41
  • Yes, but the delicate balance as in French, Italian, German and Dutch is not. – Adam Bittlingmayer Jul 09 '16 at 06:13
  • Re-worded to be hopefully more clear – Adam Bittlingmayer Jul 09 '16 at 07:09
  • 1
    Standard Ukrainian has future tense with "to have", the map does not reflect this. – Constantine Geist Apr 19 '17 at 22:31
  • 1
    @ConstantineGeist The map is for the past tense. English has auxiliary be and even do for some random tenses, the map is not intended to show that either. – Adam Bittlingmayer Apr 20 '17 at 07:54
  • "... but most do not use any translation or cognate of habere or haben." That's a tall claim without a solid etymology of both words, the ga-prefix, or even the auxiliary usage that is in question. Oddly enough, **kap- / *keh₂p-* is suspect of a supstrate, and the pedigree of **gʰeh₁bʰ- or *ǵʰeh₁bʰ* isn't impeccable either. But it isn't #badlinguistics if it's CO. It's just linguistics being bad. – vectory Dec 14 '22 at 02:53
  • It is worth to notice that this map is not very accurate by painting Portuguese as green. To express past perfect, Portuguese uses only its conjugation, no auxiliary. If you happen to use "ter" + participle you will make a different tense that does not correspond to other Romance Languages! – Ergative Man May 09 '23 at 12:09
  • @ErgativeMan Historically Portuguese even had haver, the point of the map is that when an auxiliary is used for the past, it is from a verb that means to have, not to be, even for verbs of motion or state. – Adam Bittlingmayer May 09 '23 at 16:10
  • I'd be interested where the line is, is Occitan like Iberian Romance in this regard? – Adam Bittlingmayer May 09 '23 at 16:14
  • @AdamBittlingmayer, yes, I know that. What I am saying is that Portuguese does not do that like the other ones. – Ergative Man May 09 '23 at 23:19
  • @ErgativeMan There are differences within Germanic and Slavic too, even within variants or registers of one language. – Adam Bittlingmayer May 10 '23 at 06:12
  • Yes, there are, the map is just not very accurate – Ergative Man May 11 '23 at 00:24
  • 1
    @ErgativeMan The fact that the exact form created by ter + past participle has a different nuance in Portuguese does not change the fact that Portuguese can use a verb meaning ‘to have’, but not one meaning ‘to be’, to create active verb forms whose meaning is past-based (as opposed to present- or future-based), which is what the map shows. As such, dark green is the correct colour for Portugal. – Janus Bahs Jacquet Nov 18 '23 at 20:21
  • 1
    What is inaccurate on the map is colouring Ireland blue, because Irish does not use a verb meaning ‘to be’ to form past-based forms – like the other light green areas, it uses a verb meaning ‘to be’ to form past-based forms (the perfective aspect) of certain verbs of motion, and a construction meaning ‘to have’ for other verbs. The complication is that Irish has no verb meaning ‘to have’, and the construction used contains the verb meaning ‘to be’ (I have X = X is at me). – Janus Bahs Jacquet Nov 18 '23 at 20:25
  • And just for accuracy: the light-green areas do not generally (or at all?) use be for intransitives and have for transitives. The exact rules differ (e.g., only French uses be for all reflexive verbs, as far as I know), but the distinction most have in common is between verbs of motion (or change of state) vs everything else. It is true that when normally intransitive verbs that use be are used transitively, they usually switch to have, but transitive verbs as a general category use have in the light-green areas. – Janus Bahs Jacquet Nov 18 '23 at 20:31
  • @JanusBahsJacquet I don’t have a strong feeling on motion/state vs. intransitive, but is there really no motion happening in “Er hat Auto gefahren.” (vs “Er ist gefahren.”)? – Adam Bittlingmayer Nov 19 '23 at 11:04
  • @Adam Yes, there’s motion there – that’s why fahren normally uses be. I may have been a little unclear, but the base rule is that be is only for verbs that describe motion or change of state and are used intransitively – everything else uses have (including, usually motion or state verbs used transitively). So you don’t say, “Er ist geschlafen” or “Er ist gelacht” even though those are transitive. – Janus Bahs Jacquet Nov 19 '23 at 14:06
  • @JanusBahsJacquet True. Feel free to edit my answer. – Adam Bittlingmayer Nov 19 '23 at 16:39
  • 1
    Brief edit added! – Janus Bahs Jacquet Nov 19 '23 at 16:49
0

Portuguese has an original pluperfect from Latin and a compound pluperfect tense, e.g. partíramos / tinhamos partido = we had left. The simple pluperfect is gradually becoming a literary form. 'ter' is the usual auxiliary, but 'haver' is also found.

Terry Walsh
  • 1