Why did English choose “have” as the auxiliary verb for perfect tense?

[selfzhouxinrong]

hello ！

-I have finished my homework
-I have fallen in love

Why did English choose “have” as the auxiliary verb for perfect tense?

Does this “have” has the same meaning as “I have a pen”？

Is this a metaphor ？
I think the essence behind any language phenomenon is Metaphor.

Does this “have” mean “ have a complete、 perfect action”，
just like I did something， then this action has belonged to me.
I made it so it belongs to me.
Am I right?

Thank you.

 

[berndf]

The periphrastic perfect (i.e. perfect with an auxiliary verb) developed out of constructs that can be traced back to late republican Latin. These construction became grammaticalized in Romance and most likely loaned from there into Germanic languages.

It exploits the property of the past participle to express completed actions. For transitive transitive verbs the meaning has also passive meaning and therefore transitive and intransitive verbs used different auxiliaries. In French you can still see traces of this:
Il est arrivé = he has arrived (intransitive)
but
Il a lu un livre = he has read a book (transitive)

The rationale behind these constructions are:
Il est arrivé = he is [an] arrived [person]
Il a lu un livre = he has a read book [=he has a book which has been read]

The past participle works like an adjective expressing the property of having undergone a now complete action.

In modern languages, many intransitive verbs have switched to using have instead of be to construct the perfect tenses. In English, the last be-perfects became obsolete in the 17th or 18th centuries, I am not sure when exactly.

 

[gburtonio]

As berndf mentioned, this feature is shared in many European languages, and is one of the features of the suggested European Sprachbund (what Whorf called 'Standard Average European'. Does anybody on this forum know of any non-European languages that share this feature?

A related use of 'have' is that used to express obligation, i.e. 'have to do'. I've never seen it analysed in this way, but you might pair have done vs. have to do in this way:

I have …

• spoken to him. (completed action)
• to speak to him. (action not completed but necessary)

You can sometimes do more or less the same thing in Italian (ho fatto vs. ho da fare) and Greek (έχω κάνει, έχω να κάνω). I'd imagine also in other European languages.

 

[Awwal12]

These construction became grammaticalized in Romance and most likely loaned from there into Germanic languages.

Is it assumed that Proto-Germanic lacked perfect tenses, or just that it originally used a different auxiliary verb (like it happens with some verbs in German, e.g. "ich bin gekommen"; Romance languages also use both "to have" and "to be" anyway)? Despite being pretty common cross-linguistically, perfects are actually a pretty alien concept to the speakers of a language which lacks them (Proto-Slavic did have perfect tenses, but it takes quite a while for a modern Russian speaker to get the basic idea). Are such loans conceivable outside of creoles and tight sprachbunds (which, I believe, wasn't the case) at all?

 

[Zec]

Hittite formed the perfect exactly the same way as Standard Average European: auxilliary 'to have' for transitive verbs, 'to be' for intransitive + past participle - this is the closest parallel that we have.

Ancient Indo-Iranian formed the perfect in an analogous way, but quite different in detail: instead of an auxiliarry 'to have', it used 'to be' in, probably, a possessive construction, with the subject in the dative case: "To him (is)* a book written" = He has written a book. ~ "To him (is) a book" = He has a book.

*the verb 'to be' could be freely omitted, as detailed in this thread.

The gist is the same: the past participle indicates the completion of the action, and the subject is somehow constructed as being its "possessor". The languages only differ in how exactly the possessive construction is formed.

 

[berndf]

Is it assumed that Proto-Germanic lacked perfect tenses, or just that it originally used a different auxiliary verb

It is assumed that the only native tenses in Germanic are present and past or, more precisely, non-past and past. All other distinctions are taken over from other languages. In some Germanic languages, like German or Dutch, this is to some extend still the case: past vs. non-past is the only required distinction (where past can be expressed by the preterite or by the perfect). All other distinctions are optional.

 

[Awwal12]

Surely taken over and not naturally developed (sometimes probably basing on the constructions which had already existed in the proto-language but hadn't been grammaticalized yet)?

 

[berndf]

Surely taken over and not naturally developed (sometimes probably basing on the constructions which had already existed in the proto-language but hadn't been grammaticalized yet)?

Hard to see how. Why would a language develop tenses in order to not use them? Germanic languages do not really need more then 2 tenses and popular language still largely needs only two. The 6 tense system (future, future perfect, present, present perfect, past, past perfect) is largely literary or academic and not popular; never has been.

English is a bit special; maybe because has much more Romance influence than other Germanic languages.

 

[Awwal12]

Why would a language develop tenses in order to not use them?

And would it actually *loan* them in order not to use them?
Languages do occasionally develop tenses, normally by grammaticalizing some general phrases; after all, "I am writing" isn't fundamentally different from "I am tall" (it's the origin of the very English gerunds which looks more interesting - but mostly irrelevant here), and the core meaning of the usual Germanic present perfect tenses is also transparent enough from their composition. The idea of loaning tenses, on the other hand, intuitively seems much more exotic and requiring some very unusual circumstances.

 

[berndf]

And would it actually *loan* them in order not to use them?

Yes, because the main use of early written German was to translate Latin (I am concentrating now on German because this is the Germanic language I know best and because it was the most widely spoken Germanic language of the time (late first millennium)).

"I am writing" isn't fundamentally different from "I am tall" (it's the origin of the very English gerunds which looks more interesting - but mostly irrelevant here)

That's how it looks but it isn't. If it were analogous to I am tall then writing would be a participle and not a gerund because the participle is an adjective but the gerund is a noun. The real origin of the continuous form is from the gerund: I am on writing -> I am awriting -> I am writing. Compare German Ich bin am Schreiben.

The big difference to our case is that this progressive form is popular in origin. The perfect tense isn't.

 

[Awwal12]

Yes, because the main use of early written German was to translate Latin

But we're speaking about the times when the majority (which apparently was defining the main trends of linguistic development) was simply illiterate.

 

[berndf]

But we're speaking about the times when the majority (which apparently was defining the main trends of linguistic development) was simply illiterate.

Yes, that is exactly the point. That is why it can't have developed naturally because it originally wasn't a construct of popular but of scribal language.

 

[Awwal12]

If it were analogous to I am tall then writing would be a participle and not a gerund because the participle is an adjective but the gerund is a noun.

True, it seems that (surprisingly) the gerunds proper and the participles have different etymologies in English, which I wasn't aware of. I stand corrected.

 

[Awwal12]

Yes, that is exactly the point. That is why it can't have developed naturally because it originally wasn't a construct of popular but of scribal language.

I actually cannot get how you deduced that it didn't come from popular language.

 

[berndf]

I actually cannot get how you deduced that it didn't come from popular language.

The development is sufficiently attested to show how precursors of the perfect tenses radiated from translations of Latin texts into the general language. The grammaticalization of these tenses happened only in MHG.

PS: Apart from that, there is no indication that popular language ever needed more than two tenses (apart from a grammaticalized version of the am-progressive in some dialects). Where the present perfect entered popular language it quickly became a mere replacement for the preterite and past perfect and future (normal and perfect futures) are all but unused in popular language and if past perfect is used in popular language then it serves as emphatic past. Ordinary German has absolutely no need for any perfect tense, because the concept it conveys has never been grammaticalized in German.

 

[ahvalj]

A few notes.

These perfect constructions tend to lose their resultative meaning with time, and so languages may develop fresher, originally more expressive, replacements instead. For example, Spanish and Portuguese at their current stage use both the older perfect with “to have” and a newer one with the etymological “to hold” — like for example the Spanish lo he hecho “I have done it” and lo tengo hecho ‘the same’, etymologically “I hold it done”. The exact relationships between both variants are different across the varieties of Spanish and Portuguese, but eventually the newer form is going to win, I guess.

On the other hand, these perfects with the passive participle and ‘to have’ are an option when the colloquial language (where such forms evolve) lacks a formalized distinction between active and passive past participles, like in Germanic and Romance. Where such a distinction is available, another perfect may evolve, the one with ‘to be’ and the past active participle, “I am who-has-done”, like in Baltic, Baltic-Finnic and historical Slavic languages (a millennium ago Slavic even had two past active participles, one of which was specialized for the perfect tense while another one remained independent). In modern Slavic languages past active participles have either disappeared or are no longer freely used in the colloquial speech, so these languages may develop a newer perfect, this time with ‘to have’ and the past passive participle. Macedonian has both variants, the older ‘I am who-has-done’ and the newer ‘I have it done’, codified in the standard language as synonyms.

Overall, I don't think that influence between languages is of any special importance here: such constructions are natural, they easily develop in the human speech, and easily disappear or get renewed with more expressive synonyms as time passes. For example, colloquial Russian doesn't use ‘to have’ and hardly has had any Romance or Germanic influence, yet it too has developed a newer perfect, of the type ‘at me it is done’, that is (1) with a past perfect participle on the one hand, but (2) with a Russian counterpart of the ‘to have’ construction on the other.

 

[berndf]

For example, colloquial Russian doesn't use ‘to have’ and hardly has had any Romance or Germanic influence, yet it too has developed a newer perfect, of the type ‘at me it is done’, that is (1) with a past perfect participle on the one hand, but (2) with a Russian counterpart of the ‘to have’ construction on the other.

This comparison has little significance. The Russian construct is formally different but semantically similar to the late Latin/Romance perfect while German has a perfect that is formally equal to that of the Latin/Romance perfect but does not express any perfect aspect.

 

[ahvalj]

But that's because the German perfect is old and has lost its resultative meaning on the way. The Germanic preterite comes from the post-Proto-Indo-European perfect as well, but it had lost its resultative meaning even earlier. That's basically the same story as with the French passé simple (< Latin perfect, originally a merger of aorist and perfect, that is with a partial resultative meaning sometime in the prehistory) and passé composé (< Romance perfect), both of which aren't resultatives anymore. The question is that some languages may have lost the old resultative meaning while not having developed a replacement, but that's another story (the Russian form I have mentioned is not literary, so the standard language doesn't have a resultative as well, and the old resultative is just a preterite nowadays). Southern Romance languages are simply more alive, they keep producing numerous periphrastic forms, which Germanic and Slavic aren't able to do at the moment.

 

[berndf]

But that's because the German perfect is old

No it is not.

The Germanic preterite

We are talking about the perfect and not the preterite. The preterite is inherited from Proto-Germanic. The perfect developed in late OHG and MHG.

 

[ahvalj]

But the compound perfect and past perfect are productive tenses in Old Norse.
An example from Barnes M · 2008 · A new introduction to Old Norse. Part I. Grammar: 156:

Þorfinnr vissi eigi, at Brúsi haþi upp gefit ríki sitt
‘Þorfinnr knew not that Brúsi had up given realm REFL. POSS.’
‘Þorfinnr did not know that Brúsi had surrendered his realm’

I don't think we'll ever be able to demonstrate the spread of early Romance patterns as far as to Old Norse.

 

[berndf]

But the compound perfect and past perfect are productive tenses in Old Norse.
An example from Barnes M · 2008 · A new introduction to Old Norse. Part I. Grammar: 156:

I don't think we'll ever be able to demonstrate the spread of early Romance patterns as far as to Old Norse.

The problem is that Olafs Saga Helga, like virtually all others to make the point of a "natural" perifrastic perfect in early Germanic languages, is a Christian text which is of course heavily influenced by Latin.

 

[Awwal12]

which is of course heavily influenced by Latin

It would be nice to see some other aspects of that.
On the other hand, what could we see in runic inscriptions?..

 

[berndf]

It would be nice to see some other aspects of that.
On the other hand, what could we see in runic inscriptions?..

I would be interested as well. Or if Icelandic texts then at least pagan and not Christian texts. Unfortunately, I don't know enough about Old Norse.

 

[ahvalj]

Antonsen EH · 1975 · A concise grammar of the older Runic inscriptions:

• makija maridai ala “Alla decorated [or: made famous] the sword”
• dagastiz runo faihido “Dagastiz painted the rune”
• ek hlewagastiz holtijaz horna tawido “I, Hlewagastiz of Holtaz, made the horn”
• hagiradaz tawide “Hagiradaz made”
• swabaharjaz sairawidaz [ek] stainawarijaz fahido ek hrazaz satido [ s ]tain[a] ana “Swabaharjaz with gaping wound. [ I ], Stainawarijaz, painted. [ I ], Hrazaz, set the stone on”
• ek wiwaz after woduride witadahalaiban worahto. [me]z woduride staina þrijoz dohtriz dalidun arbijarjostez arbijano “[ I ], Wiwaz, wrought after Woduridaz, the lord. For me, Woduridaz, three daughters, the most legitimate-to-inherit of heirs, prepared the stone”
• hadulaikaz ek hagustadaz hlaaiwido magu minino “Hadulaikaz. I, Hagusta[l]daz, buried my son”
• idringaz ek wakraz : unnamz wraita “Idringaz. I, Wakraz, the untakeable, wrote”
• ek wiz wiwio writum runo aisaz “I, we, of the descendents of Wiwaz, wrote the rune. Aisaz”

So, all old perfects > preterites.

 

[ahvalj]

Against the borrowed origin of the paradigm may speak the optional use of eigan (vs. habên) in Old High German:
si eigun mir ginomanan lioban druhtîn mînan “sie haben mir meinen lieben Herrn genommen”.

Also, Gothic has occasional examples with ‘to be’ like:
jah so baurgs alla garunnana was at daura “And all the city was gathered together at the door”
silba uswahsans ist “he is of age”
duhþe þai berusjos is qeþun þatei uswahsans ist, silban fraihniþ “Therefore said his parents, He is of age; ask him”.
Note the perfective prefixed verbs.

 

[berndf]

Of course Germanic has various ways of expressing perfectness as in the Gothic construct is ist uswahsans (he is in the state of having ended growing), which exists identically in modern German (er ist ausgewachsen) independently of the perfect tense. It is very much woven into the fabric of the Germanic tense system that tenses as such are aspect neutral and that aspect is only added as and where needed. That's why I am saying Germanic does not have any use for a perfect tense and is therefore unlikely to have spontaneously developed one, especially given the attested history of a perfect tense having developed in sync with Chistianisation.

 

[bearded]

If it were analogous to I am tall then writing would be a participle and not a gerund because the participle is an adjective but the gerund is a noun.

Is that a sure thing? It seems to me that in I am writing two elements speak against 'writing' being a noun: the absence of an article (would you say ''I am baker''?) and the fact that 'writing'' - like 'tall' - can be used as an adjectivized participle in attributive position, cf. 'a writing girl'.
In a phrase like 'writing is fine', writing is of course a gerund/noun, but it's a different case.

 

[berndf]

Is that a sure thing? It seems to me that in I am writing two elements speak against 'writing' being a noun: the absence of an article (would you say ''I am baker''?) and the fact that 'writing'' - like 'tall' - can be used as an adjectivized participle in attributive position, cf. 'a writing girl'.
In a phrase like 'writing is fine', writing is of course a gerund/noun, but it's a different case.

Yes, that's exactly what I've said: writing would be a participle (i.e. adjective) and not a gerund (i.e. noun).
But contrary to the apparent structure in modern English, the true historical development is from a gerund:

The real origin of the continuous form is from the gerund: I am on writing -> I am awriting -> I am writing. Compare German Ich bin am Schreiben.

 

[bearded]

that's exactly what I've said:

You said: 'if it were analogous to ''I am tall'..
So for you it's a gerund/noun. But in #27 I kind of doubted/questioned that.

 

[berndf]

You said: 'if it were analogous to ''I am tall'..
So for you it's a gerund/noun. But in #27 I kind of doubted/questioned that.

No, I haven't expressed any doubt. I have explained that despite the apparent structure involving a participle the actual historical development was from a gerund. It seems you have stopped reading my posts after the first sentence... or I have expressed myself very poorly.

 

[bearded]

I have explained that despite the apparent structure involving a participle the actual historical development was from a gerund.

My question was/is whether that is a 100% sure thing. I,too, expressed myself unclearly. You certainly have good sources for your statement...?
Wir haben teilweise aneinander vorbeigeredet.

 

[berndf]

Wir haben teilweise aneinander vorbeigeredet.

It seems like it. Your argument is about the modern structure and I agree with you that the most sensible analysis is as a participle.

Yet the modern form is derived from an older construction that is analogous to the German Ich bin am Schreiben: I am on writing > I am awriting where writing is obviously an abstract verbal noun, i.e. a gerund.

 

[PersoLatin]

Persian uses the auxiliary verbs ‘to be/بودن’ and ‘to have/داشتن’, for past perfect & imperfect and for present & past continuous respectively, so opposite to English in that respect.

 

[merquiades]

A few notes.

These perfect constructions tend to lose their resultative meaning with time, and so languages may develop fresher, originally more expressive, replacements instead. For example, Spanish and Portuguese at their current stage use both the older perfect with “to have” and a newer one with the etymological “to hold” — like for example the Spanish lo he hecho “I have done it” and lo tengo hecho ‘the same’, etymologically “I hold it done”. The exact relationships between both variants are different across the varieties of Spanish and Portuguese, but eventually the newer form is going to win, I guess.

The normal perfect forms in Spanish are with haber and bring the actions forward in time to the minute of talking. That's a bit different than in English or other languages. The verb haber is now defective and is lacking the proper meaning of "have". It's just used as an auxiliary. Likewise, you can't really split the "haber + participle" and the participle is invariable.
The special periphrasis with tener with the participle is different in many respects. Both forms are autonomous, keep their respective meanings, they can be split, and the participle (acting as an adjective) agrees in gender and number with the subject. This construction is more forceful than the perfect tenses and stresses that the verb (participle) has been retained, owned, possessed in one way or another.
I don't think the 'tener" structure will win out. It is only used in special contexts.
The verb "haber" ages and ages ago did mean "have" but it has been replaced entirely by "tener" which no longer means "hold".

Old Galician and indeed modern Galician did not have perfect tenses at all. Though Latin is not my forte I understand there were no such forms in the classical language. There is a synthetic pluperfect tense in Portuguese inherited from Latin but this switched to become a past subjective form in Spanish. Modern Portuguese has developed them in rather recent times and uses "ter". They are still not often used though and give off more of a perfect progressive idea.

He recibido una carta (Spanish) - I (have) received a letter
Tengo recibida una carta (Spanish) - I have indeed received a letter and it's in my possession

Tenho recibido cartas (Portuguese) - I have been receiving letters (more or less)
Recebi uma carta (Portuguese) - I (have) received a letter. Past simple still can also have a perfect sense.

 

[ahvalj]

I agree about the (Castilian) Spanish constructions, but it is my point that the one with haber is older (common Romance) and rather worn out, whereas that with tener is so far fresher and stronger and recalls the original meaning that resultative constructions carry when recently created in the language. The longer these constructions exist, the lesser remains of that resultative meaning. That's (virtually) universal across languages. Brazilian Portuguese has advanced much further in replacing haver with ter.

Consider the English case. Sat now is a simple past without any perfective meaning. However, its post-Proto-Indo-European ancestors, *soda (singular 1) and *sode (singular 3) were perfects, with the meaning “I have sat down (and am sitting)” (singular 1). In Proto-Indo-European, this was most probably even not a resultative but a special form expressing state, so that *sodhₐe meant “I am sitting”. So, the development in this particular case was state → attained state → past action with result continuing in the present → actual past → plain past.

 

[berndf]

Consider the English case. Sat now is a simple past without any perfective meaning.

You are aware that this is not what this thread is about? It is about the origin of the modern tense expressing the perfect aspect and not about the origin of an ancient tense expressing the perfective aspect.

 

[ahvalj]

The perfective aspect was conveyed by the aorist. The post-PIE perfect was — at some point of its evolution — a complete counterpart of the modern English present perfect. I am trying to illustrate the idea that this meaning is not stable, and exists as an evolutionary step developing most often towards simple past and sometimes (as in the Portuguese example in #34) to something more peculiar.

 

[berndf]

OK. I am still trying to understand how this relates to the question of this thread?