Accūsātīvus / Datīvus cum Īnfīnītīvō

Gregorius Textor

Animal rationale

  • Civis Illustris

  • Patronus

Location:
Ohio, U.S.A.
Ørberg has this example of ACI (Fam. Rom. XI, 179):

(a) Puerum dormīre necesse est.

When I blanked out the endings (Puer-- dorm-- necesse est) I wanted to answer with dative case:

(b) Puerō dormīre necesse est.

I think that would also be correct (though of course not ACI), but with a slightly different meaning.

He gives a similar construction in XX, 4:

nam longus somnus īnfantī tam necessārius est quam cibus.

I understand (b) as "It is necessary for (i.e., for the benefit of) the boy to sleep."

I understand (a) as "It is necessary that the boy sleep", which could also be said as "It is necessary for the boy to sleep", but although using the word "for," does not necessarily mean that it is "for the boy's benefit": it might be, e.g., that a gang of criminals need to put the boy to sleep so that he doesn't witness their nefarious deeds.

Comments?
 

kizolk

Civis Illustris

  • Civis Illustris

Location:
Bourgogne, France
I had asked a question about that to ScorpioMartianus/Polymathy's Luke Ranieri a few months ago, so I might as well share it:

Me: The combination accūsātīvus cum īnfīnītīvō + necesse esse is tricky: in the previous chapter, it said that necesse esse called for dative (cibus hominī necesse est), but in that chapter we can see things like "puerum dormīre necesse est". Spontaneously I would have used puerō, and that's typically the kind of mistakes I can easily see myself doing in the future...

Him: Right. The difference is “it’s necessary for a person to breathe” as opposed to “it’s necessary that a person breathe”



From his answer, I came to pretty much the same conclusion as you:

it might be, e.g., that a gang of criminals need to put the boy to sleep so that he doesn't witness their nefarious deeds.
But I'm still not entirely sure, and I think if I needed to say something like what's in this quote, I would try to word it differently.
 

Pacifica

grammaticissima

  • Aedilis

Location:
Belgium
cibus hominī necesse est
That's wrong, or at the very least extremely unusual. The subject of necesse est normally isn't a regular noun but an infinitive or infinitive clause or a neuter pronoun that stands for it (e.g. hoc necesse est where hoc means some previously mentioned, yet-to-be-mentioned, or implied action).

As for the difference between acc. and dat., it's subtle but I think it's along the lines of what's been said above. With an accusative-and-infinitive clause, you're stating that something must happen, be done, etc. The dative makes it more personal in a way: the person in the dative is forced to do so and so.
 
The first example

a) Puerum dormīre necesse est.

is an impersonal construction, whose subject is the entire phrase necesse est, with necesse in the neuter singular. My grammar book tells me impersonal constructions generally use Acc c Inf.

But in the second example

nam longus somnus īnfantī tam necessārius est quam cibus.

the subject is somnus. So the construction is not impersonal and there is nothing to trigger Acc c Inf.
 

Gregorius Textor

Animal rationale

  • Civis Illustris

  • Patronus

Location:
Ohio, U.S.A.
kizolk dixit:

cibus hominī necesse est

That's wrong, or at the very least extremely unusual. The subject of necesse est normally isn't a regular noun but an infinitive or infinitive clause or a neuter pronoun that stands for it (e.g. hoc necesse est where hoc means some previously mentioned, yet-to-be-mentioned, or implied action).
So here one would have to say cibum ēsse necesse est, I guess.

The first example

a) Puerum dormīre necesse est.

is an impersonal construction, whose subject is the entire phrase necesse est, with necesse in the neuter singular.
Thanks for your comment, but I don't understand. I would think the subject of this sentence is Puerum dormīre, and the predicate is necesse est.

Maybe because I do not understand the distinction between personal and impersonal constructions.
 

Pacifica

grammaticissima

  • Aedilis

Location:
Belgium
Kizolk, I took my explanation from my Latin grammar book, which states (though in Swedish): Acc c Inf is used with impersonal expressions, e.g.:

Insidias factas esse constat. - It is well known that deceits have been performed.

Non utile est captivos reddi. - It is not favourable that the prisoners are returned.

As you see, the English (and Swedish) main clauses use the dummy subject it, which makes these clauses impersonal, that is, they require a dummy subject and a verb in the neuter singular.

This way of reasoning works well for me, but I see now that you make me think about it that the Latin sentences don’t actually meet the definition of impersonal. So the explanation in my book might in fact be plain wrong, even though it works well as a strategy to translate from (quite a few) modern languages. Thank you for pointing it out!
 

Pacifica

grammaticissima

  • Aedilis

Location:
Belgium
Those Latin constructions are indeed called impersonal. Where did kizolk say the contrary?
 

kizolk

Civis Illustris

  • Civis Illustris

Location:
Bourgogne, France
Yeah I think you might be replying to something Gregorius Textor said, Linnaeus, not me!
 
He didn’t. He just made me realise that the concept impersonal expression as I am used to defining it, is not applicable in Latin, at least not in the examples in this thread or in my book.

So they are called impersonal in Latin after all? Why? I mean, what is the proper definition?
 

Pacifica

grammaticissima

  • Aedilis

Location:
Belgium
So they are called impersonal in Latin after all?
Yes.
Why? I mean, what is the proper definition?
The definition of an impersonal construction is a construction where the subject isn't a person or thing denoted by a noun but an action, fact or the like (practically speaking that can be an infinitive, a clause, or a neuter pronoun standing for such).
 
Aha, and so it seems my book’s explanation is compatible with Pacifica’s and kizolk’s.

I’ve read somewhere that the subject of an infinitive is always in the accusative. Is this a generale rule? If so, it also describes the Acc c Inf in indirect statements and the like (Dico/Video te venire), and it quite a handy rule to remember.
 

Pacifica

grammaticissima

  • Aedilis

Location:
Belgium
I’ve read somewhere that the subject of an infinitive is always in the accusative. Is this a generale rule?
It is usually in the accusative. But not absolutely always. The subject of a historical infinitive (have you heard of that?) is nominative.
 

Gregorius Textor

Animal rationale

  • Civis Illustris

  • Patronus

Location:
Ohio, U.S.A.
As you see, the English (and Swedish) main clauses use the dummy subject it, which makes these clauses impersonal, that is, they require a dummy subject and a verb in the neuter singular.
I was worrying about this, because it seems that it would be risky to rely on the vagaries of English or Swedish (where we use "it") to resolve a Latin grammar question. Would "It's raining" then be impersonal?

So they are called impersonal in Latin after all? Why? I mean, what is the proper definition?
I'm glad you asked this.

The definition of an impersonal construction is a construction where the subject isn't a person or thing denoted by a noun but an action, fact or the like (practically speaking that can be an infinitive, a clause, or a neuter pronoun standing for such).
Glad to hear this. To check my understanding, I'll propose a few examples (positive and negative), but in English, even though we are talking about Latin grammar (or maybe the concept applies to English too?):

Positive (impersonal constructions):
1. Eating vegetables is good for you.
2. It's good for you to eat vegetables.
3. To eat vegetables was his highest joy.
4. (The fact) that Bob ate no vegetables caused his illness.

Negative (not impersonal constructions):
5. Bob ate only meat and potatoes.
6. He couldn't digest the meat well.
7. It's raining and thundering.
8. There's a tornado coming.
 

Pacifica

grammaticissima

  • Aedilis

Location:
Belgium
Hm, well, I've heard the "it's raining" kind of construction called impersonal as well, because the "it" doesn't really stand for anything in particular. My definition was probably not quite complete.
 

Pacifica

grammaticissima

  • Aedilis

Location:
Belgium
1. Eating vegetables is good for you.
3. To eat vegetables was his highest joy.
I'm also not sure those sentences would usually be called impersonal, though they're certainly similar to "it's good for you/it was his highest joy to eat vegetables", which I believe would be classified as such. It's a bit weird and, well, I guess the term "impersonal", like much terminology, has its ambiguities and is ultimately of limited usefulness. It is however useful to know which Latin verbs and constructions are exclusively impersonal, at least if you want to write in Latin, so that you don't end up wrongly using them with a personal subject like in cibus necesse est above. In Latin, for a verb or construction to be exclusively impersonal does mean, as a practical matter, that its subject can only be one of these things: 1) an infinitive; 2) a clause; 3) a neuter pronoun standing explicitly or implicitly for one of those; 4) some vague, unnamed "it" not representing any tangible thing. Some verbs, however, can be used in both personal and impersonal ways; and some usually-not-impersonal verbs can be found in constructions that can be argued to be impersonal. Sum, for instance, isn't instrinsically impersonal; but many teachers would call a sentence like melius est discedere impersonal (just like necesse est discedere, for that matter; but unlike necesse, melius can be used with sum in personal constructions). Now what about errare humanum est? Is it impersonal? It's not really different in its structure from melius est discedere, and yet it feels a little different somehow. Perhaps this is where the question becomes academic and of no real interest (because who cares as long as you understand the sentence and/or can correctly use the words that compose it?)
 
Last edited:
Top