Fairly frequently in Latin one encounters a "backwards" comparison, in which the relative pronoun in the ablative precedes the term of comparison.

...philosophiam ad te adlegem, qua nec mihi carior ulla umquam res in vita fuit ... (Cic. Fam. 15.4.16)

There are many similar phrases: quo nihil optatius, qua nihil utilius, etc.

My question is whether this construction ever occurs without the negative. A&G seems to imply that it does, but the example given in fact contains a negative:

Section 407

Note 3— Relative pronouns having a definite antecedent never take quam in this construction, but always the ablative

Rēx erat Aenēās nōbīs, quō iūstior alter nec, etc. (Aen. 1.544) Æneas was our king, than whom no other [was] more righteous, etc.

(I'm not even sure what "in this construction" means, unless it's just referring back to "having a definite antecedent.")

The reason I think A&G implies that there are such positive constructions is that the very next subsection deals with negatives.

However, I have not been able to find one.

  • 1
    This is a fascinating question: what I thought would be an easy search has, so far, turned up no results. Part of the explanation is that quo, without a negative, strongly suggests either a comparative purpose clause or a correlate, e.g. quo...eo or quo...hoc. It takes a negative to reverse the assumption for me.
    – brianpck
    Jan 29, 2018 at 3:10
  • Right? I had no idea I was stumbling onto a thorny question. I thought of it b/c I was composing in Latin and wanted to say, referring to the entirety of the last clause, X is easier than that. quo facilius X est. It makes logical sense, but maybe it's just not done. Jan 29, 2018 at 5:47

1 Answer 1


According to Gildersleeve and Lodge, Latin grammar §296, Remark 1.b.2:

The Abl. [of comparison] is very common in negative sentences, and is used exclusively in negative relative sentences.

So, although a noun or pronoun other than a relative pronoun is often used as an ablative of comparison in negative sentences, the relative pronoun is used in this way only in negative sentences.

So according to the second clause of this rule, if the sentence is both negative and relative, only the ablative is used, never quam; and according to the first clause, if the sentence is negative but non-relative, either quam or ablative can be used (and the ablative is 'very common'). Of course, this rule doesn't address the remaining two permutations: a sentence that is both non-negative and non-relative, and a sentence that is non-negative but relative; in both these cases, the implication is that either quam or ablative can also be used.

Here are two examples of a non-negative relative sentence that uses the ablative:

Cicero, Pro Milone 33.90:

(The relative here is a so-called 'connecting relative.')

an ille praetor, ille vero consul, si modo haec templa atque ipsa moenia stare eo vivo tam diu et consulatum eius exspectare potuissent, ille denique vivus mali nihil fecisset cui mortuo unus ex suis satellitibus curiam incenderit? quo quid miserius, quid acerbius, quid luctuosius vidimus?

Would that man when praetor, much more when consul, provided only that these temples and these walls could have stood so long if he had been alive, and could have remained till his consulship; would he, I say, if alive, have done no harm, when even after he was dead he burned the senate-house, one of his satellites, Sextus Clodius, being the ringleader in the tumult? What more miserable, more grievous, more bitter sight have we ever seen than that?

Cicero, Philippicae 1.4.10:

exque eo primum cognovi quae Kalendis Sextilibus in senatu fuisset L. Pisonis oratio: qui quamquam parum erat – id enim ipsum a Bruto audieram – a quibus debuerat adiutus, tamen et Bruti testimonio – quo quid potest esse gravius? – et omnium praedicatione quos postea vidi magnam mihi videbatur gloriam consecutus.

And it was from him that I first heard what had been the language of Lucius Piso, in the senate of August; who, although he was but little assisted (for that I heard from Brutus himself) by those who ought to have seconded him, still according to the testimony of Brutus, (and what evidence can be more trustworthy?) and to the avowal of every one whom I saw afterwards, appeared to me to have gained great credit.

(Translations by C. D. Yonge; from the Perseus website)

  • 1
    Thanks. This doesn't quite settle it for me, because the wording is ambiguous. "used exclusively" here can mean that, in negative relative sentences, it is the only acceptable form. You cannot use quam. I'm reading it this way b/c that exact rule is in other grammars. Jan 29, 2018 at 13:56
  • @Kingshorsey. Sorry, I don't see the ambiguity; can you explain? The 2nd clause of the rule states that if the sentence is both negative and relative, only the ablative is used, never quam; and the 1st clause states that if the sentence is negative but non-relative, either quam or ablative can be used (and the ablative is 'very common'). What's taken as granted here, of course, is that either quam or ablative can also be used in any non-negative sentence (either relative or non-relative). Certainly this conforms to my own experience from reading Latin.
    – cnread
    Jan 29, 2018 at 22:31
  • OK, I agree with that. I thought you were reading G&L as requiring a negative in order to use the relative pronoun. "What's taken as granted here, of course, is that either quam or ablative can also be used in any non-negative sentence (either relative or non-relative)." This is what I'm challenging. I can't find any examples a relative pronoun being used apart from a negative, when the pronoun precedes the term of comparison. I can find plenty examples of quo nihil optatius, but none of quo optatius, in which quo is functioning as a comparative. Jan 29, 2018 at 23:20
  • @Kingshorsey. Ah, I see. I'm updating my answer with 2 examples, just from Cicero, of non-negative relative sentences that use the ablative.
    – cnread
    Jan 30, 2018 at 5:45
  • 2
    @brianpck, Good point. It's also interesting that the use of ablative in non-neg. rel. sentences seems to be largely (or perhaps exclusively) confined to interrogative sentences; and if you look at my the two examples, at least, they're both 'virtually negative' – Cicero is really saying 'quo nihil miserius, nihil acerbius, nihil luctuosius,' and 'quo nihil...gravius.' So, getting back to the original question, the construction actually does, in a way, seem to be reserved for negatives. That overlap or affinity of negatives and interrogatives would be worthy of separate investigation, I think.
    – cnread
    Jan 30, 2018 at 18:11

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service and acknowledge you have read our privacy policy.

Not the answer you're looking for? Browse other questions tagged or ask your own question.