When including the following poetic examples from Horace and Ovid in what turned out to be a long answer to a previous post on datives of agent, I made this hesitant remark: Perhaps I'm wrong but I'd say that the following (non-canonical) datives of agent that can be found in poetry are not (expected) to be found in classical prose authors like Cicero or Caesar. Basically, only canonical datives of agent are typically found in these two authors (e.g. Haec vobis provincia est defendenda; Hoc mihi deliberatum est; cf. Habeo hoc deliberatum). Could anyone share his/her experienced intuition about examples like the following ones with me?

Prima dicte mihi, summa dicende Camena … Maecenas (Hor. epist. 1,1) ‘Proclaimed by me in my first poetry, to be proclaimed in my last, Maecenas, …’

adde preces castas inmixtaque vota timori, // nunc quoque te salvo persoluenda mihi. (Ov. epist. 6, 73-74) ‘Add chaste prayers and vows mixed with fear, which now I must fulfill, since you are safe.’

carmina ... quae scribuntur aquae potoribus (Hor. epist. 1,19,3) ‘The poems which are written by water drinkers.’

This poetic usage has been said to be traced back to Greek (for this proposal, see Brenous, J. (1895). Étude sur les hellénismes dans la syntaxe latine. (Edizione Anastatica, L’Erma di Bretschneider, Roma 1965.) Paris: Klincksieck); see also Tillmann, H. (1881). De dativo verbis passivis linguae Latinae subiecto qui vocatur Graecus. Acta Seminarii Philologici Erlangensis, 2, 71–140. For a brief summary of this tricky issue, see Calboli, G. (2009). Latin syntax and Greek. In P. Baldi & P. Cuzzolin (Eds.), New Perspectives on Historical Latin Syntax. Volume 1 Syntax of the Sentence (pp. 65-194). Berlin/New York: De Gruyter Mouton.

As noted above, I'd find it surprising if in Cicero's texts one could find examples like orationes dictae mihi or carmina scribuntur Terentiae, with these datives having an agentive interpretation (cf. the ablatives of agent a me and a Terentia). Clearly, in classical prose authors like Cicero or Caesar these datives in these examples can only have a beneficiary reading but not an agentive one. But how about with other verbs?

The three following alleged "datives of agent" provided below by cmw, Cerberus, and myself, respectively, are found with a productive set of acquisition verbs (e.g. expetere, quaerere, sumere, conciliare, emere, etc.), which crucially can also take a dative in the active. These examples can be easily found in Cicero but, unlike the ones above found in poetry, should not be considered as "Greek datives of agent", i.e. they are not an influence from Greek authors (of course, they are also different from typical/canonical datives of agent like the one found in Carthago nobis delenda est).

Sic dissimillimis bestiolis communiter cibus quaeritur (Nat. Deor. 2.123)

Sumatur nobis quidam praestans vir (Cic. Tusc. 5.68)

nos hunc Heracliensem multis civitatibus expetitum (...) de nostra civitate eiciemus? (Cic. Arch. 22)

cmw also provides an interesting example from Tacitus (see also the interesting answer by Cerberus to a related post). I should say that I deliberately included Caesar & Cicero in this Q to restrict the classical period a bit (e.g. later "datives of agent" can be found depending on non-verbal adjectives or even on active (!) verbs (Pinkster 2015: 248/9; vol. 1. Oxford Latin Syntax)). Non-canonical examples of datives of agent from an author like Sallustius would also be welcome.

  • Upon reading the previous thread, I'm not sure I'm understanding the distinction you make here. Can you clarify what's so special about those particular examples? The Ovid example is a fairly straight gerundive of obligation, is it not?
    – cmw
    Commented Jul 9, 2023 at 19:22
  • @cmw Thanks. Yes, the relevant context is provided in the related Q&A, which are too long to be summarized here: hence the link. Basically, the point is that datives of agent are typically found in contexts with the verb esse (e.g. Mihi hoc est deliberatum // Carthago nobis delenda est). Many/some of the examples that do not fall under this "typical" pattern (e.g. the ones included in the present post) are found in poetry and are probably influenced by Greek (v. previous linked Answer). Here I wonder if these "atypical" cases can also be found in Latin classical prose.
    – Mitomino
    Commented Jul 9, 2023 at 19:45
  • 1
    Got it, so just that third example I provided then. You can find it edited now having gotten rid of the extraneous bits.
    – cmw
    Commented Jul 9, 2023 at 19:49
  • Hmm in what way are the datives you provided different from the normal Latin dativus auctoris, why suppose a Greek influence? latin.stackexchange.com/questions/10584/…
    – Cerberus
    Commented Jul 9, 2023 at 22:30
  • @Cerberus Thanks for the link, which contains an interesting case of "dative of agent". What is a "normal" dative of agent? Well, if by "normal", one means "any dative whose meaning can be related to agency", the ones above are "normal". But that is not the point made here. Again, "normal"/typical datives of agent are those ones included in my reply to cmw's comment above. Are they different from the Greek ones exemplified in the Q above? Yes, probably they are. Please take a look at the link I provided in my Q above and read the previous post (both Q & A) to follow the thread. Thanks!
    – Mitomino
    Commented Jul 9, 2023 at 23:40

1 Answer 1


This is indeed found in Cicero:

  • Sic dissimillimis bestiolis communiter cibus quaeritur (Nat. Deor. 2.123)

Gildersleeve and Lodge mention that it's more frequently found in Tacitus, so that might be a good starting point to find more examples.

Edit: Found a clearer example in Tacitus:

  • sibi tamen apud horridas gentis e contuberniis hostem aspici.

It's not in a participial form, and aspicio doesn't take a dative. I don't see how it differs in form from your Ovid example:

dative of agent subject of the verb present passive
potoribus carmina scribuntur
sibi hostem aspici
  • Thanks for the answer! Note that the dative of your example is different from that of carmina scribuntur aquae potoribus (see ex. above). Verbs like quaero or expeto can take a dative in their active variants: mihi aliquid expeto/quaero, whereby in the passive aliquid mihi expetitur/quaeritur the dative can be preserved (an example similar to yours is: nos hunc Heracliensem multis civitatibus expetitum de nostra civitate eiciemus? (Cic. Arch. 22)).
    – Mitomino
    Commented Jul 9, 2023 at 20:07
  • @Mitomino I don't think this is the correct assessment. In the active, you have three nouns (nominative, included in the verb, accusative as a direct object, and dative of interest), but in the passive you only have two. The subject and the dative of interest merge, as they're the same, leaving behind the expected direct object to become the subject of the passive verb.
    – cmw
    Commented Jul 9, 2023 at 20:15
  • What you say "the subject and the dative of interest merge, as they're the same" can be said to be a good intuition but, believe me, this kind of "merge" is syntactically impossible. Anyway, your point goes well with Kühner & Stegmann's (II 1, 324-325) descriptive proposal that datives like this express, at the same time, the agent (Germ. «Urheber») and the beneficiary. For a criticism of this "blending/mixing" view, see the one by Serbat (1996) included in this post: latin.stackexchange.com/questions/18518/…
    – Mitomino
    Commented Jul 9, 2023 at 20:35
  • @Mitomino I remember that question. I haven't read Serbat in full, but no real evidence is provided for those assertions. I remember not being convinced, at least by what you presented. Maybe I'll come across an example in Tacitus at some point and revisit it then.
    – cmw
    Commented Jul 9, 2023 at 20:40
  • Ok. Many thanks for the discussion! Yes, Tacitus is a very good source to find many interesting innovations. As for Serbat's point, I agree with him in that we should not mix/blend "agent" and "beneficiary" in our definition of "dative of agent" but I don't agree with his proposal that this dative syntactically depends on the participial morpheme -t- or on the -nd- morpheme, basically because (true) datives of agent are not typically found in non-verbal contexts (except in poetry; cf. the first two exs. above and the related post linked in my question above).
    – Mitomino
    Commented Jul 9, 2023 at 20:51

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.