North & Hillard Ex. 204; Q1: the following is to be translated into Latin: "If he had not mocked me, I should perhaps have forgiven him." (Impossible conditions: past tense: pluperfect subjunctive in both clauses.)

The Answer Book: "nisi mihi irisisset, forsitan ei ignossem."

Curious as to the deployment of "mihi" given that "irrideo" does not take dative, I arrived, by circuitous route (Q: https://latin.stackexchange.com/a/12946/1982 and Q: The grammar of the expression "mihi cordi est") at "Dative of Reference" (D of R) from Allen & Greenough (p376-379 original; p234 reprint): "The dative often depends, not on any particular word, but on the general meaning of the sentence. The dative in this construction is often called the Dative of Advantage or Disadvantage, as denoting the person or thing for whose benefit or to whose prejudice, the action is performed."

This solves the above example. (Not quite: see Mitomino's answer and my answer.)

On the same page the serendipitous discovery of an interesting piece of translation: "laudavit mihi fratrem" = "he praised my brother, out of regard for me"; A & G state: "laudavit fratrem meum" would imply no such motive."

Although this is a (D of R) the "out-of-regard-for-me" part reads like a leap of faith. How, without context, is it determined? Would anyone, here, have translated in this way?

Initially, it appears, from the example from (N & H) and this one that a (D of R) is NOT translated as a regular dative (to or for something).

Adding, therefore, to the confusion, the remaining examples, from A & G, appear to do the opposite. Consider: "meritos mactavit honores, taurum Neptuno, taurum tibi pulcher Apollo" (Aen. III. 118) = "he offered the sacrifices due, a bull to Neptune, a bull to you beautiful Apollo".

Datives "Neptuno" & "tibi" are translated as regular datives. (Why isn't it "pulchro Apollini"?)

What is going on, here?

4 Answers 4


Typically, the so-called "dative of reference" is not semantically selected by the predicate (linguistically speaking, it is an adjunct, i.e., it is not an argument). In order to understand what all examples of the datives of reference that appear in the first link from A&G have in common, it is crucial to realize that they are not semantically selected by the verbal predicate (i.e., they are adjuncts, whereby they are typically omissible). See also this link from Panhuis (2006: 91-92) for useful discussion on which kind of "arguments" are typically selected by verbal predicates.

So, according to the typical definition of "dative of reference" above, note that the dative that you've found to be associated to irridere in your "Answer Book" (nisi mihi irrisisset, forsitan ei ignossem) cannot be considered an example of this type, i.e., mihi is not a dative of reference. The dative mihi, if real/attested (cf. infra), is not an adjunct but is to be regarded as semantically selected by the prefixed verb in-ridere (> irridere), whereby this nominal has argumental status. Quite importantly, please note that the dative associated to compound/prefixed verbs has argumental status (unlike the so-called "dative of reference", which, as pointed out above, is an adjunct).

Interestingly, you point out that the dative of irrisisset is not real/attested since this verb takes accusative but not dative (a nice and intriguing point, indeed!). In any case, note that the dative mihi or accusative me, when associated to this prefixed verb, is an argument rather than an adjunct. In my opinion, this pronoun is semantically selected by the prefixed verb irridere (hence its argumental status).

As for your last example from Virgil, which is the fourth example in the first link above (meritos mactavit honores, taurum Neptuno, taurum tibi, pulcher Apollo (Verg. Aen. III. 118)), note that pulcher Apollo is vocative (NB: one can add an interjection like "O beautiful Apollo!").

Finally, let me make a couple of bibliographical recommendations: in my opinion, the traditional grammar by A&G is indeed very useful. However, two more recent pedagogical (NB: not advanced) Latin grammars, which include very basic linguistic up-dates like the important argument-adjunct distinction above, i.a., are the following (NB: the former adopts a functionalist perspective, whereas the latter uses a generative approach).

Panhuis, Dirk (2006). Latin Grammar. Ann Arbor: University of Michigan Press.

Oniga, Renato (2014). Latin: A Linguistic Introduction. Oxford & New York: Oxford University Press.

  • Thank you. I understand that "mihi" is an (object) argument, it is NOT an optional extra. The "Datives-with-Compounds" seems to represent indirect objects: "praeficere imperatorem bello" = "to prefer a man as Commander-in-Chief (direct object) in a war (indirect object; dative). Both of these grammatical concepts tempt the student to use accusative, direct object, "me" instead of "mihi", in (N & H) "nisi mihi irisisset…". Sadly, I am still not 100% clear as to the deployment of "mihi". What if a student uses "me"?
    – tony
    Commented Dec 11, 2019 at 13:02
  • Re. Q: latin.stackexchange.com/q/12785/1982 one suggestion: "quomodo te habes" = "How do you regard you?" = "How do you do?" After studies on the dative, how about: "quomodo tibi habes?" = "How do you regard yourself?" Is this valid?
    – tony
    Commented Dec 11, 2019 at 13:19
  • @tony As for your first question, the verb irridere takes accusative in the attested/real examples (so the student should use "me"). The fact that N&H use dative in their invented (?) example nisi mihi irrisisset, ... is to be attributed to the fact that the dative with compound/prefixed verbs (see the A&G link above) is also a usual (and indeed quite productive!) construction in Latin.
    – Mitomino
    Commented Dec 11, 2019 at 19:14
  • @tony As for your second question, note that the verb habere is strongly transitive, whereby the example quomodo tibi habes? sounds quite bad in Latin.
    – Mitomino
    Commented Dec 11, 2019 at 19:17
  • After further studies: A & G; 362(a): "Many verbs have a transitive & intransitive use and take either accusative + dative, or dative alone." Ex. "concidere amicis quidquid velint" = "to grant to friends all they may desire"; "amicos" might have been used here. Does this parallel the N & H example?
    – tony
    Commented Dec 12, 2019 at 11:01

Miltomino has already analyzed in detail the other cases you proposed, so I'll only linger over this interesting example:

On the same page the serendipitous discovery of an interesting piece of translation: "laudavit mihi fratrem" = "he praised my brother, out of regard for me"

This is the so called dativus ethicus (or "of interest"): it indicates emotional participation by a person with respect to the action or circumstance expressed by the predicate; it is always expressed by an atonic pronoun and is not necessary for the syntactic-grammatical completeness of the sentence (Salvi 1988: 65-66). Thus, for example, in the words of Cicero:

quid mihi Tulliola agit?


tu mihi istius audaciam defendis?

This form of dative still remains in languages such as Italian (as an expression, since there are no grammatical cases for Italian), but also in English, where a suitable example could be "My dog died on me" or in Brazilian Portuguese, where the same sentence is "O cachorro/cão me morreu".

  • 1
    Thank you. A literal translation: " he praised (my) brother----for me": it begins to, almost, make sense. For full clarity, the "missing link" must be along the lines of liking/ respect/ regard--QED. Thanks for pointing the way. Cicero: "What is Tulliola doing (for me)."; "You are defending, of all things, the art of risk-taking (for me)." The Portuguese: "The dog died (for me)." Learned a new Portuguese word, "pirralha", today. Bolsanaro's description of Greta Thunberg.
    – tony
    Commented Dec 11, 2019 at 13:12
  • Ahahah, now I've learnt a new word, too! Commented Dec 26, 2019 at 20:18

Firstly, many thanks to Mitomino who has gone to some lengths, with this Q.

Further research may have produced a solution to this "me" vs "mihi" conundrum.

The prefix "ir", applied to parent verb, "rideo" makes hardly any difference (Pock. Ox. Lat. Dict.). Wiki: prefix "ir" is an alternative form of "in", used to create the opposite meaning e.g. in + reverens (reverent) gives irreverens (irreverent).

Any ideas on this?

There are some ten "special verbs" in Latin which require a noun in the dative case. Compound verbs take dative objects, though it would be more accurate to say the compounds (i.e. the prefixes) applied to the parent verbs do.

Verbs, which take dative objects (technically, indirect objects) not accusative ones, even though the English verbs most often used to translate them require accusative direct-objects, give English translations which do not correspond directly to the Latin.

The English includes an accusative direct-object; the Latin, a dative object, giving rise to the very confusion which spawned this Q.

Such verbs have an underlying meaning e.g. "credo" = "believe"; to the Romans it meant "be trusting"; which, naturally, requires a dative: "be trusting to...";

"ignosco" = "pardon"; to the Romans "grant pardon"; requiring dative: "grant pardon to...";

"impero" = "command"; "give an order to...";

"noceo" = "harm"; "do harm to...";

"parco" = "spare"; "be lenient to...";

"pareo" = "obey"; "be obedient to...";

"persuadeo" = "persuade"; "make sweet or agreeable to...";

"placeo" = "please"; "be pleasing to...";

"servio" = "serve"; "be a servant (or slave) to...";

"studio" = "study"; "be zealous (or eager) for...".

Finally, the example from North & Hillard: "nisi mihi irisisset, forsitan ei ignossem" =

"If he had not been mocking/ rude/ bad to me (mihi), I should perhaps have forgiven him."


  • A couple of clarifications on your answer: (i) the negative prefix (e.g., the one in irreverens 'in-reverens') and the directional one (e.g., the one in irridere 'in-ridere') must be distinguished; (ii) the prefix of irridere does make a difference wrt ridere. Note that, unlike the unprefixed verb, the complex verb becomes directional. This said, you're right when saying "compound verbs (sc. like irridere) take dative objects, though it would be more accurate to say the prefixes applied to the parent verbs do".
    – Mitomino
    Commented Dec 18, 2019 at 16:28
  • @Mitomino: The conclusion, therefore, is that the directional (prefixed) verb requires the dative "mihi irisisset"; the non-directional (parent) verb requires accusative direct-object "me risisset". Is this correct?
    – tony
    Commented Dec 20, 2019 at 12:18
  • What I've pointed out is that the prefixed verb irridere can take accusative and, less frequently, dative, the latter being due to the directional prefix. The reason why some prefixed verbs can take a "directional dative" is quite complex (ddd.uab.cat/pub/cjol/cjol_a2017v16/cjol_a2017v16p19.pdf ). Furthermore, another conclusion is that it's not the case that all datives are created equal: some are selected by the directional prefix, others are selected by the verbal root (e.g., nocere), others are not selected but are adjuncts (e.g., so-called 'datives of reference'), etc.
    – Mitomino
    Commented Dec 21, 2019 at 0:39
  • According to dictionaries, the unprefixed verb ridere can take accusative and dative but, for the purposes of the present question, I would not analyze the latter as a "dative of reference".
    – Mitomino
    Commented Dec 21, 2019 at 1:03

There is an alternative way to approach "irrideo". The prefix "ir" could be a contraction of infinitive, "ire" = "to go". (Is it?) NO, IT ISN'T (The student tries things: sometimes they work; sometimes they don't. See Mitomino's comments. The logic is correct: prefix "ir" is directional, so the following is good.). A verb-of-motion (or a directional prefix) provides direction (to/ for [dative-governing prepositions] something/ someone), by definition. Therefore, the application of "ir" to parent verb "rideo", gives the directional compound-verb, "irrideo". This takes a dative (indirect) object.

In "nisi mihi irisisset, forsitan ei ignossem" the direction is clear: the mockery was being directed to the subject argument (I/ me/ myself), requiring dative (indirect) object, "mihi".

Attestation of the link between the dative and verbs-of-motion:

"it clamor caelo" = "the cry went up to heaven" (Verg. Aen. 5. 540).

  • 2
    The prefix ir- is not a "contraction" of the infinitive ire. It's just an assimilated form of the preverb IN-. For the so-called "assimilation" of prefixes, see department.monm.edu/classics/courses/CLAS224/Handouts/…
    – Mitomino
    Commented Dec 24, 2019 at 16:33
  • 1
    One indication that it isn't the infinitive ire (and it definitely isn't) is the fact that unassimilated forms of this verb (and related words) that start with inrid- occur in great abundance in the extant literature.
    – cnread
    Commented Dec 24, 2019 at 18:03
  • @Mitomino: Does preverb "in" ("ir") provide the required sense of direction?
    – tony
    Commented Dec 25, 2019 at 0:21
  • @tony Yes, it's the same directional prefix you can find in, e.g., irrumpere (in-rumpere 'to break in').
    – Mitomino
    Commented Dec 25, 2019 at 1:03

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.