The so-called “double dative construction” contains a "dative of purpose" (e.g. maxumo terrori in ex. (1) below) and a personal dative (e.g. Numantinis in (1)) that turns out to be affected by the former. The "dative of purpose" is typically an abstract noun that functions as a predicate, whereby it is often referred to as a “predicative dative”. As shown by Baños (1995: 15; 2021: 224) and Pinkster (2015: 780), i.a., examples like (1), (2) and (3) can be taken as clear evidence for the predicative status of the (non-problematic) "dative of purpose".
(1) in tantam claritudinem brevi pervenerat ut nostris vehementer carus, Numantinis maxumo terrori esset (Sall. Iug. 7,4)
(2) M. Antonius ad me tantum de Cloelio rescripsit, meam lenitatem et clementiam et sibi esse gratam et mihi voluptati magnae fore. (Cic. Att. 14, 19, 2)
(3) imperio quod… ipsis principibus prosperum vel exitio fuit (Tac. Hist. 2, 1, 1).
In contrast, the syntactic status of the personal dative in these constructions is much more disputed in Latin linguistics. For example, the following three syntactic analyses have been put forward in the literature for the personal dative Numantinis in the ex. (1) above:
a) Numantinis is a dative of possession and depends on esset (Suárez Martínez 1992)
b) Numantinis depends on the complex predicate formed by the predicative dative plus the verb: [maxumo terrori esset] (Dahl 2014)
c) Numantinis is an adnominal dative: in (1) it depends on maxumo terrori (Baños 1995; 2021).
The a) analysis (the “dative of possession” analysis) can be said to receive support from the typical parallelism with the transitive paraphrase with the possessive verb habere: cf. Numantinis maxumo terrori esset (Sall. Iug. 7,4) with Numantini eum maxumo terrori haberent (similarly, cf. Iste ceteris Siculis odio est (Cic. Verr. 2, 4, 15) with Ceteri Siculi istum odio habent). I found the dative of possession analysis appealing until I came across transitive double dative constructions like the ones exemplifed in (4) and (5), where the possessor subject of the verb habere would not be expected to be reduplicated (?) by an alleged “dative of possession”, contrary to fact:
(4) quod viro esse odio videas, tute tibi odio habeas (Plaut. Men. 111)
(5) equitatum sibi praesidio habuit. (B. Hisp. 2.3)
Although the complex predicate analysis (see the b) analysis above) can be said to be sometimes intuitive for the double dative construction with the verb esse (e.g. see the parallelism between [auxilio esse]+personal dative and auxiliari+personal dative), it is however less obvious for non-copular constructions like the ones exemplified in (6) and (7):
(6) Ariobarzani simul cum Agesilao auxilio profectus est . . . (Nep. Tim. 1.3).
(7) Venire Carthaginiensibus auxilio (Liv. 30, 36, 8).
Rather, as pointed out by Pinkster (2021: 800), in cases like (6) and (7) it seems more intuitive to take the personal dative as an argument of the predicative dative, in tune with the adnominal analysis of the personal dative (see the c) analysis above put forward by Baños (1995, 20021)).
Finally, as pointed out by Baños (1995: 26), it is worth pointing out that one piece of evidence for the c) analysis (i.e. the adnominal analysis of the personal dative), is that the predicative/abstract noun can also appear in a non-dative case, e.g. the accusative case of exitium in (8). However, Baños (1995: 28) himself acknowledges that this situation is not frequent since, when the predicative/abstract noun is marked with a non-dative case, the personal dative is often replaced by a noun in the genitive case (e.g. see (9)). So it is unexpected in the adnominal analysis that the personal dative can only typically modify the predicative/abstract noun iff it is also in the dative case. Why should it be so? As far as I can tell, Baños provides no answer to this question.
(8) eluant hanc maculam irasque civiles in exitium hostibus vertant (Tac. Ann. 1,43, 12).
(9) triarii … in praesidio castrorum manere iussi (Liv. 40, 27, 7)
All in all, it seems that there is NO consensus in the literature about the correct analysis of the personal dative in double dative constructions. Do you know about further evidence for one of the three analyses (a, b, & c) above? Do you know about useful references on this tricky issue of Latin syntax beyond the ones given below? Thanks in advance!
REFERENCES
Baños, J. M. (1995). “El ‘doble dativo’ en latín.” In M. E. Torrego, P. J. Quetglas & E. Espinilla (eds.). Sintaxis del dativo latino. 11-34. Barcelona: UB Publications.
Baños, J. M. (2021). “7. Dativo.” In J. M. Baños (coord. ed.). Sintaxis Latina (vol. 1). 197-226. Madrid: CSIC.
Dahl, E. (2014). “On the semantics and syntax of the Latin ‘double dative’ construction.” In N. Oettinger & T. Steer (eds.). Das Nomen im Indogermanischen. 32-50. Wiesbaden: Reichert.
Pinkster, H. (2015 & 2021). The Oxford Latin Syntax (vol. 1 & vol. 2). Oxford: OUP.
Suárez Martínez, P. M. (1992). “Dono dare: el reflejo de un espejismo sintáctico en latín.” Emerita 60: 31-40.
