By taking a look at various translations of the sentence in bold below, which is excerpted from a famous portrait of Jugurtha by Sallust, one could infer that the datives luxu (cf. luxui) and inertiae seem to be interpreted as depending on corrumpendum (e.g., cf. the translations of this fragment you may have at hand). I've seen that some commentators even make the explicit (but wrong: see below) proposal that these two constituents are "datives of agent". Others do not consider luxu a dative but say that there is a variatio involved here. E.g. see the following comment in French: "Deux constructions différentes pour le complément d’agent de l’adjectif verbal: un ablatif: «luxu» et un datif: «inertiae»" (see the source here). A similar view is expressed by a different commentator here. As for the dative form luxu, see also Dario's relevant comment below: unlike what is often said, it is not an archaic form but an analogic innovation (used also by Caesar, Vergil and others).
Qui ubi primum adolevit, pollens viribus, decora facie, sed multo maxime ingenio validus, non se luxu neque inertiae corrumpendum dedit, sed, uti mos gentis illius est, equitare, iaculari; cursu cum aequalibus certare et, cum omnis gloria anteiret, omnibus tamen carus esse; ad hoc pleraque tempora in venando agere, leonem atque alias feras primus aut in primis ferire: plurimum facere, et minimum ipse de se loqui. (Sal. Jug. 6).
However, it seems to me that, to the extent that so-called "datives of agent" cannot appear in non-verbal contexts (see here and here) and to the extent that these datives are typically animate, a better analysis would be to interpret them not as "datives of agent" depending on corrumpendum but rather as directly depending on the verb dedit (i.e., literally: 'he didn't dedicate himself to the luxury and indolence to be corrupted/to corrupt'), corrumpendum being then equally a predicative gerundive of se with the typical meaning of purpose but lacking any complement. As pointed out in the last link on predicative gerundives, corrumpendum could be claimed to be an adjunct here but seems to be obligatory in a similar example like Epaminondam pecunia corrumpendum suscepit (Nep. 15,4,1) ‘He undertook the bribing of Epaminondas with money’. However, it is not obvious (at least to me) how one should analyze corrumpendum in these two examples: e.g. to what extent can the adjunct/"non-dominant" vs. obligatory/"dominant" distinction be maintained for them? Traditional grammars just provide typical lists of verbs that go with these predicative gerundives, which are indeed useful to learn Latin but do not provide a clue to syntactic questions like the one I've just raised (e.g. see cnread's list below based on Gildersleeve & Lodge or the list included in my last link above on predicative gerundives, which was drawn from Woodcock (1959)).
Another interesting grammatical issue regarding Sallust's example above is whether the gerundive corrumpendum must be interpreted as passive or active. I'd analyze it as passive but see here for another view.
Could you please let me know your (reasons for your) preferred analysis and translation of the sentence in bold above?
NB: I remember that Joonas put forward an interesting and very original analysis for these predicative gerundives but I've been unable to find it! I only recall that he argued for a unifying analysis of these gerundives, which is then different from the non-uniform approach I've sketched out and exemplified above (cf. the adjunct/"non-dominant" analysis involved in Sallust's example vs. the non-adjunct/"dominant" one involved in Nepos' example).
