In a recent question it was pointed out that there is no passive gerund, and cupido amandi can't mean the desire of being loved. But I'm not sure that this is the case. I tend to view the gerund as signifying the action itself hence it may be used both actively and passively.
This is not the first time I wonder about this issue as seen in my addendum of this post. In which it seems the passive usage of the supine is fine.
Just to add more color and examples to what given in @Tyler Durden's answer which opened the door for me to find more information.
A Latin Grammar for the Use of Schools(p.368) by Johan Nikolai Madvig:
§ 418. Sometimes the gerund is employed less accurately, so as to have the appearance of a passive signification, inasmuch as it either (especially in the genitive) merely designates the action of the verb in general, and so takes the place of a substantive (e. g. movendi for motūs), or is referred in idea to some other agent than the grammatical subject of the proposition: Multa vera videntur neque tamen habent insignem et propriam percipiendi notam (Cic. Acad. II. 31), mark by which they can be known. Antonius, hostis judicatus, Italia cesserat; spes restituendi nulla erat (Corn. Att. 9), =restitutionis or fore, ut restitueretur. Jugurtha ad imperandum Tisidium vocabatur (Sall. Jug. 62), that he might receive orders. Annulus in digito subtertenuatur habendo (Lucr. I. 313), by our wearing it. (Facilis ad intelligendum; see §. 412, Obs. 3.)
The idea, though controversial, seems to be that this is "less accurate" usage and it is not a normal usage. Here (p.182 57) is how Bellum Jugurthinum interprets the above note in his own note on ad imperandum:
Some grammarians unnecessarily consider the gerund used passively here. Madvig, § 418, says, Sometimes the gerund is employed less accurately so as to have the appearance of a passive signification.
There is a discussion of this issue by Benjamin Hall Kennedy, where he brings out more examples:
Crescendi causa haec frequentia convenit (Cic.)
Ceteris, quae moventur, hoc principum est movendi (Sall.)
Alitur vitium vivitque tegendo (Virg.)
Finally, in "An Advanced Latin Syntax" (1919) By A. L. Francis, H. F. Tatum has a concise summary of gerund in p.66
- The Gerund usually expresses an action of the main subject, as vires acquirit eundo Virg. Aen. iv. 175: but sometimes it is impersonal, as nomen ipsum carendi Cic. Tusc. i 36 87. Sometimes the object of the gerund is the same as the subject of the verb, so that the gerund appears to be passive: as alitur vitium vivitque tegendo Virg. Geo. iii 454: lentescit habendo ii 250: cantando rumpitur anguis Ec. 8 72: ante domandum ingentes tollunt animos Geo. iii 206. Or the gerund and verb have the same object and not same subject: pueros excercendi causa producere Liv. v 27 2
To summarize: We have examples of the gerund used passively, though they are few in number. Grammarians, for a reason I could not understand, prefer to call it "appears to be passive" where I never see them condemn the regular case as "appears to be active"
I will quote from Himes' Grammer (1894) v2 p. 494:
The Gerund used Passively: The Gerund is occasionally used passively, e.g. Antonius hostis judicatus Italia cesserat spes restituendi ("of being restored") nulla erat. Erudiendi gratia ("for the sake of being taught") Athenas missus erat.
However, your phrase cupido amandi I think would normally mean "desire of loving", not "desire of being loved". The gerundive is the future passive and agrees with its subject, like an adjective, femina amanda = a woman to be loved.
