quid Paris? ut salvus regnet vivatque beatus cogi posse negat

[Aurifex]

But then what of the passive infinitives of the regular thematic verbs? Any complete explanation would have to explain both.

I'm not sure what your point is. In your opening post you asked specifically about the infin. passive of 3rd conj. verbs, not infin. passives generally.
Leaving aside the earlier by-forms with -ier -rier endings, passive infinitives are originally dative sing. of either a root stem (cogi etc.) or an S-stem, e.g. ama-ri (Old Latin ama-rei) from *ama-sai, vide-ri from *wide-sai.

 

[Abbatiſſæ Scriptor]

Thanks I was wondering whether they were originally 's' stem like active infinitives, or 'r' stem like most passives. The effects of rhoticism always obscure such things.

 

[Abbatiſſæ Scriptor]

Final question: An infinitive is essentially a verbal noun, and the form does look like a dative; but why should it have been the dative case of such a verbal noun that fossilised into an indeclinable infinitive?

 

[Pacifica]

I was asking myself the same thing. Would it be because it first expressed purpose? I don't know, I'd like to know too.

 

[socratidion]

While this thread has been changing direction, I've happened across an instance of oratio obliqua with ellipsis of the (effectively) reflexive pronoun:

'non possum dicere nihil perdere, sed quid perdam et quare et quemadmodum dicam' (Sen. Epist.1.1.4) = I can't say that I am losing nothing, but I will say what I am losing and why and how.

 

[Aurifex]

Final question: An infinitive is essentially a verbal noun, and the form does look like a dative; but why should it have been the dative case of such a verbal noun that fossilised into an indeclinable infinitive?

This may shed some light on it.

 

[Abbatiſſæ Scriptor]

This may shed some light on it.

Much thanks, but Fruyt's approach to the question is not all that illuminating, at least not to me. The connexion proposed between the dative and the locative makes fairly easy sense to a student of Latin. As the article goes on to discuss the 'prospective dative', however, it becomes less clear, and less convincing, especially in the argument that the reduced form of the preposition 'ad' which survives in Latin's daughter languages now governs the dative case (even though 'ad' took the accusitive for as long as case endings survived). There is an unacknowledged elephant in the room here: to wit, the latent question of why the dative, and not the much more easily expected accusative?

 

[Aurifex]

There is an unacknowledged elephant in the room here: to wit, the latent question of why the dative, and not the much more easily expected accusative?

The accusative was used, in classical Sanskrit to form all infinitives (e.g. dha-tum), and in Latin for the first supine, which, besides being used alone after verbs of motion to indicate a goal, was combined with impersonal iri to form the future infinitive passive.

 

[Pacifica]

Examples would be nice.

I've re-found one in Tacitus after searching for uros:

iuravere amici dextram morientis contingentes spiritum ante quam ultionem amissuros.

And others:

simul populi ante curiam voces audiebantur: non temperaturos manibus si patrum sententias evasisset.

Haud defuere qui certatim, si cunctaretur Caesar, vi acturos testificantes erumperent curia.

haec atque talia dicenti adstrepere vulgus, gentili quisque religione obstringi, non telis, non vulneribus cessuros.

namque ab lacu Averno navigabilem fossam usque ad ostia Tibernia depressuros promiserant squalenti litore aut per montes adversos.

quibus intrari curiam placebat, securos esse de constantia eius disserunt; nihil dicturum nisi quo gloriam augeret.

non expectavit militum ardor vocem imperatoris; bonum haberet animum iubebant: superesse adhuc novas viris, et ipsos extrema passuros ausurosque.