R A Brown wrote:

> It meant "unlimited, indeterminate, indefinite." The confusion dates 
> back to the earlier times when the difference between tense (i.e. time) 
> and aspect was not made, and the indicative tenses were seen in some way 
> to be the 'essential part' of the verb, with subjunctive, infinitives 
> and all the rest being the 'trimmings.'
> 
> The adjective _aóristos_ was originally applied to past perfective 
> indicative tense, or rather the grammarians distinguished between 'past' 
> forms, thus:
> - imperfect [past] (i.e. the past imperfective)
> - indefinite [pas] (i.e. the 'aorist' or 'past perfective)
> - perfect [past] (i.e. the _present_ perfect - the confusion is due to 
> the fact that the Latin "perfect tense" does duty for both the past 
> perfective and the present perfect - also in Koine Greek, the two ideas)
>  were often confused).
> - more then perfect [past[ ('plus quam perfectum' or, more simply, the 
> 'pluperfect', i.e. past perfect).
> 
> So we already have a good deal of tense/aspect confusion; then to add to 
> the merriment, the other moods and non-finite parts all get label by 
> names used for the indicative tenses; thus all the other imperfective 
> forms carry the epithet 'present', tho they have no reference per_se to 
> the present time! E.g. present subjunctive, present imperative, present 
> infinitives etc.
> 
> Similarly all the other perfective forms got the label 'aorist' even tho 
> none referred to an indefinite past! So we have the aorist subjunctive, 
> aorist imperative, aorist infinitive.
> 
> It was impressed many-a-time upon us schoolboys in the 1950s that 
> 'present' had nothing to do with present time and 'aorist' had nothing 
> to do with past time when it came to the non-indicative moods and the 
> non-finite parts of the verb (tho I'm fairly certain that the linguistic 
> term 'aspect' was not used by any of our teachers).
> 
> So the practice of labeling various parts of the ancient Greek verb 
> accounts for meanings (2) and (3) given by Trask:
> "2. A verb form marked for both past tense and perfective aspect.
> "3. A verb form marked for perfective aspect."
> 
> The first meaning given by Trask - "1. A verb form marked for past tense 
> but unmarked for aspect" - arose, I think, from supposed identification 
> of the Greek "aorist indicative" with the 'past simple' of English.
> 
> As for Trask's fourth meaning - "4. A conventional label used in a 
> highly variable manner among specialists in particular languages to 
> denote some other particular verb form or set of verb forms. For 
> example, Lewis (1967) uses the term to label those Turkish verb forms 
> marked for durative/habitual aspect ....." - I guess a case of "here's 
> an odd verbal label that already got one or two different meanings, so 
> let's add another one'      ;)
> 
> Comrie recommends that the term be not used in linguistic theory, to 
> which I say 'Amen!' I would FWIW also strongly recommend not using it 
> when describing the features of any Conlang       :)

Tirelat has (or rather had) an ending -šaj, labeled "aorist". It seems 
to have originally been a past tense form unmarked for aspect. In the 
recent revamping of the Tirelat documentation, I seem to have mistaken 
it for a non-past perfective (and used it as such on one of my web 
pages). At least I don't think I've used it anywhere else; I fixed the 
web page to use a more common non-past perfective form.