On Fri, Jan 24, 2020 at 12:03 PM Alan Anderson <qunchuy@alcaco.net> wrote:
Someone has already quoted the very end of TKD section 6.2.5.

> Finally, the use of {rIntaH} to indicate that an action is accomplished (section 4.2.7) is another example of the two-verb (or two-sentence) construction.

Based on this information, I treat {rIntaH} as its own sentence. I do not place it between another sentence’s verb and its subject.

paq'batlh, p. 180-181, lines 22-23:
ghIq Qo'noSDaq Qap rIntaH 'e' Sov
qeylIS
Then, the time came when Kahless saw
That his work on Kronos was done,
If rIntaH is its own sentence, then it's what 'e' would refer to, and the phrase would be interpreted as:
ghIq Qo'noSDaq Qap. rIntaH 'e' Sov qeylIS. Then he worked on Qo'noS. He knew that it was accomplished.
This doesn't make much sense, because the ghIq clearly is intended to apply to the accomplishment of Kahless's work, not the work itself. Kahless's work on Qo'noS started long before this point in the story, so it doesn't make sense to introduce it with ghIq, "subsequently, then".

If rIntaH is a part of the sentence with Qap, then 'e' would refer to everything before, and the phrase would be interpreted as:
ghIq Qo'noSDaq Qap rIntaH. 'e' Sov qeylIS. Then he finished working on Qo'noS. Kahless knew that.
This fits the translation and the context of the sentence.
 
{qar'a'} is explicitly said to be able to follow the sentence *or* to appear immediately after the verb, and is not mentioned in the section on complex sentences. I won’t extrapolate from it to {rIntaH} (though Andrew Strader obviously did).

I suspect the primary reason {qar'a'} is not mentioned in the section on complex sentences is that it was introduced in the Appendix to TKD, and thus did not exist when the section on complex sentences was written.

Like {qar'a'}, {rIntaH} is also explicitly said to follow the verb.
That is, instead of using the suffix {-ta',} a special verbal construction can follow the verb which indicates the accomplished action.
I know that Okrand's use of terminology in TKD was sometimes imprecise, but I think he could probably tell the difference between "follow the verb" and "follow the sentence".

Another piece of circumstantial evidence: In section 6.2.5, there are explicit explanations or glosses pointing out that phrases with 'e', net, neH, and verbs of saying are considered as two sentences:
{qama'pu' DIHoH 'e' luSov} <They know we kill prisoners.>
This sentence is actually two: (1) {qama'pu' DIHoH} <We kill prisoners> ({qama'pu'} <prisoners,> {DIHoH} <we kill them>); (2) {'e' luSov} <They know that> ({'e'} <that,> {luSov} <they know it>). The pronoun {'e'} refers to the previous sentence, <We kill prisoners.>
 
{Dalegh vIneH} <I want you to see him/her.>
({Dalegh} <you see him/her,> {vIneH} <I want it>)
 
{qaja'pu' HIqaghQo'}
or {HIqaghQo' qaja'pu'} <I told you not to interrupt me.>
This is literally <I told you, "Don't interrupt me!"> or <"Don't interrupt me" I told you> ({qaja'pu'} <I told you,> {HIqaghQo'} <don't interrupt me!>).

This isn't done for rIntaH. The example sentence
{luHoH rIntaH} <they have killed him/her> ({HoH} <kill>)
is not followed with an explanation like
This is actually two sentences, literally meaning <They killed him/her. It remains finished.>
further suggesting that despite being a two-verb construction, it's not considered the same sort of thing as 'e' and verbs of saying and so on. 

Lastly, section 6.2.5 also says
In complex sentences of this type, the second verb never takes an aspect suffix (section 4.2.7).
rIntaH is the second verb, but it's 50% aspect suffix, suggesting that the rIntaH construction is not a complex sentence of that type.