There are a couple of points which I find interesting to add to
Thorsten Renk's "The Q(u)enya Past Tense" at:
About the _-ie_ past tense:
Its presence in suffixed pa.t. forms of "The Lost Road" (e.g.
_kárielto_ 'they made' vs. _káre_ '[he] made'), is comparable to the
development of the pa.t. explained at "The Qenya Verb Forms". There, pa.t.
of "regular verbs" is normally marked by _-ie_ plus the suffixes which
denote gender, number and voice; but impersonal singular active past
tense has _-e_ < _-ie_ (PE14:31, _i_ glide and breve over the _e_ in
the primitive form). Thus, _tul-_ 'to bring, come' is _tûle_ 'came'
(imp. sing.), but _tûlien_ or _tûliendo_ 'he came', etc. (PE14:28,
macron over the _u_ in all cases).
Additionally, though it is not a great finding at all, it may be worth
noting that the _-ie_ past tense did not really disappear, but became
a perfect tense (cp. _tûlien_ 'he came' cited above vs. _utúlien_ 'I
am come' in _Eärello Endorenna utúlien..._, LR:967).
On the other hand, in the last paragraph just before the
acknowledgements, Thorsten says:
"In the QL, all such verbs [with stem in _-r-_] form past tense by
vowel lengthening, best examplified by #_kar-_ 'to make, do' pa.t.
_káre_ (PE12:45). But already in the EQG, this has been changed to a
suffix _-ne_, cf. _kar-_ 'to make' pa.t. _karne_ pr.t. _kare_
(PE14:58). In the Qenya of 'The Lost Road', the past tense form
_ohtakáre_ (LR:47) appears, suggesting a change of mind back to vowel
I think, however, that the scenario of these facts may not be Tolkien
trying to settle whether pa.t. of _kar-_ should be _káre_ or _karne_,
but something closer to what Thorsten said previously, in the
introduction about cases from _Etym._:
"verbs are not limited to one past tense, we find an example in which
alternative past tenses exist without apparent differences in meaning,
cf. _onta-_ 'to beget, create', pa.t. _óne_, _ontane_ (LR:379)..."
In relation to this, it may be interesting to look at the
"English-Qenya Dictionary", where there is a handful of verbs
conjugated in present and pa.t. (there labelled "preterite"), and
"Do. '_kara_': _kare_, pret. _kârie_, but pret. is also frequently
[All Qenya forms are in Valmaric script and then transliterated, with
the exception of the base _kara_, which is untransliterated by
Tolkien; _kârie_ with macron over the _a_.]
There we see the pa.t. form _karne_ already found in EQG, PE14:58 (EQG
being closely related to that dictionary, cf. PE15:65-66), as an
alternative to _kârie_, which conforms the conjugation of "regular
verbs" like _tul-_ (pa.t. _túlie_) in EQG (PE14:57). This proves that
distinct pa.t. forms were possible, though the text of EQG only showed
_karne_. In addition, though the pa.t. _kârie_ of the English-Qenya
Dictionarly does not occur in 'The Lost Road' as such, it does occur
in the suffixed form _kárielto_, which is apparently in complement to
unsuffixed _káre_, _ohtakáre_, etc. (see above).