Was Pharaoh's heart hard, or did it become hard?

Summary: Evaluating Rashi's emendation of Onkelos, from an itpa'el verb to an adjective, on the basis of the dikduk of the Hebrew word being translated.

Post: In Vaera, in Shemot 7:14, Pharaoh's heart is hard, kaved lev par'o.

יד  וַיֹּאמֶר יְהוָה אֶל-מֹשֶׁה, כָּבֵד לֵב פַּרְעֹה; מֵאֵן, לְשַׁלַּח הָעָם.

14 And the LORD said unto Moses: 'Pharaoh's heart is stubborn, he refuseth to let the people go.

This would be an adjective, it seems. But Onkelos renders it otherwise, as an itpael verb, which is thus reflective (/passive):

ז,יד וַיֹּאמֶר יְהוָה אֶל-מֹשֶׁה, כָּבֵד לֵב פַּרְעֹה; מֵאֵן, לְשַׁלַּח הָעָם.

וַאֲמַר יְיָ לְמֹשֶׁה, אִתְיַקַּר לִבָּא דְּפַרְעֹה; סָרֵיב, לְשַׁלָּחָא עַמָּא.

This seems strange, but Rashi comes to the rescue, emending Onkelos to match the apparent grammar of the verse, and giving a parallel verse to demonstrate it:

is heavy: Heb. כָּבֵד. Its Aramaic translation is יַקִיר [heavy], and not אִתְיַקַר [has become heavy], because it is the name of a thing [an adjective and not a verb], as in for the matter is too heavy (כָבֵד) for you (Exod. 18:18).		כבד: תרגומו יקיר, ולא אתיקר, מפני שהוא שם דבר, כמו (שמות יח יח) כי כבד ממך הדבר:

Unless he is really just selecting the better of two Targumim. Now, Rashi's grammatical judgement need not be the beginning and the end. Ibn Ezra disagrees with him, writing:

ז, יד]
ויאמר כבד -
פועל עבר. כמו: כי זקן יצחק. 

That is, it is a past tense, intransitive verb. Ibn Caspi, as well, thinks that it could be a verb.

כבד לב פרעה, הוא פעל עבר או תאר, והכל עולה לענין אחד:

Thus, it could either be a past-tense verb or else an adjective, and it does not really matter, because ultimately it means the same thing.

But of course, it does matter, if we care about establishing the correct girsa in Targum Onkelos. The grammatical point was the cause of Rashi's emendation, or at least selection of one girsa over the other. But if a plausible argument can be raised for either, then we could readily imagine the Targumist could have chosen it, even if in truth the word was an adjective.

If we look in our own printed Mikraos Gedolos, we see that it follows Rashi's correction. Thus, yakkir liba de-Phar'o.

Mendelssohn, in Nesivos Hashalom, writes as follows:

That is, first he explains it as a past-tense verb, like Ibn Ezra. Then, he cites Rashi that its Targum is yakkir rather than ityakkar, and explains that in old manuscripts / nuschaot of Targum Onkelos, it was yakkir, but in new nushchaot it was ityakkar. "Because it (kaved) is the name of a thing", giving examples. He explains that by this Rashi means an adjective of the heard, and that the Radak in shorashim, shorash kaved says this as well.
He continues that according to those who say it is a past-tense verb, it is of the verbs which receive action by way of happenstance, and that they are neither transitive nor instransitive verbs -- but they are closer to passive than the regular active, and so was the verb translated by them into Aramaic in the mishkal of hitpael, etc. And Rashi did not see fit so explain it as past-tense verb, for this would imply that it would occur after a time, after the performance of the wonders, and this would be against the truth, for his heart was hard even initially. But then he gives his own theory of dikduk by which this works out, and various things are resolved.

Shadal in Ohev Ger takes issue with this:

כבד לב פרעה אתיקר לבא דפרעה ( מא"ר יא"ר לסבונא רע"ה —
סביוניטה אנוירשא ג"ת וח"ג ) , ורשי הגיה יקיר , והלכו אחריו קצת
המדפיסים , אבל טעות הוא מה שכתוב בנתה"ש : בנסחאות ישנות
של תרגום אנקלוס הוא יקיר , ובחדשות אתיקר , ולפי דעתי כבד הוא
זמן עבר , בעדות מלת מאן שאחריו , שהיא זמן עבר , ואם היה כתוב
ממאן היה אפשר לומר כי כבד שם התאר [ קע"א איתקף ] .ש

That is, he says that some of the printers followed in the path set out by Rashi. And thus he argues with Mendelssohn who claimed that this was indeed an earlier girsa in Onkelos. In terms of actual dikduk, because of the word מאן in the next phrase, he feels that כבד must be a past-tense verb.

In terms of the girsa which has itkaf, which is parallel to what we find in Targum Yonatan, and it rather a literal translation for chazak, a related word -- I wonder whether this might play into the development of various girsaot, or if it itself derived from there. Switch the tav and the yud to arrive at the itpael form. And many of the same letters are present, and the peh sofit resembles, slightly, a resh.