Why should Ibn Ezra assert that the word "Am" is not construct?

Summary: Ibn Ezra has a seemingly strange comment, that in am benei Yisrael it is duplication, rather than construct form. What could compel this? I examine two reactions, and consider a semantic, rather than syntactic cause. Also, the trend across Tanach such that this is a common expression. Then, I consider that it just might be the influence of trup.

Post: In parashat Shemot, Pharaoh is concerned about the nation of Israel, and relates those concerns to the Egyptians. As the pasuk in the first perek of Shemot states:

ט  וַיֹּאמֶר, אֶל-עַמּוֹ:  הִנֵּה, עַם בְּנֵי יִשְׂרָאֵל--רַב וְעָצוּם, מִמֶּנּוּ.

9 And he said unto his people: 'Behold, the people of the children of Israel are too many and too mighty for us;

Note that the JPS translation is "the people of the children of Israel". That is the construct form. Ibn Ezra rejects this interpretation, and instead asserts that it means "the nation, the children of Israel":

הנה עם -
מלת עם איננה סמוכה רק בני ישראל פירוש ותוספת ביאור כמו: אהיה אשר אהיה.
אשר התהלכו אבותי לפניו אברהם ויצחק.

Thus, bnei Yisrael is commentary and additional explanation, like Ehkeh, that will Ehkeh. And that your fathers walked before him, {namely} Avraham and Yitzchak. Thus, it is not the construct form, but two phrases conveying the same meaning.

This is strange. From where does Ibn Ezra know this?
Is there something about the pasuk that conveys it, or is it something about the Hebrew language? In terms of the word am, there is a patach under it in both the absolute form and the construct form, such that nothing about the nikkud could help differentiate.

Indeed, Ibn Caspi sort of challenges Ibn Ezra on this. He writes:

ח) הנה עם בני ישראל. מלת עם, אם סמוכה או לא, הטעם בזה הנה עם מבני ישראל

The word 'am, whether or not it is in the construct form -- the intent in this is behold, there is a nation, from the children of Israel.

 Thus, he does not even see any semantic difference between one and the other. Such that he appears to argue with the idea that it is tosefet biur.

Mechokekei Yehuda, in his commentary Karnei Ohr, writes that the Mevaer to Netivot HaShalom writes regarding והנה עם, that even though the word 'am is found quite often with a patach even though it is not in construct form, here it seems more plausible to explain that it is a construct. And there is to explain Bnei Yisrael as referring to the 12 sons of Yaakov, and its import is the nation of the twelve sons of Yisrael, who were born, and that there increased from them such that they became a nation.

Thus, by seeing the contrary position, we might understand what Ibn Ezra sees, and what Ibn Ezra is trying to argue against. If indeed 'am is synonymous with benei yisrael, then benei yisrael must mean the nation called that, rather than the children of Yaakov. And conversely, if bnei Yisroel means the nation, it does not make sense too much sense to say that nation of the children of Israel, for they are themselves the nation. Thus, it is tosefet biur.

Perhaps. But how does he know this? Perhaps he did not consider the alternative, namely children of Yaakov, and so was forced into this grammatical assertion.

I think another possibility is that he considered the phrase hineh 'am as it occurred across Tanach, and while sometimes there is elaboration of the 'am, having it in construct form just messes up the resonance of the phrase. Thus, we have our pasuk:

שמות פרק א

• פסוק ט: וַיֹּאמֶר, אֶל-עַמּוֹ:  הִנֵּה, עַם בְּנֵי יִשְׂרָאֵל--רַב וְעָצוּם, מִמֶּנּוּ. 

but also

במדבר פרק כב

• פסוק ה: וַיִּשְׁלַח מַלְאָכִים אֶל-בִּלְעָם בֶּן-בְּעֹר, פְּתוֹרָה אֲשֶׁר עַל-הַנָּהָר אֶרֶץ בְּנֵי-עַמּוֹ--לִקְרֹא-לוֹ:  לֵאמֹר, הִנֵּה עַם יָצָא מִמִּצְרַיִם הִנֵּה כִסָּה אֶת-עֵין הָאָרֶץ, וְהוּא יֹשֵׁב, מִמֻּלִי. 

and

שופטים פרק ט

• פסוק לו: וַיַּרְא-גַּעַל, אֶת-הָעָם, וַיֹּאמֶר אֶל-זְבֻל, הִנֵּה-עָם יוֹרֵד מֵרָאשֵׁי הֶהָרִים; וַיֹּאמֶר אֵלָיו זְבֻל, אֵת צֵל הֶהָרִים אַתָּה רֹאֶה כָּאֲנָשִׁים. 
• פסוק לז: וַיֹּסֶף עוֹד גַּעַל, לְדַבֵּר, וַיֹּאמֶר, הִנֵּה-עָם יוֹרְדִים מֵעִם טַבּוּר הָאָרֶץ; וְרֹאשׁ-אֶחָד בָּא, מִדֶּרֶךְ אֵלוֹן מְעוֹנְנִים. 

and finally

ירמיהו פרק ו

• פסוק כב: כֹּה אָמַר יְהוָה, הִנֵּה עַם בָּא מֵאֶרֶץ צָפוֹן; וְגוֹי גָּדוֹל, יֵעוֹר מִיַּרְכְּתֵי-אָרֶץ. 

ירמיהו פרק נ

• פסוק מא: הִנֵּה עַם בָּא, מִצָּפוֹן; וְגוֹי גָּדוֹל וּמְלָכִים רַבִּים, יֵעֹרוּ מִיַּרְכְּתֵי-אָרֶץ. 

Ibn Ezra might be thinking about all those examples, and might have concluded that hinei am must stand my itself. If so, Bnei Yisrael must be an interjection, explaining just which nation.

Another possibility which struck me was that the trup influenced his explanation. The trup on the pasuk is:

Thus, on hinei 'am benei Yisrael, the trup is revia' yetiv munach zakef-gadol. Within that, the revia first divides off, and then the yetiv. Therefore, the division is:

 הִנֵּה | עַם בְּנֵי יִשְׂרָאֵל

and then within the latter phrase:

עַם  | בְּנֵי יִשְׂרָאֵל

Perhaps because of the conjunctive accent of yetiv, Ibn Ezra thought that it could not be the construct. That is, we should have a mesharet there instead. And if this is indeed what the trup is stating, perhaps it got it in turn from the semantic content, or else how the phrase appears elsewhere. However, I don't think this trup actually necessitates, or supports, this reading. The rules of continuous dichotomy demand a continuous dichotomy so long as there are three words in a phrase or subphrase left undivided. And this is just what 'am benei Yisrael was. And the best place to subdivide such a phrase is on the word 'am, despite the semichut.