In my recent studies of the Hebrew verb ויהי, I came across a familiar tension between textual and linguistic studies. In a number of places (1 Sam 10:5; 2 Sam 5:24; 1 Kgs 14:5; 1 Chr 14:15; Rut 3:4), the MT uses the imperfect verb וִיהִי when we would expect the weqatal verb וְהָיָה, which normally precedes fronted temporal clauses in non-past contexts. Textual critics commonly emend these forms to their more common forms, whereas linguists commonly insist on explaining them as the only surviving evidence. In this methodological gridlock, who is right?
First, it is important to note that our corpus and understanding of Biblical Hebrew is comparatively limited, so we must remain open to challenging received wisdom with new linguistic insight. It is all too easy to emend away difficult evidence which, properly understood, might shed light on the language and text. We must remain open to exceptions to our grammatical rules and to new insights from modern linguistic studies. We must not simply create the evidence for our linguistic studies with premature emendations.
That said, we must understand the biblical text in its historic context as a text manually transmitted over a vast period of time by scribes of varying characteristics. Some scribes were careful to preserve the text exactly as it lay before them, whereas others felt free to update the text with contemporary spellings and linguistic conventions. We can see this dynamic even in the examples mentioned above, now that we have the Qumran evidence. 4QSam(a), for instance, apparently reads the expected והיה for the MT ויהי in 1 Samuel 10:5. 1QIsa(a), on the other hand, occasionally changes the older form והיה to the updated form ויהי, as in 29:15 and 56:12. Such changes as proposed above, therefore, did in fact occur in the transmission of the text. Examples like these show that linguists are wrong simply to assume the MT text in their analyses, as many of the unique and exceptional forms may have arisen in the course of transmission. Linguists who do so run the risk of proposing explanations that unrealistically merge different temporal stages of Hebrew into a single incoherent mush.
So in the end, I would say neither approach will be consistently right, but rather that they should mutually inform one another. At times, linguistic research will clarify difficult textual problems. At other times, examination of ancient manuscripts, the transmission of the text, and the historical development of the language will provide linguistic studies with a more solid evidentiary basis. The two should always remain in a cooperative tension.