Jonathan Culler's On Deconstruction contains a very interesting review of the literature on readers and reading. The idea is to figure out the relationship between the reader of the text and the text itself. Does the text control and condition the reader's responses, or does the reader construct her own text? This question has been central to much recent literary theory, since any act of interpretation is, first and foremost, an act of reading.
The study of historical documents--primary sources--is also an act of reading. It operates in a different way than the reading of literary texts; the historian does not read in the "fiction" mode--she doesn't try to figure out the traps the author may have set for her, or the underlying themes or message of the text (at least, not themes in the same sense as in literature). Yet she does not usually read texts as if they were simply shopping lists, mere vehicles for information devoid of subtext or stylistic content.
Historians, especially after Evans' In Defense of History, have taken to dismissing the theoretical approaches developed in the past three decades as either something they'd mastered a long time ago or something that's frivolous and not worth paying attention to. This is somewhat unfortunate. While few professional historians now adhere to a rigid notion of the objectivity of historical facts and interpretations, a certain defensiveness has set in; it reacts against Derridean versions of the "linguistic turn," declaring adamantly that in the case of history, there is something outside the text.
The objection may well be valid, and certainly the Derridean case has been overstated with regard to its pragmatic implications. Still, it is important for historians to take into account insights developed in other disciplines about the process of research and interpretation. Applying theories of reading to the historian's craft can provide a way to sidestep acrimonious metaphysical and broadly epistemological disputes about whether there is a knowable historical truth, and focus more narrowly on the tactical epistemology of reading sources. Historians would benefit from a more judicious view of their own reading strategies; if they are made aware of the role of the reader's contribution to the understanding of a text, or of the text's intrusion into interpretation, they can make their conclusions and analyses more "accurate," or at least develop a grounded notion of their own margins of error.
A study on Reading History would probably demand an interdisciplinary collaboration between literary studies scholars and historians. The former would have to be more ecumenical about their notions of historical truth, and the latter would have to be willing to question their own methods. The result could be very positive, perhaps leading to broader cooperation between the disciplines (in this regard, I think literary studies and history have a much greater collaborative potential than, say, literary studies and analytic philosophy). The project would benefit literary studies as much as it would history: the application of reading theories to this particular edge case may reveal dimensions of the problem hitherto unacknowledged.