In Understanding Poetry: An Anthology for College Students, Cleanth Brooks and Robert Penn Warren (New York: Henry Holt and Company, 1938) dictated the order in which a poem of historical or political significance should be treated in the classroom. What was "prior" was the poem as a poem; then and only then could it "offer illumination" as a "document." Such illumination was possible, at least in the abstract, but literary significance must be comprehended first (and never mind the notion — which was in fact just then of interest to a number of poets ranging from Reznikoff to Rukeyser to Pound to Norman Rosten — that a poem could consist of documents or be a kind of document itself.)
Notwithstanding this statement of priority, the anthology is full of poems (Andrew Marvell's, for instance) that at least initially draw our attention to them because of their compelling historical subject matter; our sense of the beauty of the poem as a poem could follow secondarily. Really. Why not? The order in which these two sorts of value are ascertained does not affirm or refute the New Critical ban on historical readings, for, at least here, the interpretation of historical significance is permitted right from the start.
Here's a passage from the prefatory "Letter to the Teacher," written in 1938:
This book has been conceived on the assumption that if poetry is worth teaching at all it is worth teaching as poetry. The temptation to make a substitute for the poem as the object of study is usually overpowering. The substitutes are various, but the most common ones are:1. Paraphrase of logical and narrative content;
2. Study of biographical and historical materials;
3. Inspirational and didactic interpretation.
Of course, paraphrase may be necessary as a preliminary step in the reading of a poem, and a study of the biographical and historical background may do much to clarify interpretation; but these things should be considered as means and not as ends. And though one may consider a poem as an instance of historical or ethical documentation, the poem in itself, if literature is to be studied as literature, remains finally the object for study. Moreover, even if the interest is in the poem as a historical or ethical document, there is a prior consideration: one must grasp the poem as a literary construct before it can offer any real illumination as a document.