“Poetic expression,” says the sugar-coated-pill theory, “is the honey that makes palatable the medicine of content, be it philosophical, moral, or scientific.” It’s an old theory, evident even in Greek and Roman theory and practice. For its defense, a team of lawyers might submit as their Exhibit A, the Roman philosopher Lucretius, who chose, as the saccharin for his treatise De rerum natura, the rhythms of dactylic hexameter. Explaining himself, he wrote that, because his Epicurean doctrine will seem “harsh to those who have not used it,” “I have chosen to set [it forth] to you in sweet-speaking Pierian song.”
This sweet story, told by Peter Kivy, reminds us of how new is the idea that the “form” and “content” of a poem are united, inseparable, one and the same. Lucretius had the content of his doctrine worked out well in advance, without thinking he’d thereby foreclosed any of the myriad forms arrayed before him like so many aisles of candy.
Having proved “form/content unity” foreign to some poetic traditions, Kivy marshals additional indirect evidence against it, by constructing a speculative genealogy. He attributes to the idea’s emergence, motives not well-tuned to the truth:
by the time the eighteenth century rolled around, poetry as the source and conveyor of scientific, philosophical, or any other categorizable kind of human knowledge was a dead issue. … It would have seemed almost as absurd in the Enlightenment to assert that the poem was a vehicle for the expression of scientific or philosophical knowledge at the cutting edge as it seemed sensible and commonplace in the ages of Parmenides, Plato, and Lucretius. Poetry, it would seem, had lost pretensions to knowledge.
If the vibe was now against using verse to make philosophy tolerable, poetry needed a new function, or risk obsolescence. But—all the domains of knowledge had been divvied up among the various prose-writing disciplines, physics, psychology, etc, and nothing seemed left for poetry. Surely something had been overlooked? “We must secure for the poet a kind of knowledge that only he can command.” Ah—what about the special knowledge that can be expressed only in verse?
if...only the poem can say what it says, then what the poem says...is an expression of content that only the poet can have “discovered.” The poet is the world’s greatest expert, the world’s only expert, on the kind of knowledge his poem expresses, because it is the only example of that kind: it is sui generis content.
If this was right, poetry’s problems would have been wiped away. Too bad, then, that it’s all a bunch of hooey. The “special content” thesis entails that the content of a poem—what it says—cannot be paraphrased, for
that paraphrase would inevitably fall into one of the categories of human knowledge populated by resident authorities who perforce would outrank the poet in expertise.
But a poem’s content can be paraphrased, Kivy thinks—or anyway, it is no less paraphrasable than the content of anything else.
In one sense, an attempt to paraphrase even the top article in today’s Times will fail to reproduce the “total effect” of reading the original—obviously, since that total effect includes the experience of reading the exact words making it up. But neither is poetry special on less demanding criteria for paraphrase. If what a news article says could have been said in different ways using different words, a poem is no different. Milton himself prefaced each book of Paradise Lost with—a prose summary of its contents.
But if poetry is not good for making philosophical knowledge pretty for the cameras; and if it’s not good for conveying some other kind of knowledge to which science, philosophy, and all other human inquiry is blind; if poetry is not a “special conduit to the font of wisdom”; if
the practice of poetry is not a way of knowing some particular kind of thing but, in one of its offices, one of the various ways we may have of expressing all kinds of things we know or believe, wish or hope, fear or value,
then the question remains: why bother expressing any of those things in poetry, rather than in some less demanding medium?
Kivy answers by quoting Arthur Danto: art, and therefore poetry,
uses the means of representation in a way that is not exhaustively specified when one has exhaustively specified what is being represented.
Yes, a poem may convey a content that could be conveyed in another way, but part of the significance of the poem lies exactly in the way; one has not fully explained or understood the poem, if one has only grasped and paraphrased that content:
one must also specify the way in which the form, the medium, is employed...the way in which the artist employs the medium is, in effect, part of the content, because it expresses something in the artist’s point of view about the content.
But this point, finally achieved at the end of so many pages, is just the beginning. The problem simply reappears. If what the words of the poem say may be paraphrased, then so also may be what the poet or speaker expresses about it. If the “content” of
Gather ye rosebuds while ye may, Old Time is still a-flying
may be re-phrased as “don’t waste your too-short life,” then, similarly, what Robert Herricks expressed about that content (a big thumbs-up) may also be said out loud, rather than left implicit. The Kivy/Danto distinction between what is said and what is expressed does not, by itself, give poetry a job only it can do.
But we can help it along, by bringing back to the stage the total effect of the poem. An attitude as expressed in a poem may engage responses in us far different, and far richer, than those we may have to the bare information that it has been expressed. Say I express my sadness in wailing and tears; whether and how you are moved would be different if, instead, I had stated flatly how I felt from behind a blank-faced screen. If Kivy is right that what is expressed in a poem matters, he is wrong to suggest that this is just another “part of its content,” for an encounter with some “content”—some body of information—is far different from an encounter with a human expression of an attitude or emotion.
Still, still—in prose too the way may matter; in prose as well, what is expressed may be lost in a paraphrase of what is said. The question of why poetry has not been answered.
See also: William Wordsworth Book Report.
I love this exploration! Poetry, like melody, feels like it can be held in a different way than other media. When Auden writes "If equal affection cannot be,/Let the more loving one be me" it has softness and substance, the hook of a great melody. It can be carried indefinitely, and it invites one to grow into the open space around it. Why poetry? Because heart.
I loved this! I'm quite interested in these questions.
"[. . .] then, similarly, what Robert Herricks expressed about that content (a big thumbs-up) may also be said out loud."
When I read this, I immediately thought of Dewey's distinction between "the languages of science" and "the languages of art." From what I gather, the former "languages" can tell you what you would experience if you were in the position to experience it, whereas the latter "languages" can actually put you in the position to experience, if not give you the experience. I'm not a Dewey scholar, but the thought of mine he's inspired is that to read a linguistic paraphrase of a "way," no more and no less than reading a description of that "way," simply is not to experience that "way." This fact remains, despite the further fact that it is POSSIBLE — of course! — to linguistically paraphrase and describe the "way."
And then I read this: "Say I express my sadness in wailing and tears; whether and how you are moved would be different if, instead, I had stated flatly how I felt from behind a blank-faced screen."
This addressed my concern, and went some way toward answering it. You claim — rightly, I think — that "an encounter with some 'content'—some body of information—is far different from an encounter with a human expression of an attitude or emotion." But you also seem to suggest that the difference is located in "whether and how you are moved."
I'd like to propose the idea that the difference does not lie, or does not only lie, in the difference in *effects* brought about by the encounter with "content" and encounter with expression. I think there's something about the form of the encounter, the nature of the experience, that makes the difference. The argument-sketch I have for this idea would take us into the dark territory of disputes about different forms of knowledge — too harrowing a venture for a Substack comment.
Anyway, if you have a shareable response, I'd love to hear it. Thanks for the great post.