On the Ad Hoc Critic

By: Rishi Janakiraman

There was a wet slice of evening at the West End Poetry Festival in Carrboro, NC when I was on the balcony, catching a city-wide view: the clouds evaporated to leave a dark-jean sky, the town smelled of frying oil and cut grass, and an older gentleman who’d been leaning on a cane recited Donne to me. In particular, these four lines from the infamously-erotic Elegy 19 (“To His Mistress Going to Bed”):

License my roving hands, and let them go,

Before, behind, between, above, below.

O my America! my new-found-land,

My kingdom, safeliest when with one man mann’d

I had read the same poem a few weeks earlier in my English class, where our discussion was organized around R.V. Young’s influential essay about the poem. Young argues that Donne is subverting the familiar image of the New World as a naked woman; instead of the trope of “America as a mistress,” the mistress becomes America, purportedly revealing Donne’s hidden anti-imperialism. In class, Young’s argument was well-liked. Donne, the story goes, is not just a raunchy lyricist but a poet who, hiding his politics under the threat of Elizabethan persecution, managed to invert the prevailing metaphors of exploration in the form of a lewd parody of colonial avarice. 

The man on the balcony did not know Young, and did not have to know him; he only called the lines “genius” but “with unfortunate colonial sympathies.” If he had read Young, I wonder if he’d accept the image of an anti-imperial, subversive Donne, if the pull of the critic would push his instinct over the edge. The contrast—my seminar-room reading of criticism, replete with archival caveats and theoretical modesty, and the festival man’s instinctive verdict—made salient a problem I’d been noticing: how do critics arrive at their conclusions, and how do those conclusions reconfigure the text they claim to be interpreting?

When I read Young, I was among the least convinced in that class, or at least I was the most skeptical of his methods. It fell into a kind of retrodictive criticism, as I saw it, where the critic begins with a high prior probability in his own view—Donne is subversive, or this poet traffics in reactionary tropes, or that collection secretly endorses a political project—and treats the text as a field in which to collect confirmatory instances while explaining away inconvenient data as strategic disguise, censorship, metaphorical displacement, or ironic inversion. When you start with a conclusion and then read the poem to show how its oddities are really signs of your conclusion (“obscure” = “secretive protest,” “erotic” = “political mockery,” “ambivalence” = “strategic concealment”), you invert the relationship between evidence and hypothesis. In other words, rather than letting the text have a say in adjudicating between rival explanations, the critic immunizes the thesis against counter-evidence by shoehorning anomalies into ad hoc explanations.

Going back to Young, his premise—that Donne’s image of the mistress as “America” is an inversion that exposes colonial avarice—reads elegantly and seems to make sense of an otherwise ribald erotic conceit; Young even frames the concealment as prudential and tactical, suggesting that Donne utilizes the “minor [offense]” of bawdy imagery to mask the “major offense” of radical political critique. That claim is rhetorically potent because it reconciles two seemingly incommensurable facts about Donne: his sexual exuberance and his supposed political project (i.e., why doesn’t he just write a political poem? does he have to talk about undressing his mistress?). Young’s hypothesis, though—that Donne is anti-imperial in these poems—is treated like a regulative principle, and what ought to count as disconfirming evidence (for instance, the poem’s pleasure in domination, its alignment with explorers’ rhetoric, or the simple literary fact that erotic conceits often function for their own aesthetic reasons) is converted, by rhetorical fiat, into evidence of concealment. To the extent a datum resists the thesis, it is reinterpreted as proof of steganographic subversion; to the extent a datum supports the thesis, no extra argument is required. 

The pattern mirrors the well-known philosophical problem of ad hoc rescuing. In science, a hypothesis that survives only by proliferating auxiliary assumptions—this experiment failed because of contamination, that observation is merely an artifact of measurement—is weak, since it loses the ability to be falsified and therefore to be informative. The literary analogue of this is criticism that centers authorial intent in a way that produces an explanatory or biographical cause for each recalcitrant line. In other words, it introduces more auxiliary assumptions than it actually supports the hypothesis. 

I want to be clear, though, that I don’t see this as a problem of authorial intent inherently. Even if we can’t directly retrieve authorial intent, good analysis can still emerge from parsing it. I’m not arguing against attending to political context or an author’s intent; on the contrary, I want to salvage those concerns from the ad hoc function they often serve. Authorial intention and historical circumstance are legitimate data points in an argument about meaning, but they too often serve as sidelined justifying strands that support an already-held conclusion. A responsible interpretive practice treats claims about intent comparatively: does the appeal to Donne’s recusant biography increase the explanatory power of the hypothesis more than it multiplies ad hoc assumptions? Does historiography about Elizabethan prosecution produce specific, testable alignments with the poem’s lexis, syntax, and trope-structure, or does it just provide a suggestive backdrop in which one can comfortably project? The virtue to cultivate is proportionality: the stronger the claim—”this poem is an anti-imperialist tract disguised as pornography”—the stronger and more specific the evidence required, and the less we may rely on opaque suppositions like “the poet had to hide for fear of persecution.”

The necessity of a proportional evidence base is especially true in accusatory situations like the controversy over Toby Martinez de las Rivas’ poetry. For context, in 2018, Dave Coates published a long essay arguing that Martinez’s poetry displayed a persistent fascist aesthetic, naming particular imagistic patterns that Coates read as ideological markers. In particular, recurrent invocations of “coherence” turned into a demand for racial or national “purity,” the image of a “black sun” was read as a racialized symbol, and references to falconry and cyclical ruin were correlated with Yeatsian doom-mongering (the falcon, in Coates’ account, does not just hunt; it heralds the fall toward authoritarian mythopoesis). Coates’ piece performed the noble work of vigilance: languages and images have been co-opted by violent ideologies before, and critics should interrogate whether a seemingly aesthetic register has been estranged into a political program. The problem, however, is that Coates frequently begins from a political diagnosis and converts ambivalence into incriminating consistency. 

Henry King’s rebuttal to Coates’ essay in Poetry Nation Review dissects those specific points, showing that his interpretive leaps often rest on equivocal slippages, and that many of the motifs Coates construes as ideological actually have plausible alternative pedigrees within devotional, metaphysical, or folk traditions. Where Coates hears an invocation of national rebirth, King points to the devotional lexicon of sanctification; where Coates hears a coded call to ethnic homogeneity in the “black sun,” King points to a broader tradition of apocalyptic imagery and color symbolism that is not, on its face, racially programmatic; where Coates reads falconry as Yeatsian heralding, King excavates a surrounding set of references—often from medieval narratives—that materially alter the interpretive calculus. King’s corrective, here, is an insistence on evidentiary sequencing: if you posit ideological intent, you must show that this hypothesis explains more of the data with fewer gratuitous assumptions than competing hypotheses. Both Young and Coates begin with an interpretive priority (Donne as clandestine anti-imperialist; Martinez as covertly fascist) and retrofit the poems to prove it. 

To be emphatic, it is both noble and necessary to interrogate authors for fascist or illiberal messages like Coates did with his 2018 essay. We all have a public responsibility to trace how aesthetics can become complicit with violence and to call out genuine doctrinal affinities where they exist. The ethical urgency of that project cannot be overstated. But noble ends don’t excuse a bad interpretive strategy that dives in conclusion-first such that it risks false accusation. As Rob Mackenzie writes in Dark Horse Magazine, Coates’ essay was “one which could threaten to destroy the poet’s reputation and livelihood.”

If the stakes of interpretation are sometimes merely academic, they are at other times life-changing, which means that when we point our critical finger we must first disclose the evidentiary chain that lets us point; when we name a poem “complicit” we must first show which lines make that charge more probable than the alternatives. In short: continue to be vigilant about aesthetics that might naturalize violence, continue to parse the intentions of what poets write, but do so under a discipline of evidence that is premise-first, conclusions-last. If we can hold that double posture, we can convert the fevered thrill of declaring “guilty” into the steadier, harder work of saying, with reasons and reservations, “here is what the poem most likely does, and here is what would change that judgement.”