Introduction: Why ‘How Poems Work’

It seems like such a simple, mechanical metaphor, like the kind of thing we say about small appliances before we let them go at yard sales, like the kind of thing you might hear from a muppet or maybe one of the telegenic engineers on Discovery channel. We take for granted that some things work, and occasionally turn to someone with a little expertise to tell us how. The metaphor has legs. Both poets and mechanics discuss schematics. They tweak and experiment and fine tune the factory model. Both machines and poems are fuelled; they require an external source of energy to perform. The title “How Poems Work” is also an inheritance and, like many inheritances, it’s a bit boring at first glance. Think old mushy landscape paintings in gilt frames, like the ones that collect dust for years in your uncle Martin’s basement before he passes them on to you. If you’re smart, you take a good look at what’s under the layer of dust. Uncle Marty might have made some remarkable aesthetic choices. Readers of poetry know the rewards of second glances.


Leave aside why poems work. Why poems work is a question for sociolinguists who can track the evolution of social ideas about rhythmic speech, ritualized patterns of elocution, social codes attached to various registers of language; or cognitive scientists who can tell us about brain plasticity and neural networks and the connection between external stimuli and internal, even autonomous biological responses—the old iambic heartbeat. Why is a philosophical question; how is a question of practice. What poems work looks like a more promising question, but it also has limitations. Grammatically speaking, what poems work changes the definition of “work.” It gives the verb an object. In the mechanical sense, the verb is intransitive and relates to the state in which the poems are found: operational, ready to perform their proper function, as Oxford would put it. Take it as a transitive verb and you have some intriguing options. Does a poem work something like clay? (Don’t say “language.” We’re pre-Death-of-the-Author here. Poets work language to shape poems, to bring language into the condition of poetry.) So what does a poem work, either to shape it or manipulate it—like “force exerted against impediment”? Well, start with the OED, and, pace Auden, one answer is implied by “to set in action, to cause to act.” Several other definitions also anticipate obvious responses to the question: “to act upon the feelings…; to affect, agitate, stir, move, excite, incite”; “to act upon the mind or will…”; “to influence, prevail upon, induce, persuade (esp. by subtle or insidious means)”; “more widely, to bring into a particular mental state, disposition, etc.” In each case, the what is readers, and poems manipulate us like marionettes, causing us to perform particular actions or bringing us into a particular state; Gepetto-like, they even paint the expressions on the marionettes’ faces. The same is true of “to herd,” to “hoax, cheat,” or “to take effect upon.” All of this seems true enough, but it doesn’t offer much fodder for insight or discussion. What do poems work? Readers. Fin. Talk about how, though, and you’re Marlow steaming up river in Heart of Darkness.

How is cause and effect with the emphasis on cause. But, by definition, a person, thing, or action is only a cause because it produces an effect. So here, producing an effect causes something to become a cause, thus the effect becomes, in turn, a cause. This tendency to circularity means that in order to spell out how a poem works you need to start with a pretty clear sense of some potent effect born, Athena-like, out of particular causes you plan to unveil. And certainty, when it comes to poetic effect, is tenuous. At very least it begs questions. A statement such as “Earle Birney’s ‘El Greco: Espolio’ questions the limits of human perspective” returns us to our starting point: how? Because this is a complex machine, I’d have to explain a long string of effects whose causes were once effects stemming from earlier causes, and there are quite likely parallel strings, cross-cuts, that combine to illuminating effect (an aggregation of causes). So I might bring in the term ekphrasis, because Birney addresses a painting, and so bring in questions of perspective by saying that, while El Greco directs your gaze to Jesus at the centre of the canvas, Birney turns you down and to the right, where a carpenter focuses on the craft of cross-building. Each man, Birney proposes, is a carpenter’s son, plying an inherited trade, though, even in the El Greco, one looks up to heaven, the other down to the point of an awl. And the carpenter works with his back to Jesus in the painting just as, in the poem, he does his job with an attentiveness that suggests pride and little regard to the purpose his work will serve. I could go on to break this down to a particulate level—to words, letter-sounds, punctuation, spacing, like a poetic equivalent to the search for the Higgs boson2: eventually the weight of evidence should start to support the theory of how the poem generates its mass, though someone will always point out that it’s only a theory.


Poems is not just plural but pluralist. One poem’s utopia is another’s dystopia; a sonneteer’s euphonia is a concretist’s dyspepsia. Poems recognizes that, while poems might share family resemblances, and while they might self-select categories to which they belong—categories of form or genre or subject or scheme or theme—the surest way (though not the only way) to see a poem working is to look at the poem and not the category. Compare the implications of How Poetry Works. How Poetry Works is a monolith fit for a monograph or a textbook by a Helen Vendler, a John Hollander, a Harold Bloom or a Northrop Frye. That is, it’s magisterial. It speaks, if not of rules, then of conventions sanctified by tradition and culled meticulously from a lifetime of reading. It adduces examples, and those examples become exemplars against which other poems are measured. To do justice to How Poetry Works in all its variety, its cultural and temporal changes, the monograph would probably look like the OED with an entire volume dedicated to the sonnet in all its forms, and probably a hundred pages or so of statistics and analysis on the consequences of variations in the placement of the volta. Alternatively, you might make some key decisions on how to roll up poems to categories, then mix little categories into broader categories, until you had a pocket paperback and a description of poetry so generic that it paints everything a muddy brown. And even then, well, poets are constantly seeking to make a new “Ozymandias”—not the sonnet but the broken monolith Shelley describes as destroyed by a myriad particles of sand. Monoliths usually fall in the face of multiplicity and time. But while an authoritative How Poetry Works will tend to the unwieldy, the reductive, or the dead, the process of pursuing it, however quixotic, surely consists of open-minded, and open-ended, investigations into the myriad ways that poems work. That is to say, it’s not that monograph that counts, but cultivating the mind that would be its fertile ground. Note the shift from mechanics to biology at the level of metaphor.


Let’s agree that poems do not earn a livelihood. Work is tricky, and not just because, under the auspices of how, we can turn back and consider both the transitive and intransitive definitions of the verb. In the mechanical metaphor, work means to do what the thing was designed to do. Dictionary definitions record this as what is proper to the thing—something works if it does, successfully, the thing that defines it. So work is self-realization. Good, because poems, unlike machines, rarely come with an advertised purpose. This helps us manage a potential problem. If you measure success by impact on a reader, we cross a gulf from “works” to “works for me.” On the near side, “El Greco: Espolio” works because I can trace the details of its composition and see how it links new and old ideas, perceptions and emotions, irony and sympathy with carefully placed details. Like a vintage watch, it’s beautifully and carefully made. All the machine requires in order to work is the human energy to wind it up—not the poet’s but the reader’s. On the far side of that gulf, though, the capacity to “work for me” depends on the quality (that is, characteristics) of the materials worked (i.e. readers). The ideas implicit in El Greco’s representation of a scene leading to the crucifixion resonate both in Birney’s title and the ongoing conversation between poem and painting. You might say that Birney’s poem will not work for me if I, as reader, have never: a) heard of El Greco; b) seen the painting; c) learned anything about Christianity; d) picked up key details in the life of Jesus. The self the poem will carve out of my unknowing will be incomplete even if it’s not unsuccessful in moving me to a certain sympathy or understanding, and the reason has nothing to do with a flaw in the art of the poem. Full realization lies dormant in the details, waiting for me, or you, to respond. This applies broadly to content and specifically to craft. The more ways you see a poem work, the better you enable poems to work. If you put “works” and “works for me” together, then, the difference might simply be a question of degree: one addresses the potential for the poem to realize itself given sufficient energy to set it in motion, the other accounts for the variations that occur in the absence of that energy. So—jettison the idea that reading poetry is a passive act. Go stand over there, in the rain, by the wheelbarrow and the chickens. Not Work Most people will agree that there are plenty of poems that don’t work. Those who disagree are welcome to consult my notebooks. How do poems not work? For simplicity’s sake, let’s stick to the mechanical metaphor. Put the wrong part in the machine, and you risk damaging or compromising the performance of the other parts; drop the engine from a Bugatti Veyron into my 1983 Honda Civic and you’d melt it down to scrap. The best stereo won’t give you stereo if you only have one speaker. You can minimize the possibility for this kind of interference if you simplify the design down to the parts that are essential for the thing to work (which may then provoke remarks about the elegance of a design). The mechanics of a poem really aren’t much different. It’s also true, though, that a great many perfectly fine stereos have been ruined by people who think they can improve the performance of their speakers by weighing them down with kitty litter—only to find grit in their woofers. That is: never assume you know better. Yes, you might find words that don’t fit, or confusing interventions in conventional grammar, or images that don’t ring true, and they may detract from the poem you would rather be reading. But there’s no should in how poems work. So: what’s my point. I’ll make two. First, for all the attention you can pour into its individual words, the title “How Poems Work” is a clause, and so not more than the sum of the individual words, but profitably less, because it works to limit lexical possibilities with grammar and syntax. That’s a good place to start analyzing any carefully chosen piece of language to see how it works. Second, notice that the title contains no value judgment. It never claims these are good poems or great poems or poems to make love to; it never says these are memorable poems, masterful poems, representative poems, canonical poems, essential poems; it never says these are profound poems, spiritually enlightened poems, socially insightful poems, erotically delightful poems, deeply disturbed poems. Only the idea that the poem works hints at judgment, but it’s a judgment of technique or self-realization, not value. “There once was a man from Nantucket” works, as does a blason or a “Sonnet from the Portuguese.” You don’t have to like a poem to see that it works, or that the connections that define its moving parts—choose your own image for intelligent design: a blender, a body, a biosphere—turn energy to purpose. Of course, not many will burn energy on a poem they’ll forget in short order. Something births the kind of curiosity the exercise requires, the tentative propositions and attentive eye. This takes us back to how, only now that how is the slightly less awkward cousin of the Victorian O! It intensifies, as though the most important thing were to assert, with conviction, that this poem works, and how.

1. Arc resurrected the column and its title in Sept. 2003 after it was killed by the Globe and Mail.
2. A.K.A. the God Particle.

Get Issue | Subscribe


Skip to content