The poet and novelist James Dickey saw poetry as a two-step surprise. You go in expecting your excited, deep, associational mind to yield earthly paradise. Not so, Dickey warns. What you find is hell. You realize it’s “better to be stupid and ordinary,” he writes in Sorties (1971). The novelist Saul Bellow toes the same line. America, he says, loves its “dead poets” for their madness, drunkenness, and despair, for the fact that they just can’t make it here, for the way they light up the “enormity of the awful tangle” that is life. In the pungent and frequently wise We Begin in Gladness: How Poets Progress (2018), poet and critic Craig Morgan Teicher takes a less-expected path. His focus is the ascent, the delicate inward speaking, the ways in which poets climb toward profundity and specificity. Refreshingly, his argument isn’t that poetry cures. What’s refined and perfected, he suggests, is language itself.

Teicher is the author of three books of poems, including The Trembling Answers (2017) and To Keep Love Blurry (2012). He’s written for the New York Times, the Los Angeles Times, and NPR. This interview was condensed and edited.

My dad used to say poetry was "an indescribable perfection of utterance." I think he got that line from Housman, but I'm not sure. Is what makes a bunch of words poetry, or what makes poetry good, definable, or is it truly indescribable?

That's a great question. I've staked a good deal of my writing career on the notion that poetry is "describable" and that I can be trusted to describe it in a meaningful and useful way, an activity I think is deeply important to the life of literature: the attempt to articulate what it is, materially, and what it seems to be about. Poetry is speech, cast in written language, that demands a response; it’s part of a conversation. Readers either write other poems or go “mmmhmm” or write about the poem, if that’s something they like to do.

I would modify your father’s phrase: poetry is an un-paraphrasable precision of utterance. I’m not much interested in perfection, though of course a well-made poem can’t be said any other way, but a poem can be described. My favorite metaphor is a Richard Serra sculpture, one of those huge ones that take up a vast room. Viewers “interpret” those works by walking around them and describing what it feels like to be in their presence, to feel their shadows looming, to walk inside them and feel their immensity. That’s how I try to describe poetry: I ask myself what it feels like to be near it, to be in the presence of a particular poem. It’s not about whether I get the poem right or wrong.

Is the poet talking mostly to the reader or some other liminal presence?

A poem always has an addressee who can be reached only by means of poetry—someone dead, distant, and, finally, imaginary. If the addressee is alive, the poet uses a poem to address the person, who might not be able to understand or take in or accept what the poem is saying if the poet were to simply call on the phone or send a text. The poem says precisely—not necessarily perfectly, because, to me at least, perfection implies a kind of neatness that is often not useful in poems—what the poet needs the addressee to hear because there are many things a person needs others to hear that can’t simply be said. They must be shown through words that engender a response, like those Serra sculptures—overwhelming or sprawling or more than one thing at once.

One more point on this question is that I think one way criticism, or conversation, about poetry goes wrong is when critics don’t try hard enough to be precise about their response to a poem or use the language or the tone of the poem to describe how the poem affects them. I call this “signing on” with the poem. The most exciting thing about criticism or the discussion of poems is the struggle to run the foreign substance of someone else’s language through the currents of one’s own, so that what comes out isn’t really an accurate description of the poem but an interaction with it. Criticism should create a kind of distance from the poem under consideration so that readers of the criticism can run the critic’s language through their own sensibility, creating yet another response. Maybe in doing that, we get further from the original poem, but we are creating all this interaction around it, which is what every poem is designed for.

I was looking through an old anthology of poetry by Cleanth Brooks and Robert Penn Warren. It's a rather odd book. At one point, they cite “Trees,” by Joyce Kilmer ("I think that I shall never see / A poem lovely as a tree"). They say it's greatly admired. They say it's popular. Then they conclude, "But it is a bad poem." Not to be a downer, but what makes poems bad?

That poem is bad because it makes a bunch of assumptions about what its readers think that just aren’t true. Lots of people think trees are lovely, and few think about poems at all, but fewer still think it’s particularly relevant to compare the two. I’m being a little facetious, but I think this gets at the heart of what does make bad poems bad: poets’ misunderstandings of whom they are addressing. The real art, I think, is in calibrating a poem’s language so that it anticipates what poet and reader might share and what they might not.

In reference to the subtitle of your book, are there common ways in which poets do not progress? I don't mean because they must work jobs, etc. I'm thinking about internal aesthetic factors and choices.

There are lots of ways, the most common being self-imitation. Poets find a kind of poem that readers seem to like, that they know how to write, and just keep iterating versions of it for decades. Meanwhile the world and the language change around them, and their poems do not. Or maybe they change, meaning their present selves use different words than their past selves did, but they don’t let their poems change too. Throughout a lifetime, we come to care about things that we didn’t care about before, for instance our children, if we have them, meaning we will write different kinds of poems, perhaps ones our readers, or even we ourselves, will like less. If we don’t let our poems change along with our concerns and our words, it’s all bullshit, bad poetry.

You start your book with Wordsworth’s line that poets begin in gladness but end up in madness. Then you add Delmore Schwartz's take—that poets begin in "sadness." I might agree with Schwartz more. There's a line attributed to Philip Larkin: "happiness writes white." Do you think some degree of "gentle melancholy" tends to make poems more powerful? What's the better stimulant, happiness or sadness?

I believe poets start writing poems because there is something they need to say, or a kind of thing they need to say, that they can’t get across by more direct means. I think people tend to write poems because they feel otherwise silenced, at least sometimes, about certain topics. Maybe they want to talk to everybody, or nobody, or maybe the poet is talking to the self, who is so often resistant to change and suggestion. Along with those circumstances comes a certain amount of unhappiness or frustration, the longing to say something one can’t, to live in a world where one can.

Truman Capote used to say "style is personality," and you can't mess with your style any more than you can mess with the natural color of your eyes. I feel like he's a little right and a little wrong, but in any case, his comment makes me think of "voice," a subject I find fascinating. How, do you think, does voice manifest? How do you find an authentic voice as a poet?

A poet’s voice is really the substance of a poem; it’s true material. It’s the sum of the poem’s raw materials—a poet’s vocabulary, the associations present in particular words at a particular time thanks to a poet’s life in the culture at large, the mannerisms of grammar and phrasing a poet picks up or adopts, and predilections toward subject matter, which inflect and are inflected by these other elements. What makes a poem not just a pile of words is the interaction of all these things at once, which adds up to a sensibility, a poetic voice. Poets' voices have roots in whatever circumstances shaped their earliest speech: family, community, TV, movies, and, more recently, the internet. And books! One imitates the sentences one is attracted to. Then, as one begins to write more and more deliberately, one makes conscious choices and adapts grammar, inflection, and phrasing to subject matter.

Can a poet fake voice?

A voice is really limited or enabled by one’s linguistic bones, as it were. A poet can wear a mask, but I don’t think it is possible to entirely eradicate or cloak one’s speech mannerisms. Nor should one want to. That one may not always like one’s voice or think it’s hip or contemporary is irrelevant. “Finding your voice” is about listening, identifying those “bones,” coming to terms with them, accepting them, and learning to control them to some extent in the service of poetic effects.

One may be a talky, plainspoken poet, thanks to where and how one got one’s sense of language and wish that one were more naturally adept at more complex kinds of sentences. Alas, I say. Those weaknesses or perceived families are one’s truth, aspects of the voice one has been searching for. I don’t think writers preoccupied with finding different voices than the ones they were born with get very far or are able to go very deep.

Plath is a poet you single out as an example of a person who wrote in creative "surges," as with the Ariel (1965) poems. You talk about how she went from control—being withheld, formalist, bloodless—to dyscontrol, or complete abandon. I see many of the Ariel poems as practically psychotic—or, as you say, "unmoored." What are your thoughts on this dynamic of control/dyscontrol in art making? Is being lopsided in either direction ever a good thing?

You are certainly more qualified than I am to make observations about psychosis evidenced by Plath’s, or anyone else’s, poems, so I merely want to say that, for me, the “unmooring” I treasure in Plath represents a freedom from imaginative decorum, meaning from writing the way a poet of her time ought to write. I’d like to think a sane person could get as far out as she got—in poetry. I like to think that’s what poetry is for: it’s a safe corral for one’s insanity, somewhere to put it so it does less harm to oneself and to others. In life, Plath did not manage that. But I think a sane person can benefit from and deeply enjoy her art without going insane.

But that’s not what you’re asking. I think poets seek, in poetry, a forum in which they can be out of control, via the associations and meanings they can find when words bump up against one another. But a poet can achieve that and be perfectly sociable in real life, I hope.

The poetry I love most seems to have dropped the reins of the horse it’s riding, but there’s a fence in the near distance and, maybe, the risk that the mad horse will jump over it.

Your book, refreshingly, is about a poet's ascent, the ways in which poets climb to the greatest profundity. What, do you think, is the most underrated element contributing to such ascents? What's the most overrated element when it comes to reasons for descent?

This is such a good question and fun to think about. Somewhere in the book, I share a quote from Paul Muldoon that my wife, the poet Brenda Shaughnessy, shared with me: he said something like, “Poets dis-improve with age.” I’d like to say I think he’s wrong, but his theory does bear out with many poets. I’m not sure poetry is something it’s wise to “master,” meaning sometimes poets achieve a kind of virtuosity in their own voices and cease to surprise themselves, and once we’re not surprised by our own language, we’re lost.

But that’s not what you asked. Or maybe it’s the answer to the second part of your question: the answer to part two is virtuosity.

For part one, I’d say the most underrated element of a poet’s ascent is imitation. Trying to be original is ruinous. A poet doesn’t have to sound utterly different from every other poet to say things that are true and resonant and useful. And why not reach for the voice of a poet whose insights and associations tower over you? I should want to be Emily Dickinson (whenever I read her, I wish that wish); I’ll fail, finally, to be her, and even if I try to write poems that sound like hers, I’ll infect them with my voice, but that is, I think, something new and useful: Emily Dickinson via Craig, Craig via Emily Dickinson. One could do worse.

It seems as though a great many people believe that making art—poetry, painting, novels, whatever—is somehow, mysteriously, automatically therapeutic for the artist. That idea has always struck me as complete bullshit, for lots of reasons. I think we need to get away from this art-as-therapy notion. It’s as mystifying and simplistic as the notion that artists are tortured. What do you think?

I suspect few artists share the belief that art is therapeutic. But maybe people who do are mistaking art for insight, which is sometimes one of the side effects of art. What is therapeutic is maintaining an ongoing and reflective conversation with oneself in which one object is to apply or make use of any insights one has gleaned by various means. But art making in itself does not equal an ongoing conversation. Writing a poem is much more akin to performing a theater piece or playing a show in a band than it is to attending a therapy session. It transforms the artist momentarily, the way putting on a costume sort of does. After the fact, one is left with one’s old habits and ways of being. That said, one’s art may be a place to go mining for insight. But insight in itself doesn’t help. When art making is beneficial for me, it’s because I have simply kept the promise I made to myself that I would make art, and I think it’s good to keep one’s promises to oneself.

We’ve covered a lot of ground. Is there a question you want to ask yourself? A question you wish someone would ask you about this book or about you?

I do not have a question for myself: there's nothing I want to know about myself that I'd be willing to tell me.