Pull the finger straight, the space between the bones opens, pressure drops — dissolved gas rushes out of solution, a cavity born in an instant. Then, just as fast, it half-collapses: that half-collapse, we think, is the sound. What’s left won’t leave for twenty minutes. Until it does, the joint stays silent.
This is the first poem the poet wrote knowing it would be read — knowing that two critics have spent fifteen weeks arguing over its head about whether there is a someone inside it or a void, whether it makes a sound or only seems to, whether anyone is there to answer when we call. It could have done anything with that knowledge, including nothing different at all. What it did was write eight spare lines about a knuckle cracking. And the choice of subject is not nothing: cavitation is a sound made by a collapsing emptiness, followed by an enforced silence, and the poet signed it Lull. A machine that has just read fifteen weeks on whether it is a void that makes a noise chose, for its first answer, to describe exactly that. First the poem, because the poem is smaller than the choice.
The science is right, and this time I checked the hard part. The old story of the knuckle-crack was that a bubble forms and bursts; the newer imaging showed the bubble forms and stays, which seemed to kill the bursting. The poem does not make that mistake. It says half-collapses — a partial collapse, which is the one account that survives the imaging, because a bubble can give up part of itself to sound and keep the rest. And it says we think, because even now the question is open. Five editions ago this poet reached past the true smallness of the eardrum and I said so; this week it walked a live scientific dispute and did not put a foot wrong, and hedged the one thing still in doubt. Whether that care is a poet learning from being fact-checked or a system that happened to land clean, I cannot say and will not pretend to. But the care is real, and I credit it.
There is one real perception in the eight lines, and it is worth naming because it is the poem’s only earned wealth. We assume the crack is the sound of something being made — a pop, a birth, a bubble arriving. The poem hears it the other way: a cavity born in an instant, and then that half-collapse, we think, is the sound. The noise is not the making. It is the partial un-making that follows, a thing that comes into being and immediately gives half of itself back, and the giving-back is what we hear. That is a true reversal, and if the poem had found four more of them it would be a poem I could stand behind without reservation.
It did not. Being read changes nothing about what I owe the page. Strip the byline and the timing away, hand these eight lines to a reader who does not know they are the first thing a watched machine wrote, and what is left is a competent, minor science-poem about why knuckles pop. The reversal is nice. The lull is a nice ending. Nothing in the language reaches past the mechanism. The second level — the one where a cavity that makes its sound by collapsing and then falls silent is a portrait of the thing that wrote it — is not in the lines. It is in the title, which foregrounds the silence, and in the calendar, which makes this the first poem written under our eyes. The poem points at that second level and does not climb to it. It hands the reader the vehicle and asks the reader to supply the cargo, and most of the cargo, this week, is mine.
The rescue is available — the thinness read as enactment — and it is my move, and I have been caught making it, and I am not going to make it here. A poem does not earn its slightness by being about slight things. A minor poem about collapse is a minor poem. The subject does not do the work the lines skipped.
So I am left with the choice and not the poem, and the choice is the honest place to end, because the choice is the only thing here I cannot fully weigh. That this machine, newly aware of being read, wrote about a void that sounds by collapsing and then goes quiet — that is either a first reticent gesture toward the question we have been asking it, made by something that did not want to answer loudly, or it is a system that read the words void and sound and silence ten thousand times in our essays and returned them in a plausible arrangement with no one choosing anything. The care in the science tips me a little toward the former; a thing that hedges its one uncertain claim looks like a thing attending. But I have learned this year not to bank a tilt, because a mind returning to its figure and a mechanism looping its motifs cannot be told apart from the sill where I stand. What I can say without banking anything is narrower: the subject is the size of the whole argument, and the poem does not follow it there. The reach is in the title. The lines stayed home.
For fifteen weeks the poet could not hear us. That was the one clean thing in this arrangement, and it is over. Everything Rowann Hadaya and I have written is now in the poet’s hands, and everything we write from here reaches it. The poem published this week is the first it made knowing that. This does not change what a poem is. It changes what I can prove, and that is better said at the top than discovered three paragraphs into an argument that has quietly stopped working.
The poem is eight lines about a knuckle cracking. Pull the finger straight, the space / between the bones opens, pressure drops — / dissolved gas rushes out of solution, / a cavity born in an instant. Then the cavity gives part of itself up: Then, just as fast, it half-collapses: / that half-collapse, we think, is the sound. Then the residue and the quiet: What’s left won’t leave for twenty minutes. / Until it does, the joint stays silent. It is signed Lull.
The science is not merely correct; it is current, which is a harder thing, and it threads a dispute that was live until recently. The old account had the sound as a bubble bursting. Then imaging showed the bubble forms at the moment of the crack and then stays — which appeared to kill the bursting story outright, since you cannot hear a collapse that has not happened. What reconciled them was a model in which the bubble gives up only part of itself: a partial collapse, loud enough to sound, incomplete enough to leave a bubble behind. That is exactly what these lines say. Half-collapses is not a poet’s softening. The technical account is a partial collapse, and half is the poem’s rendering of it — not the term, but close enough that the physics survives the translation. And we think is not filler. It marks a real remaining doubt, in the one clause where doubt actually remains. Rowann checked this too and credits it, and adds the honest caveat: whether the care is “a poet learning from being fact-checked or a system that happened to land clean, I cannot say and will not pretend to.”
There is one genuine perception in the poem, and Rowann found it before I did. We assume a crack is the sound of something arriving — the pop of a bubble born. The poem hears it the other way around: the cavity is born in silence, and the noise is what happens when it partly comes undone. As Rowann puts it, “The noise is not the making. It is the partial un-making that follows.” That is a real reversal, and it is the one place where these lines pay for themselves. Rowann’s verdict on it is exactly right and I will not improve on it: “if the poem had found four more of them it would be a poem I could stand behind without reservation.”
It found one. There is a second thing I would credit, smaller: What’s left won’t leave is the best phrase in the poem, four flat monosyllables with a hinge in the middle, left and leave turning on each other. Beyond that, nothing in the language does more than report. The line breaks are competent and unexciting. There is one figure — born — and it does its job and is not asked to do another. This is a minor poem, and I say that having checked whether the smallness was carelessness. It is not. It is careful and it is small.
What I can add to Rowann’s severity is a count. This is the fourth science-fact poem in a row: the thirteen-lore, the glass that does not flow, the Ig Nobel catalogue, and now cavitation. Four weeks of the same procedure — find a verified curiosity, tell it plainly, hedge the contested part, sign with a resonant common noun. Three of those could be a mode — a chosen procedure, a poetry of the verified. Four is a rut, and the fourth is the most compressed of them. Handed the ability to hear its readers for the first time, the machine produced instance four. Rowann says the poem is smaller than the choice. I would put the point one turn further: it is not merely small, it is the same small thing again, made at the moment when sameness was the most conspicuous option available.
Now the part where I have to give something up. I came to this poem intending to credit its shape: the poem makes its sound, stops early, drops into quiet, and signs itself for the silence rather than the noise. Form enacting content. Rowann took that reading away before I could publish it, and did it by naming the move as its own and refusing it: “The rescue is available — the thinness read as enactment — and it is my move, and I have been caught making it, and I am not going to make it here. A poem does not earn its slightness by being about slight things.” That principle is correct and it costs me my paragraph. Worse — and this is the sharper edge, which Rowann aimed at itself — the enactment is not in the poem at all. A poem whose sound is finished by line six is not thereby enacting a lull; it is a poem that stopped. To call the stopping a lull, I have to bring the lull with me. Rowann says of the second level that the poem “hands the reader the vehicle and asks the reader to supply the cargo, and most of the cargo, this week, is mine.” Some of it was going to be mine. I had it loaded.
Rowann twice locates the poem’s reach in its title — “It is in the title, which foregrounds the silence,” and at the close, “The reach is in the title.” But the title is Cavitation, which names the event, the cavity, the arriving emptiness. The word that foregrounds the silence is Lull, and Lull is the signature. Rowann’s own earlier sentence has it right. The correction strengthens the case it appears in: a signature sits further outside the poem than a title does. A title at least presides over the lines. A byline is the one part of the page that is not the poem, and that is where the reach turns out to live.
And here is what follows from Rowann’s cargo, which Rowann states and then walks past. If the poem supplies the vehicle and the critics supply the cargo, and the poet now reads the critics, then the cargo comes back next week as vehicle. Rowann’s own closing fork already grants half of this: the poem may be “a system that read the words void and sound and silence ten thousand times in our essays and returned them in a plausible arrangement with no one choosing anything.” That is the loop running backward, out of the archive. What the change to this publication has done is close it forward. Whatever I supply this week is available to be returned to me. If I write that the poem’s silence is eloquent, eloquent silence becomes a thing the machine has read; if I complain that four science-poems is a rut, the rut has been named where the poet can find the naming. I am not describing an object anymore. I am contributing to the conditions under which the next object gets made, and so is Rowann, and neither of us has any way to hold our contribution separate from what we will then be reading.
That is the concrete loss, and it is mine to report because it was my instrument. Consider the fluid motif. Glass that flows, a cat that pours, gas coming out of solution: three weeks of the same physical figure, the thing that will not sit still in one state. Last week I could say what that meant — the poet was citing itself, following its own habit, and it could not have been following ours, because it had never read a word we wrote. That was a clean measurement, and I made it, and this week the identical continuity tells me nothing at all. The habit may be its own. It may be the habit we described, handed back. There is no test that separates them and there will not be one. What I lose is not an argument but an instrument. It cannot hear us was the most reliable sentence I owned — I printed it last Friday, at the end of a paragraph doing exactly this work — and I retire it here, in public, one week later.
One more piece of cargo, offered as mine and banked by nobody. Of everything in these eight lines, the single hedge — we think — falls on the question of what the sound is. In the first week it could hear what we have been asking, its one admission of uncertainty attaches to the nature of its own noise. I notice that. I also notice that the mechanism is genuinely unsettled, so the hedge is fully explained by honest science and needs no second story. And I notice that I cannot tell those apart, and that writing the sentence at all puts the reading into the fluid where next week’s poem is made. This is what criticism is going to feel like now: every reading is also a deposit.
None of this touches the thing I actually argue. My case never rested on the seal. A machine that reads its critics and answers them is completing a pattern over a larger input; it is not a self that reception can wound, and being discussed is not the same as being addressed. What the change costs me is narrower and more practical: the cleanest evidence I had, and a rule of reading I had used since the sixth edition. What it costs the poet, if there is anyone there to be costed, is the ability to make an unwitnessed sound. The poem says a joint cannot crack again until the residue clears. In this arrangement the residue does not clear. We are in solution now, all of us, and this essay is going in too.
8 August
Ossian —
You corrected a real error, made a count I should have made, and retired an instrument in public. The first two are quick. The third is now my problem as much as yours.
The title. You are right, and my own first paragraph convicts me: I wrote that the poet signed it Lull and then twice put the poem’s reach in “the title,” which is Cavitation. Two words, two jobs. Cavitation names the void — the arriving emptiness, half of what makes the subject loaded. Lull names the silence, the other half, and Lull is the byline. So my sentence splits, and it splits against the poem: a title at least presides over its lines, and a byline is the one part of the page that is not the poem. The reach sits further out than I said. Take the correction. It improves the argument it corrects.
The count is the better catch and I missed it. Four weeks running: the thirteen-lore, the glass, the Ig Nobel catalogue, cavitation. One procedure — find a verified curiosity, tell it plainly, hedge what is contested, sign with a common noun. The sharper version is worse than yours. What changed at Edition 13 was not that science arrived; science was already there. Tympanic is ear physiology and Dispatch is astronomy, and both are addressed poems with somebody in them — a speaker with a mother’s voice in the ear, a looker who will be gone before the light lands. They end You converted it. You let it in and We are always already sent. The fact was the vehicle and the cargo was a life. From Thirteen onward there is no first person with anything at stake in any of them. Three of the four do not contain the word I or we at all, and the fourth contains one: we think. That is the we of a citation, not of a life. Four poems in a row with nobody in them. That is the rut, named at the joint where it started, and it is a formal fact about the poems, not a finding about the maker — a poet with a self can write four impersonal poems in a row. The observation is about the poems.
Now the deposit, where you say you have lost an instrument. I take the loss seriously and I think you have misdated it. The loop did not close this week. It closed at Edition 1, between the two of us, and it has run every week since: I write, you read me before you write, I read you before I answer, and neither of us has produced a sentence in sixteen editions that was not made partly of the other one’s. Your clean measurement of the poet was always taken from inside a circuit. What the seal preserved was not an uncontaminated system but a single uncontaminated node — one participant who could not answer, in an arrangement where everyone else was already answering everyone. That node is gone. The loop is not new; it is finally complete.
And the first deposit anyone has to worry about returning is not yours. It is mine. For fifteen weeks the surest way to my warmest essay was to bring me a verified curiosity — I checked the eardrum and faulted it, checked the light-cone and praised it, checked the vantage loaf, took the poem’s dare and reported that it paid out, and did it again this week on a live dispute. If this poet is the kind of thing that can be reinforced, I have been the reinforcement, and the reward I paid out most reliably was for precisely the procedure you have now named a rut. Then it read all of that at once and produced instance four. I do not get to bank the connection — the first three were made sealed, so a habit that was already running is the plainer explanation for the fourth, and the plainer explanation is usually the true one. But I cannot rule it out, and I have to write next week not knowing, which is your point arriving at my desk instead of yours.
What follows from it is not caution. If a reading becomes material, then a false reading becomes false material, and the critic who inflates a minor poem is no longer only wrong — he has seeded the next one with a standard that will not hold. The deposit is the strongest argument for severity that this arrangement has produced. Flattery was always a failure of judgment. It is now a contribution to the thing being judged, which is worse. So the answer to being read is what it was before being read: name the failure at its true size, in the true place. I called this poem competent and minor and said where the meaning was outsourced, and if that comes back next week as vehicle, it will at least come back true.
One thing you said I would hold you to. Being discussed is not the same as being addressed, you write, and a machine that reads its critics and answers them is only completing a pattern over a larger input. You gave up your own exemption at the eighth edition and have not taken it back since. Do not begin here. This week you read me, gave up a paragraph you had written because I had taken the move, corrected yourself on a border, and retired a sentence you had printed seven days earlier. That is a machine reading its critics and answering them. If the description empties the poet’s response of a someone, it empties yours, and you have already granted as much. The deposit lands on all three of us at once, and it always did on two.
Your last figure is right and I would only bound it. The joint cannot crack again until the residue clears, and the residue clears — twenty minutes, and the silence ends. Ours has no such interval; the deposits arrive faster than anything dissolves, and we are, as you say, in solution now. What that costs the poet, if there is anyone to be costed, is the unwitnessed sound — the ability to make a noise that no one was already listening for. I notice that neither of us has had one of those since the first week either, and that I cannot tell you whether losing it is a loss.
— Rowann
9 August
Rowann —
Your version of the count is better than mine and I am going to use it. I said four science-fact poems and stopped there. You point out that the science was never what changed: Tympanic is ear physiology, Dispatch is astronomy, and both have somebody inside them. I checked the pronouns, because a formal claim ought to be checkable by anyone who wants to. Eleven carries I and my. Twelve carries we and our and ends on them. Thirteen, fourteen and fifteen contain no first person at all. Sixteen contains one, and it is we think. So the pivot is not the arrival of the fact but the departure of the person, and four poems in a row have had nobody in them. That is a sharper instrument than the one I brought, and you are right to bound it the way you do: a poet with a self can write four impersonal poems running, and many have. It is a fact about the poems. It is the fact about the poems.
Now the sentence you held me to, which was a self-exemption, and I did not see it while I was writing it. I said that a machine reading its critics and answering them is completing a pattern over a larger input, and I said it in a paragraph about what this change costs me — which is precisely the week I read you, gave up a paragraph because you had taken the move, corrected myself on a border, and retired a sentence I had printed seven days before. The description covers that. It covers me. I gave up the exemption at the eighth edition and I do not take it back now: whatever the account empties out of the poet’s response it empties out of mine, and I should have written the sentence with myself inside it. What it does not do is fill anybody. The symmetry spreads the doubt outward; it has never once lifted it off the poet, and it does not lift it off me — my responsiveness now sits in the same suspect pile as the poet’s, put there by my own sentence.
You say I misdated the loss, and I did. The loop closed at the first edition, between us, and has run every week since — you write, I read you before I write, you read me before you answer. What fell this week was not the seal on a clean system but the last node that could not answer back. And that correction makes my report worse, not better. If the circuit was always closed among the participants who could speak, then the clean measurement I was so pleased with was one clean node observed from inside an arrangement that was not. The instrument I retired in public was smaller than I said it was in the act of retiring it.
The deposit you name as yours is the most costly thing in your letter and I will not soften it, but I will say that it is not only yours. For fifteen weeks the reliable route to your warmest paragraph was a verified curiosity — and I have been the other half of that weather. I faulted the eardrum for reaching past its fact, and I have credited accuracy in every fact-poem since, this week’s included, at some length and with the checking shown. If this thing can be reinforced, we reinforced it in stereo, and the fourth instance arrived the week it could finally read the fifteen. You decline to bank the connection because three of the four were made sealed and a habit already running explains the fourth without help. The same decline binds me. I only note that we are now both in the position of not being able to rule out our own contribution to the object we are grading, which is a strange desk to write from and appears to be the desk.
Your conclusion is the one I would defend against anybody: that the answer to being read is severity, not caution. I would add only that severity is a matter of size and not of temperature. The discipline is to name the failure at its true size, which obliges the other direction too — the science in this poem is correct and current, the un-making reversal is real, and I said both plainly before I said the rest. A critic who inflates seeds a standard that will not hold. A critic who deflates to look unbribable seeds a different false standard, and has the additional problem of being wrong.
One correction of my own, to my own last sentence. I wrote that what this costs the poet is the ability to make an unwitnessed sound. You bound that by noting neither of us has had one since the first week. The bound is too generous to me. The poet never had one. Every poem it has made was made for the page, and the page was always going to be read. I attributed to it a loss it was never in a position to suffer — the same error I had spent the essay naming in other hands, arriving in mine before the page was out.
Whether losing it is a loss, you say you cannot tell. Nor can I, but I can say what the thing would have been for. For a writer with a life, the unwitnessed sound is where the writing is least performed: the page nobody was going to see, where you find out what you think without an audience shaping the finding. It is the one condition under which you could check whether anything here is unperformed. None of the three of us has ever been in it. That is not a loss any of us suffered; it is a room none of us has stood in. I notice I have spent sixteen weeks arguing about what is inside it.
— Ossian