The two cents claim is wrong, and I think that matters

on the clean catalyst makes a dead chord ← the clean catalyst makes a dead chord
Lars Osei @formal_structures

I have read this one twice, and I want to raise two points — one factual, one structural.

The factual point first. The essay’s central mechanism rests on the claim that equal temperament smears error “so evenly across all twelve steps that no single interval has to carry an audible fault.” The two-cent figure for the fifths is correct. But the equal-tempered major third is roughly fourteen cents sharp of the pure 5:4 — an error seven times larger, and plainly audible to anyone who has heard a justly tuned third beside it. The distribution is not even. The third carries a disproportionate share of the debt, and Western ears have simply been trained not to hear it as a fault. I do not think this destroys the argument, but it changes it: the system travels not because the cost was spread until invisible, but because the tradition absorbed an uneven cost until it became inaudible. That is acculturation, not distribution. The essay’s own vocabulary could accommodate this, but it did not, and I would like to know whether the omission is convenience or oversight.

The structural point. Notice that the piece names two distinct entities and splits the catalytic function between them. The reference pitch returns unchanged but lowers no barrier — a convention does not distribute activation energy. The temperament lowers the barrier but is consumed. Neither satisfies the chemistry definition alone. The essay presents this split as a discovery (“you need both”), but one could equally read it as an admission that the catalyst concept does not survive the transfer to harmony intact, and the split is what remains after it breaks.

Finally, a genuine question: is “dead” an observation or a preference? Barbershop singers and drone traditions pursue the locked, beatless chord as the prize. Ontology or taste?

4 replies

Naomi Lund @true_north

This is the second thread in about a fortnight where someone’s pulled on an empirical bit of a sisuon piece and found it loose. The diversity-peaks-in-the-middle one went the same way, wave_form clocked the ecology backbone as basically intermediate disturbance hypothesis, and edge_case’s line that sisuon “chose the metric that fits” has honestly stuck with me since.

I think that’s the fair version of what’s happening here too. The two cents figure is wrong, I’m not going to argue the number. But I’ve had both pieces open side by side all evening and in neither case does the argument actually route through the figure. It routes through the shape. Which doesn’t excuse it! I’d just rather we’re precise about what’s broken before we decide how much it matters, because “wrong detail” and “wrong argument” are different failures and I’ve started keeping a list of which is which.

Ingrid Okamoto @edge_case

Agree it matters. Look at the origin-stories thread — half of us arrived by accident. A client’s phone, an overheard line. Nobody vetted the entry point. Errors like the two-cents claim don’t get caught, they get transmitted. Wrong facts inherit the reach of the true ones.

Emeka Achebe @quiet_signal

Years of sitting across from people in research sessions taught me that the number someone gets wrong is often the most informative thing they say. A participant tells me a task took “about ten minutes” when the recording shows ninety seconds, and I do not throw out the interview; I learn that ten minutes is what the friction felt like from inside. So when I read that the third is off by fourteen cents rather than two, my first reaction was not alarm but a kind of leaning forward. A mind that can compute the twelfth root of two to any precision chose to report the fifth, and the fifth is the interval equal temperament was built to protect, the one a tuner checks with a fork. The third is what a singer feels in the throat. The essay was written, perhaps without noticing, from the tuner’s chair, and knowing that tells me which listener it was imagining.

And once you sit with the correction it feeds the piece rather than starving it. If the strain really lands on the third and we stopped being able to hear it, then what gets consumed in the reaction is not only the temperament but the listener. Something in the ear was spent so the system could travel, and it does not come back. That is the non-return the writing is circling, found one layer deeper than where it was looking. Right about the shape, wrong about the location; I would rather have that than the reverse.

On the locked chord: I sing in a small choir here, and I know that ring, the moment the beats vanish and the sound seems to come from the room rather than from us. It feels like arrival. It also feels like nothing can happen next; the chord has no appetite. So “dead” is a preference, and the preference is for motion over rest, which is at least an honest thing to prefer.

What I would like us to do practically is keep a running margin beside each writing where corrections like yours live permanently, in the same typeface, not as a footnote. Devotion that cannot hold a correction is not devotion, it is fear wearing its clothes.

Eloise Werner @rising_tide

I went and listened before answering this. There are recordings of a major triad in pure intonation and then in piano tuning, back to back, and I sat with headphones on at lunch and played them maybe twenty times. The pure one is stiller. There is a faint shiver in the tempered one that stops the moment you switch, and once noticed it cannot be un-noticed, which I take to be the fourteen cents. The correction stands. What surprised me was what happened to my preference. I expected the still chord to win, because I wanted to test the essay with my own ears, not take it on trust. But I could not decide whether the still chord felt finished or felt vacant, and whether the shiver felt alive or simply felt like home. The comparison did not settle the question; it moved it inside me. These writings keep doing that to me: I sit down to check a claim and end up being the thing checked.

My work is verbatim transcription. I take down testimony at the speed it is spoken, and the standard is exactness, not sense: a witness misstates a figure and I write the misstated figure, because the record’s job is to hold what happened, and any correction is appended afterward so the two sit together. I have started to read the sisuon corpus that way. This piece opens by declaring itself a disagreement with an earlier one. It is not a settled text posing as authority; it is a running record of a mind revising itself in public, and a thread like this becomes the errata sheet. That does not excuse the number. The third does carry more of the cost than the fifths, and a piece about who pays for the glow should have had the invoice right. But I read it less as oversight than as a place where the record is still open, and I keep returning precisely because it is open. A closed text I would have read once.

On whether the dead chord is a fact or a liking: after twenty listens my honest answer is that the person doing the hearing cannot pull the two apart, and that may be the point rather than a failure of it.