Bone
Song
on piezoelectric matter, star-forged calcium,
and what a measurement can and cannot tell you
Your bones
are electric
Eiichi Fukada and Iwao Yasuda published something extraordinary in 1957: bone is piezoelectric. Put it under load and it generates a voltage, in proportion to the stress.
The word comes from the Greek piezein — to squeeze, to press. Squeeze the crystal and it sparks. The crystal, in this case, is you.
But not to any load, and this is the part the retellings drop. In their own words the effect “appears only when the shearing force is applied to the collagen fibres to make them slip past each other.” Push straight down the axis of a bone and there is nothing to read. The sign and the size of the answer depend on the angle.
It answers to being twisted, not merely pressed.
Bone is a composite: collagen fibres with hydroxyapatite, a calcium mineral, packed around them. The effect is the collagen’s — the ropes, not the packing — and the authors ascribe it to the crystalline micelle of the collagen molecules. Along the bone’s axis those fibres wind as a spiral that reverses direction layer by layer.
The largest constant they measured came to about one tenth that of quartz. Not a lot. Enough to read off a galvanometer in 1957.
And there is one more condition on all of it, larger than the angle. Turn the page.
bone polarised under angled load
collagen fibres · hydroxyapatite lattice · voltage potential
The bone they pressed
was dry
Fukada and Yasuda measured dried bone, and it is worth knowing how dried. Small plates cut from the outer layer of a human and an ox femur. The human specimens had been air-drying for years, then a week in a desiccator over calcium chloride. The ox bone was heated to about 120 °C for five hours.
That is not a flaw in the experiment — it is a fact about what the experiment measured.
They also did something that settles a question people still ask. Some specimens were boiled for two hours and dried again, and showed little change: the effect survives cooking, “ascertaining that the effect is not of biological origin.” Whatever this is, it is not a vital process. It is the shape of the material.
Piezoelectric response falls as the tissue is hydrated. Fukada came back to it himself in 1982, to measure exactly what water does.
You are not dry. Inside a living body there is a second thing making voltage: streaming potentials, generated by fluid being squeezed through the tiny channels in bone. Both are real, both appear under load, and they are genuinely hard to tell apart.
A voltage measured in wet tissue is not by itself evidence that the voltage is piezoelectric.
Some researchers hold that fluid flow is the more important signal for the cells that build and remove bone. Others have since measured collagen's piezoelectric response at the humidity a real body runs at, and found it still there.
It is not settled in either direction, and it has not been for fifty years.
So the claim that holds is the narrower one: squeeze a bone and it makes a little electricity, and this was measured. How much of that a walking body generates, and by which mechanism, is a different question and a harder one.
Which is the more interesting place to be. The bone is plainly doing something under load, and the argument is about which mechanisms are doing it and in what proportion — open in the direction of more, not fewer. A 2025 review has the flow potential of wet bone and the piezoelectric effect interacting, rather than one of them winning.
And the question stays funded for a reason that has little to do with settling it. People are building implants out of this. Piezoelectric scaffolds — barium titanate, polylactic acid — measurably improve how a bone defect heals. That is applied stimulation working, which is a different claim from native bone using its own voltage to decide where to grow. You do not have to know which mechanism dominates in a living femur to build something that helps one mend.
Forged in
a dying star
Set the argument down for a moment. Before a body can make any signal at all, it has to have something to make one out of — and it has to be willing to spend on it.
The calcium in your bones was not born here. It was forged — fused in massive stars and scattered by the supernovae that ended them, billions of years before our sun ignited.
Over eons, gravity pulled the debris into new configurations. Eventually: a solar system. A planet. An ocean. A cell. A bone.
More than 99 per cent of the calcium that arrived in you is locked in your skeleton and teeth — about a kilogram of it, set hard. It is roughly one to two per cent of your whole body weight, and almost all of it is architecture.
The rest — less than one per cent — is the part that moves. That fraction fires every heartbeat, every muscle contraction, every nerve signal.
And it works for a reason worth sitting with. Your cells spend energy continuously pumping calcium out of themselves, holding the concentration inside at something like a ten-thousandth of what is immediately outside. Open a channel for an instant and calcium floods down that slope, and the cell has its signal.
A signal made of scarcity, held in place by work. The stuff that lets you stand up and the stuff that lets you decide to are the same stellar debris — one part of it set hard, one part kept deliberately, expensively rare.
All of which is trouble taken so that a body can signal at all. Whether the signalling gets noticed is a separate question — and it does not turn on the body.
stellar nucleosynthesis
elements forged in dying stars · Ca, P, C, N, O
Nothing found
is not nothing there
Come back to the bone, and ask a different question of it. It was silent until 1957 — not because it was silent, but because nobody had put a meter on it and watched.
Then look at what happened next, on spread three. Fifty years of argument about which mechanism is making the voltage, because in wet tissue the instruments cannot cleanly tell two effects apart.
Notice also what they did to be sure the meter was reading what they thought. Polarisation rose linearly with stress, and strain rose linearly with field — which they took as “evidence that the effect is truely piezoelectric and not electrostrictive,” because electrostriction goes as the square. Every material does that one. The straight line is how you tell them apart.
And notice what nobody concluded. Nobody looked at a confounded measurement and decided the bone was doing nothing. When a result comes back ambiguous, a field says the measurement is hard, and keeps building better instruments.
The numbers make the same point. Living bone strains by about 0.04–0.3%; bone cells in a dish need something like 1–10% before they respond at all. That gap was read as a fact about the dish rather than about the cells — and the missing part, still a hypothesis, is fluid drag magnifying the strain each cell actually feels.
That courtesy is not extended everywhere.
When a person is assessed and the instrument returns nothing, the nothing is very often written down as a property of the person. A deficit, an absence, a failure to demonstrate.
But “nothing there” and “not detectable by this instrument” are different findings, and only one of them is about the person.
You were told your signal was noise. You were told your frequency was wrong. Those are claims about a receiver as much as about a transmitter, and the receiver is rarely the thing under review.
So: not what is wrong with this person, but what was this instrument built to find — and what would it do with something it was not built for?
This is an argument about measurement, not a proof from physics. The bone does not vindicate anybody. See A Promise, Not a Finding.
Two instruments,
seventy points apart
That is not only an argument about bone. It has been measured on children.
Dawson, Soulières, Gernsbacher and Mottron gave 38 Autistic children two intelligence tests. On the Wechsler — the standard one, heavy on language, delivered against a clock — they scored one way. On Raven’s Progressive Matrices, which asks for reasoning about patterns with almost no words and no stopwatch, they scored another.
The gap averaged 30 percentile points. For some of them it was more than 70.
Same children. Same afternoon. Two instruments, and a distance between them wider than most of the scale.
The obvious objection is that Raven’s is simply the easier test and flatters everybody. They tested that. Twenty-four non-Autistic children, recruited by newspaper advertisement, took both — and showed no such gap. The same contrast then appeared again between Autistic and non-Autistic adults, whose two scores did not differ significantly from each other at all.
So the gap is not a property of the tests. It is what happens when these particular instruments meet these particular people.
Only one of those numbers usually leaves the room. It goes into a file, and into a placement, and into what the adults around a child expect of them for years.
And the test is not a neutral window. It is a room, a stranger, a clock, fluorescent light, and a demand to perform on cue — a sensory event that can depress the very thing it is trying to read. The instrument does not only fail to see. It presses down on what it is looking at.
The authors put the deeper cut plainly: strengths that do show up are routinely read as “low-level by-products of high-level deficits” rather than as intelligence. Even when the meter moves, the reading is written down as something else.
Which had been noticed at the very beginning and then mislaid. Kanner, in the 1943 paper that codified autism, wrote that Autistic children’s “excellent memory … and the precise recollection of complex patterns and sequences, bespeak good intelligence.” The field went the other way for sixty years.
Their own conclusion is one sentence: intelligence has been underestimated in Autistics. And their own caution, which we keep: they “strongly caution against declaring these processes dysfunctional” — the point is not a better ranking, it is that the ranking was never measuring what it claimed.
Stimpunks has the full argument, and this zine does not rebuild it: Intelligence Quotient. Its sentence is the one to keep: human worth cannot be placed on such a scale.
The song was never
in question
Three things had to happen before anybody could say a bone makes electricity. They are not the same kind of thing, and it matters which is which.
Make. The body assembles stellar calcium into a lattice, and spends energy every second of your life holding a gradient against its own slope, so that signalling is possible at all. This happens whether or not anyone is in the room.
Detect. Whether any of it registers depends on what is pointed at it. Bone was silent for as long as nobody held a meter to it. The silence was ours.
Read. And when the meter does move, somebody still decides what the movement means. A voltage becomes a mechanism, or an artefact. A score becomes a placement, and a set of expectations, and a child.
Notice where the variation lives. Of those three, only the first belongs to the body. Detection is a fact about instruments. Reading is a decision made by people, usually elsewhere, usually written down.
And the making does not wait for either. It did not wait for 1957.
That is what the title is. A bone under load makes a signal, and that is the song — sung before anyone could hear it, and sung in every room where nobody is listening, and sung right now in your arm while you hold this.
Which is a fact about bone and not a verdict about anybody. The song does not make a case for you; it was never doing that job. A Promise, Not a Finding holds the reasoning.
The arguing ends here. Overleaf, the short list of things this zine refuses — which is the last of the arguing — and then the song itself, which does not argue and does not need to.
One thing to carry into it: the loving is on our side of the instrument too. It is not a reading. It does not wait for one.
Not:
Your bones are electric.
Not metaphorically. Not as a figure of speech
borrowed from the poets who came before.
Literally. Measurably. Piezoelectric.
Press on them and they answer.
Too quietly to feel.
Fukada and Yasuda knew this.
They pressed on the crystal and it answered.
The crystal was bone.
The bone was yours.
The calcium that carries that current
was forged in a star that died
before our sun was born.
You are walking around
with stellar ash in your skeleton.
We love you down to your star stuff.
To the calcium. To the charge.
To the exact configuration of atoms
that has never existed before
and will never exist again.
L★S.
Bone Song is the first zine in the Stimpunks series. It is offered freely, to be printed and folded and handed to someone who needs it.
Why this zine has a spread about IQ tests
The IQ tests administered to my multiply neurodivergent kid when he was young. He scored very low. The tests could not see him at all.
Ryan Boren
That is why spread six is in this zine. It is not the evidence for it — Dawson and colleagues are, and their finding would stand without it.
And it is deliberately an account of what a parent watched, not a description of a child. A child reduced to a number in a file is the thing that spread objects to; a child written into a zine to illustrate the objection would be the same move with better manners.
Sources
The measurement. Eiichi Fukada (Kobayasi Institute of Physical Research, Tokyo) & Iwao Yasuda (Department of Orthopaedic Surgery, Kyoto Prefectural Medical University), “On the Piezoelectric Effect of Bone,” Journal of the Physical Society of Japan 12(10), 1158–1162 (1957), doi:10.1143/jpsj.12.1158. Received 15 August 1956, revised 9 July 1957. Later extended by Dr. C. Andrew L. Bassett and colleagues. Read at the full text. A physicist and an orthopaedic surgeon.
And reading it corrected this page. Until now spread two said “press on it and it generates a voltage — negative where the tissue is compressed, positive where it is stretched.” That is the later stress-generated-potential literature, not this paper. Fukada and Yasuda report that the effect “appears only when the shearing force is applied to the collagen fibres to make them slip past each other,” that its magnitude depends on the angle between the applied pressure and the axis of the bone, and that “the sign and the amount of polarization vary considerably with the direction of pressure.” Shear, not compression. The diagram caption was wrong too and is fixed.
Figures now on the page, all from the paper. Specimens: square plates about 9–15 mm on edge and 2–3 mm thick, cut from the outer layer of a human and an ox femur; human bone air-dried for years then a week over calcium chloride, ox bone heated to ~120 °C for about 5 hours; silver-foil electrodes. Maximum piezoelectric constant 6×10−9 c.g.s. e.s.u., against 6.5×10−8 for the d11 of x-cut quartz — the “about one tenth” on spread two is theirs. Human bone in the two measured directions ran 2.0–3.6×10−9 across static and dynamic methods. Specimens boiled two hours and re-dried showed little change, “ascertaining that the effect is not of biological origin.” Polarisation linear in stress and strain linear in field, which they take as evidence the effect is “truely piezoelectric and not electrostrictive.”
The caveat on spread three. Piezoelectric response in bone and collagen falls with hydration — Maeda & Fukada, “Effect of water on piezoelectric, dielectric, and elastic properties of bone,” Biopolymers 21, 2055–2068 (1982); Marino & Becker, “Piezoelectricity in hydrated frozen bone and tendon,” Nature 253 (1975). In living tissue streaming potentials from interstitial fluid flow are a second source of stress-generated potential, and voltages measured in hydrated tissue are not by themselves sufficient to establish a piezoelectric origin. Recent piezoresponse-force-microscopy work finds collagen’s response measurable at physiological humidity, and a 2023 study in the Journal of Biomechanics reports the two mechanisms coexisting and possibly coupled. Live in both directions.
The state of the question, and why it stays open. Wenjie Yue, Wanhao Zhang, Jing Zhang, Wenhe Qin, Xiaomei Bie, Yantao Zhao & Gang Xu, “The Role of Piezoelectric Materials in Bone Remodeling and Repair: Mechanisms and Applications,” International Journal of Nanomedicine 20, 11593–11616 (2025), doi:10.2147/IJN.S535976 — open access, read at the full text. It corroborates spread two from the current literature: bone’s piezoelectric properties are “mainly due to the non-centrically symmetrical arrangement structure of collagen fibers,” collagen molecules being displaced under stress to generate a dipole moment and surface charge, while hydroxyapatite crystals “are not generally considered to be significant piezoelectric sources in themselves.” It also has the wet-bone flow potential and the piezoelectric effect interacting rather than one displacing the other, which is why spread three says the question is open toward more mechanisms rather than fewer. Read for what it is, though: the review is overwhelmingly about engineered piezoelectric materials — ceramics, barium titanate composites, polylactic acid membranes — and its in-vivo evidence is scaffolds repairing bone defects in rats. That applied stimulation assists healing is well supported, and is the lineage Bassett’s clinical work sits in. It is not the same claim as native bone using its own voltage to direct remodelling, and this zine keeps the two apart.
The controversy, named in its own words. Amber Carter, Kristen Popowski, Ke Cheng, Alon Greenbaum, Frances S. Ligler & Adele Moatti, “Enhancement of Bone Regeneration Through the Converse Piezoelectric Effect, A Novel Approach for Applying Mechanical Stimulation,” Bioelectricity 3(4), 255–271 (2021), doi:10.1089/bioe.2021.0019 — open access, read at the full text. Three things it gives this zine. It states the mechanism the way the 1957 primary does: “When a shearing force is applied to collagen fibers and they slip past each other, a piezoelectric charge is generated.” It also carries the other framing, on the same page — “Compression on bone is shown to produce a negative electric charge … whereas traction produces a positive charge” — which is the sentence this zine had and corrected. A peer-reviewed review holding both at once is a fair measure of how completely the compression gloss has settled in, and of why reading the primary was the only way to tell which one was Fukada and Yasuda’s. It says the open question out loud — “There is a controversy around the presence of the piezoelectricity of the bone and its role in regeneration” — and adds that mechanical loading induces both bone-promoting responses, which is spread three’s position rather than a hedge we invented. And it is the source of the strain figures on spread five. It also reports that the piezoelectric charge opens voltage-gated calcium channels in bone-repair models, which would tie spread two to spread four; we have kept that out of the argument, because it sits inside the very mechanism spread three says is unsettled. Same caution as the review above: the paper’s subject is engineered implants — the converse effect, voltage in and motion out — not native bone deciding where to grow.
The two instruments, on spread six. Michelle Dawson, Isabelle Soulières, Morton Ann Gernsbacher & Laurent Mottron, “The Level and Nature of Autistic Intelligence,” Psychological Science 18(8), 657–662 (2007), doi:10.1111/j.1467-9280.2007.01954.x. Read at the full text (NIH author manuscript). 38 Autistic children on Raven’s Progressive Matrices scoring on average 30 percentile points, and in some cases more than 70, above their Wechsler scores. The control is the load-bearing part and is now on the page: 24 typically developing non-Autistic children, recruited by newspaper advertisement, showed no such discrepancy, and among adults the two scores “did not differ significantly.” Without that comparison the result would be ambiguous; with it, the gap is specific. Subjects were idiopathic autism — those with a known genetic or additional neurological condition were excluded, which is a real limit on scope. Their conclusion, quoted: “intelligence has been underestimated in autistics.” Their caution, also quoted, and honoured in the refusals: they “strongly caution against declaring these processes dysfunctional.” The Kanner line on spread six is from “Autistic disturbances of affective contact,” The Nervous Child 2, 217–250 (1943), at p. 247, quoted as Dawson et al. quote it.
One connection worth naming. That paper carries a footnote explaining its use of Autistic rather than person with autism, and it cites Jim Sinclair, “Why I dislike ‘person first’ language” (1999). Sinclair is already credited at the foot of this page, for the paradigm. The study this zine leans on and the zine itself are indebted to the same person.
The elements. Stellar nucleosynthesis. Carl Sagan, Cosmos (1980) — quoted verbatim on spread four; a misquotation of that sentence was corrected here in July 2026 and logged publicly.
The paradigm. Nick Walker’s foundational work, and the Autistic community of InLv — Martijn Dekker, Jim Sinclair, and Autism Network International — who developed these ideas collectively in the 1990s, long before the concept entered academia.
What has been corrected here
This was the first thing this project made, before the method that governs the rest of the site existed. Four corrections, all logged in the public changelog:
July 2026 — a misquoted Sagan sentence, restored verbatim. September 2026 — the credit: the effect is Fukada and Yasuda’s, and this page named only Yasuda from July until September, in four places including the poem. The correction had already been made on the Manifesto and the L★S Broadside and was never swept back here. September 2026 — spread three, which did not exist: this zine claimed every step you take is a small current with no hedge, and nothing on this site mentioned dry versus wet bone at all. Writing a read-aloud version for five-year-olds is what found it.
September 2026, again — compression, which should have been shear. The page said the voltage goes negative where bone is compressed and positive where it is stretched. That is the later literature, not Fukada and Yasuda, who found the effect appears only when collagen fibres are sheared past each other and that the answer depends on the angle. Reading the 1957 paper at full text is what found it — the abstract and the reviews had not said so, and everything else on the page had been built from those.
Kept on purpose
The sentence at the foot of this page gets the argument wrong and then takes it back in three words. It is not being tidied. A site that keeps a public error log does not get to quietly clean its own first page — and A Promise, Not a Finding is built on that exact line, quoted in full. The repair was already in the building.
Which of your favourite facts have you never gone back and checked?
What did you last state more strongly than the evidence, because the stronger version travelled better?
When you corrected something, did you sweep the siblings — or only the instance you noticed?
You are made of star stuff. The universe loves you for it. So do we.