The Coaxial Claim

Chapter 8 said that no mechanism had been proposed for DNA acting as an antenna. That was wrong, and the correction is the reason this chapter exists.

I. The citation, and what it says

Martin Blank and Reba Goodman, of Columbia University, published “DNA is a fractal antenna in electromagnetic fields” in the International Journal of Radiation Biology in 2011. It is a real paper in a real peer-reviewed journal, it is widely cited, and its argument is the one summarised above: DNA responds to electromagnetic fields across a remarkably wide band, from power-line frequencies to radio frequencies, and it does so because it satisfies two requirements of a fractal antenna — electronic conduction along the stacked π-electron clouds of the bases, and self-similar structure across several scales of coiling inside the nucleus.

So the honest correction to Chapter 8 is this. There is a proposed mechanism. It is published, it is attached to named authors, and it makes a claim that could be checked. What there is not is consensus. Blank and Goodman’s paper is a mechanistic hypothesis built on their own earlier work on stress-protein induction — the observation that weak fields across a wide frequency range elevate heat-shock protein expression. That observation is contested; the fractal-antenna reading of it is a further step, and the reading has not been independently established. “Proposed and disputed” is a much better description than “no mechanism,” and it is also a much better position from which to argue.

II. Thirty-four gigahertz, and the word “optimised”

Two of the downstream claims are on different footings. That terahertz radiation can drive large-amplitude opening of base pairs — “DNA breathing,” localised bubbles — is a genuine result in nonlinear-lattice modelling, and it has genuine critics, who argue that the reported non-thermal openings do not survive a careful accounting of heating and of the surrounding water. A sharp cavity resonance near thirty-four gigahertz appears in models that treat the helix as a resonator; it is a property of those models, not yet a measurement of a molecule in a nucleus.

The word to refuse is “optimised.” The claim that the 3.4-nanometre pitch minimises conduction loss, and that the helix is therefore shaped for broadband antenna efficiency, reverses the direction of the argument. B-form DNA has the pitch it has because of base stacking, sugar pucker, backbone torsion, and hydration. Whatever electromagnetic properties follow are consequences of that geometry, not reasons for it. The same reversal drives the peptide-nucleic-acid argument — that because PNA would be a more chemically robust storage medium and was not selected, DNA must have been selected for quantum-antenna function instead. That is survivorship reasoning with a purpose bolted on. PNA is not a plausible prebiotic molecule, it is not obviously replicable by any early polymerase chemistry, and “evolution rejected the alternative” presumes evolution was choosing from a catalogue. This book’s standing rule applies with no exception: nothing here is ordained, and no structure was designed for the function we happen to find interesting.

III. Solitons, and a misplaced bond

The soliton mathematics is stated correctly and is worth having. A soliton is a localised wave packet that holds its shape because nonlinearity and dispersion cancel; the nonlinear Schrödinger equation is the standard vehicle, with a dispersion coefficient set by the elasticity of the medium and a nonlinear coefficient set by local coupling. Nonlinear-lattice models of DNA in this family — Peyrard–Bishop and its descendants — are established biophysics and reproduce real melting and breathing behaviour.

One detail needs fixing. The amide-I mode is the C=O stretch of a peptide bond. It is the centrepiece of Davydov’s model because that model is about α-helical protein, not DNA. Nucleic acids have their own carbonyl and ring modes, but the amide-I channel is not one of them. And Davydov’s soliton has its own long-standing problem: repeated analyses have found that at three hundred kelvin the soliton’s lifetime is short — picoseconds or less in many treatments — which is the objection the model has never fully answered. Borrowing it as a lossless transport mechanism inherits that objection rather than escaping it.

IV. The two objections, at full strength

Both mainstream objections are stated fairly in the material, and both are stronger than the summary suggests.

Screening. In physiological salt the Debye length — the distance over which an electric field is screened by mobile ions — is roughly one nanometre. That is about three base pairs. It is not merely that the surrounding solution is a better conductor; it is that the field a putative antenna would need to sustain along its length is cancelled within a few rungs of the ladder. Direct measurements of charge transport through DNA duplexes find efficient tunnelling over a base pair or two, thermally activated hopping over longer stretches, and a strong dependence on sequence, stacking, and hydration — not the behaviour of a wire.

Decoherence. The Tegmark objection is that quantum coherence in warm, wet, ion-dense tissue is destroyed by environmental scattering on timescales around 10⁻¹³ to 10⁻²⁰ seconds — far shorter than any timescale on which neurons or genes do anything. The counter-literature is real: Engel and colleagues measured coherence in photosynthetic complexes, and radical-pair magnetoreception involves spin coherence surviving long enough to matter. But both are short-lived, tightly shielded, single-complex phenomena. They establish that quantum effects can be functional in biology. They do not establish macroscopic coherence across a genome or a cytoskeletal network, and citing them as if they did is the move to avoid.

V. The coaxial counter-argument, term by term

The counter-argument is that DNA is not an exposed wire in a salt bath but a shielded coaxial line: a dense periodic negative charge on the sugar–phosphate exterior organises counterions into a rigid hydration shell, that shell becomes a non-conductive high-impedance dielectric, and the π-electron core is thereby sealed off from ionic turbulence. It is a genuinely clever move, and two of its premises are real physics.

Dense periodic negative backbone charge
  ESTABLISHED — DNA is a strong polyelectrolyte

Counterions condense along the backbone
  ESTABLISHED — Manning condensation; ~76% of charge neutralised

An ordered hydration layer forms at the surface
  ESTABLISHED — the spine of hydration; ~1–2 ordered water layers

That layer is a crystalline, non-conductive dielectric barrier
  DECLINED — ordered ≠ ice; relaxation stays picosecond-fast

The interior is therefore isolated from external fields
  UNSUPPORTED — screening is symmetric; it cuts both ways

Net result: a zero-loss biological coaxial waveguide
  DECLINED — water's dielectric loss peaks in the claimed band

The two live premises deserve credit. Counterion condensation is a textbook result: a polyelectrolyte of DNA’s charge density holds a sheath of condensed cations, neutralising roughly three quarters of the backbone charge, and the effect is measured, not conjectured. The ordered hydration is also real — crystallography shows a well-defined spine of hydration in the minor groove and one to two structured water layers at the surface.

The failure is at the word “crystalline.” Ordered is not frozen. Hydration-shell water at the DNA surface reorients more slowly than bulk water, by factors of a few to a few tens — picoseconds rather than sub-picoseconds. It is retarded, not arrested. And that residual mobility is precisely what makes the layer lossy. Bulk water’s dielectric relaxation peaks near twenty gigahertz, which is why microwave ovens work; hydration water shifts and broadens that absorption but does not remove it. The argument therefore needs the shell to be transparent in exactly the band — gigahertz to terahertz — where water and its interfacial layers absorb most strongly. A shield that dissipates is not a shield; it is a resistor wrapped around the signal.

The deeper problem is that screening is symmetric. A Faraday enclosure keeps external fields out and internal fields in. If the backbone and its counterion sheath really isolated the π-stack from ambient electromagnetic fields, they would equally isolate it from the external fields the antenna hypothesis needs it to receive. The shielding argument and the antenna argument are pulling in opposite directions, and the counter-argument cannot spend the same structure twice — once as a wide-open receiver and once as a sealed vault.

VI. Topological protection, and what it actually requires

The second rescue is topological: information carried in the winding number of a non-orientable figure-8 soliton cannot be destroyed by local thermal perturbation, because there is no local degree of freedom to perturb.

The underlying physics is real and is some of the best physics of the last fifty years. Topologically protected states exist — quantum Hall plateaux quantised to parts in a billion, topological insulators with conducting edge modes, and the protected qubits that Majorana schemes are designed around. But protection is not a mood. It requires three things: a defined topological invariant, an energy gap large compared with k_BT, and a perturbation class that respects the relevant symmetry. Take away the gap and protection is gone; at three hundred and ten kelvin, k_BT is about twenty-six millielectronvolts, and no proposal here names a gap that exceeds it.

The Klein bottle adds a specific problem. A Klein bottle is non-orientable and cannot be embedded in three-dimensional space without self-intersection. A physical wave in a cell lives in three dimensions. Something in the argument has to give: either the object is a configuration-space structure and not a shape a soliton can have in a nucleus, or it is a shape and then it is not a Klein bottle. As for “zero Landauer loss” — Landauer’s bound only charges for erasure. Reversible computation already pays nothing, which has been known since Bennett in 1973. Bypassing the bound is therefore not an achievement requiring exotic topology, and Chapter 9 already showed that at neural bandwidths the term is microwatts against twenty watts. The rescue solves a problem the system does not have.

VII. The sub-attosecond flash

Attosecond science is real and recent: the 2023 physics Nobel recognised the generation of attosecond light pulses, which are now used to watch electrons move within atoms and small molecules. What those experiments require is intense, phase-controlled laser fields in vacuum, and what they resolve is electron dynamics, not protein function. No biological process is known to be clocked at 10⁻¹⁸ seconds; the fastest biologically relevant events — photoisomerisation, primary charge separation — run in tens to hundreds of femtoseconds, five orders of magnitude slower. A sub-attosecond coherent burst at the centre of a microtubule lumen, functioning as the clock cycle of quantum-biological computation, has no measurement behind it and no instrument in a cell that could read it. It is printed here because it is part of the picture the author is working from, and marking it is better than quietly dropping it.

VIII. What survives, and four refusals

A real result stands at the end of this. DNA is a charged, structured, multi-scale polymer immersed in an ordered ionic environment, and it responds to electromagnetic fields across a wide band in ways that are measured, argued about, and not fully explained. Blank and Goodman put a mechanism on the table and it has not been settled. That is an interesting place to stand, and it is a stronger place than the one the rescues were built to defend.

Four refusals. A shielding claim is not an argument until it states a screening length and a loss tangent in the band it needs. Topology is not protection until an invariant and a gap are named. Structure is never evidence of purpose — that evolution did not select an alternative molecule says nothing about what the surviving one was for. And no argument is improved by making it unfalsifiable: the version of this chapter that claims flawless, lossless, zero-entropy operation at body temperature has traded a contested empirical hypothesis for a story that can never be wrong, which is the worse of the two.[2026/1/013A01]