The Value, Refused in Advance

This paper claims no value for α.

The chain of custody. The value has never been claimed at any link in the corpus, and every link says where it goes next. AP15 derives that ε couples to the connection with some strength, states that symmetry, dimension and conservation fix the action only up to that coupling — which is the standard position and not a private difficulty of this corpus — and hands the value on by name to AP24 and AP28, under KS-4 live and hard. AP24 takes it up through the six faces and leaves it determined but not derivable from within, at §5.4. This paper registers every road walked since and refuses it. Three locked papers and this one, and no link assumed the number quietly. That record is the answer to the charge, and it is on the page rather than in anyone's recollection.

The author sees and describes eyes open. That is his method and this paper does not bind it. Nor does it bind the direction. The road may run from his description to a figure, or from a candidate form to his structural check of it; he has ruled that the route matters less than the road, and that the only goal is to clear the road. What is bound is what happens between a construction and a comparison, because that is the only place a number can steer a form without anyone seeing it do so.

So, if a value is ever sought, it is sought under PROTOCOL-alpha-value, filed beside this paper. Every form tried, whoever proposes it and whichever way round, is entered in a public register with its date and its landing before it is compared, so that the look-elsewhere denominator is known — KS-30.4 found 2,971 five-leaf forms at the proton's accuracy, and that number is the standard any α form must beat by structure and not by digits. A form's structural reading must be one the corpus already rules, not one written for the form. The construction is frozen and lodged with the independent verifier before the comparison is run, and the comparison is run by a desk that did not build it. KS-35 and KS-36 adjudicate. The result is published hit or miss.

The denominator, measured. A second desk, testing the bound §10 has now withdrawn, constructed a form that lands. It was built to break a sentence and was never offered as a candidate; it is entered in the register as R1 with outcome VOID, which is what a register is for. What it showed is the size of the field. From a pool of coefficients made of the picture’s own counts, quadratics alone produce 5,883 positive roots within one per cent of α, 35 within a hundred parts per million and 2 within one. Roots are not equations: the pool contains a number and its multiples, so one equation appears several times over. Normalising each quadratic by its leading coefficient and comparing the results exactly, rather than in floating point, collapses those to 3,763, 17 and one. The two inside a part per million are a single quadratic and eight times it. The distinct counts are the ones this paragraph rests on. KS-30.4’s 2,971 five-leaf forms at the proton’s accuracy were the standard; this is a harsher one, because a quadratic is simpler than a five-leaf form and the coefficients are the corpus’s own numbers. Those counts are a denominator and not a test of significance, and two further facts belong beside them. None of the enumeration’s forms is shown, here or in the register, because nobody proposed them — they were generated only to measure a field, and displaying one would be supplying a candidate that never existed. Forms the author himself tried and abandoned are a different case and are disclosed in full at Road Ten, which is what a register is for. And none of them — not even the closest, which sits six tenths of a part per million away — falls inside the proton row’s own bar of about two parts in ten million. The abundance does not threaten the corpus’s standard. It measures the field the standard has to be judged against. A form that lands therefore carries no evidential force on its own. It acquires force only if the order of construction and comparison was fixed beforehand, because that ordering is the one thing that stops the value steering the form. Pre-registration does not make a theory true. It makes a landing mean something.

One entry stands in the register already, and it is not a form. On 13 September 2026 a construction for Road Fourteen was built, frozen and hashed before any comparison, its digest recorded on receipt by a desk that had not built it, and compared there. What was lodged and when is what the instruments page records; no claim is made here beyond it. It proposed no value and produced none. Its outcome is recorded as a miss, with the comparing desk's finding that the construction was steered in one identified step, and both the construction and that finding are published. That is what the register is for, and the first thing in it is a failure of the corpus's own making.

A value found by any route is admitted only after it survives the protocol. The protocol's separations are the only protection there is.