Galactic DX · Behavioral SETI

Looking for the behavior of a source, not for a transmitter

Every search for a technosignature so far has begun by naming a carrier — the 21 cm line, an optical pulse, a waste-heat excess — and then searching it as deeply as instruments allow. All of them presuppose the channel. This asks what is left over if you assume no beam, no beacon hardware, and no carrier named in advance.

The idea

A sender who shares no units with the receiver can only use dimensionless quantities — ratios and normalized combinations of things both parties measure. A sender minimizing transmitted energy will not broadcast; it will pose a gate that is cheap to set, expensive to solve, and trivial to verify, so that no energy is spent on receivers who could not act on the message anyway.

Put those two constraints together and the carrier is not a channel at all. It is a relationship among observables — and unlike a channel, that space can be enumerated, searched, and have its coverage measured.

The signal, on this account, is structure imposed on light the star already emits, by redistributing that flux rather than generating any. There is nothing to build, nothing to see as an occulter, no byproduct and no object to encounter. The receiver's problem stops being sensitivity and becomes not knowing which series to examine.

What was searched, and what was found

29
searches of public archives
25
null
3
void — the detector or the null failed its own validation
1
detection — a known signal, as a positive control
13
limits verified by injection

Every archive used was collected for space weather or helioseismology, none of it for SETI. The deepest verified limit is 1.4×10−6 in fractional Lyman-α irradiance at two minutes. Two channels were searched here for the first time.

Sensitivity turned out not to be the binding constraint. The same instrument and pipeline are 245× more sensitive at two minutes than at one day; what limits the search is how much of the combination space has been looked at, and that is a compute problem rather than a telescope problem.

Read it

Five pages
Summary →
The argument, the twelve-step gate chain, the 29 searches, and where it leaves the Fermi debate.
20 slides
Slide show →
The same story in plain terms, without the statistics.
Full paper · interactive
The paper →
Complete, with nine live figures.
PDF · 75 pages
Download →
Typeset for print.

What it does to the Fermi paradox

The paradox usually rests on an assumption that is rarely argued for: that contact would be self-announcing. A signal gated on the receiver's own capability is not. A civilization holding the right archives, at the right cadence, that has not yet thought to search relationships sees exactly what a civilization with nothing to find sees.

That is a smaller claim than it sounds, and it is worth being precise about its limits. It shares a prediction with the zoo hypothesis — silence despite presence — and shares no mechanism with it: that hypothesis needs many civilizations to abstain indefinitely and breaks on a single defector, while a difficulty-bounded gate needs nobody to abstain at all. It informs the debate. It settles neither question, and none of the twenty-five nulls is evidence for the framework that motivated looking.

What would close it

The search measured something about the archive rather than about the sky: coverage is the binding constraint, and it responds to the number of jointly measured observables and to nothing else. Not sensitivity — the same pipeline is 245× more sensitive at two minutes than at one day. Not compute — the entire programme is under a millisecond of Frontier.

What limits it is that the long solar records were made by instruments that never observed together. A relationship between quantities is only measurable if the quantities are simultaneous and share a calibration chain, and most of ours are neither. Three things would change that, and the order matters because the first is a precondition for interpreting the second.

WhatScaleWhen
1 A null distribution for higher-order solar statistics — nobody knows the false-alarm rate of a bispectrum on a Sun-like system, because it needs an ensemble and we have one Sun. A dynamo simulation campaign supplies it. a compute allocation1–2 years
2 One Sun-as-a-star instrument, four channels, one clock — spectropolarimetry, line-profile ratios, p-modes and core g-modes measured together rather than by four separate instruments across handovers. one instrumentmid-term
3 Two in-situ measurements no telescope can supply — the interplanetary electric field, never measured continuously; and high-latitude solar wind, which existed only while Ulysses flew and has had no successor. two missionsdecadal

None of it is proposed for this search’s benefit. Core g-modes have been a helioseismology objective for fifty years. Disc-integrated polarimetry calibrates every inference drawn from unresolved stellar observation. A solar polar mission has been asked for repeatedly. The argument here is that they are also, jointly, the constraint on a separable question — and that costs nothing extra.

White paper
The co-registration gap →
The full case, with measured precisions and shortfalls, addressed to people who build instruments.
Source
Markdown →
The same document, plain text.

Code and data

The analysis code is public. It carries the search scripts, the surrogate library, the validation suite and the audit trail — including the negative controls, the injection-recovery curves behind every stated limit, and the record of corrections made along the way.

GitHub · MIT
behavioral-seti →
Analysis code, surrogates, validation suite and audit trail. Results under CC BY 4.0.