We've Been Listening for Alien Signals. I Think We Should Also Be Listening for Alien Silence.

Composite of Chandra X-ray and Webb infrared data: star cluster NGC 602 in the Small Magellanic Cloud, showing a wreath-like ring of gas and dust.

This is a companion article to the paper ‘Cessation Signature Search: A Statistical Framework for Detecting Technosignature Loss in Multi-Epoch Survey Archives’ (DOI 10.5281/zenodo.19454687)

SETI has spent more than sixty years asking whether anyone is transmitting. That question has shaped every instrument, every pipeline, every target list in the history of the field. I want to ask the one it has not asked: has anyone stopped?

The paper I published proposes a methodology for doing exactly that. I call it the Cessation Signature Search — CeSS, or Negative SETI (N-SETI). The premise is simple: if technological civilisations exist, and if their emissions contribute detectably to the sky’s aggregate electromagnetic background, then when one of them goes dark, that background should change. The change would not be dramatic, and probably not obvious. But it could be measurable, if we built a method specifically designed to look for it. Nobody has yet systematized that search as a technosignature framework for cessation. I am arguing that we should — and that the data needed to begin may already be sitting in existing archives.

The two approaches sound symmetrical. They are not. A cosmos in which civilisations regularly destroy themselves looks, to an emergence-only search, exactly like a cosmos in which they never existed at all. The cessation record would tell them apart. Once you see that asymmetry, it is difficult to set aside.

We have been asking whether anyone is out there. We have never asked whether anyone used to be.

The Background Is Not Empty Noise

When a radio telescope points at a patch of sky, it receives everything — a superposition of signals from billions of sources, almost all too faint to identify individually. Standard SETI treats this composite as a nuisance to subtract, a noise floor to see past. But if technological civilisations are distributed across a galaxy, they are not contributing isolated bright dots to the sky. They are contributing to that composite. Their radio leakage, their waste heat, the electromagnetic fingerprint of their existence — all of it folded into the background. If one of them goes dark, this background changes.

No programme has been designed to look for it. CeSS, or N-SETI, proposes to fill that gap. The detection range runs from the subtle — a statistical shift in the character of a sky position’s background with no single-source explanation — to the dramatic: a known, catalogued emitter that was there and then was not. Both ends matter. Neither is currently the target of any search.

The Shape of an Ending

The shape of an ending carries information about its cause. That is the observation the entire framework rests on.

A violent end — war, large-scale weapons use, catastrophic technological accident — would produce a brief anomalous multi-band emission burst immediately before permanent silence. Broadband electromagnetic pulse signatures, hard X-ray and gamma-ray emission from the physics of the event, extended ionospheric effects. The signal might appear, briefly, as an anomalously bright source in the historical record before going permanently dark. That specific sequence — a brief high-energy burst followed by enduring silence at the same sky position — is not how natural astrophysical sources behave. Natural transients appear in characteristic high-energy bands and do not precede the permanent extinction of a low-frequency background at the same sky position.

A civilisation building a Dyson sphere — progressively enclosing its star to capture energy — produces the opposite in terms of pace: a slow, observable transition. Optical flux from the star declining as the structure grows; infrared waste-heat emission rising correspondingly. This is the one cessation category with a visible warning sign, a precursor that can be watched unfold over decades. We already maintain catalogues of infrared-anomalous stars for which no natural explanation has been found, compiled specifically as Dyson sphere candidates. If any of those objects have since disappeared from the record, that disappearance is a cessation event. The data may already contain it.

Voluntary withdrawal — a civilisation that chooses to go quiet — produces a different profile again: potentially abrupt, with no correlated astrophysical destruction, and possibly followed by re-emergence on long baselines. How something ends is not incidental. It is the most informative thing about it.

Every Theory About Why We’re Alone Predicts a Different Kind of Silence

The Fermi paradox has generated an enormous family of proposed explanations for the cosmic silence. Most discussions treat these as equally consistent with the present-epoch absence of signals — and they are, because an emergence-only search cannot distinguish between them. CeSS could, in principle, because each hypothesis predicts a structurally different pattern in the cessation record.

Robin Hanson’s Great Filter proposes that some step on the path from dead matter to spacefaring civilisation is nearly impossible to navigate. If that step lies ahead of us — if most civilisations destroy themselves at roughly our current level of development — the cessation record would show signals that rose, persisted for a characteristic duration, and went dark at roughly the same civilisational age. The timing distribution across cosmic history would encode where the filter operates. Cessation events clustering within a narrow technological window says something different from cessation events distributed uniformly across cosmic time.

Self-termination through war, ecological collapse, or technological accident predicts abrupt, spatially uncorrelated dropouts — randomly distributed across the sky, without clustering, without propagating structure, without re-emergence. The violent subset would carry the precursor emission bursts described above. Predatory suppression — the Dark Forest hypothesis — also produces abrupt dropouts, which is why at first glance it is hard to distinguish from self-termination. The discriminators are spatial. A diffuse predatory regime produces random dropouts with no re-emergence anywhere in the record; the persistent total absence of re-emergence, over a large enough sample, breaks the self-termination model. A territorial predator leaves geographic structure: regional clustering of cessation events bounded to a domain. Alexander Berezin’s first-mover variant predicts the most geometrically specific signature of all — a directed wave of cessations advancing outward from a single origin at sub-relativistic velocity, with nothing behind the front. That pattern is inconsistent with self-termination, inconsistent with natural causes, and not produced by any other hypothesis in the taxonomy.

The Zoo hypothesis — John Ball’s 1973 proposal that advanced civilisations have deliberately set Earth aside as a wilderness preserve, observing without interfering — is one of the few Fermi resolutions that leaves almost no mark in the cessation record. The watchers remain active. What the Zoo hypothesis predicts is directed silence toward us specifically, not electromagnetic absence from the sky. CeSS is largely blind to it. Worth stating honestly.

The aestivation hypothesis proposed that advanced civilisations choose dormancy and wait for the universe to cool, on the argument that computation is thermodynamically cheaper at lower temperatures. That specific argument was formally challenged — Bennett, Hanson, and Riedel published a rebuttal in Foundations of Physics in 2019, showing the thermodynamic reasoning was flawed. But the broader class of idea — deliberate reduction of electromagnetic footprint, for whatever reason — has a clear CeSS signature: gradual wind-down rather than abrupt dropout, and a concentration of cessation events in recent cosmological epochs if dormancy is a strategy civilisations converge on at a particular stage.

The transcension hypothesis argues that advanced civilisations universally turn inward rather than outward — toward miniaturisation and computational density — until they become electromagnetically invisible from outside. Machine succession follows a parallel logic: the transition from biological to post-biological substrate changes the emission profile before it ends it, a spectral shift before the silence. Both predict structured transformation in the signal record before the final dropout. That is a different observational signature from an abrupt dropout with no precursor, and the distinction matters for any attempt at causal attribution.

Rare Earth — the argument that complex life requires a highly improbable convergence of planetary and galactic conditions — predicts a cessation record that is simply sparse, because few beginnings means few endings. And not every cessation is permanent: a civilisation that collapses and recovers would produce a dropout followed by re-emergence at the same sky position; one that relocates would produce cessation at the origin and emergence at the destination after a travel-time delay; von Neumann probe dispersal would produce the only signature in the taxonomy that links a single cessation to multiple correlated emergences in a branching spatial pattern. These are detectable in principle — but only if cessation and emergence detection programmes run in parallel and compare notes.

Before the Silence, There Is Usually a Change

The framework is broader than the word cessation implies. It includes transformation events — structured changes in the character of a civilisation’s electromagnetic output that are inconsistent with natural processes, even if the output does not end.

Earth is the only example we have. Over roughly a century, our aggregate signature has changed substantially: broadcast radio introduced strong modulated carriers in the medium-wave band; television and radar expanded the profile upward in frequency; the shift to digital broadcasting replaced analogue carrier signatures with noise-like spread-spectrum profiles harder to distinguish from natural background; satellite communications redirected large amounts of transmitter power away from the planetary surface into directed orbital geometries. We have been getting progressively less detectable as our technology matures — not because we are declining, but because we are communicating more efficiently. A civilisation that survives long enough may become quieter for exactly the same reason.

If this trajectory is at all representative, the cessation record will contain transformation events — positions whose composite character changes in ways inconsistent with known astrophysics — alongside clean dropouts. Distinguishing maturation from voluntary withdrawal from terminal decline is genuinely hard. But they are distinguishable from each other in principle, and learning to read the difference is what this programme is actually for.

It Has Happened Before, in a Different Register

The Chinese treasure fleet under the Yongle Emperor was, in the early fifteenth century, the most extensive maritime contact programme in premodern history — enormous ships, vast distances, deliberate projection of presence across the known world. Under the Hongxi Emperor, it stopped. The voyages were terminated, maritime activity restricted, the records by some accounts deliberately destroyed. A technologically capable civilisation withdrew from contact entirely, for reasons that were internal and political. From the perspective of a distant port that had received those ships and then stopped receiving them, this was indistinguishable in the short term from collapse. It was not collapse.

Communities facing epidemic disease produce the same pattern more abruptly: overnight severance, roads cut, movement halted, external contact refused as deliberate survival strategy. Not death — choice. In both cases a civilisation that was present became unreachable, for reasons that made complete sense internally, without dying. That is not an edge case in the history of intelligent behaviour. It is a recurring pattern — and the cessation framework has to take it seriously, not as a footnote to extinction, but as a category in its own right.

The Data Already Exists. The Question Has Not Been Asked.

Decades of radio survey archives exist — NVSS, VLASS, TGSS, MeerKAT — along with years of infrared maps from WISE and NEOWISE. These datasets were processed with pipelines built to find new point sources above a detection threshold. None were processed to ask whether the statistical character of the composite background had remained stable across epochs. That question has not been asked of data that already exists.

The methodological model comes from pulsar timing arrays. NANOGrav’s 2023 report of evidence for a nanohertz gravitational wave background involved no individual detection: the signal existed only in cross-correlated statistical residuals across tens of millisecond pulsars over fifteen years of continuous monitoring. Not a signal in any single observation — a pattern only visible across time. CeSS asks for the same kind of longitudinal analysis from archives that are already there.

The hardest problem isn’t collecting data — it’s knowing how reliable that data is over time. Instruments drift, the atmosphere interferes, and those effects can create patterns that look real. Understanding that background requires years of consistent observation. If you didn’t monitor the sky during a certain period, you can’t go back and reconstruct it later — no matter how good your future instruments are.

What a Positive Result Would Actually Mean

Every candidate has to clear multiple evidentiary hurdles: cross-matching against known astrophysical variables, calibration checks, multi-band confirmation, cross-reference against catalogued high-energy events — gamma-ray bursts, magnetar flares, nearby supernovae — that could explain an extinction without invoking technology. The pipeline is conservative by design.

If something clears all of those hurdles, you have a change in the statistical character of a sky position that cannot be explained by any known natural process, any instrument artefact, or any catalogued event. You have not confirmed a civilisation. You have confirmed an anomaly that demands investigation.

The shape of that anomaly carries information. Spatially random abrupt dropouts distributed across cosmic history are consistent with civilisations that destroy themselves. A region of clustered cessations advancing from a single point is consistent with something considerably less comfortable. A cessation followed by a correlated emergence at a displaced position, after a delay consistent with travel time, is evidence that something moved rather than ended. Science does not get to choose only the comfortable results. The framework has to be ready to read whatever the record shows.

The Question We Have Not Yet Asked

Imagine watching a neighborhood at night. Traditional SETI is mainly concerned with lights switching on: a signal appearing where before there was none. CeSS is interested in the neighborhood as a changing system. It watches not only for lights appearing, but for lights going out, for changes in how long they stay on, and for patterns that only become visible over time. In some cases, it may not even resolve a single house directly; it may detect a subtle change in the neighborhood’s combined glow and infer that something in the system has changed. Then, we analyze if the change matches known, natural phenomena or if it’s something more interesting.

Both are valid searches. Only one has been systematically conducted. The absence of results from the first tells us very little about what we would find in the second, because a cosmos in which civilisations regularly destroy themselves, are eliminated, or go quiet by choice, looks identical to an emergence-only programme as a cosmos in which they never existed.

The universe is very old. A great deal has happened in it. Some of those events may have left traces—not as signals arriving now, but as subtle deviations from the background patterns we expect to observe today.

The universe may be full of endings. We have not, until now, been looking for them.

Media

NASA Missions Spot Cosmic ‘Wreath’ Displaying Stellar Circle of Life — NASA/Chandra and Webb composite of the star cluster NGC 602 in the Small Magellanic Cloud.

Credit: X-ray: NASA/CXC; Infrared: ESA/Webb, NASA & CSA, P. Zeilder, E.Sabbi, A. Nota, M. Zamani; Image Processing: NASA/CXC/SAO/L. Frattare and K. Arcand.