The Corpus That Doesn't Exist: A Notice to Both Sides of UAP Research

There is a question about UAP that cannot be answered by any single case, no matter how compelling: what is the residual fraction of properly-investigated reports that resists prosaic explanation, and what does that residual look like at scale?

This is not a metaphysical question. It is a data-quality question. It has a numeric answer. The answer requires a corpus — a body of cases, curated to consistent standards, cross-referenced against the relevant background data, with the methodology made auditable so that any disagreement about a verdict is a disagreement about evidence rather than about which tribe the investigator belongs to.

That corpus does not exist.

It has not existed for over sixty years. The technical and methodological tools that would build it have been routine in adjacent fields for decades — sensor-data fusion, statistical anomaly detection, witness-reliability protocols, geometric back-projection from cameras to angular coordinates, automated cross-referencing against satellite ephemerides and aircraft tracks, integrity hashing of evidentiary records. None of these are exotic. All of them are applied, every day, to questions of comparable empirical weight in fields where the question is considered respectable to ask.

They have not been applied at scale to UAP. This is a fact in the record. It is not a controversial fact. It is the most consequential fact in the field, and almost no one in the field treats it as a fact at all.

This post is a notice — to two communities at once, because the failure runs both directions.

What kind of question UAP is

The dominant framings in public UAP discourse are mystery and disclosure. Mystery says: there is something out there, and we are trying to identify it. Disclosure says: the government knows, and we are trying to find out what. Both framings position the question as a thing to be solved — by finding the smoking gun, the alien body, the leaked memo, the verifiable artifact.

The actual question is duller and more important. It is: given a population of reports of unidentified aerial phenomena, what fraction of them survive rigorous investigation, and what are the properties of the survivors as a population?

Individual cases cannot answer this. Any individual case, no matter how compelling, is always explicable by some sufficiently elaborate prosaic interpretation, and the prosaic interpretations cannot be ruled out without statistical context. The Tic Tac was a balloon. The Gimbal was a sensor artifact. The Nimitz radar return was a propagation anomaly. For each individual case, a prosaic explanation is constructable. The honest question — do the prosaic explanations, in aggregate, exhaust the population — cannot be answered case by case. It can only be answered against the population.

The population requires curation. The curation requires methodology. The methodology has to be auditable. The audit has to be possible across both prosaic and anomalous interpretations, which means the corpus has to be built to standards that both communities can criticize without finding it pre-cooked.

This is the corpus that does not exist.

What it would take to build one

The protocols are not new. They are not specific to UAP. They are the protocols any rigorous empirical field uses to handle messy observational data with mixed interpretive overlays:

Curation to consistent standards. Each case enters the corpus with the same minimum metadata: time, location, witness count, instrumentation present, sensor types, environmental conditions, observer geometry. Cases that cannot meet the minimum are not investigated; they are flagged as below-threshold and held in a separate pool. This is what every observational science does. It is not done in UAP.

Methodology consistency. A case is investigated against a fixed protocol that runs every prosaic check before any anomalous interpretation is entertained. Aircraft tracks are pulled from ADS-B. Satellite passes are computed from public ephemerides. Weather is pulled from gridded reanalysis. Astronomical objects are computed from observer location and time. Sensor artifacts are evaluated against the known failure modes of the specific instrument. Witness reports are evaluated against known reliability factors. The protocol runs the same way for every case, regardless of how interesting the case looks at intake. This is not done in UAP either. Most rigorous-sounding investigations apply some subset of these checks, ad hoc, with no commitment to running the full protocol on every case.

Auditability. The corpus's reasoning is preserved. Every verdict points back at the evidence and the rule that produced it. A reader who disagrees with a verdict can see exactly which step of the protocol they would re-run differently, and can see what the verdict would become under their alternative. The corpus is reproducible. Hash-locked. Citable not as a single source of truth but as a specific moment of methodology applied to specific cases. This is also not done.

Burden of proof in the right place. The default classification is that a case is unexplained because it has not been adequately investigated, not because it is anomalous. Adequate investigation can move a case to identified, possible-match, or genuinely unexplained-after-rigorous-checking — and the third category is the one that, in aggregate, answers the actual question. The third category has to be earned, not defaulted to. The skeptical community would call this not jumping to conclusions. The advocacy community would call this epistemic humility. They would be agreeing about the same protocol, and they would not realize they were agreeing.

These four properties are the corpus that doesn't exist. None of them is exotic. None of them requires technology that wasn't available in 1990, let alone 2026. The reason they have not been combined into a working corpus is not technical. It is social.

Why the corpus doesn't exist: the structural reading

The charitable reading — and the one we believe the field should examine first — is that the corpus doesn't exist because the people who could have built it were filtered out by structural pressure.

For most of the period from 1953 to roughly 2017, taking UAP seriously was a career-ending move in mainstream science. Not formally — there was no rule. Just the soft pressure of professional embarrassment, sustained over generations, doing what it always does: diverting talent away from a question. A young astronomer who wanted to apply observational rigor to UAP learned, often without anyone telling them, that doing so would mark them as unserious, that their grants would dry up, that their colleagues would gossip, that their tenure case would become more complicated. So they did something else. Multiplied across thousands of would-be researchers, this is how a field gets built — or doesn't.

Meanwhile, on the advocacy side, methodological rigor was also filtered out. Not by professional embarrassment but by the opposite force: a community that wanted evidence of something specific and treated investigators who failed to find that thing as obstacles rather than collaborators. A serious investigator who concluded that a famous case had a prosaic explanation faced not gossip but accusations of betrayal. The structural pressure ran the other direction, but it was just as effective at filtering rigor out.

Both communities filtered for allegiance and against methodology. The corpus that would require methodology from both sides could not get built because it did not have, in either community, a sufficient population of researchers willing to pay the cost of being methodologically committed in a field organized around tribal alignment.

This is the structural reading. It does not require malice. It requires only the ordinary operation of professional incentives, which are powerful enough on their own to explain a sixty-year absence.

Why the structural reading may not be sufficient

The post you are reading is a notice. The notice is to both communities, and the substance of the notice is this: the structural reading is on probation.

The structural account is the charitable account. It is the account that does not require anyone to have done anything wrong on purpose. It explains the absence by appeal to incentives that operate everywhere in science. We believe the structural account is largely correct — for the period in which it was the only available account.

But the structural account makes a prediction. If structural pressure is the cause, then removing the structural pressure should produce the corpus. The pressure is now substantially lower than it was a decade ago. Government acknowledgment of the phenomenon, congressional interest, peer-reviewed publications in respectable journals, named tenured researchers willing to engage publicly — all of these have reduced the cost of being methodologically committed in either direction. The corpus should now be possible to build. It is being built, in pieces, by a small number of researchers and projects.

The question raised by the next decade is whether the rest of the science community — the people who, under the structural account, were prevented from engaging by costs that have now substantially fallen — will engage now. If they do, the structural account is vindicated. If they don't, the structural account becomes harder to defend. The continuation of the absence, under conditions where the structural pressure has fallen, becomes a different kind of evidence than the original absence.

We are not claiming the alternative readings — that the absence has been actively maintained by some interest, that the suppression has been intentional rather than emergent — are correct. We are claiming they cannot be ruled out structurally if the structural account stops fitting the data going forward. The behavior of the science community in the next several years will tell us which account survives.

This is the load-bearing claim of the notice: whichever account is correct, the variable is what happens next.

The people doing the work

Despite both kinds of pressure, a small number of researchers have applied methodological rigor to UAP-adjacent questions, and have paid the costs we've described. Their work is part of the evidence that the corpus is possible. Their treatment by their respective communities is part of the evidence that the structural pressure is still active.

Avi Loeb at Harvard, through the Galileo Project, has built a multi-sensor observation network and committed to publishing all data and results in peer-reviewed venues. The project's first commissioning paper, covering analysis of half a million tracked objects, was posted in late 2024. Loeb is also the easiest of the named researchers to criticize — and not all the criticism is unfair. Some of it is directed at specific methodological choices in his analyses of 'Oumuamua and the IM1 spherules, and that kind of criticism is the system working. But a meaningful fraction of the criticism Loeb receives is not about his methods. It is about his willingness to take the question seriously at all, framed as if engagement were itself the error. The line "he has nothing at risk, he's a tenured Harvard professor" — actually published in a profile of him — is a tell. The implicit claim is that engagement on UAP is a thing only people with nothing at risk can afford. That is the structural pressure speaking, and it is speaking even now.

Garry Nolan at Stanford, an immunologist who publishes prolifically in his actual field, has applied biomedical analysis to materials and to the neurological injuries of intelligence and diplomatic personnel exposed to anomalous events. He has reported, on the record, that he received "more grief for where he appeared than what he talked about" after a Fox interview, and that he was "more than willing to take the pelting." A scientist of his standing should not have to take a pelting for engaging with a question. That he does is information.

Beatriz Villarroel at the Nordic Institute for Theoretical Physics leads the VASCO project, which compares pre-Sputnik astronomical plates to modern surveys. In October 2025, two new VASCO papers passed peer review and were published in Scientific Reports and the Publications of the Astronomical Society of the Pacific — three independent peer review processes including the team's earlier MNRAS paper. The papers report statistically significant correlations between transients on 1950s plates and contemporaneous UAP reports, and a 22-sigma deficit of transients inside Earth's shadow. Villarroel's methodology is unimpeachable and her results are heterodox — exactly the kind of work that, in a healthy field, would generate vigorous methodological scrutiny rather than dismissal-by-association. The field's response to her work, going forward, is one of the variables we are watching.

Kevin Knuth at SUNY Albany, applying Bayesian analysis to UAP kinematics. Jacques Vallée, methodologically rigorous since long before the current cycle. The Sol Foundation, the Scientific Coalition for UAP Studies (SCU), and the academic working groups gathering under the Sol symposium and adjacent venues. These are not a unified school. They disagree about interpretations, methods, and emphases. But they share a commitment to running the protocol regardless of where the protocol leads. That commitment is the thing the corpus needs.

And on the other side:

Mick West, who has done more methodologically careful debunking of high-profile UAP videos than any other public investigator. His Sitrec tool — open-sourced in 2024 — does the kind of geometric reconstruction and sensor-artifact modeling that any rigorous investigation requires, regardless of which conclusion the investigation reaches. West has been called "the villain of the ufologists" in print. He has been the target of sustained personal attacks from believer-side communities for the offense of doing rigorous work that produced prosaic explanations. The attacks are wrong. The methodology is the methodology. A community that attacks a careful skeptic for finding what the evidence supports is a community that has substituted allegiance for inquiry, and is doing exactly the same thing it (rightly) accuses establishment science of doing in the opposite direction.

The position we are taking, as carefully as we can: the people on both lists are doing the work the field requires. They are paying costs from their respective communities for doing it. The costs are wrong in both directions. And the people on both lists, despite arriving at different conclusions about specific cases, share more methodologically with each other than with the loudest voices in their respective tribes.

The notice

To the science community — particularly astronomy, atmospheric science, sensor engineering, statistics, and the relevant adjacent fields:

The structural account predicts that as the cost of engagement falls, engagement will increase. The cost has fallen. We are watching. There are now corpora to engage with — the NUFORC dataset (now expanded with disclosed gaps), the Galileo Project's published data, VASCO's archival catalog. Engaging means applying your methodology, pointing at specific weaknesses, proposing specific improvements, building competing corpora to higher standards if ours don't meet them. Continued absence — now, when the structural pressure has fallen — is a different kind of data point than absence under conditions of pressure. Going forward, the absence becomes the message.

To the advocacy community — particularly the segment that treats methodological scrutiny as betrayal:

Mick West is right more often than you want him to be. So is every careful skeptic who finds prosaic explanations for cases you wanted to be anomalous. That is the methodology working. The corpus the field needs is not the corpus of cases that survive your filtering. It is the corpus of cases that survive both filterings — the skeptical and the anomalous — applied honestly. If your community attacks rigor when rigor produces conclusions you don't want, you are doing the same thing the establishment did to your researchers for sixty years. You have inherited the failure mode you came into existence to oppose.

To the methodologically committed minority on both sides:

You exist as a coalition. You have not yet noticed yourselves as one. You share more with each other than with the loudest voices in your respective communities, and what you share is exactly the thing the field requires. The corpus that answers the actual question is buildable. It requires standards both sides can audit, treatment of skeptical scrutiny as a feature rather than a betrayal, treatment of anomalous cases as data rather than as ammunition, and a refusal to organize around tribal alignment in either direction.

We will be in this for the long run. Other rigorous projects exist alongside us and we are glad of it. Go build the thing the field needs, in whatever venue you can build it. The question of whether UAP has a real residual after rigorous investigation is answerable. The variable is whether the people who could answer it will choose to.

That variable is now what we are all collectively writing into the record. Every paper published, every protocol applied, every shaming refused, every case investigated honestly across the tribal line is a vote. The vote that's being tallied is the answer to a question more important than what UAP turn out to be: what kind of field is this.

The corpus is buildable. The field is choosable. The notice is filed.

— UFO Index Team