Interstellar Civilization Models

An interstellar civilization model is an attempt to say how many technological civilisations there are, where they are, and how they are distributed in space and time. The family includes the Drake equation and its descendants, settlement-wave and diffusion models, percolation models, and the counting papers that convert assumptions into a headline number. The framing under test is that the distribution of civilizations can be modelled.

It can be modelled. The question this brief settles is what the model is a model of. The Drake equation is an identity: it multiplies seven terms, and four of them — the fraction of habitable worlds where life starts, the fraction where intelligence follows, the fraction that becomes detectable, and the lifetime of the detectable phase — have exactly one known instance each, drawn by an observer whose own existence is conditioned on the outcome. Nothing in the identity is wrong. The identity simply does not constrain its own output.

The single most important result in this subject is fifteen years newer than most of what circulates about it, and it is arithmetic rather than astronomy. Sandberg, Drexler and Ord took the published scientific estimates for each Drake term, propagated them as distributions rather than as point values, and reported what comes out. What comes out is not a number of civilisations. It is a probability distribution so wide that a substantial probability mass sits at zero — before any appeal to a Great Filter, a zoo, a dark forest or a dormancy hypothesis.

This brief therefore organises around a distinction the popular literature collapses. There is a small established core — exoplanet occurrence rates, the measured coverage of the technosignature search — and everything else is inference from a sample of one. Civilization Timelines has already written the seam in its own text: this brief “owns the Fermi question, the Drake equation, colonisation-wave models and technosignature search status as astrobiology”, while that one “takes the Kardashev scale, the Great Filter and the dark forest only as periodisation claims.” Interstellar Archaeology owns what has actually been searched; Exoplanet Colonization owns what is known about destinations. This brief owns the population and distribution modelling that sits on top of both.

Frontier The keystone result, and it is quantitative. Sandberg, Drexler and Ord, Dissolving the Fermi Paradox (arXiv:1806.02404, 2018), take published estimates for each Drake parameter, represent each as a distribution over the range the literature actually supports, and propagate. Their headline outputs, from the paper body: a 30% probability that there is less than one civilisation in the Milky Way (confidence interval 27–52%), and a 10% probability that we are alone in the observable universe (CI 0.2–20%) — before any update on the observed silence. Median N with no update is 0.32.

Frontier Then they condition on the silence, and the numbers move hard. After a Bayesian update on the fact that we have detected nothing, the probability that the Milky Way contains fewer than one civilisation runs 53% to 99.6% depending on which evidence model is used, and the probability that we are alone in the observable universe runs 39% to 85%. Under settlement-scenario updates the median N falls as low as 8.1 × 10⁻³⁵. The authors' conclusion is that the silence requires no exotic explanation at all: “there should be little surprise when we fail to detect any signs of it,” eliminating “any need to invoke speculative mechanisms.”

Established The mechanism behind those numbers is the part that is not contested, because it is a claim about arithmetic rather than about aliens. The log-uncertainty on the abiogenesis-rate term alone is at least 200 orders of magnitude. Multiplying point estimates instead of distributions, in the authors' own toy model, produces errors of more than 42 orders of magnitude. Their diagnosis of the underlying error is exact and worth quoting: Drake-style models “implicitly assume certainty regarding highly uncertain parameters.” That is why the correct description of the Drake equation is a framework for organising ignorance, not a calculation.

Established A note on where these figures may and may not appear on this site, because it is a real editorial decision rather than a formality. Civilization Timelines cites this same paper and deliberately prints no probability from it, because that brief obtained only the abstract and the abstract carries none. This brief obtained the body. Printing 30% / 10% / 53–99.6% / 39–85% here is the handoff working as designed.

Frontier The settlement-wave literature, and what it actually predicts — which is the opposite of what it is usually cited for. Newman and Sagan, Galactic civilizations: population dynamics and interstellar diffusion (Icarus 46:293, 1981), solve the expansion problem with potential theory and reach two conclusions: population growth is “limited by the carrying capacity of the planetary environments,” and Earth is uncolonised not because spacefarers are rare but because “there are too many worlds to be colonized in the plausible lifetime of the colonization phase.” They conclude that galactic empires of extreme age and conquest-oriented policy do not exist. This is a deflationary result, and the popular reading turns it into a paradox-sharpening one.

Frontier The modern replacement for Hart's argument is Carroll-Nellenback, Frank, Wright and Scharf (AJ 158:117, 2019), and it contains the most robust single number in the modelling literature. The settlement front has a lower limit on its speed of about 100 km/s set by stellar motions alone — that is, a civilisation with no propulsion advantage whatever over the drift of stars through the galaxy still spreads, because the targets come to it. 36 doubling times suffice to settle 10¹¹ systems, of order 270 million years on typical parameters, so the galaxy can be filled in under a billion years. The velocity floor is the interesting part, because it is independent of everything Interstellar Probes is trying to solve.

Frontier The same paper produces the two results that break the inference from silence. The steady-state settled fraction is X_eq = 1 − T_l/T_s, the ratio of probe launch period to civilisation lifetime, and there exists “a range of parameters for which 0 < X < 1” — permanently partial settlement, a galaxy that never fills. And with a one-million-year horizon on geological evidence, “it is possible to find situations in which Earth may not have experienced a settlement event for longer than some horizon time.” The authors state their own conclusion plainly: “Earth might remain unvisited in the midst of an inhabited galaxy.” The observation (no visitors) is decoupled from the conclusion (therefore nobody exists).

Frontier Their third contribution is a term this brief does not own the value of, and it matters more than its obscurity suggests. The Aurora effect: “even though a world was formally habitable it was not what we would call settleable.” The settleable fraction is smaller than the habitable fraction by an amount nobody has measured. The model needs a number; Exoplanet Colonization is where the number would come from, and that brief's finding is that it is unmeasured for every candidate world in the catalogue.

Speculative Percolation, and why “the galaxy fills up” is a modelling choice rather than a result. Geoffrey Landis, The Fermi Paradox: An Approach Based on Percolation Theory (NASA Lewis / Sverdrup, NTRS 19940022867; JBIS 51:163), drops the implicit assumption that every colony goes on to colonise: “any given colony will have a probability P that it develops a colonizing civilization, and a probability (1−P) that it will develop a non-colonizing civilization.” The critical threshold for the lattice he uses is P_c = 0.311 for a cubic array with connectivity N = 6, and his baseline parameters are P = 1/3, N = 5 — essentially at criticality. At criticality, “arbitrarily large filled regions exist, and also arbitrarily large empty regions.”

Speculative The payoff is that a void needs no explanation. A galaxy that is simultaneously heavily settled and locally empty is the generic output of a near-critical percolation process. No Filter, no zoo, no interdiction is required to produce a quiet neighbourhood; voids are what percolation makes. The weakness must be stated with the result: the near-critical baseline is chosen, not derived. Move P well above P_c and the galaxy fills; move it below and settlement dies out locally. The model converts one unknown into another and displays the consequences honestly, which is more than the counting papers do.

Frontier The three-dimensional successor tests the two competing deflations directly. Hair and Hedman, Spatial dispersion of interstellar civilizations: a probabilistic site percolation model in three dimensions (International Journal of Astrobiology 12(1):45–52, 2013), run a probabilistic cellular automaton outward from a single origin to 250,000 years of simulated time, examining whether emigration is simply too slow for anyone to be within detection range yet, and whether “large volumes of habitable space may be left uninhabited... and Earth is located in one of these voids.” Both remain live under their model.

Speculative Rare Earth is a position with named holders, and it has had a twenty-year audit. Ward and Brownlee's Rare Earth (2000) multiplies nine factors and argues the product is “no greater than 10⁻¹⁰ and could plausibly be as small as 10⁻¹²”, with a galactic habitable zone confined to a ring 7–9 kiloparsecs in radius holding no more than 10% of Milky Way stars — reduced to 5% by later authors. The book itself was not obtained for this brief; the account here is from a named expert review and a tertiary synthesis, and is labelled as such. Named holders include Ward, Brownlee, Stuart Ross Taylor, Stephen Webb, Simon Conway Morris, Barrow and Tipler, John Gribbin and Marc Defant.

Frontier The audit is the useful part, and it goes badly for several load-bearing premises. Ramses Ramirez of the Earth-Life Science Institute reviewed the hypothesis at twenty years (Centauri Dreams, 26 June 2020). What survived: the Moon's stabilisation of Earth's obliquity, though with smaller variations than originally modelled; plate tectonics; magnetic fields. What did not survive: the claim that only 5–6% of stars have planets — “nearly all stars have planets”; Jupiter as an impact shield, where modelling suggests Jupiter “has caused more impacts on Earth than it has prevented”; and water delivery, where formation simulations “often always accrete too much water, tens of oceans.” Ramirez gives an occurrence rate near 15–45% for M stars. The core claim — that complex life is rare — is untouched in either direction, and that is the honest statement.

Frontier The named methodological objections are separate from the audit and older. Jack Cohen argues the hypothesis is circular, assuming Earth-like conditions are necessary; David Darling that it is “a description of how life arose on Earth,” not a prediction; James Kasting disputes habitable-zone width; Karen Masters points out that spiral-arm crossings every hundred million years or so cut against the galactic-habitable-zone argument. None of these is a counter-measurement. They are objections to the form of the argument, and they should be reported as that.

Speculative The counting papers are this slot's cautionary tale. Westby and Conselice, The Astrobiological Copernican Weak and Strong Limits for Extraterrestrial Intelligent Life (ApJ 896:58, 2020), impose a condition that intelligence emerges on a habitable world within a fixed time and derive, under the Strong condition, 36 (+175 / −32) communicating civilisations in the Galaxy, with the nearest at most 17,000 (+33,600 / −10,000) light-years away, assuming an average communicating lifetime of 100 years — “based on our own at present.” Most such civilisations would orbit M dwarfs.

Frontier The criticism came from two named and independent directions within a week of publication. Caleb Scharf (Columbia) attacks the selection: “If, for example, intelligence came along much earlier or much later it would look at its history and also select its time of existence as 'the most likely.'” And on the core assumption: “Charting the development of civilizations as a simple function of the age of life on a planet seems positively ludicrous.” Ethan Siegel points at the interval: the 68% range runs 4 to 211, so zero remains plausible at roughly 15%, which makes the headline scientifically empty. Scharf builds civilisation models himself — he is a co-author of the settlement paper above — so his interest runs against the result he is attacking, which raises its weight.

Speculative One hypothesis in this space makes a physical claim rather than a psychological one, and deserves separating from the rest. Sandberg, Armstrong and Cirkovic, That is not dead which can eternal lie: the aestivation hypothesis for resolving Fermi's paradox (arXiv:1705.03394, 2017), argue that advanced civilisations lie dormant because computation is vastly cheaper in a colder future universe — the paper cites a 10³⁰ multiplier on achievable computation. Its virtue over the zoo hypothesis is that the authors state “physical law and observational evidence constrain the motivations of aliens compatible with the hypothesis”: it is at least the kind of claim that could be argued against on thermodynamic grounds. No test has been proposed.

Handwave The zoo hypothesis, by contrast, has no number attached to it anywhere. Ball's 1973 paper was not obtained for this brief and no figure from it appears here. It is named as a position held in the literature, with the observation that it is unfalsifiable by construction: any absence of evidence is predicted by it. The Great Filter and the dark forest belong to Civilization Timelines as periodisation claims and are named here in one clause each, not re-argued.

Established Finally, the two genuinely measured inputs, because the whole point of this brief is to separate them from the rest. Bryson and colleagues (arXiv:2010.14812, 2021), from Kepler DR25 with Gaia astrometry, give 0.37 (+0.48 / −0.21) to 0.60 (+0.90 / −0.36) rocky habitable-zone planets per star for G and K dwarfs in the conservative habitable zone, with “95% confidence that, on average, the nearest HZ planet around G and K dwarfs is about 6 pc away” and roughly four within 10 parsecs. And Wright, Kanodia and Lubar (AJ 156:260, 2018) put the fraction of the eight-dimensional technosignature search space examined at 6.0 × 10⁻¹⁸. The first constrains one Drake term. The second constrains nothing about N at all yet. Interstellar Archaeology owns the coverage figure and its caveats.

Frontier The honest summary of the observational contribution is one sentence. Exoplanet science has moved exactly one Drake term from unknown to measured, to within a factor of about ten — and that term was never the binding one. The binding term is the abiogenesis rate, whose log-uncertainty exceeds 200 orders of magnitude, and no instrument now operating or proposed measures it.

Frontier Position one, and the strongest: the paradox dissolves under honest uncertainty propagation, and we may simply be alone. Held by Sandberg, Drexler and Ord. Quantified, published, and uncontested as arithmetic — nobody has argued that multiplying point estimates is defensible. What is contested is whether the published parameter ranges they draw from are the right ones, which is a question about the literature rather than about their method.

Speculative Position two: expansion is fast, so the absence of visitors implies the absence of civilisations. This is Hart's 1975 inference, and it is the premise most of the popular Fermi literature still runs on. Hart's paper was not obtained for this brief and no figure from it is printed here; the inference is reported as Carroll-Nellenback and colleagues state and then dismantle it. Position three: expansion is slow and capacity-limited, and Earth is uncolonised because there are simply too many worlds. Newman and Sagan, 1981; a model result with no observational test.

Frontier Position four: settlement reaches a steady state in which the galaxy is inhabited and Earth is unvisited. Carroll-Nellenback, Frank, Wright and Scharf, with explicit parameter regions identified. It is unfalsified and, on present instruments, unfalsifiable. Position five: the galaxy sits near percolation criticality and voids are generic. Landis; Hair and Hedman. Coherent, and resting on a chosen baseline. These two positions arrive at the same place from different mathematics, and that convergence is the single most underrated fact in the subject.

Speculative Position six: complex life is rare and the filter is astrophysical and geological. Ward, Brownlee and a substantial roster of named holders. Several load-bearing premises have been falsified since 2000 — planet frequency, the Jupiter shield — and the core claim has not been tested at all. Position seven: Rare Earth is circular, or is a description of Earth rather than a prediction. Cohen, Darling, Kasting, Masters. A methodological objection, which is a different kind of thing from a counter-measurement and should not be scored as one.

Speculative Position eight: there are about 36 communicating civilisations and the nearest is within 17,000 light-years. Westby and Conselice. The 68% interval runs 4 to 211, zero sits inside it at roughly 15%, and the load-bearing assumption is defended by nobody but the paper. Position nine: advanced civilisations aestivate until the universe cools. Sandberg, Armstrong and Cirkovic, with a stated 10³⁰ computational gain and no proposed test. Position ten: a zoo, an interdiction, a deliberate silence. Named without a number, because none exists.

Handwave Position eleven: the Great Filter, and the dark forest. Both are named here and handed to Civilization Timelines, which takes them as periodisation claims and records that no peer-reviewed dark-forest formulation was located. This brief's contribution to that question is negative and specific: on the Sandberg–Drexler–Ord result, no filter is needed to explain the silence, so the filter literature is answering a question that may not have been posed.

Established And the position that unites the sceptics: none of the eleven above has been eliminated. That is not a rhetorical flourish. It is what a 200-order-of-magnitude uncertainty on one parameter means in practice: the space of models compatible with every observation ever made is essentially the whole space. The frontier question that would actually move this subject is not “how many are there” but “what observation would distinguish any two of these positions,” and the honest answer at present is a second, independent origin of life — which is why biosignature work matters more to this brief than any technosignature result.

Established The binding bottleneck is a single parameter, and it is not an engineering problem. The abiogenesis-rate term carries a log-uncertainty of at least 200 orders of magnitude, because the only evidence bearing on it is one instance observed by a party whose existence depends on that instance. Every other improvement in this subject — better exoplanet statistics, deeper SETI, more sophisticated settlement models — operates downstream of a term that dominates the product by hundreds of orders of magnitude. Reducing that uncertainty requires a second origin, and nothing else does it.

Frontier The second is that the models' predicted observable is the same in almost every case. A near-critical percolation galaxy, a steady-state partially settled galaxy, a capacity-limited slow-diffusion galaxy and an empty galaxy all predict a quiet sky at Earth. Carroll-Nellenback and colleagues identify parameter regions where the galaxy is fully settled and Earth is still unvisited within the geological evidence horizon. A model family whose members cannot be told apart by the observable they were built to explain is a taxonomy of possibilities, not a distribution.

Established The third is coverage, and it is quantified. At 6.0 × 10⁻¹⁸ of the eight-dimensional search space examined, the null result constrains loud, continuous, beamed-at-us transmitters within tens of parsecs and essentially nothing else. Any model that treats the silence as strong evidence is using an observation that has not been made. Interstellar Archaeology carries the bounds, the caveats and the instrument dependence.

Frontier The fourth is the settleable fraction, which is a hole in the models rather than in the instruments. Every settlement-wave model needs a value for how many formally habitable worlds are actually usable. The Aurora effect names the term. No published value exists, because establishing one requires knowing whether a specific world has an atmosphere, a surface, a magnetic field and a tolerable stellar environment — and Exoplanet Colonization reports that the count of habitable-zone terrestrial exoplanets with a confirmed atmosphere is zero.

Speculative The fifth is a sociological bottleneck and it shows up in the publication record. The counting papers get headlines and the uncertainty papers do not, so the number that enters public circulation is systematically the one with the smallest error bars quoted and the largest assumptions buried. “36 civilizations” travelled worldwide; “the 68% interval is 4 to 211 and zero is inside it” did not. This is a real obstacle to the subject's credibility and it is created inside the subject.

Established The first adjudicated dependency is Interstellar Probes, and the seam is sharper than it looks. That brief owns probe engineering — propulsion, mission architecture, what can actually be sent and how fast. This brief uses probe velocity and range only as model parameters. And the interesting result runs the other way: Carroll-Nellenback and colleagues show that the settlement front has a floor of roughly 100 km/s set by stellar motion alone, which is independent of anything I-16 achieves. A civilisation that never builds a fast probe still spreads, because the galaxy stirs itself. Any statement about how to build the probe belongs to I-16; this consequence belongs here.

Frontier The second is Exoplanet Colonization, and it is a dependency in the strict sense: this brief needs a number that brief would have to produce. The Aurora term — the settleable fraction, distinct from the habitable fraction — is a model parameter here and a measurement there. II-14's finding is that it is unmeasured for every one of the twenty-nine worlds in the conservative habitable-zone sample, and that the sample itself is a cut on radius and insolation and nothing else. The model term is currently a free parameter with no empirical anchor, and stating that plainly is more useful than assigning it a value.

Established The third relationship is not a dependency but a division of labour, and Civilization Timelines wrote its half first. Its section 8 states the seam in its own words. This brief accordingly does not re-run the Great Filter as a stage model, does not re-run Kardashev, and does not treat the dark forest as a periodisation. It names each in a clause and links. The concrete, checkable version of the seam is the probability figures in section 2: X-24 declined to print them for a stated reason and pointed downstream; this brief prints them.

Established The fourth is Interstellar Archaeology, and the division is measurement against inference. Null results, EIRP bounds and haystack coverage are II-20's to report. What they do and do not imply about the population is this brief's to argue. This brief carries exactly one coverage number and links for the rest. The reverse direction matters too: II-20 should not argue about N, and it does not.

Speculative And one dependency that is nobody's on this map: a second, independent origin of life. It is the only observation that would reduce the dominant uncertainty by more than a rounding error, it is not a technology, and no brief here owns it. Biosignature spectroscopy of outer-system ocean worlds and of exoplanet atmospheres are the instruments that bear on it. Recording it as an unowned scientific dependency is more honest than attaching it to a brief that would then be responsible for a result nobody is producing.

Frontier Experiment one, and it is the only one that touches the binding parameter: find a second origin. Not a second civilisation — a second instance of life arising. A confirmed biosignature on an outer-system ocean world, or an independent abiogenesis event in Earth's own record, would convert the abiogenesis-rate term from a distribution spanning 200 orders of magnitude into something with a floor. Nothing else in this subject has that leverage, and the instruments that could do it are being built for other reasons.

Established Experiment two: extend the technosignature search far enough that the null becomes informative, and state in advance what “far enough” means. At present coverage the null carries almost no information about N. The useful move is not more observing time but a published, agreed statement of what coverage level would make a continued null a constraint — a pre-registration of the inference. Interstellar Archaeology owns the search; this brief owns the inference rule, and the inference rule is currently unwritten.

Frontier Experiment three: the settleable fraction, measured on real worlds. The observation that would do it is atmospheric characterisation of terrestrial habitable-zone planets — not biosignatures, just whether there is air. That programme exists and is running, and its first results (reported in Exoplanet Colonization) went against habitability for two of the nearest candidates. Either direction is a genuine model input; this is the one place where the models can be fed rather than argued with.

Speculative Experiment four is computational and unglamorous: publish the parameter distributions, not the products. The Sandberg–Drexler–Ord method is reproducible and cheap. A standing, versioned, openly licensed set of parameter distributions with provenance for each — who estimated it, from what, in what year — would let every subsequent paper report a distribution instead of a point estimate. The obstacle is not compute. It is that a distribution does not make a headline.

Frontier Experiment five: model discrimination as an explicit exercise. Take the settlement-wave, percolation and empty-galaxy families and ask what joint observation would separate them — a distribution of arrival times in the solar system's geological record, a spatial correlation in technosignature nulls, an artefact census in the outer solar system. The literature has not systematically done this, and until it does, “the models are compatible with the data” is a statement about the data's weakness rather than the models' strength.

Speculative Experiment six, cheap and self-directed: run our own civilisation through the models. Sheikh and colleagues did a version of this and found that an observer at Mars during a conjunction would have had a 77% chance of catching one of our Deep Space Network transmissions, 4 × 10⁵ times better than a random observer at a random time. As a model input that is devastating: it says the detectable-phase term is not a property of a civilisation but of a geometry. Interstellar Archaeology owns the result as a search finding; this brief carries it as a parameter finding.

Established The engineering object in this subject is software, and its quality is the subject's quality. Every result in section 2 is the output of a simulation or a numerical propagation: a Monte Carlo over parameter distributions, a three-dimensional site-percolation lattice, a cellular automaton run to a quarter of a million simulated years, a potential-theory diffusion solve. The engineering requirement is therefore the ordinary one for computational science — published code, published inputs, and results a third party can reproduce — and it is unevenly met.

Frontier The specific requirement that would change the field is provenance on parameter distributions. The Sandberg–Drexler–Ord approach takes its ranges from the published literature; that literature is itself a mixture of measurements, models and educated guesses that are not distinguished when the ranges are quoted. A parameter file that tagged each distribution with its evidential kind — measured, modelled, asserted — would make the difference between the established core and the speculation machine-readable, which is exactly the distinction this brief exists to draw.

Established The second requirement is that model outputs be reported as distributions with their sensitivity to the haystack or lattice definition attached. Wright and colleagues do this for search coverage, stating that their fraction scales as the fifth power of maximum distance and is “very sensitive to our haystack definition and boundaries.” The order of magnitude of the deficit is robust; the digits are not, and the paper says so. Percolation results need the same treatment: the lattice geometry and connectivity are choices, and P_c = 0.311 is a property of a cubic array with N = 6, not of the galaxy.

Speculative A third, which is a genuine gap: no standard benchmark problem exists. The models do not share test cases, so a new settlement model cannot be scored against an old one. In a field where every result is a simulation and no result is testable against nature within a human lifetime, an internal benchmark is the only quality signal available. Its absence is why the literature accumulates without converging.

Established Four boundaries, and the first is already written from the other side. Civilization Timelines states in its own section 8 that this brief “owns the Fermi question, the Drake equation, colonisation-wave models and technosignature search status as astrobiology” while it takes “the Kardashev scale, the Great Filter and the dark forest only as periodisation claims — statements about where along a timeline a civilisation sits or fails.” That is the seam, quoted rather than paraphrased, and this brief holds to it.

Established Second, Interstellar Archaeology: measurement against inference. That brief owns the technosignature search as a programme — coverage, EIRP bounds, methodology frameworks, the RFI problem, artefact proposals. This brief owns what a null does and does not license as a conclusion about the population. One coverage number crosses the seam and nothing else.

Frontier Third, Exoplanet Colonization: the model term against the measurement. II-14 owns the catalogue, the occurrence rates, the JWST atmospheric results and the transit arithmetic. This brief does not re-list exoplanet counts; it carries the occurrence rate once, as the one Drake term that has moved, and the settleable fraction as the term that has not.

Established Fourth, Interstellar Probes: engineering against parameter. Probe speed is an input here and a design problem there, and the load-bearing finding — that the settlement front's floor is set by stellar drift rather than by propulsion — is a statement about the model's insensitivity to the engineering.

Speculative And two briefs this one deliberately does not reach for. Dyson Swarms and Stellar Engineering own megastructure-specific observational results. This brief mentions infrared waste heat once, as one axis of what a technosignature could be, and links out. Megastructure candidate lists, their contamination analyses and their resolutions are not this brief's material, and importing them would duplicate work done properly elsewhere on this map.

Frontier The adjacent scientific fields, named without claiming them: astrobiology and origin-of-life chemistry, which own the term that dominates everything; stellar and galactic dynamics, which supply the 100 km/s floor; statistical physics, which supplies percolation; the philosophy of probability, where the anthropic and observation-selection arguments live and where this brief's central methodological point — that a sample of one drawn by a conditioned observer is not a sample — actually comes from.

Established The institutional fact that explains the state of this literature is a funding fact, and it is stated in an agency document. The NASA technosignatures workshop report records that “the field of searching for technosignatures has had no regular source of federal funding for decades,” that fewer than ten PhDs have been awarded in the area, and that the last comprehensive review was published in 2001. A field with no standing funding line produces theory papers, because theory papers are what one person can produce without a budget — which is exactly the shape of the modelling literature described above.

Frontier The second institutional requirement is a norm, not a budget: report intervals, not headlines. The Westby and Conselice episode is the case study. The paper's own 68% interval spans a factor of fifty and includes zero at roughly 15%, and none of that reached the coverage. A journal norm requiring the interval in the abstract, and requiring the probability of the null result to be stated where it is non-negligible, would have prevented the entire episode at no cost to the science.

Established Third, an archival requirement that costs almost nothing and would compound. The parameter distributions used in any Drake-style propagation should be published as data, versioned, with provenance for each term. The method is cheap and reproducible; the inputs are the whole argument. At present a reader cannot check whether two papers disagree about the world or merely about a prior.

Frontier Fourth, and this one has a live institutional owner: the decision about transmission. Whether to deliberately signal is a search-programme question with no international governance, and the arguments for and against are drawn from exactly the models in this brief — which is a problem, because the models do not discriminate. A policy resting on a model family that cannot distinguish an empty galaxy from a settled one is resting on a preference. Interstellar Archaeology owns the search-programme half and Deep Space Communications the engineering half.

Speculative Fifth, an interest disclosure norm. Several of the sources in this brief are produced by people and institutions whose funding depends on the search continuing. That is not disqualifying — it is unavoidable in a small field — but it should be stated, and the cases where a researcher's result cuts against their institutional interest should be marked, because those are the ones that carry extra weight. Two such cases appear in this brief: Scharf attacking a counting result he is professionally placed to benefit from, and Wright publishing the anti-pendulum caution against the most self-serving reading of his own coverage figure.

Frontier The first ethical question in this subject is about epistemic honesty rather than about aliens. A number with a fifty-fold interval reported as a number is a misrepresentation, whatever the intention behind it, and it enters the culture as a fact. “There are 36 civilizations in the galaxy” is now more widely believed than any of the careful results in this brief, and the correction — that zero sits inside the paper's own interval — travelled to almost nobody.

Speculative The second is the transmission question, and this brief's contribution to it is a negative one. Arguments for and against deliberate signalling are routinely grounded in claims about how many civilisations there are and what they are like. Those claims come from models that, on their own showing, cannot distinguish an empty galaxy from a fully settled one. A risk argument that rests on an undiscriminating model is a preference dressed as an estimate, in both directions.

Frontier Third, and it is underexplored: the models carry assumptions about what civilisations want. Settlement-wave models assume expansion; percolation models assume that expansion is contingent; aestivation assumes patience and a utility function over computation. Each is a claim about motivation dressed as a parameter, and each is unfalsifiable with present evidence. Naming them as such is more useful than choosing between them.

Speculative Fourth, the anthropic problem is a genuine ethical hazard as well as a statistical one. Every term in the equation with a single instance was measured by a party whose existence required that instance to come out as it did. Reasoning from that sample to a population is the observation-selection problem in its purest form, and it is exactly the error the counting papers make when they treat the timing of human intelligence as typical. Scharf's objection is the ethical one as well as the statistical one: an argument that cannot help but conclude that we are typical should be distrusted precisely because we cannot occupy any other vantage point.

Established And a fifth that is small and real: public expectation management. A field whose null result carries almost no information, and whose most careful paper puts a 30% prior on our being alone in the galaxy, is not in a position to promise contact. Saying so does not damage the science. The alternative — a cycle of headline estimates followed by quiet retractions — damages it a great deal, and the record of the last decade shows the cycle running at roughly one turn every three years.

Frontier If the Sandberg–Drexler–Ord result is right, the most likely single fact about our situation is that we are the only technological civilisation in the Milky Way. Not certainly — the whole point of the paper is that certainty is unavailable — but at a prior of 30% rising to somewhere between 53% and 99.6% once the silence is conditioned on, it is the modal outcome of the published parameter ranges. That is a claim of enormous consequence derived from nothing but arithmetic done properly on other people's numbers.

Speculative The consequence usually drawn from it — that the future of value in the universe rests on us — is a value judgement, not a result, and this brief marks it as one. The measured content is narrower: on present evidence, no observation distinguishes a galaxy with one civilisation from a galaxy with many that are quiet, distant or dormant. What follows for policy is close to nothing, which is itself the finding.

Frontier The second implication is about expansion, and it inverts a common intuition. Carroll-Nellenback's 100 km/s floor means that a civilisation does not need fast propulsion to spread across a galaxy; stellar drift alone supplies the transport, and 36 doubling times suffice. The rate-limiting step in galactic settlement is therefore not propulsion but the probability that a settlement itself launches settlements — Landis's P. That is a claim about institutions and motivation on timescales of millions of years, and nothing in physics constrains it.

Speculative Third: the silence is compatible with a settled galaxy, and that possibility is rarely priced. Under the steady-state result, Earth can sit unvisited within a one-million-year evidential horizon inside an inhabited galaxy. Under the percolation result, a void of arbitrary size is generic at criticality. Two independent formalisms produce a quiet local sky without requiring anyone to be absent, hostile, hidden or extinct.

Established Fourth, and it is the implication most worth carrying away: the models cannot yet support any civilizational decision. They do not tell us whether to transmit, whether to expand, whether to expect company or expect solitude. What they do is show, with unusual clarity, exactly which parameter is doing all the work — and that parameter is the origin of life, measured once. A subject that has correctly identified its own binding uncertainty is in better shape than one that reports a number.

These horizons track what could actually change the inputs to a population model — not when contact might occur, which nothing in this brief supports forecasting.

• 10 yr: Frontier Atmospheric characterisation of terrestrial habitable-zone exoplanets continues and either finds air on one or does not; either result is a genuine input to the settleable fraction. Established Technosignature coverage rises by orders of magnitude from a base of 6 × 10⁻¹⁸ and the null remains uninformative about the population at every plausible endpoint in this window. Speculative The Drake-style literature either adopts distributional reporting or does not; the method is cheap, public and eight years old, so adoption is a sociological question rather than a technical one.
• 25 yr: Frontier Outer-system ocean-world missions return the first data capable of bearing on a second origin. A positive result is the only foreseeable event that would compress the dominant parameter's 200-order-of-magnitude uncertainty. Speculative A negative result compresses almost nothing, because absence in one location under one instrument's sensitivity is weak evidence about a rate. Speculative The Rare Earth position either acquires a test or continues as a stance; nothing currently planned tests its core claim about the rarity of complex life.
• 50 yr: Speculative If coverage of the technosignature haystack rises by ten or more orders of magnitude and the sky stays quiet, the null starts to constrain the loud-and-continuous corner of the parameter space and only that corner. Speculative Settlement-wave and percolation models remain unfalsifiable on this horizon, because their discriminating observable — the arrival history of the solar system over hundreds of millions of years — is a geological question this brief cannot forecast. Handwave Any statement about detected civilisations at this horizon is a forecast about an event with no base rate.
• 100 / 250+ yr: Handwave Beyond forecasting. The subject's own most careful result puts a substantial probability on there being nothing to find, and a model family that cannot distinguish that case from its opposite cannot support a century-scale expectation. Speculative The defensible structural statement is that the binding parameter is the origin of life, that it will be constrained by chemistry and by biosignature astronomy rather than by SETI, and that until it is constrained the population question stays open across every order of magnitude it currently spans.

• Depends on Two edges, and both are unusual in that this brief needs a parameter from each rather than a capability. Interstellar Probes owns propulsion and mission architecture; this brief uses probe speed and range only as model inputs, and its most robust finding is that the settlement front's ~100 km/s floor is set by stellar drift and is therefore insensitive to what I-16 achieves. Exoplanet Colonization owns the destination question, and this brief needs one number from it — the Aurora term, the fraction of formally habitable worlds that are actually settleable — which II-14 reports as unmeasured for every world in the conservative sample. Both edges are real dependencies in the strict sense: without them the models run on free parameters, which is what they currently do.
• Requires (not on this map) Two constraints that are not briefs on this map. The first is a scientific result nobody is producing: a second, independent instance of abiogenesis. It is the only observation that would compress the term whose log-uncertainty exceeds 200 orders of magnitude, and every other advance in the subject operates downstream of it. The second is institutional and cheap: a versioned, openly published set of parameter distributions with provenance for each term, so that two papers disagreeing about the number can be shown to disagree about the world rather than about a prior. Neither is a technology, and neither is owned by any brief here.
• Enables No typed enabling edge is claimed. A model family that cannot distinguish an empty galaxy from a fully settled one does not enable a downstream capability, and recording an edge on the strength of thematic proximity would be exactly the failure mode the tech tree exists to avoid. Civilization Timelines reasons from long-run trajectory and cites this brief for the astrobiology, but it takes the periodisation claims as its own and neither brief waits on a result the other produces.
• Adjacent Interstellar Archaeology owns the search programme, its coverage and its bounds; this brief owns the inference from them and carries exactly one coverage figure. Civilization Timelines owns Kardashev, the Great Filter and the dark forest as periodisation claims, and wrote the seam first. Dyson Swarms and Stellar Engineering own megastructure observational results, which appear here only as a named technosignature channel. Deep Space Communications owns transmission engineering, and Mega-Telescopes owns the instruments whose collecting area sets the coverage this brief reasons about.

Established “The Drake equation tells you how many civilisations there are.” It tells you what its inputs tell you, and its inputs span more than 200 orders of magnitude on one term alone. Established Multiplying point estimates instead of distributions produces errors above 42 orders of magnitude in the authors' own toy model. Any single N quoted from the equation is a choice of priors wearing a result's clothing. The equation's value is as a decomposition — it names what you would have to know — and that value is real and is not a calculation.

Frontier “The Fermi paradox is evidence that something strange is going on.” It is a tension between an observation and an estimate, and the estimate is known to be computed wrongly. Frontier Propagate the published ranges honestly and a 30% prior probability of being alone in the Milky Way falls straight out, rising to 53–99.6% once the silence is conditioned on. On that account the silence needs no Great Filter, no zoo, no dark forest and no aestivation. The paradox is a fact about how the estimate was computed, not a fact about the universe.

Speculative “Astronomers estimate there are 36 civilizations in our galaxy.” One paper derived that under an assumption that intelligence follows deterministically after a fixed interval. Established Its own 68% interval runs 4 to 211, and zero sits inside it at roughly 15%. Frontier Two named critics attacked it immediately: one on selection bias — any civilisation would judge its own moment typical — and one on the interval. The headline is the most-repeated quantitative claim in this subject and it is the least defensible.

Speculative “Exoplanet discoveries have confirmed the Rare Earth hypothesis.” They have moved against several of its load-bearing premises. Frontier The claim that only 5–6% of stars have planets has been replaced by nearly all stars having planets; the Jupiter-as-shield argument is contradicted by modelling suggesting Jupiter has caused more impacts than it prevented; and water delivery, far from being rare, comes out excessive in formation simulations. Speculative The core claim — that complex life is rare — remains untested in either direction, and reporting the audit as a refutation would be as wrong as reporting the exoplanet count as a confirmation.

Speculative “Percolation and settlement models tell us where civilisations are.” They demonstrate the opposite: that a wide range of settlement densities produce the sky we see. Speculative Landis's near-critical percolation makes arbitrarily large voids generic; Carroll-Nellenback's steady state leaves Earth unvisited inside an inhabited galaxy. Handwave A model whose predicted observable is identical across its entire parameter range is a catalogue of possibilities. That is a legitimate and useful thing to be, and it is not a distribution of civilisations.

Established “Seventy years of silence is evidence that nobody is out there.” At 6.0 × 10⁻¹⁸ of the eight-dimensional search space examined, the null bounds loud, continuous, beamed-at-us transmitters within tens of parsecs and essentially nothing else. Established NASA's own workshop report states that “current upper limits are quite weak.” The searchers say this in print; it is not a sceptic's complaint about the field but the field's own assessment. Interstellar Archaeology carries the bounds in detail.

Frontier “If they existed and could travel, they would be here.” This is Hart's inference and it is the one the modern literature explicitly rejects. Frontier Carroll-Nellenback and colleagues identify parameter regions in which the galaxy is settled and Earth has still not experienced a settlement event within the geological evidence horizon, and Newman and Sagan reached a compatible conclusion from different mathematics forty-five years ago: there are too many worlds to colonise in the plausible lifetime of a colonisation phase. The observation is decoupled from the conclusion, and the decoupling is a published result rather than a rhetorical escape.

Handwave “The zoo hypothesis explains the silence.” It is consistent with the silence, which is a different property and a much cheaper one. Handwave It predicts every possible observation equally and therefore constrains none. Speculative The aestivation hypothesis is the better-formed member of this family precisely because it makes a physical claim — that computation is enormously cheaper in a colder universe, with a stated 10³⁰ multiplier — and its authors argue that physical law constrains which motivations are compatible with it. It still has no proposed test, and this brief flags it accordingly.

Speculative “A civilisation like ours would obviously be detectable.” Not by our own methods. Frontier An observer at Mars during an Earth–Mars conjunction would have had a 77% chance of catching one of our Deep Space Network transmissions — 4 × 10⁵ times better than a random observer at a random time. Our brightest deliberate emissions are narrow in beam and narrow in time. As a model input this is severe: the “detectable lifetime” term is not a property of a civilisation at all but of a geometry between it and whoever is looking.

Established And the framing itself. “The distribution of civilizations can be modelled” is true and much less useful than it sounds. Frontier The models are well-posed, the mathematics is sound, and the output is a statement about the width of our ignorance rather than about a population. Established The one term exoplanet science has genuinely measured — the occurrence rate of rocky habitable-zone planets, 0.37 to 0.60 per star for G and K dwarfs, nearest instance about 6 parsecs away — was never the binding one. The binding one is the origin of life, and it has a sample size of one, drawn by an observer selected by the outcome.