Scientific research and archaeological evidence
43 source annotated catalogue | 10 October 2026 | Updated 13 October 2026
Research does not establish one universal maximum number of people who should live in a place. It does identify constraints on relationships, housing conditions, resources and coordination, alongside benefits from cooperation and larger networks. Those are distinct questions and must be measured separately [1–6, 32–41].
What the evidence supports
Question | What can be concluded |
How large is a personal social group? | Network sizes and camp sizes show patterns, but means and uncertain extrapolations are not residential limits [1–6]. |
Can settlement growth cause harm? | Crowding, resource demands and exposure to disease can matter. Their effects depend on living conditions and organisation [13–15, 32–34]. |
Is a larger settlement always worse? | No. Some studies find benefits in production and knowledge; their size and distribution vary [7–8, 16–17, 35–37]. |
Did ancient populations exceed capacity? | Specific cases show growth, decline or dispersal. Causes are contested, and familiar collapse narratives can fail empirical tests [10–12, 18–22]. |
At what number do people become redundant? | These studies do not establish such a threshold. Resource shortages, unmet needs, exclusion and limited work opportunities require separate measures. |
How to use this catalogue
Start with the featured settlement-growth papers by Ortman [16] and Hirst [43]. Then read Hill [4] for residential groups, Larsen [13] for settlement history, Shennan and Bevan [10, 12] for demographic change, Crema [30] for methods, and Cohen [38] for carrying capacity. Bracketed numbers refer to catalogue entries.
This is a scoped evidence map, not an exhaustive systematic review or a new meta-analysis. Records were checked against scientific journals, PubMed and institutional repositories. Annotations use available full text, abstracts and indexed primary-paper text; full access varies. Original studies, models, reviews and scientific criticisms are explicitly labelled.
Priority evidence for settlement growth
The two PLOS ONE papers below are central references for the later population-design argument. Their importance lies in the relationship between settlement organisation, infrastructure, resource conditions and growth. They should receive substantive discussion in the main text.
Design intent and room for growth
The proposed settlements begin with 5,000 people and development space allocated for 50,000. The final footprint is planned in advance, with permanent green zones. AL chose 50,000 because it is half the approximately 100,000 threshold remembered from earlier research. That reasoning must remain explicit; the remembered universal ceiling remains a hypothesis to verify.
Geographical boundaries and population policy are intended to guide development while leaving room for natural growth, ecological regeneration and progress through learning. Population can grow towards the planned cap, while knowledge, technology, enterprise and social organisation can continue to develop. The intention is to enhance the process through learning and allow adaptation as conditions change.
Ortman and colleagues 2014
The Pre-History of Urban Scaling
Across approximately 1,500 settlements in the Basin of Mexico, population and occupied area show scaling patterns resembling modern cities. The framework connects social interaction with movement costs and infrastructure. It cites Teotihuacan at approximately 100,000 people and Tenochtitlan-Tlatelolco at approximately 200,000. This makes it central to questions of organisation and scale, rather than evidence for a universal 100,000 ceiling. See entry [16].
Hirst and colleagues 2025
Constraining the population size estimates of the pre-Columbian Casarabe Culture of Amazonian Bolivia
The estimated peak of 10,000 to 100,000 people covers a settlement network across 5,020 square kilometres, occupied between 400 and 1400 CE. It is a close match to the remembered growth wording and is relevant to the proposal for connected settlements within a wider landscape. The relationship to the present design is an interpretation, not a tested result of the study. See entry [43].
Use in the later write up
Discuss both papers prominently when explaining the population cap and the scope for growth. Connect their findings to the design questions, retain the original geographical and temporal scales, and state where the proposal goes beyond the evidence. Neither study establishes that 50,000 is an optimum or that 100,000 is a universal natural roof.
Reading archaeological population evidence
Radiocarbon measures the age of organic material. Population size is an additional inference built from dates, archaeological context and assumptions. A date on one sample does not itself tell us how many people were present [24–31].
Evidence | What it contributes | Main caution |
Radiocarbon dating | Calendar-age probability after calibration. | Old wood, contamination, reservoir effects and calibration plateaus can alter interpretation [24–25]. |
Stratigraphy and Bayesian chronology | Ordering, occupation phases and duration. | A more precise estimate remains conditional on correct archaeological assumptions [25]. |
Buildings and settlement area | Potential households and residential extent. | Buildings may belong to different generations. Household size and contemporaneous occupancy must be estimated [13, 20]. |
Summed radiocarbon distributions | A proxy for changing activity or relative population. | Research effort, preservation and calibration can create peaks or gaps [26–31]. |
Human skeletons and burials | Age profiles, health and some evidence of violence. | Burial samples are selective; many skeletal lesions have multiple causes [9, 13]. |
Stable isotopes | Diet, cultivation, mobility or other environmental information, depending on isotope and material. | Stable carbon and nitrogen isotope analysis is distinct from carbon-14 dating [14]. |
Ancient DNA | Ancestry, kinship and modelled demographic history. | Genetic effective population size differs from census headcount [22]. |
Tree rings and climate archives | Independent chronology and environmental context. | Climate stress and settlement decline can coincide without climate being the sole cause [18–19]. |
Chronology and comparability
BP conventionally means before 1950. “Cal BP” indicates calibrated calendar years; BCE and CE are calendar eras. Keep dating uncertainty when comparing events. An island, a settlement region, a village, a household and a personal network have different boundaries and cannot share one meaningful population threshold [20, 24, 36].
For a defensible settlement estimate, specify the place, time period, simultaneous residents, usable space, resource supply, imports, infrastructure and outcome of concern. Report an interval and sensitivity to assumptions. Separate measured evidence, reconstructed population, proposed mechanism and planning judgement.
Research catalogue
Sources 1 to 3
1 Dunbar, R. I. M. 1992
Neocortex size as a constraint on group size in primates
Journal of Human Evolution 22, 469–493 | Comparative analysis and evolutionary hypothesis
Evidence Compares primate neocortex measures with social-group sizes and extrapolates to humans.
Finding Foundational source for the proposed human relationship-network size commonly rounded to 150.
Limit A regression-based hypothesis about social relationships is not a maximum residential population. The human extrapolation and its uncertainty have been challenged directly [2].
Original paper or authoritative record
2 Lindenfors, P., Wartel, A. and Lind, J. 2021
‘Dunbar’s number’ deconstructed
Biology Letters 17, 20210158 | Statistical reanalysis
Evidence Reanalyses primate brain and group-size data with updated datasets and comparative statistical methods.
Finding Human estimates vary markedly; reported 95% intervals include 4–520 and 2–336 in two analyses.
Limit These are uncertainty intervals for extrapolations, not recommended group-size ranges. The paper challenges a precise universal number; it does not show unlimited social capacity.
Original paper or authoritative record • Free paper
3 Hamilton, M. J. et al. 2007
The complex structure of hunter–gatherer social networks
Proceedings of the Royal Society B | Cross-cultural quantitative analysis
Evidence Analyses 1,189 social groups in 339 hunter–gatherer societies using a compilation of ethnographies and fractal network theory.
Finding Finds statistical organisation across several social scales rather than a single isolated residential group.
Limit Ethnographic categories and historical sampling affect the results. Modern foraging societies are not unchanged replicas of Palaeolithic populations; compare the critique in [5].
Original paper or authoritative record • Free paper
Research catalogue
Sources 4 to 6
4 Hill, K. R. et al. 2011
Co-residence patterns in hunter-gatherer societies show unique human social structure
Science 331, 1286–1289 | Cross-cultural empirical analysis
Evidence Studies 32 present-day foraging societies, covering 5,067 individuals. Mean experienced band size was 28.2 adults.
Finding Most adults in residential groups were genetically unrelated, with flexible residence patterns connecting people to larger networks.
Limit The 28.2 figure counts adults and is a sample mean, not a maximum population or a total including children. Camp size must be distinguished from the wider society.
Original paper or authoritative record • Abstract and source record
5 Bird, D. W., Bliege Bird, R., Codding, B. F. and Zeanah, D. W. 2019
Variability in the organization and size of hunter-gatherer groups: Foragers do not live in small-scale societies
Journal of Human Evolution 131, 96–108 | Ethnographic analysis and model comparison
Evidence Examines flexible residence and foraging networks, including Western Desert Australian social organisation.
Finding Argues that small co-resident groups can draw on extensive networks of relationships rather than nest neatly inside fixed small societies.
Limit A strong corrective to equating a camp with an entire society. It is not a universal replacement population threshold.
Original paper or authoritative record • Abstract
6 Lobo, J. et al. 2022
Scaling of Hunter-Gatherer Camp Size and Human Sociality
Current Anthropology 63, 68–94 | Ethnographic dataset and spatial model
Evidence Compares population and residential area across mobile hunter–gatherer camps and models social interaction and scalar stress.
Finding Shows that how much space a group occupies changes with group size and differs from the pattern in permanent settlements.
Limit Camp layout, mobility and institutions matter. A fitted area–population relationship supplies neither a universal optimum nor an occupancy regulation.
Original paper or authoritative record • Institutional manuscript
Research catalogue
Sources 7 to 9
7 Derex, M., Beugin, M.-P., Godelle, B. and Raymond, M. 2013
Experimental evidence for the influence of group size on cultural complexity
Nature 503, 389–391 | Controlled human experiment
Evidence Uses a dual-task computer game to test cultural transmission and skill accumulation in differently sized groups.
Finding The authors report better preservation of complexity, more improvements and greater trait diversity in larger groups.
Limit Task performance cannot establish an ideal town size. The interpretation was disputed by a published reanalysis [8]; the authors also published a reply.
Original paper or authoritative record
8 Andersson, C. and Read, D. 2014
Group size and cultural complexity
Nature 511, E1 | Published reanalysis and scientific correspondence
Evidence Reexamines the experimental data and interpretation of Derex and colleagues [7].
Finding Argues that a different performance assessment reverses the apparent advantage of larger groups.
Limit This disagreement illustrates why outcome definition matters. It does not establish that smaller communities are universally superior. Read alongside the original experiment and its response.
Original paper or authoritative record • Derex and colleagues reply
9 Bocquet-Appel, J.-P. 2011
When the world’s population took off: the springboard of the Neolithic Demographic Transition
Science 333, 560–561 | Comparative archaeological synthesis
Evidence Examines demographic signals in cemetery age profiles across the transition from foraging to farming.
Finding An increased proportion of juvenile skeletons is interpreted as evidence of rising fertility and population growth.
Limit Burial selection, preservation and age estimation affect cemetery profiles. A larger juvenile fraction is interpreted through a demographic model, not read as a direct census.
Original paper or authoritative record
Research catalogue
Sources 10 to 12
10 Shennan, S. et al. 2013
Regional population collapse followed initial agriculture booms in mid-Holocene Europe
Nature Communications 4, 2486 | Radiocarbon proxy analysis with simulations
Evidence Tests demographic fluctuations in Central and Northwest Europe between 8,000 and 4,000 calibrated years BP.
Finding Finds regional growth and decline after agriculture spread. The climate comparisons in this study did not explain the observed patterns.
Limit The proposed internal causes remain speculative. Relative radiocarbon activity is not an exact headcount, and this study does not identify a universal carrying-capacity ceiling.
Original paper or authoritative record • Free paper
11 Downey, S. S., Haas, W. R. and Shennan, S. J. 2016
European Neolithic societies showed early warning signals of population collapse
PNAS | Retrospective time-series and model analysis
Evidence Calculates early-warning statistics from summed archaeological radiocarbon date probability densities.
Finding Reports statistical signals consistent with declining resilience before some reconstructed population collapses.
Limit Retrospective signals do not provide a validated forecast for a modern town. Calibration, sampling and dependence in the proxy series require careful treatment [30].
Original paper or authoritative record • Free paper
12 Bevan, A. et al. 2017
Holocene fluctuations in human population demonstrate repeated links to food production and climate
PNAS 114, E10524–E10531 | Radiocarbon and subsistence analysis
Evidence Compares population proxies across Britain and Ireland with climate patterns and directly dated plants and animals.
Finding Finds repeated downturns coinciding with climate reorganisation and shifts in food procurement. Challenges wholly internal explanations of Neolithic boom and bust.
Limit Temporal correspondence and plausible mechanisms do not establish one cause for every decline. A correction exists and should accompany detailed reuse of the study.
Original paper or authoritative record • Data and analysis
Research catalogue
Sources 13 to 15
13 Larsen, C. S. et al. 2019
Bioarchaeology of Neolithic Çatalhöyük reveals fundamental transitions in health, mobility, and lifestyle in early farmers
PNAS 116, 12615–12623 | Bioarchaeology and settlement reconstruction
Evidence Combines human remains, settlement context and biogeochemical evidence for approximately 7100–5950 cal BCE in Anatolia.
Finding The paper reports a Middle Period peak estimate of about 3,500–8,000 inhabitants and changing health, labour, mobility and interpersonal conflict across the settlement’s history.
Limit The population range is reconstructed, not counted. Some skeletal lesions are nonspecific; co-occurrence of crowding and harm does not isolate density as the sole cause.
Original paper or authoritative record • Free paper
14 Schlütz, F. et al. 2023
Isotopes prove advanced, integral crop production, and stockbreeding strategies nourished Trypillia mega-populations
PNAS 120, e2312962120 | Stable-isotope and archaeological analysis
Evidence Analyses stable isotopes in archaeological food-system evidence from Copper Age Trypillia communities in eastern Europe.
Finding Supports intensive cultivation, manuring and integrated livestock practices capable of nourishing large settlements.
Limit Stable isotopes inform diet and cultivation; they are not radiocarbon dates or direct population counts. Contemporaneous house occupancy remains a separate issue. A 2024 correction accompanies the paper.
Original paper or authoritative record • Correction
15 Bentley, R. A. et al. 2024
Modelling cultural responses to disease spread in Neolithic Trypillia mega-settlements
Journal of the Royal Society Interface 21, 20240313 | Epidemiological simulation grounded in site layout
Evidence Runs susceptible–infected–recovered–susceptible models on clustered networks and a Trypillia settlement plan, relevant to the fourth millennium BCE.
Finding Lower occupation density or fewer interactions between residential clusters can reduce modelled disease spread.
Limit Model plausibility is not direct proof of an ancient epidemic or deliberate distancing policy. Thresholds depend on the specified pathogen and contact assumptions.
Original paper or authoritative record • Data and code
Research catalogue
Sources 16 to 18
16 Ortman, S. G., Cabaniss, A. H. F., Sturm, J. O. and Bettencourt, L. M. A. 2014
The Pre-History of Urban Scaling
PLOS ONE 9, e87902 | Archaeological settlement dataset and scaling model
Evidence Tests estimated population and settled area across approximately 1,500 pre-Hispanic settlements in the Basin of Mexico, spanning four cultural periods.
Finding Finds spatial scaling patterns resembling modern urban systems, connecting social interaction with infrastructure and settlement area. The paper cites Teotihuacan at about 100,000 people and Tenochtitlan-Tlatelolco at about 200,000.
Limit Population estimates and settlement boundaries influence the fitted relationships. Average scaling is not a maximum-size rule or an individual growth trajectory. Priority for the later write-up is recorded near the front of this catalogue.
Original paper or authoritative record
17 Ortman, S. G. et al. 2015
Settlement scaling and increasing returns in an ancient society
Science Advances 1, e1400066 | Archaeological quantitative analysis
Evidence Examines monuments and house-size evidence from settlements in the pre-Hispanic Basin of Mexico.
Finding Finds patterns consistent with increasing returns to socioeconomic production as settlement population rises.
Limit Output proxies do not measure everybody’s welfare. The findings show why a larger population may confer benefits as well as costs; they do not prove unlimited growth.
Original paper or authoritative record • Free paper
18 Axtell, R. L. et al. 2002
Population growth and collapse in a multiagent model of the Kayenta Anasazi in Long House Valley
PNAS 99, 7275–7279 | Agent-based model tested against archaeological patterns
Evidence Combines settlement records, geomorphology, pollen and tree-ring climate evidence in a 96 km² Arizona valley, focusing on CE 200–1300.
Finding Models households, maize production, settlement decisions and nutritional constraints; reproduces major features of population and spatial change.
Limit A close fit is not unique causal proof. The assumptions about farming, household behaviour and resources restrict transfer to other places.
Original paper or authoritative record • Free paper
Research catalogue
Sources 19 to 21
19 Kennett, D. J. et al. 2012
Development and disintegration of Maya political systems in response to climate change
Science 338, 788–791 | Palaeoclimate and historical archaeological comparison
Evidence Compares a precisely dated, subannual cave-climate record from Belize with historical sequences recorded on Maya monuments.
Finding Associates wetter conditions with expansion and later drying with political fragmentation and decline during parts of the first and early second millennia CE.
Limit Regional timing varies. The study’s causal interpretation should not be treated as a universal account of Maya history or evidence that a particular headcount was exceeded.
Original paper or authoritative record
20 Klassen, S. et al. 2021
Diachronic modeling of the population within the medieval Greater Angkor Region settlement complex
Science Advances 7, eabf8441 | Lidar and archaeological demographic modelling
Evidence Combines lidar, excavation, radiocarbon dates and machine learning to reconstruct settlement growth in medieval Cambodia.
Finding Estimates approximately 700,000–900,000 inhabitants in the Greater Angkor Region at its thirteenth-century peak.
Limit This refers to an extensive settlement region, not a single tightly packed neighbourhood. Household assumptions and spatial boundaries affect the estimate; it is not a demonstrated optimum.
Original paper or authoritative record • Free paper
21 Davis, D. S. et al. 2024
Island-wide characterization of agricultural production challenges the demographic collapse hypothesis for Rapa Nui (Easter Island)
Science Advances 10, eado1459 | Remote sensing and agricultural-capacity reconstruction
Evidence Uses shortwave infrared satellite imagery and machine learning to estimate island-wide rock-garden extent.
Finding Finds less agricultural infrastructure than some earlier estimates, challenging the assumed large pre-contact population underlying a common collapse narrative.
Limit Crop yield, diet, classification error and other food sources affect capacity estimates. This is a specific test of Rapa Nui claims, not proof that resource overshoot never occurs.
Original paper or authoritative record • Open dataset
Research catalogue
Sources 22 to 24
22 Moreno-Mayar, J. V. et al. 2024
Ancient Rapanui genomes reveal resilience and pre-European contact with the Americas
Nature | Ancient DNA and radiocarbon dating
Evidence Sequences genomes from 15 ancient Rapanui individuals, radiocarbon dated to CE 1670–1950, and models earlier population history.
Finding Finds no genetic evidence for the proposed severe seventeenth-century population collapse.
Limit Genetic effective population size is not census population. Inferences depend on sampling and demographic models; absence of a bottleneck signal does not exclude every local or short-lived hardship.
Original paper or authoritative record • Free paper
23 Turchin, P. et al. 2018
Quantitative historical analysis uncovers a single dimension of complexity that structures global variation in human social organization
PNAS | Comparative historical database analysis
Evidence Uses Seshat historical and archaeological coding and principal-component analysis to compare dimensions of social organisation across world regions.
Finding Finds strong covariation between social scale, governance, economic organisation and information systems.
Limit Results depend on coding, missing evidence and variable choice. A shared statistical dimension does not prove an inevitable evolutionary ladder or establish a population ceiling. Published online in 2017; journal issue in 2018.
Original paper or authoritative record • Free paper
24 Reimer, P. J. et al. 2020
The IntCal20 Northern Hemisphere Radiocarbon Age Calibration Curve (0–55 cal kBP)
Radiocarbon 62, 725–757 | Calibration reference and methods
Evidence Integrates independently dated evidence, including tree rings and other archives, to calibrate Northern Hemisphere radiocarbon measurements.
Finding Provides the conversion from radiocarbon age to calendar-age probability over a range extending to 55,000 calibrated years BP.
Limit Calibration does not remove all uncertainty. Geographic and marine-reservoir effects require appropriate treatment. Dating an organic sample does not directly date every associated structure or count residents.
Original paper or authoritative record
Research catalogue
Sources 25 to 27
25 Bronk Ramsey, C. 2009
Bayesian Analysis of Radiocarbon Dates
Radiocarbon 51, 337–360 | Chronological modelling methods
Evidence Presents Bayesian models that combine radiocarbon evidence with archaeological ordering, phases and other prior information using OxCal.
Finding Allows explicit estimation of event timing, occupation phases and durations rather than reliance on isolated date midpoints.
Limit Modelled precision depends on the correctness of context and ordering assumptions. Check outliers, model agreement and sensitivity before deciding that buildings were occupied simultaneously.
Original paper or authoritative record
26 Surovell, T. A. et al. 2009
Correcting temporal frequency distributions for taphonomic bias
Journal of Archaeological Science 36, 1715–1724 | Preservation-bias model
Evidence Uses dated geological deposits to investigate how time-dependent loss of evidence changes archaeological frequency distributions.
Finding Shows why an apparent increase toward the present can arise partly because older evidence has been destroyed or buried.
Limit A general correction may not match a particular local landscape or material. Preservation processes must be tested, not treated as a universally solved adjustment.
Original paper or authoritative record • Institutional PDF
27 Williams, A. N. 2012
The use of summed radiocarbon probability distributions in archaeology: a review of methods
Journal of Archaeological Science | Methods review with empirical testing
Evidence Uses 2,996 Australian archaeological radiocarbon dates to explore sample size, calibration effects and preservation corrections.
Finding Demonstrates that date-series construction choices can materially change inferred population patterns.
Limit Its proposed numerical rules, including minimum sample sizes and smoothing windows, are recommendations from a particular analysis, not timeless universal standards. Read with later work [29–31].
Original paper or authoritative record
Research catalogue
Sources 28 to 30
28 Timpson, A. et al. 2014
Reconstructing regional population fluctuations in the European Neolithic using radiocarbon dates: a new case-study using an improved method
Journal of Archaeological Science 52, 549–557 | Statistical methods and archaeological application
Evidence Tests population-proxy fluctuations in 12 European regions between approximately 8,000 and 4,000 years BP.
Finding Improves statistical testing of radiocarbon distributions and regional departures from demographic expectations.
Limit Significance is conditional on the chosen null model and sampling treatment. It does not convert a probability distribution into exact population counts.
Original paper or authoritative record • Institutional PDF
29 Contreras, D. A. and Meadows, J. 2014
Summed radiocarbon calibrations as a population proxy: a critical evaluation using a realistic simulation approach
Journal of Archaeological Science 52, 591–608 | Simulation-based methodological critique
Evidence Simulates the relationship between known demographic histories and recovered summed radiocarbon distributions.
Finding Tests whether population signals survive the production, preservation, discovery and dating processes that create archaeological samples.
Limit Visible peaks and troughs require formal validation. A demographic explanation cannot be assumed merely because a curve looks like growth and collapse.
Original paper or authoritative record • Author manuscript
30 Crema, E. R. 2022
Statistical Inference of Prehistoric Demography from Frequency Distributions of Radiocarbon Dates: A Review and a Guide for the Perplexed
Journal of Archaeological Method and Theory 29, 1387–1418 | Critical methods review
Evidence Compares reconstruction, null-hypothesis tests and model-fitting approaches to radiocarbon population inference.
Finding Explains why sampling, calibration and ambiguity about what is being counted can generate misleading demographic patterns.
Limit A useful quality-control reference for this whole catalogue. It explains both the promise and the limitations of the methods rather than rejecting all archaeological demographic inference.
Original paper or authoritative record
Research catalogue
Sources 31 to 33
31 Crema, E. R. and Bevan, A. 2021
Inference from Large Sets of Radiocarbon Dates: Software and Methods
Radiocarbon | Reproducible software and methods paper
Evidence Introduces rcarbon tools for calibration, aggregation, statistical tests and spatial analysis of radiocarbon datasets.
Finding Provides practical methods for transparent analysis and comparison of archaeological date collections.
Limit Software does not establish the validity of a population proxy. Preserve raw data, context, calibration settings, binning choices and code; test how reasonable alternatives change conclusions.
Original paper or authoritative record
32 Stokols, D. 1972
On the distinction between density and crowding: Some implications for future research
Psychological Review 79, 275–277 | Conceptual scientific paper
Evidence Separates objective density from the psychological experience of crowding.
Finding Establishes why people per unit area and experienced lack of space or control should be measured separately.
Limit This is a foundational conceptual distinction, not an experiment establishing a threshold. It should guide the selection of empirical outcomes and explanatory variables.
Original paper or authoritative record
33 World Health Organization 2018
Report of the systematic review on the effect of household crowding on health
WHO Housing and Health Guidelines | Official systematic evidence review
Evidence Reviews studies relating household crowding to infectious disease and other health outcomes to support WHO housing guidance.
Finding Provides a direct evidence base for evaluating inadequate space within dwellings and its health implications.
Limit Definitions and study quality vary by outcome. Household crowding is not the same as neighbourhood density or total town population; the review does not set a universal town-size limit.
Original paper or authoritative record • WHO publication
Research catalogue
Sources 34 to 36
34 Lederbogen, F. et al. 2011
City living and urban upbringing affect neural social stress processing in humans
Nature 474, 498–501 | Human experimental neuroimaging with observational exposure measures
Evidence Tests social-stress responses in healthy volunteers and relates brain activation to current urban living and urban upbringing.
Finding Reports different associations with amygdala and regulatory brain-region responses.
Limit Residence was not randomly assigned. Urbanicity bundles many exposures, and brain activation is not itself a diagnosis or a count of residents beyond a safe threshold.
Original paper or authoritative record
35 Bettencourt, L. M. A. et al. 2007
Growth, innovation, scaling, and the pace of life in cities
PNAS 104, 7301–7306 | Comparative urban data and mathematical modelling
Evidence Examines how infrastructure and socioeconomic indicators change with city population across urban systems.
Finding Reports economies of scale for some infrastructure and faster-than-proportional increases for some social outputs and costs.
Limit These are statistical relationships, not physical laws guaranteeing outcomes in every city. Boundary definitions, variation and contrary results matter [36].
Original paper or authoritative record
36 Arcaute, E. et al. 2015
Constructing cities, deconstructing scaling laws
Journal of the Royal Society Interface 12, 20140745 | Empirical robustness and boundary-sensitivity analysis
Evidence Constructs thousands of city definitions using commuting and population-density thresholds in England and Wales.
Finding Finds that population alone does not sufficiently describe cities; many indicators scale approximately linearly and nonlinear exponents can vary with boundaries.
Limit A direct warning against applying one urban scaling equation everywhere. Administrative towns, built-up areas and commuting regions are not interchangeable units.
Original paper or authoritative record • Institutional paper
Research catalogue
Sources 37 to 39
37 Ahlfeldt, G. M., Redding, S. J., Sturm, D. M. and Wolf, N. 2015
The Economics of Density: Evidence from the Berlin Wall
Econometrica 83, 2127–2189 | Natural-experiment setting and structural economic model
Evidence Uses thousands of Berlin city blocks in 1936, 1986 and 2006 and variation arising from division and reunification.
Finding Estimates local production and residential externalities and how accessibility changes urban activity.
Limit The historical setting strengthens identification, but structural assumptions still matter. Productivity and property prices are not complete measures of welfare or ecological capacity.
Original paper or authoritative record • Author PDF
38 Cohen, J. E. 1995
Population growth and earth’s human carrying capacity
Science 269, 341–346 | Analytical synthesis and mathematical models
Evidence Examines how natural constraints interact with human choices in population-support capacity.
Finding Concludes that human carrying capacity is dynamic and uncertain, depending on economic, environmental, cultural and demographic choices.
Limit This provides a framework, not a single verified maximum for Earth or a settlement. Any numerical estimate must specify consumption, technology, distribution, trade and environmental conditions.
Original paper or authoritative record • Abstract
39 Gerten, D. et al. 2020
Feeding ten billion people is possible within four terrestrial planetary boundaries
Nature Sustainability 3, 200–208 | Spatially explicit food-system scenario modelling
Evidence Models agriculture against constraints on biodiversity, land use, freshwater and nitrogen flows.
Finding Under the analysed boundaries, the model supports 3.4 billion people with the existing system versus 10.2 billion after specified production and consumption changes.
Limit These are conditional model outputs, not observed carrying capacities or a recommendation about population. The study covers four boundaries and particular dietary and agricultural assumptions.
Original paper or authoritative record • NASA source and abstract
Research catalogue
Sources 40 to 42
40 O’Neill, D. W., Fanning, A. L., Lamb, W. F. and Steinberger, J. K. 2018
A good life for all within planetary boundaries
Nature Sustainability 1, 88–95 | Cross-national resource and social-outcome analysis
Evidence Compares resource use and indicators of basic needs and wellbeing across more than 150 countries.
Finding Examines whether social outcomes are achieved within allocated biophysical limits, connecting human welfare with resource consumption.
Limit Country averages hide inequality. Boundary allocation and wellbeing thresholds involve explicit analytical choices; the paper is not a settlement-population calculator.
Original paper or authoritative record • University research and data site
41 Ostrom, E. 2009
A general framework for analyzing sustainability of social-ecological systems
Science 325, 419–422 | Research synthesis and analytical framework
Evidence Organises evidence on resource systems, resource users, governance and interactions in social–ecological systems.
Finding Explains why resource outcomes depend on institutions and local conditions as well as the number of users.
Limit The framework guides comparative studies; it supplies no universal numerical capacity. Governance should be measured rather than assumed to improve or deteriorate automatically with size.
Original paper or authoritative record
42 Calhoun, J. B. 1973
Death squared: the explosive growth and demise of a mouse population
Proceedings of the Royal Society of Medicine 66, 80–88 | Captive mouse experiment
Evidence Reports growth and eventual demographic failure in an enclosed mouse population, including changes in social and reproductive behaviour.
Finding Relevant to the history of ideas about crowding, social roles and the so-called behavioural sink.
Limit Captive mice could not migrate or redesign institutions as humans can. This experiment does not establish a human village or city limit, and its social-role interpretation is not a measure of human redundancy.
Original paper or authoritative record • Original source record
Research catalogue
Source 43
43 Hirst, J., Singarayer, J. S., Lombardo, U. and Mayle, F. 2025
Constraining the population size estimates of the pre-Columbian Casarabe Culture of Amazonian Bolivia
PLOS ONE 20, e0325104 | Archaeological reconstruction and population modelling
Evidence Combines construction-labour estimates, carrying-capacity calculations and agent-based population modelling. Radiocarbon evidence underpins the occupation chronology; it does not directly count residents.
Finding Estimates a peak population of 10,000 to 100,000 within a 5,020-square-kilometre study area during occupation between 400 and 1400 CE.
Limit This is a conditional regional estimate covering multiple settlements. Model assumptions and resource productivity matter. It establishes neither a universal city ceiling nor a modern design optimum.