5D Library › Scaling Laws, Allometric Growth and the Statistical Physics of Organisational Metabolism · 中文
Preface: Mice, Elephants, Cities and Companies
Note
Please read this first. This book introduces several mathematical models of biological allometry, urban scaling and corporate survival, together with their empirical evidence, and uses short programs to make its arguments. Unless a source is given, the “species”, “cities”, “companies”, “teams” and data in the book are fictional or synthesised by the programs, and the parameters are illustrative. The book discusses mathematical models of biological allometry and of the size of cities and firms; it is not investment advice and promises no results for the management of any business. Its models of firms and teams are schematic models written to show a structure, not estimates or forecasts for any real company. The material on heart rate and lifespan in Chapter 1 is a statistical regularity of comparative biology and not health advice; it does not imply that slowing one’s heart rate would lengthen one’s life.
When concepts from biology (metabolism, growth limits, fractal supply networks) are applied to cities, firms and teams, they are models and analogies: they help us ask better questions, but a company is not an animal and a city has no heart.
A mouse weighs about one hundred-thousandth as much as an elephant, yet it burns far more energy per gram. In 1932 Max Kleiber, who studied livestock nutrition at the University of California, plotted the basal metabolic rates of different animals on logarithmic axes and found that they fell roughly on a straight line with a slope close to 3/4: a ten-thousandfold increase in body mass brings only about a thousandfold increase in metabolic rate (Kleiber 1932). More than sixty years later the physicist Geoffrey West and the ecologists James Brown and Brian Enquist proposed that this three-quarter power arises from the fractal branching networks that supply the whole body (West et al. 1997); a decade after that, West, Luís Bettencourt and colleagues carried the same reasoning over to cities and found that urban infrastructure grows sublinearly with population while wages, patents and crime grow superlinearly (Bettencourt et al. 2007). They went on to argue that companies resemble animals more than cities, and that companies therefore “die of old age” (West 2017).
This book asks four questions. First, how reliable is the three-quarter power law of biology, and where does its theoretical explanation hold and where is it disputed? Second, why does a system whose supply grows sublinearly while its maintenance grows linearly stop growing at a certain size, and does this conclusion carry over to cities and firms? Third, why is the size of teams and firms limited by communication and hierarchy, and which claims rest on experiments and which are rules of thumb? Fourth, can these ideas be tested in a controlled experiment?
What this book covers
Chapter 1 discusses Kleiber’s law, the fractal-network model of West, Brown and Enquist, and the statistical dispute over 2/3 and 3/4; Chapter 2 discusses the model of ontogenetic growth and its asymptotic size; Chapter 3 discusses urban scaling, the finite-time singularity of the superlinear growth equation and the criticisms of urban scaling; Chapter 4 uses survival analysis to discuss the “mortality” of listed companies, and a schematic model of administrative overhead and survival; Chapter 5 discusses the Ringelmann effect, social loafing and Brooks’s law; Chapter 6 discusses span of control, managerial hierarchy, executive pay and the dispute over Dunbar’s number. Chapter 7 reads this research beside the “Momentum” chapter of the Sunzi, the “Eight Canons” of the Han Feizi and the “Examining Divisions” section of the Lüshi Chunqiu, and states where classical experience of organisation ends and modern scaling laws begin. Chapter 8 offers a “four-step checklist for scale and organisational metabolism” and a pre-registered group-randomised controlled experiment in which 12 people divide the writing of code by module and must finally pass a hidden integration test suite written in advance, comparing the completion times of “three squads of four plus standardised interface documents” and “twelve people in a flat structure communicating freely”, with both arms communicating only through logged, structured text chat.
The book borders on several other volumes of 5D Library, which it recalls only briefly without repeating their programs: Volume 20, Credit Cycles, Self-Organised Criticality and the Statistical Physics of Market Phase Transitions, covered power-law distributions and critical phenomena, and Chapter 4 only cites the distribution of firm sizes; Volume 25, Social Network Topology, Structural Holes and Complex Contagion, covered network structure, and Chapters 5 and 6 discuss teams only in terms of size and the number of communication channels; Volume 28, Phase Transitions, Critical Slowing Down and Resilience in Complex Dynamical Systems, covered critical transitions and early-warning signals before collapse, which border on the conclusion of Chapter 3 that growth stagnates or collapses unless innovation resets it.
Claims corrected after checking
Several statements in the author’s original framework needed correction or qualification after checking.
First, the three-quarter power is not an undisputed universal law. Kleiber’s 1932 data comprised only 13 groups of animals, and he proposed the 3/4 power of body mass as a unit for comparing metabolism. Later large-sample analyses gave different results: White and Seymour (2003), analysing 619 mammal species, concluded that basal metabolic rate is proportional to the 2/3 power of mass (White and Seymour 2003); Dodds and colleagues (2001) found an exponent of 0.668 for mammals up to 10 kg, rising systematically when larger animals were added (Dodds et al. 2001); Glazier’s 2005 review showed exponents ranging roughly from 2/3 to 1 across groups and physiological states (Glazier 2005); and Kolokotrones and colleagues (2010) found that the relation is itself curved on logarithmic axes (Kolokotrones et al. 2010).
Second, “about 1 to 1.5 billion heartbeats in a lifetime” needs qualification. The mammalian data compiled by Levine (1997) average about 0.73 billion, with a standard deviation of about 0.56 billion, and the mean the abstract gives when the data are taken together is about 1 billion (Levine 1997); “about a billion” is therefore acceptable as an order of magnitude, but “1 to 1.5 billion” lies above the average of the compiled data and hides a scatter of an order of magnitude; at 70 beats a minute for 80 years a human has about 2.9 billion, four times this average, and bats live much longer than mammals of similar mass (Austad and Fischer 1991). That heart rate scales roughly as mass to the −1/4 and lifespan as +1/4 is a theoretical expectation and a rough empirical trend with wide scatter.
Third, in the West–Brown–Enquist model the exponent 3/4 holds exactly only as the number of branching levels tends to infinity; the smallest vessels are in the Poiseuille regime, with radii shrinking by an area-increasing ratio, and networks of finite size give exponents that differ from 3/4. Savage and colleagues (2008) calculated that over the range of mammalian body sizes the model should predict about 0.81 (Savage et al. 2008). The model’s derivation and biological assumptions have also been questioned (Kozłowski and Konarzewski 2004).
Fourth, the ontogenetic growth equation is formally a special case of von Bertalanffy’s growth equation (Bertalanffy 1957), and the collapse of all species onto one “universal curve” cannot distinguish the 3/4 power from the 2/3 power (Banavar et al. 2002).
Fifth, the urban exponents “about 0.85 and 1.15” in the framework are convenient representative values. In the 2007 table of Bettencourt and colleagues, infrastructure indicators have exponents between 0.77 and 0.87 and socioeconomic indicators between 1.08 and 1.34 (new patents 1.27, for example) (Bettencourt et al. 2007). The arithmetic that doubling the population multiplies the totals by 1.80 and 2.22 is correct, but per person this is only about −10 % and +11 %, not the “15 %” often quoted in popular accounts. The definition of city boundaries changes the exponents (Arcaute et al. 2015), and the choice of noise model changes the conclusion about whether scaling is nonlinear at all (Leitão et al. 2016); the finite-time singularity is a mathematical consequence of the growth equation, not an observed fact.
Sixth, company mortality. The abstract of the 2015 paper by Daepp and colleagues refers to “more than 25,000” North American listed companies between 1950 and 2009 and a typical half-life of “about a decade” (Daepp et al. 2015); the figures “28,853 companies” and “10.5 years” in the framework are those given in Chapter 9, “Toward a Science of Companies”, of West’s 2017 book Scale (West 2017); they are consistent with the round figures of the abstract, and the book is cited here by chapter only. More importantly, “death” here means disappearance from the database of listed companies, most commonly through acquisition or merger, and is not the same as bankruptcy.
Seventh, Ringelmann’s data, including his rope-pulling trials, were collected in the 1880s and published in 1913; the much-cited table of “force per person declining with group size” summarises many observations without naming the task, and its description as a rope-pulling table comes from later retellings; Ringelmann himself attributed the loss mainly to a lack of simultaneity in effort (Kravitz and Martin 1986). Ingham and colleagues (1974) found that in real groups the output per person no longer fell noticeably beyond three people, and used “pseudo-groups” to show that much of the loss is motivational rather than a matter of coordination (Ingham et al. 1974).
Eighth, Brooks’s law is a rule of thumb. In the 1995 anniversary edition Brooks himself called it an “outrageous oversimplification”, and later discussion holds that it applies only when tasks are hard to partition, newcomers are slow to become productive and the project is near its end (Brooks 1995; McConnell 1999).
Ninth, Simon’s 1957 paper is about why executive pay rises with company size, not about the optimal span of control; the model of optimal span and hierarchy comes from Williamson (1967) (Simon 1957; Williamson 1967). Dunbar’s figure of 147.8 is a regression extrapolation from 1993 with a 95 % confidence interval of 100.2 to 231.1 (Dunbar 1993); the Bayesian analysis by Lindenfors and colleagues in 2021 gives a 95 % interval of 4 to 520, so the framework’s figures check out (Lindenfors et al. 2021).
Tenth, the classical texts. The line quoted in the framework, “when the ruler likes to act himself, those who hold office abandon their offices and cater to what the ruler does”, comes from “Guarding the Ruler’s Position” (《君守》), the second chapter of the “Examining Divisions” section (《审分览》) of the Lüshi Chunqiu, not from its first chapter, “Examining Divisions” (《审分》); and the final words read 阿主之为 (“cater to what the ruler does”), not 之己. A passage of “Examining Divisions” that bears more directly on this book is “when the land is worked in common, the work is slow, for there is somewhere to hide one’s strength; when the land is divided, the work is fast, for there is nowhere to hide one’s slackness.” The quotations from the Sunzi and from the Han Feizi’s “Eight Canons” check out; the latter comes from the canon on “the way of the ruler” (主道).
Eleventh, the channel counts in the Chapter 8 trial, 3 × 6 + 3 = 21 against 66, are correct; but “three squads plus liaisons plus standardised interface documents” combines several ingredients, and the trial can test only the combination, not the separate effects of the squad structure, the liaisons and the interface documents; since both arms communicate only by text chat, the conclusions also hold only under that condition of communication.
Each chapter states these points in place, and they are collected in the editorial notes published with the book.
How to read the boxes
As in other volumes of 5D Library, different kinds of statement are marked separately.
Evidence
Evidence boxes report published research findings, with sources.
Calculation
Calculation boxes give formulas and short Python programs. Every program runs unchanged in a standard Python 3 environment using only the standard library, and its actual output is printed beneath it. The “species”, “cities”, “companies”, “teams” and data in the book are fictional or synthesised by the programs; parameters are illustrative values, not estimates for any real system.
Hypothesis · not established
Hypothesis boxes contain the author’s ideas, are marked “not yet established”, and state what result would show them to be wrong.
Source text
Source boxes quote classical texts, with the author’s explanation.
Note
Note boxes give cautions and explain choices.
About this edition
This is version 1.0.0, the first complete edition: a preface and eight chapters, in Chinese and English. The book contains 15 programs and 4 vector figures. The text, boxes and code were drafted with an AI writing tool from the author’s title, framework and notes; the author is responsible for reviewing and revising them and for checking every source. Corrections and missing references are welcome through the 5D Library website.
Disclosure
The author uses the works of 5D Library, including this book, as the methodological basis of his community courses, the 5D Ascension Cards (Dimensional Shift Card Decks) and the Yafu AI knowledge base, and therefore has a commercial interest in the subject of this book. The author also founded and runs Yafu Global and Asia Influencer X, whose business involves the organisational growth and management of firms and teams, and therefore has a direct commercial interest in the subject of this book. The book discusses mathematical models of biological allometry and of the size of cities and firms; it promotes no course, card deck or service, is not investment advice, promises no results for the management of any business, and uses no client or employee data from these businesses; the Chapter 8 trial is designed to be run and analysed by researchers independent of the author’s business and does not recruit from the author’s community, students or customers, including the clients and employees of these businesses; all the code and the study protocol are published openly for anyone to check.
References
Fan, Y. (2026). Scaling Laws, Allometric Growth and the Statistical Physics of Organisational Metabolism: From Kleiber's Three-Quarter Power Law and Fractal Supply Networks to the Mortality of Firms, Read Beside Classical Chinese Organisational Thought [规模法则、异速生长与组织新陈代谢的统计物理] (Version 1.0.0). 5D Library. https://doi.org/10.5281/zenodo.23788553 · Licensed under Creative Commons Attribution-NonCommercial-NoDerivatives 4.0 International (CC BY-NC-ND 4.0).