Some weeks ago I was approached by an editor from Cambridge Scholars Publishing (an imprint I’ve never heard about before) with a request to review a book that presented a new theory of history. I agreed, the book arrived in the due course, and so here are my thoughts as promised.
First, about the author. Jim Penman received his Ph.D. in history in 1985, but he didn’t become an academic, instead going into business. He started a servicing franchise, called Jim’s Group, which eventually become very successful. According to a recent interview, a career in business wasn’t his first choice; he went into it because he couldn’t get a job in academia.
Jim Penman at the Jim’s Group headquarters in 2012. Photo: Luis Enrique Ascui. Source
But he kept his interest in history through the years, and now he is out with the book, Biohistory, which summarizes his thinking and theories about such Big Questions as, why are some societies are much wealthier than others? Why do civilizations fall? These are, of course, questions that have been asked many thinkers in the past, and ones that we are addressing, or plan to address, in our own work within the Seshat project.
The book itself is clearly written and well-researched. Penman is up on current literature and cites most important recent research in theoretical social history, including my own (on which more below).
So what is Biohistory? Basically, it’s the idea that the main moving force of history is the biological characteristics of individual people. Penman comes up with several such characteristics, assigning them letters C, V, S, and so on. For example, C is “a physiological system that adjusts behavior to conditions of limited food. … Characteristics of C in humans include hard work, discipline, and willingness to sacrifice present consumption for future benefit.” Penman then comes up with a variety of factors that affect how C changes with time in a particular society.
The book is quite long (over 600 pages) and presents many different ideas, all of which I lack space to discuss. So let’s take one of the more dynamical theories in the book, which Penman calls “lemming cycles.” I chose to address this particular theory, because I have some expertise in lemming cycles (in fact, my greatest scientific coup was publishing a paper in Nature that was featured on the journal issue’s cover about lemming cycles).
Perhaps more relevant is that I have also done a lot of work on the structural-demographic dynamics, which is an alternative theory to what Penman proposes.
Here’s a summary of Penman’s theory (from the Glossary at the end of the book):
A fluctuation in a trait known as ‘G’, believed to be a mammalian mechanism triggered by occasional food shortages which causes an alternation between V and C. Reflected by changes in population growth, political cohesion, and habits of thought. Normally around 300 years in humans but lengthens in ‘Dark Ages’ following the collapse of civilizations.
There are two major problems with this theory. The first one is that it is of the kind that I call ‘the exogenous driver’ mechanism. As far as I could determine (there are no explicit mathematical equations), at various phases of G—G+30 years, G+60 years, etc.—various events are supposed to happen. It’s like a clock, in which the hour hand sweeps around the clock face and triggers various events (e.g., an alarm that wakes you up). This is not a truly mechanistic, dynamical theory, because it doesn’t explain why different events are triggered at various parts of the cycle.
In contrast, the structural-demographic theory uses the systems approach. It represents different parts of the social system—general population, elites, the state—as mutually interactive compartments. The secular cycles arise as a result of nonlinear feedback loops relating these compartments to each other. It’s a very different kind of theory. It is also formulated as a suite of explicit mathematical models. The theory generates quantitative predictions that can be, and have been tested with dynamical data (see our book with Sergey Nefedov, Secular Cycles).
The feedback loops in the Structural Demographic Theory
In Chapter 8, devoted to the ‘lemming cycles,’ Penman doesn’t cite our work, although he goes much over the same data for England and France. Our book Secular Cycles is cited in a long footnote on p.382. There Penman acknowledges our rigorous mathematical model and empirical time-series, but he criticizes us as follows.
However, there are a number of basic flaws with their model, starting with the obvious fact that Malthusian propositions do not hold in the state of industrial innovation. Their model oversimplifies the nature of human motivation, dealing only with the reality of sexual reproduction beyond the carrying capacity of the land. Biohistory provides a far richer understanding of human motivations based on actual experimental case studies.
This critique misrepresents our theory. Although we are happy to acknowledge our intellectual debt to Thomas Robert Malthus and David Ricardo, it is by no means a neo-Malthusian theory, because it pays much more attention to the social structure—the relations between the elites, the general population, and the states—than to demographics. And it translates surprisingly well to the industrializing and industrial societies (see my article, it’s also will be dealt in a forthcoming book).
‘Biohistory’, on the other hand, with its focus on individuals and their biological characteristics, completely misses out on the all-important social dynamics. Penman never addresses the biggest puzzle of human social evolution—how humans evolved the capacity to cooperate in huge groups of genetically unrelated individuals. And why do some societies cooperate much better than others? That is the key to understanding why some countries are wealthy and peaceful, and others are poor and convulsed by internal warfare.
I am sympathetic to the valiant efforts of those, like Jim Penman, who have the courage to propose radically new theories. But the science of human history has moved beyond the valiant attempts of brilliant individuals to propose new explanations. We are now in the less glamorous and more laborious phase, when new theories need to be translated into mathematical models, which can generate quantitative predictions that are subsequently tested with rich databases.
Very nice post, Peter, with clear exposition of the situation..quite convincing too.
All the best,
L.
Thanks, Lee! It’s good to be back to blogging, after an unintended time away.
It would be an interesting use of Seshat, to look for epigenetic effects. If a situation of high stress, or more plentiful food, or what-have-you had an epigenetic effect, it would take a generation or so to see the impact. Hopefully Seshat would have data that reflects the obvious potential causes of epigenetic changes, such as food scarcity, level of violence, etc. If you have other variables that can serve as proxies for pro-social behavior, impulse control, etc. then you could empirically test this kind of hypothesis, to see if there is an impact of one on the other with a 20-year lag.
One more reason to look forward to Seshat being available! When is that, again? 🙂
We certainly have a number of relevant variables already being collected in Seshat. But it will take a substantial effort to collect data on additional variables needed to test the Penman hypothesis. If we get the funds, we can do it. This may sound mercenary, but the way I think about it is it’s evolution. We write proposal addressing various Big Questions. Some are funded and others declined. Whatever proposals are funded are the ones with the greatest ‘fitness.’ 🙂
This is a very poor and self-serving review, if it can even be called a review, as there is almost nothing about the book, and more about the reviewer’s theory. It is hard to believe the reviewer read the book; it seems like Turchin jumped right to the chapter on lemming cycles without reading the first seven chapters. Why focus on lemmings, chapter 8, when the central tenets of the book’s theory are contained in the first four chapters? I have never encountered a review in which the basic theoretical starting points are ignored and a subsidiary, albeit important, chapter is taken as containing the essential elements of the theory.
I actually find that Penman’s theory teaches a lot about historical cycles, more than any current theory, leaving the lemming’s cycles out. What Jim Penman calls “C” and “V” behaviors are far superior in their appreciation of human motivation than what we have today; and what is interesting is that its arguments are in line with some of the observations about human behavior by past civilizational theorists such as Vico, Pareto, and others, except that Penman has a fully developed argument backed by some experimental research. Why did this review ignore the heavily documented chapter 2 on “food restriction” based on laboratory research?
I actually agree with Penman’s assessment of Turchin’s model, and the counter argument offered in this “review” reinforces my sense that Penman is correct that Turchin’s “model oversimplifies the nature of human motivation,” considering that Turchin says nothing about human motivation but thinks it is all a predetermined system-dynamic driven by Malthusian drives, or as Turchin calls the starting premise of his model, “general population”; so what if this general population has feedback loops with the state and elite; the question is what are the motivational elements driving human behavior?
I would like to see a unified simulation, that incorporates both theories, worldwide, on an acre-by-acre basis. I believe a societies attitude to debt and forgiveness, social mobility, etc has a huge impact. Jim correctly identified Trust as one of the big items of importance also.
Peter, the previous commentator is right and you really ought to read the first four chapters. The ability of people to cooperate in large groups is a direct consequences of epigenetic changes based on childrearing patterns such as control (especially in infancy) which makes possible more impersonal loyalties
The behaviour of people at each stage of the lemming cycle flows directly from the model. For example, the feudal chaos of the G-90 period (mid 15th century England or the Warlord Era of China in the 1920s) results from the lowest ebb of the child V factor (punishment in late childhood) which predisposes people to accept wider authority.
The key criterion for a scientific theory is that it leads to hypotheses that can be tested in the laboratory. My theory predicts that lemming cycles could be produced in the laboratory, and that each stage will have a distinct epigenetic pattern. We are currently ordering equipment from Israel which will allow us to do exactly this for a multi-male community of mice.
I appreciated ‘Secular Cycles’ and am pleased to see other people discussing these very interesting phenomena.
Hi Jim,
I agree with you that scientific theories need to be tested. But a laboratory experiment on mice is not going to convince (or most other people). Even if you could demonstrate the effect in the lab on humans, that would only suggest that punishment in late childhood could be a contributing factor to the cycle. But there are many other things going on, and testing your theory requires more than a lab experiment. In particular, is there historical evidence showing that the English punished their children less in mid-15th century?
And even that wouldn’t be enough. The ultimate test of a theory is its ability to predict novel data better than its rivals. This is what we do in the Seshat project. To test your theory properly its predictions should be compared with the predictions of, for example, the structural-demographic theory on historical data.
Let me know if you’d like to do it – I think that would be a better way to spend your money than buying lab equipment.
To Jim Penman: I read your book with great interest.
Having read “The Fate of Empires” by Sir John Glubb, may I recommend that you call lemming cycles “Glubb cycles”? Glubb’s book was all about the 300 year cycles you label lemming cycles. In a way, I see your work as a great expansion and development of his initial work.
Peter Turchin: I now have your book on my Amazon wishlist, and will buy it soon. I like your approach, since my background is computer simulations. I ask though, does your model take this into account:
Climate. Specifically, little ice ages, and the dramatic reduction in carrying capacity of the land.
Jim Penman and Peter Turchin: little ice ages causing hunger and forcing population reduction are historical facts; sure if your models include them, some of the anamolies in Jim Penman’s work will disappear. Tree Ring expert Mike Baillie has a good summary of the little ice ages over the past 3000 years in his book “New Light on the Black Death”
Now, this might cause problems with fund raising to include in your research, but there are vineyards today that don’t produce; but 1000 years ago the climate was warmer, and they did produce. Today, there are farms in Britain that are too cold to put to work, but were formerly productive. So, the sliding effect of climate on carrying capacity of land, if you can incorporate that, I’d be very interested to see the results.
Jim Penman: I’m also wondering if your work incorporates (yet) the behavioral sink research of John Calhoun, of “Rat Utopia” fame. Effect of overcrowding, etc.
Also, from your book I couldn’t tell if you feel America is in 100BC, or is Byzantium 500AD.
Emperor Augustus may have tried to make moral reforms, but the stories of his adulteries (and the even worse adulteries of his daughter Julia) indicate the moral decay was well advanced by his time. When I reread Seutonius this year, I was shocked at what a horrible man Augustus Caesar was. Violating men’s wives right at the dinner table. Performing mass human sacrifices with his own two hands in honor of Julius Caesar. Rome wasn’t so different from the Aztecs and Mayans.
Peter, my background is simulations, and I’ve done my own population simulations with startling results. Would you be interested in collaborating on a simulation that breaks the earth down on an acre by acre basis, so as to model changing climate and population, incorporating travel distance and eventually the effect of improved technology. Ie, better logging technology accelerates rate of deforestation; less forest means faster travel. Mountains block travel; waterways both block and enable travel, depending on capital outlay.
I think eventually such a simulation would be valuable by hiliting potentially underutilized and undervalued pieces of land. Simulation would have to include the cost of improving the land to increase its fertility, livability. This would allow a Holland scenario to evolve naturally. And increasing technology plus increasing size of government would make land reclamation more practical and cheaper.
Mycroft, thank you for the offer. However, currently I am completely occupied with steering data collection and analysis within the Seshat project. I am planning to return to modeling in 2016. Send me an e-mail next Spring, if you are still interested at that time.