An evolutionary theory of history: science fiction and scientific fact in the real world

Author: Michael Mernagh

What are the best questions to ask a homo sapiens statisticus who analyses history? Does God play dice in the fourth dimension? How can we measure the uncertainty of events that channel the course of history? How are the affairs of humanity inter-connected between local, national and global levels down through history? Does an evolutionary factor play a part in that process?

A decreasing exponential mathematical model fits the sequences of major events in the histories of western Europe, Ireland and County Wexford.

Diagram 1 (a). Actual timespans of the major historical eras in western European history from 753BC to 1492AD

Diagram 1 (b). Actual timespans of the major historical eras in western European history
from 1492AD to 1916AD

The literary works of Isaac Asimov

Mathematical physicists enjoy the stories by Isaac Asimov (1920–1992). His Foundation novels are classics of the science fiction genre. Asimov suggested a macro-statistical approach to history in order to predict the future. He envisaged a universal model based on the theory of gases. The molecules that make up a gas move randomly in three dimensions at a variety of speeds. Nevertheless, one could fairly describe what those motions would be on average and thus infer the gas laws.

Asimov applied that concept to human beings. Each individual has a free will, but large populations behave quite predictably. The probabilistic analysis of mob behaviour was his theory of psychohistory. Three conditions would be required:

a. He invoked an enormously large population. His galactic empire would have a total population of 100 quadrillion people.

b. He retained randomness. Individuals might not behave randomly if they knew what was expected of them.

c. Just as gases consist solely of molecules, psychohistory would work only if all intelligent inhabitants in the Milky Way galaxy were either humans or robots subject to humans. Any genetic mutation would alter Asimov’s conceptualised design.

Conclusion

In the past, economic periods of boom and bust usually coincided with episodes of war and peace, respectively. The decreasing exponential model charts the general trend of those eddies in the onward swirls of the stream of time. The tides of major eras heave in the ocean of world history. They can have knock-on effects that can generate turning points in local history.

The year 1859 was an important date in modern history. Darwin published his Origin of Species in that year. It has its place on the graph of the general model of history.

The graphs converge rapidly towards the baseline. This suggests that the structures of human society are evolving towards ultimate equilibrium. The progression of history does not decay into unpredictable chaos. Human society continuously evolves from imperfect conditions towards more humane social structures. A better era of social justice has usually followed the preceding era.

A statistician’s best answers to questions about history might be:

• at the micro-level of individuals, the future is unpredictable; but
• at the macro-level of large populations, events are more certain to turn out as expected.

Statistical analysis reveals that an evolutionary dynamic underlies the general trend of history.

Diagram 2.1 The theory holds well on the macro-level of European history1

Diagram 2.2 The history of Ireland2 was similar to that of western Europe

Diagram 2.3 The general model also fits the local history of Wexford3 quite well

Diagram 3.1 The adjusted model of historical events in western Europe

Diagram 3.2. Composite graph of key historical dates

References

  • 1. P. Van Ness Myers, 1923, A Short History of Medieval and Modern Times. Ginn and Company.
  • 2. J. Carty, 1964, A Class-Book of Irish History, Macmillan and Co. Ltd.
  • 3. The Past, Journal of the Uí Cinsealaigh Historical Society of County Wexford.
Bookmark and Share

Comment on this article

Submit your comment
  1. Image of unique ID

Comments

Rumba

What a neat ariclte. I had no inkling.

reply to this comment

Michael Mernagh

Thank you, Antonio.

I tested the null hypothesis that the truncated data were uniformly distributed. I rejected that null hypothesis.

However, I also tested to see if the truncated data were decreasing linearly. I accepted that null hypothesis.

Interestingly, I then used y = (1−0.2811)^x to model a cumulative loss of memory (= y) at a constant rate of 28.11% from x = 0 to 10 by 1. The resulting model is identical with y = exp(−0.33x) from x = 0 to 10 by 1.

Best regards,

Michael.

reply to this comment

Antonio

Take off the few oldiest points and carry out a test to see if the other ones come from the uniform distribution. All my best wishes for your health.

reply to this comment

Antonio

Joke or junk? How have I to consider such a totally non-sense article? It has not been published on April 1, so apparently I have no reason to think of it as a joke... Besides all the remarks that Jordi has already pointed out, I have to note that according to the proposed model God did create Europe on 2762BC (or at least, he made the European history started on that date) and did destroy it on 2009AC (or at least, he did make the European History ending that year). But maybe the author will claim that his model is True apart an oppurtune shift of both the axis. In the meantime, I'm waiting for the end of the world in 2012.

reply to this comment

Michael Mernagh

Excuse me while I transmutate my dermal layers into rhinoceros hide. I think I now know how Galileo must have felt.

First steps are always tentative in sowing the seeds of a new idea. “From tiny acorns, mighty oak trees grow.”  Breaking new ground always carries the risk of being misunderstood. True scientific method must be open to original thought.

I first conjectured that the main eras of history tended to get longer as one looked further back in time. I then selected events and dates to represent the beginnings of the main historical eras. I sampled some other dates.  The model seemed quite robust.

As the house rules of this website set an upper bound of 800 words on articles, I was limited as to how detailed I could expand on my hypothesis. My health is not good, so I do not have the strength to write that book.

Does time accelerate as time elapses? The sample of 2,762 years of recorded history is not long enough to test that idea in the context of the age of the universe (which is about 14 billion years, give or take a margin of error). Applied cosmological physics might answer the question.

Philosophically, is eternity a frozen state of time standing still, or is it a dynamic state of infinitely accelerating time? Time in the universe as we know it is somewhere in between.

reply to this comment

Tom Fanshawe

Jordi, I'm impressed with your restraint in making your comments as I find the entire premise of this article to be more than a little absurd.  The so-called 'evolutionary dynamic' is entirely an artefact of the particular events chosen, and the choice appears at best arbitrary and at worst biased towards more recent events in an attempt to justify the theory.

In addition, and as Jordi also points out, even if the choice of events can be justified objectively, the obvious conclusion is towards a trend of more turbulence, not less, completely invalidating the claim that society is obtaining some kind of idealised 'ultimate equilibrium'.

reply to this comment

Michael Mernagh

Is the “evolutionary model” of history simply a particular case of a more general property underlying the distribution of numbers in relation to their rank order?

Here are two examples.

Firstly, the Benford distribution predicts the proportional frequency of the leading digits of numbers. Using logs with base 10, the expected proportion (P) of numbers beginning with any digit from 1 to 9 (inclusive) is P = log[(n+1)/n], where 0 < n < 10 is the numeric rank of the leading number.

Secondly, when the populations of counties in the Republic of Ireland (as per this year’s Census) are sorted into descending numerical order (y) and plotted against their ranks (x), the model  y = {1,127,616x −1.77 } + {151,651 − 4,170x}  fits the data almost exactly.

Michael Mernagh.

reply to this comment

Michael Mernagh

Quote:

I think that the question whether History -and Natural History- has a direction is an intriguing one and I thank Mr. Mernagh for the article and the reflections it gives birth. However, I have a substantial methodological question regarding the data presented here. I wonder how major historical events are defined. For example, let's examine European history. Shouldn't we include in the analysis the Punic Wars that gave preeminence to Rome over the rival Cartaghe? Or under which criterion is French Revolution left out? Also, if we include the publication of The Origin of Species by Darwin, why don't we also include the publication of Capital by Marx? These are just some examples, but I think that the way you define major historical events might change the observed patterns and so the definition should be stated explicitly.

Then, I do not quite agree with the conclusions. In my opinion, the data doesn't show that human societies are evolving towards equilibrium. In fact, equilibrium is usually associated with an absence of change and stability. What the data shows is that as history goes on major events occur more frequently. That is, historical changes occur more often and history becomes more volatile, dynamic, unstable...

And finally, it could all be an artefact due to the increasing lack of information for more remote historical periods or to a lack of perspective. In the first case, note that we have much more data (and a higher number of potential major events) for the last century than we have for the whole Roman Empire. And regarding perspective, it helps putting events in their place in history. For example, failing your Probabilities exam at college might be a tragedy when it happens, but many years later, comparing it with other events such as losing your job or a dear one, you may think it was not really that important. Many recent events that are now judged to be major, might be downgraded with the passage of time.

Anyway, I still find the idea of analysing history with statistical tools worthy, but also extremly difficult.

 

Thank you, Jordi, for your feedback.

The exponential model of history paints a generalised picture on a broad canvas with very large brush strokes.

The conclusion that that the structures of human society are evolving towards ultimate equilibrium was inferred from the asymptotic flattening of the decreasing exponential curve towards the baseline.

On the matter of subjective selection of historical events, the proposed model seems to be quite robust to the inclusion of extra events. (Communism was not included in the model of Western European history.)

The model illustrates how events on the international stage can have knock-on effects at a local level.

“History repeats” is a popular cliché. Why, then, do people never seem to learn any lessons from it? Suitable models of how history continuously evolves may help to find an answer to that question.

Mathematically, the general negative exponential distribution is notorious for its “lack of memory”. In the real world, people do not remember old events as well as more recent. It could be useful if theoretical statisticians derived a distribution whose memory of more distant events decreases as one goes back in time. In other words, what is the best model of memory?

There is always some variability in history. By minimising that variability, we may get a better vision of where we are going and how we may best control our destiny. On a lighter note, if we could apply the model to the financial world, then we would all be rich.

My mind is a bit like history. I am getting old and my memory is failing.

Best regards,

Michael.

reply to this comment

Jordi

I think that the question whether History -and Natural History- has a direction is an intriguing one and I thank Mr. Mernagh for the article and the reflections it gives birth. However, I have a substantial methodological question regarding the data presented here. I wonder how major historical events are defined. For example, let's examine European history. Shouldn't we include in the analysis the Punic Wars that gave preeminence to Rome over the rival Cartaghe? Or under which criterion is French Revolution left out? Also, if we include the publication of The Origin of Species by Darwin, why don't we also include the publication of Capital by Marx? These are just some examples, but I think that the way you define major historical events might change the observed patterns and so the definition should be stated explicitly.

Then, I do not quite agree with the conclusions. In my opinion, the data doesn't show that human societies are evolving towards equilibrium. In fact, equilibrium is usually associated with an absence of change and stability. What the data shows is that as history goes on major events occur more frequently. That is, historical changes occur more often and history becomes more volatile, dynamic, unstable...

And finally, it could all be an artefact due to the increasing lack of information for more remote historical periods or to a lack of perspective. In the first case, note that we have much more data (and a higher number of potential major events) for the last century than we have for the whole Roman Empire. And regarding perspective, it helps putting events in their place in history. For example, failing your Probabilities exam at college might be a tragedy when it happens, but many years later, comparing it with other events such as losing your job or a dear one, you may think it was not really that important. Many recent events that are now judged to be major, might be downgraded with the passage of time.

Anyway, I still find the idea of analysing history with statistical tools worthy, but also extremly difficult.

reply to this comment

Skip to Main Site Navigation / Login

Site Search Form

Site Search