Dr Isaac Asimov, head-and-shoulders portrait, facing slightly right, 1965 (Photo credit: Wikipedia)
Toss a coin into the air. What are the chances it lands heads-up?
The conventional answer is 1 in 2.
The convention is wrong.
The conventional answer simply reflects what we can easily calculate: that if the coin is perfectly balanced, and there are no other environmental factors affecting the toss, then if you toss it a very large number times, the average chance of it landing heads-up will be once in every two tosses. An equivalent way to calculate this is to take a very large number of identical, balanced coins and toss each once: half would land heads-up.
However, it is exceptionally difficult to calculate the actual outcome of a single coin toss with any degree of accuracy. Certainly, it is a virtual impossibility before the coin is in the air and only slightly more feasible (at least theoretically) once the coin is in motion. That theoretical feasibility would likely involve cameras/sensors able to track the movement of the coin in the air, and a sufficiently fast computer simulation of the physics involved to be able to project an outcome an instant before the coin lands.
What does this little thought experiment tell us about probability, and about making predictions?
- Tossing a single coin thousands of times yields equivalent predictive power to tossing thousands of coins simultaneously, assuming the coins in the second experiment are sufficiently similar (homogenous) to not bias the results.
- Even once a suitably mathematically rigorous model is developed to predict the average behaviour of an entity (or of a sufficiently homogenous population), predicting what actually happens becomes harder and more complicated the narrower the timeframe and the smaller the unit being observed.
While it may seem that I am belabouring very simple mathematical & scientific principles, these two points are highly counter-intuitive to most attempts we make to predict human behaviour. The internal model of people that we tend to adopt is that individual members of society (including ourselves) are unique, and that extensive study of their singular natures is required to predict how they might act. Even then, we assume that our predictions will only hold in the short-term where fewer variables have time to interfere. An expression of what happens in the mind when these incorrect beliefs coalesce is the vanity of the fundamental attribution error, wherein our own motives seem rational & pure and those of others irrational & base.
In fact, people are both simpler and more similar than we like to believe. The reason for this is that we apply conscious thought to our actions in a much more limited fashion than is generally accepted. This should not be controversial: just try remembering the exact details of your last commute to work, and then tell me you are were acting in a fully conscious manner at the time! And yet, most people rail against such an interpretation of their lives.
It does however have advantages. Unconscious actions tend to be more automatic, and therefore more predictable, than conscious ones. Give people time and opportunity to actively choose a course of action, and they can react in surprising ways. But keep them moving on a path that requires relatively less choice, and their potential range of action diminishes. This theory of behaviour becomes extremely useful when thinking about how large groups of people may act. The complexity of societal structure itself is an exceptionally good narcotic agent on the conscious mind: it funnels and moulds our actions in a deep manner, subtly constraining our choices at every step while still providing the illusion of choice.
The beauty of this effect is that it leads to a surprising conclusion: societies – and the people within them – are MORE predictable the larger and more complex they are. Simple small societies bring us closer to the level of a free individual, and so create more opportunity for a single rationally-thinking person to make changes. Large, intertwined societies tend to blur the effect of a single person, reducing the noise affecting the signal indicating the direction of travel of the society.
Irrationality thus becomes an asset to predictability, and rationality a liability. Situations are harder – not easier – to predict when too many people in a group actively attempt to cognitively reason out their next course of action. By constrast, irrationality tends to be the product of the unconscious and automatic mind, and that is much more straightforward to anticipate. The irrational action may or may not actually be helpful in achieving the group’s aim, but it is almost certainly more predictable. Any student of the human mind will confirm this pattern, whether they work as psychiatrist, psychologist, politician, salesperson, negotiator or diplomat.
Of course, I do accept that if absolutely everyone acted rationally all the time, society would be perfectly predictable. This is why the behaviour of an algorithmic computer is theoretically predictable. The difficulty is where both rationality and irrationality co-exist i.e. the human context. Here, it is the presence of rationality that creates a more disruptive effect on predictive power than irrationality. The larger and more complex the society, the less impact a small amount of rationality can have. Trends, conventions and accepted boundaries, however arbitrary, are more effective constraints of behaviour than rational argument ever can be.
Psychohistory is the fictional science created by Isaac Asimov for his Foundation series, which postulates a mathematical predictive model of society. Today, the concept is being explored (albeit in rudimentary fashion compared to Asimov’s fantastical version) through the work of a range of behavioural modellers in many disciplines, though few would apply the term psychohistory to what they do or indeed even realise that their work may in future apply to such a field. The few who consciously strive towards such a distant grand ambition have renamed themselves students of Cliodynamics. Psychohistory as a term survives instead as the study of historical and sociological trends through the lens of psychodynamic principles.
The fictional science of psychohistory, or modern cliodynamics, faces massive practical challenges. Fortunately, two of them are being gradually solved simply by the passage of time: firstly, our ability to computer simulate complex models grows exponentially with increases in processing power; and secondly, our societies become ever more complex and intertwined, limiting the impact of disruptive, rational, behaviour.
While the idea of a fully developed psychohistory of society may induce feelings of anomie or nihilism, it should be emphasised that it does not negate the individual. Individuals will still exist, exercising both rational and irrational thought just as they always have. And they’ll still have all the same illusions around choice and freedom that they always have. It would be exceptionally hard to eradicate such a useful evolutionary trait. Just like today, each of us will still prefer to believe we can control the world, while minimising the degree to which it influences us. And just like today, it will be those with a strong sense of their own identity who will be most able to adopt quiet & discreet lives where such influence is indeed minimised.
