中国A片

Things fall apart

Death and taxes: even nature is in on the act. Omar Malik discusses entropy's effects on human organisations ... and what can be done to combat the constant risk of system failure

七月 1, 2010

Philosophers have long pondered the big question: why is life such a bugger? Lesser orders have similarly wondered why life is just one damn thing after another. Why do things always go wrong? The answer is short and simple: the laws of nature.

Some like to believe that we are the highest form of life, blessed with free will. Maybe. But as far as nature is concerned, we are just another of her countless products and, like the rest of them, serve sentence under her laws.

Francis Bacon said that we cannot command nature except by obeying her; sadly, he omitted to say that to obey her, we must first understand her. In macro terms that is surprisingly easy: all we have to do is identify her laws. The micro task of combating them is much more difficult, well beyond the scope of libraries of regulations, however vast.

However, the guiding principle is clear. The imperative is to fight, to strive to win every battle, undaunted by the knowledge that in the end nature will triumph and humanity will lose. The key motivation is the knowledge that any battle can be won. Lose sight of that and you lose the battle, and perhaps the war.

The laws of nature are the laws of battle. Some, perhaps all, are against us, because nature writes the laws for her benefit, not ours. Her first law is that we must obey her. Her next two are tax laws. We note that nature thus provides the blueprint for many hierarchies.

Nature's two tax laws are our subject. They are important. They tax our endeavours, top-slicing all that we do, thereby weighting the scales to nature's advantage. Any activity that we undertake can be viewed as a system of operation, and any such system - your office, your department, any large organisation, the nation - is at heart energy organised to perform work. In performing that work, some energy is always lost by the system. It is not lost to the world, because energy cannot be created or destroyed, only converted into other forms of energy.

As your car motors along, heat energy is lost through the exhaust and the flexing of the tyres. In an office, energy is lost when staff walk to the photocopier, read impenetrable documents, pursue pointless management edicts and shelter from the incontinence of the ever-active regulatory volcano. In each and every action, some energy is lost by the system.

Such frictional losses are the first of the costs of converting energy into work. The bigger the system, the more difficult it is to identify and address these losses.

Some commercial enterprises are big organisations. Their managers are often competent, experienced and do their best: some succeed, many do not. Nations are very big organisations. Their managers - the politicians - often have little idea of the source or extent of the frictional losses in the operating system. Few seriously try to understand them. Very few have the necessary managerial competence and experience to reduce them. Little wonder that many nations fail to balance their books. Frictional losses of energy are nature's first tax, Tax #1, a Tax On Trying. Worse is to come.

Tax #2 is a little more complicated. It is not a tax on doing work, but on the system itself. The energy in any system - your office, your department, your nation - will seek to escape from its control. In so far as it succeeds, it will remain within the system, but no longer harnessed to the system's ends.

The simplest example of this is the behaviour of the individuals employed by the system. Each will bring their personal agenda to work. These agendas will almost never be precisely congruent with that of the system.

In any decision, the first concern of many involved will be to safeguard their jobs; therefore, they may studiously avoid taking any responsibility, positive decision or action in case it proves contentious. They will be mindful of their other interests, of how they wish to pass the time, of the need to please their boss, of the need to bring work to their department, and of their conviction that their approach is the right one. To a patient, surgeons will recommend surgery, pharmacologists will recommend drugs, psychiatrists will recommend psychotherapy, personal trainers will recommend exercise, epicureans will recommend red wine, steak and a double bed.

Beyond normal vested interests, there may be unworthy interests: the discomfiture of a rival, the misappropriation of wealth or the service of an undeclared, different master. All constitute energies that are not harnessed to the system's purpose.

This phenomenon is common to all systems, however tightly controlled. In some situations it may be minimal, as in an armed forces unit in contact with the enemy. In others, it may be so great that it seriously prejudices the performance of the system, as in the active obstructiveness of the disenchanted, the self-serving of the undisciplined, and in the kleptocracy or promotion contortions of the highest ranks.

Unlike Tax #1, the energy lost to Tax #2 does not leave the system, but does its own thing. Tax #2 is the Tax On Systems.

Taxes #1 and #2 have the same effect: they remove energy from the system's effective output. They degrade the system.

Thermodynamicists and others will have correctly recognised that Tax #2 is a manifestation of entropy (energy not subordinate to the system). Unfortunately, most non-thermodynamicists feel the need for a deep breath and a strong coffee before embarking on any discussion of this elusive concept. For our purposes, all we need to know about entropy is that it is a measure of Tax #2. The concept of entropy has been applied to physics, chemistry, biochemistry, information theory, computing, cosmology and doubtless elsewhere. These applications need not trouble us. We are concerned with the application of the concept to operating systems in general.

To recap, there is a simple distinction between Tax #1 and Tax #2. Tax #1 is the energy lost in doing work for the system; Tax #2 is the energy lost in doing its own thing. Both are unavoidable.

For system managers seeking to minimise the taxes, Tax #2 is a much more difficult problem. It arises from system components arranging themselves to their own advantage - and to the disadvantage of the system.

To return to our earlier example, there is an inherent tension between any system and the individuals it employs. This is most easily recognised in large, badly managed organisations. Remember that we are discussing energies that are not deployed to achieve the system's design output.

The first symptom is the absence of control: the human components are doing their own thing and getting away with it. Look for the absence of discipline: managers unable or unwilling to confine subordinates to their primary duties; subordinates contemptuous of or over-familiar with managers. Often this is accompanied by much jollity and overt self-satisfaction of the employees with their performance and their easy life.

The second symptom is poor system performance. In commercial organisations, overall performance is relatively easy to measure and its deterioration ends with bankruptcy. In state organisations, performance is neither technically nor socially easy to measure. As with states themselves, there does not appear to be an intrinsic limit on its deterioration. We can see why. State organisations (by common observation: no names, no angry letters from the NHS or the local council ... oops) tend to be high in entropy. High entropy is likely to reflect the comfort of the components of the system, its employees; they may be expected to be resistant to the measurement of entropy and positively active in their opposition to its reduction (Greece's economic meltdown is an obvious example).

The same is true to some extent in all organisations. It is a knotty problem. Managers seeking to reduce entropy need the stomach for the fight. Unless their own entropy is under control, they will opt for an easier life.

System managers have three tools available to them to tackle entropy. One is arguably strategic, because it addresses the root of the problem and may be decisive for as long as it lasts. The other two are tactical; essential to the battle but not decisive in the war.

The first tool is employee motivation. If they can be persuaded to commit themselves to the system, employees will do all that is possible to direct their energies to its ends and to prevent their unproductive dissipation. In the short term, it does not matter how the leadership achieves this commitment, whether by sweet reason or by the BIFSS (Because I Say So - the F is silent in polite society) style of management. BIFSS may lead to violent revolution and is therefore not a good long-term option.

The second tool is the terms and conditions by which the system's employees work. Again, there is an inherent tension between the perceived needs of the system and those of its employees. At one extreme is the rapacious employer who treats staff as assets to be stripped and discarded. At the other is the old-fashioned trade unionist who treats the employer as a mine to be excavated to the last nugget. The manager has to strike a workable balance between these extremes.

The third tool is a valid, accurate means of measuring employees' contributions to the system's output. This tool is subject to the fourth law of nature, which states that every component of any system will always seek to exploit the system to its own advantage. Thus, new performance measures displace measures of system output as active objectives. Hence the need to find performance measurements congruent with system objectives.

The recent history of too many organisations in this country reflects counterproductive performance measures. Some are simply ill-conceived, being either easily manipulated or providing no real measure of system output. Others are designed by those who are being measured, considerably to their advantage, with negative effects on the system - and in the case of the financial sector, disastrous consequences for everyone else.

Such counterproductive measures offer a good example of the malign effect on a system of a boost to its entropy. They offer another illustration of the fourth law of nature: anyone left unsupervised or free of sanction will utilise the situation to their own advantage.

There are two creatures in the system managers' room. Both were nice when they were small and deserving of nurture. Both are now too big, and no longer nice, and need to be tamed.

The first is the elephant of entitlement. Fed on a sustained diet of easy affluence or state munificence, and a creaking dessert trolley of political correctness, this obese creature will accept no hardship or criticism. It has long forgotten that there is a direct connection between productive work and eating. The system manager has to re-educate this elephant and restore the lost connection. Otherwise it will see no reason to lend its weight to furthering the system's ends.

The second creature is the great hippopotamus of regulation. It consumes everything that comes within reach of its capacious jaws. It excretes red tape as widely as it can with each flap of its tail. It reaches into every corner of the system: ridiculous, static micromanagement of not only every dynamic detail, but also of situations unknown and unknowable. These regulations, produced by a distant bureaucracy, govern the ability of the system manager, the person on the spot, to manage the system.

Regulations that govern details that cannot be seen from a distance in space or time are certain to be seriously counterproductive. They add to the burden of both Tax #1 and Tax #2. Any regulation of system output becomes a component of the system, absorbing or diverting energy from its output. The greater the degree of national regulation, the lower the productive output: that is all the government needs to understand. Happily, it does not need to understand entropy.

Both taxes are universal. They apply to each and every system and they degrade output. They are the means by which nature makes 100 per cent performance an impossibility. We must understand that this is a fact of life.

We must bestir ourselves to minimise the malign effects of these taxes. This we can do by applying ourselves to the problem of system degradation and by exerting ourselves to prevent the possibility of system failure.

The classic theory of system failure is American organisational theorist and sociologist Charles Perrow's Normal Accident Theory. It propounds that accidents are normal in tightly coupled complex systems. "Tightly coupled" means that if one bit fails, it takes other bits with it. "Complex" means that all possible interactions in the system, and thus the ramifications of all possible failures, are incomprehensible and cannot be foreseen. We can see why Perrow says that accidents are normal in such systems.

My Revisionist Normal Accident Theory goes further than Perrow's. It propounds that accidents are normal in all systems. There is good evidence for this. One key point is that every safety measure brings with it a new source of risk. Furthermore, as we can now see, nature's two taxes remove energy from the system's designed output. Therefore, no system can run perfectly. Failure is thus the default mode of all systems, the norm imposed by nature. In every system, accidents are waiting to happen. For that reason, we must exert ourselves and fight every battle against failure. The proper study of mankind is the prevention of system failures.

And finally, we have proof of God's existence. Those who postulate a God will prefer the term "Laws of God" to laws of nature. We have seen that His laws impose taxes. We cannot doubt taxes: they are one of two certainties in life (alongside death). Therefore, we cannot doubt the existence of the writer of the Universal Tax Laws. Over to you, Richard Dawkins.

请先注册再继续

为何要注册?

  • 注册是免费的,而且十分便捷
  • 注册成功后,您每月可免费阅读3篇文章
  • 订阅我们的邮件
注册
Please 登录 or 注册 to read this article.
ADVERTISEMENT