Why this is so seems to have much to do with the instinctive behaviour of humans under pressure: do what used to work before, but do it harder. Since organisations change only in response to pressure, real change necessarily involves noticing and questioning these instinctive reactions. It is far easier to duck the challenge, even at the price of failure.

The instinct to hold on to the known is hardly surprising. However commercially dangerous not changing might be, it feels safer - from the individual's point of view - than moving off down a new behavioural path. The nervousness is understandable because nobody in the organisation has yet been down the new path, and that knowledge adds to the inner tensions.

The image that best captures the problem is walking. In order to walk, we actually start by toppling forward; only then do we put a foot beneath the body's new centre of gravity (it is easier to notice the process if you try walking backwards). The point is that we cannot move at all unless we are willing to accept losing our balance at least temporarily.

So attempts to remove in advance all the uncertainties of change - by planning, proceeding only by consensus, or any other means - have the effect of slowing the process dramatically or stopping it altogether. Ironically, that prolongs and magnifies the tensions which make change so difficult in the first place.

As the economist JK Galbraith once pointed out: "Faced with the choice between changing one's mind and proving there is no need to do so, almost everyone gets busy on the proof" - usually without noticing what has driven the choice.

The phenomenon is apparent across all organisations and at all levels. The chief executive of a national supermarket chain, for instance, made his name and won his promotion in the '80s as a hard-nosed cash controller. So much so that all spending, down to expense-account claims for less than £10, would wind up on his and other directors' desks for final approval - even though they already had signatures from five or six lesser managers.

As a way of signalling an end to earlier sloppiness, the controls worked splendidly. But in the early '90s turnover began to fall off, and the need for greater productivity and local flexibility became urgent. In these new circumstances the controls worked only to slow decision making, sometimes by months - damaging the organisation's morale and its ability to put local sales-boosting initiatives into place. Nevertheless, the chief executive's response was to set up still more central controls. Why? Because he was following unconsciously the instincts that had worked for him before.

In professional firms, the same patterns are apparent. Look, for instance, at marketing habits. In the late '80s, consultants - like accountants and lawyers - produced lots of brochures, and lots of work flowed in. No firm did any serious testing of whether there was any connection between the two events, or which compared the results from brochures with the results from other approaches. But the belief took hold that brochures work. So what are most firms doing, now that competition for work is tighter? More brochures.

Two examples may help to show how insidious can be the effects of following old instincts unquestioningly.

• A partner in a large firm was put in charge of an under-performing regional office with a brief to turn it from a "coaster" into a "racer". He is clear that the local partners need to sell more (most of their work is referred from London) and to act as a team rather than as sole practitioners (there is little cross-selling and less trust). That must mean leading by example and launching new approaches - ideally several at once to improve the odds of success. But his personal instinct is to avoid risk and to find "the right answer" before acting. So, apart from exhorting the partners to change - which has not worked - he is still looking ... a year after moving in.
• The chief of a medium sized practice nursed it through a time when its survival was in doubt. It is now growing, albeit slowly and the chief wants to let go of the daily management burden. He has brought in new heavyweight practitioners for that purpose and urged them to get on with it. But his unnoticed instinct is to keep tight personal control of everything -because the firm is "his" baby - and to undermine anyone who acts independently. The result: he has yet to build an effective management group .... after five years of trying.

Leaders of any business have to fight hard to spot and break free of such instincts, because they have few peers to lean on for help, and because, if a leaders falters, his subordinates will run for cover. But leaders of professional practices face three special difficulties:

First, their fellow partners are better placed than outside shareholders to block change. And, like their leader, they will instinctively want to repeat the behaviour patterns that got them to partnership, whether or not those patterns are still appropriate.

Second, because professionals are among the brightest people in the country, the arguments they throw up in defence of their instincts are likely to be highly sophisticated. The arguments may be only rationalisations, masking a deeper uncertainty, but that won't stop them seeming plausible.

Third, knowledge-based businesses largely depend for their success on their critical and analytical skills. For that reason, they tend to apply the same faculties internally. That can mean the organisation has an instinct to see negatives and weaknesses rather than strengths, and a preference for blame rather than praise. These tendencies cannot and should not be eradicated; they have a valuable restraining function. But they do increase the inertia of the organisation's culture.

The effect of all three difficulties is to create pressure on practice leaders to adopt an incremental approach to any change, or to delay action until they have cast-iron assurance that they can achieve precisely the change they want.

The trouble is that, in the nature of things, nobody knows for certain what will work until after it has been pilot-tested. Nor is it possible to predict all the results of change in advance; the road ahead becomes clearer only when you are moving along it. In addition, an incremental approach runs a greater risk of leaving more openings for resistance ("if the boss isn't sure, why should I be?") or delay ("I don't need to make any real changes yet").

Arthur Radford, an American admiral turned management guru, defined the dilemma this way:

"A decision is the action an executive must take when he has information so incomplete that the answer does not suggest itself". Or, as the French author Andre Gide wrote: "One doesn't discover new lands without consenting to lose sight of the shore for a very long time."

The only armour that can withstand the external pressures and uncertainties is internal: the leader's own conviction in the necessity for change, along with a willingness to start moving anyway, and the confidence to handle unforeseen problems as they arise. And the fire which forges such armour is the personal ordeal of breaking through the inner fears which drive instinctive reactions.

The ordeal is never easy, but it is necessary. Until it happens, nothing much else will.

In the words of the German author and poet Goethe, a professional himself (he trained as a lawyer): "Whatever you can do, or dream you can, begin it/Boldness has genius, power and magic in it".