During the inquiry into the exams meltdown in 2000, the members of the Scottish Parliament's education committee were given a guided tour of the Scottish Qualifications Authority's Dalkeith headquarters, myself among them. I well remember being stunned at the size of the premises and the acres of metal shelving. I was even more stunned to be witness to a search for a missing script. Boxes were moved, racking levered from the wall and staff got down on their hands and knees. When eventually the item was located, great relief ensued.
What worried me most was the fact that the poor student whose life blood had gone in to that piece of paper would never have known if it had been lost beyond recovery. No doubt some method would have been found of assessing his grade from previous work and the whole incident swept, with the lost script, under the carpet for ever.
This is no way to run an exam system. Douglas Osler, in his contribution on this subject on this page last week, claimed that "there is nothing wrong with Scottish education that can't be cured by what is right with it". In the generality, that is true - Scottish education has a lot going for it, not least the dedication and professionalism of teachers. But there are things that are inefficient and which need new thinking. The exam system is one of them.
What is a national exam system for? In part, it is a means of measuring attainment and ability, and the interplay between the two. In part, it assesses the effectiveness of our education system and those who work in it. And in part it provides a bedrock of certainty about the level of skill and learning on which we can build national success.
The problem with this tripartite obligation on our exams and those who administer them is that the requirements do not sit easily with each other.
The more attempts are made to ensure that all three are met, the more bureaucratic and cumbersome becomes the entire edifice.
An example can be found in the recent passion for primary testing. Teachers have always assessed the progress of their own pupils, and usually done it effectively. However, when political fears about the overall quality of education raised their head, the political solution was to create a structure in which the performance of pupils, teachers and schools was required to be tested at the same time.
The result was unsatisfactory for all involved: teachers began to "teach to the test" to secure results that met national and local benchmarks. These benchmarks were constantly upgraded, until it became obvious - as it should have been from the start - that the entire system was fundamentally flawed.
There is no such thing as perpetual improvement - at least not in exams.
There is such a thing as perpetual effort to ensure that young people aspire towards the best they can do, and are given the help and encouragement to achieve that goal.
Consequently, the assessment of the effectiveness or otherwise of teaching in any school, let alone the progress of its pupils, should be dependent on a number of indicators, some of which have to be subjective - the satisfaction of parents, the enjoyment of the experience by students and the morale of teaching staff are three that spring to mind. The number and quality of exam passes are only part of that mix, and a part that has to be considered alongside the others, not as the only significant issue. Taking that pressure out of the exam system would be desirable. Then a real effort can be made to make the system as effective as possible for its primary purposes - the purposes that serve the individual student and society as a whole.
The development of Standard grade and Higher Still courses was an attempt to do just that - to set a rising path, understood nationwide, on which progress could be marked and which could raise attainment by degrees.
Unfortunately, that now looks almost as old fashioned as its predecessor and is in need of some refreshment. Standard grade can be achieved earlier than in the fourth year, and the wider use of Intermediate studies as building blocks towards Highers needs to be encouraged. Presently in many schools it is just an impediment, or a sop to those who are not doing well enough.
The use of personal learning plans (PLPs) - no matter the huge resource implications they have - should provide the context for this process as well as provide routes to noted success in other forms of education, particularly for those who will never find the formal structures appealing or appropriate. The individual plan, and its individual completion, would be the best indicator of all of personal achievement, within which exam passes would simply be part of the whole. PLPs within a national framework and given national status, would be a better passport to both tertiary education and the world of work and we should, as a nation, be considering how we make such a system work.
We should also be considering how we apply technology to the process. It is almost unbelievable that, in this day and age, vast volumes of paper are shifted by post from place to place and stacked on creaking shelves in between times just in order to provide assessments of individuals. The proposals for taking exams online need to be fleshed out and speeded up, not delayed by the inevitable Scottish fear of change.
If education is to succeed, it must affect each child and must develop them as individuals in ways that are most effective and appropriate for them. To make the need for national comparisons, and the need to "guard the guards", more important than that task is not only to misunderstand the purpose of learning: it is also to misunderstand the needs of a flexible, innovative and individualistic modern society.
Michael Russell is a writer and broadcaster.