A razor of disagreement
We are all familiar with Occam's Razor: "One should not increase, beyond what is necessary, the number of entities required to explain anything." But it seems from the previous section that scientists also obey quite a different kind of razor as well, whereby one should not increase, beyond what is necessary, the number of disagreements required to conduct any inquiry. Furthermore, this is so for reasons that are parallel to those for Occam's razor. Occam's razor is essential for model building because for any given set of data, there is an infinite number of possible models to explain it. Occam's razor eliminates the vast majority of these competing models, making the resulting list more manageable. Conversely, the razor of disagreement is essential for arriving at a non-coercive convergence of opinion. For any inquiry, a seemingly infinite amount of disagreements can be introduced at every level of the inquiry. Occam's razor can get rid of some theories on the basis that they are unnecessarily complex, but this will still leave open an enormous number of possible points of contention. Part of the power of science consists in its ability to avoid disagreement about things that will impede its legitimate progress.
There is an obvious problem with this conception, however. It is where the analogy starts to break down between Occam's razor and the razor of disagreement. Even after one has eliminated all unnecessary assumptions from an explanation of the universe, it will be obvious that some assumptions will still be necessary in order to fit the facts. But it is not at all obvious what disagreements are necessary in order to arrive at a non-coercive convergence of opinion. It is quite possible, through an intolerant enough culture, to stifle all disagreement whatsoever. This would seem to make science no different from religion, in that it would have simply become another repository of dogma.
Perhaps this is missing the point, however. Religion has in the past used coercion as a means of arriving at a consensus, at least to the extent that anyone is willing to admit to what they believe in. On the other hand, science is committed to a non-coercive convergence of opinion. This means that they cannot in principle forcibly silence somebody who chooses to propose a new theory when they think that the current paradigm is already working correctly. However, they can do their best to resist adoption of the new idea for themselves. If the scientist with the new theory wants her own theory to gain general acceptance, therefore, she has to work much harder than if the dominant paradigm is in a crisis. In practice, it simply does not happen, because the majority of scientists will prefer to adhere to the paradigm to which they are already accustomed.
It seems to me that this conservative attitude towards new theories is fair enough. If science were to indulge every new theory that came along, they would never achieve the kind of non-coercive agreement that they naturally seek. A new theory constitutes a considerable expense to develop. A large amount of research has already been invested in an existing paradigm. It therefore does not make sense simply to scrap such a large investment without giving it every opportunity to prove itself to be correct. In this way, then, the razor of disagreement can be seen to form part of a cost-benefit analysis in scientific progress. While the current paradigm is still explaining the facts in a satisfactory manner, it is far too expensive to take on board another theory. But as the current paradigm starts to lose its value as more disparate facts come in that it cannot explain, the introduction of a new theory becomes more and more cost-effective. Looking at the situation in an idealised way, one can claim that there is a critical threshold for the adoption of new theories. Below the threshold, the benefits of the old paradigm outweigh the costs of the new one. Above the threshold, the benefits of the new paradigm outweigh the costs of the old. The threshold itself can be seen as the defining point in time in which a Kuhnian scientific revolution will take place.
There is an obvious problem with this conception, however. It is where the analogy starts to break down between Occam's razor and the razor of disagreement. Even after one has eliminated all unnecessary assumptions from an explanation of the universe, it will be obvious that some assumptions will still be necessary in order to fit the facts. But it is not at all obvious what disagreements are necessary in order to arrive at a non-coercive convergence of opinion. It is quite possible, through an intolerant enough culture, to stifle all disagreement whatsoever. This would seem to make science no different from religion, in that it would have simply become another repository of dogma.
Perhaps this is missing the point, however. Religion has in the past used coercion as a means of arriving at a consensus, at least to the extent that anyone is willing to admit to what they believe in. On the other hand, science is committed to a non-coercive convergence of opinion. This means that they cannot in principle forcibly silence somebody who chooses to propose a new theory when they think that the current paradigm is already working correctly. However, they can do their best to resist adoption of the new idea for themselves. If the scientist with the new theory wants her own theory to gain general acceptance, therefore, she has to work much harder than if the dominant paradigm is in a crisis. In practice, it simply does not happen, because the majority of scientists will prefer to adhere to the paradigm to which they are already accustomed.
It seems to me that this conservative attitude towards new theories is fair enough. If science were to indulge every new theory that came along, they would never achieve the kind of non-coercive agreement that they naturally seek. A new theory constitutes a considerable expense to develop. A large amount of research has already been invested in an existing paradigm. It therefore does not make sense simply to scrap such a large investment without giving it every opportunity to prove itself to be correct. In this way, then, the razor of disagreement can be seen to form part of a cost-benefit analysis in scientific progress. While the current paradigm is still explaining the facts in a satisfactory manner, it is far too expensive to take on board another theory. But as the current paradigm starts to lose its value as more disparate facts come in that it cannot explain, the introduction of a new theory becomes more and more cost-effective. Looking at the situation in an idealised way, one can claim that there is a critical threshold for the adoption of new theories. Below the threshold, the benefits of the old paradigm outweigh the costs of the new one. Above the threshold, the benefits of the new paradigm outweigh the costs of the old. The threshold itself can be seen as the defining point in time in which a Kuhnian scientific revolution will take place.
Comments