During the eighteenth and nineteenth centuries we can see the emergence of a tension that has yet to be resolved, concerning the attitude of scientists towards the usefulness of science. During this time, scientists were careful not to stress too much their relationships with industry or the military. They were seeking autonomy for their activities. On the other hand, to get social support there had to be some perception that the fruits of scientific activity could have useful results. One resolution of this dilemma was to assert that science only contributed at the discovery stage; others, industrialists for example, could apply the results. ... Few noted the ... obvious paradox of this position; that, if scientists were to be distanced from the 'evil' effects of the applications of scientific ideas, so too should they receive no credit for the 'good' or socially beneficial, effects of their activities.
Co-author with Philip Gummett (1947- ), -British social scientist

To me there never has been a higher source of earthly honour or distinction than that connected with advances in science. I have not possessed enough of the eagle in my character to make a direct flight to the loftiest altitudes in the social world; and I certainly never endeavored to reach those heights by using the creeping powers of the reptile, who, in ascending, generally chooses the dirtiest path, because it is the easiest.

What these two sciences of recognition, evolution and immunology, have in common is not found in nonbiological systems such as 'evolving' stars. Such physical systems can be explained in terms of energy transfer, dynamics, causes, and even 'information transfer'. But they do not exhibit repertoires of variants ready for interaction by selection to give a population response according to a hereditary principle. The application of a selective principle in a recognition system, by the way, does not necessarily mean that genes must be involved—it simply means that any state resulting after selection is highly correlated in structure with the one that gave rise to it and that the correlation continues to be propagated. Nor is it the case that selection cannot itself introduce variation. But a constancy or 'memory' of selected events is necessary. If changes occurred so fast that what was selected could not emerge in the population or was destroyed, a recognition system would not survive. Physics proper does not deal with recognition systems, which are by their nature biological and historical systems. But all the laws of physics nevertheless apply to recognition systems.