A book by F.A. Hayek: The Counter-Revolution of Science, get it from Mises.org, see also the full text online.
Some time passed since I have finished the book, so it's a good idea to look back at what remained from it. The basic lesson seems to have sunk in, though I have to refresh me on the details. The lesson comes later, first the impressions.
At round 400 pages (minus index), the book looks very readable. But don't be mistaken, it is quite a heavy reading, particularly the first part. There are three parts, actually three essays on one subject. It can be clearly felt, that they were written at different times.
The first part, it is noted, is quite difficult to process, so the reader is advised in the preface to consider reading the second part first. However, it seemed to me the most important and telling of all. Suffer through the theory, it is worth it. (I had real difficulty reading it, long, strangely complicated sentences - and I have had my share of Austrian economical literature. Weird. Maybe they were written in German first, then translated into English?)
The second part explains roots of the phenomenon described, with the third part following up. I admit they are in the end of little use for me, the unknown members and founders of long-dead schools of thought. By the time you get to names like Schmoller or even Marx, the huge impact of these ideas through half-forgotten thinkers is already clear. What is it, then?
Once upon a time, the science of economics, as other social sciences, was analyzed through certain methods particular for it. The rise of the "scientific method" in the natural sciences has but opened a new way of research and thinking, promising the same dramatic success. As Hayek notes, this view advanced not rarely by thinkers, that had some difficulty in applying the method consistently anyway. Let it be declared however, that there is nothing wrong with the scientific method, if it properly applied. Now to the when.
The rising Science (capital S here) had to fight against existing preconceptions, especially the interpretation of all events in the world as if it was animated by a mind of its own. A scientist should not look at what people thought of something, or how some other mind would direct the events, but replace it by objective facts and "reconstruct the concepts formed from ordinary experience on the basis of a systematic testing of the phenomena, so as to be better able to recognize the particular as an instance of a general rule". Science breaks up and replaces the system of classification which we use to classify and learn about our world. Things may appear to be the same, but are really something very different; things appearing different may be in all other respects the same (consider water and ice for a short example). Science becomes the way we explain the world as we perceive it; and appearances can be deceiving.
And so does Science frown upon individual observations, they are at most the start of true knowledge, never really the end. But until everything in the human mind is explained to the last detail, what men think of things, events and other men and their perceptions (be it right or wrong), remains very important; and Science tends to overlook this with its otherwise efficient methods.
Social sciences deal with the relationships between man and objects; and man and other men. Not all of them are inviable for analysis by the scientific method. But the social sciences in the narrower sense of the word, the "moral sciences", "are concerned with man's conscious or reflected action, actions where a person can be said to choose between various courses open to him" are different. People "will react in the same way to external stimuli which according to all objective tests are different" and will react differently to the same stimuli at different times. Our procedure is based on the experience that other people (mostly) classify their sense impressions as we do. A purely scientific description reduces the knowledge of the situation, or makes its explanation downright impossible. Many, if not most objects of human action are not "objective facts" (in the narrow scientific sense) and cannot be defined in physical terms at all. "So far as human actions are concerned the things are what the acting people think they are."
If you wanted to define a tool, for example, in the end you will have to refer to what it was intended for. And there may be basically nothing in common between items belonging under the definition (e.g. a hammer and a steamhammer), besides what men think they can be used for. As another example serves the archaeologist, that tries to find out, if something is an actual artifact, or merely a funny piece of stone - to do that, he must try to understand the mind of the prehistoric man and how he would use this item.
A point to mention is the subjective vs. objective assessment of facts: while in one sense beliefs, opinions, etc. of the people are certainly subjective, in another sense are they objective, hard facts of the social sciences; no matter if right or wrong otherwise. These opinions etc. can't be directly observed in the minds of people, but can be interpreted from their actions, because we have a similar mind as they have (but not the same). The knowledge, that guides the actions of any numbers of people, never exists as a consistent and coherent body. It is dispersed, incomplete and inconsistent, appearing in many individual minds - those are the basic facts of social sciences. Unless we can understand what the acting people mean by their actions any attempt to explain them... is bound to fail. It is the good old subjective nature of value, but much more strongly restated. There rises subjectivism.