In a blizzard way, it is simply the dominance of the privileged. The reality is that science attract s more money, is more quantifiable and most importantly, produces most of the decision makers in research. As a result, the conditions associated to the so called impact factor are science orientated. You see, science research often lead to financial outcomes and indeed, very concrete outcomes as in the case of medical related sciences. Be it in the area of research or anywhere else, the god we worship is money. 

Dear Juha,

Please search the net (or Web of Science/JSR methodology, or specialised datasets) for "Normalised Impact Factor" publications.

I attached here three "quick search" freely available articles on the net. Please start with the one titled "Impact Factor".

In addition, I think the Thomson Reuters services (WoS, JSR, InCites?) have tools/menus/options for direct normalisation and comparison of IFs accros scientific fields and disciplines.

Best regards.

In some countries, researchers in the humanities and social sciences do not worry about impact factors at all - researchers are supposed to read papers, and papers are deemed influential, if they influence people's thoughts. In addition, many texts are still written mostly in the native languages (because you cannot express all thoughts in all languages and as there are different discourse traditions), and other formats such as books and edited volumes are deemed as important as peer-reviewed papers. The latter do not appear in impact factor calculations at all, because typically, only Anglo-Saxon peer-reviewed journals are used for calculating impact factor - thus simply not measurung the "real" discourse properly.