Science theorises by proposing ideal types and deducing ideal relationships between them. In nature there is no ideal sphere touching an ideal frictionless plane in an ideal single point. Instead of these ideals, nature gives us avalanches. Yet to study real avalanches the theory derived from the unreal ideal is required. Presumably, reality is too chaotic to theorise directly. Does all useful theory depend on ideal types? It does seem usual. Economics creates idealised relational theories from idealised constructs such as homo economicus, market clearing, perfect information and other things which do not and cannot exist in reality. Presumably, this idealisation approach is one reason for the relative success of economics compared with other social sciences. In the natural sciences measurement is also ideal. For example, a temperature noted as 23.59 degrees is not real: the reality will be plus or minus some small amount. The recorded value, like any exact number, is a mathematical abstraction. In the social sciences there are no units of measure so the only available ideal value for any concept is its total presence or total absence. In economics, the concept of rationality is perfect rationality, market clearing is absolute. Economic theories are constructed from all-or-nothing ideals. Why don’t the other social sciences develop relational theories based on ideal types? For example: total cooperation, perfect competition, utterly forgiving nature, perfectly nice human nature, perfect selfishness, total equality, perfect justice, utter domination, complete freedom, etc, etc. Shouldn’t this be the normal approach? Am I reinventing a philosophical (square) wheel or is this a new thought?

Your question is excellent. Though I am afraid your proposal is not completely novel insofar as Plato initiated a philosophy of ideal forms in all areas of life (the good, the true, the beautiful, the just, and so on), though of course he was working long before we began carving up inquiry into the different natural and social sciences. At many points in the history of ideas, philosophers have worked with ideal or what has come to be called paradigm cases. So, in the theory of knowledge, a philosopher might describe an ideal or paradigm case of what it is to know some internal state (the feeling of pain) or see a remote object and then use that paradigm to assess different, more controversial knowledge-claims. So, one might entertain an ideal case of what it is to see a person, and then ask whether claims to see or perceive a sacred reality (God) in religious experience is similar or too remote to count as evidence. And in ethics we often use thought experiments to try to capture the different values that come into play with, for example, the duty to keep promises or refrain from killing and so on. So while there has been a healthy role for ideal or paradigm cases, there has also been a focus on all the messy details in non-ideal situations. To go with ethics again, the reason why philosophical disputes persist on abortion is not (in my view) because of disagreements on ideal cases, though there actually is disagreement on that level. But more often (I suggest) it is because we continue to wrestle with problems in philosophy of mind (when can we say with confidence that there is mental life..), reverence for life principles, and so on.

You may find the work of John Rawls interesting. He proposed two levels of reflection when carrying out political (and essentially ethical) theory. Balance your abstract reasoning with a study of concrete, case studies and work out your position in what he called reflective equilibrium.

Read another response by Charles Taliaferro
Read another response about Science