Coming from the natural sciences, I typically see a scientific model as follows: A scientific model takes all available evidence and attempts to ascribe a relating structure to all this information that will allow one to make a set of predictions about the natural world. When new evidence is introduced or when the model fails to accurately predict universally, a new model must be developed or an existing one must be modified. Because of the inherent complexity of the world, a model has to compress information through structure. It cannot simply describe everything - it has to have sufficient structure such that very complex processes can be explained through simpler systems. This is a very general definition and can vary if you are talking about mathematical/statistical models or biological models.

In less quantitative fields such as philosophy/social sciences, it seems to me things are different. Many philosophers seem to start with observations or subjective "axioms" (sometimes including empirical data) about the world and then attempt to find an underlying structure that relates these observations. The resulting model is not always testable. It may simply collect observations together to try to unify them. Many philosophers that have done this an invent very detailed terminology to accomplish this (many of the Continental Philosophers such as Hegel, Sartre, and Heidegger come to mind). In fact, even in this post I am attempting to define how other philosopher develop models through my own personal observations of how I have seen models developed in other authors I have read.

In many cases, I will read about an author's personal model and disagree entirely about how the observations are organized. This can happen in the case of scientific models too (after all, the results may be objective but the interpretation can be subjective) but because it is working on testable hypotheses, it is possible to evaluate the model's success or failure to predict reality.

My question: Are there any standards for how models are developed in philosophy for rigor and accuracy? What guidelines are typically followed when developing a model? How does one ensure that the model is accurate universally and not subjective when it cannot be tested in a formal, scientific sense? How has this been treated historically (1850-1900 science) versus the modern scientific era?

  • imvho each great philosopher redefines something like this, how to do philosophy
    – user6917
    Sep 26, 2016 at 17:38
  • 1
    I think your setting for philosophy is too close to science. This may fit some segment of analytic philosophy, at least aspirationally, but overall philosophers do not see themselves as constructing "models", and do not value "rigor and accuracy" very highly. As for universality/non-subjectivity many consider pretenses to it a vice rather than a virtue, and even those who strive for it have their particular versions of what it means. Originality, novelty, subtlety and unity of ideas, familiarity with tradition (especially in continental philosophy) are valued much more highly.
    – Conifold
    Sep 26, 2016 at 18:09
  • 1
    You're making quite a few controversial presuppositions such as philosophy being any less quantitative than science. Predicate calculus is no less quantitative than the math done when analyzing the results of this or that experiment. You probably mean that philosophy is less empirical in nature but even that is a controversial claim. I would recommend first looking at how philosophers look at scientific models. Van Fraasen's The Empirical Stance, Quine and Feyerabend are great starting points. Sep 26, 2016 at 19:57
  • Also for a putative first person perspectives on developing philosophical models: read Descartes Meditations, Husserl's Cartesian Meditations, Hume's Discourse on Human Nature, and Wittgenstein's Tractatus. Sep 26, 2016 at 20:02
  • @ Lee Malatesta "You're making quite a few controversial presuppositions such as philosophy being any less quantitative than science". This is more from ignorance on my part. I am less familiar with those fields of philosophy and, as I understand it, they overlap largely with mathematics so I wasn't so sure about where the demarcation lies. However, aren't the majority of more quantitative techniques from the 20th century? That still leaves a lot of other well-respected philosophy that does not build models in this way. Thank you for the references, I'm looking forward to learning more.
    – syntonicC
    Sep 26, 2016 at 20:08

3 Answers 3


Which standards? Well, I think there are two main standards:

Logical Coherence

In philosophy, every model (I will rather call them "theories" from now on) has to be reasoned carefully. Most lines of attack are against the logical coherence of a theory.

Just as in science there should be no falsifications of the model, in philosophy there should be no problem of coherence either. That is, the beliefs, statements and findings of a philosophical theory should be without explicit and implicit contradictions or, even worse, fallacies.

For example, Kant's ethics have been attacked not in the first place because of incoherence of the argument, but moreso because it is supposed to be applicable, while most people consider it inapplicable in particular situations.

Philosophers love to find contradictions and fallacies almost as much as pointing them out.

Historical coherence

By this, I basically mean 'Know your precedessors.' That is what universities prepare to, basically.

This does NOT mean that there can be no totally original, new theory of someone who doesn't know the history of philosophy. It does mean, however, that you should not redo the mistakes of former generations of philosophers.

This is, in my understanding, the main problem of (contemporary, that is!) philosophy: Because it is such a vast field and we are forced to publish relatively early within our educational development, philosophers tend to make mistakes that have been made centuries ago, often without realising it, because they did not even read the most important ones, not even thinking of understanding them.

  • I think this answer my question. While philosophy may not have any expressed standards for "model building" (constructing theories), the "testing" comes from testing against alternative interpretations of the same idea both past and contemporary. I get the sense that it's sort of like "does this interpretation bring us closer to understanding compared to alternatives?". Is this a fair interpretation of your answer?
    – syntonicC
    Sep 29, 2016 at 14:24
  • @syntonicC: Yes, it pretty much is. There are of course, as in science, meta-principles like Occam's Razor. That is also why philosophy so heavily relies on reference to other authors: They provide the foundation and strength of a theory by showing the relative merits. Science can do this relatively short, as there are few groups working on the very same specific problem, if any. The main part is the interpretation of data. But in philosophy, there is no data, there iss interpretation. For the most part. Good philosophy has both and is interdisciplinary, imho.
    – Philip Klöcking
    Sep 29, 2016 at 14:35

There is no one universally acclaimed standard or set of criteria. One possible way of conceptualizing this is that philosophy is an earlier step on the road to understanding a subject than science. The philosopher may either be considering a topic that has never previously been formulated, or one that has proven resistant to analysis.

Sometimes the philosopher's chief contribution is just posing the right questions --it falls to someone else to organize the concepts in a testable manner. In retrospect, it may look like the philosopher's ideas are ridiculous, because we've all gotten used to the later, more developed version. But the real brilliance of the philosopher was thinking such a thing was even possible.

  • 1
    I guess the logical coherence is a minimum criteria for philosophical theories, isn't it? Objections are normally put forth by showing fallacies or explicit as well as implicit contradictions regarding arguments/beliefs/propositions/etc.
    – Philip Klöcking
    Sep 26, 2016 at 17:51
  • @PhilipKlöcking Theoretically, but how well do Plato or Descartes arguments really cohere? And those are two of the biggest names ever. Nevertheless, I'd welcome you offering your version as a competing answer. Sep 26, 2016 at 18:06
  • Logical coherence is overrated. We still have problems with most of the ancient paradoxes concerning logic. We can't even successfully reduce mathematics to logic in a satisfying way. And, somewhat embarrassingly, the success (or failure) of a philosophical model is typically a function of its popularity - or at least its popularity within certain groups - rather than any sound philosophical basis. Sep 26, 2016 at 20:07
  • @LeeMalatesta: It is the same as stating that empirical data coherence is overrated. There is no such thing as absolute coherence, nevertheless it is the standard in which the quality of a model is measured. What you are talking about does not really hold in the long run, although tragic exceptions can be found.
    – Philip Klöcking
    Sep 26, 2016 at 21:09
  • Read Feyerabend's treatment of the Galileo controversy in Against Method. If we are to consider Galileo a scientist and his work important in scientific progress, then we have to account for his actual methods: trial and error, ignoring results that contradicted his theories, engaging in propaganda. There is an unwarranted assumption going on here that empirical knowledge (or scientific knowledge) is somehow epistemicly privileged and that the success of modern science is due to something more than its cumulative nature. Sep 27, 2016 at 14:54

Excellent answers Philip Klocking. I have been working on theorizing (Weick et al) and personally think that philosophical methods, doing philosophizing is theorizing. It employs certain aspects of the processes of theorizing. Theorizing, theory-construction or theory-building is not only important in the the discourses of 'natural' sciences. See this link Theory in Social Science www.ssc.wisc.edu/~jpiliavi/357/theory.white.pdf Theory in Social Science. P I. What is a theory? < A. Definition from Schutt: A logically interrelated set of propositions about empirical reality. These propositions . It sets out 'theory' in details. I have written a lot on theorizing and the methods, methodology of doing philosophy, for example here -https://philpapers.org/profile/342710

  • 1
    Welcome to philosophy and thanks for the answer! I notice it reads a little bit like a comment on another answer... --Is there any chance you could spell out a little more clearly why you think it is a persuasive answer to the question in its own right?
    – Joseph Weissman
    Mar 5, 2017 at 20:35

You must log in to answer this question.

Not the answer you're looking for? Browse other questions tagged .