Prominent theoretical physicist and science educator David Hestenes begins his widely cited paper Modeling Games in the Newtonian World (American Journal of Physics 60, 732 (1992)) with this:

The great game of science is modeling the real world, and each scientific theory lays down a system of rules for playing the game. The object of the game is to construct valid models of real objects and processes. Such models comprise the core of scientific knowledge. To understand science is to know how scientific models are constructed and validated.

My question -- Where does contemporary philosophy of science stand on this view -- that the business of science is modelling? Both personal opinions and references to supporting, opposing or other views are welcome.

Note that the emphasis of my question is somewhat different than and narrower than these...

Is Science about Truth or Adequate Models?

Is the primary aim of science predicting outcomes of experiments or understanding the nature of our world?

... which are about the metaphysical status of science and scientific assertions and their relationship to reality and truth (at least that is how they got answered).

My question here is more about what scientists are actually doing when they conduct science -- the "business" of science, as it were, how it is carried out rather than its ultimate aim or meaning. I realize those are not necessarily easy to disentangle, but I am seeking references that lean toward the business side rather than the meaning side. Maybe ignoring the world "real" in the cited paragraph will help.

  • If you are interested in what scientists are actually doing then this is not the right SE to ask, it is best answered by scientists themselves, or perhaps by sociologists or historians. If, on the other hand, the question is what philosophers of science argue is the "business" of science then this question is indeed a duplicate. Predicting outcomes of experiments or "understanding" some phenomena sound like "business" to me, not "ultimate aim or meaning".
    – Conifold
    Commented Oct 17, 2017 at 19:42
  • I see your point. But I still think philosophers can and often do weigh in on issues of methodology in various fields. It's not so much "what scientists are actually doing" -- that would be more sociology or history, as you say, but what their methodology amounts to. For example, could you contend that Linnaean biological taxonomy is a model, much less a mathematical model. I think that both could be argued (or opposed), and there are philosophical issues involved. There's also the question of what the methodology ought to be, and that's clearly a philosophical issue. Commented Oct 18, 2017 at 1:55
  • 1
    There's an ongoing debate that started in the 70, not about the business of science, but about what a theory is (independently of the question of realism). Different authors argue that theories should be viewed as collection of models rather than statements about the world as was traditionally considered. This is the semantic view of theories, which emphasise the role of models in scientific representation and empirical confrontation and it's now dominant. I was about to answer something in this line but as the question is about the "business"of science, not about theories, I put a comment. Commented Oct 18, 2017 at 6:13
  • 1
    @Conifold: Many contemporary academic philosophers of science are also historians of science or STS researchers. More broadly, I would argue that case studies of actual scientific practice have become the standard method in philosophy of science. For the purposes of a short comment, we can think of this as empirically analyzing and articulating the unstated norms of actual scientific practice, which cuts right across the disciplinary distinction you're trying to make.
    – Dan Hicks
    Commented Oct 18, 2017 at 13:42
  • 1
    @DavidLewis: As Quentin Ruyant says, there's a large literature on models in science. (Stanford Encyclopedia article here.) Margaret Morrison and Michael Weisberg have both argued that models are, in Morrison's language, "autonomous" of theory, which goes to answer your question. But I don't have time right now to give a full answer. Check out the Morrison references in the SEPh article, Weisberg's [Who is a Modeler?](file:///Users/dhicks1/Downloads/download.pdf), and ch. 5 of Weisberg's Simulation and Similarity.
    – Dan Hicks
    Commented Oct 18, 2017 at 13:49

1 Answer 1


Models were quite neglected by philosophers of science and even scientists before the second half of 20th century: science was more thought of as in the business of producing statements about the world, but the view was criticised in particular by Suppes and Suppe in the 60's and 70's, who emphasised the central role of models in scientific representation and empirical confrontation. This has come to be called the semantic view of theories, according to which a scientific theory is just (or is better presented through, analysed as...) a collection of models, and it as been endorsed by many authors, realist or anti-realist alike (e.g. van Fraassen, French, Ladyman, ...). Models are conceived of by these authors as mathematical structures that represent features of the world, and the laws of a theory are just a way to describe this collection of structure.

These authors think that this view has several vitues: it is close from scientific practice, where empirical confrontation is made through comparison of data models and theoretical models (e.g. van Fraassen "the scientific image"), mathematical structures are more or less language-independent which allows one to bypass semantic issues in science (ibid), it provides mathematical tools such as isomorphisms to analyse intertheoretical relations (reduction, theory change), etc.

Not everyone agrees though: some think that theoretical laws are more than descriptions of collections of models (but represent natural laws, e.g. Maudlin in "metaphysics within physics"), or should be viewed as tools to produce contextual models (Cartwright, Suarez). Some think that language is still important to apply models (Chakravartty). But in any case, the importance of models in scientific activity is widely recognised today.

For a review of this specific debate you can read Lutz "what was the syntax-semantic debate about?" Or SEP entry on the structure of scientific theories https://plato.stanford.edu/entries/structure-scientific-theories/

Note however that "model"is a polysemic word. I only addressed one aspect here: models as mathematical structures built from theoretical laws. There are other types of models: scale models (a reproduction of DNA), analogies and metaphors (model of the mind as a computer) or thought experiments... And a large literature on their status and role.

For a larger view on models in general you can read the dedicated SEP https://plato.stanford.edu/entries/models-science/ or Bailer-Jones "scientific models in philosophy of science".

You must log in to answer this question.

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