Skip to content

Domains and theories in science

Last updated on 22 Jun 2018

We are getting to the tail end of this series. Here are the previous posts:

The false analogy between species and art

Pattern cladism and the myth of theory dependence of observation

Species, phenomena and data

More on phenomena

Disambiguating the Theory-Dependence of Observation thesis (TDOT)

In this post I want to discuss the theoretical relativity of domains. If observation is hostage to Theory, the question immediately occurs: which Theory? One answer is that it is either the Theory of the domain in question, say, biology or immunology, or it is from a Theory external to the domain (Hacking’s point about optics). Theories are either of that domain or not, or perhaps the domain is part of a wider domain defined by a general Theory. How do we tell?

In part, this is made more difficult because it is not entirely clear either what a Theory is in the literature (there are a number of conceptions of theories), or what a domain is. The latter question used to be a major topic in discussions of science in the latter half of the nineteenth century. In the period since it seems to be discussed under the rubric of reduction of Theories, and so in the modern literature, a domain is defined by its Theory. Hence, on this account, biology is a domain, or not, because it has (or doesn’t) a ruling Theory. Usually this is supposed to be Evolution (which is, in my view, at least six Theories, not one, but that’s for another time). Opponents of the evolutionary hegemony will offer Genetics, Development, or Ecology as defining global theories. Of course, if you do not have a Theory, but have a domain, then it gets interesting.

But this is a debate over contingencies of scientific development. It matters not a whit to nature that we choose to isolate biology from, say, chemistry (talk to a biochemist about that) or from thermodynamics (again, there are biological thermodynamicists) and so on. It must not be thought that the disciplinary structure of our present science is itself a natural classification and carves nature at any significant joints. In the usual parlance, it is unclear they are natural kinds themselves. Yet, studies like Okada and Simon’s (1997) on collaborative discovery in a domain do presuppose that the domain matches up to some sui generis field of discovery and investigation.

If we presume that a domain is defined by Theory, at least when it is a nascent field, then it is hard to see how a field might evolve in the first place. But if we think, as seems to be the received view, that a domain is a psychological, social or arbitrary field of enquiry, then it is hard to make out arguments about domain-specific theories, counter-reductionism, and the naturalness of a subject such as, for example, biology. If biology is simply what we prefer to identify, because of our prior conceptual dispositions (consider Gelman’s argument for the innate essentialistic disposition of children regarding living things in her Essential Child 2003), no inferences other than about our psychology may be made about domains (and domains are indeed largely the grist for educational psychology, e.g., Lawless and Kulikowich 2006).

So the bounding of natural domains by global Theory is a way to ensure that we can make inferences about objects within that domain. This leads rather directly to the Theory-dependence of ontology (TDOO). We appear to have a domain conundrum:

If TDOT5 is correct, and TDOO is correct, then no new scientific domain may arise from a prior state of ignorance and lack of Theory regarding that domain.

The only way to resolve this is, in my opinion, to deny either TDOT5 or TDOO. I deny both, in the sense that I think they are not necessary for all domains (they may be true of some domains). But first, let us consider under what conditions a domain may develop.

For a domain to identify some phenomena that are natural, we may approach these phenomena is one of two ways. The phenomena may be derived from Theory that is external to the domain (TheoryE), which would mean the ontology of these phenomena, and the data on which they are based, is not an ontology/data suite from the domain in question. For example, atoms and the ordinary chemical kinds of abiotic chemistry are not biological entities even though they play a crucial role in explanations of the aqueous media of cells and interstitial fluids. So one may begin to examine chemical phenomena in biology without there being a biochemical Theory (yet). This permits one to iteratively refine one’s categories and observe phenomena in biochemistry to the point where they can be classified in terms of some kind of Theory independently of the assumed chemistry kinds.

But if the TheoryE is too distant from the data and phenomena of the domain itself (quantum mechanics relative to most biology, for example), then the kind terms in a domain without Theory cannot be bootstrapped this way. Of course, it may turn out as a contingent historical thesis that no domain has ever begun like this – though I think it has in several instances (taxonomy in biology and mineralogy being two cases, and arguably also in nosology – medical classification – and even in the case of the development of the periodic table, which did rely upon laboratory techniques of refining elements and weighing them, but lacked much other Theory, Scerri 2006) – in which case we would treat this as a limiting case.

Moreover, it is likely that the notion of a domain is itself relative to background knowledge anyway: is mammalogy a domain independent of biology? What about protein chemistry? The nineteenth century is replete with examples of debating such questions (e.g., Jevons 1878), leading to the view that it was, after all, merely a matter of convenience and librarianship (Richardson 1901). Yet, we still debate whether or not biology is reducible to physics or has some irreducible core (Dupré 1981, 1993, Rosenberg 2006); the domains are taken seriously in this debate.

The domain conundrum may be resolved if we accept external Theory and ontology, but if we have no recourse to this, what then? I think the answer is to understand that domains can evolve both culturally and cognitively-psychologically in order to reliably demarcate natural phenomena. Psychological evolution is basically about perceptual and cognitive evolution: we see objects that are there in the ordinary sense because if we did not, we tended not to pass on our cognitive and perceptual traits – “Creatures inveterately wrong in their inductions have a pathetic, but praiseworthy, tendency to die before reproducing their kind” as Quine (1953) wrote so eloquently but normatively (see also Griffiths and Wilkins forthcoming and Wilkins and Griffiths forthcoming). So domains may begin through the construction of our evolved psychological traits, and then be refined. Moreover, a domain may begin as a social construct (for example herbalism in the middle ages, Wilkins 2009), and then be refined by experience to become more natural, leading to Theory in the domain.

In these cases, TheoryD does not determine what is observed initially, because there is no TheoryD. We begin with “ready-made” phenomenaD, but they are not handed to us on a plate by Nature. We surely have to construct them; only we do not need to already have observation languages and techniques unique to the domain under investigation. We can even construct the domains and be relatively sure they are more or less natural without a theory for that domain.

To finish, I offer this diagram, based on a metaphor of Papineau (1979). Note that there are no hard and fast divisions between TheoryD and observationsD, nor even (despite the sharp lines) between the domain itself and the external influences on observation. Theory and Observation are asymptotic ideals, possibly never to be reached.

Papineau's pyramid


We began by asking what the species concept referred to and whether or not it was necessarily a Theoretical concept, and hence either a unitary concept based upon a monistic Theory, or a plural concept based upon many Theories. We see now there is a third option: they are phenomenal objects that form the explananda for Theory in some cases. Given that there are many ways in which species is applied across the living world, and the lack of a viable rank or level of organisation that species (and only species) apply to, we now can explain why biologists nevertheless continue to identify and strive to explain species (Wilkins 2003, 2007).

We have three conclusions: The recognition that there are phenomenal objects in domains that might not be constrained by theories of that domain leads to the possibility of theory-independent classification. This suggests that scientific theories might develop in a manner that is more like classical empiricism than usually thought. Finally, phenomenal objects in a domain form the target of explanation by theory rather than necessarily being derived from it Duhem-style.


Dupré, John. 1981. Natural Kinds and Biological Taxa. The Philosophical Review 90 (1):66-90.

Dupré, John. 1993. The disorder of things: Metaphysical foundations of the disunity of science. Cambridge MA: Harvard University Press.

Gelman, Susan A. 2003. The essential child: origins of essentialism in everyday thought, Oxford series in cognitive development. Oxford; New York: Oxford University Press.

Griffiths, Paul E, and John S. Wilkins. In Press. When do evolutionary explanations of belief debunk belief? In Darwin in the 21st Century: Nature, Humanity, and God, edited by P. R. Sloan. Notre Dame, IN: Notre Dame University Press.

Jevons, William Stanley. 1878. The principles of science: a treatise on logic and scientific method. 2nd ed. London: Macmillan. Original edition, 1873.

Lawless, Kimberly A., and Jonna M. Kulikowich. 2006. Domain knowledge and individual interest: The effects of academic level and specialization in statistics and psychology. Contemporary Educational Psychology 31 (1):30-43.

Okada, Takeshi, and Herbert A. Simon. 1997. Collaborative discovery in a scientific domain. Cognitive Science 21 (2):109-146.

Papineau, David. 1979. Theory and meaning. Oxford; New York: Clarendon Press; Oxford University Press.

Quine, Willard Van Orman. 1953. From a logical point of view: 9 logico-philosophical essays. Cambridge MA: Harvard University Press.

Richardson, Ernest Cushing. 1901. Classification, theoretical and practical I. The order of the sciences. 2 vols. Vol. 1. New York: Scribner.

Rosenberg, Alexander. 2006. Darwinian reductionism, or, How to stop worrying and love molecular biology. Chicago: University of Chicago Press.

Scerri, Eric R. 2006. The periodic table: its story and its significance: Oxford University Press, NY.

Wilkins, John S. 2003. How to be a chaste species pluralist-realist: The origins of species modes and the Synapomorphic Species Concept. Biology and Philosophy 18:621-638.

Wilkins, John S. 2007. The dimensions, modes and definitions of species and speciation. Biology and Philosophy 22 (2):247 – 266.

Wilkins, John S. 2009. Species: a history of the idea, Species and Systematics. Berkeley: University of California Press.

Wilkins, John S., and Paul E. Griffiths. In Press. Evolutionary debunking arguments in three domains: Fact, value, and religion. In A New Science of Religion, edited by J. Maclaurin and G. Dawes. Chicago: University of Chicago Press.


  1. Markk Markk

    Speaking as someone who hasn’t put a lot of thought in this.
    Don’t we as humans come up with models almost instantly, even at the level of just categorizing things? We have qualitative models right away, for example, that this oak tree is different from that because it has different color leaves and its shape is different. So they are different kinds of trees. Isn’t that a theory? As we see more that are each color and shape we start quickly slotting them in the categories. Then we see one in between.
    But the theory is there essentially from the very start of observations or immediately in place.

    That doesn’t seem to quite match what you are saying I think.

  2. J. J. Ramsey J. J. Ramsey

    What’s with the “x-blogpost” in most of the links to the previous posts?

    • Cut’n’paste problem with my blog editor. Now fixed. Thanks.

  3. I know you mostly work with biology. However, I would like to know how you apply these ideas in physics, and particularly in electricity and magnetism.

    The early researchers began with very informal “observations” of hair standing on end, and of sparks. Much of the history of that research is a history of a struggle to find reliable ways of making observations. And our contemporary electromagnetic theory is, to a large extent, an account of how we now make observations.

    I am inclined to think it is a counter-example to your domain conundrum, but perhaps you are using “observation” in a broader and less precise sense than I would use it.

  4. This reminds me of the time I was reading Herodotus, and kept coming across serious references to flying serpents: as if he had actually seen them personally.

    I finally figured out he was talking about insects like dragonflies and such with long slender bodies. Not the giants of the Carboniferous, but modern ones: 15cm max.

    With my own “theory” for contrast, once I understood what he was talking about I could see the “theory” he had built into his observations.

    • Many thanks, Eric. I will find that piece, as I am in the process of working up a discussion of disciplines and domains.

      • bob koepp bob koepp

        John – If you’ve got a copy of Fred Suppe’s “The Structure of Scientific Theories” handy, that’s where to find Shapere’s essay. Starting from his work, I’ve come to think of an intellectual discipline as (roughly) constituted by a distinct set of problems (the domain of the discipline) and internal standards for what counts as a solution to a problem in that domain.

  5. varshitha varshitha


Comments are closed.