What is science?

Vox Day asks a question: what is my definition of science? It’s a bit weird coming from him — he is not usually that lucid or civil — but OK, I’ll take it seriously.

Unfortunately, “science” is one of those hugely polymorphic terms that carries a tremendous amount of baggage, and any one definition is going to be inadequate. This is one of those subjects where a smart philosopher (Janet? John?) could go on at amazing length, and even then, everyone will argue with their summaries. I’ll just charge in, though, and give a couple of shorter definitions off the top of my head.

This first one isn’t entirely satisfactory to me, since there’s too much implication that science is a thing that can be examined and treated as an at least briefly static object, but it probably accords better with most people’s naive concept of science.

#1: Science is a changing and growing collection of knowledge, characterized by transparency (all methods are documented, and the lineage of ideas can be traced) and testability (prior work can be repeated or its results evaluated). It is an edifice of information that contains all of the details of its construction.

When I teach introductory biology, a substantial part of it addresses this definition: there are conventions of the scientific literature that a new student must learn in order to be able to efficiently extract information and follow the chain of evidence, and also in order to some day be able to add to it. To the novice, science can appear to be a huge database, and what they have to figure out is how to tap into it.

If you’ve been in science for a while, you know there’s another pragmatic definition:

#2: Science is what scientists do. We have institutions that train people and employ them in the business of generating new knowledge — contributing to that edifice in definition #1 — and we have procedures like the bestowal of degrees and ranks that certify one’s membership in the hallowed ranks of science.

This is science as a social construct, a tradition, a meat grinder, a tool that shapes people into useful configurations for churning out new data and ideas. It isn’t pretty (any grad students or post-docs or struggling faculty out there? Yeah, you know it’s ugly) but it’s part of the reality of science.

There is also an ideal of science that we try to live up to, and which is the other significant component of our introductory biology courses here — it’s what we aspire to in good science.

#3: Science is a process. It is a method for exploring the natural world by making observations, drawing inferences, and testing those inferences with further experimentation and observation. It isn’t so much the data generated as it is a way of thinking critically about the universe and our own interpretations of it.

A more philosophically inclined writer could then go on at length about induction and reductionism and materialism and the scientific methods and so forth; they’d all be part of that definition if made more thorough. I suspect the commenters here will cheerfully expand on everything and tell me where I’m wrong — and that’s part of it, too. Science is a matter of throwing out ideas and refining them in the crucible of reality.


  1. #1 David Marjanovi?
    March 6, 2007

    Practice: Science rests on a tripod of theory (ideas), observation and experiment.

    No, experiments are just a way of arranging opportunities for observation whenever and wherever you want to have them. Observations made elsewhere can suffice, however. Astrophysics and geology are sciences.

    But biodiversity surveys and taxonomy are also science–

    Biodiversity surveys are the gathering of data — that’s a prerequisite of science, but not the whole thing. Taxonomy is an art, and that’s why it has started dying out.

    Similarly, Science could handle divine interventions and miracles, if there were any.

    Only if they were predictable. (If a miracle is predictable, is it still a miracle? Or does it become just another law of nature?)

    Never miss an opportunity to quote Terry Pratchett. ;)

    Well said!!!

  2. #2 David Marjanovi?
    March 6, 2007

    The elegance of Einstein as compared to Aquinas, is like comparing Picasso to a beer commercial.

    Great quote except for the extra comma. I need to remember it!

    (And to find an image of one of Picasso’s early paintings, from when he didn’t make fun of his customers yet.)

    DNA peers to the beginning of time

    It won’t get you farther than the beginning of life.

    Why is “accepting” the premises of science more justified, or more likely to lead to true beliefs, than “accepting” the premises of religion, if neither set of premises is rational?

    Let’s save frog from his poetry here. :o)

    Hypotheses derived from the premises of science are testable. Whether they correspond with reality can simply be observed. Hypotheses derived from the premises of religion are usually not testable — and therefore useless, because they can explain everything and its opposite.

  3. #3 David Marjanovi?
    March 7, 2007

    We are trying to make a foundation for reason. We can’t use reason to do that!

    Maybe not, but we can make a really beautiful argumentum ad lapidem about it. In some medieval caliphate (sorry — I’ve forgot the century and the names of the protagonists) there was a discussion whether philosophy should be done at all or whether religion should be left alone. An advocate of the latter view wrote a book “The refutation of the philosophers”. An advocate of the former answered in a book called “The refutation of the refutation” which makes the basic point that you can’t argue against reason — if you use reason to argue against reason, you’re contradicting yourself; if you don’t use reason, you are unreasonable.

    That’s the beauty about methodological naturalism. I haven’t put enough thought into it, but I think it makes argumenta ad lapidem work. =8-)

    DNA peers to the beginning of time

    It won’t get you farther than the beginning of life.

    That depends on your TOE. If you go with Max Tegmark’s TOE == Ultimate Ensemble Theory, you might interpret it to mean that time outside of self-aware entities isn’t quite the same as observable time.

    Philosophers have no right to make stuff up and call it a TOE. That’s what physicists are for — the important difference being that their TOE will hopefully be testable.

    You could also go with a biological panspermia hypothesis that could push you pretty close to the Big Bang.

    If a couple hundred million years (and that’s optimistic) are “close”…

    I think I read in one of Feynman’s books this definition: experiment is the sole arbiter of truth.
    Succinct, yet complete.

    Too narrow. Exchange “experiment” for “repeated observation”. (And maybe “truth” for “reality”.)

    G.Tingey: Science assumes, indeed MUST assume, that only natural forces are at work, and that there are no special cases or exceptions.
    Very nice but quite useless if you cannot provide a RELIABLE method to sort out “natural forces” from (hypothetical) “non-natural” ones.

    Interesting question, but I think it has a very simple answer: if we observe it so regularly that we can develop a hypothesis that successfully predicts it, it is part of reality — of nature — natural. If it can’t be predicted, it’s a miracle… or we just need to look and think harder. Greg Byshenk is right (both times).

    Can you elaborate on your Metaphysical Naturalism?

    That’s the methodological one.

    Science is not just correlating data. It must have a theoretical construct to interpret that data. Saying that the Sun rises every morning is just the bare beginning of science.

    Well, we come up with hypotheses, and use data to test them. Whether we gather the data first (as in your example) or last doesn’t really matter. (Gathering them first may save us a lot of time otherwise spent speculating idly, but it’s more boring.)

    The placebo effect can be studied scientifically. (Of course hypotheses have already been suggested that it boils down to hormones and the like.)

  4. #4 David Marjanovi?
    March 9, 2007

    Thanks for the explanation of Tegmark’s idea.

    On science, I was being prescriptive, not descriptive. Having only just started the Ph.D. program, I have never written a grant application. :-)