Data on "Facts" and Facts on "Data"

By Sean Carroll | October 14, 2009 9:44 am

A philosophy professor of mine used to like to start a new semester by demanding of his class, “How many facts are in this room?” No right answer, of course — the lesson was supposed to be that the word “fact” doesn’t apply directly to some particular kind of thing we find lying around in the world. Indeed, one might go so far as to argue that what counts as a “fact” depends on one’s theoretical framework. (Is “spacetime is curved” a fact? What if spacetime isn’t fundamental in quantum gravity?)

Nevertheless, people sometimes use the word. A recent post by PZ reminded me of how it comes up especially in arguments over evolution, which is occasionally accused of being “just a theory.” I’ve tried to make my own view clear — when we as scientists use these words, we shouldn’t pretend they have some once-and-for-all meanings that were handed down by Francis Bacon when he was putting the finishing touches on the scientific method. Rather, we should be honest about how they are actually used. “Theory,” in particular, isn’t cleanly separate from words like “law” or “hypothesis” or “model,” and doesn’t have any well-defined status on the spectrum from obviously false to certainly true. And “fact” — well, that’s a word scientists hardly use at all. We use words like “data” or “evidence,” but the concept of a “fact” simply isn’t that useful in scientific practice.

But you know what would really be useful here? Some facts! Or at least some data. There’s one repository of professional scientific communication that I know very well — SPIRES, the high-energy physics literature database run by SLAC. (My hypothesis guess is that any other field would turn up similar results.) I don’t know an easy way to search entire papers, but it’s child’s play to search the titles. So let’s ask it — how often do scientists (as represented by high-energy physicists) use the word “fact”?

find t fact or t facts
120 records

Okay, they clearly use the word sometimes. What about some competitors?

find t data
9909 records

Ha! Now that’s the kind of word scientists like to use. And the others?

find t evidence
4396 records

find t observation or t observations
10924 records

You get the picture. Scientists prefer not to talk about “facts,” because it’s hard to tell what’s a fact and what isn’t. Science looks at the data, and tries to understand it in terms of hypothetical models, which rise or fall in acceptance as new data are gathered and better theories are proposed. Just for fun:

find t theory
42285 records

find t model
45977 records

find t hypothesis
578 records

find t law
1293 records

So I’m happy to say evolution is “true,” or is “correct,” but I’ll leave “facts” to Joe Friday.

  • The Science Pundit

    Scientists prefer not to talk about “facts,” because it’s hard to tell what’s a fact and what isn’t.

    Is that a fact? 😛

  • Robert Cumming

    OK, so why do the general populace think scientists are the still biggest fact merchants out there?

    My guess: because we don’t like discussion of the validity of our ideas unless its on our terms (read: background reading, clear thinking, the scientific method). For most people, that’s dogma and fact-mongering, however reasonable our terms seem to us.

  • Michael T.

    So Sean, do you think it is correct to say then that a “theory” is simply the best explanation supporting a given body of evidence ?

  • Bee

    The issue isn’t with the facts, but with the “is.” Spacetime “is” curved (as a matter of fact) in your book. But what “is” spacetime “really?” Oh-oh. You get around that by talking about data. Data is data. You don’t say the data “is” spacetime. Or is it ;-p

    The reason why we use the word hypothesis so rarely is that nobody knows what the plural or genitive is.

  • FUG

    What about something like the Atomic Theory of matter, the Germ Theory of disease, or any other number of old hypothesis that are in the 99.999% chance of being a fact? Is there not a “limit” when a theory becomes what we think of as facts?

    Evolution is on par with those theories. As long as we are comfortable with thinking of the atomic theory as a fact, then we should feel just as comfortable with evolution.

  • Igor Khavkine

    The arXiv has allowed full text search of its contents for a while now (can be found under Advanced Search). However, the results would have to be filtered for rhetorical uses of some words, as in the expressions “the fact that” or “in theory”.

  • Sean

    Michael, I don’t think that, no. The phlogiston theory of combustion is certainly not the best explanation for the evidence we have.

  • Anne

    @6: I quickly tried some searches with the Arxiv, and unfortunately the abstract search doesn’t list the number of hits if there are too many, and the full-text search does but crashes or reports syntax errors on some of the words of interest. So I wouldn’t trust the results too much even for the results that do turn up.

    @2: It seems to me that people think that scientists are fact merchants because in science classes what we teach is almost always the results of science, not science. In an introductory physics course, for example, you might be taught Hooke’s law (extension of a spring is linear in the force applied). Hooke’s law is the result of Hooke doing some science. But you are not doing any science, or even being told how to do science; you are being told a fact. You are then required to reproduce and apply this fact on the assignments, midterm, and/or final exam.

    If you have a lab section of the course, you may be asked to perform a procedure to show that the spring you are handed follows Hooke’s law. This is slightly closer to doing actual science: in principle you might actually find out that Hooke’s law was not appropriate for this particular spring. But in most physics lab courses, what you are expected to do when your data fail to match the model is mumble something about “experimental error” and hope the TA doesn’t take off marks. It’s not until much more advanced courses, or maybe graduate school, that you are actually given a question and asked to do some science: to figure out the answer for yourself by observing the world.

    So for people who haven’t reached that point, it looks a lot like science is a big collection of facts, and the job of scientists is to convey those facts to you. The idea that people are figuring out new facts as we speak, or that some of those new “facts” are actually hotly debated, is very rarely conveyed.

  • Sam Gralla

    Nice post =)

  • Arun

    Anne said it regarding science.

    As to other things, we deal in facts all the time. E.g., at a particular period on Oct 13, 2009, policeman X either fired his weapon or didn’t. Or the President said such and such or didn’t. Obama was either born in Hawaii or not. And so on. Seemingly independent of theoretical framework :)

  • Eugene

    find t crazy

    9 records

    Clearly, we are not crazy enough.

  • Jdhuey

    WRT @8 : Funny you should mention Hooke’s Law in this regard because in my high school physical science class my experience was very different. The first day we were given a bunch of different strength springs and a set of standard weights and instructed to measure the spring displacements. We then spent the last half of the class aggregating the data from the different students and then derived Hooke’s law from the results. Admittedly, the teacher was herding us toward the correct result but the lesson that was being taught was not what Hooke’s law was but on how to extract regularities from experimental data.

    I don’t know how high schools and colleges teach science nowadays but way back in the 60’s and 70’s the emphasis, at least in my school, was primarily on the process.

  • Mike Perkins

    Someone, I think it was Stephen Jay Gould (and he may have been quoting someone else), said the “a fact is something for which there is so much evidence that to deny conditional acceptance would be perverse”. Always sounded like a workable definition to me.


Discover's Newsletter

Sign up to get the latest science news delivered weekly right to your inbox!

Cosmic Variance

Random samplings from a universe of ideas.

About Sean Carroll

Sean Carroll is a Senior Research Associate in the Department of Physics at the California Institute of Technology. His research interests include theoretical aspects of cosmology, field theory, and gravitation. His most recent book is The Particle at the End of the Universe, about the Large Hadron Collider and the search for the Higgs boson. Here are some of his favorite blog posts, home page, and email: carroll [at] .


See More

Collapse bottom bar