How would you define "science"?

My definition is the systematic study of all realms of natural activity and processes with the intention of formulating descriptive narratives that lead to predicting either future outcomes, or past outcomes that have not yet been fully described or analyzed.

[[Notable Caveat:
Particularly challenging is the analysis of what some would call the super-natural, with the understanding that for something to come under fruitful scientific study requires the phenomenon to be regular enough to enable prediction. By some definitions, the very term “super-natural” means beyond nature, beyond prediction, beyond regularity - - in other words, in the realm of mysticism, magic or metaphysics - - and thus beyond the realm of scientific study!]]


Since this got such an overwhelming response (thanks @gbrooks9) let me start out with how I talk about “what is science?” at the start of all my introductory science courses (majors and non-majors).

I start out defining science in Webster’s-type terms “systematic study of the physical universe” but of course that’s so vague as to border on useless. I then go through the following points.

The Goals of Science

  • reliable knowledge - explanation of how the world works, this is what got me into science
  • prediction - warning (think earthquakes, hurricanes) and verification (testing for reliability)
  • manipulation - most scientists aren’t content to just sit on knowledge, they want to manipulate the universe (see prediction and reliable knowledge) in some way. Things like medicine, human flourishing, stewardship of resources (I also talk about the negative side of this).

The Scientific Method
I dislike giving the impression there is a “the” scientific method. Instead I like to use a progressive disclosure case study about Ignaz Semmelweis, who was early pioneer of antiseptic medical procedures, to give a feel for how observation, experimentation (including design), and refinement based on analyzed results can work.

Then we talk about the scientific method using the following figure:

Products of Science
I think it’s important for students to get an idea of what the goal is and what the difference between laws and theories is. Also, I emphasize models because it is very important down the road (climate change, etc.) to talk about how scientist develop a model of how things work and then test and analyze them to get at that reliable knowledge.

  • Laws: well-tested description of observed phenomena. “When this is the situation, X ALWAYS happens”
  • Models: conceptual or mathematical analogy
  • Theories: tested hypotheses & models
  • unlike facts, laws are broadly applicable
  • unlike theories, laws do not give a mechanism

Key Elements of the Scientific Method
Instead of focusing “the scientific method” diagram, I focus more on the “features” of science, which I break down into properties and processes:


  • Open to change, unfinished, creative
  • Able to make predictions, testable, reproducible, falsifiable
  • Shared


  • Observation
  • Hypothesis
  • Prediction
  • Testing
  • Dissemination

Science is a Human Endeavor
We talk about investigator bias and how science has some built-in ways to deal with it:

  • Multiple investigators - science is usually done in teams (which surprises most students)
  • Independent peer-review
  • Sample size

I talk about careful selection of variables and controls and show a bunch of goofy “correlation does not equal causation” graphs like how the divorce rate in Maine correlates well with per capita consumption of margarine.

Certainty vs Scope
The last thing I do might be the most controversial in this group and the one I’d like to develop more. I say that in order for science to be able to generate reliable knowledge (the primary goal) it limits its scope and methods.

I go through an exercise where I ask the class “how certain are you”:

  • that you ate breakfast? this one is easy, we do a poll of who and who didn’t. We talk about our confidence being in that we made an objective observation.
  • of your birthday? they are very confident, and then I tell them about someone I know who doesn’t know their birthday or even birth year. We take this one based on authority (my parents told me) or reliable evidence (birth certificate) but it’s not really a question of science.
  • that you are human? we talk about how this one is trivial, but more tricky than they might think depending on how one defines “human”. Science may have an answer, but it may not be the full answer.
  • that One Direction was the greatest boy band of all time? this one can be the most hotly contested. We talk about various measures (most album sales, most #1 hits, etc.) but in the end, we know that it is a personal question, not a scientific one.

We then talk about the difference between opinion, reasoned arguments, historical events, and what can really be addressed by science. Science self-limits to questions that can lead to certainty (at least quantifiable certainty) and reliable (repeatable, consistent) knowledge.

Any thoughts?

1 Like

@jordan, aren’t there multiple understandings if what science is? How do you for that in here? Are you just presenting one of many views? Or an over arching meta proposal?

Yes, there are certainly multiple understandings. Part of this is just practical. I can’t really go into a science classroom and say “well, there are so many different understandings of science that it’s just sort of whatever you think it is”. So what I’m after, in the spirit of common ground, is to see if we can’t come up with a decent overarching understanding what science is. Or at least get it down to 2-3 statements that cover the vast majority of people.

Part of this is also motivated by constant arguments between ID proponents and scientists over whether it’s really science or not. It really seems like they may be working from different definitions and I’d like to sharpen that up. I’d like to do the same with “natural” but that’s a bigger nut to crack. I’m not really sure how to even begin that one.

This has come up in origin subjects because creationism(s) are accused of not science and opponents say they are.
In truth there is no such thing as science.
All there is in human affairs is people figuring out things using intelligence.
Yet we must use this outdated concept of science which replaced a former concept called natural philosophy.
The best, I say, the best they can say is that SCIENCE is a high standard of methodology that can demand confidence in its conclusions.
(So YEC/ID and evols can say YOU are not doing the methodology right and so your conclusions are suspect)
So inventing/confirming new medicines must show a methodology before put on the market. they do and this is science.
That includes testing, predicting, etc etc.

In invisible things like origin matters, possibly lots of physics stuff, or anything ITS harder to use methodology. ITS ABOUT INVISIBLE things. Only somewhat mote then the supernatural.
So it must be done in the spectrum of forensics.
Yet proving things, so invisible, is close to impossible.

YEC has Gods word(limited boundaries established) but otherwise does science in debunking the other side or proving our hypothesis.
ID debunks and hypothesis. Evolutionists debunk and hypothesis.
Its all figuring things out.
Is for all, any, one, methodology not been applied to justify the title SCIENCE.
i say its on a case by case issue.
i always, with satisfied success, ask evolutionists to show scientific biological evidence for evolution.
they never can. Instead they bring up comparative anatomy and genetics, geology/fossils, biogeography, limited in-species selectionism, and classification concepts.
They misunderstand this is not science backing up a evolution hypothesis/theory.
Its not science because its methodology is not using biological evidence of biological processes.
They truly don’t do a high standard of method. They can not put their conclusions on the market I say.

Science doesn’t mean something unless it means something different then other methods of figuring things out.
It must be a high standard of investigation or go home.
Origin matters are slippery for everyone relative to PROVEN science.

I.D. supporters clearly conceive of science as the study of anything “real”… even if by “real” they include how the parts of Christian metaphysics affects the natural world.

Unitarian Universalists are probably pretty light on the idea that Christian metaphysics can “push” or “pull” any part of the natural world.

Others choose the opposite view.

But even with the opposite view, if Science requires the Scientific Method, how does one “control for the variable of God” ?

Similarly, if ESP is REAL … is REAL and REGULAR enough that it can actually be studied? All the experiments suggest NO… ESP cannot be studied. So Scientists say it isn’t real.

But there are plenty of people who think ESP is real… even if it can’t be studied.

This differs from the I.D. community: they think God is real (check!). They think God can be studied scientificially (Nope!).

Its Gods creation that can be studied. It shows brilliance within its essence. The other side says the universe shows happanhance.
Something got to give with someone!.

Somewhere in the description of science we need “indepent” and “empirical”. I tend to look at the debate between the Rationalists and Empiricists during the infancy of science as the real turning point. I will probably get a lot of this wrong, but from my understanding the Rationalists thought the truth of the universe could be deduced from human reason while the Empiricists thought we had to test our ideas against the objective facts found in the universe. At the heart of science is the idea of putting your ideas at risk by judging them against empirical measurements that can themselves be checked by others. It’s the risk of being proven wrong that runs down the center of the scientific method, at least in my humble opinion.

I think it is also important to point out that science makes inferences, not deductions. Scientific conclusions are based on the limited evidence we have, and are always tentative.

1 Like

@T_aquaticus, Do you think “empirical” and “testable” are the same thing? What do you mean by “independent”?

I sometimes struggle with the tentative nature of science. Sometimes, while technically tentative, we are really quite certain something is a certain way. How do we retain the inductive nature of science without making the public think we don’t know what we’re talking about. I will often hear from non-science majors something to the effect of “science doesn’t prove anything, it’s always changing, so why should we believe it?”

1 Like

No, they don’t think that – except in the sense that one could draw conclusions about what an ancient architect, whose biography is unknown, must have been thinking regarding a building which he designed, which is still standing and available for us to study, measure, etc. That would give us some idea of the architect’s mind, range of intellectual ability, etc.

Thus for Newton, science could teach us something about God, because science showed us how the universe – which God designed – worked. But neither Newton nor any early modern scientist colleague, nor any ID proponent today, ever asserted or asserts that we could or can put God under a microscope, subject him to chemical and physical tests, etc. We cannot make God an object. God is a subject. But we can learn something about subjects from what they do, and in that indirect manner, scientific knowledge can potentially teach us something about God.

It is one thing to infer the existence of an intelligent designer, and another to claim to have spotted the designer, captured the designer, strapped the designer to a table, and examined the designer and learned all of his properties, powers, secrets, motives, etc. ID has never been so foolish as to claim the latter. Anyone who thinks so either hasn’t read ID literature carefully, or hasn’t read it at all.

I don’t expect to change George’s mind on this, since his purely ideological orientation to ID was long ago fixed and since he has steadily refused to read even one book by an ID author all the way through and therefore is incorrigible by means of literary evidence. I’ve offered this rebuttal purely to encourage those silent readers who may suspect something wrong with statements like George’s, and are looking for confirmation that their hunch is right.


Rather than rely on an “if and only if” definition for science, I think it is better to start of with agreed goals for science, namely prediction, explanation, and control of its domain of study. Then define a scientific research program as one which successfully meets these goals.

Many measures of success are then available to public scrutiny: do its the theories underlie technology we use? Is their ongoing government and private industry funding for the research program? Are there new students eager to enter the that scientific research program? What do scientists in neighboring domains of science think of the research program?.

Rather than defining, we then study the practices of that program as characterizing science in that domain. We do find commonalities across science but we also find differences of emphasis. Furthermore, we can recognize that the scientific community itself decides how to apply the practices, how to change them over time, and how to weight their importance for each scientific decision.

Recognizing theories are fallible is then seen to be part of the nature of successful inquiry, for successful inquiry must always be open to new evidence and new ideas. But we can also study how science maintains continuity with past theories, eg by requiring that any new theory be able to explain as well as past theories (the generalized correspondence principle).

FWIW (and philosophy is generally not worth much on internet forums!), the nature of science is covered in philosophy of science under the topic of the demarcation problem.


One thing I have not noticed in the preceding discussion is the role of the Merton norms in ensuring that the objectivity of the scientific process…

As per my previous post, these emerge from studying the successful practices across scientific domains. From the linked Wiki article:

The four Mertonian norms (often abbreviated as the CUDOS-norms) can be summarised as:

  • communism : all scientists should have common ownership of scientific goods (intellectual property), to promote collective collaboration; secrecy is the opposite of this norm.[3]
  • universalism : scientific validity is independent of the sociopolitical status/personal attributes of its participants[4]
  • disinterestedness : scientific institutions act for the benefit of a common scientific enterprise, rather than for the personal gain of individuals within them
  • organized scepticism : scientific claims should be exposed to critical scrutiny before being accepted: both in methodology and institutional codes of conduct.[5]

I like this set of goals. I’ve been using “reliable knowledge, prediction, manipulation” which are similar I think.

Yes, I’m very interested in the demarcation problem. I took a History & Philosophy of Science class as an undergrad (in an Environmental Science program, which seems odd now) and we studied Gould’s NOMA.

As a science professor at a Christian liberal arts university I find positivism a non-starter and Kuhn and Popper have always seemed helpful but not quite complete given the speed of science these days and the fuzzy boundaries in theoretical physics (falsifiable is perhaps ambitious). NOMA and methodological naturalism are OK and sort of what I use now, but I really like critical realism and the work of Alister McGrath and T.F. Torrance (both theologians and not philosophers, but they address the interaction between science and theology quite a bit) talking about science and theology as conversation partners.

I think the demarcation problem becomes important if we want to allow science to be science and theology to be theology, while also having them interact in meaningful ways. I see that to some degree here on Peaceful Science, it’s just really hard to do well.

1 Like

Independent and empirical are close to the same thing, but testable is a separate entity.

  1. Empirical means measurable through the senses, and includes the use of tools (e.g. mass spectrometer). The absorbance spectra of carbon dioxide or the density of 4C water at sea level are empirical measurements.

  2. Independent means it doesn’t solely rely on “because I say so”. The Royal Society is the oldest modern scientific institution founded in 1660, and its motto is “Nullius in verba” which is Latin for “take nobody’s word for it”. In other words, show me. Evidence needs to be repeatable so that others, in priniciple, can repeat your work and confirm your data.

  3. Testable means your hypothesis can be proven wrong. A hypothesis must make predictions about what you should see, AND what you should not see. These need to be risky predictions. By risky I mean that the potentially falsifying evidence should be attainable in principle. A good experimental test of a hypothesis should be capable of producing the potentially falsifying observations if the hypothesis is false.

Perhaps my brain is tweaked in the wrong way, but I think the ever changing nature of science is one of its most wonderful attributes. As you say, there are things that we are very certain of and those ideas are not really challenged or considered tentative, even though they really are. However, the ideas at the fringes of science are the most tentative, and that is where science is the most exciting.

As to those non-science majors, tell them not to believe it, and then encourage them to look at the evidence for themselves and draw their own conclusions. Of course, a few science classes wouldn’t hurt. :wink:

I am fond of the curmudgeon Steven Weinberg’s view of philosophy and science:

Good science is a bit like pornography–we know it when we see it. We learn through experience what is and isn’t good science.


Out of respect for your views, let me adjust my initial wording, and amplify it with some context:

Perhaps not every I.D. supporter thinks the presence of God’s design and/or engagement can be scientifically revealed… but the majority appear to have no other way to understand the natural world other than it can be systematically categorized by Science, even if the systematic categorization being sought is God’s role in design!


Let’s hear your version of an answer to the following question:

“If God can front-load cosmic creation so that he arranges the creation of one-celled life forms, some of which are equipped with flagellum (as per his plan), is there a way to scientifically determine that God arranged for these flagella to evolve?”

The example above is what I MEAN when I might say that I.D. supporters are putting God under the microscope… it’s a figure of speech. I obviously don’t think I can put any part of God under a microscopic investigation.

For example, in Behe’s new book, he seems eager to point out that Evolution without God usually leads to Devolution. And yet we also know he feels God is responsible for the successful (and timely) emergence of flagella used by single-celled life forms.

So…between the extremes of “brilliant design” and “terrible results from the lack of brilliant design” - - in this vast gray area - - how does Behe propose that we will know when God was not doing the “design”, but things still worked out surprisingly well?!

That is something I have noticed as well. You would think that after all of their snide comments about “scientism” and “materialism” that ID supporters would eschew scientific support for their beliefs, but reality is just the opposite. ID supporters struggle immensely to fit ID into a scientific and materialistic framework as if science is the final arbiter of truth. They even go as far as to try and discredit scientific theories by attaching religious epithets to them (e.g. the “Church of Darwin”) as if being religious disqualifies a belief. ID supporters even go further by claiming that ID isn’t even religious or supernatural because it could conceivably include physical beings, like Niven’s Engineers. Again, an appeal to materialism.

I think it makes a lot more sense to meld theology into science as TE’s and EC’s do rather than pit nature against religion as ID supporters do.

1 Like

I tend to think, and @TedDavis would be the person to ask, that the history of the ID and YEC/OEC movements shows that they are modernist in approach, seeing science as the final arbiter of truth. That is why they work so hard to make science to prove God and the Bible inerrantly true through extensive concordism. What confuses me is, sometimes they seem to combine that with presuppositional apologetics if the science doesn’t work our right. Basically, science will prove the existence of God, unless it doesn’t, but that’s only because you’re not presupposing the existence of God. That kinda how I feel many of the methodological naturalism discussions go.

1 Like

Again we must come back to the distinction between what some people in “the ID movement” may do, and what ID as a theory holds.

Many people in the ID movement (though not too many of the leading lights) are concordists, as you point out, and many of them overrate the importance of science. This is true of most creationists as well, and there is some overlap between ID and creationist folks. All of this is true.

Nonetheless, ID as such has no position on concordism, the Bible, Christianity, etc. It does, however, have a position on the nature of science, and its position is opposed to that of many in the atheist and TE/EC scientific communities. It holds that the purpose of natural science is to get at the truth about nature, following the evidence wherever it leads, unhampered by arbitrary rules such as “methodological naturalism.” Thus, if reason and evidence seem to point to design as the explanation for the origin of something, most atheist and TE/EC people reflexively respond, “No, we can’t go there,” ID people say “Why can’t we go there, if that’s what reason and evidence seem to indicate?”

I don’t see any resolution to this problem any time soon. If a ruling is made that design inferences can’t, even in principle, belong to science, because of “methodological naturalism,” then ID cannot proceed as a scientific project. If a more nuanced ruling is adopted, i.e., that design inferences might in principle be acceptable in science, though the particular design arguments offered by ID proponents are weak or invalid, then ID can keep on trying to improve itself. But if the ruling from on high is that design inferences by their very nature can’t be allowed within science, then ID people and their opponents are playing two entirely different games (as if one is playing by the rules of ping pong and the other by the rules of tennis), and all discussion must end.

George’s position is clearly that design inferences are ruled out of court by science, by the very nature of science, and therefore that ID was the pursuit of a chimera from the beginning, and ought to be abandoned. He’s entitled to that belief, but it rests on a particular account of the nature of science, the nature of knowledge, the nature of nature, etc. I’m not obliged to accept that account, and I don’t. I go with a more nuanced account in which design inferences are generally not relevant in natural science, but are not absolutely forbidden. This was the attitude of Newton, Boyle, and Kepler. The ID folks are mostly with Newton etc. The TE/EC folks, on the other hand, go with Descartes, Bacon, and Kant, and of course the atheists do, too. The differences thus go back to major disagreements in early modern philosophy – though most of the proponents of the various modern schools are completely oblivious of the history of those differences, and how they play out in modern arguments.