In Theory

Of all scientific terminologies, "theory" is surely the most misunderstood. I've covered this before many, many times, but somehow I don't feel I've quite done it justice. Let me attempt to put this right. First the short version, the one you can link to if someone on the internet is shouting, "it's only a theory !!!" and you want them to shut up and go away.


Theory : Too Long Didn't Read Version

Sorry people, "theory" is ambiguous and has many meanings. It's context dependent. Yeah, I know, that's a right bugger for arguing on the internet. Alas, the notion that a theory has a very specific meaning to scientists (e.g. a model that's been especially well-tested) is simply not true at all, and even if it did, that wouldn't mean that everyone else is being an ignoramus by using it to mean something different.

A theory is a model that describes how the world works... but not all theories are equal. Sometimes it's fine to say "it's only a theory", sometimes it isn't. This depends on the degree of testing which the theory has survived. It's not about "theory", it's about the prefix "only". Rough guide :

• "Anyway, the round earth / evolution / relativity / dark matter /  is only a theory" - people who say this tend to be ignorant savages who live in huts and sacrifice goats to avoid angering the Great Duck. I've tried to place those examples in a rough order of how well-tested they are, but in all those cases, saying, "it's only a theory" is directly equivalent to saying, "I'm a pillock."
• Something you or your mates just thought up of the top of your head. As in, "You know me mate Dave ? I've got this theory right, that he's actually, like, secretly a ninja. Just a theory though." That one mostly definitely is only a theory, because it isn't well-tested at all.

So, calm down people. Take your time and consider carefully how well-tested your theory is. If it gets a lot of things right - even if it also gets some things wrong - then you can't say "it's only a theory", because that ignores how well it does. In general, forget whether or not the theory makes any sense to you and concentrate on the results. But if the theory only explains a few things and not very well, and gets a lot of things wrong, or hasn't been tested at all... then you can absolutely say, "it's only a theory". Or better yet : it's only a model.

The only other detail is that someone might say, "it's only a theory" when what they really mean is, "it's not a fact". That's a bit tougher. Gravity isn't a theory, it's a fact - but models of how gravity works (like relativity) are very definitely theories, not facts. Similarly the spherical shape of the Earth is a fact, but we don't know its shape with infinite precision, so in that sense it's also a theory. And we know with certainty that evolution happens, even if much remains unknown about genetics. So if you want to say, "it's not a fact", it very much depends on which theory you're discussing.

That's really all there is to it. You may stop reading at this point if you like. But if a more detailed analysis is needed, read on. There's nothing new here for regular readers, I just want a go-to post. I've got this theory that collecting these ideas together may be useful...

Theory, Defined

The dictionary offers no less than seven entries for theory. No wonder it's so misunderstood ! But it's not quite so bad as it first appears, and in fact we might get away with grouping them into three :

1. A coherent group of tested general propositions, commonly regarded as correct, that can be used as principles of explanation and prediction for a class of phenomena (e.g. Einstein's theory of relativity; synonyms : principle, law, doctrine)
2. A proposed explanation whose status is still conjectural and subject to experimentation, in contrast to well-established propositions that are regarded as reporting matters of actual fact; contemplation or speculation; guess or conjecture (e.g. the theory that there is life on other planets; synonyms : idea, notion, hypothesis, postulate).
3. A body of principles, theorems, or the like, belonging to one subject; the branch of a science or art that deals with its principles or methods, as distinguished from its practice; a particular conception or view of something to be done or of the method of doing it; a system of rules or principles (e.g. music theory, number theory).

Over-simplifying :

1. Things that are true.
2. Things that might be true but, then again, might not be.
3. The set of things we currently think might be true or might not be true and the methods we use to establish if we think they might be true or not.

The third one need not concern us today as it's a bit different from the other two, which are quite complex enough to be getting on with. Let's start with the second one.


Definition 2 : Hypotheses 

This one is actually the easiest because it also falls under another, more rigorous definition : hypothesis. As in, "I've got this hypothesis that if I sneeze too hard my eyes will pop out." That fits a few known facts (that if you sneeze pressure builds up) but it hasn't been tested. Or, "I keep getting attacked by pigeons because I've got this bag of bird seed in my pocket, so if I get rid of it they'll stop". That one offers both an explanation and a prediction; its prediction hasn't been tested yet but it could be.

A hypothesis is an idea that's consistent with what you know. It can contain within it an explanation of how the world works, sometimes implicit ("my eyes must be connected to my lungs, somehow") and sometimes explicit ("pigeons attack me to get my bird seed"). Usually it should also make a prediction, but it must always be broadly consistent with known facts. If not, then it's not a hypothesis, it's just a wrong idea. Once you learn about anatomy, you'll realise that no amount of sneezing will cause nasty things to happen to your eyeballs.

Already we see that there are very good reasons for confusion since even the dictionary doesn't ruthlessly distinguish between hypothesis and theory. And in everyday speech, hypothesis and theory are used interchangeably by scientists and everyone else alike.


Definition 1 : Theories or Laws ?

This definition is very much stronger and is what angry people on the internet like to insist is the correct usage of the word, ignoring the fact that a hypothesis is also a type of theory. In this version a theory is our very best explanation for what's going on. It must, by necessity, have undergone extensive testing and made predictions in agreement with observations to within the measurement errors. But while it may be "commonly regarded as correct", it's usually distinctly different from a fact.

Cats provide some nice examples. For example, "the cat doesn't like being kicked, so if I keep kicking the cat it will become angry and attack me". That contains a testable prediction and an explanation. The prediction can become factual - either the cat does or doesn't attack you. Even better would be to generalise it to all cats rather than one specific feline. So if a hundred people kick a hundred cats, and ninety eight people come away slashed across the face, then that theory is doing pretty darn well on the testability front. It doesn't have to be perfect, but it has to do very well.

But what about the explanation ? Can you ever really be certain about what the cat likes and doesn't like, or if its reaction you describe as anger is the same sort of anger that you sometimes feel ? Not really. Oh, you could stick the luckless animal in an MRI scanner or something and watch its brain activity, or even dissect it and see if human and feline brain structures are similar, but this wouldn't ever tell you what it feels like to be a cat. Your explanation for the cat's behaviour is a very good one, but it's not a fact.

On the other hand, if you'd said, "the cat will not be able to walk across the ice on the pond because the cat is too heavy", then that statement makes a prediction and contains an explanation but is entirely factual. It can be proven to be correct, just as the spherical nature of the Earth transmuted from theory to fact as observations became irrefutable. Theory and fact cannot be distinguished so easily in this case.

The point is that even this stronger definition of theory doesn't automatically mean, "we think this is a fact". Sometimes it does and sometimes it doesn't. It's all very grey and murky and deeply unsatisfying, but that's life for you.


But where's the dividing line ?

Forget the line and remember that it's drawn in sand. It's not exactly a 700m high wall of ice, is what I'm saying.

Even if we were to insist that hypothesis only means, "explanation with little or no testing" (which it does) and theory only means, "well-tested explanation" (which it doesn't), then it wouldn't be easy to distinguish between the two. No strict criteria of what "well tested" means exists. It's probably impossible anyway, given the incredibly diverse nature of theories. You can't equate cat emotions with the distortion of spacetime around a black hole, or at least you shouldn't.

The reality is, though, that the vast majority of theories fall somewhere between these two extremes. They aren't just speculations based on limited data, and they aren't so convincing that no other explanations are plausible. They've had some testing and they generally work, but they have room for improvement. Some of them might turn out to be completely wrong, others just need tweaking.

So if everything is fuzzy and there aren't any black and white definitions of "theory", can we at least appeal to common usage to clear up this mess ? Is it at least true that the scientific use of "theory" is distinct from the bloke down the pub who thinks his mate Dave is secretly a ninja ?


Actual usage

Nope. No such luck, this is purely mythical. After ten years in academia - fourteen if you count undergraduate studies - no-one has ever told me there's this specific scientific definition that we have to use. Scientists use the word in just the same way that everyone else does - interchangeably with "hypothesis". And this is by sheer necessity, because yes there are theories that have withstood very high levels of testing indeed and yes, there are mere speculative hypotheses... but most models fall between these two extremes. What else are we supposed to call them if not "theories" ? We can't just call them "models" because they have undergone some testing.


Laws

The myth didn't come from nowhere, of course. Clearly there are some theories which do extraordinarily well - sometimes so well that theory and fact are indistinguishable. It might be fair to start to describe these as laws, not theories - the law of gravity, the law of evolution. Both of these things are established factual processes. Yet even these are like Russian dolls : within them we find detailed theoretical models of how they occur, and within those we find competing hypotheses as to how particular aspects proceed and even rivals to the theory - but not the laws. Gravity is a thing. Evolution happens. It's the mechanisms by which these things occur that's open to debate (at least a little bit), not their very existence.

So does the dictionary's statement that "law" is a synonym of theory hold up ? Only in very limited, carefully defined circumstances. In general I prefer to relate laws to observation, i.e. they are descriptions of what happens, not why it happens. Newton's laws of motion give excellent descriptions of how objects move under certain conditions, but they don't explain the causal mechanism. Just like legal laws, really - which tell you that there's a fine for kicking cats, but they don't go on to explain about empathy and why animal cruelty is unethical*.

* Plato had the rather interesting notion that legal laws should do this, but that's another story.

We should probably use the term "law" more often in these cases where it really does apply, especially for evolution. But this doesn't solve everything, because while evolution is certainly a law, and gravity is a law, relatively is, err... just a theory. But, like, a mega-awesome theory. Or something.

EDIT : Unfortunately there's an additional complication that "law" is not used in a consistent way. For example Freeman's Law describes how galaxies have the same central surface brightness, while the Kennicutt-Schmidt Law describes how star formation relates to gas density. Not only do these completely lack an underlying explanation, but they have a large scatter (e.g. the "same" surface brightness, plus or minus a bit) and sometimes aren't even true at all. What we really mean is that these are relations which are extremely well determined, but they are not perfect and inviolable in the sense that gravity and evolution appear to be.


Principles

I don't think anyone would much quibble over the definition of "law" in the scientific sense, caveats notwithstanding. Another term which is perhaps under-used in the popular vernacular is "principle". This one's more ambiguous in terms of strict definitions and common usage. My geology teacher explained it quite nicely as an explanation you can't prove (though it may have supporting evidence) but you have to assume in order to make progress. For example, by studying volcanic activity today we may be able to understand volcanic activity millions of years ago - we assume they're basically similar. I like this version better than the dictionary's, which is as follows :

A fundamental, primary, or general law or truth from which others are derived (e.g. the principles of modern physics); a fundamental doctrine or tenet; a distinctive ruling opinion (e.g. the principles of the Stoics); a rule or law exemplified in natural phenomena, the construction or operation of a machine, the working of a system, or the like (e.g. the principle of capillary attraction).

Although this is consistent with the earlier definition in which theory, principle and law are all synonyms, I don't find this useful. They are too close to one another. I prefer a version which mimics how the words are used outside the scientific realm. Principles, then, would be our best guess as to the true explanation for a law or other observations, which we then extrapolate to say holds true in other, untested and/or untestable situations (or, similarly, a statement that behaviour we observe in one set of conditions will remain similar in others).

If, by analogy, a man believes it's wrong to kick cats, then that's one of his principles. If the lawyers tell him he'll be imprisoned for kicking cats, then that's a law. If he finds himself confronted by an angry lion and refuses to kick the cat in self-defence, then that man is tremendously principled and stupid in equal measure; his principles have cost him his life whereas the law would have seen him temporarily incarcerated at most (and probably not at all, since few laws would be framed as stupidly as to allow people to get eaten by lions).

In astronomy we have the Copernican Principle - the idea that we're not located in a particularly unusual part of the Universe. That's somewhat supported by observational evidence and makes life a hell of a lot easier - but we can't prove it absolutely. In geology the Principle of Uniformitarianism  says that current processes are much the same as they were in the distant past. This too is an evidenced-based assumption that enables progress, just as assuming believing that kicking cats is wrong avoids running foul of the law. It perhaps can't be proven with certainty, but it could be disproven. I think my high school geology teacher had a much better idea than the dictionary's notion that principle and law were synonymous - I think it's more useful to give these words different meanings.

So could we declare that some of our very best theories, while falling short of the high standards of laws, are at least our principles ? We can't say "the law of relativity" because relativity is a complex theory which has many distinct observationally-proven aspects : those are individual laws, whereas the theory as a whole is a proposed explanation for those laws. That is, the curvature of spacetime explains gravity and time dilation. Unfortunately, we can't just co-opt the term "principle of relativity" because that already has a different, widely accepted definition. But we might be able to get away with talking about, "the principle of curved spacetime".


Yeah, but what about common usage ?

Actual scientists worry an awful lot about this, honest.

What ? Oh, yeah, sorry, I guess I got a bit distracted. Principles, laws, hypotheses and theories all have distinct meanings. The first three are not quite iron-clad, totally unambiguous terms, but "theory" is especially vague. Both in practise and in the strict linguistic definition, it can mean anything from, "something me mate Dave said one time over a pint" right up to, "something that's so utterly certain you could knock yer teeth out with it."

It's worth repeatedly emphasising that there isn't any special scientific definition of theory. String theory is a nice example of how strange things can get. It's got all the trappings one normally associates with a theory : an underlying physical interpretation, detailed mathematical models, consistency with known facts and laws, improvements in (especially philosophical) problems with standard theories. What it doesn't offer is any predictions - at least nothing that can be used to objectively test and distinguish it from other theories.

Some people, scientists included, treat the requirement for testability as an absolute, so regardless of how well a model performs in general, if it lacks this one condition then they deem it to be unscientific. Personally I disagree. I just don't think that things are ever so black and white as that.

For example, there are scientists who are pure theorists (and yes that's the term that's used). That is, they never perform observational checks themselves, they spend their whole time playing with mathematical models and numerical simulations. And not just ones based on the super-strict variety of theories, but all kinds of wildly speculative models. I think many theorists would be rather surprised to be told that they were supposed to be working only within the confines of known laws, much less facts, let alone that they or anyone else should ever actually try and test their ideas.


But surely this specific definition of theory must have some basis in reality.

So is there any benefit in trying to define "theory" to have the specific meaning that many angry people on the internet insist that it already has, and why are they trying to do this ?

To the first part of the question I'd have to say that it certainly would be nice if there was a word that meant this, but "theory" definitely isn't it. It already has its own meaning and it's very broad, encompassing laws, principles and hypotheses with open arms. Trying to pin it down to only mean something like, "a well-tested model based on observed laws and explanatory principles, which is consistent with so much observational data that it's just not even funny and if your mate Dave says 'it's only a theory' then you should kick 'im inna ribs" just doesn't work - the word is too complex. It would be a bit like trying to redefine, "dairy products" to only mean, "cheese". Sure, with enough effort you could do it, but it wouldn't be worthwhile and you'd lose a valuable term in the process.

As far as I can tell the reason people want the word to have this more specific (admittedly very valuable) definition is because a very small but influential group of scientists do think that's what it means. From the National Academies of Science, Engineering and Medicine :

The formal scientific definition of theory is quite different from the everyday meaning of the word. It refers to a comprehensive explanation of some aspect of nature that is supported by a vast body of evidence.

But again, this isn't true. Unless there's some Secret Scientist's Special Dictionary that everyone forgot to tell me about, this is just rubbish (the rest of the article, mind you, is quite good). It's true that scientists do adopt a different common usage amongst themselves for many everyday words (as we've seen, principles, laws, also bias), but no such formal definitions exist anywhere. Moreover, not only do we utterly lack this equivalent of a dictionary, but we just plain don't use the word that way. And string theory isn't some weird exception either...


"Theory" as used by actual scientists doing actual proper research

Look, I'll prove it to you. If there really was a rigorous scientific definition of the term, then the best place to look for it would surely be in the pinnacle of scientific literature : peer-reviewed papers. I ran a query for "theory" in the abstract and/or title of all peer-reviewed papers available on NASA's ADS service. This is heavily biased towards astronomy, but still a lot of manual filtering had to be applied to select things I even vaguely understand. Still, the results are pretty clear, so here are some examples (I could easily find a lot more, but please don't make me) :

• This article discusses the author's recent "united theory of planet formation". It's obvious from the abstract that they're actually just describing a model, and nothing at all comparable to the "commonly regarded as correct" almost-law version of the popular media.
• This article talks about "several competing theories" of atomic collisions, and you can't really have several theories which are commonly regarded as correct.
• This one uses the term "theory" in the title but swaps it for "model" in the abstract.
• This one talks about a very recent "theory" of gravity which is not even accepted by a minority of people, let alone a majority.
• Here's another one that interchanges "theory" in the title with "model" in the abstract.
• This one explicitly describes itself as developing a new theory, which is by definition impossible with the "well-tested model" description, let alone the idea that it should be commonly accepted.
• This one talks about a theory of gravity I haven't even heard of. Widely accepted ? I think not.
• Another one talking about a bunch of competing theories, this time with "theory" an explicit part of their names.
• Another one developing a new theory.
• This one discusses "preliminary comparison of theory and observation", which is not possible in the "well tested model" definition.

So let's nail that coffin shut on the idea that "theory" has a special definition to scientists. That's just not true. I almost wish it was true, because it would be useful to have a word with that meaning, but it isn't.

Why do these people claim that it means something so very different to the way the rest of us use it ? I find it hard to believe that any scientist really goes around being so rigorous - it would make daily conversations very difficult.


Sinister forces ?

If I was feeling cruel, I might suggest that such people are followers of scientism : the idea that scientifically-obtained knowledge is the only true kind of knowledge and that anything else is just opinion. This is a very black and white, downright silly version of science that's popular on the internet because people like simple answers.

Was Neil merely over-simplifying ? Perhaps, but I doubt it. There is no one theory of gravity; even Newton's theory of gravity is still useful despite the fact that Einstein's is better. And even if he meant Einstein's, it's crazy to call it a fact when virtually everyone accepts that it's flawed, because basically no-one believes that singularities are real. No, I'm afraid this is an attempt to redefine knowledge : science has the facts, no-one else is allowed any.

Now I should re-iterate that the NASEM link is quite good, apart from its mistaken formal definition of "theory"; I don't wish to ascribe them as being scientismists. But this tribalist, us-against-them approach is a big problem in science communication*. It is at least a plausible explanation why this very strange definition of theory is sometimes deployed rather aggressively, because I simply cannot believe that it's always an honest mistake. I just don't see how so many people could believe that such a commonly-used word has such a different, specific meaning that it simply doesn't.

* I suppose it's rather ironic to say, "I don't agree with the way those tribalists do outreach !". Certainly we should be very careful not to fall into the trap of just becoming another variant of the problem, defining ourselves by hating tribalists... much as the Welsh are largely defined by not being English**.
** I say that as a Welshman, you understand.

Did this popular but mistaken idea really originate from scientism ? I don't know for sure, though it would be consistent. Of course, there are legions of internet-based idiots who have this wrong-headed idea that science is all about facts, not uncertainties, and that we should trust the scientific consensus unconditionally. If their favourite pet scientist said it wasn't true, then it definitely isn't. These acolytes have never done an hour's honest research in their lives (the classic problem of sciolism, presuming much greater understanding of a topic than is actually the case). You could say the same about any field, of course.

What you get with scientism, though, is a number of people who are also extremely intelligent and technically proficient, who are genuinely good scientists. But instead of using their knowledge to educate, they (also) use it to beat people and demonstrate how clever they are and how stupid everyone else is. It's not that they attack the genuinely stupid or the wilfully ignorant (for such people do exist), it's that they automatically assume everyone who disagrees with them must be stupid by definition. They make no effort to engage or understand the other, sometimes uncomfortable perspectives, they just declare them to be wrong. Shouting, "theories are the same as facts !" is an oh-so-perfect way to declare victory without actually having fought the battle.

It doesn't really work, of course. It just alienates people.

Although its origins are murky, it's much easier to understand how this idea of the special meaning of theory can spread in the popular media : it has a certain degree of "truthiness" to it; why wouldn't scientists have a special, more rigorous definition than everyone else ? That's just the sort of thing scientists do. More than that, it even has some basis in reality. The very best theories really are distinct from the run-of-the-mill variety : relativity, evolution, and climate change are tantamount to being actual laws or at least compromised in no small part of laws more than mere ideas. But we simply don't have any single special word for those kinds of theories, and that's the end of it.


Knowledge

I don't want to go on to develop an entire theory of knowledge, because that would probably take at least twenty years. That's why this section is only under a subheading. But, while we're here, I shall offer some crude thoughts as to why scientism, to me as a scientist, just seems wrong-headed :

Facts : Are what you can objectively measure and verify in the world with hard numbers and repeated observations. If you can't repeat it, then you still have a ,"fact that this thing was measured and these were the values", but you don't really know if those values are true.

Knowledge : Can be comprised of facts but not always. You have an opinion about what it's like to be a cat, but you know what it's like to be you. Vice-versa, your cat probably has an opinion - most likely a very unflattering one - about what it's like to be a human, but only it knows what it's like to be a cat. It's not an opinion, but neither is it a fact. Similarly, I know what I imagine and emotionally feel, but I can't objectively prove it. These things are subjective, yet I know them to be true.

Opinions : Is a catch-all term for what you think is happening or emotionally approve of. As in, the movie Star Wars factually exists, you know you like Star Wars, in your opinion it's a good movie. That example is purely subjective but they can also be evidenced-based, e.g. your opinion is that it will rain today because the forecast said so, but later this can be objectively tested.

Conclusions : The collective result of opinions and beliefs - that is, things you genuinely think are true, regardless of the state of the evidence (or lack thereof) supporting them. Thus you can form a conclusion by entirely scientific, logical reasoning that is utterly wrong because you didn't go back to check the facts : "The cat loves tuna, cats should eat as much tuna as possible, therefore I conclude it would be a good idea to buy the cat a whole tuna every day."

Un-science : Deals with matters that cannot be scientifically verified, e.g. whether Scarlet Johansson is prettier than Salma Hyek or Sean Bean has more macho charisma than Ron Perlman. Subjective opinions and unverifiable knowledge; moral issues and the like. Not always a bad thing (often unavoidable and indeed a good thing, as with humour) unless it becomes...

Anti-science : Uses unscientifc reasoning applied to scientific, factual matters. Denies the facts on the basis of opinion and unverifiable inner knowledge : "But relativity is way sexier than MOND, so it's clearly true !" Not always about replacing science with God or a magic duck or whatever (though often that's the case) - can also occur within science. Distinct from, but not mutually exclusive with...

Scientism : Applying scientific reasoning to unscientific matters, believing that science will answer all questions eventually. An unjustified insistence that the world is essentially nothing more than atoms bashing about and that even questioning this assumption is anti-science and eventually we'll be able to objectively determine if The Lord of the Rings is better than Star Wars and anyone who keeps disagreeing at that point will just need to be shot or drowned or fed to a flatulent bear or something.

(What really ticks me off about scientism is not the idea that reality is only what we can observe and measure - I'm fine with this, it's a perfectly sensible assumption to make... but dammit, it's an assumption. You can't just say, "this inherently unprovable assumption is a fact because I really like it", because that's exactly the same sort of stupidity that afflicts the blindly faithful.)

And to round off that little tirade about what's wrong with the world, it needs to be emphasised that people can merge the scientific, unscientific and anti-scientific into one glorious paradox. Yes, there are people out there who are truly anti-science, but there are plenty more with just one or two anti-scientific beliefs. Even Creationists occasionally just have an unscientific blind spot, performing entirely worthy pieces of scientific analysis but never truly accepting them on the grounds that it's all God / Satan / a magic duck playing an elaborate joke. And scientismists can insist that only the scientific approach is the absolutely correct approach on one very specific issue (particularly atheism) whilst happily being anti-science on others.

Most people, I think, are a bizarre and inscrutable mixture of these characteristics. They are walking puzzles wrapped in enigmas wearing woolly hats and sandals - God knows why - that utterly defy rational analysis, though many of them would vehemently deny it. Internet-based communication tends to filter out all that baffling self-contradictory complexity and replace it with only the opinions they held strongly enough to write about, thus ensuring that we perceive everyone in a bizarre and unflattering light. Yay.

Then again, a good many people are just monkey-fondling arsewipes with the cerebral capacity of crack-addicted donkey that's been forced to join an elite team of Morris dancers. They wouldn't know a persuasive argument if it convinced them to donate all their life savings to rescue a Nigerian astronaut. But I digress. What was I talking about ? Ah, yes, theories. We're almost done. Just one last point remains.


Variability

Facts don't change, except in the trivial and boring sense of measurement errors. Laws and principles are more subtle. Laws aren't based on just any old observational facts, but on those subjected to repeated, independent testing under carefully controlled conditions by multiple experimenters. That's as good as we can ever establish anything observationally. Like facts, the only way to deny them would be deny objective reality. And that's fine, but it's unscientific and often anti-scientific as well*.

* Those who acknowledge when they're being rational or not tend to be able to make sensible calls as to when this is appropriate. To re-emphasise, it's those who apply science and un-science inappropriately who cause trouble.

Laws are based on facts, so they cannot be wrong - but they might be incomplete. Newton might have performed his experiments on motion on a flat horizontal smooth surface, but he could never have envisaged that he'd be operating within curved spacetime. Principles, as we've already seen, can actually be disproved completely.

Theories can be modified by changing evidence. Where you draw the line and decide, "this simply doesn't work" is, just as with the supposed condition of being, "well tested", incredibly murky and can't be rigorously defined. And again it's deeply unsatisfying, but that's just tough on us.

Theories can also wax and wane in credibility and popularity as evidence changes. However, while the sort of day-to-day investigations of actual front-line research churn out different ideas and hypotheses every five minutes, it's very hard to shift those at the extreme ends. Astrology isn't going to be welcomed back into the scientific fold just because a Capricorn felt anxious about an upcoming decision on Tuesday; relativity won't be overturned because some retired engineer has an equation that they like. Saying that relativity is only a theory is to make a mountain into a molehill : bloody daft.

You would think that this is a powerful argument against the anti-science crowd : look, we change our minds all the time, but not about these things because we've tested them really, really well. Alas this underestimates the human capacity for bullshitting and a tendency towards really extreme beliefs. Deniers insist on believing that because we're changing our minds that means we're incompetent, and when don't change our minds we're dogmatic.

It's a similar story for that famous "97% of scientists believe in human-induced climate change", or whatever it was. For those who understand how science actually works, as a series of international competitive collaborations where major discoveries are rewarded not punished providing the evidence is compelling, this statistic is a powerful validation of the CO2-based theory. For those who don't, it causes bullshitting in the extreme. Suddenly every halfwit under the Sun professes to be a statistical expert because that poll just can't possibly be right and even if it was right it's just evidence of dogmatism apparently on a universal scale so it's useless anyway.

Likewise, assessing where various ideas stand on a plausibility index of credible theories is fine as a methodological teaching aid, to show people how ideas change, how the evidence level varies, and most importantly how murky the real world is (especially at the forefront of active research) rather than the rote-learning version of science sometimes emphasised in schools. But I wouldn't want it adopted en masse as a communication tool. People would say, "hah hah, this theory's best, anyone supporting any of the others must be stupidy stupidy stuuupid !". Or they'd accuse anyone who supports the most popular theory (regardless of why it was popular) of being a dogmatic tool of the establishment. Or that because hardly anything on the scale was truly certain, all of science must be just useless. And if you really forced their hand they'd say, "it wasn't like this in the old days, mind you." Some people insist on absolutes because they are unable or unwilling to think in any other way.

Such people demand objectivity where little or none can exist. There's no point arguing with such people. While every effort must be made to explain as much research as possible in a public-friendly way, ultimately there are aspects of research that requires years of training to understand - and often then isn't fully understood by anyone. Yet the public still have decisions to make : at some level it comes down simply to trust. For that to work we need to explain the methods and practise of science, to clearly explain which parts are really objective and which parts are murky. Taking the scientisimists approach of pretending that there is no murk doesn't work because people know full well that it's hardly ever as simple as that.


Summary

This has been a long-winded and somewhat complex post, but the main points boil down to :

• "Theory" does not have a specific meaning to scientists. It's a very ambiguous word all-round.
• "Hypothesis" is a much safer word. You can't really go wrong with this one, but it's limited.
• Some theories are clearly very much better than other theories. No satisfying word exists to distinguish the very best theories from the everyday ones that theoreticians routinely investigate.
• In a very few cases, we might - and I stress, might - be able to get away with using the word "law" instead, e.g. "the law of evolution" and perhaps, "the law of CO2-induced climate change". We still don't know all the details of these, just enough to say that the basic explanatory mechanism* behind them is fundamentally correct.
• ... but these are rare exceptions. Most of the time we should be incredibly wary of declaring the underlying principles to be true, even in very strong cases like relativity.
•  It would be great if did have a word describing the really elite theories, the ones which have been extremely well-tested and where the explanation is highly likely to be correct (but not quite certain). But "well tested" cannot be defined rigorously; we must accept a degree of subjectivity.
• The idea that "theory" itself means only these most polished gemstones of all possible models has, perhaps, been spread about by scientismists seeking to redefine knowledge. But it has spread simply because it's a plausible idea which sounds believable.

* And I mean basic. As in, we know that greater mass equates to stronger gravity, even without knowing how gravity operates.

Yelling, "it's only a theory !" is generally done out of ignorance and misunderstanding. But yelling back, "that's one of the dumbest things you could possibly say !" is also a woeful misunderstanding. Listen up people, and listen good : it's not that simple. There is no simple, special definition of theory reserved for members of the Junior Science Club. This crude answer helps no-one, and in actuality conveys a damaging message that reinforces the kind of science people learn when they're five : the idea that science is only ever about right and wrong answers. 

Facts are of fundamental importance in science, of course. But they are very far indeed from the whole story, and reinforcing this child-like idea that many people have of science is damaging. Conversing with strange people from the magical land of The Internet sometimes leaves me thinking that we're between a rock and a hard place : on the one hand, we have people who think science is useless because it's forever changing its mind, and on the other we have the equally idiotic who think that all science is dogmatic groupthink pseudo-religious doctrine.

It isn't either of those things, it's the happy middle group. Possible sinister origins aside, crude definitions of words only add to the confusion. Of course the scientific world view changes, it's supposed to do that because it's supposed to look for new and challenging evidence. People need to understand that science is all about those shades of grey, much more than it's about those rare streaks of black and white. This does not mean that all theories are equal. It does not mean that some theories don't have literally mountains of evidence in support of them. And it certainly doesn't mean that you're entitled to pronounce judgement after 30 seconds of reading silly newspaper headlines that you don't like.

I understand the point people are trying to make when they say that there's a special scientific meaning to "theory" : they want to emphasise the extraordinary degree of testing that some theories have survived, and that these particular cases are very much harder to shift than the run-of-the-mill stuff that's part and parcel of everyday research. That's very much a good thing. But to steal a commonly-accepted word in this way is counter-productive : a) this special meaning in science is a myth; b) it's (sometimes) an attempt to redefine knowledge; c) it's bloody confusing; d) when people hear scientists using this word in a variety of different ways, after being told it has this one special meaning, they aren't going to be best pleased. The problem here is the method, not the goal.

I'm all for rigorous definitions wherever that's possible and appropriate. But in the case of "theory" I think that neither is the case. The simple truth of the matter is that science isn't always purely objective. It's a murky, messy business of turning facts into models, testing those models, rejecting some while provisionally tolerating others. Pretending that it's more objective than it actually is won't work, because it simply isn't true. Would it be nice if it was ? Sure ! But that's not what it's like, and that murkiness is sometimes what makes it fun.

No definition will stop the most ardent from bullshitting about science, because these people simply do not care - and you can't argue with someone who doesn't care, you can only have shouting matches. But for the rest, let's not set ourselves up for disaster by pretending we know things we do not. Simply admit the plain truth of it - that we know hardly anything for certain, but we're far, far more confident about some things than others. If this leaves people feeling lost and insecure, then that would be a good start. Perhaps (and I say this cautiously, knowing how damaging bullshit and stupidity can be) then they'd stop the chest-thumping for a moment, begin to realise that not everything can be quantified, and actually learn how to think.

Communicating science is as difficult and perilous as actually doing research. Over-simplifying the process leads people to believe they understand it as well as the experts do, which they don't. Under-simplification leads to accusations of elitism. It's a murky path indeed, and sometimes the murk is full of werewolves and crocodiles - but in the end, the choice is actually very simple. We can either try and wend our way through all the rich murky complexities, risk being mocked or making mistakes, the anger of bullshitters and idiots who have no real clue what they're talking about... or we can stay at home, quiet and ignored and learning nothing.