whats the definition (or difference) between a theory, a law, and a principle?
EX: einstein's theory; keppler's law, archimides' principle.

A law describes what is observed. A theory attempts to explain why or how it is happening. If you ask me.

Fred

__________________
"For shame, gentlemen, pack your evidence a little better against another time."
-- John Dryden, "The Vindication of The Duke of Guise" 1684

Semantics. . .and a dash of politics. But this is my 'off the cuff' definitions:

A law is a static observation. If you have 'A' then 'B' will be the observation.

A theory is a process observation. If you do 'A' then 'B' will be the result.

A principle is a direct correspondence. If you observe 'A' you will also observe 'B' at the same time.

The English language being quite flexible, your terms ultimately mean nothing in terms of actual science.

__________________
Keeper of the Jabberwock

i think i read many years ago that a theory is a statement which has not yet been proven to be correct at all times, in all circumstances.. if and when the statement is proven therein, a "theory" may be upgraded to a "law" by a governing body representing the science for which the theory was written.. is this antwhere near familiar, or just a crock?

not as easy a question as it might appear ...

as I understand it:

a scientific (or natural) law is a universally consistent description of an observed process of nature - a confirmed observation which falls outside of a law undermines that law ...

a scientific theory is a testable and consistent description of the mechanisms which underpin scientific (or natural) laws, and from which predictions regarding observed processes can be drawn - the how, why, what, and when to expand on Fred's description ...

a scientific principle is a rule which underpins the practice of science ...

I'm not sure about that ... but a law can be formulated from a theory where it is shown to be universally consistent (eg, the Law of Conservation of Mass and Energy) ...

__________________

It is central to science that no statement is ever proven to be correct.

I look at it like this: a law is a summary of a bunch of observations. The Law of Gravity says that stuff with mass attracts other stuff with mass. It doesn't say why or anything like that. It just says that it happens, and maybe describes how it happens, maybe that F = Gmm'/d^2.

A theory is a testable explanation of how or why. So the Theory of Gravity might say that the Law of Gravity is true because it's a limiting case of the Theory of General Relativity, and here are a bunch of experiments we can do to test that. So a theory might encompass a number of laws and theories.

A principle is more like an assumption, or an untestable theory, something that you just say is true (usually based on some observations) and then derive theories from it. It's like a law, but maybe was developed before there were enough observations to really call it a law. Maybe you don't know the limitations on the statement, so you call it a principle rather than a law, and then the name sticks.

But of course those aren't actual definitions or anything. It's just what pops into my head when I see the terms.

__________________
"It's turtles all the way down."

This is illustrated by the fact that many laws are named differently in other languages. Eg Steiner's theorem (parallell axis) is in Norwegian named using a word that roughly translates to "assertion" in English (Steiners sats). It's really as you say just politics and history.

This is an incorrect statement that has unfortunately been taught in school (I vaguely recall hearing it somewhere in grade school.) It is false. Theories do not become laws. Laws do not become theories.

Laws are descriptions of observed behavior. Boyle's Law, Charles' Law, Ideal Gas Law, Law of Gravitation, Laws of Conservation. They are typically expressed mathematically.*

Theories are explanations of how things work. They take Laws as their foundations and attempt to build a cohesive explanation. They typically incorporate several laws, plus other observations. They usually are built of words, because the intent is understanding, not calculations.

Hypotheses are educated guesses, starting places, concepts that haven't been validated for accuracy. A hypothesis may be a pre-theory or a pre-law, but as typically thought of they are pre-theories.

As I understand it, Principle was an alternate word choice for people who felt "Law" was too hubristic. There may also be truth to the description as an assumption or untestable description of behavior, similar to a Law but on less-solid footing. However, I don't think the Heisenberg's Uncertainty Principle is considered unsupported anymore. I think it's treated as a Law.

Einstein's theory of General Relativity is an explanation behind gravity as a bending of spacetime.

Kepler's Laws are mathematical descriptions of the motion of orbiting bodies as determined from observations.

Archimedes' Principle is alternately called the "law of upthrust".

* It is important to note that most laws are constrained by assumptions that must be valid for the law to be valid. E.g. the Ideal Gas Law only applies to an ideal gas, which is a string of assumptions about the make up of the gas.

you're right ... but I have no idea who first came up with that term for it ...

I thought that Archimedes' Principle (derived from observation and experiment of relative buoyancy and fluid displacement) was one of the guiding principles which led to Newton's 3rd Law?

__________________

Yes, it's not terribly useful to think of "laws" as "more certain" theories. Indeed, often the opposite is true-- we tend to keep the laws even after the theories invented to explain them have been compromised, because laws are fundamental building blocks that theories attempt to explain, so have a "life of their own". In other words, when a law is found to have limitations, we just modify the law, thereby deflecting the blow, whereas the theory invented to explain that law receives a body blow that can knock the wind out of it.

I see the general distinction in the OP on a scale of how complex and abstracted each is from the observed phenomena, and how far it advances our understanding (understanding also being a key dimension mentioned by Irishman). If all it does is quantify a phenomena using some simply defined measurables, while still applying in a variety of situations, it is a law, even if it does not generate a lot of simplification because it still just explicitly describes what we see. If it unifies a number of seemingly separate laws, then we are gaining unification and understanding, at the price of additional abstraction-- that's a principle. And if we go whole hog, piecing together several principles using abstract assumptions about how they interact and when and why they hold, then we have a theory.

A simple analogy might appear if you imagine a future archaeologist digging up one of our houses and investigating how it was built. The laws are like the nails, there for all to see, that the archaeologist quickly notices are very important but may not understand how they got there. Then the principles are like the concept of the hammer that the archaeologist at some point realizes must have been used to hammer the nails in, and the theory is like the whole process of the right way to use hammers on nails.

remoat,

A law is the way that it is. A principle is a description of how the law works or and example. A theory wants to be a law but lacks the principle.

I think Irishman did a nice job of summarizing the difference between a law and a theory. Carlo Lastrucci in his book "The Scientific Approach" (1967) states that historically a Principle was synonymous with a natural law.

I think if a distinction is to be drawn between a "principle" and a "law" it might be along the lines of what snarkophilus said - principles often have a component of "assumption" that is not present in laws. Laws such as those listed by Irishman tend to be mathematically represented, but lack the ability to explain and therefore are not theories.

However, Principle's often are simply taken as true so that science can function. The best example is the principle of uniformitarianism - or uniformity of natural laws. Uniformitarianism is typically discussed in the context of geology, but it really applies to all science. The scientist must assume that the processes and natural laws that operate today have also operated in the past (though not necessarily at the same rate). So there is a component of observation to this principle (what we observe happening today) and a component of assumption (what we see today can be extrapolated to the past).

What I really wish we could abolish is the two word definition "educated guess" for hypothesis. Students all memorize this - but they haven't a clue what it means or what a hypothesis really is. A hypothesis is a proposed explanation for an observation/phenomenon or set of observations/phenomena that can be experimentally tested.

__________________
"The scientist who asks the right question reconnoiters a new patch of the unknown, and may, with luck, bring it within the constricted but expanding boundaries of the known."

~Timothy Ferris (The Red Limit) 1982

I agree-- there's a misconception that the point of the exploratory part of science is to guess right, when in fact scientists don't care if they are right or not when they test something (unless it was their own theory to begin with!). The hypothesis is just a way to focus the test, and the result of the test stands on its own. I suppose one could argue that one purpose of science is to make correct predictions, so the hypothesis could be viewed as a prediction, but that doesn't sound like basic science discovery, it's more like science application. A tricky distinction to make, but I agree that "educated guess" is misleading as to what the scientist is really trying to do-- which is basically to ask nature a question and listen real hard for the answer.

Science does not explain. It predicts. Explaining is for frauds like astrology.

__________________
"A witty saying proves nothing" Voltaire.
"All your bias are belong to us" Ara Pacis.

I think it explains. It just explains in terms of predictions, instead of in terms of handwaving and lies.

For instance, heat moves through a solid at some rate. Predicting is just getting that rate based on the structure of the solid. Explaining is saying that the heat moves because of atomic collisions, lattice vibrations, et cetera. The difference between science and astrology is that in science, you'd have actually observed (or had a solid theoretical basis for believing) that those mechanisms are plausible, and there'd be some way of predicting them, too. In astrology, that second step wouldn't be taken.

__________________
"It's turtles all the way down."

My problem with that formulation is that plausibility is not observable. An explanation is either testable or non-testable. In the latter case, it's a nonscientific explanation, with (perhaps) nonscientific plausibility. The only reason you know your explanation is plausible is because you derive accurate predictions from it.

__________________
"A witty saying proves nothing" Voltaire.
"All your bias are belong to us" Ara Pacis.

You're quite mistaken on this. Natural Laws describe what is observed (usually by some mathematical relationship) whereas theories attempt to provide a mechanism that explains the observations. That mechanism is an attempt to explain the reason for some observed phenomenon. If the mechanism of a theory is viable, then of course it will make testable predictions.

Moti Ben-Ari seems to agree. From his book "Just A Theory" (2005):

__________________
"The scientist who asks the right question reconnoiters a new patch of the unknown, and may, with luck, bring it within the constricted but expanding boundaries of the known."

~Timothy Ferris (The Red Limit) 1982

Most real-life natural laws arise from some sort of rudimentary "understanding" of the mechanism which they describe. And no so-called "explanation" would ever be accepted unless it made testable predictions. You yourself equate "viable mechanism" with "testable predictions":

The two notions, of "describing/predicting" and "explaining", are joined at the hip, sorry.

__________________
"A witty saying proves nothing" Voltaire.
"All your bias are belong to us" Ara Pacis.

Really? Newton's laws require no understanding of the mechanism. The laws of conservation of mass and energy require no understanding of the mechanism underlying them.

Two different sources give similar descriptions of a scientific law:

And from Lastrucci's "The Scientific Approach"

Nowhere in these definitions is a hint of mechanism required. Scientific laws are descriptive - usually mathematically. No explanatory power or predictive power is required - other than the prediction that nature will behave according to the laws.

And ... your point? ... you're just re-stating a position consistent with what I've said and implying that I disagree with it.

Earlier you said that ... well here is the exact quote



But now you're saying that science does explain because it is linked to prediction. So you're making statements that illustrate I was right when I said theories explain, but you're trying to imply I was wrong.

Look carefully at what I wrote initially. I never said theories do not predict. I simply said that they do offer explanation - and that is part of what a theory does as you now admit.

__________________
"The scientist who asks the right question reconnoiters a new patch of the unknown, and may, with luck, bring it within the constricted but expanding boundaries of the known."

~Timothy Ferris (The Red Limit) 1982

I didn't see this post earlier. I think you're right that there is a very fine distinction in some of these terms. If a hypothesis is proposed, there pretty much has to be a predictive part to it - or how do you test the hypothesis? So it can be hard to get students to understand what a hypothesis is. Is it a guess? Is it a prediction?

I think if teachers are going to use the term "educated guess" then they need to take a different approach with the term. For example, ask any class what a hypothesis is and you'll get "educated guess" from most of them. So at that point the teacher needs to build upon what the students do know by asking something like this: "Yes, a hypothesis is often described as an educated guess, but what does that mean?" Then use some kind of discrepant event via a demonstration and start brainstorming some hypotheses that might explain the observation and means of testing those hypotheses.

I just worry that in science education there is too much emphasis on what science has uncovered and not enough emphasis on how that knowledge was uncovered. Of course that is a product of standardized tests that focus almost exclusively on the what. So the teacher really needs to infuse the how in the process of teaching the what.

__________________
"The scientist who asks the right question reconnoiters a new patch of the unknown, and may, with luck, bring it within the constricted but expanding boundaries of the known."

~Timothy Ferris (The Red Limit) 1982

I entirely agree, although possibly not for exactly the same reasons. In several threads I've found myself having to defend the idea that scientific conclusions rely on the scientific method for establishing their virtue, and have no independent meaning. Many tend to think that scientific discovery is simply "the truth", and there's no need to analyze how this truth was arrived at or what limitations are inherent in such an approach. It's like memorizing theorems and thinking one is doing mathematics.

That's nonsense. Newton didn't just record a mountain of data, and then extract his laws from the data. He got his "laws" from reasoning about how bodies fell. From trying to understand the mechanism.

Blind data crunching will get you nowhere in science.

__________________
"A witty saying proves nothing" Voltaire.
"All your bias are belong to us" Ara Pacis.

First, are you now conceding that theories involve explanation?

Second, what I said is not nonsense. It's the way it is. Newton's laws of gravity are a mechanism free mathematical formulation. They describe how bodies fall. They do not provide a mechanism that explains why bodies fall toward a center of gravity. Einstein's theory of gravity attempts to do that.

Of course Newton's effort to develop laws of gravity would require reasoning - mathematical reasoning - as well as measurement and observation. All that's critical to the process of science. I never said anything about blind data crunching. Very poor interpretation of what I'm saying on your part.

And how do you know that Newton's efforts to conceive of a mechanism to explain gravity had any important impact on the development of his laws? His theory of gravity was wrong - and his laws do not require a mechanism to predict the behavior of objects subjected to the gravitational pull of other objects. Where is the evidence that any speculations he made about the mechanism of gravity steered the development of the laws of gravity. Perhaps you're right on that, but you're going to need to back it up with something substantive.

__________________
"The scientist who asks the right question reconnoiters a new patch of the unknown, and may, with luck, bring it within the constricted but expanding boundaries of the known."

~Timothy Ferris (The Red Limit) 1982

I probably shouldn't jump in, except to say that I don't think you two are disagreeing as much as you think-- it's pretty hard to say anything in regard to scientific theories that is entirely right or entirely wrong. Science is more like the skin of an onion than like that of an orange-- there really aren't such clear delineations between what is an explanation and what is a description. For example, Newton's theory of gravity did provide a mechanism-- it identified gravity as a force, and that did lead to understanding. That the understanding was not exactly correct isn't that big of a deal-- happens all the time. But Einstein's theory is indeed at a deeper level-- the next skin down in the onion. So let's not split hairs about what is an explanation and what isn't, there probably isn't much value in trying to draw a line in the sand on that. Even Einstein's theory is only explanatory to a point-- why does gravity bend spacetime? At some level, all theories become purely descriptive, the real issue is how deeply do they underlie other theories and laws-- again the onion.

You make some really great points here. Though - and I'm not trying to be difficult - I've learned from debates here and elsewhere that much disagreement occurs and time is wasted when two people do not agree on the definition of a term. Sometimes those disagreements occur because one person has an incorrect definition they're working from. Sometimes the difference is because the two people are thinking of two subtly different usages of the same word.

There are certainly enough borderline examples relative to these definitions that it is critical to pin down what the standard definitions of the terms being debated are.

This disagreement between myself and DA started with DA suggesting that theories do not explain. I supported my statement that theories do have an explanatory component. I'm hoping DA will confirm that we've reached agreement on that aspect of what is meant by a "theory".

And now we seem to be disagreeing slightly over the scope of a natural law. But I think you are right. I don't think I disagree with DA as to what science is, but I think we may disagree as to the exact boundaries of these definitions.

__________________
"The scientist who asks the right question reconnoiters a new patch of the unknown, and may, with luck, bring it within the constricted but expanding boundaries of the known."

~Timothy Ferris (The Red Limit) 1982

I completely agree that definitions are very important. Probably the most significant aspect of any discussion is to decide, if not what is the "correct" use of the words, then certainly what is the "most productive" use of the words. And that is worth establishing, even if to some it sounds like "just semantics". I just don't want to see your discussion on this break down into battle lines based on black-and-white distinctions that don't really exist. We've all seen plenty of debates that caused time to be wasted on that score, I'm sure!
I'm not really sure the intent of that claim, but DA seems to have retreated from that position, sort of like a platoon retreating from an exposed open field under incoming machine gun fire. But the idea that Newton's laws had no explanatory component is kind of like you advancing into that same field-- it's starting to sound like trench warfare in WWI, and we all know where that got them. I'm not really sure what is actually being disagreed on-- theories are deeper efforts at explanation that link laws and principles into a kind of coherent scaffolding (which sounds like your point), yet the rock on which they rest is their testable predictions, not the sense of understanding they convey (which I think is Disinfo Agent's point). The sense of understanding is an important element that individuals use to derive value from science (your point), but that part is subjective-- the testable predictions are the objective part that allows it to be identified as science (DA's point). I don't see anything incompatible there, holster those guns!

That's certainly what the thread is all about, and I think you both have valid points to make. I'm not claiming that "law" and "theory" are arbitrary distinctions, even though they have often been mislabeled (that we have "Newton's laws" but "Einstein's theory" is to me more a reflection of the historical evolution of our appreciation for the fallibility of science, but that's unfortunate because a useful distinction can be made). Sadly, they are both poorly understood by nonscientists-- a "law" sounds like appeal to authority, and a "theory" sounds like appeal to convention, and neither are.

I could not post sooner because the real world kept getting in my way.

Well, Ken G has succeeded in explaining my point of view better than I ever could hope to. Thank you!

My feeling is that description and explanation always go hand in hand in science, and so one of them is redundant. Since "explaining" the world seems to be a favorite of certain strands of pseudoscience, or dubious science ("the reason why you lost your job was because Saturn was in the house of Aquarius"; "the reason why you can't hold a job is because of your Edipus Complex"; that sort of thing), I prefer to keep the "descriptive" label, and drop the "explanative" label. To me, the concept of science as a method for explaining the natural world has always seemed incomplete, and even misleading. Philosophy also explains the natural world in its own way; religion also explains the world in its own way; children's stories also explain the natural world in their own way...

I see the explanative component of science as no doubt inspiring, but still akin to myth-building. You gather a couple of facts, and build a plausible story around them. Which can be a very powerful thing, but the trouble is that different stories can be compatible with the same facts. We see this today in quantum mechanics. What is the explanation for subatomic interactions? The Many Worlds interpretation?... the Copenhagen interpretation?...

Less philosophically, I have my doubts that the sequence law --> theory is actually followed by scientists when they're doing research. I think that at least phychologically the motivation of scientists is always to attempt to understand (=explain) the natural world. To imagine them getting out of bed and telling themselves "O.K., this semester I'll just focus on finding the natural laws. I'll try to discover the theory next semester" seems like an oversimplification of the scientific method to me.

At the same time, though, good scientists should be aware that the most compelling explanation of all is worthless if there is no way to confront it with the data (that is, to test it). This is why I prefer to frame science in terms of making accurate descriptions and valid predictions of natural phenomena. Admittedly, this comes down to a personal peeve of mine.

__________________
"A witty saying proves nothing" Voltaire.
"All your bias are belong to us" Ara Pacis.

Newton's proposed mechanism was a magic genie in a bottle. He said gravity was some tendency for matter to mutually pull together. He said it was a "field", without providing a mechanism for the field. Yes, he identified gravity as a force, but he also defined force, and so by definition the tendency of matter to clump together creates forces on said matter. So as far as mechanism, no, Newton did not provide much mechanism. He did, however, provide some powerful mathematical descriptions of behavior of bodies, including the gravitational attraction of matter and linear motion.

And I disagree. This is a clear cut example to me of the use of the terms "law" and "theory" in the modern distinctions that I explained above. Newton's Law of Gravitation is a mathematical description of how bodies move. It does not say a thing about why they move that way, what causes the attraction. Einstein's General Relativity discusses the fabric of space time and gravitational curvature. Thus it is theory.

Except that he didn't actually admit to retreating from that position. Isn't that one of them rules of behavior, acknowledging errors and such?

I'm the one who brought up the terminology in this thread. I think it serves a valid point to distinguish between the purpose of scientific laws (mathematical formula that summarizes how things behave) and scientific theories (attempts to understand the structure that makes the laws as they are). Yes, in science both are tied to experimental validation rather than any touchy-feely emotional satisfaction. That was never the point or the use of the terms in this discussion, so your statement that science does not explain was misdirected and incorrect by the use of the terms in this thread. That is the point that dgruss23 has been trying to make.

I do agree with this summary by Ken G
You are correct that science in application is not nearly so streamlined as the "ideal" sequence layed out. That sequence is primarily to separate the concepts of the activities that go into science and show how they interrelate. In practice, the sequence is not a linear progression, but rather more of a criss-crossing spaghetti bowl. The theories and the laws are both hypotheses that the scientists come up with and continuously bounce off each other, making modifications wherever appropriate. It is the dynamic interaction that drives the valid (i.e. successful) results to the top and the invalid (i.e. conflicting with data) to the bottom.


hijack
You should apply this better to your ongoing argument in that blog thread.

Really - I'm not trying to be difficult here - but I have to disagree with you on this. While it is true that description and explanation go hand in hand, I do not agree that one of them is redundant - because they are not the same thing.

For example, I think we can agree that the scientist today would argue that evolution and natural selection go hand in hand. But natural selection is a mechanism (not the deepest level of the onion but an important layer) proposed to explain the observations (descriptions) that have led scientist to the conclusion that life on Earth has and does evolve.

My point is that description and explanation are not always the same thing. Observations and natural laws describe the universe and its components. Theories are our attempts to explain them. The key component of the theory that leads to an attempt to explain is mechanism.

I'm willing to accept KenG's proposition that we shouldn't be too black and white - that there can be an element of explanation within natural laws. His example of "force" being a mechanism within the context of Newton's laws fits that. However, the deeper you penetrate the onion the more you leave the realm of natural law and enter the realm of theory. So just as we should not be too black and white with our definitions we should not be too gray with our definitions either.

The significant difference between a natural law and a theory is that natural laws are primarily descriptive while theories rely heavily on providing an explanatory mechanism that makes testable predictions. I do not believe that we should allow the small areas of overlap between description and explanation to let us conclude that to use both terms is redundant.

But I think this is an example of attributing incorrect/pedestrian usage of the terminology by pseudoscientists to a problem with the terminology when it is really a problem with the pseudoscientists. The examples of "explanation" by astrologers you give are not scientific.

A scientist can choose to focus his/her research on the descriptive label. That is empirical science. But the scientist cannot just drop the explanative label entirely from science because the two labels are not synonymous.

Descriptive answers "what is/are" type questions: What is the form of the gravitational relationship between two bodies? What are the characteristics of spiral galaxies?

Explanations via mechanisms are attempts to answer the "why" type questions. Why does gravity pull? Why do spiral galaxies have spiral arms?

If science was purely descriptive there would be very little need for theorists.

I agree with this. Science, philosophy and religion are different ways of seeking understanding. That is why it is so important to be precise with what these definitions mean to the scientist. And that is why I don't think we should object to a term such as "explanation" being used when we talk about a scientific theory. The scientist applies a different expectation to that term than the pastor.

I agree. That is one of those misconceptions about science that can be avoided if the difference between a natural law and a theory is carefully defined and then explained.

The better your grasp of nature's "laws" and observational "facts", the better the position you are in to develop a theory. Of course scientists are aware of both theory and observation, but many scientists focus on the empirical side of the research. Others focus on the theoretical side of science.


Ooh, I really like that! I think that might be the best one sentence definition of science I've ever read.

__________________
"The scientist who asks the right question reconnoiters a new patch of the unknown, and may, with luck, bring it within the constricted but expanding boundaries of the known."

~Timothy Ferris (The Red Limit) 1982

It seems we are all in agreement on many points. Just one remark:

Can't something similar be said about General Relativity, though? Einstein says that the apple falls down to the ground because mass makes spacetime curve. Well, but by what mechanism does mass curb spacetime? That isn't explained; it's Einstein's own magic genie in the bottle. A genie behind another genie, like Russian matryoshka dolls.

__________________
"A witty saying proves nothing" Voltaire.
"All your bias are belong to us" Ara Pacis.