Pique the Geek 20100725: Corruption of Scientific Terms

(10 pm. – promoted by ek hornbeck)

Scientific terms are often corrupted, and the wingnuts often do it.  They conflate hypotheses with theories, and theories with laws.  They also reduce the value of a theory to what they make out as just a guess.

This post is an attempt to separate the words and make the scientific method more sensible to folks who are not trained scientists.  As always, if I not clear, comments and questions are always welcomed.

Words are often taken the wrong way.  Boy, do I know that!  But let us try to take some the correct way, and those are about physical manifestations around us.

There are several words in science, and amongst them are Laws, Theories, and  Hypotheses.  These are often used incorrectly.  For the sake of completeness, I have added guesses and speculations.  I have used all of these concepts in my scientific work, and now shall attempt to define them in my own words and experience.  For the sake of impartiality, I have also included other workers’ definitions of the first three in the discussion.

In reverse order from the list just above, starting with the least precise term, let us examine each of them.  It is difficult to determine whether a guess or a speculation is less precise.  My gut feeling is that a guess is a little looser than a speculation, but that is only my personal take.  Let us start with a guess.

A guess is just that.  Now, some guesses are better than others, and the better guesses come with experience and insight.  Guesses often lead to more firm areas, but to be truthful, most great scientific concepts start with guesses.  The great advantage of guesses is that they need no proof, no evidence (other than intuition), and no experimentation.  Guesses form the basis for more well thought out concepts, and most of them are discarded soon after being made.  Sometimes, though, even incorrect ones last forever.

For example, Benj. Franklin guessed that the carrier of electricity had a positive charge, later shown to be incorrect.  However, except for algebraic sign it makes no difference, and electrical circuit mathematical development in engineering still uses that convention.  It was a bad guess, but not a fatal one.  In science, guesses are often discussed with colleagues and are often discarded or modified, but guesses are rarely published because there is no evidence to support them.  In the nonscientific world, folks guess all of the time.

The problem with guesses is that they are often affected by the prejudices of the person making the guess.  This can often be troublesome.  However, when experience and previous data are consistent with the guess, they can work very well.

For example, Sir Humphrey Davy guessed that the alkali metals were elements in the very early 1800’s, and was correct. He isolated sodium and potassium using an electric battery from their molten hydroxides, then proved that they could not be further changed into more fundamental chemical substances.  It was just a guess, but it was a good one.  By the way, the scientific method was in its infancy at the time.

A speculation is, to me, a bit more than a guess.  A speculation, while not formal at all, tends to have a bit more structure to it than a guess.  In a speculation, oft times there is a bit more background information than there is for a guess, and thus a speculation seems just a bit more based in observation.  However, the division between the two is not well defined at all, and for most purposes they can be treated as pretty much the same thing.  Those of you who are scientists may think that a speculation is weaker than a guess, and I will not argue since the difference is really pretty subjective.  The best way for me to explain the difference, at least to me, is that a guess involves what is happening, and a speculation involves why it is happening.  I suspect that this paragraph may cause some lively discussion.

Now, after these informal categories come the formal ones insofar as the scientific method is concerned.  These are the ones that cause the problems when lay people (absolutely no insult intended towards readers who are not scientists, for reasons that will become clear as the discussion progresses).  The reason that they cause trouble is that even scientists do not agree completely as to their definitions and distinctions (as will be seen in Dr. Hawking’s explanation of a theory).

Generally, it is agreed that the most basic category of explanation for a phenomenon is an hypothesis.  (Note that I use “an” rather than “a”, just as some folks say “an herb” rather than “a herb”.  Neither is more or less correct, but rather a matter of personal preference).

Here is one definition of an hypothesis:

A hypothesis is a proposed explanation for an observable phenomenon.

That is not a bad definition, and much better than the one so often used, an “educated guess”.  A guess is an educated guess, but an hypothesis is much more.  Implicit in the definition are several things, in accordance with the scientific method.

First, an hypothesis must explain something that can be observed and verified.  In other words, a viable hypothesis will have predictive value when similar phenomena are examined.  If the predictions hold, then the hypothesis remains valid.  If not, then the original hypothesis must be reexamined, and modified or discarded for a better one.

The reason for this is that an hypothesis must have some amount of generality to be useful.  If an hypothesis explains only one or two observations of a phenomenon in a very limited situation, it is too restrictive to be of much value in the grand scheme of things.

Second, an hypothesis must be reproducible when conducted by other investigators under similar conditions.  This is critical, because not only does this requirement greatly reduce falsification of data (observations), it also helps to eliminate systemic error introduced by any one investigator.  Data falsification does occur, even we wish that it did not, but is much more rare than experimental errors.  Testability reduces both of those eventualities.

Third, an hypothesis must be consistent with earlier observations and hypotheses, unless those observations and/or hypotheses were incorrect.  This is often the case, by the way.  Either way, one of the two conflicting hypotheses are discarded, and sometimes both are.  As understanding of the phenomenon becomes greater, and as more precise observations are made, better hypotheses are formulated.

After a particular hypothesis has been tested many times by many investigators and found to be without conflict, it may become a theory.  However, just because something is never shown to be incorrect does not mean that there is a causal relationship between the data going into the hypothesis.  Here is sort of a silly, but telling example.

Translator’s hypothesis about carrots and criminal behavior.



Observations show that in every case, persons convicted of felony crimes had drunk beverages containing water within 24 hours of commission of the crime.

There is nothing incorrect with the observations.  The hypothesis is completely bogus, however, because it purports to show a causal relationship betwixt drinking water and criminal behavior.  This may sound silly, but such false hypotheses are used all of the time by those who try to influence opinion.  Political types use this device to a great extent (on both sides of a given issue, by the way) and it is important to think for one’s self to determine whether or not a causal relationship exists.  I hear on wingnut talk radio all the time that “tax cuts always bring economic improvement” and one or two examples are given.  However, never mentioned are the cases wherein tax cuts did NOT produce economic improvement.

After an hypothesis has been tested many times over, and no experiment has shown it to be incorrect, AND if explains physical events, AND if it can be modeled mathematically, it becomes a theory.

Here is Stephen Hawking’s explanation of a theory:

A theory is a good theory if it satisfies two requirements: It must accurately describe a large class of observations on the basis of a model that contains only a few arbitrary elements, and it must make definite predictions about the results of future observations. … Any physical theory is always provisional, in the sense that it is only a hypothesis; you can never prove it. No matter how many times the results of experiments agree with some theory, you can never be sure that the next time the result will not contradict the theory. On the other hand, you can disprove a theory by finding even a single observation that disagrees with the predictions of the theory.

Note that Dr. Hawking is of the opinion that there is not much difference betwixt a theory and an hypothesis.  In the sense that he stated it, I do not disagree, but the fundamental difference is that a theory has been tested many more times than a hypothesis (in general) and that it is more broad in scope (in general).  Whilst an hypothesis may have had scores or hundreds of verifications, a good theory may have had tens of thousands in many different situations.  Generally, a theory is more broad than an hypothesis.

Theories also tend to be complex, although they are not always so.  There are several reasons for that.  One is that theories tend to explain lots of things, to the mathematics are often complex because of that.  For example, Einstein’s General Theory of Relativity has scores of pages of extremely complex mathematics associated with it, all essential to develop it properly.  By the way, after millions of observations, it has NEVER been shown to be incorrect.  As a matter of fact, its predictions are so good that relativistic corrections have to be made for the Global Positioning System to work.  (It turns out that the passage of time slows relative to an independent reference frame as the force of gravity increases, so the GPS satellites, being further from the center of mass of the earth are in a weaker gravitational field, so time passes more swiftly for them than of the surface of the earth.  Without correction for that, accurate locations would be impossible).

Now, is the General Theory of Relativity absolutely the right explanation for our observations?  Maybe, maybe not.  What it is is the best model that we have at present to explained observed phenomena.  Perhaps the time will come that another theory that is more fundamental will replace it, or that it will become a special case of a broader theory.  This will become more clear later when we discuss scientific laws.

Now, a theory is just about the strongest statement in science that can be made.  In a good theory, as much assumption must be discarded as possible and replaced with actual observation.  (Alas, SOME assumptions are always made because our knowledge of the universe is imperfect, so we must substitute assumptions for actual knowledge from time to time).  However, when someone says, as in the case of Darwin’s Theory of Evolution, “Well, that is just a theory”, you can now counter with certainty with the statement, “Perhaps so, but this theory has been been tested thousands of times, and no exception has ever been found.  It is not a guess, but rather the best model proposed that accounts for every observation yet made.”

Jingoists like to conflate the term “guess” with “theory”.  Remember, a guess is just that, although some are better than others.  I guess that this post will be well-received, but I have no evidence to support that.  A theory is as close as we have to the real picture of how things work.  As a matter of fact, the terms “theory” and “model” are sometimes used interchangeably, and often that is justified.

It is not even necessary to know the fundamental processes about why a theory works to have a good theory.  When Darwin proposed the theory of evolution, genetic science was in its infancy, chromosomes were just some part of the nucleus of a cell with unknown function, genes as a physical objects were not yet observed, and DNA had not even been imagined.  However, with few modifications, that theory still explains what is observed.  We now just have a better handle on the why and how, rather than just the what.

Thus, a theory is pretty much the distillation of knowledge from a large collection of many experiments, over many hypotheses, that still makes sense.  A theory is NOT rigid, but is open for new input, as experimental evidence comes forth.  To use more common English, a theory is “

a model, derived from empirical evidence, that can predict future events in a reliable manner

.”  That is my definition.  A good theory has been shown NOT to be incorrect over a huge number of observations.  That does NOT mean that it is completely correct.

The next step up in most folks minds is a scientific law.  I sort of take issue with the term, as will be shown later.  Here is one definition of a law:

A scientific law or scientific principle is a concise verbal or mathematical statement of a relation that expresses a fundamental principle of science, like Newton’s laws of motion.

Actually, to my way of thinking, a law is just a theory that has very broad application and (usually) a fairly short way of stating it.  However, just because something is called a law does NOT mean that it is the be all and end all.  For example, the very law used in the definition that I found above does not always hold.

Newton’s laws of motion work fine for objects that we can see (or even some smaller ones) as long as they are not too massive, or move too fast.  For example, they describe perfectly well, within any observable experimental limit, the motion of a bullet when fired from a rifle.  For the most part, they describe planetary motion to a high degree of accuracy.  However, those laws are actually theories (I maintain that there are NO natural laws of which we are yet aware, but that only means that we are not yet astute enough to decipher them yet).

However, when one looks on the atomic and smaller distance domain, Newton’s laws break down, particularly as particles become less massive and velocities increase.  To describe those motions, quantum mechanics must be invoked (and quantum mechanics is just another theory that explains motion on the very small scale).

Likewise, on the macroscopic scale, Newton’s laws fail when velocities become significant when compared to the speed of light.  Let us go back to our bullet analogy.  Let us use the .340 calibre Weatherby Magnum cartridge as an example.  This is a really “hot” cartridge in that is has a tremendous muzzle velocity and hence energy.  The 250 grain (1.62 x10^-2 kg) load has a muzzle velocity of 2963 feet per second (903.1 m/s).  For some reference, this is 2.65 times the speed of sound.  However, because of the high velocity, relativistic corrections show that the bullet is actually a bit more massive.  Without showing the work, the mass of the bullet at that speed is actually only about one trillionth of a kg larger.  That is because the muzzle velocity is small (only around three parts per million) compared to the speed of light, which is 2.9979 x 10^8 m/s.  Thus, the effect of relativity is negligible for this bullet.  However, what if we could fire it at, say, 10% of the speed of light?  Keep with me here.

If we were able to do that, the muzzle velocity would be 2.9979 x 10^7 m/s.  Applying the relativistic mass correction, the mass of the bullet then would be 1.71 x10^-2 kg, or almost 6% more massive.  Using Newton’s F=ma would give the wrong energy for the bullet because of the mass increase.  By the way, as the speed of light is approached, the effective mass tends to infinity.  (Physics purists will argue that I should properly use momentum rather than mass, but the approximation of using mass does not injure the illustration).

This is an example of high speeds affecting Newton’s laws of motion.  There is another one.  For years, the orbit of Mercury was observed not to follow Newton’s predictions quite correctly.  Speculation was that there was an invisible body affecting its orbital period or some other gravitational anomaly.  However, relativity explains it exactly, and it is the same effect that must be corrected for by GPS.  Mercury is so close to the sun that it is affected by the huge gravitational well of the sun that time passes more slowly for Mercury than it does for us on earth, in our reference point.

Thus, I am not fond of “laws”, because it makes them seem the ultimate answers.  They are not.  The laws that we know are highly developed theories with global applications, but they are not immutable.  This gets me to my final point.

Even the best theory is likely to be replaced by a better one.  Newton’s laws of motion are not wrong, just of limited scope.  They work well in their domain, and general relatively collapses to Newtonian physics at relatively (pun intended) reasonable masses and velocities small compared to the speed of light.

This is just to point out that there is a great difference between the casual use of the terms mentioned here and the strict scientific one, and that even top scientists are not in full communion as to the definitions.  However, a real theory is NOT a guess, it is a model that describes the behavior of a given system with no exception as of yet found.

Well, you have done it again.  You have wasted many einsteins of photons reading this hypothetical post.  And even though Karl Rove chooses slate and chalk when he reads me say it, I always learn much more than I could ever hope to teach whilst writing this series, so please keep those comments, questions, corrections, and other correspondence coming.  Tips and recs are also always welcome.  Remember, no scientific or technical issue is off topic here.

Warmest regards,

Doc

Crossposted at Dailykos.com and at Docudharma.com

Copyright July 25 2010 by Dr. David W. Smith.  The titles “Translator” and “Pique the Geek” are claimed trademarks of Dr. David W. Smith.  Anyone is free to quote the text in this essay if credit is given to Dr. David W. Smith for the quotations.