SCIENTIFIC MODELS AND JUST-SO STORIES
“Catastrophic results” may follow when “those
who use formal models…forget that models are not theorems.”
—Eugene J. Neckar
Rudyard Kipling wrote a series of children’s stories that he called Just-so Stories. The stories often explained why something was a particular and everything turned out nicely in the end (“just so”) more because the author wrote it that way than because if might have really naturally happened so. Since the stories are fiction this is valid and since the author is making up the story anyhow it will happen any way he/she wants it to. History tends not to happen the way anyone plans or wants it to. “Reality,” the “Laws of Nature,” or physics limit the way things can and do happen.
Fiction, or the art of creating a story based on real facts or possibilities is one of Man’s great achievements, and has a valuable role in human culture. This creative ability is also important in the fields of science when it is used to assemble several facts, or a multitude of facts into an hypothesis or model to try to explain them.
Construction of “stories:” hypotheses, models, scenarios, is important to the progress of science. They serve a valid function. A scientific hypothesis is a model that explains how/why certain observed facts are related. Based on this model, predictions are made which are tested. If the predictions are found to be correct the model is considered to represent reality. If the predictions are not correct, the model may be discarded or modified to reflect the new facts.
A scientific model is meant to represent nature, reality. As such they are only an approximation of nature and not absolutely identical with it. Generally I suspect that the model is much simpler, they do simplify very complicated relationships and processes. Models are always an approximation, similar to but not identical with the reality that they are a description of. They do not explain the phenomena but rather describe possible relationships, and provide a way to think about a phenomena, to help visualize what is happening and can help illuminate the problem(s), and suggest ways of solving or explaining it. Models are a way to economically organize data and observations in an conceptual way it replaces the data and observation is a way that is easier to remember. Models also help to communicate ideas to others by providing and drawing upon analogies1 and metaphors. A model is not a theory although it could be converted to one. A model is the interface between theory and data.
The heliocentric solar system of Copernicus is model of the Universe with the Sun at the center and the planets revolving in circular orbits about it. (It is possible to built or buy a physical model of the solar system). Kepler demonstrated that the motion of the planets did not fit with the prediction of circular orbits. Slightly elliptical orbits produced predictions of the planets’ locations that matched those actually observed. Copernicus’ model was modified, circular planetary orbits were changed to slightly elliptical orbits. Since then further changes have been made to the basic model, moving the Sun from the center of the Universe to the edge of a galaxy somewhere in the Universe. The basic planets-revolving-around-the-Sun has remained unchanged. This model is believed to represent reality, in words.
Thinking about this led to the prediction that Venus and Mercury would show phases like the Moon because their orbits were between the Sun and Earth. Telescopic observation of those phases helped win support for the model.
Any hypothesis about how the Universe began and evolved into what it now is, has to explain why planets revolve around stars. The story of the evolution of the Universe has to have planets in elliptical orbits around stars.
A “scenario” can be defined as: a model in narrative form, an outline of a hypothesized chain of events. As such a scenario is a model linking various facts and events together in sequence, often proposing a cause-and-effect relationship between the facts/events. Archduke Franz Ferdinand and his wife are assassinated in Sarajevo by a Serbian nationalist. The Archduke is the heir apparent to the throne of the Austro-Hungarian Empire. Austro-Hungary decides to punish Serbia, with whom they have been having problems with for years and see a chance to solve the problems once-and-for-all. The Russian Empire is not likely to accept Austro-Hungary’s attack on Serbia. So Austro-Hungary asks their ally, the German Empire, if they will honor the treaty to help Austro-Hungary if it is attacked by Russia, even if the attack is in response to an offensive move by Austro-Hungary against a third party. The German Empire looks at its plan for a war against Russia which has a defensive treaty with France. If Russia is attacked by Germany, France will attack Germany. If France is attacked by Germany, Russia will attack Germany. The Germans had planned, to avoid a two-front war with France and Russia, to attack France, before Russia can mobilize, and quickly knock her out of the war before turning on Russia. To defeat France quickly the plans called for the German Armies to march through Belgium (with or without their permission). Therefore, if Austro-Hungary’s declaration of war against Serbia led to Russia’s declaration of war against Austro-Hungary and Germany honored its treaty to assist Austro-Hungary, Germany would attack France. Complicating this was the British Empire’s guarantee of Belgium’s borders. This was a fatal chain of events for at least ten million people as this is the scenario for the “cause” of World War I. (I wrote “cause” because this is the immediate, proximal, cause. The ultimate causes of World War I lie much further back in time.) It has been supported in much detail by historical documents, etc. It was plausible when first proposed and has been thoroughly tested. Prior to the assassination had this result been proposed for such an event it probably would have been considered a far-fetched chain of events. Yet when it happened, it seemed to be inevitable.
Models and scenarios can also be proposed as a “possible” answer that is not thought to be the real answer but is plausible enough to deal with “misgivings” and will suffice until more evidence is found and a model based on the evidence is proposed. A plausible answer has to have a logical consistency and to rationally explain or account for all the relevant known facts.
When Darwin proposed his theory of evolution by natural selection many opposed the entire idea because, for instance, how could a fish evolve into a land animal or how could a four-legged animal evolve wings and fly? Scenarios were proposed without any solid fossil evidence. The models were not proposed as the way it did happen but as a plausible way it might have or could have to show that “evolution” could change fish to reptiles of reptiles to birds. This allowed scientists and non-scientists to accept the concept of evolution before all the evidence had been found to explain all the details. All this “airy theorizing” formed a basis for future work. There was no harm, and a lot of value, in these models/scenarios as long as everyone remembered that they were just vague ideas unsupported by evidence, just-so stories. Problems have arisen when these scenarios have become accepted as tested models. They become entrenched in textbooks and people’s minds as if they were true and influence the interpretation of newer data and the construction of other scenarios, forming a house of cards.
In paleoanthropology models are more often in the form of a story2, a narrative, perhaps all models are a story—a word description. Critics often complain that they are just-so stories. This is a reference to Rudyard Kipling’s “Just-so Stories,” deliberately fanciful stories meant for children about how some animal got the trait that is characteristic of it (e.g. the elephant’s trunk, or the spots on the leopard). Just-so stories stand or fall on their plausibility or the author’s ability to tell it. These ideas are just tossed out to explain facts, no attempt is made by the author to evaluate or test them. True just-so stories are not testable, defended by ad hoc excuses, they are analogies and/or anecdotal. Calling some model a just-so story is an implicit criticism, implying that the model is fictional, unverifiable and unfalsifiable.
Narrative stories are justified if they function to rebut claims that something could not have happened. A plausible story3 (whether true or not, but could be, cf How the Giraffe Got Its Neck) rebuts the claim of implausibility.
Every model, hypothesis, and ad hoc hypothesis should be plausible—consistent with known facts, etc. “But plausibility does not render the interpretation true or accurate (italics in original, Binford 1983:75). This “plausibility” can lead to trouble when people begin to think that the plausible is true or some how supported by facts and then use the hypothesis as a basis for other hypotheses or models. There have been several occasions when some scientist put some very hypothetical idea down in print, sort of thinking out loud (in print actually) about what the facts might, just might, mean or in answering some critic’s objection that there was no way such a thing could occur. Then a few years later discovering that the idea has become “set in stone,” and assumed to be a well supported theory.
It is always important to distinguish between facts on the one hand and models and hypotheses on the other. Facts will probably not change but models and hypotheses will. Any additional model of hypothesis based on a plausible hypothesis has a foundation built on sand.
Just-so stories can serve as a hypothesis, as a model that can be tested by further, if it is constructed properly. A proper model has to be a plausible account based upon the evidence. Some models are more plausible than others, and as new facts, observations, and ideas are gathered, some models will cease to be plausible, and new ones will become plausible. Science progresses by testing hypotheses. Archaeologists and paleoanthropologists have been accused of creating such models to explain what they have found. Just-so stories that sound great but are actually without tested evidence to support them. Actually the criticism isn’t that the scientists make up “stories,” but rather that the stories are accepted without supporting evidence and then used as true and verified models to explain other facts.
Creationist stories (“theories” in the colloquial sense) are almost nothing but just-so stories, based on no facts with numerous ad hoc claims and miracles. The supporters never test the stories.
While we’re at it, I might as well discuss “ad hoc,” because another this is another criticism often used. Ad hoc means for the specific purpose and when it is used as a criticism it is often meant to imply that the reason was made up specifically for the particular situation, on the spot, and that it is untested.
The real problem with ad hoc hypotheses is not that they might have been made up “on the spot,” or that they were made up to answer a specific problem, and only that problem. Some great scientific discoveries were made up on the spot to deal with a specific problem. Inspired genius is one way to describe it. No, the problem is not even that it is, at the moment, untested. It can be tested later, if it remains untested then that does become a problem. The real problem with ad hoc claims is why it applies in that one case and not in other identical or similar cases.
As an example: airplanes have been landing all day at an airport as they have been doing for years without incident. Today, two identical aircraft are in the landing pattern, one behind the other by the FAA required 60 second interval. The first plane lands without incident, the second landing in what seems to be the same identical conditions is just short of the runway and about touch down when it drops suddenly and violently to the ground.
Everybody in the vicinity begins to think “what happened?” Why did the previous plane land safely and the second crash? Some will come up with a hypothesis, a reason why. Since the reasons were made up on the spot for the specific incident and are untested, the reasons are ad hoc. One of them may be right. The NTSB will investigate the crash to determine the cause. Whatever caused he second plane to crash was in some sense unique to that plane at that time. The cause needs to explain why the second plane crashed, why the cause was operative then, and why none of the other planes did not crash, why the cause was not operative at those times. Wind shear was discovered under such conditions.
The ad hoc hypotheses that scientists do not like are the ones that for no rational, natural reasons only occur at “convenient” times. Science is a search for regularity in the world around us. Even an irregular event should have an explanation that is regular, unique in its occurrence. Not only should the explanation explain why the second plane crashed it should also explain why the first did not, why it was not operative then but was a moment later.
Trust is a great thing. I like to believe most humans are inclined to believe that most people are trustworthy and I am willing to extend some trust to people I have just met. However, I have learned that in too many cases the trust has to be conditional, limited until I have more knowledge about that person. Complete trust is earned, Salespeople, advertisers, politicians, and scam artists want everybody to be trusting, to trust what they say without question. They do not want you to think about what they say, what they claim. People who think critically about their statements may begin to question and doubt them. They don’t like that. It makes it hard to convince you, to get you to buy something, vote for them, or believe what they want you to.
I find it helps a lot to find out what the person in question is trying to “sell me.” Every statement made in support of fact presented needs to be thought about and not uncritically accepted. Statements and facts about their opposition is also suspect. What they don’t mention can be more important than what they do.
We have actors endorsing various products, or politicians, and athletes also doing endorsements. We are supposed to think that a famous person has “something” that makes their endorsement more valuable. Well they do, they have their fame, this brings recognition and more attention to that which is being endorsed. But an actor endorsing a product or campaigning for a politician leaves me cold, totally unimpressed. Now Lance Armstrong endorsing a bicycle tire, or biking shoes (or something that he actually uses bicycling) might lead me to consider it. If he uses the product, it might be better than the others. He’s an expert in bike racing so he should know, but for politicians, insurance plans he’s endorsement is no better than mine. There is currently an ad on TV for a particular golf putter. At the end of the ad a well-known professional golfer (I think he is but I haven’t heard of him, not that golf interests me) endorses the putter. The ad also touts the number of tournaments he has won (that is why I assume he is a well-known golfer). I am supposed to think that this club is really great because he says it is and look at all the tournaments he has won (he also shot a record low score during one of them. He says it will improve your golf game. It is quite possible the club was not available when he was competing so he couldn’t have used it. But I cannot help thinking that he doesn’t use the club now when he plays, either in tournaments or for fun, so what does he really know about it. I know he got paid to stand in front of a camera and say it is great but it is not like he actually used it to win golf games, the ones he played for money. He says it will improve your golf game. He never a says it improved his.
There are few if us who have such a broad base of knowledge to be expert in every field. It is possible that when reading a book the author may cover a range of disciplines that include one or more areas that we are not knowledgeable about. I have adopted the principle that if an author has a poor understanding of the areas I have a good knowledge of, then I do not quite trust his arguments in the areas I don’t have a good grasp of.
I first read Immanuel Velikovsky’s Worlds in Collision when I was in college. I found the idea of near collisions with a comet as a basis for many of the stories in the Bible interesting, if a bit far-fetched. The idea sounded as if it might be plausible. I re-read the book a few years later, paying more attention to some of the “evidence.” The astronomy I had some basic understanding of and very little knowledge of the literary sources (Biblical, Sumerian, Akkadian, Egyptian, Chaldean, Chinese, Aztecan, Incan, Hindu, etc.). However I had better knowledge when it came to the part about the frozen mammoths.
Over the years a few frozen carcasses of mammoths have been found in Siberia. Velikovsky explained these as the result of the Earth’s poles being shifted by the near miss of a large comet (which became Venus or Mars). The shift of the poles moved Siberia north about 20° where it instantly became colder and the mammoths were frozen and buried in snow and ice. (They had died by asphyxiation when the comet had burned all the oxygen out of the atmosphere.)4 This also marks the end of the last ice age. He dates all this to Biblical times, several hundreds or thousands of years BC, rather than 15 or 20 thousand years ago according the radiocarbon dating and other evidence. According to him the mammoths lived in a temperate environment and died suddenly. We know this because their stomachs were full of grass and various flowering herbs that they were eating at the time of their deaths. The problem here is he seems to believe that the flowers grow in warm climates and Velikovsky seems to think that mammoths are elephants (which they are a type of) and elephants live in Africa and India where it is hot, so when mammoths lived in Siberia it must have been warm there. The fact that they were eating flowers when they died is further evidence of the warm climate. I read this and thought he has got this part all wrong. The particular flowers grow in Siberia, in the tundra regions in the spring today, and that mammoths are very well adapted to cold climates. The mammoths froze because they fell in to crevasses while crossing ice fields or were overwhelmed by a mud flow which then froze (permafrost), not because of a sudden sharp drop in temperature caused a glacier or monster blizzard to bury them. At this point my doubts about the validity of the rest of his evidence increased sharply.
Years later I read Carl Sagan’s Broca’s Brain and came across the following incident (page 101):
…I can remember vividly discussing World’s in Collision with a distinguished professor of Semantics at a leading university. He said something “The Assyriology, Egyptology, Biblical scholarship and all that Talmudic and Midrashic pilpul is, of course, nonsense; but I was impressed by the astronomy.” I [Carl Sagan] had rather the opposite view.
It only reinforces my own opinions.
1 An analogy is a method of illustrating a situation by comparing it with another which has some features in common with the first. The second situation may be chosen because it is easier to understand, perhaps because it is more familiar. When I use an analogy it will be used more in the line of illustration (to clarify or help to clarify a point) and not argumentative (to compare dissimilar things to exemplify or prove a point).
2 “Only by telling a story can you tell if an idea is valid.”—Mishia Landau?
3 Plausibility simply demonstrated that given a line of research is a rational endeavor. Research stemming from such arguments of plausibility ought to result, one hopes, in the production of reliable methods for inference. — L. R. Binford 1983:75
4 I have just realized that this “cause” has a number of implications, such as: why did not all the other animals suffocate?, when oxygen burns it releases a lot of energy as heat, why did not the Earth heat up? and, how long does it take the oxygen levels in the atmosphere to return to breathable levels?