What is Science? – Part 2 of a Series

Before we start reading scientific papers we should all agree what science is. We all know what science is, right?

So what is it?

From the Oxford English Dictionary (Shorter)

“Theoretical perception of a truth, as contrasted with moral conviction (conscience).”

Sounds a little loose a definition for my purposes, and like it is a derivation of the argument between rationalists and theologians at the beginning of the Enlightenment. Perhaps I should have bought the full OED.

How about Wikipedia?

“Science is a systematic enterprise that builds and organizes knowledge in the form of testable explanations and predictions about the universe.”

That is more like it, and “…in the form of testable explanations and predictions…” sounds like an allusion to the scientific method. That builds and organises knowledge… so it is the practice of acquiring knowledge, but also of recording it, or documenting it. Otherwise it would not be very organised.

What exactly is the definition of that, of the scientific method? We have all read about it, but what does it mean? From Wikipedia again:

“The scientific method is a body of techniques for investigating phenomena, acquiring new knowledge, or correcting and integrating previous knowledge. To be termed scientific, a method of inquiry must be based on empirical and measurable evidence subject to specific principles of reasoning. The Oxford English Dictionary defines the scientific method as “a method or procedure that has characterized natural science since the 17th century, consisting in systematic observation, measurement, and experiment, and the formulation, testing, and modification of hypotheses.”

Interestingly they quote the OED. More interestingly it is a statement of principles, not a recipe or method. It does not have the rigour or specificity of a law, but it is flexible, so long as one agrees about what methods conform.

Note that there is no mechanism for making a theorem part of the accepted canon of scientific knowledge. Length of time in use, and probably how widely used in place of any available alternatives, for the theorem appear to be what is required for acceptance.

Which suggests as the converse that the opposite is true, too. For an accepted and widely held theorem to be replaced takes a long time, perhaps in proportion in some respect to how long the idea to be overthrown has enjoyed acceptance. Someone described it to me once as waiting for the current generation (of scientists) to die.

But why so slow?

Experience, such as with Newton’s Theory of Gravity, demonstrates why it is customary within science to call ideas, or propositions, theorems, and not laws. Why they are tested, and retested, why the process takes so long, and why they stick around so long when confronted with a replacement. A theorem may possess predictive power across an enormous range of situations and circumstances, but often break down in unusual or extreme situations. Or, when instruments improve and allow us to observe or experiment with more phenomena or with much greater precision. Newtonian physics and his equation describing gravity held true, had the appearance of law, for about two hundred years. If it works, why replace it?

Enter Einstein’s version of physics about two centuries after the adoption of Newtonian physics, and the use of quantum theory, relativity, etc., to describe the fundamental properties of matter, space, energy and time, and the ongoing search for a unifying theory that works across subatomic as well as interstellar distances. Einstein’s theories were not just new, and revolutionary. They were hard, opaque, and seemingly nonsensical in relation to everyday life and observations. If it so conflicts with my beliefs, why replace it?

Facing such resistance to new ideas then, for good reasons and bad, custom became that multiple, different, repeatable experiments be conducted by different people, at different times, and in different locations before a theorem is widely accepted. Nor is there any guarantee that such acceptance be universal, or happen at once.

Hence perhaps the scorn of “hard” scientists for the social side of science, which is not at all amenable to the scientific method, and one of the reasons I believe the scientific method requires adaptation.

Repeatability, as well as rationally deriving conclusions from experimentation, demands certain things. In practice it is very hard to derive cause and effect if more than one factor varies. Repeatability is also made harder.

Any factor that can vary, and that could affect the outcome, which in a strict practice of science has to be accepted as any if one is not to prejudge*, is called a variable. Because of this experiments that have full rigour under the scientific method isolate one variable to test. While all possible variables must be controlled for, kept constant. Or discounted by accepted mathematical or experimental methods.

One measurable phenomenon is made to vary, say temperature, usually in a uniform way, in regular increments. In this way the impact of that variable on the other phenomena being observed can be measured. More to the point the degree of variability correlated to measurable changes in the phenomena observed can be derived. You can see the importance of this with drug testing, it would not do much good to give them three experimental drugs if you want to find out if any are effective. Nor would it be any good to know that a microgram didn’t do much, but a kilo of it killed the subjects. Nor should the range of the variable be too narrow. Nor the steps too large. One can always discard extra data.

So it seems we have the rough basics for methodology for research.

But science also includes the process of publication and of testing of an experiment by others.

So science must also include the dissemination of ideas, theorems, and of experiments and their observations. The process of the testing of experiments and theories by others, and of the arguments that must follow, and the recording and documentation of that process, publishing.

The custom of testing and of peer review, and the demands of the scientific method, saves us from investing time and resources in pursuing the suggestions or ideas springing from theories that lack predictive power. But also why science moves slowly, and only adopts new ideas long after a new theorem has been proposed.

Perhaps the most infamous example of the medical community’s, doctors’, resistance to new ideas is the edifying story Ignaz Semmelweis. His story also illustrates why the business of communicating and sharing scientific research, and subjecting it to wide spread review and consideration, to peer review, is so important.

As well as why being able to read and understand it can be so important. I would not like to have delivered a baby on the pavement because my doctors were so hidebound and because I could not argue otherwise.

The story of Newtonian physics also shows how a theorem may have utility, be right in most observable instances, while being strictly “wrong”.

In terms of predictive power we got along fine with Newtonian physics for centuries, we managed the industrial revolution, powered flight, rocketry, aqueducts, dams, suspension bridges, steam power, and the internal combustion engine just fine without Einsteinian physics. Although Newton’s equations describing gravity are not “right”, we stayed stuck to the earth. So there is clearly good enough science that may not be perfectly correct in all circumstances. This is an important concept for later.

What we think of as modern science and it’s methods grew out of the Enlightenment, and Newton was alive during the period. The Enlightenment was a movement away from ideas derived from faith and belief, characterised and embodied by the christian church in Europe in the 17th and 18th centuries, and towards one of beliefs obtained from reason and argument. Newton himself embodied this nicely. He successfully resisted the requirement at the time that members of Cambridge University, his college was Trinity after all, be ordained ministers in the Church of England. The origins of science in a movement against faith based beliefs probably helps explain the Christian Church’s antipathy towards science.

This coincided with the widespread use of printing, specifically of movable type and the relative democratisation of the access to knowledge, in Europe. The Chinese and Koreans had enjoyed movable type for half a millenium by the time it came to Gutenberg. It’s impact was enormous, the internet of the day. As printing evolved and printed materials became more widespread, the dissemination of scientific ideas and research spread via the new medium. As it did so it adapted and adopted various new forms as a result. Newton published his work in book form. But one cannot write a book about every idea or experiment. So the practice of bundling papers from more than one author evolved.

From this process, still ongoing, we inherit the scientific journal. Specialist publications devoted, for a profit of course, to publishing science and research deemed worthy of consideration in the first place. That evaluation process often excludes as much as sixty or eighty percent of material submitted. It then undergoes revisions during a process called peer review, which is just as it sounds, before publication. The journal system is the primary way science is shared, evaluated, developed, refined, and accepted so that it becomes part of the library of human knowledge in the sphere of science today, flawed as it is.

I think we have worked our way to a complete definition of science. Here we go:

“Science is both the study of natural phenomena required to obtain an understanding of the universe and its functioning, and the canon of accumulated knowledge derived from that activity.

Scientific research uses methods and practices congruent with the conventions of the scientific method. Those methods, and observations derived, should be reproducible and repeatable, and made in accordance with the traditions of reason. Experimentation should be in a controlled environment, and deal with isolated variables so that cause and effect may be clearly derived. The conclusions drawn should be supported by experimental observations, and analysis of the observations and measurements should itself also be repeatable and congruent with accepted practice and methods.

Scientific knowledge consists of research that has been published and disseminated after a process of peer review, in journals usually dedicated to the branch of scientific knowledge aligned with the subject at question. Science demands critical thinking, evaluation of any claims, and healthy skepticism. It should therefore be read critically, not just accepted at face value because that is antithetical to science itself. With time and use theorems become accepted as part of the canon of scientific knowledge.

Theorems and knowledge derived from science are never be regarded by a scientist as the final word. Further research will almost invariably produce a successor theorem that improves upon its predecessor, but older theorems still often have utility, and often times the virtue of simplicity in comparison to their successors.”

My next post will deal with the topic of “Where science is published and how to find it”, and that is largely the story of journals and how papers get published, where they are stored, and why you might have to pay to read some of them.

Next will be “Where science is published and how to find it”.

Index to this series of posts:

  1. Overview
  2. What is science (October 18th, 2014)
  3. Where science is published and how to find it (October 23th, 2014)
  4. Peer review (October 29th, 2014)
  5. Evaluating science (November 3rd, 2014)
  6. Type of scientific paper, their structure and interpretation (November 7th, 2014)

Changed list of planned posts on October 17 to accommodate increase in number of planned posts and compressed publication schedule, moving dates up.

*Bias is an issue I will address when I discuss evaluating science.

Please do not email me asking for papers, you can do it. In fact send me your favourites, thanks.