vetenskapsteori.se

Important parts of science

Documented observations

It makes it possible to estimate the probability for that the observation was correctly performed, or if errors could have influenced the observation result.

The description also makes it possible to use the result from the observation in the future, even it a later scientist does not agree with the implementation made of the first observer.

A third advantage with the descriptions is the pedagogic value. Later scientists may use the same methods as the initial report, and do not need to learn them from the basis

Reproducible observations

A criterion demonstrating that an observation is well documented is that it is possible to be reproduced.

Reproducing the observation should also give the same result as in the original description. This requirement increases the probability that the first observer makes efforts to eliminate random factors in the report.

Verified
results

An additional factor, that has added to the reputation of science in society, is that scientists review their colleagues results. A scientist verifies or falsifies other's observations regularly. New and important results even  r e q u i r e s   verification by independent scientists to be regarded as reliable.

In case a reproduction of the observation does not give the same result as in the original report, the new result may be published. This may imply that the original author is criticized in public, which of course is a negative experience. Openness hence leads to a personal motivation to deliver results of high quality.

Controlled
reports

A scientific article may of course contain obvious errors. To decrease the risk of such errors, the article is reviewed prior to publishing by one or two referees, persons skilled within the area in question. The procedure is called "peer review".

The system with "peer reviews", may unfortunately give rise to possibilities of abuse in areas where a few people dominate their research area and often are chosen as referees.

My own experience of referees is that their suggestions for completions have been well motivated. In one case, the referee was maybe more seeking to demonstrate his own excellence than to provide an objective review. In that case, it was enough to defend the original results for the magazine editor, to get the article published.

Building
further

The high requirements for documentation and verification have led to an incredible development within scientific disciplines. Instead of "starting from the beginning", the documents provide a possibility to "build further" on previous results. Hypotheses based on the results may rapidly be judged as plausible or be rejected.

A similar type of building on previous experience can be found within almost any human activity. Experience is transferred e.g. through journals, handbooks or through personal contacts.

What is significantly separating science from other activities are the higher requirements of reproducibility. High standards of documentation are also found within e.g. house building. Within this area, however, it appears that the normative information (regulations) is dominating over the describing.

Interpretation
of results

Results from observations are interpreted, meaning they are placed in a context. Interpretation may i.e. mean that visual observations are claimed to be reliable.

Interpretation is, according to my opinion, a part of scientific work, in cases when the process is so well documented that it can be reproduced by an independent colleague.

Hypotheses,
theories
and laws

A hypothesis is a statement, an idea, a fantasy or just some simple though of no specific type. It may, or may not, show a connection to what we call reality.

In scientifically documented publications, a hypothesis is often an attempt to explain or generalize observations. Hypotheses are interesting and important components within scientific publications.

Inductive - deductive method

Already Aristotle discussed the inductive-deductive method that still often is applied within scientific work:

One or several observations or previous hypotheses results in a belief of something, i.e. we have created a hypothesis. If we want to investigate if the hypothesis is in accordance with our perceived reality, we reproduce or control the observation. Was it only possible to make our observation together with other factors that we did not consider? Is it possible to verify the hypothesis further? Is it falsified by some observation? Are the consequences of the hypothesis verified or falsified by observations? In case additional observations are in accordance with our hypothesis, it is verified.

When a hypothesis is verified enough, as indicated by that is it commonly accepted, it may be used as a starting point during investigations of our perceived reality. Is it possible to use it, together with other verified hypotheses be used to create hypotheses with still higher degree of generalization? Or can it be used to interpret phenomena within other areas than from where it was created?

This procedure is sometimes called the hypothetico -deductive method, and then somebody have forgotten that the hypothesis was originally created from observations.

Verification, falsification, theory, law

As observations results in that a hypothesis is verified, or that its consequences are verified, the belief in the hypothesis increases. A falsification, i.e. a verification of a negation to the hypothesis or to its consequences, is of course disturbing; it decreases the belief in the hypothesis and may result in that the hypothesis must be discarded or modified.

When observations that verify a hypothesis are generally accepted, the belief in the hypothesis is increase and it may become termed as a theory. The border between hypothesis and theory is very diffuse. Hypotheses that have been widely discussed and are generally believed to be in accordance with our perceived reality may be called "laws".

Verified hypotheses, theories and laws are, together with observations, important parts of science. But scientists normally do not express the opinion that something within science should be "true" or should represent "absolutely certain knowledge". Every scientist is, sometimes painfully, aware of that an article may be published that will claim other hypotheses than therr own belief.

An example of this skepticism about hypotheses is that "energy cannot be created", a fully tested and accepted law, is called for "postulate" or "law".

Theoretical publications

An additional and common method to create hypotheses that gain confidence, is to connect previously accepted hypotheses. Often mathematical relations are used for connection, which imply that the result may be quantified and be compared to observations.

When such a connection has been verified by observations, or by traceable logic like mathematics, a very strong belief in the starting hypotheses and in the result hypothesis is created. If we believed strongly in the starting hypotheses, an emotion that we call understanding may be experienced.

Science is largely structured using such connections and this may contribute to that philosophers, that are unaware of that the original hypotheses described results from observations, claim that theories are the basis for what we believe about our world.

"Scientific" and
"non scientific" hypotheses

A hypothesis is hence a more of less credible statement. The credibility is based on observations and how these are reported. In cases they are reported according to scientific methodology the hypothesis may be accepted in a scientific magazine..

A discussion about a hypothesis being "scientific" or "non-scientific" is hence basically illogical and unnecessary - a hypothesis is not in itself neither of these two. The question is whether the observations, forming the basis for the hypothesis, are reported using something being in the neighborhood of scientific methodology.

The interesting feature of a hypothesis is if it is entertaining, stimulating or if it, through its origin or consequences, can be shown to conform to what we call our reality.

Hypotheses that are not based on observations

Very often, hypotheses are expressed that are neither based on carefully described observations nor can be verified through it's consequences. The source may as examples be observations that are not possible to reproduce during testing, analogies to other areas, assumptions that are not generally accepted, or a wish of the hypothesis' creator to receive a favor of some kind, without having to care too much about our perceived reality.

Such hypotheses may later be found to be verified by scientifically described observations, or may be shown not to be in accordance with such observations.

In order to avoid mistakes, the community of (natural) scientists views hypotheses that are not supported by observations, as unreliable. This skeptic view is sometimes criticized, e.g. by people advocating alternative reality.

A vast majority of not verified hypotheses have been forgotten or have been shown to be erroneous. But there also exist examples of hypotheses that have not been accepted by the scientific community, but later have been shown to be in accordance with observations and have been accepted.

Examples of creators of hypotheses, that unfortunately only after their death were acknowledged, include Ludwig Boltzmann (thermodynamics), Alfred Wegener (tectonics) and Milutin Milanković (climate).

Logic and mathematics

Logic and mathematics are tools that help us to organize observation results. As tools they contribute to what we call understanding. We may for instance understand that different observations have the same cause, that experiences are possible to transfer from one area to another, or that a certain result always follows from a cause.

Use and development of logical tools, e.g. mathematics, are very important within scientific work.

Is mathematics science?

It is sometimes discussed if mathematics is a part of science, as mathematics is not considered with empiric methodology. In my opinion it is similar to a discussion if letters are a part of science. Both are tools used in description of what we call reality.

My opinion is that a well documented development or discovery of a mathematical relationship is a scientific result, similar to e.g. the discovery of a new species of fish. The observation consists of the relationship and the documentation determines if the development or discovery should be considered as scientific.

Use of mathematics and logic to connect different observations, and by this to create increased acceptance and feeling of understanding, is according to my view also a scientific contribution.

Adjustable
parameters

Use of known mathematical relations does not necessarily imply that a publication becomes "scientific". It is unfortunately common that "models", or mathematical constructions that contain adjustable parameters are used in order to increase the belief in reasoning. When the values of the adjustable parameters are not coupled to observations, this type of reasoning becomes as scientifically uninteresting as the phrase "I guess that...", it may only appear as being more scientific.

	With four parameters I can fit an elephant and with five I can make him wiggle his trunk.
	John von Neuman (according to Enrico Fermi)

 ver. 4.5