Readings

Developing and testing scientific explanations

When we make observations we may propose a theory which accounts for them. We judge theories on the basis of the match between their predictions and what we observe. An observation is often explained by relating it to a particular scientific theory or theories. A scientific theory proposes an underlying process that results in the observations we have made. Scientific theories do not ‘emerge’ from data by a process of logical deduction; proposing an explanation always involves using imagination and conjecture; informed guesswork. An explanation is not just a summary of the data, but tries to go further.

A common way of explaining a specific event is to claim that it is an instance of a general law. A scientific law is a claim that something always happens in situations of a certain kind (for example that the pressure of a sample of gas always gets bigger if it is forced into a smaller volume).

Scientists test an explanation by seeing if specific predictions based on it agree with data from observation or from an experiment (a deliberate intervention to generate data). If data agree with predictions that are very novel or unexpected, this is particularly influential. The aim is to rule out alternative explanations, and so reach a single explanation that most scientists can agree on.

Scientists are more confident about theories that include a plausible mechanism for causing the events observed. It is also important that a new theory is consistent with existing theories that are well established and generally accepted.

It may be possible to interpret data from complex equipment in more than one way. Interpretations need to be checked and discussed with others working in the field. Data can show that a scientific theory or law is incorrect (falsification) but cannot prove conclusively that it is correct.

Paradigms and scientific revolutions

Scientists test a hypothesis by seeing if predictions agree with their observations or the results of their experiments (deliberate interventions to generate data).  A scientific theory should lead to predictions that are precise and detailed enough for it to be possible that they can be shown to be false (falsification). A theory is ‘non-scientific’ if it does not make any predictions that could possibly be falsified.

If data agree with predictions that are unexpected that is particularly useful.  The aim is to rule out alternative explanations and reach an explanation which most scientists can agree on.

Scientific belief systems are called paradigms. These dictate what models and theories are used and what questions are asked.  If new observations or theories seem to be inconsistent with existing well-established theories this may lead to a paradigm-shift where the assumptions of a whole field of science are called into question and new laws and models developed within a new paradigm.  This is a revolutionary process which often leads to rapid progress.

Establishing causal links

Scientists often want to find the cause of an event or phenomenon. A first step is to show that there is a correlation between a specific factor and an outcome. This does not prove that this factor is the cause, but it can stimulate further work to establish a causal link.

If something happens only when a factor is present, we say there is a correlation between the factor and the outcome. If one factor varies steadily in value as the value of another factor increases, this is even stronger evidence that the two variables are correlated. If both variables increase together, the correlation is positive; if one goes down as the other goes up, the correlation is negative. “Correlation does not imply causation” (cum hoc non propter hoc, Latin for “with this, not because of this”) is a phrase used to emphasize that a correlation between two variables does not necessarily imply that one causes the other.

Examples of scientific methods in action

1. Semmelweis and antiseptic practice

Ignaz Semmelweis (1818-1865) was a Hungarian doctor now known as an early pioneer of antiseptic procedures. Semmelweis discovered that the incidence of puerperal fever could be drastically cut by the use of hand disinfection in obstetric wards. Puerperal fever was common in mid-19th-century hospitals and often fatal, with mortality at 10%–35%. Semmelweis came up with the idea of washing with chlorinated lime solutions in 1847 while working in Vienna General Hospital, where doctors’ wards had three times the mortality of midwives’ wards.

Despite publishing his results showing that hand-washing reduced mortality to below 1%, Semmelweis’s observations conflicted with the established scientific and medical opinions of the time which was that infection was caused by miasma or “bad air” and his ideas were rejected by the medical community. Some doctors were offended at the suggestion that they should wash their hands and Semmelweis could offer no acceptable scientific explanation for his findings. Semmelweis’s practice earned widespread acceptance only years after his death, when Louis Pasteur confirmed the germ theory and Joseph Lister, acting on the French microbiologist’s findings, practiced and operated, using hygienic methods, with great success. In 1865, Semmelweis was committed to an asylum, where he died at age 47 after being beaten by the guards, only 14 days after he was committed.

2. Genes and inheritance

As well as proposing the theory of evolution by natural selection, Charles Darwin (1809-1882)  also proposed the idea of “gemmules” to explain inheritance. These were imagined particles of inheritance as part of his Pangenesis theory. Gemmules were assumed to be shed by the organs of the body and carried in the bloodstream to the reproductive organs where they accumulated in the germ cells or gametes. No one was ever able to isolate these gemmules.

In the 1860’s Gregor Mendel worked with peas and discovered that each parent provided a “unit of inheritance” to their offspring and that the effects of one of these could mask the effect of the other. For instance, a pure breeding yellow pea bred with a pure breeding green pea produced a first generation that were all yellow. However, roughly 1 in 4 of the second generation bred from these offspring were green. This proved the particulate nature of inheritance but did not explain what these particles consisted of. Prior to Mendel’s work, the dominant theory of heredity was one of blending inheritance, which proposes that the traits of the parents blend or mix in a smooth, continuous gradient in the offspring.

In 1910, Thomas Hunt Morgan showed that genes reside on specific chromosomes. He later showed that genes occupy specific locations on the chromosome. A series of subsequent discoveries led to the realization decades later that chromosomes are made of DNA, a molecule found in all cells on which the ‘discrete units’ of Mendelian inheritance are encoded. In 1953, James D. Watson and Francis Crick demonstrated the molecular structure of DNA. Together, these discoveries established the central dogma of molecular biology, which states that proteins are translated from RNA which is transcribed from DNA. This dogma has since been shown to have exceptions, such as reverse transcription in retroviruses.

3. Insulin and diabetes

In 1889, the Polish-German physician Oscar Minkowski, in collaboration with Joseph von Mering, removed the pancreas from a healthy dog to test its assumed role in digestion. Several days after the dog’s pancreas was removed, Minkowski’s animal keeper noticed a swarm of flies feeding on the dog’s urine. On testing the urine, they found there was sugar in the dog’s urine, establishing for the first time a relationship between the pancreas and diabetes.

In 1901, another major step was taken by Eugene Opie, when he clearly established the link between the islets of Langerhans and diabetes: “Diabetes mellitus . . . is caused by destruction of the islets of Langerhans and occurs only when these bodies are in part or wholly destroyed.” Before his work, the link between the pancreas and diabetes was clear, but not the specific role of the islets. In 1916 Nicolae Paulescu, a Romanian professor of physiology, developed a pancreatic extract which, when injected into a diabetic dog, had a normalizing effect on blood sugar levels.

In October 1920, Canadian Frederick Banting was reading one of Minkowski’s papers and concluded that the pancreas’s internal secretion, which, it was supposed, regulates sugar in the bloodstream, might hold the key to the treatment of diabetes. A surgeon by training, Banting knew certain arteries could be tied off that would lead to atrophy of most of the pancreas, while leaving the islets of Langerhans intact. He theorized a relatively pure extract could be made from the islets once most of the rest of pancreas was gone.

Banting’s method was to tie a ligature around the pancreatic duct; when examined several weeks later, the pancreatic digestive cells had died and been absorbed by the immune system, leaving thousands of islets. They then isolated an extract from these islets, producing what they called “isletin” (what we now know as insulin), and tested this extract on dogs. Banting was able to keep a pancreatectomized dog named Marjorie alive for the rest of the summer by injecting her with the crude extract they had prepared. Removal of the pancreas in test animals in essence mimics diabetes, leading to elevated blood glucose levels. Marjorie was able to remain alive because the extracts, containing isletin, were able to lower her blood glucose levels.

Task:

Briefly re-tell each of these 3 stories of scientific discovery using as many of the following words as possible: causation, conjecture, correlation, data, deduction, experiment, explanation, falsification, hypothesis, imagination, interpretation, law, mechanism, model, observation, paradigm, prediction, process, theory.