Question

Christiaan Eijkman shared the Nobel Prize in 1929 for the discovery of vitamins. Eijkman went to Java to isolate the pathogen causing beriberi, an illness that appeared to be contagious as it was common in prisons and among ship crews. Eijkman noticed that chickens fed a diet of white rice also suffered from a similar disorder. When the chickens were fed a diet of red rice, the chickens quickly recovered. Eijkman returned to the Netherlands and persuaded a prison to feed the prisoners red rice, and the incidence of beriberi decreased. What would the next step for Eijkman be?

Answer 1

He can infer that all prisoners should get the new diet right away.

He can infer that humans and chickens have very similar dietary needs.

He can infer that humans should eat a more varied diet.


He can infer that beriberi may be related to diet, but more and better studies are needed.

Answer 2

All of these are reasonable, and it depends a lot on Eijkman's responsiblity and interest. I think the first is probably a necessary step on simple humanity grounds: if you find a particular intervention saves lives or cuts short disease, it is unethical not to broadly and immediately recommend it.

The second and third are both reasonable tentative conclusions, or perhaps additional hypotheses to be tested next.

The last is a perfectly reasonable statement, the kind that would very often end a paper on this subject in the modern era. But it also has a bit of a useless wishy-washy character to it. He could have easily made it before he did his experiment as after, and as it stands it really doesn't do any good to any other researcher. A much better formulation would be something more precise and unqualified, viz.:

"He can infer that a contagious pathogen is not necessarily the only possible cause of beriberi."

That has the advantage of being precise and categorical. There's no "may be" about it -- it's a categorical statement, the kind that can rule out entire classes of experiments, or open up new classes.

Science generally proceeds not be constructing a lot of "may be" statements, and then striving to turn them into "probably is" statements, although it is often portrayed that way, and scientists carelessly speak that way, often enough. Science actually proceeds by making very sweeping categorical statements, usually of the negative variety, and then putting them to the test, which usually generates more sweeping negative statements.

For example, you are trying to find out why and how people get the plague. By a series of experiments, you could proceed from categorical statement to categorical statement, like so:

(1) You think the plague comes from wicked behaviour and evil thoughts. You do an experiment, and find that even babies get the plague, and people widely acknowledged to be virtuous die of it no less than those widely condemned as evil. Now you say:

"Evil thoughts are not the only possible cause of plague."

(2) You think it might be related to immigrants, so you do an experiment, and find that villages that haven't had an immigrant in 50 years succumb to the plague just like villages with lots of immigrants:

"Evil thoughts and immigrants are not the only possible causes of plague."

(3) You think it might be related to hygiene, and you do an experiment, and find that indeed households that are much cleaner get significantly fewer cases of plague:

"Evil thoughts and immigrants are not the only possible causes of plague, and the cause must be connected with hygiene."

(4) You figure it might be personal cleanliness, so you do an experiment where everyone in a given village agrees to wash every 12 hours carefully, and in a similar (control) village they do not. You find that there is no significant difference in plague rates:

"Evil thoughts and immigrants are not the only possible causes of plague, and the cause must be connected with hygiene in some other way than personal cleanliness."

(5) Finally, you hit upon the presence of rats in the house, and find that households without rats don't get thep plague, while those with rats do. This is what we often call a breakthrough discovery, because it allows a significant simplication and sharpening of your negative categorical:

"Households without rats don't get the plague."

And on we go. Eventually you'll narrow the categorical even further:

"Individuals who have not been exposed to the Y. pestis bacteria that typically lives in the guts of fleas who infest black rats do not get the plague."

That's the more precise formulation of what we would typically somewhat carelessly say as "Plague is caused by the Y. pestis bacteria."