The Importance of Humility in Science — A Philosophical Musing For the Weekend

Visit Us
Follow Me
Socrates once said, “All I know is that I know nothing.” Centuries later, St. Paul, the great expounder of Christian theology, ethics, and mysticism, said that “any man who says he knows something does not yet know as he ought.”

A very wise faculty member and department head of a science program I once spoke with told me that the more we learn as scientists the more we learn how little we know, and that he tells all his PhD students when they graduate to never brag about what they know but to go out into the world bragging about how little they know.

This type of humility is essential to science. Not simply to avoid the arrogance that often comes with knowledge, but because our understanding of the world around us often is, in fact, much more limited than we realize or want to believe.

In quantum mechanics, Werner Heisenberg (1901-1976) offered the Heisenberg Uncertainty Principle. Heisenberg showed that we cannot simultaneously determine an electron's position and momentum. The more we study the position, the less able we are to determine the momentum; the more we study the momentum, the less able we are to determine the position.

There is a similar principle that exists in nutrition. The greater the certainty with which we can determine cause and effect, the less certain we can be that our knowledge pertains to the real world. The more certain we are that our knowledge pertains to the real world, the lesser the certainty with which we can determine cause and effect.

Consider for a moment the “gold standard” of evidence for nutritional and medical treatments: the randomized controlled trial. It takes place, at least in part, in a laboratory or hospital setting. The subjects know they are being studied. They know they are receiving some type of treatment or placebo. They are getting their blood drawn. If we really want to understand cause and effect, the subjects may have to be placed in a metabolic ward, where their exercise and food intake is strictly controlled.

If the treatments really need placebos, this alone is a profound admission that every aspect of the study has a psychological impact with physiological consequences. Are the patients thinking the same things? Feeling the same way? Doing the same fun things, the same boring things? No. They may be bothered by blood drawings; they may be hopeful about treatments. They may be analyzing how they eat and exercise more closely. They may be conforming their behavior to the expectations of physicians and scientists.

Is the nutritional or medical treatment that proves effective within this type of setting effective in the real world? Is the one proven ineffective under this model without any effect in the real world? We simply don't know, and can't prove it one way or the other.

If we wish to study with great precision the detailed mechanisms of cause and effect, we turn to studying cells isolated in a petri dish, or we react molecules with molecules in a test tube. We thus establish with great certainty the laws of biochemical reactions but must then begin asking, to what extent does a test tube approximate the living environments within which such biochemicals ordinarily find themselves? To what extent does bathing a cell in a compound approximate the effect of eating it? Do immortalized, usually cancerous, cells behave the same way in a petri dish as normal, healthy cells behave as part of human tissue?

Consider, on the other hand, the observational study. Scientists observe free-living people doing what they would ordinarily do. In some cases the people may be enrolled in a study while in other cases they may not even know they are being studied. In these cases, we can be highly confident that the truths we obtain are applicable to the real world because it is that real world we are studying. Yet at the same time we lose our ability to determine cause and effect.

When all of the different types of evidence agree with each other, we may become more confident that we have discovered a universal truth. But there is always a degree of uncertainty underlying the knowledge we obtain in the field, and the more closely we study a phenomenon, the more we risk distorting it.

The job of the scientist is, of course, to develop models of studying phenomena that produce the least distortion of the real phenomenon. But the scientist also must exercise a healthy dose of humility, and admit that her or his knowledge is but a drop in the ocean of truth.

Visit Us
Follow Me

You may also like


  1. If we did figure some things out with an absolute certainty, then we would be robbing future generations some imagination and mystique. Why be dull?

  2. Thanks for the good reminders. As Lao Tze said, "those who talk do not know, those who know do not talk." Always makes me want to shut up…

    When I was in academia, we used to joke about MDs doing research, since they are not trained, as you point out, in the scientific method, and tended to do sloppy research. I'm not sure there is such a difference when I look at the literature though. Researchers from all fields seem perfectly willing to let their biases influence what observations and conclusions they make. Einstein suppsedly said that theory determines what can be observed. If a researcher is looking for a certain effect or correlation, then they may not even notice other important factors, and if they don't find what they are looking for, they may not even publish, as you mentioned in your recent post about Krauss's metaanalysis of saturated fat trials. This is not science, and I don't believe much of what is published these days passes for science either. People are under pressure to publish, publish and publish, and get money for their labs and departments, and hobnob with their peers and thought leaders, and the money/prestige/speaking engagements etc are all contingent on playing the game and not going too far afield unless you have true "cajones". And even that usually just leads you to a deadend. I've heard too many stories about people getting into disputes with the powers that be in their field of endeavor and not being able to get published or get grants.

    I wish the true spirit of scientific inquiry was rewarded and some of that humility could be recovered.

  3. Nice post, Chris.

    (Isn't the PhD mentality 'ooh, I think I'll stay on at university a bit longer so I don't have to get a job..'??
    Just kiddin'! I did mine because I was totally blown away by quantum mechanics.)

  4. Great points, Jim!

    Many physicians of course are in emergency-oriented situations where the "MD" mentality is a great asset and even a necessity; others less so, and many somewhere in between. Certainly the primary care physician would best serve people by having a little bit of the "PhD" mentality. And we certainly need open communication between both types of people.

    Thanks for commenting!


  5. We certainly must not overlook the major role played by the scientists doing the observations.

    Just as beauty is said to be in the eye of the beholder, the observational findings are in the mind of the observer.

    This is a terribly over-broad generalization, but I think there is a fundamental difference between the thinking processes of Ph.D.s and M.D.s.

    The Ph.D. is trained to take lots of time to seek new information, and to patiently go wherever the data lead.

    The M.D., on the other hand, often has to make very quick decisions based on very limited information. If a dying patient appears in the ER, action has to be taken quickly, without the luxury of lots of time. Even in a doctor's office, time is limited because patients in the crowded waiting room often have to treated rapidly, in assembly line fashion.

    It would be great if we could have health care provided by people with PhD. mentalities.

  6. Chris, Sorry to nitpick, but it's position and momentum that you can't determine simultaneously for an electron (or if you prefer, position and velocity – since momentum = mass x velocity).

Leave a Reply

Your email address will not be published.