This blog is the fourth entry in a series by Steven Benner (the first post can be found here). Throughout the series, Benner discusses the nature of scientific progress and the difficulty of defining what is and is not science. Discussion questions are included at the bottom of each post.
In my previous post, I argued that science is a thoroughly cultural activity, fraught with potential bias. The power of science, compared to other disciplines, consists of its ability to challenge paradigms while keeping the primary evidence in view.
In most fields of science, the person who sets up an experiment is generally the same person who analyzes its results. Thus, a part of the discipline of training scientists is to get them to understand and manage how they participate in a scientific enterprise that is their own. It is, we teach, not wrong that they have an interest in the outcome of an experiment. Indeed, it is impossible for them not to have an interest. What is wrong is not to acknowledge their own interest, and not to mitigate it using the processes particular for their field.
The processes are different for different fields. In medicine, we insist that pharmaceutical trials be done "double blind", with neither the patient nor the physician knowing who is receiving the drug and who is getting the placebo. In chemistry, we might run an experiment under a range of different, but carefully controlled conditions, something impossible when dealing with human patients.
But in any field, the most successful scientists establish within their laboratories a kind of dialectic. On Mondays, Wednesdays, and Fridays, we believe one thing. And act like we believe it. On Tuesdays, Thursdays, and Saturdays, we believe the opposite, and act like it. We might even cherry pick data to make the case "pro" on Mondays, just to see how strong that case is. But if we do, we make sure that on Tuesdays, we cherry pick data to support the "con". Operating throughout is the ability to do experiments, mix reagents, observe stars, or follow moose. These all give reality an opportunity to slap us in the face, to remind us to be not quite so certain that we know what we are doing.
Above all, we teach scientists to distrust all measurements, but to distrust most those that confirm what we want to believe. All experiments should be repeated to make certain that their results are reproducible, of course. But the experiments that are most in need of reproduction are those that produced data that support the proposition or theory that the student wants to support.
Scientists can also deliberately set goals to drive discovery. One way is through "synthesis", the act of creating something new following a design based on currently accepted theory. Synthesis has become especially big in biology, where "synthetic biologists" try to create new proteins, or new genetic systems, or new genetic regulatory networks based on what we think we know about living systems. If the theory guiding our design is correct, it should be empowering; the synthetic protein, the synthetic gene, or the synthetic regulators should work. But if the theory is wrong, it might not be empowering. The synthesis will then fail, and fail in a way that cannot be ignored. Thus synthesis drives discovery and paradigm-change in ways that analysis cannot.
But suppose scientists do not establish this kind of dialectic internally? Suppose the scientist, enamored with his innovation, simply becomes an advocate, publishing data to support the innovation while burying data that contradicts it, rationalizing away contradicting observations by introducing ad hoc explanations? In this case, the scientist has lost for himself the power of science to discern knowledge. His innovation must now be evaluated by the community.
Fortunately, the community of science offers the opportunity for correction. Unlike in the law, advertising, or politics, science does not have a jury, market, or voter who stands above the dispute, hears both sides, and makes an authoritative decision. Instead, there are rules that deny the existence of an authority. Data are made public. Experiments are open to be repeated in the laboratories who want to attack the innovation. New experiments are designed by others to test the innovation.
The interaction can be rough and tumble, with advocates on all sides showing little of the dispassionate disengagement of ideal scientists. Sometimes, the dispute is not resolved until the advocating scientists die, to be replaced by a new generation of scientists who can dispassionately evaluate the dispute. But as long as politicians do not intervene, science can be self-correcting.
In this activity, a community can easily become divided into warring parties of advocates. One of these disputes well known to biologists (but to few other communities of scientists) is the "neutralist-selectionist" dispute, which consumed a generation of evolutionary biologists arguing over whether or not most genetic change influenced the fitness (of a moose, for example).
A dispute well known to chemists (but to few other communities of scientists) is the "non-classical carbocation" dispute, which asked how bonds should be represented in organic molecules carrying a positive charge. Known to physicists (but to few other communities of scientists) is the dispute over the "Copenhagen interpretation", having to do with uncertainty and quantum mechanics. In each dispute, every observation presented by one side was immediately contradicted by data selected by the other. The debates were largely unresolved until the initiating protagonists had left the stage.
How should a community react when our paradigms are challenged? Certainly, we cannot drop everything every time some crackpot decides to challenge what he learned in middle school. Nor can we expect scientific communities, composed of individual humans, to be more liberal than other human communities, where a challenge to orthodoxy is generally met by a response equivalent to "kill the heretic".
But a scientist is not allowed to dismiss this challenge out of hand as ridiculous. Scientists need a process to balance the fact that settled science may be wrong against the need not to waste time with truly ignorant challenges to science.
The first element of this process is a familiarity with the "primary data", the actual observations that underlie the science that is being challenged. A good scientist is always able to answer the question: "So you believe that the Earth is 4.5 billion years old. What are the primary observations that support your belief?” Thus, when a paradigm is challenged, the disciplined scientist can say: "Well, let me see. Suppose our world view is wrong? What primary data must we have misunderstood? What else in our current view of reality would need to be revisited?"
For example, if someone challenges the current paradigm of by asserting that the Earth is not 4.5 billion years old, but rather was created by divine intervention 6000 years ago, the correct response is not: "You are crazy".
The correct response is: “Well, maybe. But if that is what happened, then much else of what we think we know must also be wrong. We will need a new explanation for how the Sun gets its energy, as our laws about nuclear physics must be wrong. As this is the physics that has manifestly empowered engineers to build nuclear power plants, we need to explain how they are doing so well even though they are operating with the incorrect laws. The same would go for the empowerment provided by science for the use of radioisotopes in medicine X-rays in dentistry.”
All of this empowering knowledge would vanish if the current paradigm concerning the age of the Earth were wrong. Ultimately, it is this interconnection between biology, physics and chemistry, the engineers who are empowered by their laws, and the breadth of observations that are accounted for by those laws, that constrain the search for paradigms in need of revision.
This search is on-going within all of these fields. Evolutionary theory as it is currently structured is not able to explain all of the puzzles that observations of natural biology present us. Physics as it is currently structured is not able to explain all of the puzzles presented by observation of the cosmos. Chemistry as it is currently structured is not able to explain all of the puzzles presented by observations of the molecular world.
It is entirely conceivable that paradigms within these disciplines are ripe for replacement. It is conceivable that the various received views will need dramatic revision. But these replacements and revisions will come from those who create dialectics within their own thinking, are fully conversant with primary data, and are prepared to revisit "settled science" whenever prudent as their views are challenged. Not from those who enter the debate as advocates.
In my final post in this series, I will argue that science is misunderstood by the lay public in large part because other voices in the public square hijack scientific imagery to take advantage of its status.
Discussion Questions: Dr. Benner describes the importance of primary data in determining when a challenge to an established paradigm might be legitimate. Do you think scientists do an adequate job of explaining the primary data behind their conclusions to one another? What about to non-scientists?