In “Why Our Children Don’t Think There Are Moral Facts,” Justin P. McBrayer observes that his second-grade son has been taught to distinguish between facts (which can be “tested or proven”) and opinions (which are just what “someone thinks, feels, or believes”).

In the category of “opinions” are placed all moral claims, including “Copying homework assignments is wrong,” and “All men are created equal.” Presumably, if a child says it is wrong to kill someone for the fun of it, that is labeled an opinion.

McBrayer notes that the same school that teaches his son to view moral claims as opinions also insists that it is really is wrong to cheat and really important to protect other students’ rights. I assume that the school not only proclaims these ideas explicitly but also builds them into its “hidden curriculum” of norms, expectations, punishments, and rewards. By teaching moral values while defining them as opinions, the school contradicts itself.

McBrayer has not just discovered an educational fad or a politically controversial agenda being pushed lately by a small group of adults under our noses. The fact/opinion distinction, as it is taught to his son, is a troubling hallmark of our age.

For instance, education is deeply influenced by standardized testing. What is tested will determine what McBrayer’s son learns in school for the next decade. I have been involved in writing exams, such as the federal government’s National Assessment of Educational Progress (NAEP) in Civics. This is an excellent instrument, supported by impressive science. Much skillful effort is devoted to identifying questions that yield good statistical results. Proposed questions that produce anomalous scores get cut. Based on their scores, the higher-performing students are labeled as “proficient” or “advanced.”

But each item on the NAEP is fundamentally a value-judgment. Should a citizen know the text of the Second Amendment, how many votes it takes to pass a law, or the history of racist violence in the US? Is a young person who understands half of these topics a “proficient” citizen, or “below basic”?

There are no scientific answers to those questions. They are matters of value, on which the entire edifice of testing rests. Yet all the official discourse about standardized tests skirts value questions and dwells on the statistics.

A Nazi civics test could be scientifically valid and reliable. It could work beautifully to identify young Nazis. It would be evil, whereas our standardized tests are at least reasonably decent—but the difference is not scientific. It is a moral matter.

Going beyond tests, the whole educational system that serves Prof. McBrayer’s son is built on techniques and practices scrutinized by science. The No Child Left Behind Act (still the governing federal law on k-12 education), favors forms of instruction supported by “scientifically-based research.” Randomized experiments count as the most scientific.

Thus, for example, experiments endorsed by the federal government show that paying teenagers to stay in school can cut their dropout rates. Another approach that also seems to lower dropout consists of “weekly after-school discussion groups … on personal, family, and social issues,” such as those arranged by a program called Twelve Together.

These very different programs are both presented as proven by science. But it is not self-evident that completing high school is a valid target, especially given the kinds of schools we actually provide. To identify graduation as the goal is a judgment. If such judgments are mere opinions, then there is nothing more to be said about them. But surely we can reason about the ends of education.

We should also reason about means. Could paying teenagers to stay in school “work” (boosting their graduation rates) yet still be wrong? Could it be an example of treating human beings as objects rather than autonomous subjects?

Finally, nothing just “works.” Ideas that are ready to be scientifically evaluated have always been designed, advocated, funded, implemented, tweaked, and refined. That implies effort by teachers or other front-line practitioners, administrators, and social scientists. A wide range of ideas can be made to work if the investment is sufficient and skillful.

But what we should invest in is a value question. We could start by paying teenagers to stay in school and work to make that a highly effective program. Or we could start by teaching them philosophy and refine our methods until that keeps them in school. Which approach we should try to make work is again not a scientific question but a moral one. All the scientific data on “effective practices” follow from our fundamental moral choices.

I have used educational examples here to connect to McBrayer’s article, but the same modes of thinking will be found in health, environmental protection, labor—indeed, all domains of policy and practice. A simplistic fact/opinion distinction influences sophisticated scholars and policymakers as much as 2nd graders and their teachers.

To be sure, budding social scientists are taught that values matter; they influence people’s behaviors and actions, and they influence social science itself. But this influence is treated as a problem. In the “limitations” section at the end of a scholarly article, the authors may confess that they have a “bias” in favor of certain values.

But moral commitments are not limitations; they are preconditions of decent scholarship. The difference between valuable and harmful social science is that the former manifests good values.

Science has achieved prodigious successes in understanding and controlling nature. It can also debunk certain assertions that are morally problematic, for example, that white people are biologically superior. But science cannot demonstrate most moral claims.

For instance: every child in second grade has the same moral value and importance. Looked at from a scientific perspective, that statement makes no sense because value is not a scientific idea. Or perhaps the statement is scientifically false, because science translates “value” into something like capacity or functioning, and not every second-grader does function at an equivalent level. We can try to equalize their capacity by devoting care and resources to the children who need it most—but science provides no reason to do that.

The influence of a simplistic fact/opinion distinction is not the fault of philosophers, who have always viewed the topic as complex. But it is philosophy’s responsibility to challenge the distinction that is so prevalent today. Otherwise, not only will we teach second-graders to view morality as mere opinion, but we will build massive social institutions on the same untenable premise.