review of Neil deGrasse Tyson’s book Starry Messenger

Shelve this under scientism, not science. I've got two basic problems with it:

  1. It paints an implausibly rosy picture of how well scientists actually live up to the ideals of the scientific method.
  2. It encourages a haphazard approach to interpreting data rather than rigorous inspection of one's implicit assumptions.

First complaint: Scientists aren't science

The prologue extols the "self-regulating system within science":

There's no law against publishing wrong or biased results. But the cost to you for doing so is high. If your research is checked by colleagues, and nobody can duplicate your findings, the integrity of your future research will be held suspect. If you commit outright fraud—if you knowingly fake data—and subsequent researchers on the subject uncover this, the revelation will end your career.

OK, but the replication crisis is a thing, so this process seems to have a lot more room for error than that paragraph might lead you to believe. Tyson himself mentions in a later chapter how the scientific establishment’s endorsement of eugenics promoted racism in the early 20th century. And those of us who lived through 2020 in the US watched the reputedly scientific viewpoint on masks shift almost overnight, from confident denouncement of anyone who dared to wear one to confident denouncement of anyone who didn’t. Perhaps the truth always wins eventually, but big mistakes can be made in the mean time.

Which makes sense. Scientists are humans, their colleagues are also their friends and community, and their science is also how they make a living. It would be astonishing if they were somehow exempt from the pressures that lead to corner-cutting, corruption, and groupthink in other parts of society.

...conformity in science is anathema to progress. The persistent accusations that we take comfort in agreeing with one another come from those who have never attended scientific conferences. Think of such gatherings as "open season" on anybody's ideas being presented, no matter their seniority. That's good for the field. The successful ideas survive scrutiny. The bad ideas get discarded.

I haven't attended any scientific conferences, and this does make me curious what they're like. I would note that pressure to conform can come in many more forms than only a demand to respect seniority. What social forces govern who is invited to speak and who is invited to attend? What topics are not raised because of cultural or political taboos?

Conformity is also laughable to scientists attempting to advance their careers. The best way to get famous in your own lifetime is to pose an idea that counters prevailing research and that earns a consistency of observations and experiment. Healthy disagreement is a natural state on the bleeding edge of discovery.

I'd be interested to see polling data among scientists—and former scientists / people who dropped out of PhD programs—regarding how often they feel pressured to conform to (or dissent from) the prominent views in their field. Some other things I've read from people in academia take a more cynical perspective, more in keeping with the famous adage that "science progresses one funeral at a time".

Regardless, my point is not to denigrate the scientific community or cast aspersions on scientific institutions. Like Tyson, I want the average American to have more faith in science, and scientists, than they do now. But if you tell people a just-so story where the average scientist is superhumanly rational and scientific institutions magically transcend all corrupting influences, it undermines your credibility. The story defies common sense and counterexamples are too easy to find.

Second complaint: Science is only part of rationality

Tyson thinks the scientific method boils down to this:

Do whatever it takes to avoid fooling yourself into believing that something is true when it is false, or that something is false when it is true.

That's a great quote! I think it would serve better as a definition for rationality in general than for science in particular—which would be a pedantic quibble, except that Tyson also (rightly) closely connects science with "data":

One can't realistically expect people to argue in the same way scientists do among themselves. That's because scientists are not in search of each other's opinions. We're in search of each other's data.

If science is both (a) the key to not “fooling yourself” and (b) mostly focused on data, the implication is that not fooling yourself is almost entirely a matter of knowing empirical facts. Life would be simpler if that were true, but it's not. We also often fool ourselves by drawing unjustified conclusions from the facts. Tyson's comments on many issues fail to advance the conversations because he expects the facts to speak for themselves. Consider this paragraph on abortion:

Let's look at recent abortion rates in the US. Of the more than 5 million pregnancies per year between 1990 and 2019, nearly 13 percent were medically aborted. Yet all by itself the uterus spontaneously aborts as many as 15 percent of all known pregnancies during the first twenty weeks. Many more miscarriages go unnoticed since they occur in the first trimester, often before you know you're pregnant. Combined, the number of spontaneous abortions may surpass 30 percent of all pregnancies. So, if God is in charge, then God aborts more fetuses than medical doctors do.

Tyson doesn't develop this into a clear argument—he says he's merely providing "some perspectives to consider as we all take sides on who controls our bodies"—but presumably the reader is supposed to think, oh, if abortion happens so often naturally, it's weird to prohibit doing it on purpose.

I'm pro-choice and I think convincing more people to be pro-choice is important. You don't convince someone by lobbing half-baked weak arguments at them, which is what this is. Theists have already bitten the bullet on believing that a good God presides over a world riddled with horror; they already believe God is allowed to do things that would be grave sins for humans to do: they don't think the existence of cancer makes it OK to murder people, or the existence of natural disasters makes it OK to bomb cities. Given their beliefs about the foundations of morality and the expressed will of God, it would be nonsensical for them to see any connection at all between facts about the prevalence of spontaneous abortion and the question of whether intentional abortions are morally permissible. Reciting facts like this and expecting them to influence anyone's opinion merely reveals to your opponent that you have not made any serious effort to understand and engage with their point of view. They will, therefore, feel justified in ignoring you.

(Pro-lifers make the same mistake when they recite facts about when a fetus has a heartbeat or feels pain, etc., as if those facts should settle the debate. Most pro-lifers would not accept hey, cows feel pain! as an immediately decisive argument for veganism, nor burglars have heartbeats! as a decisive argument against allowing people to use lethal force against a home invader. If you don't make an effort to understand why pro-choicers think killing a fetus is morally different from killing an adult, you can't expect to change our minds. But that requires engaging with a complicated tangle of questions about the basis of morality, about what makes human life sacred and how we determine the boundaries of that sanctity, and about how to resolve conflicts between different individuals' rights. If you think you can say everything important about the issue in a soundbite, you don't understand the issue.)

I don't want to caricature Tyson; I don't think he literally means to equate data and reason. He brings up plenty of non-empirical considerations in passing. But generally the book gives the impression he thinks the first argument that came into his head totally resolves the issue (not bothering to consider counterarguments); or else that he thinks because it can't be resolved that quickly, any opinion is as good as any other. The chapter on vegetarianism is particularly awful; for example: could choose not to eat animals at all, living life as a vegetarian, but when you think about it, that's being speciesist against plant life.

There's a pretty obvious reason to be speciesist against plants and not animals: most of us don't think plants feel pain. Tyson hints that maybe that's incorrect, noting some "behaviors ... have been analogized by botanists to various human emotional states such as pain, joy, fear, and anger." But do those botanists mean it literally, or just as an analogy? If plants literally feel pain, are there more and less painful ways for us to make use of them? Tyson doesn't show any interest in trying to get to the bottom of things. Rather, he follows a line of thinking just far enough to get the conclusion he wants—in this case, that meat-eating is fine—then sort of throws his hands up. But it's entirely possible to think through these issues in detail and reach considered, defensible positions. I recommend reading books that make an effort to do so—like Practical Ethics—instead of this intellectually lazy chapter.