Science Magazine is Failing Us

Science journalism is failing us in important ways. This post will be far shorter than I might like it to be, but I want it to be readable, and in any case I plan on following up soon with more information and also, I hope, a detailed action plan.

Here I simply want to point out a single article. In another post, I’ll also be discussing an article on the dismissal of Francisco Ayala from UC Irvine and the pattern of sexual harassment that led to that dismissal. But right now, let’s tackle an interesting article with a headline that is … terrible, in ways we will investigate later. The headline reads thus:

The little piece of DNA that makes girls boys

We know, of course, that gender identities and gender roles are subject to change over time. While Will Roscoe has documented a small part of the gender diversity of indigenous nations of North America in Changing Ones: Third and Fourth Genders in Native North America, we also know that non-binary genders largely disappeared from the US landscape during the middle decades of the 20th century (in part because of the genocide of indigenous peoples) only to make a strong and quick comeback since the early 1990s. What is the explanation for the lack of people publicly identifying as non-binary during those lost decades? Could it really be that a sequence of DNA was lost and then re-emerged? Of course not. Not only can gender identities change with some persons’ new understandings of themselves, but entire gender roles can appear and disappear in ways that do not depend on DNA at all.

Then, in regular type, the body of the article describes some decent science, but with far too many failures:

What if you could flip a single DNA switch and make a world of only women? That sci-fi vision is unlikely to become reality anytime soon, yet such a switch—one near the gene that prompts the development of male body parts in embryos—has just been discovered in mice. The finding could help explain why some human babies with a male chromosome are born female ….

“[The switch] seems alone to be able to do the job” of making a man.

If left to their own devices, all human embryos would develop into girls. But a gene on the Y chromosome, named SRY, brings about a change in early development, causing testes, a penis, and other male traits to form. This gene indirectly turns on another gene called Sox9, which kick-starts the construction of the testes.

Note, please, that the research is about gonads and genital anatomy, but the author repeatedly describes gender as genetically determined. The odd thing isn’t so much that these conflations exist, but that they exist at a Science Magazine owned outlet for science journalism. This isn’t PopSci. This is Science. They can and should and do strive for a high level of scientific accuracy in their reporting. So how could such errors show up in such a respected science journalism outlet?

Well, first, we should recognize that this is the website and not Science Magazine itself. Yet I doubt that Science Magazine is entirely free of such errors. Second, we have to consider whether these really are errors – and if they aren’t, should they be called lies?

It’s this second bit that is so troubling to me. Research that shows gender is separate from sex is well established. An easily readable yet rigorous account of the basic separation of sex from gender and qualities of gender such as gender identity from from gender attribution has been available for 40 years now, and it’s grounded in research older still.

Is it possible that the writers and editors at Science simply don’t know what gender is or that it differs from sex? Given that Science reports on not only biology and genetics but also anthropology, sociology, and psychology, that seems highly unlikely. Worse, if they really don’t know what gender is then they are simply not competent to report on important topics that are addressed in Science-owned publications on a frequent and regular basis.

Does that mean that they’re necessarily and fundamentally lying, then, when they report that a single genetic regulator reliably produces not only penises and testes, but also men? Well, no. It doesn’t have to be that. I’m quite sure that when the more accessible accounts of physics research are written for ScienceMag.Org that these include misleading simplifications at which any physics doctoral student might scream but which I skim over, content, believing that I’m getting an accurate picture of what physics researchers have recently learned. ScienceMag.Org could be making a judgement that their readership – outside those who have specialized in a related field – simply don’t understand and can’t appreciate a distinction between boys and young male humans.

The thing is that I don’t believe that their readership is so incapable, and I also don’t believe that this is what actually motivates and justifies the editorial decision to conflate sex terms with gender terms. Rather, my first guess is that because “male” is a categorization used far beyond the human species while “boy” communicates information about age and species, the single word boy can, if sex and gender can be validly used interchangeably, replace three words, as seen above: young male humans. While the word boy might be objectively wrong in some contexts, it’s sure as heck shorter and more accessible!

So, then, what’s the harm? If all popular science journalism has to balance brevity and accessibility against rigor and accuracy, are the failings of ScienceMag.Org here significantly different than whatever failings it makes in describing plasma physics to an ethicist and lawyer with undergrad degrees in psychology and gender studies?

The truth is that I can’t know, but I don’t have to know. When science journalists fail in their reporting on physics, it will be up to the knowledgeable physicists to describe that failing, to analyze it, and to provide feedback on whether the level of error is an acceptable trade off for the gain in popular, albeit imperfect, understandings of physics. My job is to analyze and report back on areas where I do have some expertise.

Here, I think its clear that ScienceMag.Org is failing us. The level of approximation is simply not necessary. Few if any of its readers are going to fail to understand the words “penis” or “testes”, and if any do, they’re going to have much more trouble with

The researchers found 16 good candidates for the Sox9 enhancer. With other tests, they homed in on one that was 557 bases long and located half a million bases away from the gene itself. To turn on its target gene, such a distant enhancer is brought in contact with the gene by the looping of the chromosome they are both on.

Moreover, in making the approximation that sets “males” equal to “boys + men”, science journalists aren’t misrepresenting the details of a particular model, they are misrepresenting fundamental and terribly important concepts that are used constantly in at least 5 major fields.

Then, of course, there is the issue of the actual social impact: many people are asserting the opposite of the consensus of scientific understanding on the distinct nature of the concepts of sex and gender in order to advance political and social agendas that harm real persons. The writing in this article takes a side in that conflict. By leaving terms relating to gender entirely out of the article, ScienceMag.Org could have taken no side. By placing those terms in the article and making sure that they were distinguished from terms relating to sex, they would arguably be taking one side in the conflict, but also arguably could simply be fulfilling its mission to accurately educate the public on scientists, the scientific process, and their findings.  But in choosing how they have chosen, they cannot possibly fall back on the defense that they are merely reporting the science, because they have instead provided reporting contrary to known science.

Whether or not you like the political and social policies of the people who deny any distinction between concepts of sex and concepts of gender, ScienceMag.Org should oppose basing those – or any! – policies on fundamental misrepresentations of scientific facts. The reporting here should be rejected in the same way and for the same reasons we would reject any Science-related media property reporting on global climate by asserting that we don’t yet know for sure if mean global surface temperatures are warming.

The failure here is every bit as serious. We wouldn’t quibble about the meaning of “for sure” in that (hypothetical) assessment of the climate science and whether that misrepresentation could be proved beyond reasonable doubt to be somehow malicious, mendacious. Likewise, we can understand that there is room for quibbling about whether this conflation of sex and gender is entirely unintentional (and thus casts doubt on competence grounds) or egregiously malicious (and thus casts doubt on moral/honestly grounds) or simply an excessively lax example of compromising accuracy for pithy writing, we don’t have to hash that out to the a scientific certainty (and we shouldn’t) before we agree that however this conflation came to be used in this article, it is simply not acceptable.

We should demand better of respected science journalism outlets. That leaves us with two primary options: to relegate Science-related media properties to the category of disreputable science journalism, or to raise our voices in a challenge to Science to do better. I choose the latter.



  1. Allison says

    we also know that non-binary genders largely disappeared from the US landscape during the middle decades of the 20th century … only to make a strong and quick comeback since the early 1990s. What is the explanation for the lack of people publicly identifying as non-binary during those lost decades?

    Um, there was a general lack of people publicly identifying as any non-cis gender during that time. Gender policing was very, very strong starting after World War II, and anyone who didn’t fit the mould learned pretty fast to stay deep, deep in the closet. Moreover, non-binary people are even today routinely invalidated and erased by both binary cis and binary trans culture. I suspect that in the days when even binary trans was an outlawed underground subculture, non-binaries were doubly closeted and, when caught out, misgendered. (After all, we — gay, trans, lesbian, bi, binary or not — were all the same: queer, perverted child molesters, right?)

    Moreover, people with non-standard genders have been erased from histories. For example, how many USAans are aware that the Stonewall riots and the early gay rights movement were driven by drag queens, transgender people (mostly trans women), and homeless gay hustlers, with little to no involvement by the kind of middle-class gay men who write the histories?

    We have always been here; what has varied from decade to decade and century to century has been how (cis-dominated) societies have chosen to frame our deviances and how we have been able and/or allowed to express them (or not express them.)

  2. invivoMark says

    These article summaries in Science are meant for a lay audience. Actual scientists don’t read them – they skip straight to the actual publications. So they’re under the exact same pressures as National Geographic or Scientific American to make the language as accessible and the headlines as attention-grabbing as possible.

    I don’t say that as an excuse. I agree that the language needs to change, and places like Science should be leading that change. But I don’t think that we, as a society, have done a good job establishing the rhetoric so we can discuss these topics in an accessible way. Doing so is difficult and dangerous: include too many articles with the word “gonads” and people might stop reading because they feel the writing has become stiff and technical (or at least that’s the fear).

    The problem sometimes also extends to the technical writing in the publications themselves. In medical research, “male” and “female” almost exclusively refer to body types, and gender nonconforming individuals are usually pretended not to exist. Of course, most studies on the genetics of gender are also based on mice or fruit flies, which generally aren’t given much of a chance to express a gender identity that a scientist would notice. Still, it would be nice for authors to be mindful of their language.

    In the article in question, the study is entirely about mice. The only reference to human biology in the entire manuscript is, “[H]umans heterozygous for null mutations develop campomelic dysplasia, a severe syndrome where 70% of XY patients show female development.” This language works fine, in my view, since “female development” clearly refers to female-typical body arrangements and shouldn’t be confused with female gender identity.

    It might be a fun and productive exercise to think of alternative ways to write the headline for the pop-sci summary of the research. Is there a good way to do it while being accurate to the research but without throwing gender non-conforming people under the bus?

Leave a Reply

Your email address will not be published. Required fields are marked *