Anxiety about science journalism

04cover-395This past weekend, the science writer Robin Henig had a fascinating article in the New York Times Magazine on the propensity of different individuals to be more or less anxious, and how research in Jerome Kagan’s laboratory has shown that such traits can be identified early in infancy.

There.  I think that I have summarized the article fairly, without hyperbole, without an overload of geek-speak, and without any obvious untruths.  The problem is that the article itself, as interesting as it might be, is guilty of hyperbole at a minimum, potentially of untruths, and in an effort to reach out to a mass (if educated audience) or completely avoiding geek-speak.  David Shenk over at Atlantic Correspondents has begun a very useful debate with Henig about the proper role of science writing, using the article of anxiety as an example of what is proper and what is not (part II is yet to come).  It is both refreshing and healthy that Henig and Shenk are having this debate publicly; scientists, journalists and (most of all) editors should take note.  Here are a couple of relevant snippets.

Robin says:

I know that your own particular beef, David, is the language we use to describe the gene-environment interaction, which makes words like “innate” and “inborn” especially loaded.   But I’d argue that what these words mean is just “present at birth” – an observation, not a presumption about provenance.  Maybe it’s time for you and me to go back to the article itself, and see the full sentences in which I used the words that bother you — and if a casual reader would have made the same assumptions you did.  I’d also like to hear how you would have me re-write those sentences in a way that makes it clearer that neither I nor the scientists are biological determinists.

To which David replies:

I completely agree that this is a very difficult task, to distill very complex science and ideas into easily-understood and hopefully memorable phrases. And it’s that much harder to do on deadline, which you had to worry about and which I effectively didn’t when I was writing my book. I also think you and I are swimming upstream here: we’re working to dispel long-held notions, using tools (specific words) that come loaded with fixed meanings in the public mind.

And so it goes, back and forth, with the two of them spending some valuable time trying to arrive at some reasonable conclusion of how to balance the pressures that science writers face with the insistence of scientists upon not overstating the case.  But before scientists criticize science writers for such sins, perhaps we should look in our own backyard.  As it turns out, scientists (and the associated scientific publishing establishment) are hardly blame free.

A recent Letter to the Editor of the journal Molecular Psychiatry has the title of  “The human serotonin transporter gene explains why some populations are more optimistic.”  The letter goes on to describe two observations: (1) that Brazilians are “highly motivated and enthusiastic”; and (2) that one particular polymorphism in the serotonin transporter gene is 2.5 times more prevalent in the Brazilian population than in the UK.  From this, they conclude that

the association of this variant with personality traits should be validated in other population studies, it would provide an interesting biological explanation for being more or less optimistic.

Really?  I am hardly an expert in population genetics, but even a cursory examination of the data reveals that the authors have overstated their case by a fairly wide margin.  OK, so scientists are people too, and perhaps one could argue that they have a natural desire to draw the most important conclusion possible from their data (meagre as it may be).  But what about the editors of the journal Molecular Psychiatry?  Don’t they have discretionary control over which letters get published and which do not?  Isn’t this what editors are supposed to do?  After all, Molecular Psychiatry is the premiere journal dealing with issues of genetic variation and behaviour.  Shouldn’t they know better?  Methinks the answer is self-evident.

Sadly, hyperbole in scientific journals is not a rarity.  An even more egregious example arose recently in a news item (OK, it is not a scientific study in and of itself) by Cassandra Willyard in Science with the title “Early risers are Mutants.”  The article is based upon a scientific report in the same issue of the journal Science with the admittedly geek-speaky title, “The Transcriptional Repressor DEC2 Regulates Sleep Length in Mammals.”  It turns out that the authors found a point mutation in the human gene hDEC2 in two affected individuals in one family, and these individuals were regularly early risers (full disclosure: I am an early riser).  They then generated transgenic animals (both mice and fruit flies) with the relevant mutations, and found that the resultant animals also had a short sleep phenotypes.  The science is elegant, the approach thoughtful, and the conclusions given by the scientists couched in appropriate language.

Our results demonstrate that DEC2 plays an important role in regulating daily total sleep time in mammals and that the control of sleep-like behavior may be conserved and regulated in a similar manner as far back in evolution as invertebrates.

Notice that the scientists did not suggest that early risers such as myself harbour a DEC2 mutation.  Why?  Because they know that the data in humans is based on only 2 individuals.  But the headline of Willyard’s news article suggests something completely different.  It is both sensational and misleading (even if it turns out to be correct in future studies).  Again, the journal Science is one of the most influential scientific journals on this planet, and the editors seemed to have abdicated responsibility for responsible reporting by allowing this headline to go forward.

The sad conclusion from these examples is that scientists and scientific journals (and in particular their editors) are sometimes guilty of allowing hyperbole to invade the scientific enterprise.  I have no data to suggest that findings regarding the brain are more subject to being overstated than other fields, but it is worth remembering Deena Weisberg’s paper, “The Seductive Allure of Neuroscientific Explanations” in which they found that,

Explanations of psychological phenomena seem to generate more public interest when they contain neuroscientific information. Even irrelevant neuroscience information in an explanation of a psychological phenomenon may interfere with people’s abilities to critically consider the underlying logic of this explanation.

The data from the Weisberg paper suggests that individuals with expert training are less susceptible to such pitfalls; perhaps it is worth probing the extent to which science writers (and editors), working in the real world, are falling under the sway of such seductions.


6 thoughts on “Anxiety about science journalism

  1. Typically wonderful post, Peter! While you well know my frustration with the triumphalism and hyperbole of both lay and professional discourse as to neuroscience, I get even more rabid when it comes to genetics. I continue to be amazed at how many people — again, lay and professional — seem to engage in the fallacy Jeremy Freese observes of conceptualizing gene expression as the combination of x% “genetic causes” and 1-x% “environmental causes.” This could not be more wrong, since, as Lewontin and Evelyn Fox Keller have amply demonstrated, genetic information only expresses itself as part of a complex, nonlinear, multifactorial process involving hundreds, perhaps even thousands of different variables. There is, quite simply, no such thing as a genetic component of anything that is independent of social, cultural, and environmental factors. The belief in some hermetically sealed concept of “biology” separate and apart from “society” is one of the most pernicious and persistent beliefs I perceive in discourse on genes and health (it is also well-documented in the literature). It is what Lewontin refers to as the “doctrine of DNA.”

    It drives me crazy, honestly.

  2. Daniel – in response to your post. I am a mildly informed layperson of a wide variety of science. The nature/nurture debate I agree is an unfortunate holdover from an era where neither side knew nearly as much about the interaction as they do now. I’ve read Lewontin and Oyama concerning their cautions about the use of language and the “doctrine of DNA.”.

    Peter’s post, however, concerns how difficult it is to write well, accurately, and for a mass audience. It is so easy to make assertions that can be undone by professionals. In your comment to Peter you say that there is no genetic component of anything independent of other factors. I really would like to know if there is any exception to this comment – birth defects, genetic mutations that are truly transcription errors – not due to outside influences. I’m not trying to be provocative – I really want to know – are there sources I can read about this?

  3. Ron,

    I apologize for the delay, but I have been traveling. I suppose the obvious answer that might be presented to your question would be monogenically-caused conditions. Social conditions or not, if you have the alleles for Huntington’s chorea, you are going to develop the illness. But even in those cases, we have no ability to rule out the presence of social and environmental factors that affect the distribution and prevalence of these conditions.

    What’s more, as Stephen Pemberton and Keith Wailoo have suggested, the fact that 99% of all illnesses are affected only by polygenic contributions suggests that even if there are exceptions to my claim that genetic factors NEVER operate independently, those exceptions are exceedingly few and of little importance. There is more that can be said here, of course, so please feel free to email me if you’d like to continue the discussion.

  4. Thanks for the response. I do see your point about not being able to totally rule out social conditions even for something as “environmentally independent” Huntington’s chorea. That leads me to consider the flip side. If Huntington’s (I’m using this example as a proxy for other similar monogenic conditions) is not dependent upon social or environmental conditions – or at least unknown external conditions, I wonder if there are examples of gene expressions that require a very specific set of social or environmental conditions for an interaction to trigger. Do you know of any? Again, I’m no scientist, only an interested layman. (And I think my question could contain a pretty interesting screen play scenario.) I will review what I can of Pemberton and Wailoo.

    • Ron – there is a rather famous example: Caspi et al. Role of genotype in the cycle of violence in maltreated children. Science (2002) vol. 297 (5582) pp. 851-4. Here is the abstract:
      We studied a large sample of male children from birth to adulthood to determine why some children who are maltreated grow up to develop antisocial behavior, whereas others do not. A functional polymorphism in the gene encoding the neurotransmitter-metabolizing enzyme monoamine oxidase A (MAOA) was found to moderate the effect of maltreatment. Maltreated children with a genotype conferring high levels of MAOA expression were less likely to develop antisocial problems. These findings may partly explain why not all victims of maltreatment grow up to victimize others, and they provide epidemiological evidence that genotypes can moderate children’s sensitivity to environmental insults.

      The essence of the finding is this: if you have the MAO polymorhpism and are treated badly as a child, you grow up aggressive with antisocial behaviour; if you have the MAO polymorphism and are treated well as a child, you grow up normally socialized.

      The paper remains controversial, but represents a paradigm that is relevant to this discussion.

  5. Thanks, again Daniel. I appreciate your thoughtfulness. I am familiar with the MAOA controversy – ironically enough it has shown up in a current episode of the CSI television show. While the jury is still out, it may be this is one of those very high correlation instances. Thanks for your thoughts.

Leave a Reply

Fill in your details below or click an icon to log in: Logo

You are commenting using your account. Log Out / Change )

Twitter picture

You are commenting using your Twitter account. Log Out / Change )

Facebook photo

You are commenting using your Facebook account. Log Out / Change )

Google+ photo

You are commenting using your Google+ account. Log Out / Change )

Connecting to %s