(Originally posted on November 6, 2023)
Once upon a time, there was Scientific American. Douglas Hofstadter, who introduced me to feedback and recursivity, and Martin Gardner had regular columns back then. And if you browse through the pages of this 1981 issue, you'll realize how radically different the editorial content was compared to today's approach—regardless of considerations about the caliber of the authors, especially in the Italian edition, which would also be worth discussing...
You see, when I think of "science writing," I envision a standard set by Scientific American and New Scientist until the late 1980s. As for addressing current scientific topics in public debate, I cannot help but recall the masterful classic by Robert May and Roy Anderson, The Logic Of Vaccination (https://ilchimicoscettico.blogspot.com/2018/04/la-logica-della-vaccinazione-1982-uk.html). A classic that marks the abyss between the level and quality with which these topics were treated then and what is mostly called "popularization" or science writing today: no one, especially in Italy over the last five years, has wanted or known how to address the topic of vaccinations in this way. In fact, five years ago, when I reintroduced this text, I was mostly accused of heresy (the most dim-witted classified it as obsolete material). The fact is that May and Anderson knew and understood the subject well because they had been working on it for some years. The science writer of recent years, especially in Italy. not only lacked specific knowledge on the matter but also the conceptual tools to understand the topic, so they relied on a "medical community" that was, unfortunately, equally incompetent on the subject. And the result has been evident over the years, from the time of the furious italian debate on measles and mandatory vaccination to the last two COVID years when the only word allowed was "exponential," whether the curve was bending or declining. And every ripple in the baseline became the beginning of a new epidemic peak.
But let's return to "science writing." If I read some titles suggested by Nature (https://www.nature.com/articles/d41586-022-04236-9), I don't see anything that interests me (pandemics and climate are political issues, science has nothing to do with it), but they appear significant to me: they follow the market. But this is generally what is referred to in Italy when discussing scientific literacy: education on topics (and their current vulgate), not education on method—despite the vast audience of scientifically incompetent people who "spread science" and fill their mouths with the phrase "scientific method." Educating on method means, first and foremost, educating on practice, and sorry to say, in many high schools, the science laboratory was an unused relic of the past already in the 1980s.
When those who write about science come from scientific practice, it's usually noticeable, or at least I perceive it well. An emblematic case is Giacconi's "The X-ray Universe." But I also found Lisa Randall's writing skills fascinating. Forget for a moment the context of her Warped Passages (2006), that is, the factions for and against string theory, quantum gravity, and all the rest. And absolutely forget the book's marketing ("Randall has opened the doors to the multiverse"). It was, first and foremost, a brilliant attempt to write the history of recent physics without a single equation but minimizing the loss of information, a successful attempt, in my opinion (according to others, who later found themselves perfectly at ease with Carlo Rovelli's Seven Brief Lessons on Physics Randall's text is very difficult).
Her subsequent book, Knocking on Heaven's Door (2012), was released almost simultaneously with the detection of the Higgs boson at the LHC. This forced the author to write a long preface and an even longer introduction. Writing books on contemporary physics from a historical perspective during that period was an extremely risky work. Perhaps this is also why Randall's text initially disappointed me at the time. After more than 10 years, I reopened it and found it significant in more than one aspect.
First, reiterating the role that physical laws have, or do not have, depending on the scale at which the phenomenon occurs, which is far from common in science popularization, especially in Italy (I briefly discussed this here: https://ilchimicoscettico.blogspot.com/2021/10/chimica-gravita-posizione-del-sole.html). If people have lost count of those who have launched against quantum mechanics inappropriately, on the other side with crossbows drawn and the like, they have fallen exactly into the same pattern, bringing up Einstein where relativity has no significant role (phenomena described by classical mechanics and thermodynamics).
Second, the account of the author's relationships with interlocutors outside the "science profession": screenwriters, journalists, politicians. And it's something that reminds us that in the USA, certain phenomena begin about ten years ahead of Italy (the difficulty of relating to the "general public," which usually receives scientific knowledge in a badly distorted way).
Third: attention to the experimental part, from the construction of the LHC and the experiments to the problem of data and how to process it.
Fourth: the attitude of someone who has measured themselves against experimental evidence, respecting it. And here I cannot help but quote:
"Making precise and reliable predictions is a difficult undertaking. Even when one does one's best to model everything that is important, the input variables and assumptions introduced in a certain model significantly influence the conclusions. Thus, predicting a low risk makes no sense if the uncertainties associated with the initial hypotheses are greater than the calculated risk value. If a prediction is to have any meaning, it is important that the problem of uncertainties introduced into the calculation be carefully weighed and duly taken into account.
Before moving on to other examples, let me tell you a small anecdote that helps frame the problem. At the beginning of my research career, I observed that the Standard Model, for a certain quantity that was intended to be studied, had a more extensive range of values than previously thought: this was due to a quantum contribution, the magnitude of which depended on the surprisingly high value of the top quark mass, as recent measurements indicated. When I presented my results at a conference, I was asked to plot the trend of the new predictive data as a function of the mass attributable to the top quark. I refused, because I knew that several contributions would have to be taken into account and that the residual uncertainties involved too much variability in the result for such a curve to be drawn, as had been requested of me. It happened, however, that an 'expert' colleague underestimated the uncertainties and drew such a graph (not unlike what happens in predictions in non-physical domains).
The 'expert' colleague was quickly contradicted by experimental evidence and exhibited a behavioral pattern that should not be unfamiliar to you: I was wrong but I was right (and she was right but she was wrong). As for 'not unlike what happens in predictions in non-physical domains'... well, no need to rake up the events of two years of pandemic, from models to drugs.
I find this passage from 11 years ago or thereabouts frighteningly relevant. Let's skip over the difference in specific weight between those who write about science based on hearsay and those who do so based on competence and experience. It's also a matter of attitude: in the short term, respect for data doesn't pay off, treating it nonchalantly does (cf., again, two years of pandemic). Someone should have learned, even in the vast chaos of the web and social media, to distinguish between the ones and the others, in both categories (including hands-on experience and its absence, I mean). Then look at who writes about science on social media (or on the italian edition of Scientific American) and draw the appropriate conclusions, because it's about time.
.png)
.png)
.png)
.png)
