So I want to talk about flossing, not because I have very strong feelings about dental hygiene, but because there’s been some interesting developments in what we know about flossing in the last few weeks that I think serve as a good model for one of the biggest challenges in talking about scientific and scholarly knowledge.
I’ve talked before about how powerful conventional wisdom is – as humans, and especially as scholars, we like to believe that we approach each new question with fresh eyes, weighing up the evidence on both sides to come to a conclusion, but in reality, we measure new knowledge against what we already know. It’s a particularly flawed system because the information we’re given when we’re younger is often oversimplified, if not flat-out false, and obviously the more we study a question, the more we should find out, so future information should always be better than past information.
Which brings me to flossing – we’ve been told for decades that flossing is ‘good for you,’ by which is generally meant that it reduces long-term plaque buildup on our teeth better than just brushing and regular dental screenings and cleanings. A few weeks back, an Associated Press journalist ‘took down flossing’ (as Poynter puts it), publishing a lengthy review of available studies and a freedom of information request which revealed that the DHHS had actually quietly removed flossing from its recommendations. The article even implies that the decision to remove flossing may have actually been the result of the FOIA (which the AP reported had submitted the previous year, citing that all DHHS guidelines need to be based on scientific evidence and requesting the evidence base for the flossing recommendation). As news agencies tend to do, the story blew up in newspapers, talkshows, and online, with headlines bordering on just announcing, “EXTRA! EXTRA! EVERYONE STOP FLOSSING IMMEDIATELY!”
The people who actually work in dental health then started responding, pointing out that clinical studies on flossing are difficult to execute because they rely on self-reporting and at-home treatments, which are always the weakest options for getting consistent results, and that any long-term study might actually violate the rules for human testing because if, after 20 years of telling a control group not to floss, it turns out that it did cause them to develop periodontitis, the study itself would be liable for their condition. They also noted that there’s still sound logic for flossing, that it breaks up plaque formation on parts of the tooth not reached by brushing, and that the description of available evidence as “weak” and “very unreliable” by the Cochrane Library (the group that reviews scientific studies, whose analysis started this whole discussion) are precise, hierarchical categories used to describe and rank scientific studies, and that the repetition of these terms in news reporting about the Cochrane Library’s report are unrepresentative because the general public takes these terms to mean something different.
Now, to start with, there are a ton of things wrong with the current system for funding and reporting scientific studies, and a ton more in how the media reports scientific reports. There’s a great primer on this from John Oliver’s show, and some more info specifically on how the media reports healthcare/nutritional information from Adam Ruins Everything. But these issues don’t really explain why we believed in flossing or what we should do about it because, as any good scientist will tell you, science is incremental, and scientific studies only focus on one or a few specific aspects of a much bigger issue or question.
As someone who works in research administration, one of my jobs is writing and copy-editing scientific grants, and every grants I’ve ever worked on opened with something to the effect of, “____ is a huge public health threat in the US today,” often without even a citation at the end (insert the general topic under which this grant falls – obesity, diabetes, heart disease, etc.). I’m sure most studies on flossing read the same way – I’d bet most of them opened with the statement “regular flossing seriously reduces the risk of periodontist” and then went on to discuss some particular aspect of flossing that that study would test. The big issue gets treated as an axiom – something assumed to be true for the sake of the argument which follows – but without anyone admitting that it’s axiomatically true, not evidentially true.
It’s important the general public understand the serious problems with the state of scientific research today, but I’d argue it’s just as important – if not more so – that we all understand just how limited we are in addressing these axioms. Again, the debate about flossing hits most of the high points – on the one hand, we can use literature studies to highlight gaps in significant evidence, and on the other hand, we can use logic and anecdotal evidence to try to explain deductively why the axioms might be true, but both options are hindered by our natural predisposition towards conventional wisdom.
Literature studies are often rejected by the very fields they investigate because even scholars and scientists have problems overcoming the sway of conventional wisdom. In the case of flossing, the Cochrane Library report dates from 2011, but obviously wasn’t widely circulated or reported on until this one journalist highlighted it five years later. The responses to the AP article by dentists and other oral health experts all have more or less the same tone – the Cochrane Library findings are interesting, but not interesting enough to overcome my feeling (as an expert) that flossing works, and besides, the risk of us being wrong is too severe for us to even consider behaving differently. But that’s not the scientific method – that’s intuition and habit, albeit intuition based on relevant personal experiences.
A similar debate has (very quietly) taken place across the last several decades about the efficacy of dieting and weight loss. Literature studies dating back to the 1990s have suggested that long-term weight loss is just not possible, and these findings get revived every 5 to 10 years (like from UCLA and Melborne), and yet nothing could be farther from the message most of us hear about weight loss and health, not only from the media, but from the medical establishment, as well. Why? Well, the six billion dollar dieting industry probably has something to do with it, but in responding to these studies, scientists often take the same tone as with flossing – this doesn’t feel true, and besides, the risk is too real if we change things.
The reliance on logic to bolster axioms goes hand-in-hand with the power of conventional wisdom. Again, as humans, we often think of logic as an objective, self-regulating *thing* – that if something is logical, that means something about its realness that extends beyond our own perception. However, logic is a system that we invented to try to systematize understanding things that aren’t easily perceived, like abstract concepts. It’s still grounded in our perception, and it can vary greatly from person to person. Indeed, if you listen to people argue, often the argument will boil down to both people stating their logic, and then getting angry that the other person doesn’t accept it or continues to believe their logic is a better fit. That’s because logic isn’t abstract or objective – it’s based on what axioms we presume at the outset, what laws we lay out to work within. If these axioms aren’t specified, or if we want to question the axioms themselves, the whole system breaks down.
To take the example of flossing, it may sound logical to say rubbing a piece of waxed filament between your teeth breaks up plaque formations, but it’s also logical to argue that since plaque forms first on a molecular level, using floss to break it up would be like using the side of a skyscraper to push a balled up piece of paper across a street. Being able to formulate an argument that sounds logical to support an idea doesn’t make the idea true – that’s the very definition of a justification.
Unfortunately, there’s no real solution here, but at the very least, these issues about testing or understanding axioms demonstrate that how desperately we need new and better language for discussing scientific and scholarly findings. We often talk about scientific findings as if they were the very definition of facts – indisputable, consistent, not open to interpretation or variation by perception, but that’s just not true. All scientific ‘facts’ are based on some axiomatically thinking, and like with any logical proof, we need a way to go back to the start, lay out exactly what those axioms are, and discuss whether they are still sound and reasonable assumptions. At the very least, we need a way to be able to identify them as assumptions. In the short term, it’s probably fine to carry on flossing, but not if it means we all stop asking why exactly we’re doing it.