On November 25, I gave a talk with the above title at an event hosted by SAAFoST* and ADSA**. Unfortunately, the proceedings weren’t recorded, so you won’t be able to hear the superb presentation that preceded mine, by Dr. Celeste Naude, who focused on an evidence-based approach for differentiating between varying macronutrient-focused diets.
Those of you who are interested in the topic of diets, and specifically the role Prof. Noakes has played in popularising the LCHF approach to diet, might already know of the recent study by Naude and others, which found that low carb diets showed a similar reduction in weight to other diets. Noakes’ response to that study was to say that the “researchers have no clue”. I leave it to you to determine who you find more persuasive.
You won’t be able to watch my talk as presented either, but in case it’s of interest, I decided to record a version of it in any case, accompanied by the slides I showed on the day. By contrast to Dr. Naude, who focused on science, I focused on rhetoric, hyperbole, and sound scientific reasoning – or, the lack of it.
You can find that recording immediately below, followed by the approximate text of the presentation. It hasn’t been edited into essay form, so is telegraphic in places. Lastly, I’ve embedded the presentation slides at the end, for no particular reason.
*SAAFost: the South African Association of Food Science and Technology
**ADSA: the Association for Dietetics in South Africa
Is Noakes the North Korea of epistemology?
Betteridge’s Law – any headline that ends with a question mark can be answered with “no”.
Of course, “no” is part of the answer here – if we are asking the question of whether Prof Noakes is a propagandist who leads a repressive state, and is implicated in various human rights abuses, the answer is clearly no.
I’m also make no claims with regard to their relative levels of sincerity. In the case of Prof. Noakes, I’m of the view that he is utterly sincere, and desires nothing but to enhance the health and wellbeing of those he engages with. He’s received far too much abuse related to claims regarding a profit motive, for example, and I think that sort of abuse unjustified, and deplorable.
But that’s not what my title is alluding to. Instead, I’m highlighting the fact that the North Korean press machine has a habit of making hyperbolic claims, and Kim Jong Un for appearing in various baroque, grandiose, and sometimes merely perplexing situations, all to buttress his mystique and support a particular narrative.
This narrative is of him being misunderstood, a maverick, and a person who has privileged access to knowledge and opportunity that he is able to share with the enlightened or anointed. He serves as an inspiration, and in doing so, the impression created is more important than the evidence – marketing is the point, rather than content.
This is the sense in which Noakes is the North Korea of epistemology. As I will show, he displays a pattern of what philosophers and psychologists call “motivated reasoning”, which can be defined as confirmation bias turned up to 11.
You all know what confirmation bias is, I’d imagine: our predisposition to take evidence that confirms what we believe seriously, while discounting contrary evidence. What motivated reasoning adds is a substructure or foundation to this, in which the agent develops background rationalisations to justify holding the beliefs that others argue are false, or at least not settled conclusively at this point.
The motivated reasoner might see conspiracy instead of disagreement, and tends to react defensively to contrary evidence, seeing conspiracy, or dismissing it out of hand for other reasons instead of responding to it on its merits.
“I ignore what I consider not to be evidence” – Noakes.
The goal of my talk today is quite simple. I want to suggest to you that regardless of any debate on the virtues of the Banting diet – which I’m not interested in discussing, and haven’t expressed any public view on – there’s a language, method, and character that we should all value in scientists and scientific enquiry.
I believe that any of us who work in fields including science, education, or journalism have a responsibility to encourage a responsible epistemic approach, rather than to aim for persuasion above all else.
By this I mean an approach that is objective as regards the evidence, where we are willing to be wrong, and where we resist attacking the character or motives of opponents when arguments are the relevant issue.
Prof. Noakes has frequently set a bad example in these regards, and my concern as an educator – particularly one active in the field of critical reasoning – is that with 50 000 Twitter followers, and as an engaging and hard-working media personality who has garnered as many accolades as just about anyone you can think of in South African science, he has a powerful influence on how people perceive scientific activity.
One of the virtuous traits I mentioned a moment ago was a willingness to be wrong. Defenders of Noakes might immediately retort that of course he’s willing to be wrong – after all, he famously changed his mind on carbohydrates! And while this is a notable change of mind, it doesn’t (and shouldn’t) reassure anyone that it’s representative of a general disposition. As you’d know, it’s a single data point, and we don’t find a pattern in a single data point.
More to the point, perhaps, is that his own language regarding that change of mind comes with significant warning flags regarding his commitment to sound reasoning. Take this example:
At a public discussion with his erstwhile supervisor, Lionel Opie, in May 2014, Noakes told the audience “I’ve said one thing, and now I’m saying the exact opposite. And they can’t both be wrong. And that’s key.”
While a statement like that plays well to a crowd – and in this case, did result in some appreciative chuckles – it’s nonsense as far as logic is concerned.
He’s presenting his change of mind as evidence that he’s right now. And there are two immediate problems here – one is that the argument only gains traction if you agree with him that he was wrong before. If you think he was right then, then you’d think that he’s just strayed into error now.
The second way in which the logic is fundamentally flawed is that he’s suggesting that the audience embrace the logical fallacy of the false dilemma. What I mean is that the truth might actually be undiscovered, and exist somewhere in the middle – it’s not the case that one of the two extremes he’s suggested at various points have to be correct. We have other options, and he’s misrepresenting the choices available to us in leaving them out.
We should all – maybe A-rated scientists, teachers, and public figures in particular – have a concern for good scientific thinking and clear reasoning in expressing the conclusions we’d like to see adopted. Science does not work in absolute truths – it’s an inductive process, whereby we chisel away at falsehoods to arrive at a clearer understanding of what’s most likely to be true. The example above recommends absolutism, despite that being manifestly incompatible with the messy world of empirical data.
At various points in today’s talk, I’ll be showing you examples of statements like these, made by Prof Noakes on public media. There are many more such examples, but it would be tedious to belabor the point through repetition.
But I mention this to offset concerns you might have that I’m indulging in cherry-picking here – the pattern is unmistakeable in itself, and more to the point, many of the examples I’ll show you are examples of Noakes responding to critics accusing him of over-simplifying. In other words, even after applying the principle of charity and seeking clarification from him, his responses validate the concerns I’ll be highlighting.
A taxonomy of trouble
For ease of reference, I’ve loosely categorized the issues into 5 groups:
- Indiscriminate use of sources
- Ad hominem & double-standards
- Exaggeration and immunity to error
- Salesmanship over science
- Errors in scientific reasoning
INDISCRIMINATE USE OF SOURCES
Chemicals that may cause diabetes – Noakes links to Natural News.
This is the site that argues that Microsoft are developing eugenics vaccines. And that HIV doesn’t cause Aids. And Icke? The world is controlled by reptiles from outer space, who live in underground tunnels and take on human form (Thatcher, Bush)
“Next they say it has been discovered before. Lastly they say they always believed it. Louis Agassiz.” – Noakes links to evolution-denier.
This from a series of tweets explaining how his conclusions will be vindicated in the end. But if you’re going to suggest that there are better and worse ways of proceeding in science, perhaps better examples than Agassiz could be chosen. He is, after all, an evolution-denier, and a proponent of scientific racism – does one want to cite him as an authority on the scientific method?
“@natachab Weston Price http://amzn.to/15c8xRz believed superior nutrition of trad societies protected against TB. Need LCHF HIV/TB trial” – and to a “holistic dentist”.
Weston Price – “holistic dentist” whose treatments included homeopathy. The site carries numerous articles arguing that vaccines cause autism. Current board member Joseph Mercola has received at least three FDA warnings for making misleading and/or unsubstantiated claims regarding the products he sells.
AD HOMINEM & DOUBLE STANDARDS
On Anthony Dalby, a more recent critic: “Noakes said the doctor who said that happened to weigh 120kg.”
Prescription? A mirror.
OVERSTATING THE CASE/IMMUNITY TO ERROR
Franschhoek: “If you’re insulin resistant, you do not have to get any disease whatsoever. If you eat a high-fat diet all your life, you will not develop diabetes, you will not get cancer, you will not get dementia. That I can guarantee you” – that’s where, and when (audio).
“Sweden becomes first nation to reject high fat dogma!” Noakes was making this claim in 2013, when the SBU report wasn’t even available in English, RT’ing AuthorityNutrition & Diet Doctor.
The authors of the (independent) body eventually had to step in and tell people to stop saying this. Report just on obesity, not necessarily generalisable, and:
Two mis-interpretations have, in our opinion occurred in the wake of the publication of the SBU report. One is that low carb high fat is by far better. Yes, during the first 6 months you lose weight faster on low carbohydrate diets. But after one and two years that diet has no advantage than other diets for obesity.
After having this pointed out to her, Teicholz blocks me and others. I couldn’t point it out to Noakes, because contrary voices also get blocked from being heard on his timeline. Contrary to what I think the ideal approach – of seeking out ways in which you could improve your arguments – motivated reasoning can involve simply shutting out dissent.
Then, addiction, where I fear that the LCHF movement is doing great harm to public understanding of the difference between compulsive and destructive behavior, and lifestyle choices which can be better or worse but are not intrinsically problematic.
From Real Meal Revolution:
The final blow to the gut: because carbohydrates are nutrient-deficient and often packaged with salt and sugar, you feel the need to eat more of them, thereby putting yourself into a near-perpetual cycle of weight gain.
Unless, of course, you break the addiction…
There’s plenty of “addiction” talk on his Twitter feed, as well as a partnership with Harmony Clinic in Hout Bay, that offers in and outpatient treatment for sugar addiction. Well, Harmony Clinic now liquidated, so perhaps not anymore.
The problem is – there’s no compelling evidence for sugar addiction, yet, and the case is being overstated in the service of promoting Banting.
What should we then say about so-called “addictive” foodstuffs? The first thing to remember is the point Paracelsus made in the 15th century – “the dose makes the poison”.
While there might be no safe number of cigarettes to smoke, there will be a dosage of carbohydrates, or sugar, that’s unproblematic in all but the most rare of cases.
Let’s look more closely at sugar addiction, and addiction in general. Two papers are typically cited as evidence for sugar being addictive. But what they mostly reveal is that science journalists no longer read or understand the journals, and that the public – and some professionals – are far too trusting when it comes to the sensational headlines that convey elements of those studies to us.
First, the Avena study, published in Neuroscience & Biobehavioral Reviews in 2007:
“we suggest that sugar, as common as it is, nonetheless meets the criteria for a substance of abuse and may be “addictive” for some individuals when consumed in a “binge-like” manner.”
Pause there – who might be inclined to consume in a “binge-like” fashion? Perhaps someone with a pre-existing impulse control disorder, who happens to latch on to sugar – the reverse inference from the bingeing to the sugar might get the causal direction entirely back-to-front. We’ll get back to the neurochemistry later, but also, notice the scare-quotes – the author is hedging her bets, with the text only weakly supportive of any claim to sugar addiction.
One is perhaps reminded of a line from Lewis Carrol’s “Through the looking glass”, where Humpty Dumpty said: “When I use a word, it means just what I choose it to mean—neither more nor less.”
Then, there’s Johnson & Kenny’s paper in Nature Neuroscience (2010) on junk food and addiction, also conducted on rats.
“Notably, it is unclear whether deficits in rewards processing are constitutive and precede obesity, or whether excessive consumption of palatable food can drive reward dysfunction and thereby contribute to diet-induced obesity.”
As in the Avena study, we don’t know whether an impulse control disorder is simply being expressed – rather than discovered as an effect, resulting from the junk food – in this experiment.
Yes, if you grow to like something (or find it rewarding), you’ll seek it out. This does not mean the thing is innately addictive. In fact, Hebebrand’s recently published paper in Neuroscience & Biobehavioral Reviews concludes that if anything, “eating addiction” rather than “food addiction” best captures what’s going on when people compulsively over-eat. The food is an expression, not a cause of the impulse control disorder.
We can easily engage in a battle of link-sharing – but the point is that the truth is complex, and not settled by individual studies. In this case, there’s one clear issue with using these studies as definitive, and this is that both of these studies use brain imaging to support their conclusions.
As Sally Satel (who works as a psychiatrist in a methadone clinic) puts it, brain scanning is “a perfect storm of seduction”. The technology promises great revelations and great objectivity. More to the point of my presentation today, it offers the possibility of eliminating your responsibility for what’s wrong with you – we can say, “it wasn’t me, it was my brain!”
This image is interesting is that it neatly summarises why you can’t reach firm conclusions from fMRI data. This fish is in fact dead, yet the scanner showed signs of brain activity.
fMRI data are suggestive, and weakly so at that, in that they reflect neural correlates of various stimuli, and nothing of the perceived and subjective mental responses to those stimuli.
In slightly more detail: Increased blood flow and a boost in oxygen are treated as proxies for increased activation of neurons, and from there we induce to what those neurons are doing. We compare that data to a baseline, and subtract the one from the other, averaging out over the many data points of all participants in a study, with software filtering out background noise, and creating the seductive images.
But our experimental conditions are imperfect – think of the difficulties of creating appropriate baseline tests, for one – and large sample sizes cost a lot of money. Add to that the fact that our brains can process the same stimuli in different regions – no one specific area can reliably be said to perform the same task for all of us – and it should be clear that it’s far too soon to reach definitive conclusions from fMRI data.
The philosophical problem is one of reverse inference – we reason backward from neural activation to subjective experience. But if identified brain structures rarely perform single tasks, one-to-one mapping between activation in a region and a mental state is very speculative.
To avoid the false positive of the fish brain activity above, we need to use multiple comparison fMRI, which comes at far greater expense in terms of cost and time. But headlines don’t have space for subtleties, and furthermore, novel and exciting claims get the public’s attention. If your fMRI scans can be said to show that sugar is more addictive than cocaine, you’re guaranteed some prime media attention, and who can blame you for trying to capitalize on that? Well, perhaps nobody can blame you if you’re trying to sell newspapers. And perhaps we can blame you, or be rather concerned, if you’re presenting yourself as a responsible scientist.
We can’t tell – yet – whether fMRI scans indicate an impulse that is irresistible, or one that simply hasn’t been resisted. But it’s easier to make choices when you believe that there’s a choice to make, rather than a forced one, such that an “addiction” narrative might support. Diminished expectations of agency lead to diminished agency – if you’re not aware of your choices, it’s more difficult to make choices. So, it’s politically useful to say that carbs are addictive – but that isn’t equivalent to it being true.
SALESMANSHIP OVER SCIENCE & SENSITIVITY
“Robin Williams was a vegan.Vegetarians are twice as likely to suffer mental illness.Humans are designed to eat meat.” @ProfTimNoakes RT’ed, the day of Williams’ suicide.
Response when challenged:
“@theviscountess @HermanBeukes Was Mr Williams informed about association between meat-free diet and mental health? http://1.usa.gov/1t2janN”
Then when people point out how poor that paper is, he appeals to the association, even though when anyone cites associational data in favour of low fat, he rejects it.
“Dr Bill Wilson wonders if the Carbohydrate Associated Reversible Brain (CARB) disorder played a role in Newtown tragedy http://bit.ly/URArgp”… Where a 20-year old fatally shot 20 kids, and gun control might be a more interesting conversation.
ERRORS IN SCIENTIFIC REASONING
What does the future hold for a pastry chef?
“@PastryKeegan The public will decide. In era of social media, public will eventually discover what works for each, independent of “experts”
Why are “experts” being trivialized here? Experts do exist, and the public are often misinformed. If experts disagree with you, then you defeat them in the battleground of expertise – peer reviewed journals. An army of laypeople doesn’t make the scientific case.
Why is journo Gary Taubes pushing for scientific studies into #LCHF diets (http://blogs.scientificamerican.com/cross-check/2014/10/07/gary-why-we-get-fat-taubes-speaks-out-on-diet-studies-including-his-own/ …) while @ProfTimNoakes focuses on populism?
— Tim Noakes (@ProfTimNoakes) October 9, 2014
“.@sarahemilywild Because change will only come when public understands truth expressed in @garytaubes book. New science will not change that”
“New science” will not change the truth? And again, a scientist calling for less science?
— Tim Noakes (@ProfTimNoakes) October 9, 2014
September 2014, Australia
So, I watched it. Here’s one rather interesting bit (the text below is copied from a previous post):
And you must never trust consensus guidelines, because they are anti-science. Science is not about consensus, it’s about disproof, disbelief and skepticism. It’s not about consensus. When you’ve got consensus, you’ve got trouble.
This conflates two very different stories into one, to serve the rhetorical purpose of granting credence to the underdog-story. The two stories are first, that yes, dogma is antithetical to science. The second is that if a preponderance of evidence points in a consistent direction, consensus guidelines could be well-justified, and it would only be irrational or inattentive people who would not believe in that consensus.
In the second story, you’d have been rational to believe in the consensus account even if it later turns out to be false. The point is that denying a well-justified consensus doesn’t make you a better scientist – it makes you a conspiracy theorist, or simply wrong about the facts.
In other words, consensus guidelines that emerge out of honest engagement with the evidence, and that are open to correction, are not anti-science at all. They are the product of good science, and their later overturning (if that happens) in favour of a new consensus is also the product of good science.
You don’t measure or identify good science from its conclusions – because we don’t know that those will survive future data – but by method, and by openness to correction in light of evidence. The first kind of story mentioned above, involving dogma, is of course an example of bad science. That doesn’t mean that consensus is by definition bad.
Science is indeed about “disproof, disbelief and skepticism” – but all of these serve to challenge any existing view and replace it with a better one. They are tools, or methods, for reaching a better consensus, not for rejecting consensus in general.
The simplest way of putting the point is this: Noakes would like it to be the case that medical practitioners and educational programmes see the light, and teach the same message he professes. In other words, he’d like his own views to be the basis of a new consensus, because he believes that the existing consensus is wrong.
When you’ve got dogma, you’ve got trouble. And when you’ve got consensus, you might have dogma. But you might also have a bunch of responsible people agreeing that yes, that’s what the data imply, and until we learn something to overturn our view, the evidence leads us – as rational, responsible scientists – to a certain consensus.
In short, while the quote above can play as a sexy soundbite for undercutting received wisdom, it’s another instance of Noakes playing scorched earth with understanding of the scientific method.
He might say that the public health concerns are too significant to care about niceties like the ones I’ve been talking about today. To that, there are two immediate responses. First, sloppy thinking should arguably never be encouraged. For someone who is regarded as an inspiration by many budding scientists, and who is one of South Africa’s most decorated scientists, one might even argue that he has a moral obligation to encourage sound scientific thinking.
The second response is that even if this were true – that we should misrepresent the strength and consensus behind a certain dietary position, in order to save lives – we should be able to debate how far we’re allowed to take the misrepresentation.
But this would require agreeing that misrepresentation occurs, and Noakes insists that it’s others who are being obtuse rather than himself. More disturbing, perhaps, is that those who do disagree are ignored or blocked, or characterized as shills, or victims of “groupthink”.
In conclusion, two tweets that show the best and worst of Prof Noakes:
Yes, it certainly is, and I wish he’d take his own counsel on these matters.
This is the worst, obviously, in that the consequences of him being wrong could be far more acute. The logic here is entirely circular, in that his conception of being “wrong” simply ignores all the harms that competing views say could result from a high-fat diet.
He’s assuming he’s right, even while speculating about the consequences if he were wrong.
Even Noakes’ supporters should expect more of him, for two reasons:
- The scientific method deserves better.
- If he’s right, he’s impairing the credibility his viewpoint garners. The same dietary advice could be given without the “aid” of examples like these, and that might well get the revolution taken seriously in a far more widespread fashion.
Also published on Medium.