Science & The Problem Of Complexity
Updated: Jul 7, 2019
Controlled clinical trials are among the gold standard of nutrition science. So how do they fare against tradition and experience as a reliable source of knowledge ? Not well. In his classic book ‘The Fat Of The Land’ Vilhjalmur Stefansson tells the story of how a meticulously designed and implemented trial arrived at completely the WRONG conclusion in a study of pemmican as a field ration. This was being assessed over its suitability for US troops in the field during World War II.
Their adversaries, the Wehrmacht, did not conduct any such tests and DID supply their infantry with this item, one factor that contributed to the general superiority in combat effectiveness (1.5:1)* of the German soldier. Physical fitness and stamina often make the difference between life and death on the battlefield.
In other words, US soldiers died as a direct result of their Army’s reliance on science.
Stefansson describes this episode as ‘The Second Pemmican War’, the first involving a contest between the two trading companies operating in Canada during the 19th century. This was a fascinating example of the use of food as a strategic weapon of war. In this case the Hudson Bay company attempted to deal its competitor a knock-out blow by installing farmers on the prairies of North Dakota and Manitoba, in doing so they displaced the bison who were the source of the pemmican their rivals the North West company relied on and from which they gained a clear competitive advantage in the fur trade. This led to a fatal shoot out in 1814 that left 21 men dead. The conflict was later resolved when the two merged to form a single corporation. (pp.181-9)
The ‘Second War’ did not come to physical blows but consisted of, “an internecine strife within the framework of the Second World War,” and which “appears in retrospect to have been primarily a struggle between theory and testimony.” (p.227) (This is the source of the phrase I used to discuss Scott of the Antarctic’s demise here) The combatants in this struggle were the ‘dietitians’ on the side of theory, and officers in the field together with historians who had studied the experience of polar explorers, fur traders, and Native American peoples in the North on the other. Officers within the Air Transport Command in particular, for who energy to weight ratio was critical when determining what foods to transport by air, were adamant that pemmican be adopted for the men under their command. (p.227)
In Stefansson’s account the aggressors in this war were the theorists. It was,
“a revolt of modern dietitians against the traditional dominance of pemmican as an emergency ration and was in defence of the right of laboratory technicians to prefer the results of their experiments to the testimony of experience.” (p.228)
Here was a familiar scene of the ‘experts’ knowing better.
“The technicians felt they knew from its chemistry that pemmican was not a good food (just as today’s chemists know ‘alternative proteins’ are – see here), they had tasted it and did not like the taste. It was their duty as they saw it to protect our soldiers against this unscientific and disagreeable ration.” (p.228)
The scientists were also subject to certain prejudices, including “racism” (p.228) Pemmican was an indigenous food, and “The Red Man of the Americas was obviously far inferior to us.” The ‘truth’ of this was proven by the very fact of conquest, a form of ‘might equals right’ transferred onto the nutrition field. Stefansson goes into this aspect of the question in some detail throughout his book, and makes some very interesting observations about the early encounters between explorers, settlers, and natives across the continent. He describes “that early period when hunger was an ever-ready fillip – a period too when the colonists were, to an extent, receptive to new things. The realised they were in a new world and so had the inclination to taste new fruits and learn if they were good.” (p.229). This receptivity came to an abrupt end, however, once their women arrived, “who, in turn, brought with them the idea that Indian food was not good enough for them. They insisted it was not good enough for their husbands, either, and certainly not for their children, who were not going to grow up like savages if their mothers could help it.” Stefansson goes on to cite a book on the subject by Irene D. Paden, herself a pioneer woman, detailing how the women of the day ensured the continuity of European habits within their families, above all when it came to food. (pp.232-3) This discussion in ‘The Fat Of The Land’ was one inspiration for my post earlier (see here) where I characterise the nutrition space as a feminine one, and highly conservative as a result.
The point of this digression is to show the extent to which the ‘scientific’ knowledge of chemists and technicians actually rests on cultural influences, and has no basis in science whatsoever. These experts ‘knew’ pemmican was bad before they ever set foot in their laboratories. This kind of ‘knowing’ is even more widespread today, and even less understood at a conscious level. In itself there is nothing objectionable to it, we all belong to a culture, and are shaped by it, the contradiction only lies with Modern ways of thinking, which have broken with the traditions that gave birth to them in the self-belief that they are ‘rational’, ‘scientific’. My argument is that this is pure self-delusion, much better that a culture recognises itself, including its ‘prejudices’ and world-view, takes ownership of these, and by doing so is able to recognise that other cultures and world views can exist elsewhere and be perfectly legitimate on their own terms. This is a point I will return to in other posts, it is an ethical outlook and an important counter to our current drift towards ‘liberal’ totalitarianism in the West.
Pemmican put to the test
The US Army’s response to the disagreement over rations was to settle the matter through rigorous science. They did this by reporting on a trial of the food conducted on a platoon of combat hardened Canadian riflemen under the supervision of their scientists. The testing was done under strict conditions so as to give the food the fairest chance possible, for any result to be as objective as it could be. Stefansson describes these in great detail, here is a brief taste,
“No report of an experiment could be more convincingly set up than the one which is about to be reviewed… There is precise information of every conceivable phase of the trials. There are tables of the results and graphs… the platoon which tested pemmican was particularly chosen because of its exceptional showing in physical fitness tests and in morale during previous manoeuvers in the same area.” (this provided the control element for the trial, the unit’s track record) (p.243)
Careful instructions were provided on how to prepare the food and in what quantity to eat. “In order to avoid prejudice the word pemmican was never used in front of the troops.” (this was an effort to make the test as ‘blind’ as possible) (p.243)
The result was a disaster. Pemmican failed miserably as a ration.
“At the end of the (first) day morale was not high, there were complaints about the food, and 11 of the 17 men complained of nausea. ‘The second was a trying day, with the men hauling fully loaded toboggans 13 miles in deep snow over bush and trail. In contrast to their usually cheerful adaptation in previous tests the troops showed very poor morale and felt the cold so keenly that wood fires had to be started at the noon halt. The platoon ate very little. One soldier vomited early in the morning, one was dizzy all day, 4 were nauseated all day, and 12 felt weak and tired. Pathological fatigue was evident…The third day brought the platoon to the point of disintegration as a military unit… At the noon halt the commanding officer decided that to complete the day’s work would be to risk too many casualties, so he turned back for camp… the platoon was completely useless operationally. The line officers of the company agreed that if there had been actual combat… the (platoon) would have all been casualties by the end of the second day” (pp.250-1)
That was the end of pemmican’s involvement with the US Army, a hardly surprising outcome given the way the trial went. “With such a condition facing them, what else could military nutritionists do but find against pemmican as a military ration ?” (p.255)
What we have here, therefore, is the application of a rigorous scientific experiment with an unequivocal result. While it is possible to argue the nuances of the test’s design, it is hard to see how in practical terms it could have been, or indeed needed to be, improved. Today we consider double blind, placebo, randomised controlled trials (RCTs) as the ultimate gold standard, but in the circumstances of this trial it is difficult to see how that would have been superior to what was actually done. This WAS science at its best.
The trouble is – SCIENCE GOT IT WRONG.
Stefansson states the issue in these terms, “We are in a dilemma, then, between two clear cut and diametrically opposed verdicts.” On the one hand the product of scientific inquiry, and on the other the testimony of explorers, traders, sportsmen, and indigenous peoples collected over hundreds of years. The written record of this testimony amounted to several thousand books, all of which were in principle available to Army decision makers had they taken the trouble to consult them. Instead they relied on their own ability to be “meticulous about their field studies.” (p.255)
In other words, the experts knew better.
Except that they didn’t, and their error cost soldiers’ lives.
The limits of science in the face of complexity
The problem here is a methodological one, and it applies equally to RCTs. The strength of these trials consists in their ability to isolate a single variable, and test its effects, positive or negative. This has its place, without question, at times it can be very useful. Its application, however, is extremely limited. This is because it fails in the face of COMPLEXITY. In complex systems causality is non-linear, this is an insight of systems engineering, and also of holistic forms of medicine. This means that the tiniest variation in the state of a system can have an exaggerated impact on any single factor. Pemmican CAN be an extremely effective ration source, however only if a whole number of other conditions are in place, what is needed is to test not only the pemmican in isolation, but in CONTEXT, in a whole number of contexts in fact, far more than is feasibly practical in a clinical environment.
This is the dilemma Stefansson is refering to. This method of conducting science DOES lead to diametrically opposed verdicts, NEITHER of which are of ANY use whatsoever. If the same experiment had been conducted on another unit, perhaps under different leadership, and in this case had been a success for pemmican, this would still have been an EQUALLY FALSE outcome.
Pemmican is not the problem, the method is.
The only way around this problem is to approach it from the point of view of the totality, as a system. This is why experience, and the tradition it contributes to over decades and centuries, trumps science EVERY TIME. We can see this in the contrast between Scott’s expedition and Shackleton’s, discussed previously here. The former failed and the latter succeeded, for a whole number of reasons. One of the differences between the two was that Shackleton’s crew spent the winter lazing about doing as little as possible, while Scott’s men regularly exercised. Does this mean physical fitness is not a required condition of success for polar exploration ? That would be an absurd conclusion, but it follows LOGICALLY from the kind of scientific method we are witnessing in the pemmican example. Isolating a single variable does NOT HELP. Learning from the entirety of Shackleton’s experience, from Amundsen’s, from Nares’, on the other hand, DOES.
In the nutrition space, we face complexity ALL THE TIME. The scientific methods currently in use, even the best of them, have limited application in the face of this. Usually, far too much is asked of them, which is why we get the farce in debates of studies being thrown at one another by both sides, all proving the exact opposite. Every now and again we get lucky, we CAN isolate some variable or other and show direct causation, but this is rare. Sometimes it is worth the effort anyway, my point is simply to recognise the limitations of this kind of approach.
When compared to science, tradition is infinitely superior, because it rests on the totality of experience. Our problem is that we have severed all connection with our traditions, in the name of ‘progress’, science, rationality, the promise of Modernity (see here). This presents us with a dilemma. Certainly we should attempt to reconnect with our past, the ways of our ancestors, not only from hunter gatherer times but also more recent, where specific cultures are involved. In countries like the United States and Australia, however, this is very difficult, as these are not our lands historically. Nevertheless, we should try, and at the same time seek to create NEW TRADITIONS, ways of eating and of living that WORK, that support and embody our values, what matters to us. This is why the n=1 experiments so many are conducting on keto, LCHF, and carnivore DO matter, not as individual efforts but for their potential to generate a new way of living, one our forefathers would recognise and embrace, but also in accord with our own times, the possibilities it presents, as well as the realities it involves.
My purpose here is to explain the task that lies ahead, the challenge before us. I hope this post helps to achieve that, I suspect it will take many more before we really start to get some clarity, and in the meantime of course, a whole body of experience, all the n=1’s, will be accumulating. It is the latter that will be decisive, not some blog, but if I can assist with the process then that will make it well worth the effort.
* Trevor N. Dupuy, A Genius for War: The German Army and the General Staff,1807-1945 (1977). His precise evaluation is not accepted by everyone, but the combat effectiveness of the German soldier in WWII is not really in dispute. Dupuy’s figure is discussed online here https://www.washingtonpost.com/archive/opinions/1985/05/05/their-wehrmacht-was-better-than-our-army/0b2cfe73-68f4-4bc3-a62d-7626f6382dbd/?noredirect=on&utm_term=.9d98baeed027