Solving scurvy through deus ex machina: How a scientific theory is born
Searching for serendipity
Now also available as an audio post!
Last time, I posed a question: how does scientific progress work, if most humans are so eager to come to beliefs they already hold?
Today, I’ll jump straight to my best attempt at an answer, which comes from the winding history of scurvy and its cures. I like this example because it is obscure enough that people don’t have a preconceived idea of how it went or what it means, and it’s such an undeniable success story that it’s clear something went right there.
On sources:
I first learned of this whole story through the tale of Scott and Scurvy, written by Maciej Cegłowski. His article is a well-written narrative culminating in the demise of Robert Falcon Scott in 1912, but that turns out to be the Darkest Hour of the larger scurvy history, so it makes for a misrepresentatively pessimistic ending.
Cegłowski doesn’t provide many citations. I found a very helpful book in The history of scurvy and vitamin C by Kenneth Carpenter. This book covers all of Scott and Scurvy, provides boundless citations to historical documents spanning five centuries, was written with obvious interest, and is readable without a chemistry or history background — I found the account of how the structure of vitamin C was discovered interesting from both chemical and historical perspectives, and presented clearly to a non-chemist, even showing all the calculations in interpreting the experiments. The final chapter is particularly useful, since it reviews all the historical data from the modern, “correct” perspective.
The hook of the tale is that though the correct theory of how scurvy works was known by the early 19th century, it was later discredited in favor of a theory that fit new evidence better, and this new and wrong theory contributed to the disastrous end of the expedition of Robert Falcon Scott to the South Pole as late as 1912. But two dei ex machina intervened and brought us to the current world where scurvy is so rare most doctors never see a case in their life.
A quick(-ish) telling of the story as it appeared to people at the time
Scurvy was first known as a sailor’s sickness, even though it also sporadically struck on land. For centuries, it was practically the sailor’s sickness, often more destructive than enemy action, pirates, shipwrecks, or any other of the numerous dangers of the sea. It starts after several weeks at sea with gums growing over the teeth and crippling weakness, then the arms and legs swell, old wounds open, and the entire organism just falls apart. Untreated, the disease has a 100% mortality, but it can also go away incredibly fast and completely, often with a landing and some fresh food.
Speculation about the causes of this disease and ways to cure it were plentiful; common suspects were bad air (thought to come from wherever the expedition was voyaging to), idleness (probably a case of mistaking cause and effect), unclean conditions; it was sometimes thought to be contagious. Often some local plant was known as a miracle cure (aneda, watercress, multiple plants called scurvy-grass... all of which do indeed work).
One repeatedly discovered cure were citrus fruits, especially lemons and oranges. They helped already Vasco da Gama in southeast Africa in 1498, French explorers in the East Indies in the 1600s (as recorded by François Pyrard), English settlers in Massachusetts in 1628...
The most famous rediscoverer of lemons as a scurvy cure was James Lind, a Scottish doctor who ran a small controlled experiment in 1747, and who is still sometimes credited with discovering the scurvy cure.1 He had six pairs of sailors try different curatives: apple cider, sulfuric acid (yum!), vinegar, seawater, citrus (two oranges and a lemon per day), and a herbal medicinal paste. The citrus group was the only one cured.
Lind then wrote a 400-page treatise on scurvy reviewing older work, criticising its overreliance on theory (and the ancients), and then... providing his own convoluted theory (comments in brackets): Scurvy is caused by the cold wet air at sea clogging pores in skin — pores through which some kind of waste product is normally excreted,2 but now it accumulates and causes the symptoms of scurvy. Moist air also cannot assist the lungs in converting crude chyle (lymph) into blood (‽), so food cannot be turned into blood (‽).3 Acids can break down some of this waste (here Lind contradicts his results where sulfuric acid was useless, so he has to invent another effect to correct for that), and vegetables combine with chyle to produce soap that helps break the waste particles into smaller pieces that can pass through partially blocked pores (‽‽). This, by the way, is why eating too much citrus causes diarrhea — the body humors melt down.
Based on this theory, Lind suggested possible cures:
Ventilate and dry the air.
Failing that, in damp seasons, put a red-hot loggerhead in a bucket of tar to produce wholesome antiseptic vapour.
Go well clothed and with dry linen.
Eat a bit of raw onion or garlic before being exposed to rain or spray (to promote perspiration).
Move around to support the digestion.
Improve your diet:
Salt is not in itself either helpful or harmful.
Salted meat might be bad (“rendered improper to afford that soft, mild nourishment, which is required to repair the body”).
However, it could be “corrected by bread, vinegar or vegetables.”
Flour well leavened and baked into fresh bread is an excellent anti-scorbutic.
Wait, it’s actually fine as it is:
“Such hard dry food as a ship’s provisions, or the sea-diet, is extremely wholesome; and no better nourishment could be well contrived for people, using proper exercise in a dry pure air.”
However, for those at sea, fresh vegetables and greens “correct the quality of such hard and dry food as they are obliged to make use of.”
His Treatise just goes on and on like that with more and more haphazard details. Eventually he recommends, for long voyages, the use of lemon juice. To stop it from spoiling, he suggests slowly evaporating it down to a syrupy concentrate.
After writing his Treatise, Lind moved on to other duties, but occasionally he would try to return to his original experiment, without success. He did not find any rot in the blood of scurvy patients, his proposed concentrate turned out to be less curative than fresh citrus fruit, scurvy broke out in armies at land during summer (with no cold wet air), and in another experiment (in Haslar naval hospital) he again divided patients into groups with different cures, but all of them got better; in another, nobody got better. In 1772, he would republish his Treatise in an updated version, still claiming that his concentrate “preserves the virtues [of lemons]“, but also including the results showing it doesn’t work as a cure — altogether making it hard to draw any conclusions from the Treatise.
So as you can see, this wasn’t really a step forward. I’m dissecting it here in gruesome detail for a single reason: I must convey the horrible mess this field was in to prepare a foil for what comes next.4
In 1780, there was a scurvy outbreak at the siege of Gibraltar that was cured by a cargo of fresh lemons. This was witnessed by Gilbert Blane, another Scottish doctor, who had a more practical approach. Blane collected tables of information on the fleet under his control and kept pushing for a single reform — using citrus juice preserved with brandy — with all the energy and persuasion and political finesse he had. At first he was ignored, because experiments with Lind’s lemon concentrate had failed and the Admiralty didn’t think fresh fruit would be different. Eventually (in 1793) he convinced an admiral who was a personal friend and had success.5 In 1795, he was appointed to the Board of the Sick and Wounded Sailors and used this prior success to turn lemon juice into a regularly issued supply for the entire British Royal Navy, which caused an instant dramatic decline in scurvy cases.
This was a big deal. By Blane’s statistics, in 1781 the deaths in a fleet of 12,000 men had amounted to nearly 1,600, of whom only 60 were killed by the enemy, the rest by disease. The regular issue of lemon juice cut the total disease burden from all causes in half, and it became possible to stay at sea for long periods of time and maintain naval blockades. This happened at the best possible moment for Britain, since the Napoleonic wars were about to start, and it is largely thanks to the prevention of scurvy that Napoleon got beaten at sea. (But more about that in a future post…)
On the topic of why the lemon cure worked, Blane admitted ignorance (emphasis mine):
For the cure of the scurvy, lemons and oranges are of much the greatest efficacy. They are real specifics in that disease, if any thing deserves that name. This was first ascertained and set in a clear light by Dr. Lind. Upon what principle their superior efficacy depends, and in what manner they produce their effect, I am at a loss to determine, never having been able to satisfy my mind with any theory concerning the nature and cure of this disease, nor hardly indeed of any other.
— Gilbert Blane, Observations on the Diseases of Seamen (1789)
By 1831, scurvy was a medical rarity associated only with prisons and famines, and a more-or-less correct theory of how it worked was presented in a regular medical lecture for students: Scurvy is caused by a deficiency of some substance, an antiscorbutic, which the human body gets from food, and which is present in most fresh vegetables and highly concentrated in citrus fruit.
But the story does not end here. As the years went by, the Royal Navy modernised from sailing to steam power and ships became faster. Scurvy takes several weeks to develop, so you could now safely cross the ocean even without a cure. In the 1860s, lemon juice was swapped for lime juice sourced from British colonies in Montserrat, which was considered a more reliable source than foreign lemon merchants.
Then scurvy appeared on polar expeditions. One, led by George Nares in 1875, tried to sail along the west coast of Greenland, then sledge north when the ice set in, but had to turn around when there was barely anybody left able to pull the sledges. Four people died before a relief party from the ship reached them.
_(14765552602).jpg){kind=link}
This embarrassing failure started several years of heated and detailed debate. Just to give an idea of the scale: It began with a full parliamentary inquiry dedicated to finding the cause of the disaster. The resulting report (issued two years later in 1877) is a fascinating document to skim through, also to get an idea of the scale of a polar expedition — they documented everything.6 Its 571 pages list 9,387 questions asked of 50 witnesses and 34 appendices of supplementary documents, all publicly available. This is dedication to finding the truth on a level rarely seen and everybody involved has my respect, even if all they could conclude was that the sledging expedition did not carry lime juice (because it would be difficult to thaw), that this was a mistake, and that Captain Nares was to blame.
In the wider debate it quickly became clear that the subject was not settled. Other captains protested that lime juice was not reliable as a cure in their experience, and evidence from Eskimo diets showed that it is possible to live for years on a purely-meat diet without developing scurvy.
The common experience on later Arctic expeditions was that eating fresh meat tended to prevent scurvy, while lime juice was ineffective. Scurvy was associated with eating preserved meat, dried or salted.
From the sum effect of all this, a new theory developed where scurvy was actually caused by poisoning by “ptomaine”, some kind of bacterial toxin from badly preserved meat. This was a move from missing antiscorbutics to poisonous ptomaine, neither of which substances anyone had ever seen directly.
Neither of those theories worked particularly reliably, but when Robert Scott was preparing for his South Pole expedition of 1912, he used the ptomaine theory and went to great pains to ensure none of his meat was spoiled. Scurvy soon appeared on a preliminary expedition, and though he tried various modifications of the rations on subsequent expeditions (e.g. replacing all canned meat with freshly hunted seals — boiled to remove moisture and save weight), he never found a reliable solution. He would die on the return trip from the Pole from the generally horrible conditions, among which there likely was also scurvy.
At this point the field was again an utter mess with no reliable cure and conflicting theories of how the disease worked.
In 1907 came the first deus ex machina of this story. Two Norwegians, Axel Holst and Theodor Frølich, were studying the disease of beriberi that had recently started to affect Norwegian sailors. This outbreak seemed connected to a change in rationing, so they looked for effects of diet and tested various diets on guinea pigs. On top of beriberi, they found that some of the guinea pigs had died showing symptoms very much like scurvy. Nobody had done successful animal experiments with scurvy before, because scurvy seemed to be a disease unique to humans — other animal models such as dogs or rats can’t get it.7 Once this simple animal model was available, the pair quickly established scurvy as a deficiency disease and characterised different foods for their anti-scorbutic strength (by feeding them to guinea pigs and seeing whether they developed scurvy). (Sadly, their work didn’t become widely known in time for Scott to catch on.)
Meanwhile, other controlled animal dietary experiments showed that there were some as-yet uncharacterised vital nutrients out there, to which a Polish chemist, Casimir Funk, gave the name vitamines. This was based on his wrong assumption that they were chemically all amines, so it was later shortened to vitamins to prevent confusion.
Around WWI, a research group led by Harriette Chick continued the work of Holst and Frølich, and in 1918, they discovered that lime juice is a less powerful antiscorbutic than lemon juice. They also found that representative samples provided by the Navy often had no curative power at all. Also in 1918, Alice Henderson Smith found the lemon-lime switch in naval archives and matched it to the reappearance of scurvy.
At this point the race was on for isolation of the antiscorbutic compound (now named Vitamin C), but the race would still be on ten years later, since every chemical step (starting from lemon juice and hopefully ending with something crystallic and pure) meant waiting two months for a guinea pig assay to show whether the antiscorbutic compound is still there.
In 1928 came the second deus ex machina. Albert Szent-Györgyi was trying to find the biochemical pathways of respiration and noticed a strong reducing (hydrogen-donating) agent in the adrenal cortex. Finding this “reducing factor” was a chemistry problem without the need for two-month waiting times. In 1928, it was isolated, in 1932 it was found that this chemical is the antiscorbutic factor and its chemical structure was solved. In 1933, a chemical with this structure was synthesised from scratch in Switzerland and verified to also work as a scurvy cure.
This is strange history: lots of effort made seemingly no progress and the correct solution came out of apparently entirely unrelated efforts. Now, we will move back to the present and characterise the change that happened at the end.
Working with and without understanding
The shift here was from a mess to something understood. The problem that killed Scott was that though the wrong theories are full of excuses and long convoluted cause-effect chains, so is the correct theory. People weren’t just missing that “scurvy is caused by a lack of vitamin C”, because on top of this simple fact was a tangle of confounding factors that had to get untangled. An incomplete list:
There’s more than one nutrient deficiency disease (there are 13 vitamins and then there’s trace elements etc.) and their symptoms sometimes occur together, because most bad diets lack more than one nutrient. This meant that different populations with bad diets presented combinations of symptoms that partially, but not entirely, overlapped. This created the problem of whether a given set of symptoms is one disease or multiple.8
The name “scurvy” was sometimes used for multiple different diseases (e.g. “wet scurvy” for wet beriberi, a variation of B1 deficiency).
Scurvy in infants fed on pasteurised milk (who, being toothless, would
not have the characteristic inflamed gums seen in adults) was first known as “Barlow’s disease” with its own parallel debates over causes and cures.
The absorption of any nutrient can be affected by other nutrients and the general state of the organism. In particular, different people take different amounts of time to develop scurvy depending on their diet before going to sea.
There are multiple forms of vitamin C that animals can interconvert, which was a problem in matching the antiscorbutic factor of animal experiments and reducing factor of chemical experiments.
Vitamins differ in their properties:
Some vitamins (e.g. A) are toxic at high quantities while others (e.g. C) aren’t, adding to the confusion between poisoning and deficiency theories.
Vitamin D (whose lack causes rickets) also needs sunlight for conversion into an active form.
Humans can’t produce vitamin C but most animals can... except for guinea pigs.
Vitamin C is present all over the place in plants and animals, but it’s also reactive and easily destroyed (e.g. by cooking). The details of the process often make a big difference — in particular, the killer problem of the 1860s switch was not so much the limes (which still have some vitamin C in them) but the copper tubing used in the production of their juice, which catalysed the breakdown of vitamin C.
It can even go the other way: letting seeds sprout before processing them can make them antiscorbutic as the plant starts producing its own vitamin C.
This also affected the lemon cure, because it’s not just about the lemon juice. Lemons rot easily and can be preserved in different ways (e.g. Lind’s suggested concentrate, juice kept under oil, combination with brandy, lemons bottled in brine...). Some of those work (brandy) and some don’t (Lind’s boiled concentrate).
Lemon/lime juice was often sourced from foreign merchants who may or may not have adulterated their products. This was part of the reason for the 1860s switch: switching from foreign lemons to colonial limes meant better control over the quality of the juice.
Without access to the actual chemical of vitamin C, everything could only be inferred from indirect effects, which meant you couldn’t isolate the problem.
The cures found before vitamin C was synthesised (such as the use of lemon juice) could be replicated, somewhat, by general principles that help repeatability: try to get the same raw materials and use the same methods of preparation. This worked well enough to beat Napoleon, but it’s also an approach that limits what you can do — you will run into logistical issues when trying to use cooking methods from transatlantic ships on a South Pole expedition — and change is inevitable: lemons become expensive, ships become faster, diets change. For the cure to still work reliably, you must be paranoid about any changes — though most are probably fine, you don’t know which ones aren’t. If you stray, sometimes you get lucky and sometimes you don’t. Sometimes multiple changes combine when each individual one wouldn’t be a problem. Sometimes one change (faster ships) hides the effects of another (limes). In this environment, it’s possible for knowledge to become unreliable and methods to stop working because you import the wrong kind of lemon or prepare it in the wrong way.
But this changed when vitamin C was discovered, isolated, and synthesized. Once it was known how scurvy works, people could actually find that mysterious antiscorbutic in various types of food, study its stability, find the metabolic pathways broken by its lack, and establish a reliable connection between its content in food and scurvy symptoms. This all together constitutes an understanding that goes far beyond the Napoleonic era.
Understanding means “ability to predict whether scurvy could be a problem in any future new thing we might want to do, like building a base on Mars, without having to try it out first”. Even if someday we find somebody eating plenty of vitamin C and still suffering from scurvy symptoms, our understanding of scurvy gives us lots of immediate “debugging” tools, like tracking whether vitamin C is actually absorbed from the food, whether its metabolic byproducts are present in the body, etc. The best Scott had was threatening the cook and thoroughly cleaning his ship:
Of course there is no good blinking our eyes to the fact that this is neither more nor less than scurvy, but whence it has come, or why it has come with all the precautions that have been taken, is beyond our ability to explain. The evil having come, the great thing now is to banish it. In my absence, Armitage, in consultation with the doctors, has already taken steps to remedy matters by serving out fresh meat regularly9 and by increasing the allowance of bottled fruits, and he has done an even greater service by taking the cook in hand. I don’t know whether he threatened to hang him at the yardarm or used more persuasive measures, but, whatever it was, there is a marked improvement in the cooking.
(...)
The determination to have everything above suspicion, and not to give our dread enemy another chance to break out, has kept all hands pretty busy of late. With the idea of giving everyone on the mess-deck a change of air in turn, we have built up a space in the main hut by packing cases around the stove. In this space each mess are to live for a week; they have breakfast and dinner on board, but are allowed to cook their supper in the hut. The present occupants enjoy this sort of picnic-life immensely.
We have had a thorough clearance of the holds, disinfected the bilges, whitewashed the sides, and generally made them sweet and clean. As a next step I tackled the clothes and hammocks. One knows how easily garments collect, and especially under such conditions as ours; however, they have all been cleared out now, except those actually in use. The hammocks and bedding I found quite dry and comfortable, but we have had them all thoroughly aired. We have cleared all the deck-lights so as to get more daylight below, and we have scrubbed the decks and cleaned out all the holes and corners until everything is as clean as a new pin. I am bound to confess there was no very radical change in all this; we found very little dirt, and our outbreak cannot possibly have come from insanitary conditions of living; our men are far too much alive to their own comfort for that. But now we do everything for the safe side, and from the conviction that one cannot be too careful.
— Captain Scott, The Voyage of the Discovery (1901–1904, published 1905), through Cegłowski
Understanding also lets others build on that knowledge in other, independent research. For example, humans need vitamin C because they can’t make the enzyme necessary for its production, which is because the relevant gene became dysfunctional for them. In 2013, the Neanderthal genome was sequenced, so it should now be clear whether the Neanderthals could have gotten scurvy, which can be checked against other archeological findings.
Now, let’s look in more detail at the two dei ex machina that created this understanding. How did that happen?
Repeatability: connections, cores, and variables
The actual criterion for science to be useful, for a scurvy cure to work, is repeatability. There has to be enough specific guidance on what to do to be able to apply the cure in varying circumstances with predictable results. That requires separating the relevant from the irrelevant.
This is hard, because what’s relevant isn’t always obviously connected. Lind connecting his theory to scurvy to a theory of diarrhea was wrong, but the correct theory of scurvy involves copper tubing and genetics, which are also not obviously connected to swollen gums and weakness.
Any established scientific theory roughly divides the world into three parts: a core universal phenomenon of interest (here: a deficiency disease / vitamins), important variables differing from situation to situation (here: all the details listed above, e.g. the reactivity of vitamins), and everything else that is irrelevant (here: e.g. the cleanliness of the lodgings). The core is that which unifies otherwise different situations, the variable is that which separates similar situations. For example: Scurvy connects common sailors and babies of rich families, while copper tubing (and the difference between limes and lemons) separates the Royal Navy of the 1830s and the Royal Navy of the 1870s. Understanding means knowing both the core and variables well, which is best done by first getting a handle on the core in one situation where it’s relatively undisturbed and then using it in other situations to find the variables.
In the first deus ex machina, the core of scurvy being a deficiency disease with an animal model came from the study of beriberi, which was more unambiguously connected to diet. Since there are multiple vitamin deficiencies (a connection), there was a good chance studying one of them would provide little gifts to the study of others.
In the second deus ex machina, the structure of vitamin C came from biochemical pathways — Szent-Györgyi also played a large role in discovering the citric acid cycle that is at the center of energy metabolism.
The core discovery often comes from a different field than where the original problem (or eventual application) lies. But you don’t know ahead of time what the core might be, and you have no guarantee there even is one. Knowledge of this sort often comes not from systematic controlled trials (because you don’t know what to control for ahead of time), but from a new connection between adjacent (but previously thought to be unrelated) fields.
A similar principle in Newtonian mechanics
Now, allow me a quick excursion into a different field where the same dynamics played out.
I’m a physics student, and already in high school I’ve always found explanations of Newton’s laws unsatisfying. In part this might be because mathematically speaking, Newton’s laws say very little: only that acceleration of objects is linked to something called a “force”. But the laws themselves don’t actually specify what forces are, and any motion whatsoever can be explained in this framework by postulating the right forces.10
Newton’s laws only become interesting when coupled with a tacit understanding that forces in the world of various origins exist, can be characterised and understood, and are useful in describing a vast array of phenomena, and this is hard to communicate to somebody just learning Newton’s laws, because it’s a potentially boundless list of examples: Gravity, the Lorentz (electromagnetic) force, contact friction, normal (constraint-enforcing) force, all kinds of complicated drag forces in aerodynamics, spring/deformation forces, radiation pressure, air pressure, muscle tension, piezoelectric motors... (here the different force origins play the role of variables)
In other words, the main takeaway of Newton’s laws is that acceleration is a better concept for describing dynamics than velocity — it’s more universal, easier to explain, and less dependent on the history of particular objects than e.g. Aristotelian physics, which says that velocities, not accelerations, require an attributable origin. But this view is non-obvious and often doesn’t seem to apply in practice: we know that things seem to come to a halt (a preferred velocity state) on their own given enough time (an amount that is highly circumstances-dependent), that cars need fuel just to maintain a constant speed, or that some objects seem to change their velocity instantly (like a puck getting hit by a hockey stick).
Because of the long list of complicated forces, the understanding had to come from a far-away field — celestial mechanics — where it is in particularly clear view because there is only one relevant force (gravity). Once the concept is articulated, all the seeming contradictions can be understood within the same framework: thus people learned of dynamic friction (forces that tend to stop moving objects), contact mechanics (very strong and short-lived forces occurring on impacts), or normal forces (originating from the Pauli exclusion principle but in practice just postulated such that solid objects don’t overlap) and could use Newtonian mechanics on Earth.
A similar principle in math
Very quickly: As noted by Scott Aaronson here, big, long-standing math problems ultimately get cracked not by a direct approach and hard work, but when arbitrary-seeming statements get connected to large frameworks, which are often built for another purpose entirely. His example is that the proof of Fermat’s Last Theorem came shortly after it was connected to the broader research program of elliptic curves, and the final proof came in the form of a one-to-one connection between elliptic curves of number theory and modular forms of complex analysis.
Lessons learned
Serendipity
The language of deus ex machina and chance discoveries might make it look like I’m saying scientific discovery is random and impossible to affect. That would go too far! Our dei ex machina appeared during sensible study of something else. Beriberi was an existing problem of the Norwegian navy, so it already existed in a space potentially occupied by other existing problems. Respiration and reducing agents are part of the huge network of metabolic pathways whose study is important for all biological production of chemicals and a big chunk of drug discovery, so it made sense to study them.
To find new clearly-visible cores, it helps to be doing something new and interesting, even if it doesn’t have a clear application. I’m not the first to make this suggestion: This is why Steven Weinberg encouraged young physicists to go for the messes, for the areas that seem inelegant and finicky, because that’s where creative work (and breakthroughs) could happen.
Richard Feynman made a similar recommendation in one of his lectures, audible in the recording here (click the casette icon in the top right), starting at 66:50. He’s talking about a book on atmospheric electricity and complaining that the field has not yet managed to get their findings to be reliably repeatable:
It’s very difficult to read because this is one of those books where the man is not at all critical, he’s working in a field where nobody knows the answer to anything, so everything is allowed in the book. In a section on “the charge carried by rain on calm days”, it will say “Schmutz has measured the value and found it to be negative so-and-so much per drop”, “Spugglehead measured the value and found it to be...” — different value, different sign...
In the same recording (at 59:40), he recommends a student that it’s a good subject to study, precisely because it’s such a mess and there’s a lot of opportunities to think of something new.
Searching for dei ex machina is also an argument for seeking all kinds of extremes and challenges. In a perverse way, it’s a justification for polar expeditions, and by analogy for manned space exploration. Though the expeditions of Nares or Scott were disasterous, they were still one piece of the puzzle that ultimately brought biochemistry into existence. Cegłowski is against humans in space, but even if there were no astronomical value in that research, this kind of extreme situation can create unexpected problems that move unrelated fields forward. (It is true that this only works if you try to go to the Moon, and then Mars, and then... and not just back and forth to the ISS for decades.)
Read any press release about a scientific proposal today and at the end there will be a list of possible applications. This, I think, can be harmful in that it suggests that all science should be working with applications in sight. Often this is the case in the more systematised/industrial areas (like drug discovery), but there is a point in basic research whose discoveries are unpredictable in advance, and then the questions are rather: What is this doing that hasn’t been done before? Is there a plausible happy path to settle a current controversy? Are there any people who don’t care about the work process (i.e., not fellow researchers in the same field) but who would still find its results interesting?
Doubt
In the 1800s, if you were an admiral, the best you could do really was to note that lemons tend to work and try to record when. Further study wouldn’t have helped you, because the answers were just that much out of reach, and there was no way you could have gotten from “my sailors keep dying” to inventing the entire field of biochemistry in time to help with that problem. Indeed, the 1877 report or Blane did their best, can’t be said to have missed something for lack of effort, and still didn’t come up with the correct theory. In that situation, it’s remarkably hard to declare that ignorance, and tempting to speculate — perhaps with an addendum that you’re obviously not sure, but speculate nonetheless. Those who didn’t do that deserve praise.
I already mentioned Blane admitting he doesn’t know how lemons cure scurvy. There’s no way he would have given the correct answer with all the caveats, so he was wise to remain silent.
Carpenter’s book was written in 1986, when there was a theory in the air that large doses of vitamin C could cure diseases from colds to cancer — promoted heavily by Linus Pauling. Today this is considered disproven, but in 1986 it was not quite settled yet, so Carpenter deserves thanks for declaring his uncertainty on the hypothesis and presenting the evidence of his time in all its contradictoriness.
Isaac Newton notably avoided this overconfidence when he refused to propose any mechanism for the action at a distance of gravity, because he had nothing useful to contribute in that regard. Indeed, if we have an explanation today, it is that Newtonian gravity is a limiting case of Einstein’s field equations. Newton couldn’t have said anything productive about this, so he was wise to remain silent.
Archive
This one is a bit of a reach. As mentioned above, one of the factors that helped the field find its footing once the guinea pig model was found were the naval archives. Thanks to them, it was possible in 1918 to explain the difference between lime/lemon juice working in 1850 and not working in 1875. So more generally, information which does not make sense today should still be documented in as much detail as possible, since future people might find it useful.
From the other side, use all of the archives you can find! In the history of scurvy Lind and Blane are archetypes that reappear: the theoretician and the empiricist. Carpenter uses as his archetypical examples two other 18th century doctors, Boerhaave (who came up with a general theory of disease and attributed scurvy to cold and damp conditions based on his native Holland) and Bachstrom (who overviewed cases where the disease flared up — in sieges, tropics, on the coast and inland... —, discarded all effects of the air and temperature as inconsistent, and found the connection to fresh food). This is an area where I think scientific culture has made unambiguous progress since the time of Lind, and scientific discourse is way more open than it used to be.
The future
Today, barring very bad planning, one is safe from the specific fate of Scott, because spectroscopy provides unified methods to identify most small molecules and we don’t need improvisation as much, since we have satellite internet to communicate at a distance and helicopters to get to remote locations. But there are other areas where we still are flying blind, vulnerable in ways that future humans might one day look back on as horrifying. Those gaps are the purpose driving science.
Of course, in practice one doesn’t necessarily work in a field because of some long narrative that leads from a mundane or fundamental purpose to a scientific project. But most other reasons people have (an interest in something “for its own sake”, money, intellectual challenge, the chance to learn something new every day) are in one way or another held up by these ties to the unknown — that is what makes science interesting (and at times frustrating).
What makes reading about scurvy particularly intense is that today we know the disease has a very specific mechanism, and yet in the contemporary documents it looks as complicated and messy as many diseases (say, Alzheimer’s or depression) look to us. It may be that those diseases really are more complicated (scurvy is, after all, a particularly simple one), but it may be that there is a solution, and it’s as simple as lemons, but picking it out from all the noise requires concepts that would make no sense to us.
To make this more vivid: We know that lithium salts help with manic-depressive disorder, but we do not know the mechanism. For sure, Wikipedia includes a section on various partial results in that respect, but they are filled with maybes and so many alternative possibilities that I can’t imagine drawing any therapeutic guidance from them. Compare for example the mechanism of penicillin, in which the mechanism of every step from entry to cell death is clear, both down to the actual molecular interactions and up to interactions with other chemicals, and alternative mechanisms for different bacteria types are mapped out.
New hydra heads
I already teased a political connection. A question for the curious that hides a story: Why were there lime plantations on Montserrat in the 1860s?
But a more direct question you might have is: People have already said many things about how science works. For example, what would Karl Popper say about all this? Have there been falsifiable theories and predictions here? How come the “correct” theory got falsified and then reappeared? And which of the numerous activities described here are scientific and which aren’t? (And really, we haven’t much answered how people do this, only why it’s possible at all...) In a future post, I’ll overview some existing philosophies/explanations of science from the time after Descartes.
He was motivated by the disastrous expedition of George Anson where over 1,000 men started the journey and only 145 survived it, with scurvy the main culprit behind the disaster. This brought the problem to the Admiralty’s attention and various experiments were allowed in the years to come — Lind’s only was the most famous one.
This was not pulled entirely out of thin air, but ultimately it was a mistake originating in the poor understanding of gases at the time (CO₂ had yet to be described):
[Lind] was impressed (we would say overly impressed) by the quantitative experiments that were supposed to have proved this beyond doubt. They involved a subject weighing his food and drink and also his excreta over a period, and also himself at the beginning and end. If the intake weighed more than the urine and feces, together with any gain in body weight, it was said that the excess was lost by perspiration. If the subject had not visibly sweated, then this loss was entirely “insensible perspiration.” It was probably the quantitative aspect that particularly appealed to Lind, but of course, there was no measure of the carbon dioxide gas and water vapor lost in the air expired from the lungs.
— The history of scurvy and vitamin C, page 58
Part of the problem with scurvy was that theories of nutrition in general were completely wrong and people didn’t even have a good idea of why humans have to eat:
[Lind] did write:
An ounce of powdered salep, and another of portable soup, dissolved in two quarts of boiling water, become a rich thick jelly, capable of receiving any flavour from the addition of spices. This is sufficient sustenance for one man a day; two pounds of each would serve him a month; and being a mixture of both animal and vegetable food, it is more wholesome that either used alone.
The modern reader may feel that there has been some misreading of the text here, because the quantities per day are very small and we can estimate that they would have provided no more than 250 kilocalories, or less than 10% of a man’s needs. But the idea that animal heat and energy were produced solely from the energy of the combustion of food was a much later one. Provided that the soup was mucilaginous or jellylike, it was considered to be raw material for conversion into blood. Even Count Rumford, who came later and had a special interest in both heat and latent heat, used very dilute soups when he designed economical diets for the poor. He apparently believed that water itself was a food, provided that it was thickened:
Our knowledge in regard to the Science of nutrition is still very imperfect, but I believe, that we are upon the eve of some very important discoveries. (...) It is now known that Water is not a simple element, but a compound, and capable of being decomposed. (...) One single spoonful of salope, put into a pint of boiling water, forms the thickest and most nourishing soup that can be taken; (...) The barley in my soup, seems to act much the same part as the salope in this famous restorative; when it is properly managed, it thickens a vast quantity of water; and, as I suppose, prepares it for decomposition.
— The history of scurvy and vitamin C, page 67
And also because after reading about this topic for absurdly many hours, I feel the need to knock Lind off the pedestal he’s put on. I have some sympathy for him — his failures were a product of his time, he tried his best, and he saw his failure — but he was the wrong person for the job. This story has other heroes who deserve more recognition (as we will see in a moment).
This ship made a 23-week journey without a single scurvy case. That this was remarkable goes to show how widespread a problem scurvy was.
Some people (e.g. here, and even Cegłowski and Carpenter say things to this effect) claim that there was confusion in the language between lemons and limes. The terms were used somewhat interchangably, but from some questions it’s clear that the distinction between the fruit types and the suppliers was raised (esp. 5557: May I ask you whether it is lime juice or lemon juice which is really used in the merchant service? and 5570: Can you tell us where the limes and lemons come from that are used?). It was not caught as important until much later, though.
Today we know this is because there is an enzyme necessary for producing Vitamin C, and the gene that encodes this enzyme got lost (became a pseudogene) in humans, monkeys, and guinea pigs.
In this giant response, this is probably the only actually helpful step. Indeed, within the (...) the disease goes away (for the moment) — everything after it is just Scott trying to be sure.
The third law specifies that forces come in pairs, which can be reformulated as the conservation of momentum. That is a mathematical result (i.e., it constrains possible forms of motion, assuming you have a way to measure the mass of objects), but though important, I’d argue that isn’t the main takeaway from Newton’s laws. In particular, it isn’t necessary for most of Newton’s Principia (1687, about two generations after Descartes), which deals with the effects of gravity (where the third law already follows from the formula for the gravitational force), but also air resistance (where the other force acts on air and is irrelevant).