A case of analogy in scientific discovery
Keith S. Taber
How is a drafted man like a draft horse (beyond them both having been required to give service?)
"The phthisical soldier is to his messmates
Jean-Antoine Villemin quoted by Goetz, 2013
what
the glandered horse is to its yoke fellow"
Analogy in science
I have discussed many examples of analogies in these pages. Often, these are analogies intended to help communicate scientific ideas – to introduce some scientific concept by suggesting it is similar to something already familiar. However, analogy is important in the practice of science itself – not just when teaching about or communicating science to the general public. Scientific discoveries are often made by analogical thinking – perhaps this as-yet-unexplained phenomenon is a bit like that other well-conceptualised phenomenon?
Analogies are more than just similes (simply suggesting that X is like Y; say that the brain is like a telephone exchange 1) because they are based on an explicit structural mapping. That is, there are parallels between relationships within a concept.
So,
- to say that the atom is a tiny solar system would just be a metaphor, and
- to simply state that the atom is like a tiny solar system would be a simile;
- but to say that the atom is like a tiny solar system because both consist of a more massive central body orbited by much less massive bodies would be an analogy. 2
Read about analogies in science
A medical science analogy
Thomas Goetz describes how, in the nineteenth century, Jean-Antoine Villemin suspected that the disease known as phthisis (tuberculosis, 'T.B.') was passed between people, and that this tended to occur when people were living in crowded conditions. Villemin was an army surgeon and the disease was very common among soldiers, even though they tended to be drawn from younger, healthier members of the population. (This phenomenon continued into the twentieth century long after the cause of the infection was understood. 3)
(Image by Daniel Borker from Pixabay)
Villemin knew that a horse disease, glanders, was often found to spread among horses that were yoked closely together to work in teams, and he suspected something similar was occurring among the enlisted men due to their living and working in close quarters.
"…Jean-Antoine Villemin, a French army surgeon…in the 1860s conducted a series of experiments testing whether tuberculosis could be transmitted form one animal to another. Villemin's interest began when he observed how tuberculosis seemed to affect young men who moved to the city, even though they were previously healthy in their rural homes. He compared the effect to how glanders, a horse disease, seemed to spread when a team [of horses] was yoked together. "The phthisical soldier is to his messmates what the glandered horse is to its yoke fellow", Villemin conjectured."
Goetz, 2013, p.104
To a modern reader this seems an unremarkable suggestion, but that would be an ahistorical evaluation. Glanders is an infectious disease, and so is tuberculosis, so being in close contact with an infected cospecific is clearly a risk factor for being infected. Yet, when Villemin was practising medicine it was not accepted that tuberculosis was infectious, and infectious agents such as bacteria and viruses had not been identified.
Before the identification of the bacterium Mycobacterium tuberculosis as the infectious agent, there was no specific test to demarcate tuberculosis from other diseases. This mattered as although T.B. tends to especially affect the pulmonary system, it can cause a wide range of problems for an infected person. Scrofula, causing swollen lymph nodes, was historically seen as quite distinct from consumption, recognised by bloody coughing, but these are now both recognised as the results of Mycobacterium tuberculosis infection (when the bacterium moves from the lungs into the lymphatic system it leads to the symptoms of scrofula). The bacterium can spread through the bloodstream to cause systemic disease. However, a person may be infected with the bacterium for years before becoming ill. Before the advent of 'germ theory', and the ability to identify specific 'germs', the modern account of tuberculosis as a complex condition with diverse symptoms caused by a single infectious agent was not at all obvious.
The contexts of discovery and justification
Although the analogy with glanders was suggestive to Villemin, this was just the formation of a hypothesis: that T.B. could be passed from one person to another via some form of material transfer during close contact. The context of discovery was the recognition of an analogy, but the context of justification needed to be the laboratory.
Sacrifices for medical science
The basic method for testing the hypothesis consisted of taking diseased animals (today we would say infected, but that was not yet accepted), excising diseased material from their bodies, or taking samples of tissue from diseased people, and introducing it into the bodies of healthy animals. If the healthy animals quickly showed signs of disease, when similar controls remained healthy, it seemed likely that the transfer of material from the diseased animal was the cause.
Although the microbes responsible for T.B. and similar diseases had not been found, autopsy showed irregularities in diseased bodies. The immune system acts to localise the infection and contain it within tissue nobules or granuloma known as 'tubercles'. These tubercles are large enough to be detected and recognised post-mortem.
It was therefore possible to harvest diseased material and introduce it into healthy animals:
"If one shaves a narrow area on the ear of a rabbit or at the groin or on the chest under the elbow of a dog, and then creates a subcutaneous wound so small and so shallow that it does not yield the slightest drop of blood, and then one introduces into this wound, such that it cannot escape, a pinhead-sized packet of tuberculous material obtained from a man, a cow or a rabbit that has already been rendered tuberculous; or if, alternatively, one uses a Pravaz [hypodermic] syringe to instil, under the skin of the animal, a few droplets of sputum from a patient with phthisis…"
Villemin, 1868/2015, p.256
Villemin reports that the tiny wound quickly heals, and then the introduced material cannot be felt beneath the site of introduction. However after a few days:
"a slight swelling is observed, accompanied in some cases by redness and warmth, and one observes the progressive development of a local tubercle of a size between that of a hemp seed and that of a cobnut. When they reach a certain volume, these tubercles generally ulcerate. In some cases, there is an inflammatory reaction…"
Villemin, 1868/2015, p.256
Despite these signs, the animals remain in reasonable health – for a while,
"Only after 15, 20 or 30 days does it become evident that they are losing weight, and have lost their appetite, gaiety and vivacity of movement. Some, after going into decline for a certain period, regain some weight. Others gradually weaken, falling into the doldrums, often suffering from debilitating diarrhoea, finally succumbing to their illness in a state of emaciation."
Villemin, 1868/2015, p.256
In the doldrums
The doldrums refers to oceanic waters within about five degrees of the equator where there are often 'lulls' or calms with no substantial winds. Sailing ships relied on winds to make progress, and ships that were in the doldrums might be becalmed for extended periods, and so unable to make progress, leaving crews listless and frustrated – and possibly running out of essential supplies.
"Down dropt the breeze, the sails dropt down, 'Twas sad as sad could be; And we did speak only to break The silence of the sea!
All in a hot and copper sky, The bloody Sun, at noon, Right up above the mast did stand, No bigger than the Moon.
Day after day, day after day, We stuck, nor breath nor motion; As idle as a painted ship Upon a painted ocean.
Water, water, every where, And all the boards did shrink; Water, water, every where, Nor any drop to drink."
Extract from The Rime of the Ancient Mariner, 1834, Samuel Taylor Coleridge
So, the inoculated animals 'fell into the duldrums', metaphorically speaking.
Read about metaphors in science
(Image by Youssef Jheir from Pixabay)
The needs of the many are outweighed by the needs of humans
It was widely considered entirely acceptable to sacrifice the lives and well-being of animals in this way, to generate knowledge that is was hoped might help reduce human suffering. 'Animal rights' had not become a mainstream cause (even if animals had occasionally been subject to legal prosecution and sometimes found guilty in European courts – suggesting they had responsibilities if not rights).
Similar experiments were later carried out by Robert Koch in his own investigations of T.B. and other diseases soon after. Indeed, Goetz notes that when working on anthrax in 1875,
"As Koch's experiments went on, his backyard menagerie began to thin out; his daughter, Getrud, grew concerned that she was losing all her pets."
Goetz, 2013, p.27
p.27
(Image by Adina Voicu from Pixabay )
Although animals are still used in medical research today, there is much more concern about their welfare and researchers are expected to avoid the suffering and death of more animals than considered strictly necessary. 5 Wherever possible, alternatives to animal experimentation are preferred.
Inadmissible analogies?
One of the arguments made against animal studies is that as different species are by definition different in their anatomy and physiology, non-human animals are imperfect models for human disease processes. One argument that Villemin faced was that his inoculations between animals was most successful in rabbits, when, it was claimed, rabbits were widely tubercular in the normal population. In other words, it was suggested that Villemin only found evidence of disease in his inoculated test animals because they probably already had the disease anyway.
That suggests the need for some sort of experimental control, and Villemin reported that
"…despite routine sequestration and the tortures that the vivisectionists force them to endure, rabbits are almost never tuberculous. I have explored more than a hundred lungs from these rodents from markets and I found none to be tuberculous."
Villemin, 1868/2015, p.257
Indirect evidence
Villemin had made an analogy between disease transfer between horses to disease transfer between humans. His experiments did not directly test disease transfer between humans – as that would not have been considered unethical (and so "absolutely forbidden") even at a time when animal (i.e., non-human animal) research was not widely questioned:
I believe that I have experimentally demonstrated that phthisis, like syphilis and glanders, is communicable by inoculation. It can be inoculated from humans to certain animals, and from these animals to others of the same species. Can it be inoculated between humans? It is absolutely forbidden for us to provide experimental proof of this, but all the evidence is in favour of an affirmative response.
Villemin, 1868/2015, p.265
So, Villemin did not demonstrate that T.B. could be transferred between people, but only that analogous transfers occurred. So, in a sense, the context of justification, as well as the context of discovery, relied on analogies. Despite this, the indirect evince was strong and Villemin's failure to persuade most of the wider scientific community of his arguments likely reflected the general paradigmatic beliefs at the time that disease was caused by hereditary weakness, or through broad environmental conditions, rather than minute amounts of material being transferred between bodies.
Mycobacterium tuberculosis – the infectious agent in tuberculosis – could only be detected once suitable microscopes were available – Koch published his discovery of the bacterium in 1882.
(source: Wikipedia Commons)
Koch was able to be more persuasive because he was also able to actually identify a microbe present in diseased bodies, as well as show inoculation led to the microbe being found in the inoculated animal. That shift in thinking required the acceptance of a different kind of analogy: that the presence, or absence, of a bacterium in the tissues mapped onto being infected with, or free from, a disease.
| present in tissues | Mycobacterium tuberculosis a microscopic 'germ' – only visible under the microscope | absent in tissues |
| ↕︎ | ↕︎ | |
| infected | tuberculosis a widespread and often fatal disease of people and other mammals | not infected |
Sources cited:
- Daniel, T. M. (2015). Jean-Antoine Villemin and the infectious nature of tuberculosis. The International Journal of Tuberculosis and Lung Disease, 19(3), 267-268. https://doi.org/10.5588/ijtld.06.0636
- Frith, J. (2014). History of Tuberculosis. Part 1 – Phthisis, consumption and the White Plague. Journal of Military and Veterans' Health, 22(2), 29-35.
- Goetz, T. (2013). The Remedy. Robert Koch, Arthur Conan Doyle, and the quest to cure tuberculosis. Gotham Books.
- Surget, A. (2022). Being between Scylla and Charybdis: designing animal studies in neurosciences and psychiatry – too ethical to be ethical? In Seminar series: Berlin-Bordeaux Working Group Translating Validity in Psychiatric Research.
- Taber, K. S. (2013). Upper Secondary Students' Understanding of the Basic Physical Interactions in Analogous Atomic and Solar Systems. Research in Science Education, 43(4), 1377-1406. doi:10.1007/s11165-012-9312-3
- Villemin, J. A. (1868/2015). On the virulence and specificity of tuberculosis [De la virulence et de la spécificité de la tuberculose]. The International Journal of Tuberculosis and Lung Disease, 19(3), 256-266. https://doi.org/https://doi.org/10.5588/ijtld.06.0636-v
Notes
1 As analogies link to what is familiar, they tend to reflect cultural contexts. At one time the mind was referred to as being like a slate. The once-common comparison of the brain to a telephone exchange has tended to have been largely displaced now by the commparison to a computer.
2 Whilst this is a common teaching analogy, it is also problematic if it taught without considering the negative aspects of the analogy (e.g. electrons repel each other, unlike planets; planets vary in mass etc.), and if the target concept is not clearly presented as one (simplified) model of atomic structure. See Taber, 2013.
3 "During both World War I and World War II in the US Army, tuberculosis was the leading cause of discharge [i.e., from the service]. Annual incidence of tuberculosis in the military of Western countries is very low, however in the last several decades microepidemics have occurred in small close knit units on US and British Naval warships and land based units deployed overseas. Living and working in close quarters and overseas deployment to tuberculosis-endemic areas of the world such as Afghanistan, Iraq and South-East Asia remain significant risk factors for tuberculosis infection in military personnel, particularly multidrug resistant tuberculosis."
Frith, 2014, p.29
4 Some horses have been bred to be fast runners, and others to be capable of pulling heavy loads. (That is some have been artificially selected to be like sprinters or cyclists, and others to be like weightlifters or shot-putters). The latter are variously called draft (U.S. spelling) / draught (British spelling) horses (US), dray horses, carthorses, work horses or heavy horses. When a load was too heavy to be moved by a single horse, several would be harnessed together into a team – providing more power. Ironically the term 'horsepower' was popularised by James Watts – whose name has since been given to the modern international (S.I.) unit of power – in marketing his steam engines. According to the Institute of Physics,
Whilst the peak mechanical power of a single horse can reach up to 15 horsepower, it is estimated that a typical horse can only sustain an output of 1 horsepower (746 W) for three hours and, if working for an eight-hour day, a horse might output only three quarters of one horsepower.
https://spark.iop.org/why-one-horsepower-more-power-one-horse
5 Alexandre Surget (Associate Professor at University of Tours, France) has even argued that the guidelines adopted in animal experiments are sometimes counter-productive as they encourage experiments with too few animals, and consequently too little statistical power, to support robust conclusions – in effect sacrificing animals without reasonable expectations of securing sound knowledge (Surget, 2022).
Any research that makes demands of resources and the input of others, but which is designed in such a way that it is unlikely to produce reliable new knowledge, can be considered unethical.