created: 28 Oct 2018; modified: 03 Nov 2018; status: finished; confidence: likely; importance: 6
Like Engend’ring Like: Heredity and Animal Breeding in Early Modern England, Russell 1986: Expansion of Russell’s thesis, reviewing some of the early English breeding literature, focusing on thoroughbred horses then sheep. There is not much original research as far as I can tell, it is primarily review of existing primary/secondary sources, but it gives a useful overall arc from the dark ages of English agriculture up to post-Robert Bakewell. What is most interesting is the intellectual history we can extract from it in terms of inventing heritability and as important, one of the inventions of progress in the gradual realization that selective breeding was even possible.
The Invention of Heritability
I was interested primarily in the section on Robert Bakewell, the Moneyball of sheep breeding, about whom information remains difficult to get, but also in the broader question highlighted by
Heredity Before Genetics: A History, Cobb 2006: given how blatant the general phenomenon (however difficult the actual mathematical analysis of modern quantitative genetics) appear to be - averages or additive effects everywhere, equal parental effects, normal distribution of traits, selection on measured traits produces gradual but cumulatively enormous gains and leading to creation of breeds or even speciation - why were Bakewell and other breeders such an intellectual revolution? Why could they influence Charles Darwin and Mendel so much and gradually revolutionize agriculture?
People, both ordinary and men of leisure, have been farming animals for millennia and presumably interfering in their reproduction, and had ample opportunity to informally observe many matings, long pedigrees, crosses between breeds, and comparisons with neighboring farmers, and they had great incentive to reach correct beliefs not just for the immediate & compounding returns but also from being able to sell their superior specimens for improving other herds.
But surviving theoretical scientific discussions of heredity are baffling. People lurch between
only fathers matter &
only mothers matter, endlessly elaborating on wildly speculative (and wildly wrong) mechanistic explanations of how exactly sperm & eggs & embryos connected and formed, and in an example of
hard cases make bad law, the focus on
monsters and other extreme cases among humans or animals badly misguided their premature attempts to elucidate universal principles comparable to that of astronomy or physics - the examples did not
prove any rule, but baffled everyone trying to come up with a rule to prove. (Cobb remarks
Réaumur and Bonnet’s discovery of aphid parthenogenesis in the 1740s had not helped matters, to say the least - and parthenogenesis, aside from existing in animals too, is far from the oddest reproductive system or genetics in insects or plants!) We now know that here are an indefinitely long list of ways that development can go wrong with thousands of environmental insults or developmental error or distinct genetic diseases (each with many possible contributing mutations), and that for the most part, each case is its own special case; sometimes
monsters can shed light on key aspects of biology like metabolic pathways, but that requires biology centuries more advanced than was available, and that the search for universal principles was futile. There are universal principles but they pertain mostly to populations, must be investigated in the aggregate, statistically, and individual counterexamples can only be shrugged at, as the universal principles can be and often is overridden by many of those special cases.
Why didn’t farmers & breeders correct the theoreticians? Any experience with selective breeding or cross-breeding should have strangled
preformationism in its crib, and it would not have been hard to look over a good pedigree and note that, say, elite mothers had elite offspring about as often as elite fathers had elite offspring, and that it couldn’t be true that eggs were passive sustenance for an embryo (or alternatively, that the sperm was a mere spark igniting the egg).
Unfortunately, breeders & farmers were also deeply confused. Rather than carefully preserving & breeding prize animals, the best members of a herd would often be culled for eating or selling at the highest profit, and the worst left to reproduce1 (a
negative breeding strategy although perhaps
dysgenic would be more immediately comprehensible), since, after all, heredity didn’t matter as much as environment. Rather than a keen focus on the end-goal, whatever that was, farmers would choose for
fancy points, whatever was in fashion at the moment, and naturally there were fierce disagreements about what a
beautiful sheep looked like. Abhorrence of inbreeding (due to human norms or a great fear of inbreeding depression) prevented realizing that inbreeding could be used to create new inbred lines which fixed desirable traits & were man-made breeds, rather than breeds simply being pre-existing natural kinds from time immemorial. The lack of record-keeping or consistent control over reproduction, with catch-as-catch-can matings, culling, environmental shocks, ad hoc introductions from other breeds, and ultimately, tiny average populations controlled by each individual compared to modern farming - all introduced noise. (Bakewell’s own experiments required a large herd, and he either did or almost went bankrupt, sources conflict, despite the extremely high sums he was ultimately able to charge for his animals.2) The belief that only fathers mattered led horse breeders astray: they failed to race mares, and then took the ludicrously expensive imported Arabian stallions and crossed them with random mares, and then took the sub-par performance of their offspring as evidence that race performance was critically dependent on the dry Arabian environment and they simply had to keep importing & crossing.3 Many breeders, going back to the Greeks like Xenophon & Plato or Roman writers thereafter, focused on the circumstances of conception, emphasizing the critical importance of breeding a male in his prime rather than an older one, no matter how much better the older one had performed in its prime, breeding females with large temporal gaps if highest quality is necessary, and particularly avoiding any obese animals, as obesity sapped vitality, impeding the production of blood, from which semen & eggs derived - Greeks supposedly would deliberately starve cows to ensure svelteness before impregnation. And Varro held that twinning in goats/sheep was due to the parents, but that the trait could be both inherited by a male but also was communicable ie. might be carried from a female by the impregnating male to successive females, and so a breeder should try to use twin-begetting goat/sheep as much as possible. (I’ve heard of paternal age effects and epigenetics before, but this is ridiculous!) Considering all of this, I am no longer surprised that selective breeding was invented so late nor that such low-hanging fruit went unplucked for millennia. Under this statistical blizzard of contradictory noise, how was anyone to discern the simple biometric patterns?
Cobb rousingly & convincingly concludes:
It took humanity a remarkably long time to discover that there are consistent relations between parent and offspring, and to develop ways of studying those relations. The raw phenomena of heredity were sufficiently complex to be impervious tocommon-sensereasoning, to the brilliant but stifling schemas that were developed by the Greek philosophers, and even to the stunning forays of the early scientists. What was required was not a novel piece of apparatus, nor even a new theory; the key thing that was needed was statistically extraordinary data sets. On the one hand, these were composed of many reliable human pedigrees of unusual or pathological characters; on the other, they were the large-scale experimental studies that were carried out consciously by Mendel, or as a by-product of the commercial activity of eighteenth-century livestock breeders.
Like all great truths, heredity seems obvious once it is understood. But the fact that so many people took so long to realize what we take for granted does not mean that our predecessors were stupid. Instead, it indicates that, before the patterns within hereditary phenomena could be detected, society had to develop to a sufficient level for these kinds of data to be collected, examined, compared and interpreted. However, although science, written family records and large-scale agricultural production were the prerequisites for the discovery of heredity, the birth of our science was not simple, and required bold thinkers who were prepared to resolve an issue that had perplexed humanity for thousands of years. The result - the twin fields of genetics and evolution - represents one of the greatest insights in human history.
Early English agriculture & breeding
What of this progress do we see in Russell’s trace of history? In an example of The Fable of the Bees, the idle wastrel amusements of the aristocrats & aping by their lessers may have been key: because the development of Thoroughbred racehorses starts Russell’s story!
The aristocratic & government interest in racehorses & war horses gradually led to many specialized horses being kept and better record-keeping. The creation of the
stud book and the classist superstition of
blood in horses, where even distant ancestry from a famous thoroughbred elevated a horse above common horses, appears to have accidentally backed into success: by creating a reason to track ancestry carefully, and importantly, ensuring that a thoroughbred’s offspring with a horse not in
the book would be worth much less (regardless of their performance or true genetic potential), a closed breeding population under steady selection was created and ensured that what progress was made was not then immediately undone by careless haphazard matings. Arabian/Turkish stallions were permitted, and now that they were no longer being immediately diluted by outside 100%-non-thoroughbred horses, gradually
graded up the closed thoroughbred gene pool towards more Arabian/Turkish genes. Further, the mania for racing was not satisfied by the stock of mature stallions so races began expanding to include younger horses (accelerating generational turnover and thus annual gains) and also mares (finally capturing critical performance data and allowing selection on the other half of the equation). Before too long, the import of full-blooded Arabian stallions was no longer particularly necessary as thoroughbred performance had thoroughly outraced them. Russell remarks that thoroughbreds, like cats or dogs, were then (and still are, based on the crudity of the racehorse genetics papers I’ve read) bred in an unsystematic and inefficient manner, but this seems to have been enough. Other English fashions, like the demand for carriage-horses due to increasing wealth, rapidly molded various breeds of horses larger & smaller as necessary. The sudden sustained progress in racehorses and changes demonstrating malleability probably did not go unnoticed.
Early cattle/sheep productivity was low. Dairy cow yield, for example, appears to have been probably below 300 gallons a year (pg129); for comparison, contemporary dairy cow yield is closer to 2,300 gallons a year & increasing (~1%/annually genetically). Dairy cattle breeding optimized for as few males as possible, since they did not produce milk and were needed essentially only to impregnate the cows, and so calving would happen as simultaneously as possible early in the spring. A village might collectively pay for a single
common bull to
tup the villagers’ cows, but wealthier farmers might buy or simply hire individual bulls. If you only need one, and that one will be kept busy impregnating as many cows as possible, you want the best & lustiest one, and it naturally develops into a selective breeding program. (Although having only one male is far from optimal for maximizing the long-term response to selection.) After a few years, the bull slows down, one fattens it for the butcher, and buys a new one. Bakewell, in addition to his more famous sheep, was also involved in steer & horse breeding and would’ve been familiar with this. By the time Bakewell began in the 1740s, English sheep farmers had been struggling with changing market incentives: small sheep, while tastier, didn’t fetch a sufficient premium, and likewise, fine wool wasn’t premium enough to compensate for the small amount of fleece on such sheep; they had begun explicitly seeking out and buying large high-meat/wool-yielding sheep.
This perhaps formed the jumping off point for Bakewell. Bakewell began carefully measuring his animals & paying attention to the offspring of any hired-out males to gauge the males’ genetic quality, optimizing for fast growth and fattiness, even preserving joints in jars from previous specimens the better to compare with current animals, and perhaps practicing more inbreeding than other contemporary breeders. Russell is critical of the extent to which Bakewell’s Dishley sheep was really an economic success or to what extent better measurements were responsible for improvements, calling some of the later prices Bakewell charged
more to do with theatre and the cunning exploitation of fashion than any relationship with the breeding value of stock and noting that an unknown but possibly large amount of the Dishley sheep’s quality was due as much to Bakewell’s
superb standards of husbandry and assiduous investment in environmental improvements like irrigating fields & feeding his animal high-quality pasturage & being extremely kind/gentle to his animals (some of the many visitors, domestic & international, to Dishley would note that the animals were remarkably happy, calm, and good-natured, and that Bakewell was also beloved by his employees). One improvement I particularly liked was Bakewell’s construction of a canal for carrying fodder around: a worker would toss some into the canal, which would then carry it to the main house into a pool, washing it along the way. Russell concludes (pg215), after reviewing some later small-scale data from the Annals, that:
It must be doubtful if food conversion or carcass ratio were significantly improved, or that the fundamental form of the carcass was changed either in the Bakewell strain of Longhorn cow or in Lincolnshire Wold sheep. However, the animals looked much better grazier’s animals, with their tendency to fat up and round out well. In a sense, all Bakewell had done was to create a new, if somewhat more rational fancy for sheep of a particular shape, rather than merely tinkering with colour or horn form. On the other hand, it would seem that the Lincoln breeders had genuinely succeeded in breeding animals with a greatly increased fleece yield, although the death of the longwool market made their achievement a pointless one. Sensibly they reverted back towards the form of the Wold sheep from which they had started, of which the best surviving examples were the Dishleys. The use of Dishley stock by the Lincoln breeders was, of course, made much of by the Leicestershire men, although it probably did not have the significance the latter ascribed to it. Certainly Bakewell must take considerably credit for publicising the idea of selecting stock for economic performance, but whether his actual achievements in this field were of any significance remains doubtful.
Inasmuch as there do not seem to be any surviving records from Bakewell (!) and Bakewell never wrote up his data or methods, only discussing it with visitors, it is difficult to say either way. Given how many subsequent gains have come from breeding and the low initial level and the fact that response to selection is expected to be greatest at the beginning, I incline towards thinking that Bakewell did cause large genetic gains by simply being thorough and exercising some care, and of course his environmental improvements may also have been critical to his genetic success by allowing each animal to reach its genetic potential, increasing heritability/reducing non-shared-environment effects. (If the Dishley breeds had been maintained to the present day, it might be possible to partition gains from environment and genetics by common-garden experiments or by using polygenic scores, but unfortunately, they appear to have all long since disappeared or merged into other breeds which have undergone intense selection since then, and are no more available for study than the Dishley records.)
In any case, Russell makes an interesting suggestion there. If Bakewell’s true contribution was
publicising the idea (which can be further buttressed by noting that Bakewell was widely praised in the 1700s & 1800s, citing Charles Darwin invoking Bakewell in the context of natural selection as demonstrating what explicit selection can do even to the point of making a new breed, the goal of German sheep breeders to emulate Bakewell’s supposed success, etc) what can we draw out of this?
Bakewell and the Invention of Progress
One way would be to say that Bakewell played a part in the invention of Progress or the
It is not a universal belief among humans that it is possible to
progress; Whiggism must be learned. Isaac Newton, for example, regarded the recent progress as evidence that human history went in cycles of creation & destruction, and believed that his research on gravity or the Philosopher’s Stone was merely recovering what the Ancients knew & had been lost. Breeders likewise regarded selection as merely frustrating the inevitable decay of herds under inbreeding & local environments. One imported a Turkish or Spanish or Arabian stallion to try to temporarily elevate one’s horses, but that was to try to borrow some of the ancestral power of a born & raised foreign racehorse - no permanent gain was looked for nor, apparently, seen, and one simply kept importing. The idea that it is possible to almost arbitrarily improve a breed’s traits, or steer a breed in a direction to the point that it would have to be considered a new and clearly distinct breed for all intents & purposes, appears to have not been in circulation. It would have been deemed absurd, worthy of parody in the Laputa of Gulliver’s Travels, to imagine that dairy cows could one day yield >8x more milk. Most merchants & aristocrats dreamed of nothing more than to own a large landed estate with an annual income of several hundred or thousand pounds sterling, and their descendants living off the land-rents for eternity (and land prices reflected this, with prestigious full ownership costing far more than 99-year leases).
The market is a weighing machine, but where do the things weighed come from? What differentiates a complacent society from an innovative society? If chance favors the prepared mind, what is the nature of this mental preparation?
The Improving Attitude
In a recent talk, economic historian Anton Howes4, who studies a similar period of English history, specializing in the Royal Society of Arts (est. 1754), brought up many interesting points about innovation and progress, putting together a circumstantial case for the role of social imitation & elite competition in driving innovation, what you might call the Velvet Underground model of innovation. Rather than innovation & progress just sort of happening on its own or being driven by accumulating assets, progress appears to be caused at least partially simply by an attitude of progress, of people competing to be innovative, and of simply looking at age-old things and going
why do we do it that way? Why not do it this other more sensible way?
The Royal Society of Arts is a case in point: the RSA was founded to encourage technological and practical innovations by contests, funded by
subscriptions from aristocrats & well-to-do bourgeoisie, But it did not have the money to directly fund people to do R&D or pay for getting a patent or buy out existing patents, and instead, based on votes, primarily awarded medals and occasionally substantial but still relatively nominal monetary prizes. (They were not awarded to patented things. This was not as much of a restriction as it might seem because patents, being so expensive in real terms, rarely obtained, often not useful when obtained, and nearly abolished in the early 1800s, seem like a minor player at best. Corporations, likewise - just about anything you might do with a limited-liability corporation could be done with a trust instead.) And this… apparently worked really well? For example, the RSA takes credit for 60 million trees planted by the landed gentry starting in 1758, simply by awarding a gold medal to the Duke of Beaufort followed by
various other dukes, duchesses, earls, viscounts, marquesses, bishops, and members of parliament, not to mention many more untitled members of the minor gentry. Aristocrats and their neighbours engaged in a The only payment it could make was prestige, reflected from the English aristocracy, and this was apparently adequate, indeed, perhaps even more motivating than mere wealth. (I’ve often been baffled by how medals were endlessly awarded before the 1900s, and now I wonder if I over-hastily dismissed the idea that medals could be real motivation for anyone, simply because I can’t imagine being motivated by yet another symbolic medal.) The founder of the RSA, William Shipley, it is worth noting, was inspired by… horse racing: he had noticed the tremendous efforts invested in it, all out of disproportion to the awarded prizes, and the resulting progress, and sought to harness that energy for more socially-valuable purposes.5
very laudable emulation, each vying to out-do one another in the extent and quality of their plantations.
Other observations follow. Some major inventions are so simple as to defy belief they were not invented thousands of years ago - the flying shuttle, for example. Many inventors had little or no training or experience in the field they invented something in, and might be a lawyer or something else entirely (like a small boy irritated at being assigned to a steam engine), and otherwise sometimes seem incompetent; John Kay of the flying shuttle claimed to have spent only a month apprenticed before making the first of his textile inventions. Inventors who lived in a neighborhood with a high per capita patent rate are themselves more likely to file a patent. Future inventors might correspond with their heroes (shades of the
college of letters) and, if they then meet them in person, they are more likely to go on to innovate, even if it was only a single short meeting and the hero was in an entirely different field and so it is difficult to see what key fact, skill, or wealth/object they could have transmitted which might make any difference. Innovation appears to be contagious: the society of clockmakers was often hired by scientists to make instruments for their needs, and the hired clockmakers started innovating more and this spread to the rest of the guild. Immigrants (not all Scots, France/Germany/America are common origins), and religious Dissenters, such as Scots Presbyterians moving south into English Anglican territory, are constantly overrepresented, as are mentions of The Great Exhibition. For comparison, France appears to have been much less practically innovative and slower to industrialize; why? Howes suggested it reflected different elite priorities: the French aristocracy was much stronger & wealthier than the English, and had an inclination towards purer, more abstract, more universal theorizing. Technology and economic growth and health simply weren’t rewarded with prestige from a French RSA. And you get what you incentivize.
Taking Howes’s claims at face-value, we could expand on the model a little more. Passing over Girardian claims of mimesis as the most powerful force in society, it’s still intriguing to note parallels elsewhere.
Bakewell, it hardly needs to be said, follows the Howes model well: Bakewell had no special training or math or technology to offer, and his breeds have been much criticized for being useless in practice and disappeared, but what he did accomplish was endorse the idea of progress, providing a model to emulate, and a prestigious figure to cite as precedent. Bakewell may have been grossly overrated by acolytes, but from this point of view, that is a feature and not a bug - the more praised the better! His influence then spread and sparked Bakewellites elsewhere and abroad, better equipped to successfully do what they (thought) he did.
Such a strategy sounds completely absurd and it’s hard to believe the ancients would do it even if Russell tells us they usually did up to the Industrial Revolution, but on the other hand, Richard Dawkins (The Ancestor’s Tale, pg46) notes that:
…It was probably more difficult [for people] to work out that it might be a good idea to keep back the best seed for planting, rather than follow the obvious course of eating the best and planting the dross (my father, as a young man fresh out of college, taught agriculture to peasant farmers in central Africa in the 1940s, and he tells me that this was one of the hardest lessons to get across).
It’s interesting to note that despite the social stigma of being a mere tenant & not a gentleman-farmer who of course owned the property, and occupying it for many decades, and investing enormous long-term efforts in improving it, Bakewell never bought his farm, Dishley, outright. Perhaps the necessary upfront money was always more productively spent improving his animals. It’s been noticed that 99-year leases for English property were extraordinarily cheap compared to outright purchase prices in the pre-modern era (
Pricing the Future in the Seventeenth Century: Calculating Technologies in Competition, Deringer 2017).↩
For all that people assume that historical people were super-genetic-determinists about anything and everything, a remarkable amount of naive environmentalism was current: for example, the European assumption, criticized by Sir Thomas Browne, that Africans were black only because of their sunny environment! Some first-hand experience with their offspring in Europe eventually refuted that, although Russell indicates that even this was not enough to convince environmentalists:
it was thus destined to be discussed interminably in the 18th and 19th centuries, with few able to accept that it was an entirely heritable phenomenon.↩
Homepage; relevant writings:
The Relevance of Skills to Innovation during the British Industrial Revolution, 1651-1851, Howes 2016 (working paper; previously titled,
The Improving Mentality: Innovation during the British Industrial Revolution, 1651-1851)
The Spread of Improvement: Why Innovation Accelerated in Britain 1547-1851
How Innovation Accelerated in Britain 1651-1851
Immigration and the Industrial Revolution
Economic Growth in Ancient Rome
Who will watch the Watch-Men? - Celebrating the Watch-Makers of the British Industrial Revolution
The Great British (Industrial) Bake-Off
Chim-Chim-Cher-ee: When Inventions Stealing Jobs was Popular: how scandiscopes replaced chimney-sweep boys
The Ideology of Inventiontalk: video (slides)
Definite optimism as human capital;
How Technology Grows (a Restatement of Definite Optimism), Dan Wang
The success of Shipley’s project, however, is due ultimately not so much to the force of his character as to the Tightness of the idea, simple, indeed naive, though it was, which he conceived and on which he based his Society. It is fortunate, therefore, that we possess, in a pamphlet dated 1763 and entitled
A Concise Account of the Rise, Progress and Present State of the Society for the Encouragement of Arts, Manufactures and Commerce, a detailed story of how the project gradually formed in his mind.
It really grew from two ideas, the first of which came to him as he watched the horse fair which was held twice a year at Northampton, where he resided. When he saw all the buying and selling that went on at the fair, involving large sums of money, he began to enquire into the cause of its success. He was told that it was largely due to the institution of horse-racing meetings; and that many of these had recently been promoted by the King and others who had presented plates or prizes for the various races. From this it occurred to Shipley that the gift of a comparatively few prizes had thus stimulated a whole industry; and impressed by this discovery he began to ask himself whether this, same principle could not be applied to stimulate other industries. He decided that it could, and subsequent history proved that he was right, but in the meantime there remained a second question - where was the money for even a few modest prizes to come from?
The 1763 pamphlet appears to be reprinted in William Shipley: Founder of the Royal Society of Arts; A Biography with Documents, Allan.↩
Speaking of Musk, why did I have such a vivid impression in 2014 that Musk endorsing the idea of AI risk & Bostrom’s book Superintelligence was a huge turning point in public interest & intellectual respectability of the topic? After all - what does Musk know about AI risk, really?↩
Would we find only-children overrepresented? Left-handers? Does this connect to why LGBT seem to be overrepresented in similar circles? Is there some sort of