The Roots of Progress

Why did we wait so long for the threshing machine?

When ripe wheat is harvested, the edible seed is encased in an outer husk. Before the seed can be ground into flour, or boiled into porridge, or planted in the field to produce next year’s harvest, it must be removed from the husk. This process is called threshing.

As the husk is quite hard, threshing is a violent process. Traditionally, it was done with a tool called a flail, which is simply a short stick attached by a cord to a longer handle. The grain was spread out on the ground (yes, disgusting) and beaten with the stick to open the casings.

The Growth of Industrial Art, 1892

Other methods included “treading”, in which livestock trampled the grain with their hooves (yes, even more disgusting) or dragged a sledge over the grain (the Latin word for this sledge is tribulum, from which we get the world “tribulation”).

Horse treading
Horse treading The Growth of Industrial Art, 1892

Occasionally the grain would be rubbed against a wire screen, or placed in a sack and beaten with rocks. It’s no coincidence that the word “thrashing” is similar: it is an archaic spelling of the same term.

As one of the more labor-intensive stages of wheat farming, threshing was a natural candidate to be automated by machinery. And the threshing machine was a relatively simple device: like the cotton gin, the flying shuttle, or the bicycle, it was a mechanical invention that did not depend on any scientific discoveries. Still, threshing machines were not used to any significant degree until the late 1700s in the UK and the early 1800s in the US.

Once again, the question arises: why did we wait so long?

Early claims and concepts

First, it’s not because no one had the idea to mechanize the threshing process. Although we think of the Industrial Revolution as a uniquely inventive age, and the era before it seems blind to the possibility of innovation by comparison, we must remember that there were inventions prior to 1700, many of which were broadly adopted, especially when they helped with essential economic activities. The loom and the printing press predated the threshing machine by centuries, and neither seems far less complex.

The threshing machine was referred to in an English patent as early as 1636 (for historical context, this is not long after the major works of Galileo and Francis Bacon). Sir John Christopher van Berg, a knight of Moravia (part of the present-day Czech Republic), fleeing religious wars in Germany and seeking protection from King Charles I, offered to Charles and to England his observations of “diverse mechanical instruments and frames operating by weights.” The patent is extremely broad and vague by today’s standards, listing a literally incredible variety of machines: for pumping water, washing clothes, cooking meat, working bellows, dredging rivers, raising sunken ships, measuring distances and depths, and for pretty much any mechanical operation. Towards the end of the list is mentioned an invention “to be agitated by wind, water, or horses, for the clean threshing of corn [grain], whereby much corn that is now left may be saved, and the straw made near as good as hay.” At the time, patents did not require detailed descriptions, diagrams, or models, and there is no more insight than this into how the machine might have worked, if indeed the invention was even real. The point is that by the early 1600s, the threshing machine was seen as possible and desirable.

And why not? The grinding of grain into flour had been mechanized by watermills in ancient Rome. The threshing process was similar enough that some type of mill could plausibly accomplish the task. A century after Van Berg, inventors really started working on the problem, with one Michael Menzies of Scotland patenting a threshing machine in 1734, and several more efforts throughout the 1700s.

So why did none of these machines gain wide adoption?

Challenges with reliability

One clue comes from reports of trials of the early machines.

Some of the machines simply broke. Of Menzies’s machines, one historian wrote that “owing to the velocity required to do the work perfectly, they soon broke, and the invention fell into disgrace” (Somerville 1805, p. 75). Another machine was trialed in Scotland in the late 1700s, but “in a few minutes the model was torn to pieces” (Quick 1978, p. 45).

When the machines did not break, they often failed to do the job. In one trial in 1753, it was found that the machine “broke off the ears of barley and wheat instead of clearing them of the grain, and that, at best, it was only fit for oats” (Ransome 1843, p. 140). Another was attempted in the late 1770s, but: “Upon trial this machine was also found defective, as along with its doing very little work in a given time, it bruised the grain, and so materially hurt its appearance, as to lessen its value considerably in the market” (Somerville 1805, p. 76).

In general, McClelland (1997) says that threshing machines in the late 18th and early 19th century were “among the most complicated and expensive of all agricultural implements in American and Britain. High price and frequent breakdowns did not bode well for rapid adoption when ‘the least derangement … is death to the whole machine’” (p. 172). He reports that Washington and Jefferson were both interested in threshing machines, but (pp 175-6):

Washington’s machine… built using the plans of William Booker soon proved unsatisfactory; a Maryland machine constructed with the aid of Colonel Anderson’s plans developed a warped wheel and was abandoned; in 1802 an immigrant from Edinburgh introduced into the Mid-Atlantic states “six or seven” based upon “the Scotch principle,” but they soon developed problems and “common workmen” could not repair them; the machine patented in 1803 by Jedediah Turner, based “upon entirely new, and very plain principles” was never commercially produced.… Washington perhaps spoke for a number of would-be agrarian pioneers when he voiced his frustration with this prospective new technology in 1793: “I have seen so much of beginning and ending of new inventions, that I have almost resolved to go on in the old way of treading.”

What was the problem?

One challenge was to find the right basic idea for the design of the machine. As is often the case, some of the early attempts tried to mimic human motion too closely: these machines automated the motion of the flail, but continued to beat the grain on the ground, an idea known as the “beating” principle. “All machines based upon the beating principle suffered from the same problem. The violence of the mechanical action used to separate the grain soon produced broken parts and an implement in disrepair” (McClelland 1997, p. 170). This was the problem with Menzies’s 1734 machine.

Design for a “beating” type machine
Design for a “beating” type machine Thomas Paine's Pennsylvania Magazine, February 1775
Another “beating” concept
Another “beating” concept The Growth of Industrial Art, 1892

A better idea was the “rubbing” principle. This design was more like a mill or a cotton gin: the grain was fed into a rotating drum that had pegs or other protrusions to grind the husk off of the seed. While “beating” vs. “rubbing” was still debated as late as the 1830s, all successful machines were of the rubbing type.

Meikle's threshing machine, 1786
Meikle's threshing machine, 1786 Wikimedia
Pope's threshing machine
Pope's threshing machine American Farmer, 1823

But the threshing machine didn’t seem to take off after one core design breakthrough. There isn’t a single individual who is always called out as “the inventor of the threshing machine,” nor a single date that stands out as an inflection point. Instead, one finds series of important inventions across decades: Andrew Meikle built the first successful machine in Scotland in 1786, but Joseph Pope invented a popular one in the US in 1820, and the Pitts brothers made improvements to power and to winnowing (separating the wheat from the chaff) in the 1830s. Nor does there seem to have been a single design that became dominant. Instead, it’s a story of gradual improvements in effectiveness, reliability, and cost that led to gradual adoption by farmers. This pattern of iterative improvement applies to most inventions—even ones that have a “heroic” story in which a single invention and inventor are traditionally highlighted—but even so, the history of the threshing machine seems unusually long and incremental.

My hypothesis is that the threshing machine was just past a certain threshold of the combination of the amount of force required and the delicacy of the operation. A loom is a somewhat complex machine performing an intricate process, but not one that uses a high degree of force. A flour mill, or a trip hammer at an iron works, is a high-force application, but not one that is particularly subtle or delicate. Both of these were in use long before the Industrial Revolution. But a threshing machine seems to require enough of both characteristics that manufacturing quality became critical to meet the standard of reliability that was needed for practicality and adoption. A simple, reliable design helped with this, but competent workmanship was at least as important.

Challenges in manufacturing

Here’s some evidence for this from historical sources.

The American farm periodical Genesee Farmer, April 1831, in an article on the the threshing machine, writes:

… one of the great and principal causes of failures, in many kinds of machines, is the flimsy, cheap, and do-for the-present manner in which they are made. They are not unfrequently constructed by carpenters, or rather by those who are only an apology for a good one, and who could hardly construct a button to a barn-door…

By the operation of these causes the farmer often gets an ill-constructed, weak, and rickety machine, which needs wedging, nailing, and bracing, at every revolution…

He continues with specific advice on construction and especially on materials:

The machinery that generates the motion, whether horse or water power, ought to be as well constructed, and of as good materials, as a flouring mill; and it is worse than useless to make the main wheel and pinion gearing of wood. Nothing but cast iron, and that of the softest and best kind, can be depended upon.

(Using iron instead of wood has also been noted as a key step taken by Henry Maudslay to improve the precision of machine tools.)

On the British side, Ransome (1843) defends the concept of the threshing machine against the bad reputation it might acquire from poor workmanship:

It has been urged against these machines, that they are apt to break the straw, and that they bruise and nib the barley so as to render it unfit for malting; but these faults are not so much attributable to the principles of the machines, as to the manner in which they are frequently turned out of the hands of the workman; and sometimes to the want of skill and judgment in the parties who have the management of them.

Many of these machines are made by persons who possess little claim to any mechanical knowledge, and who, purchasing the unfitted castings, by the help of village artisans, produce an imitation of those which are considered good. As the perfection of these machines must depend upon mathematical accuracy in the adjustment of all their parts, and in the truth and precision of their fittings, it is unreasonable to expect that this can be accomplished where no facilities exist beyond the forge and the work-bench; and hence arises a degree of discredit, which is unfairly thrown upon the principles upon which the machine is formed.

So, if manufacturing quality was a critical factor—how were the early machines made?

Manufacturing systems

Today, agricultural equipment is manufactured by large, specialized firms who distribute their product worldwide. But in the 1700s and even into the 1800s, there was no such thing. “Agricultural equipment” meant plows, scythes, and wagons, and it was made by local craftsmen, such as the town carpenter, blacksmith, or cartwright. Not only were there very few large manufacturing enterprises in the world, but until the growth of the railroads in the mid-1800s, there were also no efficient transportation networks to distribute their products to farmers throughout the countryside. (See also MacDonald 1975 on this point.)

In the promotions of early threshing machine inventors, we can see how they expected their machines to be acquired.

The Virginia Gazette, Williamsburg, 1772

In many cases, an inventor merely proposed to supply plans and/or models, assuming that the machine would be built by a local craftsman, typically a carpenter or especially a millwright (millwrights were the closest thing to mechanical engineers in the 1700s). A 1772 advertisement in The Virginia Gazette promotes an invention that “may be carried into execution by any tolerable carpenter” or even “by gentlemen’s own servants.” The inventor, one John Hobday, does not even seem to expect to sell the plans, let alone the machines: he proposes that a subscription be raised for him. More than fifty years later, in 1823, not much has changed: an article in the American Farmer on Joseph Pope’s threshing machine quotes a letter from the son of the deceased inventor as saying that the machine “can be constructed at little expense, the materials, including the shears, cost $13, and it can be made by a good workman, (say a joiner or carpenter) in 12 days.”

In other cases, inventors offered not the plans but the parts, “some assembly required.” In 1735, Andrew Good Wright in Edinburgh advertised a threshing machine, saying: “Those who want them, must send for them from Edinburgh, and a millwright to receive them, that he may know how to set them up; which any millwright will be able to do, after he has seen one going.” He also mentions a licensing opportunity: “Millwrights recommended by gentlemen for their integrity, if they come to Edinburgh, and understand the machine, shall have liberty from the patentee to make them in the country.” A 1796 advertisement from “John Jubb, millwright and machine-maker,” offers both options: he will “send it to any part of the kingdom” for 25 guineas plus shipping and handling, or he will sell the plans along with “the machine made at Leeds, taken to pieces, properly marked, and sent off… so that any workman may with ease put it together.”

The National Gazette, Philadelphia, 1823

There is some evidence that threshing machines thrived in places and times when they were made by the inventor himself, or by a skilled mechanic who decided to specialize in them. One source (pp 3–4) reported as of 1800 that “almost all the threshing machines in England” had been built by one “industrious workman, of the name of Stevenson”. The article noted approvingly that his machines moved easily by hand and showed little friction, “a sure warrant of accurate workmanship.” As for Joseph Pope’s machine, soon after the article quoted above, he contracted with an engine maker to be his manufacturer. From May through October 1823, the Philadelphia National Gazette ran advertisements stating that “Messrs. S. V. Merrick & Co. Engine Manufacturers, Philadelphia,” had been invested with the privilege of making and selling his machine; prospective customers were encouraged to “apply” to them to obtain one. An engine manufacturer, at that time, would have had a well-equipped machine shop with the precision tools and skilled workers needed to make reliable machines.

Bottom line: In its early decades, the threshing machine suffered from inconsistent quality due to distributed, unspecialized workmanship.

Contrast with the reaper

In contrast, consider a closely related piece of agricultural equipment invented decades later: the mechanical reaper.

The reaper cut down stalks of grain in the field, automating the step just prior to threshing. As with the thresher, reliability was a crucial challenge for the reaper: early models failed to cut tangled thickets of grain, or clogged when the stalks were damp, or performed poorly on hills.

McCormick reaper, 1831
McCormick reaper, 1831 Scientific American

One of the most successful reapers was created by Cyrus McCormick. He began selling them in the 1840s, but he relied neither on local craftsmen nor on manufacturing partners. He made the first machines himself, and eventually created a large, central factory.

McCormick sought a nationwide market, but the primitive cargo networks were a barrier. Casson (1909), one of his biographers, writes (p. 62):

To get the seven Reapers [sold in 1844] to the West, they had first to be carried in wagons to Scottsville, then by canal to Richmond, re-shipped down the James River to the Atlantic Ocean and around Florida to New Orleans, transferred here to a river boat that went up the Mississippi and Ohio Rivers to Cincinnati, and from Cincinnati in various directions to the expectant farmers. Four of these Reapers arrived too late for the harvest of 1844, and two of them were not paid for.

McCormick later moved from Virginia to Chicago, setting up shop at the heart of a transportation network that integrated railroads and waterways—and in a location that was closer to his biggest market, in the Midwest (where the lands were flat and labor was scarce).

He also innovated on business practices to sell his machines: he advertised heavily in newspapers, organized field demonstrations, offered farmers a written guarantee and free credit until the next harvest, and built out a network of distributors in each region.

No one, from what I’ve read, did any of the above for the threshing machine, at least not before 1830 or so, when the machines were already common.

Why did it take so long?

The question of the threshing machine is a microcosm of the bigger question I’m interested in: Why did we wait so long for the Industrial Revolution?

Some inventions depended on theoretical concepts that were not discovered before the Scientific Revolution: the electric generator, for instance, or even the steam engine. But others did not: the threshing machine, the cotton gin, the spinning jenny. What is the difference between those inventions, and others of seemingly similar complexity and importance that were adopted centuries earlier: the loom, the spinning wheel, the printing press?

One theme we see here is the manufacturing capability required to make affordable, reliable machines. This was probably also important for the bicycle, I think even more so than I realized or emphasized in my previous essay on that topic. For another example, Robert Allen (2009) says that precision manufacturing from the watch industry was necessary for an early thread-spinning machine known as the “water frame”: “The watch industry was the source of gears—brass gears in particular—and they were the precision parts in the water frame…. Without watch-makers, the water frame could not have been designed” (p. 204).

A secondary theme here is that there are prerequisites to market creation. If specialized manufacturing is required, then a centralized manufacturer must be able to reach a wide market. This requires communications infrastructure, such as newspapers, in order to market the goods, and transportation infrastructure, such as railroads, to ship them.

One reason perhaps (I’m speculating here) that mechanization came to textile manufacturing decades before it came to agriculture is that the business model was different. Richard Arkwright made his spinning machines, not to sell them as McCormick sold his reapers, but to use them to make cheap thread. He was his own market for the machines; they were a capital investment in a new type of business. This business model wasn’t really available to early agricultural equipment makers.

Another instructive comparison is to the steam engine, which also required high-quality, precision manufacturing. Boulton & Watt adopted the centralized, specialized factory in the 1770s, decades before this model came to agricultural equipment. My speculative hypothesis here is that farmers, as a market, were a large number of smaller and more geographically distributed customers, vs. the coal mines, ironworks, breweries, etc. who were the customers for Watt’s engines. Thus the challenges of marketing and distribution were greater in the agricultural market. (Not to mention that industrial concerns such as mines might have had more ability than small farmers to finance a large expenditure on capital equipment.)

Flywheels

Maybe all this can be summarized as infrastructure. Manufacturing capability is infrastructure, as are railroads and newspapers (or today, trucking and the Internet). Infrastructure lowers the activation energy of any particular development. Sometimes these developments themselves create more/better infrastructure, creating a reinforcing cycle that generates exponential growth.

Is this effect enough to explain the pace of progress throughout history? Do we still need to posit cultural causes—a general factor of inventiveness—Joel Mokyr’s “idea of progress” vs. “ancestor-worship”, Deirdre McCloskey’s “honor and prestige for the bourgeoisie”? It seems obvious that it at least matters whether people believe that progress is possible and desirable (and if it doesn’t, then it certainly seems like an enormous coincidence that the Industrial Revolution happened not long after Bacon and the Scientific Revolution, and in the same part of the world). But now that I more vividly understand the challenges facing 18th-century threshing machine inventors, I find myself shifting a bit more towards infrastructure as a cause.

Or, more broadly, I’m seeing that there is a set of overlapping flywheels, each creating a virtuous cycle as it gets going. Better infrastructure enables more progress, which in turn creates more and better infrastructure. The belief in progress leads to actual progress, which reinforces the belief. And there are others: Surplus wealth allows us to invest in progress, which creates more surplus. Science ultimately leads to progress, which then helps advance science. And since all these cycles intersect at progress itself, they all reinforce each other, indirectly if not directly.


Bibliography

Allen, Robert (2009). The British Industrial Revolution in Global Perspective

Casson, Herbert Newton (1909). Cyrus Hall McCormick: His Life and Work

MacDonald, Stuart (1975). “The Progress of the Early Threshing Machine”

McClelland, Peter D. (1997). Sowing Modernity: America’s First Agricultural Revolution

Quick, Graeme and Wesley Buchele (1978). The Grain Harvesters

Ransome, J. Allen (1843). The Implements of Agriculture

Somerville, R. (1805). General View of the Agriculture of East Lothian

Key excerpts from the above, as well as citations from newspapers and other corroborating sources I did not reference, can be found in this summary of primary/historical sources.

Van Berg’s 1636 patent


Thanks to Anton Howes for conversations, help finding sources, and comments on this essay, all of which contributed greatly to it.

Comment: Reddit, LessWrong

Get posts by email:
Become a patron

Get posts by email