Allan Willey
June 6, 1997
For the past several years I have been engaged in the task of attempting to introduce concepts, methods, and tools to support "architecture reuse" in a product development organization at Motorola. From my perspective, I have been attempting to introduce a new technology into an organization which doesn't perceive the utility or value of the innovation as compared to its short-term risks. My frustration led me to investigate the available literature concerning technology transition, and in the course of that investigation, I discovered an article by Brad Cox. Cox is currently pioneering an industry effort to introduce the notion of "Superdistribution; Objects as Property on the Electronic Frontier" as outlined in his book of that title (1). In his earlier work, Cox was discussing elements of a "software industrial revolution," and citing as a background example the introduction of interchangeable parts (2). There Cox says:
"Contrary to what a casual understanding of the industrial revolution may suggest, the displacement of cut-to-fit craftsmanship by high-precision interchangeable parts didn't happen overnight and it didn't happen easily. The heroes of this revolution were not the cottage-industry gunsmiths, who actually played almost no role whatsoever, for or against. ...
It was actually the ultimate consumer of ordnance products, Thomas Jefferson, who found the solution in 1785 during a visit to France prior to his presidency. Congress supported his proposal with remarkable steadfastness through 25 years of unsuccessful attempts, such as Ely Whitney's pioneering effort, until John Hall finally succeeded in 1822. An additional 24 years were to elapse before 'Armory Practice' spread to private contractors."
This "teaser" lead me to explore the story of interchangeable parts further. Without education or training in historical analysis I rely on the analysis and reporting of professional historians of technology for the details. In the brief essay below, I am relying entirely on secondary sources, and on the consensus opinions of historians who have examined these issues. It is my hope that these deficiencies won't substantially detract from the points I will raise in the end.
One of the key elements in the industrial revolution was the implementation of the simple idea of interchangeable parts in manufactured goods. While we take the concept for granted today, at the beginning of the 19th century no manufactured goods were produced which were built from interchangeable parts. Some authentic heros from our elementary school days played a sometimes surprising role in the pageant of events leading to their use.
Since interchangeable parts are so ordinary to our everyday life it is hard for us to conceive that this innovation was key and pivotal to achieving the "American manufacturing system" we take for granted today. Yet the story of how this took place will be instructive to anyone who is responsible for the introduction of any process innovation. To help set the stage, we should review a time line of key events, with some of the background historians have unearthed concerning these events.
While Ambassador to France, Thomas Jefferson visited the shop of Honore Blanc, a French mechanic, who was building muskets using hand-crafted parts of such precision that they could be interchanged. Jefferson wrote to a friend back home describing the obvious advantage to be had when such arms needed repairs. He apparently attempted unsuccessfully to persuade Blanc to move to America to set up a shop and demonstrate the techniques. Thus, a key early "sponsor" for this innovation was converted, and as the story unfolds, became central to its eventual success. It is important to point out that Jefferson had not seen any quantified results, but had only his intuition to guide him. None-the-less, Jefferson was convinced of the value of the concept on seeing it applied once to musket manufacture.
At a time when contractors were paid on delivery for arms purchased by the U.S. War Department, Eli Whitney, a well-connected and proven inventor convinced the government (through the use of his Yale, eastern, establishment connections) to advance him money to mass-manufacture army muskets. On June 21, 1798 he was awarded a contract to deliver 4,000 muskets by September 30, 1800. Whitney, who had never manufactured anything let alone muskets, had three principles he planned to employ: (1) the use of water-powered machinery, (2) producing uniform or interchangeable parts, and (3) run by unskilled but "steady, sober people." The transition to water powered manufacturing, mining, and milling was well underway at this time but the notion of producing and assembling high-precision goods using unskilled labor and interchangeable parts was unprecedented. It might be safe to say that no one other than Whitney would have been given this opportunity, and he only because of the sponsorship of Jefferson and other high level government officials who believed in him and his ideas.
In January of 1801, Whitney held a demonstration of the principle of interchangeable parts--and what a show it was! Whitney brought 10 muskets to this demonstration, at which he demonstrated that he could fit ten different lock mechanisms to the same musket with the use of a simple screwdriver. The audience included President John Adams, Vice President Thomas Jefferson (then President-elect), and many other distinguished government officials. Those present were said to be astonished.
This demonstration was so enthusiastically received that the near-bankrupt Whitney was advanced further funds to continue, and the deadline for delivery was extended once again. To put it simply, Whitney was saved from sure disaster through the intervention of high-level executive sponsorship.
Only later, in the 1960s, did curious historians discover that the demonstration was carefully staged and that the muskets were all hand-built to assure that the demonstration could be carried out. Sound familiar? "In short, it appears that Whitney purposely duped government authorities in 1801 and afterwards encouraged the notion that he had successfully developed a system for producing uniform parts" (3).
Having promised to deliver a total of 4,000 muskets to the government by September 30, 1800, Whitney delivers the final set of muskets nine years late. The muskets as delivered, in fact, contained no interchangeable parts, and the cost to the government exceeded the original contract amount many times over. None-the-less, Whitney's Mill Rock (CT) Gun Manufactory continued to receive awards for more weapons up until his death in 1825, and Whitney has always been held in the highest esteem by historians of technology. One historian refers to Whitney as a "song and dance man" (4). But all historians agree that the promotion of the ideas of interchangeable parts by Whitney, even when he himself was never able to achieve the goal, was a key element on the eventual success of this innovation. In other words, a failed champion is better than none--particularly if he is a showman.
A protege of Whitney and returning veteran of the War of 1812, Roswell Lee was made Superintendent of the Springfield Armory. At that time, there were two government owned and operated armories, one located in Springfield and the other at Harper's Ferry, VA. In addition, outside contractors such as Whitney also produced weapons for the Army's Ordinance Department.
The long term significance of this event is that Roswell Lee proved to be a patient, diligent, and hard-working advocate for principles of "uniformity in manufacturing." In fact, one might say that the race had truly been entered for the first time. Lee never gave up on the dream, and he promoted the notions of uniformity through a variety of mechanisms. Pushed by the first chief of ordnance, Colonel Decius Wadsworth, Lee and the Superintendent of the Harper's Ferry Armory, James Stubblefield, maintained constant interchanges of ideas, visits, evaluations of performance, etc. Each also worked with outside contractors to promote uniformity.
One might say that while the value of this innovation was itself easy to conceive, the implementation of the vision required much hard work with the active exchange of ideas. Many intermediate steps needed to be devised and tested before implementation success could be assured. Smith comments (5):
"Nowhere was the emphasis on communication and cooperation stronger than in the small arms industry. There, as indicated earlier, private contractors as well as federal administrators regularly corresponded, visited and assisted one another in working out the basic configurations of the uniformity system. To insure that the diffusion process would not be hindered, the Ordinance Department also insisted that the national armories open their shops to visitors, who could make drawings, borrow patterns, and other information pertinent to their special interests. At the same time, the department had an implicit understanding with all arms contractors that they had to share their inventions with the national armories on a royalty-free basis if they wished to continue in government service."
The next decade saw both evolutionary and revolutionary change in manufacturing methods, tools and technologies, all of which taken together lead finally to the production of rifles made with truly interchangeable parts in 1826. However, Roswell Lee was not the innovator who achieved this success. That honor goes to John Hall who worked at Harper's Ferry. Hall headed the armory's Rifle Works, a semiautonomous adjunct to the main shops. Hall had invented a breach-loaded rifle in 1811, and after eight years had finally persuaded the government to award him a contract to build 1,000 of these rifles at Harper's Ferry using manufacturing techniques including interchangeable parts. Hall set up an entirely separate operation in 1822, and over the next four years, first built the manufacturing machinery, then built the rifles.
At this point in our story, we now know from direct evidence that the principle of interchangeable parts can and will work. Hall delivered not just the first 1000 rifles, but the means of production which could be applied to manufacture thousands more, as well as untold other manufactured goods we take for granted in our everyday life.
Is this the end of our story? Hardly. We must look ahead for at least two more decades before Hall's success is transferred outside the Harper's Ferry Armory to civilian contractors. We must look even further into the future before we see evidence that this innovation is being applied widely in manufacturing.
Hall's success at Harper's Ferry was widely acclaimed. A special Commission was appointed to investigate and examine this innovation and to report to the U. S. House of Representatives. James Carrington, head of the investigation, reported that "We would, however, further observe, that in point of accuracy, the quality of the work is greatly superior to anything we have ever seen or expected to see, in the manufacture of small arms..." (6).
Eight years later a private contractor, Simeon Hall, was able to use the methods, tools, and technologies developed by Hall to produce parts in his factory which could be used interchangeably with those manufactured by Hall. This completes the technology circle. Now parts from multiple sources could be used to assemble rifles. These same parts could be used for field repair of any rifle manufactured at any location.
By the mid 1840s Hall's innovations had been spread adequately so that it was possible for both the Model 1841 rifle and Model 1841 musket to be manufactured by both the national armories and private contractors. These were the first fully interchangeable firearms, and apparently, the first fully interchangeable manufactured goods of any kind.
The first manufactured goods using fully interchangeable parts were U. S. Army rifles and muskets. This was not a *necessary* condition for the success of this technological innovation, but was rather the result of historical circumstance. It can be persuasively argued, however, that the innovation is unlikely to have been diffused as rapidly without this setting. Other manufactured products of the day such as steam engines, sewing machines, or textile equipment could have been built using this innovation, but as the historical time line above demonstrates, even with the full support of all interested parties it was not a simple or, initially, cost-effective innovation.
Much patience, effort, and cooperation was required to see this innovation implemented. Would that same level of patience be evident in other settings where the "deep pockets" of the U. S. government were not available? Most historians argue, no, that this innovation would have languished had it not been for the peculiar circumstances found in that era. On the other hand, there are some circumstances surrounding this innovation which fit many:
If we look at the adoption of innovations in our discipline (software development) we can find analogies to this story. Let's look, for example, at the SEI CMM and see if we can find analogies.
It might be instructive to cite a counter-example, OO technology.
What is instructive about this historical example is that we can see that all of the historical circumstances which lead to the introduction of interchangeable parts are not *necessary* to insure success. On the other hand, this experience also highlights some of the fundamental difficulties all innovations face, and further, help us to see some of the problems we face as change agents.
There are many more aspects to this story which are not explored here. For example, to build interchangeable parts requires a manufacturing process which will assure that precision tolerances are maintained. The development of gauges to inspect tolerances was a supporting technology which was widely diffused for many purposes beyond this application. The work performed by workers in this type of factory was entirely different from that performed in earlier small arms factories. As Cox commented, the workers themselves had little to do or say about this transition. Precision machining of the parts required development of machining processes which were innovative. Like the use of gauges, these innovation were then widely applied to many factory settings.
One overwhelming conclusion remains, however. The diffusion of innovation is neither assured or rapid. Certain circumstances can accelerate the process, and change agents need to be sensative to these circumstances and encourage them to effect the changes sought.