Category Archimedes

Between 1891 and 1896, Otto Lilienthal in Germany made over 2000 successful glider flights. His work was timed perfectly with the rise of photography and the printing industry. In 1871 the dry-plate negative was invented, which by 1890 could freeze a moving object without a blur. Also, the successful halftone method of printing had been developed. As a result, photos of Lilienthal’s flights were widely distributed, and his exploits frequently described in periodicals throughout Europe and the US.

These flights caught the attention of Nikolay Joukowski (Zhukovsky) in Russia. Joukowski was head of the Department of Mechanics at Moscow University when he visited Lilienthal in Berlin in 1895. Very impressed with what he saw, Joukowski bought a glider from Lilienthal, one of only eight that Lilienthal ever managed to sell to the public. Joukowski took this glider back to his colleagues and students in Moscow, put it on display, and vigorously examined it. This is the first time that a university-educated mathematician and scientist, and especially one of some repute, had become closely connected with a real flying machine, literally getting his hands on such a machine. Joukowski did not stop there. He was now motivated about flight – he had actually seen Lilienthal flying. The idea of getting up in the air was no longer so fanciful – it was real. With that, Joukowski turned his scholarly attention to the examination of the dynamics and aerodynamics of flight on a theoretical, mathematical basis. In particular, he directed his efforts towards the calculation of lift. He envisioned bound vortices fixed to the surface of the airfoil along with the resulting circulation that somehow must be related to the lifting action of the airfoil.

Finally, in 1906 he published two notes, one in Russian and the other in French, in two rather obscure Russian journals. In these notes he derived and used the following relation for the calculation of lift (per unit span) for an airfoil:

L = pVT

where L is the lift, p is the air density, V is the velocity of the air relative to the airfoil, and Г is the circulation, a technically-defined quantity equal to the line integral of the flow velocity taken around any closed curve encompassing the airfoil (Circulation has physical significance as well. The streamline flow over an airfoil can be visualized as the superposition of a uniform freestream flow and a circulatory flow; this circulatory flow component is the circulation. Figure 1 is a schematic illustrating the concept of circulation.) With this equation, Joukowski revolutionized theoretical aerodynamics. For the first time it allowed the calculation of lift on an airfoil with mathematical exactness. This equation has come down through the twentieth-century labeled as the Kutta-Joukowski Theorem. It is still taught today in university-level aerodynamics courses, and is still used to calculate lift for airfoils in low-speed flows.

The label of this theorem is shared with the name of Wilhelm Kutta, who wrote a Ph. D. dissertation on the subject of aerodynamic lift in 1902 at the University of Munich. Like Joukowski, Kutta was motivated by the flying success of Lilienthal. In particular, Kutta knew that Lilienthal had used a cambered airfoil for his gliders, and that, when cambered airfoils were put at a zero angle of attack to the freestream, positive lift was still produced. This lift generation at zero angle of attack was counter-intuitive to many mathematicians and scientists at that time, but experimental data unmistakenly showed it to be a fact. Such a mystery made the theoretical calculation of lift on a cambered airfoil an excellent research topic at the time – one that Kutta readily took on. By the time he finished his dissertation in 1902, Kutta had made the first mathematical calculations of lift on cambered airfoils. Kutta’s results were derived without recourse to the concept of circulation.

Only after Joukowski published his equation in 1906 did Kutta show in hindsight that the essence of the equation was buried in his 1902 dissertation. For this reason, the equation bears the name, the Kutta-Joukowski Theorem.

This equation became the quantitative basis for the circulation theory of lift. For the first time a mathematical and scientific understanding of the generation of lift was obtained. The development of the circulation theory of lift was the first major element of the evolution of aerodynamics in the twentieth century, and it was in the realm of science. The objective of Kutta and Joukowski – both part of the academic community – was understanding the nature of lift, and obtaining some quantitative ability to predict lift. Their work was not motivated, at least at first, by the desire to design a wing or airfoil. Indeed, by 1906 wings and airfoils had already been designed and were actually flying on piloted machines, and these designs were accomplished without the benefit of science. The circulation theory of lift was created after the fact.

Contemporary with the advent of the circulation theory of lift was an equally if not more important intellectual breakthrough in the understanding and prediction of aerodynamic drag. The main concern about the prediction of lift on a body inclined at some angle to a flow surfaced in the nineteenth-century, beginning with George Cayley’s concept of generating a sustained force on a fixed wing. In contrast, concern over drag goes all the way back to ancient Greek science. The retarding force on a projectile hurtling through the air has been a major concern for millenniums. Therefore, it is somewhat ironic that the breakthroughs in the theoretical prediction of both drag and lift came at almost precisely the same time, independent of how long the two problems had been investigated.

What allowed the breakthrough in drag was the origin of the concept of the boundary layer. In 1904, a young German engineer who had just accepted the position as professor of applied mechanics at Gottingen University, gave a paper at the Third International Mathematical Congress at Heidelberg that was to revolutionize aerodynamics.7 Only eight pages long, it was to prove to be one of the most important fluid dynamics papers in history. In it, Prandtl described the following concept. He theorized that the effect of friction was to cause the fluid immediately adjacent to the surface to stick to the surface, and that the effect of friction was felt only in the near vicinity of the surface, i. e., within a thin region which he called the boundary layer. Outside the boundary layer, the flow was essentially uninfluenced by friction, i. e., it was the inviscid, potential flow that had been studied for the past two centuries. This conceptual division of the flow around a body into two regions, the thin viscous boundary layer adjacent to the body’s surface, and the inviscid, potential flow external to the boundary layer (as shown in Figure 2), suddenly made the theoretical analysis of the flow much more tractable. Prandtl explained how skin friction at the surface could be fundamentally understood and calculated. He also showed how the boundary layer concept explained the occurrence of flow separation from the body surface – a vital concept in the overall understanding of drag. Since 1904, many aerodynamicists have spent their lives studying boundary-layer phenomena – it is still a viable area of research

today. This author dares to suggest that PrandtTs boundary layer concept was a contribution to science of Nobel prize stature. Perhaps one of the best accolades for Prandtl’s paper was given by the noted fluid dynamicist Sydney Goldstein who was moved to state in 1969 that: “The paper will certainly prove to be one of the most extraordinary papers of this century, and probably of many centuries.”8

As in the case of Kutta and Joukowski, Prandtl was a respected member of the academic community, and with the boundary layer concept he made a substantial scientific contribution to aerodynamics. This was science; the boundary layer concept was an intellectual model with which Prandtl explained some of the fundamental aspects of a viscous flow. However, within a few years this concept was being applied to the calculations of drag on simple bodies by some of PrandtTs students at Gottingen, and by the 1920s, research on boundary layers had become focused on acquiring knowledge for the specific purpose of drag calculations on airfoils, wings, and complete airplanes. That is, boundary layer theory became more of an engineering science.

In retrospect the beginning of the twentieth-century was the time of major technological breakthroughs in theoretical aerodynamics. These events heralded another breakthrough – one of almost a sociological nature. Wilhelm Kutta, Nikolay Joukowski, Ludwig Prandtl were all university-educated with Ph. D.s in the mathematical, physical, and/or engineering sciences and all conducted aerodynamic research focused directly on the understanding of heavier-than-air flight. This represents the first time when very respected academicians embraced the flying machine; indeed, the research challenges associated with such machines absolutely dictated the direction of their research. Kutta, Joukowski, and Prandtl were very much taken by the airplane. What a contrast with the prior century, when respected academicians essentially eschewed any association with flying machines, thus

causing a huge technology transfer gap between nineteenth-century science and the advancement of powered flight.

What made the difference? The answer rests in that of another question, namely, who made the difference? The answer is Lilienthal and the Wright brothers. Otto LilienthaTs successful glider flights were visual evidence of the impending success of manned flight; we have seen how the interest of both Kutta and Joukowski was motivated by watching Lilienthal winging through the air, as seen either via photographs or by actual observation. And when the news of Wilbur’s and Orville’s success with the Wright Flyer in 1903 gradually became known, there was no longer any doubt that the flying machine was a reality. Suddenly, work on aeronautics was no longer viewed as the realm of misguided dreamers and madmen; rather it opened the floodgates to a new world of research problems, to which twentieth-century academicians have flocked. After this, the technology transfer gap, in the sense that occurred over the previous centuries, began to grow smaller.

On 12 November 1975, ten lawyers from nine different law firms appeared in U. S. District Court, Southern District of New York. They represented twenty clients – nineteen of the largest aerospace firms in the United States and a curious legal and business entity known as the Manufacturers Aircraft Association, Inc. (MAA). All the aerospace firms were members of the MAA; some had been members since the MAA was founded in 1917. All ten lawyers agreed with the court’s finding that the MAA violated Section One of the Sherman Anti-Trust Act of 1890. The MAA was, in short, “a contract, combination… or conspiracy in restraint of trade or commerce.”1 On behalf of their clients, the assembled lawyers agreed to “wind up the affairs and terminate the existence” of the MAA.2 They further agreed to “terminate and cancel the Amended Cross License Agreement,” the legal instrument defining the purpose and operation of the MAA.3

The consent decree captures none of the historical irony hanging over this decision. The federal government had directed aircraft manufacturers in 1917 to enter into a cross-licensing agreement and to form the MAA to administer the agreement. In spite of protests at the time and repeated challenges in the 1920s and 1930s, the Justice Department consistently found that the cross-licensing agreement did not violate the Sherman Anti-Trust Act. In 1972, however, that same Justice Department brought suit in District Court, arguing, in effect, that it had been mistaken for 55 years. During that time, the United States aircraft manufacturing industry had been arguably the most successful in the world, dominating a market in which other nations, beyond the reach of the MAA, were free to compete. Half the companies represented in District Court had joined the MAA since its founding, a membership pattern suggesting openness and inclusivity, not combination and conspiracy. The MAA had been good for its members and good for America.4

Still more ironically, termination of the cross-licensing agreement had no discernible impact on American aircraft development. The MAA went out of business in 1975. All the patents licensed by the MAA and controlled by its members were made available to any applicant. The Court arranged to adjudicate disputed royalties arising from the new dispensation. Yet American aircraft manufacturers went right on dominating the free world market for this product, just as they had done under the protection of the cross-licensing agreement. In fact, the termination of the MAA coincided with the introduction of the European Airbus,

323

P Galison and A. Roland (eds.), Atmospheric Flight in the Twentieth Century, 323-345 © 2000 Kluwer Academic Publishers.

which, in the words of author John Newhouse, appeared at first “to be an even more dismal failure than most of Europe’s other jet transports had been.”5 The Airbus went on to be more competitive, but the United States still dominates the world market for commercial airliners more than twenty years after the dissolution of the MAA.

The many ironies of this case have attracted scholars of patent law, economists, and even sociologists.6 They have not, however, stimulated much study by historians of technology, not even historians of aviation. In most histories of American aviation, patents are noted by their absence.7 Even in my own study of the National Advisory Committee for Aeronautics (NACA), patents find little place after the cross-licensing agreement of 1917, which the NACA brokered.8 The NACA formed a patents committee in 1917, but discharged it with thanks when the cross-licensing agreement was signed. The Air Commerce Act of 1926 directed the NACA to form a patents committee, but the following year the NACA converted it to a committee on “Aeronautical Inventions and Design.” Congress may have thought of aviation development in terms of patents, but the NACA did not. Even in the 1930s, when the NACA found itself under congressional pressure to be of greater service to industry, it was proprietary information, not patents, that proved the sticking point.

The importance of aircraft patents to economists and legal scholars and their apparent irrelevance to aviation historians raises several nagging questions. Have historians simply overlooked the importance of patents in aviation history? If, as historians seem to believe, patents were not important, why was there a patent pool? And why did the Justice Department believe that the patent pool restrained trade? Finally, if patents do not shape an innovative industry, such as aircraft manufacture, what do they achieve? And what, then, does drive innovation in aircraft manufacture?

This paper will attempt to answer those questions. It will first summarize the history of aircraft patents in the United States. Then it will explore the theory of patents and the application of that theory to this particular case. Next it will seek the reasons for the success of the American aircraft industry, looking especially for ways in which patents might have played a role. In conclusion, it will attempt to explain the invisibility of patents in previous accounts.

In this and the subsequent sections, specific examples of important advances in aerodynamics will be examined, primarily from the point of view of the relative roles of science, engineering science, and engineering. The state of the art in aerodynamics has grown exponentially since the turn of the century; we can not do justice to the whole story in this limited paper. Instead, only a few specific examples from each era will be considered. In the present section, we examine some developments in aerodynamics contemporary with the hey-day of the strut-and-wire biplane exemplified by the British S. E.5 from World War I (Figure 3).

With Wilbur Wright’s flying demonstrations in Europe, which began on August 8, 1908, the world truly discovered the existence of the successful airplane. With this “discovery”, the attitudes surrounding the value of scientific and engineering work in aerodynamics changed radically. Almost overnight it became fashionable, indeed critical, to learn more about the laws of nature that sustained these flying machines in the air, and to develop engineering techniques that could lead to improved aerodynamic design. We have already discussed how academic science met the flying machine at the turn of the century. Now, with Wilbur’s dramatic demonstration that the airplane was indeed an established reality, the world of professional engineering suddenly had a new and very exciting discipline to develop.

This technical awakening was accompanied by the sound of new wind tunnels revving up throughout Europe. Nowhere was this as dramatic as in the shadow of the Eiffel Tower in Paris. In 1909, Gustav Eiffel designed and built a large wind tunnel on the Champ de Mars adjacent to the famous tower he had erected 20 years earlier. Indeed, Eiffel was soon to become France’s first great aerodynamicist on the strength of his wind tunnel experiments. Today, the name of Eiffel rarely crosses the lips of practitioners and students of aerodynamics. In fact, most people in general do not associate Eiffel’s name with aerodynamics at all. However, his contributions to

experimental aerodynamics were as important in the history of technology as were his structural innovations embodied in the design and construction of the Eiffel tower. At the beginning of the twentieth century, Eiffel pioneered some of the experimental techniques which we still use today, and in the process he was the first to quantitatively measure some of the most basic aerodynamic aspects of a complete airplane configuration.

The wind tunnel experiments at the Champ-de-Mars laboratory conducted during 1909 and 1910 led to five substantial contributions:

1. First, there was the wind tunnel itself, an innovative design using a free jet in a hermetically-sealed chamber. This was pure engineering.

2. Eiffel found that drag measurements made in his wind tunnel agreed with earlier measurements he made by dropping aerodynamic shapes from his Tower. Eiffel finally proved experimentally once-and-for-all the basic principle of the wind tunnel that had been first stated by Leonardo da Vinci more than four centuries earlier, namely that “the same force as is made by the thing against the air, is made by air against the thing.” Some doubt about this persisted until the twentieth century, even though wind tunnels had been in use since their invention by Francis Wenham in 1871. Eiffel proved the validity of the wind tunnel principle. This was science.

3. He was the first to make detailed measurements of the distribution of pressure over the surface of an aerodynamic body, proving conclusively that the aerodynamic lift on a wing was due to the presence of lower pressure on the top surface and higher pressure on the bottom surface. Moreover, he proved conclusively that the majority of the lift on a wing is derived not from the higher pressure exerted on the bottom of the wing, but rather from the lower pressure exerted on the top of the wing. This was engineering science.

4. Eiffel pioneered the general principle that the net resultant aerodynamic lift on a body is due to the integrated effect of the pressure distribution exerted over the surface, and he was the first to prove it. Using his measured pressure distributions on one hand, and his direct measurements of lift using a force balance on the other, he was able to state: “The direct measurement of pressure has given us a result to which we attach great importance; viz., the summation of the observed pressures was equal in every case to the reaction weighed on the balance.9 This was engineering science.

5. Eiffel was the first to conduct wind tunnel tests using models of complete airplanes, and to show conclusively the correspondence between such tests and the performance of the real airplane in actual flight. This was pure aeronautical engineering.

Perhaps the most important contribution Eiffel made to aerodynamics was as follows. Over the course of his earlier aerodynamic work, Eiffel had measured the drag of spheres. However, in experiments at Prandtl’s laboratory at Gottingen, the data showed sphere drag to be more than twice as large as Eiffel’s measurement. Eiffel was angered by a German insinuation that he had made a mistake, and in 1914 he carried out a definitive series of drag measurements on spheres in a new laboratory at Auteuil, a suburb of Paris. Testing spheres of various sizes, he found out that for each size, there was a velocity above which the drag decreased markedly – by slightly more than a factor of two. Every student of aerodynamics today recognizes this variation, and knows that the sudden decrease in drag is associated with a transition of the flow inside the boundary layer from laminar to turbulent at a value of the Reynolds number of about 300,000. Prandtl was the person who eventually explained why this phenomenon occurs. But Eiffel was the person to first observe and publish the phenomenon. This was a purely scientific contribution.

The period during and just after World War I saw major advances in the theoretical calculation of airfoil and wing aerodynamics. Based on the circulation theory of lift, Prandtl conceived a theoretical model of the aerodynamic properties for a finite wing (a real wing with wing tips, in contrast to the two-dimensional aspects of an airfoil shape). Labeled “Prandtl’s lifting line theory”, this model allowed the calculation of lift and induced drag (a pressure drag due to the influence of vortices generated at the wind tips and trailing edge of the wing). This is a rational, engineering-oriented theory that could be applied to the wings of real airplanes. It is still used today. In a similar vein, one of Prandtl’s colleagues, Max Munk, after immigrating to the United States after the war and going to work for the National Advisory Committee for Aeronautics (the NACA), derived the first practical theory for the calculation of the lift of airfoils of any arbitrary shape, as long as the airfoils are relatively thin. Prandtl’s lifting line theory for finite wings, combined with Munk’s thin airfoil theory, represented major contributions to applied aerodynamics. This was one of the first important examples of engineering science in the twentieth century. However, again emphasis is made that these theories were after-the-fact; airfoils and wings, albeit not optimum, were being designed successfully based on empirical experience and routine wind tunnel testing long before these theoretical tools became available.

The story began where it ended, in court. On 13 January 1914, the United States Circuit Court of Appeals ruled that Glenn Curtiss and his colleagues had infringed the basic patent of Wilbur and Orville Wright.9 The implications for the nascent American aircraft industry were profound. If Curtiss, a prolific inventor and tireless entrepreneur, could not break the Wright patent, no one could. Curtiss himself and his associates could fall back on their own stock of pioneering patents, mostly in seaplanes. But other builders faced the prospect of crippling royalty payments to both Wright and Curtiss. The conflict reached crisis level in December 1916, when the Wright-Martin Company, holder of the Wright patent, announced that manufacturers would have to pay a royalty of 5% on each aircraft sold, with a minimum annual payment of $10,000 per manufacturer.10

This crisis sprang from a patent phenomenon that is rare but not unprecedented. The Wrights held a foundational patent that the courts interpreted broadly. To understand how they got that patent and why the courts might be inclined to construe it generously, it is necessary to briefly retrace their steps.

In 1899, when Wilbur and Orville Wright entered the race to fly, they discovered a crowded field. Most inventors around the world were taking the same path to heavier-than-air flight.11 They were trying to move airfoils through the atmosphere fast enough to generate lift greater than the combined weight of their airframe, engine, and pilot. None had yet overcome the fundamental dilemma: simply enlarging the engine added power and speed, but at the cost of more weight.

The Wrights embraced a different model. They emulated Otto Lilienthal, a German inventor who was conducting glider experiments (what would now be called hang gliding) in an attempt to learn how to actually fly. Like the Wrights who followed him, he was more concerned with what to do aloft than how to get there. The Wrights’ first round of experiments revealed that most other researchers were on the wrong track, producing flawed data. Thus, the Wrights began at the beginning. They built their own glider and learned to control it in flight by applying lessons from bird-watching. Only then did they turn to propulsion. They designed their own engine and their own propellers, relying on data from their own wind tunnel. The resulting airplane, the one that first achieved sustained, powered, human flight in 1903, was a product of their native genius and no small measure of what Alfred North Whitehead would call scientific method.12 They had broken a large problem down into component parts and had solved them one after another using observation, experiment, and theory.

Recognition of the Wright achievement came slowly. The world at first paid little attention to the 1903 flight at Kitty Hawk, NC, and to the follow-on flights in Dayton, Ohio. The brothers filed their first patent application in 1903, but received their foundational patent only in 1906.13 Still, that was early enough to preempt the field. They claimed that the method they had developed for controlling the airplane in flight – wing-warping, which created differential lift on opposite sides by actually twisting the wings – was fundamental to all flight. The courts agreed. Even when Glenn Curtiss introduced an airplane with a superior method of lateral control, the ailerons still used on aircraft today, the courts refused to budge from their commitment to the Wrights, for the Wrights had specifically anticipated ailerons in their patent. Curtiss had behind him the resources and prestige of Alexander Graham Bell and the Smithsonian Institution, both of whom had reason to challenge the Wright stranglehold. But nothing would avail. In case after case, the courts held for the Wrights, leading up to the climactic decision of 1914.

By this time, the matter of aviation development was slipping from the hands of lone inventors and into the board rooms of nascent manufacturers. Orville Wright, demoralized after his brother’s premature death in 1912, sold his interests in 1915 to an entity called the Wright Company, soon to become Wright-Martin Corporation. Curtiss had his own company and thirty important patents on seaplanes; his patents stood in spite of the fact that all of his airplanes infringed on the Wright master patent. With the outbreak of World War I in Europe and the prospect of lucrative government contracts, other aircraft manufactures were attracting capital and coming under the sway of major investors. An Aircraft Manufacturers Association formed in February 1917, but its prospects were dimmed by the refusal of Wright-Martin to join. Jacob Vander Meulen and others have seen in the Aircraft Manufacturers Association an attempt to bring Progressive-era associationalism to aviation. It was, in short, a bid for the middle ground between “the chaos of laissez faire… and the tyrannies of business monopoly and statism.”14 Manufacturers that found such an association congenial might also be expected to embrace patent pooling, a technique that navigated between the treacherous shoals of commercial pirating on the one hand and monopoly extortion on the other.15

Serious movement toward a patent pool, however, was not likely to issue from such voluntarism, at least not when a critical player such as Wright-Martin refused to participate. Instead, government had to intervene forcefully to impose an associationalist model on the entire community.16 Responding to the Wright-Martin royalty claim of December 1916, the Army and Navy both asked the NACA to resolve the aircraft patent stalemate. At the Committee’s request, Congress appropriated $1,000,000 to buy out the existing patents.17 With that leverage in hand, the NACA on 22 March 1917 convened a meeting of its newly-formed patent committee and representatives of the major aircraft manufacturers. They agreed to a cross-licensing agreement that would ultimately pay the Wright and Curtiss interests $1 million each, twice what Congress had appropriated.

Before the agreement could be consummated, the United States entered World War I. Realizing that orders for aircraft would skyrocket, the principals held out for greater royalties. A second round of negotiation ensued, this one including W. Benton Crisp. He was a lawyer for the Curtiss interests who, not coincidentally, had brokered a comparable cross-licensing agreement in the automobile industry that had gotten Henry Ford around the foundational Selden patent. With Crisp’s help, a new agreement was reached that doubled the payments to the Wright and Curtiss interests. The NACA committee on patents submitted its report on 12 July 1917 and the cross-licensing agreement was signed twelve days later.

The specific terms of the agreement warrant close scrutiny.18 All aircraft manufacturers who were party to the agreement were to pay to the MAA $200 for every aircraft they produced, plus royalties to be determined on licenses issued after the agreement was adopted. Of that $200, $175 would be distributed between Wright-Martin and Curtiss until they had each received $2 million on their foundational patents. The remaining $25 per airplane went to administrative costs of the MAA. A Board of Arbitration within the MAA would determine the royalties to be paid on future patents. Significantly, “engines and their accessories” were excluded from the agreement.19

Eight companies signed the cross-licensing agreement on 14 July 1917. They were Aeromarine Plane and Motor Company, Burgess Company, Curtiss Airplane and Motor Corporation, L. W.F. Engineering, Standard Aero Corporation, Sturtevant Aeroplane Company, Thomas-Morse Aircraft, and Wright-Martin Aircraft Corporation.20 These were also the charter members of the Aircraft Manufacturers Association, which had formed on 13 February. Day ton-Wright joined the cross­licensing agreement a few months later. By this time more than 283 companies had formed in the United States to produce aircraft or components other than engines.21 How many were still viable in 1917 and how many individual, unincorporated concerns were manufacturing aircraft is impossible to determine. It is nonetheless clear that these charter eight were the cream of the crop in 1917. They made up less than half of the group that attended the meeting of the AMA in Washington on 22 March 1917.22

“Aircraft trust,” howled critical newspapers when the agreement was announced.23 A patent attorney who protested on behalf of the American Aeronautical Society later described the agreement to Congress as “a most pernicious undertaking, detrimental to the interests of inventors and independent manufacturers, operating to stifle invention and thereby to retard the development of the art of aviation.”24 Another Congressional witness reported in 1918 that the MAA was “condemned by every airplane manufacturer outside the beneficiaries.”25

Suspicions about the agreement followed two tracks. One group protested that the major manufacturers were conspiring to drive out the small operations. Another group saw a “Detroit conspiracy,” an attempt by the automobile industry to comer the wartime market for engine manufacture and parley that into a foothold in airframe manufacturing as well.26 The latter suspicion was fed by the merger in 1916 of the American Society of Aeronautical Engineers and the Society for Automotive Engineers,27 by the role of Crisp in transferring his experience from the Selden patent fight, and by the subsequent role of Detroit manufacturers in wartime production.

At the request of the Secretary of War, the Comptroller of the United States reviewed the cross-licensing agreement to determine if it violated the anti-trust laws. The Comptroller certified that it did not.28 This satisfied the government and allowed the aircraft manufactures to get on with the business of producing planes for the war. That production, however, proved disappointing, by any standards. The government spent over $1 billion for aircraft in World War I but only 960 aircraft reached the front, not a single fighter among them.29 The reasons for this dismal performance had more to do with corruption and mismanagement than with patents. Nevertheless, “aircraft trust” became a familiar refrain in what NACA Chairman Charles D. Walcott called the “hymn of hate” spewing from critics of the establishment after World War I.30

By the middle of the 1920s, other voices had joined the chorus. Billy Mitchell and the advocates of a separate air force held the “air trust” partly responsible for the inadequacies of military aviation in the United States. Advocates of commercial operation believed the government was providing insufficient support for the nascent industry. Some observers believed that air safety required government regulation. And other observers believed that the government should transfer air mail operations to the private sector. Congressional critics of government policy formed the Lampert Committee in 1924 to study this array of problems and make constructive suggestions. Distrustful of the politics of this group, the Coolidge Administration appointed its own Morrow Board to make an independent appraisal. The battle of the boards produced a political stalemate, but a legislative landslide. In short order Congress passed and President Coolidge signed the Air Mail Act of 1925, the Air Commerce Act of 1926, and the Air Corps Act of 1926.

Among them, these new laws established the legal structure that would carry American aviation to the Second World War. All of them recognized the need for government support, but none of them explicitly endorsed the belief that an “air trust” existed or that the patent pool was an illegal combination in restraint of trade. Quite the contrary. An amendment of the cross-licensing agreement dated 31 December 1928 essentially validated the original pact. It, and two supplementary agreements which had preceded it in 1918 and 1923, changed the proportion of MAA income that went to the association and the two major patent holders, but it did not substantially alter the fundamental deal struck in 1917.31 The 1928 agreement was the one overturned in 1975.

By the time the cross-licensing agreement was amended in 1928, American aviation had turned a comer. Charles Lindbergh had crossed the Atlantic Ocean in May 1927, claiming the Orteig Prize first offered in 1919 and more importantly stimulating a buying spree of aviation stock. Before the market crashed in 1929, new purchases had pushed aviation securities over the $1 billion mark.32 Approximately 95% of that value disappeared in the early years of the Depression, but not before the wave of capital had financed an increase in infrastructure. With the new resources, aircraft manufacturers not only survived the Depression, they also created the so-called “airframe revolution.”

The all-metal, stressed skin, monocoque airliner with twin cowled engines and retractable landing gear was not exactly invented in the United States. The various innovative components were invented, sometimes simultaneously, in several countries including the United States. But American airframe manufacturers, especially Boeing and Douglas, combined these features in designs that transformed commercial aviation. Relying on the infrastructure purchased with the capital investment of the late 1920s, Boeing introduced the 247 in 1933 and Douglas followed with the revolutionary DC-2 and the classic DC-3 shortly thereafter. The DC-3 went on to be the most famous and most durable airliner of all time. Along with its imitators, it made possible the growth of a viable commercial airline industry in the United States.

The technical innovations that made these airplanes the best in the world came in part from the manufacturers themselves, but also increasingly from university and government laboratories. The Guggenheim Fund for the promotion of aeronautics, for example, supported laboratories at MIT, the California Institute of Technology, and other universities around the country.33 The National Advisory Committee for

Aeronautics supported what it called “fundamental research,” benefiting both the military services and the commercial industry.34 Patents contributed to this development, but the role of government especially made patents difficult to rely on. The government often insisted that developments supported by its funding be made freely available to all, or at least licensed for government purposes. Similarly, the NACA agreed to respect proprietary information whenever possible in conducting research for industry, but in its own work on the famous NACA cowling, for example, it declined to patent.35

The outbreak of World War II only accelerated these trends. Aircraft production went from less than 6,000 in 1939 to more than 96,000 in 1944.36 In some cases, designs from one manufacturer were handed over to another for production. Existing patents were licensed as necessary, not only to other aircraft manufacturers but to automobile manufacturers as well.37 As Robert Ferguson discovered in his study of aircraft manufacturing in World War II, patents and even proprietary knowledge hindered the flow of information but little. More intractable were the incompatibility of “technological cultures” in the different companies, “that is, the unique methods and traditions of practice for designing and producing aircraft that existed within each firm.”38 Some companies, for example, moved airframe components to the workers in assembly-line fashion, while others moved workers about the shop floor. Similarly, tooling varied greatly from company to company and existed beyond the reach of the cross-licensing agreement. Ferguson found that the MAA and the patent pool provided “an adequate legal and financial infrastructure” for the exchange of ideas, but new institutional arrangements were needed to provide an “organizational framework for sharing proprietary technologies.”39 A National Aircraft War Production Council was created to fill this need.

After World War II, the American aircraft industry did not return to its pre-war status. The creation of a separate air force, the elevation of strategic bombing to the first line of defense, the outbreak of the Cold War, the advent of rockets and missiles, and the commitment of the United States to a qualitative and quantitative arms race with the Soviet Union all combined to transform the aircraft manufacturing industry into the aerospace industry. Bulwark of the military – industrial complex and mainstay of the U. S. balance of payments, the aerospace industry grew to enormous proportions without a comparable growth in the number of companies. The dollar value of aircraft shipments doubled between 1958 and 1992, from $31.3 billion to $63.0 billion (both in 1992 dollars), while the number of companies rose in the same period only from 113 to 151 40 The trend was even more striking among major airframe manufacturers, where mergers in the 1990s of Lockheed Martin, Northrup Grumman and Boeing/McDonnell Douglas accelerated the concentration of the industry.

As aerospace business grew, those fewer and fewer manufacturers grew larger and larger. Their ideal business formula was some combination of military and commercial production. This allowed companies to conduct research for the government and then convert that knowledge into commercial products.41 Very often the research conducted for the military services set higher standards for quality and reliability than would have been warranted by market forces; the resulting technologies nevertheless found their way into commercial products. Patents played little role in this enterprise, in part because the government insisted on licensing the knowledge gained with its funds; in part because the success of the companies was predicated more on design, production, and marketing than it was on invention; and in part because the firms preferred the secrecy of proprietary knowledge to the disclosure of patenting. For whatever reason, the American formula worked. In 1975, when the MAA disappeared, the U. S. aerospace industry controlled 67% of the international market.42

Perhaps one of the best examples of an aircraft during the era of the mature propeller-driven airplane is the Douglas DC-3 (Figure 4) from the 1930s. The aerodynamic hallmark of this generation of airplanes is streamlining. Let us briefly examine some historical aspects of streamlining.

A clarion call for the advantages of streamlining was made by Sir Melvill Jones in 1929. Jones, a respected professor of aeronautical engineering at Cambridge University, gave a lecture to the Royal Aeronautical Society in 1929 entitled “The Streamline Airplane.”10 Jones’ engineering analysis of the drag reduction that could

be achieved by the streamlining was so compelling that, in the words of Miller and Sawers, “designers were shocked into greater awareness of the value of streamlining.”11 Jones’ paper was a turning point in the practice of aerodynamics during the age of the mature propeller-driven airplane. This work was pure aeronautical engineering carried out by a respected academic.

Two other milestones involved drag reduction during this era; they were the development of the NACA cowling, and the detailed drag clean-up work – both carried out by engineers at the NACA Langley Memorial Laboratory. The cowling, a streamlined metal shroud wrapped around the cylinders of a radial piston engine,

was developed by Fred Weick and colleagues through extensive parametric testing in the Propeller Research Tunnel at Langley in 1928. The results were dramatic – the drag for a fuselage with the engine wrapped in the NACA cowling was 60 percent less than that for the fuselage with the engine cylinders exposed to the air. The normally staid NACA knew they had something very important, and no time was lost in getting the data to industry. The Lockheed Vega was the first production airplane to incorporate the NACA cowling in 1929. The top speed of the Vega was increased from 165 mph for the uncowled version to 190 mph for the cowled version – spectacular success. The cowling was developed at that time without any understanding of the basic aerodynamic reasons for its success; the work was all based on parametric testing in the wind tunnel. The fundamental understanding of the flow associated with the cowling, as well as the derivation of a theoretical approach to calculating cowling performance, was finally achieved by Theodore Theodorsen in 1937. Theodorsen was the NACA’s leading theoretician at that time. Analytical understanding was finally achieved, but the NACA cowling had been in use for eight years prior to that. The development of the cowling, and its empirical application to airplanes was pure engineering by Fred Weich. Theodorsen’s analysis eight years later was an excellent example of engineering science.

Some of the aerodynamicist’s final touches in the quest towards Melville Jones’ ideal streamlined airplane took place in the late 1930s and early 1940s, when every effort was made to reduce or eliminate even the seemingly most innocuous sources of local flow separation (with its attendant pressure drag) on an airplane. A perfect example of this was the drag cleanup program at NACA Langley, starting in 1938 and lasting essentially through the end of World War II. Here, the whole airplane was mounted in the Langley 30 x 60 ft full-scale wind tunnel, and one-by-one various appendages and protrusions were removed, each time measuring the drag reduction. The reduction for each “fix” was usually small, but summed over 20 or more modifications, the fully streamlined airplane typically experienced a 40 to 50 percent reduction in drag. The NACA drag cleanup program was pure engineering.

Societies issue patents to promote the public welfare; they encourage individuals to innovate and they guarantee a reward when useful innovations are shared with society. The first patents in the Anglo-American legal system were granted in the 16th century to encourage foreign craftsmen to migrate to England and spread their knowledge through apprenticeships.43 Thereafter, patents served more often to promote invention. If individuals would benefit the commonwealth by developing new techniques and products, the state would reward them with a temporary monopoly on the sale or exploitation of their contribution. At the heart of all patent systems, therefore, is a tension between the public good (invention) and private gain (monopoly).44

The United States is no exception. Its first patent law appears in Article I, Section 8, of the Constitution, which gives Congress the power “to promote the Progress of Science and useful Arts, by securing for limited Times to Authors and Inventors the exclusive Right to their respective Writings and Discoveries.”45 Subsequent legislation, culminating in the Patent Act of 1836, established a national system of patent examination and registration. The essential features of a patent are novelty, invention, and utility.46 In practice this has meant that the patent applicant must demonstrate an unprecedented process or product that embodies concepts beyond mere technical skill. Criteria for utility are less demanding.

The inherent tension between public good and private profit remained tolerable through the nineteenth century and into the twentieth. In the 1920s and 1930s, however, this tension pitted American esteem for the individual against suspicion of corporations and the state. Inventors such as Thomas Edison were seen as quintessential American heroes; their patents were the fruits of their labors. Large corporations, however, appeared to subvert the American system and corrupt the political process. Slowly the impression dawned that corporations were abusing the patent system to control the marketplace, and that government was doing their bidding. The charges of trust and conspiracy leveled at the aircraft industry in the wake of the cross-licensing agreement and the debacle of World War I constituted but one of many public scandals that grew up in the 1920s and 1930s around large corporations in high-technology industries.47 The Nye Committee hearings of the mid-1930s, which attached the label “merchants of death” to aircraft manufacturers and other “war profiteers,” were followed in the late 1930s by hearings before the Temporary National Economic Committee, popularly known as the Monopoly Committee. This latter body repeated the widely-held belief that American prosperity was based on invention, and it noted with alarm that individuals had accounted for 72% of patents in 1921 but less than half in 1938.48

Law Professor Robert Merges believes that the American patent system fell under a darkening cloud in the 1920s and remained compromised for almost 60 years, just about the period when the cross-licensing agreement was in force.49 Though the cross-licensing agreement contributed little to the phenomenon, it nonetheless operated in this inimical environment. The height of the anti-patent movement, Merges believes, was the reformist era of the 1920s and 1930s. World War II brought some relief, and a post-war honeymoon produced a new patent act in 1952. But the anti-technology sentiment that he sees dominating American society in the 1960s and 1970s sparked a revival of anti-patent sentiment. During this half­century, the courts were less likely to hold patents valid,50 and even industry moved away from patent activity.

Merges’ periodization of American experience with patents casts the aircraft patent pool in a new light, suggesting that many of its problems were not peculiar to this industry but were rather part of a larger national ambivalence toward patents in general. Not until 1982, says Merges, when Congress passed the Federal Courts Improvement Act, creating a single federal appeals course for patent cases, did the situation improve. Patents are now more likely than previously to be held valid, money damages have risen dramatically, and injunctions against infringers are easier to win.

Whatever the politics of patents may have been in the era of the cross-licensing agreement, the still more important issue is whether or not patents worked. Did they, that is, promote invention? Were they useful? And were aircraft patents any different from patents in other industries? Were they different from industries that did not pool their patents?

Scholars disagree. Sociologist S. C. Gilfillan spent most of his adult life arguing that patents and inventions correlated poorly with each other. In 1935 he wrote that inventive activity was demonstrably increasing while the number of patents was decreasing.51 Almost thirty years later, in a book sponsored by the Joint Economic Committee of Congress, Gilfillan made the same claim, calling for reform of the patent system to return it to its original purpose of promoting invention.52

Other scholars, however, have found patent activity useful in attempting to measure the level of invention within a community. Jacob Schmookler, for example, argued that “for the pre-1940 period,… the behavior of patent statistics is consistent with what little external evidence exists as to the course of American inventive effort.”53 The basis for Schmookler’s opinion was a pair of articles he had written in the 1950s, attempting to correlate patenting and invention.54 Gilfillan singled out Schmookler for criticism in the summer 1960 issue of Technology and Culture, eliciting responses from Schmookler and I. Jordan Kunik, a patent lawyer.55 Kunik raised the novel argument that one could not expect a rise in patenting comparable to the increase in population because “proliferation of the population requires merely a marriage license” while patenting requires an idea that has never been patented before; the supply of children is endless but the supply of new inventions is, in his view, finite.56

The debate spilled over into the December 1960 meeting of the American Association for the Advancement of Science. In a panel sponsored by Section L of the AAAS, Gilfillan and Kunik were joined by experts from various fields and disciplines, most of whom viewed the patent system more favorably than Gilfillan. One panelist presented data purporting to demonstrate that in the chemical field “technology in those various disciplines stimulated by the patent system had advanced more rapidly than in those where the advantages of the patent system were either unavailable or were not broadly used.”57

This flurry of interest in the early 1960s, and the subsequent publication of Gilfillan’s Invention and the Patent System, produced little consensus. Historians of technology took up the matter again at the 1971 meeting of the Society for the History of Technology. Morgan B. Sherwood presented a paper entitled “Patent Nonsense in the History of Technology.” Employing rhetoric and some arguments reminiscent of Gilfillan, Sherwood argued that throughout American history the U. S. patent system had failed, as the panel chair put it, “to encourage technological progress, to reward inventive genius, and to benefit society.”58 The commentators all disagreed.

When the historical value of patents was again addressed in the pages of Technology and Culture, in a special issue on “Patents and Invention” in 1991, the contributors avoided the direct question of whether or not patents promote invention. Issue editor Carolyn Cooper reviewed the previous literature and cautioned that patents should not be used as a direct measure of inventive activity, though “patent records of various types can be valuable sources of information about particular inventions.”59 In sum, historians of technology and students of patent history are ambivalent about the explanatory power of patents. Most believe that patent records and statistics can be a useful source of information about technical development. At the same time, the best scholars caution against using patent activity as an index of invention. Their reticence suggests that aircraft patents may have been less closely related to aeronautical development than the friends and critics of the cross-licensing agreement believed.

Economic analysis is somewhat more positive, at least in the special category of “cumulative industries.”60 These are industries such as automobiles, aircraft, and computers in which fundamental, pioneering patents often control initial production. When they have run their course, the field experiences improvement patents, which are generally more difficult to win and less valuable to hold. Such industries may be contrasted with those based on discrete inventions, such as the safety razor and ballpoint pen, and the rarer fields of science-based technology, such as biotechnology. In the cumulative industries, the “broad basic patents” often have a blocking effect on commercial development and invite pooling, cross­licensing, or consolidation. This analysis suggests that the MAA was not an extraordinary intervention but rather a familiar response to a certain category of industrial patenting.

In spite of this strong endorsement of patent pooling in cumulative industries, the literature of pooling in general is ambivalent.61 Most authorities agree that there are good pools and bad pools. Most agree as well that the difference often turns on the openness of the pool. If pools accept new members under reasonable terms, then they are less likely to cross the line into monopolistic practice. Indeed, there is widespread agreement that pools can have important positive impact on a field or industry. For one thing, they can lower the transaction costs of individual licensing.62 A 1981 study found that transaction costs averaged more than $100,000 for the cases examined, and a 1976 investigation found that transfer costs accounted for more than 19% of total project costs in the international ventures studied.

But pools can also retard invention and competition.63 They smother competition in two ways. First, members of pools may be reluctant to purchase a patent from someone outside the pool, because to buy it is to share it; to abstain is without cost, for no other member of the pool will have it exclusively.64 Second, members of pools may be unwilling to develop new products, for they will have to share them with other members. Even though a pool such as the MAA offered a mechanism by which members could earn royalties for a patented invention, the royalty was determined by arbitration rather than by the market. In the MAA, total royalty payments in the first 16 years of the agreement amounted to $4,360,000. But $4,000,000 of this went to the Wright and Curtiss interests for their foundational patents. All the other patents combined earned a mere $360,000.65 That is small incentive for companies to invest in research and development with the intent of patenting or for outsiders to invent with expectation of selling to the major manufacturers.

The era of the jet-propelled airplane began during World War II, and is the era we live in today. A typical example of such an airplane from this era is the swept-wing F-86 fighter, shown in Figure 5.

One of the most pressing aerodynamic challenges during the early part of this era was the proper understanding of compressibility effects. Indeed, the myth of the “sound barrier” had materialized, wherein it was doubted that airplanes could ever fly faster than sound. The essence of the “sound barrier” was the dramatic increase in drag encountered by a body flying near the speed of sound. The gradual understanding of the physical nature of this large drag rise near Mach one is an excellent example of engineering science. The following is a brief synopsis of this story.

In 1920, Frank Caldwell and Elisha Fales at the Army Air Service Engineering Division at McCook Field in Dayton, Ohio, were the first to observe the large drag rise on airfoils beyond some “critical speed” in a high-speed wind tunnel. They did not have a clue as to what was causing it. Lyman Briggs and Hugh Dryden, working for the Bureau of Standards under a contract from the NACA, in 1926

discovered that these precipitous changes corresponded to the sudden separation of the flow over the airfoil surface. They did not have a clue as to what was causing the flow to separate. Then, in 1934, John Stack and Eastman Jacobs at NACA Langley made the first schlieren photographs of the high-speed flow over an airfoil, and observed the existence of a shock wave on the top surface of the airfoil beyond the critical speed. Suddenly, the pieces were in place. The shock wave was caused by a pocket of locally supersonic flow occurring on the airfoil beyond the critical speed. This shock wave caused the flow to separate from the surface at the point where the shock was touching the surface. The combined pressure effect due to the shock wave and the separated flow caused the drag to greatly increase. The acquisition of this physical knowledge later gave airplane designers some insight as to how to minimize the effect of transonic drag – divergence phenomena. This whole story, only briefly summarized above, is one of the most beautiful examples of the role of engineering science in the evolution of aerodynamics in the twentieth century.

The final case study we will mention here is the development of the swept wing for high-speed airplanes. The concept of the swept wing for such an application was first introduced by the German aerodynamicist Adolf Busemann at the 1935 Volta Congress in Rome. Even though the major leaders in high-speed aerodynamics were present for this meeting, Busemann’s concept went virtually unnoticed. However, by 1939 the Luftwaffe had classified the swept wing concept and was sponsoring research on its aerodynamic characteristics. Later, in 1945, Robert Jones, an extraordinary aerodynamicist at NACA Langley independently conceived the idea for a swept-wing. Many of Jones’ colleagues, especially Theodorsen, were skeptical of Jones’ idea. However, this skepticism quickly melted away when a large bulk of swept-wing data was found in Germany in 1945, and transported to the United States. Quickly thereafter, the swept wing was incorporated on the Boeing B-47 bomber and the North American F-86 fighter (Figure 5). The development of the swept wing concept is an example of engineering science.

The starting point for any discussion of technical advance in aeronautics is the landmark study of Ronald Miller and David Sawers, The Technical Development of Modern Aviation:66 One of the authors, David Sawers, had collaborated on an earlier and even more famous investigation, The Sources of Invention,67 In the latter work, Sawers and his co-authors had studied fifty cases in order to determine how industrial inventions arise in the modem world. A second edition added ten new case studies.68 Only two of the inventions, helicopters and the jet engine, were peripherally related to the aircraft manufacturing industry.69 Nonetheless, the general conclusions had wide applicability.

The study found a trend away from the lone inventor of previous ages to large, institutional sites. In these, creative genius counts for less than enlightened management. All are vulnerable to ossification as they grow large and old. Eclecticism usually triumphs over monolithic agendas and methodologies. There is no consistent relationship between monopoly and invention, but the patent system, for all its imperfections, remains important to the individual inventor or the small firm.

Miller and Sawers sought to test these conclusions in their investigation of aircraft development.70 What they found was an industry in which progress was not steady and incremental but rather episodic and pivotal. Writing in the late 1960s, amidst the U. S. policy debate over whether or not to proceed with development of a supersonic transport, Miller and Sawers found only ten aeronautical inventions after the Wrights and Curtiss to have been seminal. They are listed in slightly modified form in Table l.71 Others might argue for a longer list, but this one serves the purposes of this paper.

It also provides a chronology that helps to understand how and when aeronautics made its greatest advances. The Germans, say Miller and Sawers, made the greatest technical contributions. Partly for that reason, Europe led the United States in aircraft development until the late 1920s. Following Lindbergh’s flight, however, American aviation caught up quickly. The airframe revolution of the early 1930s, much of it based on German innovations, catapulted the United States into the lead in commercial airliners. While the Europeans devoted much of their attention to the development of fighter aircraft between the wars, the United States focused instead on bombers and long-range airliners, both powered by air-cooled engines and both placing a premium on range and payload. Depending in part on government-funded research for the military, the American aircraft industry achieved a 33% cost reduction in airliner operations between 1927 and 1933.72 The experience was repeated in the 1960s, when the first American jet airliners, the Boeing 707 and the Douglas DC-8, were introduced.

Miller and Sawers found what the previous Sources of Invention had found: there is no dominant pattern of technical development. While “technical progress… seems to have been rapid” in the aircraft industry, the inventions themselves came in many different ways.

Table 1.

Some were made in times of prosperity – which, for the aircraft industry, usually means war – and some in depression. Some were based on scientific discoveries; others were straight inventions based on the state of technical knowledge where any was appropriate, the recognition of need by an alert mind, or a more or less chance discovery.73

The aircraft industry itself produced a low level of invention; most ideas came from outside, many of them from universities and government laboratories.

Patents, say Miller and Sawers, “have rarely been important in the development of the airplane.”74 Advances in aerodynamic efficiency and supersonic flow, for example, do not often take patentable form. Manufacturers guard new designs as proprietary information until it is made public by incorporation on an aircraft. Outside inventors, in contrast, are reduced to seeking royalties from manufacturers on those developments, such as wing flaps, that are easily patentable. These conclusions highlight two characteristics of the patent system that ill suit it for aeronautical developments. It forces disclosure of information in exchange for protection and reward, a trade-off that some industries find disadvantageous. And it focuses more on mechanical artifacts than on designs and processes, which are central to aviation progress.

For these reasons and others, say Miller and Sawers, the aircraft manufacturers embraced the Manufacturers Aircraft Association and the patent pool. “The existence of the MAA,” they believe, “is mostly evidence of the unimportance of patents in the aircraft industry, but it reduces their importance still more.” Innovation has been important to the American aircraft industry, but patents have not. Furthermore, the patent pool has undermined the outside inventor by reducing the incentive of any single manufacturer to purchase rights to an invention. In the final analysis, however, “one cannot blame the MAA for the lack of invention in the American industry. It seems to be more an effect than a cause of this condition.” The proof of this, they believe, is that the British industry, which has no patent pool, has shown no greater inventiveness.75

Other analyses of the American aircraft industry have come to similar conclusions. John Newhouse, for example, attributes competition in the aircraft industry to operating efficiency, engineering integrity, configuration of aircraft, and price.76 As the industry has contracted down to fewer and fewer giant firms, there has been a tendency to associate innovation with survival. The continuous, incremental, cumulative refinement of aircraft design, usually conducted below the patent threshold, is often lost to view amidst these market forces.

Robert Ferguson’s exploration of World War II aircraft production in the United States reinforces these findings. He discovered that shop-floor practice and unpatentable, short-term, localized research often accounted for the technical gains made during the war.77 Companies did use patents, but primarily to establish postwar claims of primacy and to provide modest incentives for employees.78 For example, the Guerin patent for sheet-metal pressing of aluminum fuselage, described by one authority as “the greatest single contribution to the manufacture of all-metal airplanes,” was freely shared within the industry during the war, different manufacturers adapting it to their particular shop-floor practice and culture. “No two companies,” notes Ferguson, “designed or produced aircraft in exactly the same manner,” so that even when they used the same specifications, they often manufactured differently and produced slightly different products. For all these reasons, “during the war, patents did not play a significant role in the transfer of technology” in the aircraft industry.79

Nor do patents loom large in the case studies in aeronautical engineering that Walter Vincenti has used to explore What Engineers Know and How They Know /L80 It might be expected that patents would embody what engineers know and therefore play a large role in these stories. But “patent” does not appear in the index, and patents do not play a significant role in Vincenti’s analysis. They appear in his account of flush riveting, but mostly in the footnotes; patents seem to have little impact on the process that interests Vincenti most, the way in which knowledge of this important technique circulated within the aircraft manufacturing industry.81

Patents play a somewhat larger role in Vincenti’s account of the Davis Wing. Inventor David R. Davis invented an airfoil profile, which he then shopped around the aircraft industry. Davis creatively got around the disclosure aspect of patents by withholding from his patent application two unspecified, assignable constants, lacking which his formula was useless. The Davis wing came to be used on the 19,000 B-24 bombers built by Consolidated Aircraft in World War II.82 This, the largest production run of any bomber in history, no doubt enriched Davis and made him famous as well. But Vincenti doubts that the secret Davis wing section really contributed much to the great range achieved by the B-24. “High aspect ratio and other features of the airplane,” says Vincenti, have greater explanatory power.83 His larger point is that the real source of aeronautical development lies in the great diversity of engineering knowledge and the various ways in which that knowledge is accumulated and transferred.

Unlike Ferguson and Vincenti, Seymour Chapin addresses patents explicitly in his analysis of an important aeronautical technology, cabin pressurization.84 More precisely, he studies a costly and consequential struggle over patent interferences in the introduction of automatic rate-of-pressure change controls in the 1930s and 1940s. Patent interferences occur when two or more parties file patent applications for similar inventions that make overlapping claims. In this case, researchers at Boeing Aircraft Corporation and Douglas Aircraft Company filed patent applications for automatic rate-of-pressure change controls within two years of each other, in 1937 and 1939 respectively. Both were building on previous work that had been funded by the Army. Had the companies retained the rights to the patents, the dispute could have been handled within the Manufacturers Aircraft Association, where the arbitration board would have determined the licensing fee due to either or both of the companies. But Boeing granted licensing rights for its device to a new firm, AiResearch Manufacturing Company, which was not a member of the MAA. This dispute dragged through the Patent Office review system throughout the 1940s, finally to be decided in Douglas’s favor in 1950. AiResearch urged Boeing to appeal or to sue for reversal, arguing that it had already paid Douglas close to $1 million in royalties, with the prospect of still higher rates if the decision stood.

But Boeing had had enough. Patents were granted to both applicants, in 1951 and 1952 respectively, and Douglas’s device continued to win significant royalties. Boeing, of course, did not have to pay that market rate, for it was licensed to use the Douglas patent through the cross-licensing agreement. But AiResearch did have to pay, and Boeing had to forgo the royalties it would have received from AiResearch had its patent prevailed. Patents may not have played a major role in the technical development of American aviation, but they could still command significant sums of money in those instances where they did appear.

Analysis of U. S. aircraft patents from 1900 to 1996 reinforces the conclusions of these case studies. Figure 1, showing total aircraft patents, indicates no decline when the cross-licensing agreement was created nor any immediate increase when it was abolished. The steepest spikes accompany the public demonstrations of the Wright airplanes in 1908, the Lindbergh flight of 1927 and the airframe revolution of the early 1930s, and the introduction of commercial jet airliners in the late 1950s. When aircraft patents are taken as a percentage of total patents (see figure 2), different trends are revealed. Steep rises during the world wars indicate that other industries tended to decrease patenting activity during the wars more than the aircraft industry did. Furthermore, aircraft patenting has declined relative to other industries ever since the peak associated with the introduction of jet propulsion in commercial service. The overall patterns in these data confirm Miller’s and Sawers’ belief that major innovations in aviation have been few and far between. Most of the remarkable refinement of the modem airliner has come in small increments outside the patent system.

CONCLUSIONS

All the literature and all the data point to the same conclusion: patents have not been important in the development of aircraft manufacture in the United States. This answers some of the questions posed at the beginning of this study. Historians have neglected patents in aviation history because they did not play a large role after the early years. The patent pool appeared and survived because some such mechanism was needed in those early years, and it did little harm and perhaps some good thereafter.

But these conclusions leave unanswered the most vexing question: why was it that patents played so little role in the technical development of airframe manufacture, and what was driving that development? The answer to that question is complex and layered. It requires pulling together the findings of this study under five main headings.

First, the expectation that patents would play a role lacks warrant. Technical development does not necessarily depend upon, nor even correlate with, patenting. This is especially true in cumulative industries such as aircraft manufacture. In these cases, pioneering patents often dominate the field for a number of years. Thereafter, with the foundational ideas already introduced, the total body of knowledge at play is large and cumulative. New ideas are relatively less important. This seems to have been the case in aircraft manufacture. Technical development in this field has been continuous and significant, but it has proceeded incrementally below the patent threshold. Lucrative, important patents such as those for cabin pressurization devices, did appear from time to time, but they were comparatively few.

Second, airframe manufacture came to be conducted in a corporate environment increasingly less conducive to patenting. The industry gravitated from small companies reflecting the style and inventiveness of their founders to large corporations in which technical innovation flows from large, impersonal teams.85 Patents do issue from such institutions, but they are less important to these sponsors than they are to the lone inventor. Jacob Schmookler has found, for example, that as corporations grow larger, they pay more for research and realize fewer patents.86

Third, government funding of aeronautical research grew more important as the century proceeded. In universities and industry, as well as in its own laboratories, the government financed as much as 85% of aerospace research. Government and university laboratories were less inclined than industry and individuals to patent their developments. Even industry had less incentive to patent inventions made with government funds, for their contracts usually required free licensing to the government. Often the inventions of one industry contractor were shared freely by the government with others, further diminishing the value of patents.

This government-funded research, most often conducted under military auspices, was the principal way in which the United States subsidized aeronautical development. Many other countries adopted models of national airlines and research laboratories, directly supported by government funds. The United States eschewed such models, but found other, less obvious ways to support its aerospace industry. Research and development on military aircraft were often transferred to civilian aircraft, either through direct applications by companies such as Lockheed and Boeing that made both, or indirectly through publications such as NACA research reports that provided data applicable to both types of aircraft.87 This mechanism for the dissemination of research results became especially important during the Cold War, when aerospace research in the United States was driven by strategic and political considerations to be the best in the world. Spin-off from that research fueled the commercial aircraft industry throughout the second half of the century.

Fourth, in this competitive environment, companies often preferred proprietary knowledge over patents as a way of protecting their ideas. Indeed, patents were often avoided for the very reason that they made inventions public. An invention published in a patent could often be worked around faster by the competition than if it were kept secret until incorporated in an actual aircraft, whence the competition would have to reverse-engineer it, master its production, and redesign an existing plane to install it as a modification. As far back as the 1930s, aircraft manufacturers who came to NACA wind tunnels to test their aircraft and components were more interested in maintaining proprietary secrets than they were in patenting.88

Maintaining proprietary information in this industry was a difficult proposition. Aerospace engineers are mobile professionals. Companies can raid the staffs of their competitors to hire expertise and knowledge. Alternatively, some engineers may find themselves laid off, especially in the boom-and-bust atmosphere that came to surround the assignment of large contracts to increasingly fewer firms in the late Cold War. Furthermore, as aerospace engineering became more professional through the twentieth century, more and more members sought career advancement through publication. Walter Vincenti’s account of the development of flush riveting, for example, shows that even as companies and government employees were taking out patents on techniques of flush riveting, others were publishing the results of their research in the open literature. His account of the ways in which ideas about flying qualities circulated in the aeronautical community provides still more evidence of the multiple avenues of communication through which knowledge could travel.89 Keeping up on the state of the art, even as it advanced rapidly, was comparatively easy to do.90

But the free circulation of people and ideas did not mean that inventions in one firm were necessarily transferrable to another, at least not directly. Each of the major aerospace manufacturers developed a unique culture of management, design, and shop floor practice. So distinct were these cultures during World War II, that, according to Robert Ferguson, some firms could not replicate exactly the product of other firms even when they worked from identical specifications and plans.91 This points up the unusual tension between standardization and hand­crafting that has always shaped aircraft manufacture. Aircraft defy the kind of automation that has overtaken, for example, the automobile industry. Airframes are still put together by hand, albeit in an assembly-line style. Therefore, shop floor practice leaves its imprint on the final product far more than in other standardized industries.92 Know-how from one environment might require considerable adaptation to work in another.

This last point emphasizes a fifth and final reason why aircraft manufacture came to depend less on patenting than might have been expected. Specific inventions in this field, in the form of devices, are often less important than design and process. In aircraft manufacture, production economies and competitive advantages are more often achieved through superior design and efficient manufacture than through superior components. The cabin pressurization devices studied by Chapin, for example, were similar in their effect and both were workable. The patent fight was not over which was better but which was first. When Douglas won that fight, Boeing simply worked around the patent to install on its aircraft a functionally comparable device. This sort of skirmishing, while it entailed significant royalty payments, was not the kind of development that accounts for the rise of Boeing or the decline of Douglas. It is possible to win patents for designs and processes, but companies seldom invested their energies this way. Rather, they developed their own style and relied on that to give them economic advantage in the marketplace. Real competitive advantage in the aircraft manufacturing industry came from practice, more than it came from patentable ideas.

As the aircraft manufacturing industry in the United States contracted over the course of the twentieth century, it became both more and less competitive. It was less competitive in the simple sense that there were fewer competitors. But it was more competitive in the sense that each remaining competitor was larger, had more resources, and risked more in a market with fewer buyers. The industry moved toward oligopoly while the market moved toward oligopsony.93 In such an environment, marketing and business decisions loomed larger; the price of the aircraft and its operation and maintenance counted for more than technical features, which tended to be similar across the industry. And as always, the single most important determinant of aircraft operating efficiency was the engine, a technology that always fell outside the cross-licensing agreement.

For all these reasons, patents have grown less salient in airframe manufacture than in U. S. industry in general (see figure 2). Individual decisions to seek a patent were no doubt shaped by these considerations and others, depending on particular circumstances. This account cannot begin to explain how those decisions were made. But it does provide some conceptual tools for thinking about aggregate patenting behavior in this field.

The evolution of aerodynamics in the twentieth-century – was it engineering or science? In retrospect, this should now appear to be a rather naive question. The answer is both, and more. When the academic community embraced the idea of powered flight at the turn of the century, they found a plethora of questions about aerodynamics ripe for science to answer. Of course, airplanes were flying, and flying somewhat successfully, long before the answers came. It is this aspect that prompted the aviation historian Richard K. Smith to state: “The airplane did more for science than science ever did for the airplane.”12 Today, modem aerodynamics is dominated by the computer; the techniques of computational fluid dynamics allow us to solve complex aerodynamic flowfields heretofore dreamed impossible. And applications are being made to the whole spectrum of flight, from low-speed to hypersonic vehicles. However, it is clear that the way we have arrived at our current understanding of aerodynamics, and our ability to predict aerodynamic phenomena, is through an intellectual process that blended the disciplines of science, engineering science, and pure engineering. Perhaps, within the scope of the history of technology, aerodynamics is one of the best examples of such blending.

These conclusions about the MAA raise one final irony. The criteria used by the Justice Department in 1972 to rationalize breaking up the Manufacturers Aircraft Association soon proved to be inappropriate. The Antitrust Division had identified nine restrictions that it believed should be applied to patent licensing. Restriction number two forbade a patentee from requiring a licensee to assign to the patentee any subsequently acquired patents.94 The cross-licensing agreement did this, and it was on this basis that the Justice Department filed against the MAA. By 1973, however, the Justice Department began to appreciate that these restrictions were, in the words of the Chief of the Intellectual Property Section of the Antitrust Division, “economically counterproductive in that [they] discouraged investment in R&D and discouraged efficient licensing of patents.”95 This revelation did not, however, stop the wheels set in motion in the U. S. District Court in New York. The same Justice Department that endorsed the cross-licensing agreement in 1917 brought it to a close in 1975. The agreement probably did restrain trade in 1917; it probably did not in 1975. The Justice Department got it wrong both times.