Other Research Projects: Apollo Bioinformatics Molecular Evolution Physics of Scale

       Materials Research Activities

Bernadette Bensaude Vincent, Materials science and engineering: an artificial discipline about to explode?

[This paper was originally intended for publication on paper. We have kept the format with footnotes and an appendix intact. A version of the argument has been published in Historical Studies in the Physical Sciences. The title of the paper on this webpage was chosen by Arne Hessenbruch. It was intended to be provocative.]

Introduction Crystals and interdisciplinarity (1957-1970) From metallurgy to solid-state physics The Cold War A coherent research field 1970-1990: Composites and mixed-disciplinarity The Japanese challenge Materials by design The presence/absence of chemistry 1990-2001: Complexity and the nanoscale From micro through bio to nano A variety of approaches textbooks practices underlying view of nature Conclusion Appendix

Introduction

Although materials became an object of scientific inquiry as early as the 17th century, the academic discipline known as Materials Science and Engineering is much more recent. Galileo's 1638 dialog about Two New Sciences laid the foundations of the mechanics of elastic bodies which developed at the fringe of pure science, as a kind of mixed science. It later became an integral part of the program of experimental philosophy developed by Robert Hooke at the Royal Society, by Mariotte in France, and Jacob Bernoulli in Switzerland. However, only in retrospect can we identify these studies of wood, iron and steel as a proto-science of materials, because there was nothing like a notion of materials in general at the time. [1] This tradition, which has been part of the training of engineers since the 19th century, is not the source of modern materials science.This paper starts from a clear distinction between materials research encompassing many research activities conducted in various departments and institutional settings around the world and Materials Science and Engineering (MSE), a discipline that emerged within a specific historical and geographical context: the American universities in the 1960s. This paper is essentially about MSE. It is an attempt to understand the dynamics of the emergence of this discipline in the US academic world and to point out the local circumstances and forces at work. Unlike Robert Cahn's recent book entitled The Coming of Materials Science that provides a valuable "pointilliste portrait of the discipline, to be viewed from a slight distance", I attempt a general narrative. [2] This narrative is focused on two questions. Firstly: What can be the consistence of a field of research that includes such diverse subjects as wood, concrete, paper, polymers, metals, semi-conductors, and ceramics? The generic concept of materials is in itself a challenge because each material is a singular, if not a unique substance with specific properties adapted to specific functions. Consequently how could materials be objects of science according to the old Aristotelian saying: "there is no science but of the general". What is a material in general? Unlike matter this notion refers to a substance that is useful or of value for human purposes. Materials are usually defined as "substances having properties making them useful in machines, structures, devices, and products". The notion of materials combines natural science and humanities: it combines physical and chemical properties with social needs, civilization, industrial or military interests. From this coupling of natural and human aspects embedded in the definition of materials follows one characteristic feature of materials science. Knowing and producing are never separated; the cognitive purpose and the technological interests are intertwined. Materials science couples scientific research with engineering application of the end product. In the USA this domain is known as materials science and engineering, and the report of the National Academy issued in 1975 makes it clear that "one should speak of materials science and engineering as an it rather than them". [3] Here is a science made of a cluster of specialists from various disciplines (metallurgists, ceramists, electrical engineers, chemical engineers, physicists, inorganic chemists, organic chemists, crystallographers, and so on). What do they share?Whereas the professional activity of highly specialized physicists may be totally different, they share a common culture acquired at an earlier stage and can still present themselves as physicists. Individuals can imbibe the identity of the discipline. By contrast, few contemporary materials scientists or engineers define themselves saying  "I am a material scientist. Most of them present themselves as physicists, or chemists, or metallurgists or mechanical engineers, chemical engineers. They are defined by their respective background rather than by their current field of research. The reason for this contrast is that the diversity offered by materials science is not of the same kind as the diversity of such disciplines as physics or chemistry. It does not result from a process of diversification within a field keeping some cohesion. It is the product of an aggregation of several specialties already separate.The second concern of this paper is to try to identify the driving forces that converged into the emergence of the discipline in the USA. On several occasions over the past 50 years materials scientists wrote historical pieces to recast the past of their field. More than often the emergence of materials science is described as the outcome of an internal movement of modern science towards increasing abstraction. In this view the generic notion of materials instantiates the transition from an empirically based technology to an era when the properties of materials were understood on the basis of rational sciences. The advanced materials created by MSE can be presented as a new episode of the secular process of emancipation of mankind from nature. Because they are no longer extracted from nature, new materials are supposed to testify to the increasing power of mankind to create artifacts. Man and man's needs are the supposed driving force as indicated by a report of the National Academy of Science published in 1975 and entitled Materials and Man's Needs. However with such general considerations about human nature the identity of the various agencies that converged into the ermergence of MSE in the USA remains obscure. Did it result from an internal evolution of the various disciplines that merged into this new entity? Was it the creature of political measures coming from the top? The balance between bottom-up and top-down constraints presumably changed over forty decades. Therefore a kind of periodization may be helpful in order to get a finer understanding of the causal dynamics at work.

Crystals and interdisciplinarity (1957-1970)

Let us start with a simple remark. It is said that Copernicus founded modern cosmology, Galileo and Newton modern physics, and Lavoisier chemistry. For various reasons scientific communities like to forge the statue of their founding hero, duly celebrated by the disciplinary communities. Every science, every subdiscipline has been given a founding father. But who is the founder of materials science? To my knowledge, there is no candidate for this prestigious position - and this is not for want of genius or for lack of ambition among materials scientists. So we are faced with a founderless science, a domain that stems from no single identifiable root. In particular, although the study of the strength of materials remains indispensable for all structural materials, the mechanics of solids was not the source of the modern discipline.Metallurgy is a more obvious source. In the 1960s the departments of metallurgy of a number of academic institutions were renamed "metallurgy and materials science" and a few years later materials science emerged as an autonomous entity.

From metallurgy to solid-state physics

This linguistic change is the outcome of an evolution within metallurgy starting in the 1910s, when William H. and William L. Bragg - father and son - opened up a window on the arrangement of atoms in crystalline structures thanks to x-ray diffraction [4] . An instrumental technique for visualizing structures was thus the prime mover. And we will observe that in the following decades imaging techniques have continued to play a leading role in the advancement of materials science and technology. The study of crystals and the determination of crystalline structures became a concern of physicists or rather "physical metallurgists" as they were named in the 1920s. And from this moment on, the future of metallurgy lay in their hands rather than in the hands of skilled metallurgists, or chemists. Whatever the importance of composition and chemical bonding for the study of alloys, chemists were marginalized. Investigating the microstructure became a priority because it provided an understanding of the mutual disposition of phases and of the properties of the alloys. As Robert W. Cahn emphasized, "Microstructure as a scientific category is the peculiar contribution of physical metallurgy to the study of solid state physics, and was later extended to materials science." [5] Physicists introduced the notions of crystal lattice, dislocation, and defect. Dislocations were directly observed in the 1950s with the transmission electron microscope. The connection between microstructure and mechanical properties was thus probed and the models and theories elaborated by physicists were put at work to design new materials.Once x-ray diffraction techniques had provided precise atomic pictures of solids, quantum mechanics provided the theoretical foundations for the description of solids. Quantum theory soon reinforced the domination of physicists while the solid state became an object of investigation in itself. More directly, solid-state physics contributed to the emergence of materials science, by its focus on "structure-sensitive properties". As pointed out by Spencer Weart in his contribution to the volume Out of the Crystal Maze, solid-state physicists discriminated between the properties depending on the idealized crystal pattern and the properties dependent on "accidents" of the inner arrangement or of the surface of the solid [6] . This focus on structure-sensitive properties in the study of crystals can be seen as the main pathway leading to materials science. In addition to this theoretical influx, thermodynamics and phase diagrams are another important tributary stemming from Josiah Willard Gibbs that merged into solid-state physics with Friedrich Seitz and David Turnbull.However, even if the study of the solid state was a first step towards the emergence of the generic concept of materials and if it provided the notions of microstructure and structure-sensitive properties, a solid is not a material. The relation between structure and properties is only one aspect of MSE: the notion of a material requires that structure and properties be coupled with functions or performance.

The Cold War

The generic concept of materials first appeared in the language of science policy makers where it was represented as a bottleneck for advances in space and military technologies. During World War II, the critical needs were still addressed in terms of one strategic material (synthetic rubber, or plutonium for instance). By contrast, in the 1950s, the US President Science Advisory Committee (PSCAS) singled out materials as a priority. The advent of Sputnik in 1957 brought heavy investments in space research, with long-range programs and no concern for payoff [7] . The idea that all materials were strategic emerged in the context of the Cold War as a major condition for responding to future emergencies [8] . The Department of Defense (DoD) decided to sponsor many investigations into materials for special applications in weapons and aerospace. Through its Advanced Research Project Agency (ARPA), the DoD developed contracts with a number of universities. In June 1961, a contract was signed with 5 universities (Harvard, MIT, Brown, Stanford, and Chicago) for a total of $13,375,000.

Dr Jack P. Ruina, the director of ARPA, emphasized the critical need for new materials and a better understanding of the fundamental processes underlying their performances. He mentioned four topics: magnetic and low-temperature research; semi-conductors and their applications to devices; electronic materials development and preparation (semi-conductor, superconductor, and semi metals); and solid-state structure studies using advanced techniques. Research in 1957 was categorized as follows:

• solid-state physics 35%
• metallurgy 29%
• ceramics 10%
• smaller amounts to chemistry structural or conventional materials.

It was indeed a very narrow definition of materials research.

The ARPA program had two important features.

1. Instead of building a big unit concentrating all national resources in materials technology, the DoD encouraged research at already existing academic sites with long-term contracts for grants and instruments. This choice, which had been a major feature of all federal research programs since the early 1940s, boosted fundamental research [9] . In other words, the DoD provided academic scientists with equipment that they could never have afforded with their own funds, with the aim of tilting their research toward military needs.
2. ARPA has a credo: interdisciplinarity. The DoD's strategy was to create interdisciplinary labs (IDLs), modeled after the Nuclear and Electronics Labs at MIT and other universities. The interdisciplinary unit provided core support to a series of individual research projects and additional budgets would come from traditional sources. Twelve IDLs were funded by ARPA, three by NASA, and two by AEC (Atomic Energy Commission). From 1961 to 1970 ARPA spent $ 157.9 million on the IDLs program. [10]

In 1964, ARPA launched a "Coupling program" to gather specific industrial, governmental laboratories and university interdisciplinary laboratories. The aim was to hasten the transfer from laboratory research to industrial production but the program was still basically orientated towards defense needs. [11]

Practically the ARPA program resulted in the construction of new buildings built around a central unit of heavy equipment for processing and testing materials. [12] Instrumentation would act as a driving force to prompt meetings and collaborations, to develop a common culture or what one actor called a spiritual entity. The rhetoric surrounding the creation of IDLs emphasized flexibility and partnership between departments.

A coherent research field

After 10 years of operation, the program seemed a full success at least in quantitative terms: there was a dramatic increase in publications and doctoral and master degrees in materials related subjects [13] . However it is difficult to evaluate the success of this program since a number of achievements in materials technologies like the transistor, and photocells, soft magnetic materials preceded it. Rather than initiating creativity the ARPA program rode a wave, although in restropect, with the experience of the next decade, it appears to have encouraged imaginative science. Moreover, the optimistic annual reports sent by the IDLs to ARPA underplayed the difficulties of fitting the flexible, non-hierarchical interdepartmental centers into the academic structure. In addition to the usual struggles for power between academics, a source of difficulty was that most of the research projects were conducted outside the buildings created by the ARPA program. Frictions sometimes occurred between the research centers and the departments. More importantly, interdisciplinary research often meant that the physics community had the leadership of the scientific side and the metallurgists on the engineering side.

This program, however, created the research field of Materials Science and Engineering, at this stage mainly an American science. Although in Europe, a number of Materials Science Centers grew out of former metallurgy departments, materials did not become a political concern until the1970s. [14] This does not mean that no advanced materials research existed outside the US. [15] In France, the Netherlands, Norway, the United Kingdom, as well as Canada, there were many scattered pockets of research on subjects related to materials but no identifiable field of materials science, no central project pushed by the government.

To sum up this first period, materials science as a research field first materialized in a number of buildings with research facilities. Instrumentation played a crucial role in this early stage. Although a strong impetus towards interdisciplinarity existed in a number of places such as MIT, the discipline of MSE emerged out of a governmental decree with generous funding. It was nurtured by academics finding in materials science a good opportunity to pursue their own research interests in crystalline structures. Therefore materials science, at this stage, was essentially oriented towards fundamental science. Industrial companies were expected to take over and jump onto the bandwagon but the 1964 coupling program failed.

Before 1970, MSE was a fairly coherent field of research because it was largely dominated by physics, metallurgy and crystallography. Although a number of chemists were involved in the interdisciplinary projects, polymer science was not properly integrated into the corpus of MSE and the 1975 report, Materials and Man's Needs, expressed doubts about the feasibility of doing so. [16] The 1976 issue of the journal Materials Science and Engineering reviewing the field for the 10^th anniversary of the journal's publication provides a good overview: nearly 80% of the contributions dealt with metals and alloys most of which focussed on phase transformation. Only two essays dealt with polymers, one dealt with ceramics and one touched on oxide glasses.

1970-1990: Composites and "mixed-disciplinarity"

In the early 1960s, many metallurgy departments were renamed "metallurgy and materials science". In the 1970s they dropped the term "metallurgy" altogether and became known as "departments of materials science and engineering". Thus, materials science became more than a research field: a discipline of its own, a pedagogical entity.It is certainly possible to view this changing status as an example of the standard process of maturing disciplines mentioned above. It is tempting to argue that the field "naturally" evolved toward a more general approach to the relation between structure and properties. This is actually the standard interpretation to be found in textbooks. But in fact, it was nothing like a smooth change. The elimination of metallurgy from academic curricula raised violent reactions from both the academic milieu and metallurgy corporations. [17] More than a resistance to innovation, this opposition expressed a loyalty to metallurgy. The change was finally accepted because it was perceived as a measure to face a dramatic decline in students' enrollment. At the same time graduate and undergraduate curricula in MSE opened up with the expectation that they would prove more attractive to students than existing programs It was an odd new discipline, however, having to include as subfields well-established disciplines such as metallurgy and chemistry. A certain violence and much negotation were required for MSE to be established as an academic discipline.These university initiatives paralleled changes in government research policy. In 1972, the responsibility for the interdisciplinary laboratories was passed from ARPA to the National Science Foundation and they were renamed Materials Research Laboratories. [18] The new contract signed on July 1, 1972, following visits to all sites and evaluation of the materials centers over ten years, did not deeply change the organization but rather maintained the policy of a core funding. The NSF reinforced the interdisciplinary requirement. "Thrust groups" of research were encouraged to develop coherent multi-disciplinary and multi-investigatorial projects and the definition of an interdisciplinary group was clarified. [19] After a transition period of two years, however, the NSF developed its own program philosophy. It was clearly intended to reduce the imbalance between materials science and materials engineering and between research on electronic properties and research in other significant areas. [20] What was the impact of this new policy?In the long-term, the transfer from the Department of Defense to a civil agency meant a gradual but nonetheless drastic reduction of the budget during the next 15 years. Consequently the number of faculty involved in materials research dropped from 600 to 400 between 1970 and 1985. [21] Moreover, the aims of academic research were reoriented according to new social priorities in the 1970s. On the one hand, national committees expressed a growing concern with resources, renewability and the environment. [22] Environmental and safety legislation put new demands on material scientists. In both industrialized and developing countries, criticism arose of the overemphasis on basic science at the expense of technological needs of materials supply and demand [23] . Hence a misalignment was repeatedely alleged between university and industry. In contrast with the 1950s and 1960s emphasis on fundamental aspects wiping away the empirical approaches of the past, the new wave praised engineering skills and experienced knowledge. A characteristic feature of this period is the reassessment materials technologies' past with a valorization of ancient skills in processes.

The Japanese Challenge

In addition, a major concern of the 1980s was the economic competition with Japan. US policy makers and materials scientists realized that generous federal sponsoring could generate perverse and counterproductive effects. A number of scientists deplored that it blocked the transfer from university to industry although it certainly secured the leadership of American universities in materials science. Meanwhile Japanese companies, thanks to a vigorous government-developed technology program, had incorporated the best of American technology and were beating American companies in advanced materials. The success of Sony Inc. had the same humiliating effect as Sputnik had had a few years earlier and prompted a vigorous response from American science. Competing with Japan and regaining the leadership thus became the prime concern and even obsession of US materials projects in the 1980s [24] . Academic materials scientists such as Thomas Eagar vigourously questioned the hunt for new materials. He called for a reorientation of research towards processing and an increased commitment of industrial companies: "Designing new materials with curious properties is fun for the materials scientists and engineer but it does not often yield results of major commercial or social benefit. American companies must spend their resources learning how to manufacture existing materials economically, not searching for exciting new materials. But if we spend our resources on processing selected new products of high reliability and low cost, we will all be winners." [25]

Consequently, the governmental sponsorship of more than half of the US research that had two decades earlier been viewed as the key to success was in 1980 denounced as a major obstacle for technological innovation. A reform of patents was urged. It was argued that without the protection of an exclusive license, companies were not willing to undertake the risk of developing thousands of inventions produced by university laboratories belonging to the government. Hence only 4% of the 28,000 patents the government owned were successfully licensed. In 1980, the Patent Uniform Procedure Act established a uniform patent policy for all federal agencies authorizing the licensing of federal patents. [26] This legislative measure was aimed to encourage collaborations between industrial companies and universities.

Although pressure arising from concern with renewability of natural resources was felt as a dramatic turning point by a number of pioneers who had built the military-university programs [27] , the American academic community responded promptly to the new demands and tried to develop links with private industries. MIT, where the concept of university research as a nursery for industrial innovations had been developed in the 1930s, revived its old tradition of industrial collaborations. [28]  MIT created a second center for materials: The Materials Processing Center, inaugurated on February 1, 1980. [29]  It is in the draft of the project of this center, signed by Merton C. Flemings, that I found the first occurrence of a diagram expressing interaction among structure, properties, performance, and process - which would later become the core of basic concepts presented in courses on materials science.

Thus, after the Cold War had prompted the emergence of the generic concept of materials, the competition with Japan helped reinforce the conceptual basis of materials science. Performances (or functions) and processes entered into the terminology of materials scientists adding to the already established structure and properties [30] . MSE now acquired a coherence characterized by the interaction between four variables: structure, properties, performances and process. Since 1989, this conceptual basis of MSE has commonly been visualized with the help of a tetrahedron indicating the interrelation of each of the "four elements" with each of the three others. [31] Although there is no evidence for the origin of this visual representation, it might well be reminscent of vant'Hoff's representation of carbon valencies. Could it mean that the chemical culture became more active in the field of MSE?

Materials by Design

From a logical point of view, polymer chemistry may appear as the second route that reinforced the coherence of MSE by enlarging the conceptual basis that had been restricted to structure-properties in the early decades. For advanced materials are tailored to specific purposes, they are adapted to a set of specific tasks. In contrast to conventional materials that have standard specifications and a world market, the materials created within MSE are developed according to the functional demands of the final product in view of a set of services. In other words, instead of supplying commodities that would be finalized by the customers, new materials are the end products of a co-operation between customers and suppliers. [32] Materials by design were first developed for military and space applications requiring never-before-seen combinations of properties like light weight and high-temperature resistance for use in severe conditions. A number of these exotic materials have been successfully transferred to more common uses such as sports articles and clothing.

This new generation of materials had a strong cultural impact. The creation of materials by design has often been celebrated as the dawn of a new era. An age when mankind was no longer constrained by matter but in command of matter. Materials properties would no longer limit our possibilities. Ivan Amato, a popular writer, characterized them as "the stuff that dreams are made of". [33]

The composite material is the paradigm of materials by design, i.e. a structure made of a matrix reinforced with fibers. Composite materials combine various mutually exclusive properties into one single heterogeneous structure. Steel and iron are used as supports for toughness, plastics are useful for weight saving, and ceramics for heat resistance and stiffness. The notion of composites clearly indicates that materials can no longer be defined by the structure-properties coupling. They are more adequately defined as exhibiting certain properties for certain usages and the properties themselves are conceived of as responses to specific functional demands. [34]

In 1975, the report Materials and Man's Needs queried: "Maybe composites are the future. The dominance of crystalline materials is already being challenged." [35] Was it a good prophecy? Did composites help reorganize the field of MSE? Composites effectively became a fashionable topic among materials researchers in the late 1970s and 1980s. [36] However, the integration of this technology, developed first within chemical industry, into an academic discipline dominated by the metallurgists' culture proved difficult. Composites grew out of an established chemical technology known as reinforced plastics. [37] Glass-fiber reinforced plastics had been manufactured as early as 1938 and commercialized in 1940 by Pittsburgh Plate Glass Inc. In the 1940s, glass fibers reinforced plastics were developed for military purposes, such as aircraft noses and boats for the US Navy. [38] Once the difficulty of molding large pieces had been overcome, these early composites (laminates of polyester resins), molded at low pressures, were mass-produced from the 1950s onwards for civil applications such as electric insulators and tankers. The shift from reinforced plastics to composite materials was not a radical break. It was nothing like a revolution since the technology of composites did not overthrow the more traditional reinforced plastics. The composites gradually emerged from polymer chemistry and they undoubtedly contributed to extend the domain of plastics and their successful substitution for wood, steel, or aluminum leading to "the plastic era". [39]

In the 1980s, composites became an independent field, quite distinct from polymer chemistry. Various reasons contributed to this emancipation. First, although a vast majority of the composites are made of a polymer matrix reinforced by fiberglass, composites can also be made up of metals or ceramics. The composite principle has been successfully applied to these materials. There are metals/ceramics composites, metal/metal, and ceramics/ceramics composites (the latter were crucial to extend the uses of ceramics because the fibers prevent catastrophic failures) [40] . No matter what the nature of the constituents, as long as the structure associates two phases. The term composite definitely superseded the traditional notion of fiber-reinforced plastics when new high-modulus fibers were introduced as reinforcing components for high performance materials in aerospace industry. Carbon fibers started to be used in the late 1960s. With these long fibers, Japan reinforced its leadership in advanced technology. Japanese companies took over and held a monopoly on carbon fibers. Moreover carbon fibers required a different treatment of composites. Unlike glass fibers they are used as long fibers. They are not spread all over the resin but carefully arranged with a definite orientation according to the main efforts during the functioning of the structure. Unlike fiberglass-reinforced plastics, carbon composite materials are anisotropic structures. They are mapped to respond to specific solicitations, with specified conditions of use. Each one offers a landscape of its own. Such composite materials share with fine arts pieces the privilege of being unique creatures.

In addition to white glass fiber and black carbon fiber, a "yellow" fiber was manufactured in 1971. Kevlar®, the first of a family of synthetic aramide fibers, emphasized the role of chemical industry in the emergence of materials science, since it was invented in Du Pont's laboratories in Wilmington, Delaware. The inventor was a woman scientist, Stephanie Kwolek. Her work exemplifies the close connection between products and processes in the technology of materials. The Du Pont research laboratory was looking for new products, new nylons. [41] But it was the study of a new process (room-temperature polymerization) that opened up an avenue to a new generation of polymers like Nomex®, Kevlar®, and Lycra®. The example of Kevlar® thus demonstrated the enormous potential of processes.

The composites technology contributed to a specific approach to materials design, also within MSE. The interaction of the four variables - structure, properties, performances and processes - is such that changes made in any of the four parameters can have a significant effect on the performance of the whole system and hence require a re-thinking of the whole device. A sample success story will help understand this synergism: The substitution of a composite material for the chrome-steel bumper in automotive structures. It was initially intended to gain weight and reduce energy consumption. It turned out that it allowed a redesign of the bumper with integration of various functions in the same piece (radiator grilles, ventilators, lights, etc). Not only functionality changed but so did design: gradually the car body was redesigned with curves and a new style of automobile was created. [42] However, this commercial success is limited. Composite materials cannot invade a mass-production market like automobiles because of their impact on the environment - composite materials are not easily recycled. There were also dramatic failures; for instance the Rolls-Royce Company was unable to develop a highly promising carbon fiber composite for the compressor blades of the engine in the Lockheed Tristar because it failed to meet the requisite service-reliability for such engines.

Failures are as instructive as successes, but they are more difficult to document because they are not usually advertised. The lesson to be learnt was that the traditional linear approach - "given a set of functions let's find the properties required and then design the structure combining them" - should give way to a systems approach. The systems approach makes use of computer simulation. It requires more than new techniques, and more than interdisciplinarity. The synergy between the four aspects calls for a synergy between various specialists.

Cross- and multidisciplinarity became the motto of all kinds of reports on materials research all over the world. I prefer to forge the term "mixed-discipline" by analogy with the Aristotelian notion of "mixt", a mixt being something more than the addition of its individual components. Remarkably the composite structure of materials is mirrored by a composite arrangement of human competences. At least, many attempts have been made to replace the sequential division of labor - from suppliers through manufacturers to customers - by a partnership involving all actors from the outset of the project. Both in the advanced materials and in the social organization of production, the aim is to obtain a synergy between the various components, i.e. that the properties of the whole structure should be more than the sum of its parts.

The systems approach had to be learnt through trial and error rather than through teaching. In this process academics were less committed than industrialists. At this point a striking contrast between advanced research and teaching becomes clear: the latter lags far behind.

The Presence-Absence of Chemistry

To what extent did this new paradigm and new approach change the face of materials science? Did they really challenge the prevalence of metallurgists and reinforce the role of chemists in the field? They certainly tilted the balance between science and engineering, when performances and processes became as important as structure and properties. However there is no evidence that this new approach changed the balance of power between the academic communities. Chemists did not suddenly rise to the top of materials science. Quite surprisingly, the proportion of chemists active in the field of MSE in US universities decreased rather than increased, as indicated by this chart published in 1985.

Faculty Distribution in U.S Materials Research laboratories [43]

How are we to explain this surprising diminishing importance of chemists? One conceivable explanation is that academic chemists entering the field of MSE were so well integrated in the emerging community that they were counted among the MSE scientists rather than as chemists. But no: chemists neither integrated easily into the research field nor did chemistry enter into the curricula of MSE. Merton C. Flemings and Robert W Cahn recently pointed out this failure: "Of all subfields of MSE, the one that has proved hardest to integrate with the rest is that of polymers. Polymer chemistry has remained entirely distinct, with large numbers of its own journals (…) Many universities have found polymers difficult to integrate into MSE curricula since many concepts of polymer science seem remote from those of traditional inorganic materials. It may well be that the long-term strength of MSE as a discipline will depend on how well it eventually succeeds in integrating polymers into its academic programs." [44]

A number of chemists deeply involved in materials science deplored the lack of recognition of their importance in MSE. A feeling of marginality is clearly expressed in 1987, on the occasion of the 25^th anniversary of the creation of the interdisciplinary laboratories. [45] George Whitesides and his collaborators developed a lengthy argument to demonstrate the importance of chemistry for materials science. They first pointed out that chemists pioneered the age of materials by design since conventional organic polymers already illustrated the making of materials with desirable properties. If organic remained on the margins of the emergent field of MSE in the 1960s, it was because they did not present the desired properties such as high-temperature stability, electrical or thermal conductivity, or oxidation resistance. However the situation changed with the dramatic development of composites. The paper emphasized that chemists directly contributed to rendering interfaces and surfaces a new field of research - a molecular surface science - which is aimed at understanding electrical, magnetic, and optical properties of surfaces at the molecular level. [46]  This new focus was a direct consequence of the emergence of composite technologies. Since the mechanical properties of heterogeneous structures depend upon the quality of the interface between fiber and polymer, it was crucial to develop additive substances favoring chemical bonds between the glass and the resin. Chemists were indispensable for the synthesis of many small components of advanced materials such as adhesives, lubricants, and surfactants. More broadly, as Whitesides emphasized, the chemists' ability to manipulate structures at the molecular level should allow them to play a leading role in the shaping of MSE.

Moreover it could be argued that chemistry contributed more than organic materials. Inorganic chemists had been involved in ceramics and to a certain extent their participation in this field paralleled the role that solid-state-physics had played in metallurgy a few decades earlier. Solid-state chemistry related the macroscopic behavior of solids to their microscopic crystalline structure. This approach made possible such strategic materials as advanced solid electrolytes for high performance batteries and other novel solid-state electrochemical devices, ionic supra-conductors, and conductive polymers. Here we may find one reason for the relative marginality of chemists in the field of MSE. Solid-state chemistry remained a relatively small subject in the US academic community as compared to Germany and France. According to Frank DiSalvo, "there were still no more than 15 solid-state chemists in the US in 1985". [47]

Another plausible reason for the poor recognition of the role of chemistry is that many of the chemists involved in materials research were active in industrial laboratories rather than in university departments. Although it is difficult to know the number of chemists employed in industrial research, given the dispersion of sites, many significant contributions to materials science were made in the corporate world. Kwolek's invention of kevlar in 1971 illustrates the gap between industrial research and academic research in chemical technologies. She entered Du Pont with a bachelor's degree and was unaware of the theoretical predictions made by Paul Flory on condensation polymers. Her invention did not result from fundamental research. While basic science was not crucial at this stage, her discovery of liquid crystalline solutions (in which the molecules all line up pointing in the same direction) proved to be a tremendous advance in fundamental science as well as the starting point for the large-scale production of new fibers. Fundamental and theoretical knowledge were no longer the driving force, academic research was no longer upstream with industrial applications downstream. This new regime is also examplified by the invention of new materials for solid batteries which all came from industrial laboratories such as Ford Motor company for beta-alumina, then from Hydro-Quebec for polymers and from Exxon laboratories for titanium disulfide electrodes [48] . Not only new materials but also new instruments were discovered in industrial laboratories. For instance, two industrial researchers, Heinrich Rohrer and Gerd Binnig developed a vacuum tunneling technique at the Zurich IBM laboratories, in 1978. Their Nobel Prize winning contribution allowed the construction of a new instrument, the scanning tunnelling microscope, and opened up a new era in materials research.

To sum up, this second period enlarged and reinforced the field of MSE. First a conceptual framework was clearly identified, more closely intertwining science and engineering. All practitioners, whatever their background, had to learn a new way of thinking, the systems approach. It was during this decade that materials science became a discipline taught at school and went through the whole process of institutionalization that has become the standard of discipline formation since the 19^th century. An annual review of materials science started in 1971, insisting that it was a booming field of activity based on the general characteristics of materials. In 1973, a Materials Research Society was created in the USA organizing two annual meetings and publishing a Journal and a Bulletin. The European Materials Research Society was founded in 1983. [49] And there is now an International Union of Materials Research Societies.

At this stage it would certainly be possible to agree with most of the authors of textbooks and editors of reports that MSE was a maturing or already matured discipline. [50] However this biological metaphor, used and overused by materials scientists, is misleading. It tends to assimilate scientific disciplines to maturing organisms it suggests a natural and predetermined process of growth.

1990-2000: Complexity and the nanoscale

The emergence and evolution of MSE may be more adequately described in terms of responses to challenges than in terms of "organic growth". In the beginning, the generic notion of materials science has been forged in response to the Soviet challenge of Sputnik. Then in the 1980s, the Japanese challenge prompted a re-orientation towards processing and industrial competition.

Over the last decade, materials scientists and engineers invented (or discovered) their most serious challenger: nature itself. While striving to design high-performance, multi-functional composites, materials scientists realized that such materials existed already in nature. Optimal combination of properties and adaptive structures are exemplified in living organisms. Through the eye of materials engineers, wood is a composite material made out of long, orientated fibers immersed in a light ligneous matrix. Sea-urchin or abalone shells are wonderful bio-mineral structures made out of a common raw material: calcium carbonate. They present a complex morphology and assume a variety of functions [51] . Similarly, spider silk is a fiber extremely thin and robust that offers an unchallenged high strength-to-weight ratio. How could such efficient structures be designed? Suddenly the metaphysical entity  "Nature" surfaced among the technical and esoteric terminology of specialized journals in MSE. Nature is called upon to provide materials scientists or engineers with novel concepts for crystal engineering and materials designing.

"We can be encouraged by the knowledge that a set of solutions have been worked out in the biological domain", wrote Stephen Mann. "The challenge then is to elucidate these biological strategies, test them in vitro, and to apply them with suitable modification, to relevant fields of academic and technological inquiry." [52]

Turning from the military and industrial competition with rival nations to the rivalry with nature, the style, conditions and practices of research changed deeply. The political momentum of the 1950s to create MSE declined and is still declining. Despite the continuous pressure for miniaturization, despite an extensive mobilization around environmental issues and the cycle of materials, it is more and more difficult to present MSE as an answer to "man's needs". There are many sophisticated materials on the shelves, without applications. In a way, the supply exceeds industrial demand.

This situation was reflected in the dramatic reduction by 50% of the research staff in most industrial companies by the end of the 1980s: DuPont, Exxon, Bell laboratories only maintained programs focused on improving their company products. Many industrial researchers joined the university laboratories. Academic materials centers found new recruits. The academic community was thus reinforced but it had to rely on its own resources, on its own dynamics.

From Micro through Bio to Nano

Academic scientists successfully invented their own program. Looking back into the past, they found a prophet of the coming age of nanotechnology. A lecture delivered by Richard Feynman in 1959 and published in 1960 became their bible. "There's plenty of room at the bottom", Feynman declared, in order to invite physicists to enter the field of nanometer scale and to construct objects atom by atom. Feynman made various and often playful suggestions of technological applications such as writing the Encyclopedia Britannica on a pinhead using beams of electrons. His predictions were already inspired by a biological model: "This fact - that enormous amounts of information can be carried in an exceedingly small space - is, of course, well known to the biologists (…)The biological example of writing information on a small scale has inspired me to think of something that should be possible. Biology is not simply writing information; it is doing something about it. A biological system can be exceedingly small. Many of the cells are very tiny but they are very active; they manufacture various substances; they walk around; they wiggle; and they do all kinds of marvelous things - all on a very small scale." [53]

Feynman's suggestions were made reality thanks to advances in scanning tunnelling microscopy (STM) and the Atomic Force Microscope (AFM). Such instruments allow not only the visualization but also the manipulation of individual atoms. [54] Writing with atoms, the dream inspired by Feynman's metaphors, was achieved in 1990 when Peter Zippenfeld published the atomic logo of IBM in xenon atoms. The IBM release entitled "IBM Scientists Position Individual Atoms - First to Build Structures One Atom at A Time" epitomizes the combination of advertising and cognitive interests. [55] IBMs image was soon joined by the Japanese company JEOL's Japanese characters for "nanoworld". This was clearly intended to convince both the public and the scientific community that controlling and manipulating the small world was possible. Atoms and molecules can be placed at selected positions so that it is possible to build structures to a particular design atom by atom.

Thus MSE shifted from the microscale to the nanoscale (1nm=10^-9m). A new approach developed with nanotechnology. Until then, materials had been carved out like a statue out of a block of marble: for instance electronic chips are made by slicing a cylinder of single-crystal silicon into thin wafers, and later divided into tiny pieces. This "top-down" approach is now replaced by a "bottom-up" approach where a nanostructure is built by bonding atoms or groups of atoms. The key issue becomes the assembly of building blocks. Instead of the chemical interactions used in traditional chemical synthesis, materials scientists are looking for processes of spontaneous association of molecules similar to those employed by nature in the design of living organisms. Molecular self-assembly is thus one major reason of the interest of materials scientists in biomaterials. Nature makes the mortar and the bricks at the same time in biominerals and self-assembles them through the use of templates with a close control of the process at each level. The biological macromolecules are involved in controlling nucleation, growing, shaping the crystal and adapting its mechanical properties to the functions of the bio-mineral. [56]

A second motif of the materials scientists' interest in bio-structures is that they reveal a complex hierarchy of structures. Each different size scale, from the Angström to the micron and millimeter scale, presents different structural features. The remarkable properties of bulk materials, such as bone or tendon are the result of this complex arrangement at different levels, with each level controlling the next one. [57] Here is a form of complexity far beyond the complex composite structures that materials scientists had been able to design.

A third feature of biomaterials that scientists are trying to achieve in their own man-made materials is their adaptability to the environment. Designing responsive, self-healing structures has been one the major objectives of materials research in the 1990s. For this purpose, programs on smart or intelligent materials were launched. Basically, intelligent materials are structures whose properties can vary according to changes in the environment or in operating conditions. For example, materials whose chemical composition varies according to its surroundings are used in medicine as prostheses. Materials whose structure varies according to the degree of damage due to corrosion or radiation are thus able to repair themselves. The whole problem is to have inbuilt intelligence. This means at least having sensors (for strain, temperature, or light) and actuators embedded in the structure in order for the structure to adapt to external stimuli. In Japan attempts were made to promote smart materials as a new discipline built up around a hard core of axioms and concepts. [58]

Nanotechnology, biomimetism, and smart materials - all three trends are manifestly inter-related and converge towards similar targets. They could or should prompt the unification of the field of materials science. However, and quite surprisingly, until now they are not really acting as converging forces nor are they creating any unified set of concepts and practices.

One major reason is that not all materials scientists are involved in these recent technologies. There is still a future in exploring microstructures and designing bulk materials. Whatever their importance in the recent science policy measures [59] , nanotechnologies will never cover the entire field of materials science. Moreover, the new perspectives opened up by the nanoscale, act as diverging rather than converging forces. On the one hand, they extend the frontiers of MSE to such territories as bioengineering or artificial intelligence. Multidisciplinary cooperation is intensified. For instance, the programs of smart materials emerged from close collaboration among electronic engineers, computer scientists, and materials engineers. The efforts to understand biominerals and to mimic the natural processes of self-assembly require an interaction with structural and molecular biologists. [60] Computer science and biology have entered the field of materials science. As a result, if one were to map out the territory of MSE today one would get nothing like a center and a periphery or a mainstream split into various little branches. Rather, MSE appears as a booming field without clear borderlines and with dozens of streams flowing in various directions.

A Variety of Approaches

In addition to its undefined territory, MSE is characterized by a variety of approaches. Without attempting a general survey of the field, not even at a typology, I will show the diversity of approaches at three different levels:

1. a survey of textbooks suggests that there are different approaches to teaching
2. a survey of research articles or review articles will provide a sample of diverse practices used for designing materials
3. finally I will try to characterize the underlying views of nature.

1. From a brief survey of a few dozen general textbooks of materials science published over the past decades one gets the impression that the discipline is not stabilized (See the list in Appendix).

Certainly most textbooks emphasize the contrast between the old, traditional, empirical knowledge about materials and scientifically based knowledge. After centuries of routine, the birth of a new scientifc age took place whose origin varies from the mid-19th to the mid-20th century. This common positivistic historical scheme undoubtedly helps cement a disciplinary community.

However, the core knowledge of the discipline of MSE, that all students around the world are supposed to know, is very small. Most textbooks begin with general notions about the structure of matter and move to the presentation of the traditional groups of materials: metallic, non-metallic, polymers, ceramics. The portion of theroretical and basic knowledge about the structure of matter varies from one introductory chapter to 12 chapters over a total of 15. Indeed, such variations depend on the audience targeted because materials are involved in many diverse science and engineering curricula. Texts aimed at engineers present the theroretical notions about structure as tools for selecting materials for design while in other texts the purpose is simply to help the reader understand how and why materials behave as they do. There is a large spectrum of formulas between the two extremes of science-centered texts and engineering-centered texts, as noticed by Robert Cahn. [61]

However, the overall organization does not convey the image of unified discipline. Rather it suggests an aggregation of fragments of knowledge taken from various disciplines around a core of atomic physics. In other words they suggest that materials science is an applied science of quantum and statistical mechanics and solid-state physics. In fact, there is no inner logic presiding over the organization of the discipline 40 years after its emergence. Rather, the organization of most textbooks simply reflects the successive layers of its historical development textbooks. One could not find a materials science textbook without a chapter on metallurgy, whereas ceramics and polymers remain on the periphery. The introductory chapters providing the fundamental notions mainly deal with metallic structures, dislocation, crystal lattices, etc. The core knowledge still consists in crystals and solid-state physics. Although the textbooks published in the 1980s emphasize processes in addition to structures and properties, they still derive their basic principles from crystalline structures and simply extend their scope to include polymers and ceramics. [62] Composites are introduced in more recent textbooks but only towards the end, after chapters devoted to semi-conductors, polymers, ceramics, and structural materials. They do not induce a new general approach to materials. The theoretical foundations, borrowed from physics, do not determine a definite way of ordering the particular chapters dealing either with classes of materials or with types of properties such as mechanical, magnetic, and optical. There is no general law to govern the multitude of particular cases and few efforts are made to frame a generic discipline. Among these attempts to rethink the whole in a more general way, Allen's and Thomas's The Structure of Materials can be singled out. The authors explicitly set out the four basic concepts structure, properties, performance and process in order to describe materials in "a catholic way". [63] Rather than deriving their basic concepts from solid-state physics, they start from general notions about symmetry and bonding that allow them to deal with various states of matter.

Since the discipline of MSE does not look remarkably stabilized, it is ironic that the textbooks enjoy a remarkable longevity. Some of them were circulated for 20 and sometimes 30 years, showing no significant evolution of the genre over four decades. This stability is quite surprising given the rapid advances and the deep reorientations described above. It is also remarkable that the most recent developments did not bring about new concepts susceptible of reorganizing a didactic exposition of the field. In particular, the notion of complexity, a candidate unifying concept, remains extremely loose and no consensus on a standard definition has come into being. A number of materials scientists refer to the hierarchy of structures instantiated in biomaterials [64] , while others rely on a physicist's notion of complexity referring to dynamic processes creating order out of disorder without dissipation of energy. [65]

2. The research strategies of materials scientists and engineers are no more unified than their textbooks. Even within the restricted domain of nanomaterials - let alone the design of bulk materials for industrial uses - there is no uniformity. Some materials scientists use and advocate rational design with the aid of computer simulation. Computational materials science developed in several universities in the 1990s, and Elsevier now publishes a journal of Computational Materials Science. [66]   Its prestige was reinforced with the award of a chemistry Nobel Prize to Walter Kohn and John A. Pople for their work in quantum chemistry. Beyond the objective of calculating the properties and stability of different structures, the ambition of computational materials science is to "model the real world by computer in a reasonable amount of time" as Uzi Landman, director of the Georgia Tech Center puts it. [67] The supreme achievement is to build up a material ab initio, using computer calculations and starting with the most fundamental information about the atoms and from the basic rules of physics. Despite the strong theoretical inclinations of computational materials scientists, their effort is oriented towards industrial applications. For instance, Gerbrand Ceder (from MIT Materials Science Laboratory) has used computers to design new aluminium oxide electrodes for batteries. Because it was one of the first materials that had been predicted on a computer before it was synthesized, this achievement revived the hubris of the ancient alchemist. It has been reported in semi-popular journal in 1999 under the title "The virtual alchemist". [68]

This a priori method, dispensing with the cumbersome manipulation of real substances in the laboratory, contrasts with the combinatorial approach. First introduced in pharmaceutical industry for the design of new drugs, the combinatorial method consists in synthesizing a large array of compounds up front before screening them. [69] Peter Schultz of the University of California, Berkeley, transposed it to materials science in a quest for high-temperature superconducting materials. This kind of Darwinian strategy - given an equivalent of biodiversity artificially created by combinatorial techniques, the materials designer selecting the fittest structures - has been licensed and commercialized. [70] Combinatorial techniques are mainly developed to speed the process of looking for new materials in various strategic applications.

Both the computational and the combinatorial methods are extremely systematic in comparison with the strategies developed by materials scientists to explore new ways of designing novel materials. A number of them are trying to use genetic information in order to design new structures. There are many attempts to synthesise novel proteins not observed in nature but engineered for specific purposes. Stephen Mann and his team at the University of Bath are using proteins as processing plants. After studying the structure and properties of native proteins and finding that they form small cages, Mann realized that he could "use the protein as a reaction vessel for controlling the particle size of other materials". [71] He and his group use proteins to cage several compounds by altering the chemical affinities of the protein. Other kinds of biological resources can be exploited for designing new materials too: DNA can be used as a rigid template to direct the spatial disposition of nanoclusters; genetically engineered bacteria are used to produce biodegradable polymers, and so on. Bionanotechnology programs are developed both in Japan and in the USA with the idea that proteins could in the future replace the components of today computers. Efforts are made to make biotransistors, biosensors, and biomolecular switches.

Whereas a number of materials scientists pursue biotechnological strategies, others rely on chemical stategies for the synthesis of nanostructures. "How can one make structures of the size and complexity of biological structures, but without using biological catalysts or the information coded in genes?" [72] Such is the program developed by chemists and physico-chemists like George Whitesides, among others. Gathering all possible resources, from covalent bonds to weaker forces like ionic bonds and van der Waals interactions, and playing with the action of thermodynamics, chemists let various building blocks self-assemble into various structures of two and three dimensions. [73] In response to the challenge of biomaterials they work at the mesoscale and deploy an artillary of techniques like etching, stamping, coating. They thus generate a variety of active or passive structures, a jungle of shapes - knots, bagels, shells - out of very ordinary materials like mesoporous silica using sol-gel techniques or surfactants [74] . Whether chemical or genetic, those semi-empirical approaches have nothing in common with the former more systematic strategies. They use trial and error, through a kind of astute tinkering or "bricolage", rather than deductive method.

3. Underlying these different practices for designing manomaterials are very different views of nature and of the relations between art and nature. The scientists who use rational and systematic strategies are led by a deterministic view of nature as a rational system governed by a few simple and general laws entirely transparent to the human mind. In particular, both the ab initio and the bottom-up strategies imply that individual materials are just a specific configuration of atoms deducible from the universal laws of matter. For these scientists, the route from atoms up is smooth, without obstacles.

In the writings of the pioneers, one can find visionary views that magnify and emphasize the main features of the emerging technological culture. Several promoters of smart materials technologies seriously planned to design materials endowed with some of the functions of human mind: from sensors and actuators, they would move to consciousness. [75] Eric Drexler, devoted a number of popular writings announcing that "nanotechnology would bring changes as profound as the industrial revolution". [76] Drexler depicted atoms and molecules as nanomachines. They are "universal assemblers", to be used by engineers as machine tools to create performance-enhanced molecular machines. Improving on nature is the main objective and there is no limit to the power of those handling the "universal assemblers". Like Marcellin Berthelot one century ago, Drexler develops an ambitious program for reproducing the bio-structures and outdoing nature. [77] This program is based on a simplistic view of the complexity of material structures. Drexler does not doubt the feasibility of his program because the existing "molecular machines" serve a range of basic functions nor that they can be combined to design more and more complex machines. One remarkable feature is that although the systematic strategies for designing complex structures are generally based on quantum physics, they rely on a view of nature close to that developed by classical mechanics. The material designer is like a clockmaker, who by adjusting small units can make a complex machine allowing mankind to dominate time. In this perspective, the ideal is the mastery of nature and there is little doubt this ideal can be achieved through fundamental science. Knowledge is power.

In stark contrast to this mechanistic view of nature underlying the systematic approaches to materials design, the more empirical approaches are based on a more teleological view of nature. Nature is presented as a subtle and unrivaled engineer, a wizard who has invented elegant solutions to most of the problems that confront materials designers. [78] Nature is the mentor that guides, encourages and teaches lessons because "Nature can produce complex structures even in simple situations and can obey simple laws even in complex situations." [79] The complexity of nature is no longer viewed as a combinatorial association of the ultimate, homogeneous and universal units composing matter. Rather it exemplifies the composites principle, a strategy of mixing heterogeneous ingredients in order to obtain the optimal combination of properties. [80] The lessons of nature concern not only the design of materials but also their repair, and their destruction thanks to recycling processes. Nature is the master of time. In this view, we cannot improve on, not even rival nature since she spent billions of years for designing and perfecting high-performance structures capable of sustaining life. Steven Boxer, a chemist from Stanford who uses proteins as transistors in electronic circuits, thus expresses his strategy: "We've decided that since we can't beat them (biomolecular systems), we should join them." [81] Materials designers have no choice but to start from the building blocks provided by nature - whether they be proteins, bacteria, micelles or colloids - in order to achieve their goal. In this perspective the relation between nature and artifice is one of partnership, not of domination. This attitude, however, does not limit the Promethean ambitions of materials scientists. Some of them develop grand technocratic views, based on this cooperation with nature.

Conclusion

In conclusion, the emergence of MSE was neither the outcome of the growth of academic disciplines nor a response to "man's needs". Materials science was created thanks to a close cooperation between different kinds of agents: governmental agencies and universities working together to outline a new interdisciplinary research field. Then, while on the academic stage MSE became a discipline with its own curricula, industrial companies took over in the 1970s and 1980s.With the reorientation of the field towards processing and performance, a conceptual framework emerged along with new strategies for designing materials.

This brief historical survey reveals two major characteristics of MSE.

From the outset, from x-ray diffraction to the AFM, instrumentation acted as a driving force. Not only instrumental techniques opened the way to microstructures and nanostructures, but also they helped create a scientific community. Whatever their different topics, aims and cultures, the users of instrumental facilities meeting in specific buildings embody materials science.

Industrial and academic researches are equally important. Materials scientists do not only present their results in publications but also in patents. They are equally active on the academic stage - writing articles, books and textbooks to promote the field in engineering - and on the industrial stage where they often hold positions in companies or start-ups.

The recent evolution of MSE suggests a booming field of researches moving in many different directions. The shift from the microscale to the nanoscale acted as a diverging rather than converging force. The field's coherence can no longer be secured through the leadership of physics and metallurgy. Although chemistry played a key role in the dynamics of MSE, it did not create a new coherence. Other disciplines, computer science and biology in particular, also brought their own techniques, interests and traditions into the field. This ongoing process of hybridization shapes the identity of MSE and distinguishes it among other recent fields such as computer science or cognitive science.

Finally a major characteristic of MSE is that its dynamics is such that the future is still wide open. Despite repeated efforts to stabilize the field as an autonomous discipline, MSE remains out of equilibrium, with temporary unstable states of equilibrium from time to time. Far from gaining more and more stability the field is really exploding rather than imploding. Maybe, like physical chemistry, materials science will become a fundamental hybrid discipline with its its own concepts and a strong identity. Or, like radioactivity, it will be absorbed into physics, chemistry and biology. Maybe the promotion of an interdisciplinary science will eventually feed back into and reinforce the disciplinary structure of conventional disciplines such as physics and chemistry.

Appendix: Textbooks surveyed

• Allen, Samuel M., Thomas Edwin L., Structure of Materials, New York, Wiley & Sons, 1999
• Jacobs James A., Kilduff Thomas F., Engineering Materials Technology (Structure, Processing, Properties), Englewoof Cliffs, Prenctice Hall, 1^st ed. 1985, 2^nd ed. 1994, 3^rd ed. 1997
• Jastrzebski, Zbigniew D., The Nature and Principles of Engineering Materials, New York, Wiley & Sons, 1^st ed. 1959; 2^nd ed. 1976
• Jones, S.W. Materials Science, London, Butterworth, 1970
• Kayser Carl A., Materials Science in Engineering, Columbus Ohio, Ch Merill Publishing, 1^st ed. 1956, 2^nd 1968
• Newey, Charles and Weaver Graham (eds.), Materials. Principles and Practice, Milton Keynes: The Open University, Butterworth, 1990
• Pask, Joseph A. (ed.) An Atomistic Approach to the Nature and Properties of Materials, New York, Wiley & Sons, 1967
• Richards, Cedric W., Engineering Materials Science, San Francisco: Wadsworth Publishing, 1961
• Ralls, Kenneth M., Courtray, Thomas H., Wulff John, Introduction to Materials Science and Engineering, New York: John Wiley & Sons, 1976
• Smith, Charles O., The Science of Engineering Materials, Englewood Cliffs: Prenctice Hall, 1^st ed. 1969, 2^nd ed 1977, 3^rd ed. 1986
• Smith, William F., Principles of Materials Science and Engineering, New York: McGraw-Hill, 1986
• Thornton Peter A, Colangelo Vito J., Fundamentals of Engineering Materials, Englewood Cliffs, Prenctice Hall, 1985
• Tottle, C.R. The Science of Engineering Materials, New York: Chemical Publishing, 1966.
• Van Vlack, Lawrence H., Elements of Materials Science: An Introductory Text for Engineering Students. Addison Wesley Publishing, 1^st edition, 1959; 2^nd ed. 1964; 3^rd edition 1974
• Young James F., Shane Robert S., Materials and Processes, New York & Basel: Marcel Dekker, 1^st ed. 1944, 2^nd ed. 1954, 3^rd ed. 1985