Professional Documents
Culture Documents
Editor
VOLUME 144
TECHNOLOGICAL
DEVELOPMENT
AND SCIENCE IN THE
INDUSTRIAL AGE
New Perspectives on the
Science-Technology Relationship
Edited by
PETER KROES
and
MARTIJN BAKKER
Department 0/ Philosophy and Social Sciences,
Eindhoven University 0/ Technology, The Netherlands
PREFACE vii
v
PREFACE
PETER KROES
M A RT IJ N BAKKER
vii
P. KROES AND M. BAKKER
INTRODUCTION:
TECHNOLOGICAL DEVELOPMENT AND SCIENCE
Almost two decades ago, in March 1973, the famous Burndy Library
conference on The Interaction 01 Science and Technology in the
Industrial Age took place. 1 One of the explicit aims of this conference was
to bring together historians of science and technology to discuss the
mann er in which science and technology influence each other. In his
Foreword to the proceedings of the conference, Robert E. Kohler
remarked that a better insight into this matter required " a breaking down
of barriers between the two groups of historians" .2 Indeed, an adequate
study of a topic such as the interaction between science and technology
requires a trespassing of disciplinary boundaries, or even better the
abolition of those boundaries . Kohler already saw a hopeful sign that
something like that was happening: at universities historians of science
and technology were increasingly placed in one and the same department.
According to Kohler, this institutional development would insure that
"the history of science will cease once and for all to be the history of
isolated pieces of intellectual software, or the history of technology to be
the history of isolated pieces of hardware". 3 Twenty years later, the
cooperation between historians of science and technology seems indeed to
have intensified, but the disciplinary boundaries are still operative."
The basic issue at the conference was the relationship between
scientific knowledge and technological devices. Most participants agreed
on the inadequacy of the "technology is applied science" point of view,
that is of models postulating a linear, sequential path from scientific
knowledge to technological invention and innovation. In this way, science
is considered to be the " prime mover" of technology. From such models
it is but a small step (by adding the widely accepted postulate of an
internal developmentallogic for science) to some form of technological
determinism. Instead the "relative autonomy" of technology and techno-
logical development with regard to science was generally acknowledged
and stressed.
Apparently, semantic issues concerning the meaning of notions like
"science" , "technology", "engineering" and "applied science" heavily
dominated the Burndy Library conference .' The problem of distinguish-
P. Kroes and M . Bakker (eds.), Technological Development and Science in the Industrial
Age, 1-15 .
© 1992 Kluwer Academic Publishers.
2 PETER KROES AND MARTIJN BAKKER
that it had unique qualities to bring to the task, and each made claims on
national resources to increase its numbers and activities". 21 Apparently,
engineers and scientists considered it more opportune to emphasize their
own, specific abilities and resources for solving a given problem than to
point out their common and complementary means for dealing with it.
What we see here could be characterized as a quarrel between engineers
and scientists about whether the detection of submarines is a "tech-
nological" or "scientific" affair. 22 The interests at stake, of course, are
very high, for the answer to this question has far-reaching consequences
for the funding of research in this field. But the simple fact that science
and technology had become competitors for the solution of a certain
problem illustrates the degree to which modern science and technology
had converged since the emergence of modern science. It is also illustrated
by the fact that, around the turn of the century, industry recognised the
technological impact of science and, undisturbed by the rhetoric of
engineers and scientists, started to lure scientists into its laboratories.
The Greek idea - that science and technology are essentially distinct
activities - was not only part of the 19th century rhetoric; it still is very
influential today, and has not been replaced or adapted. Time and again
this idea is endorsed by engineers and scientists themselves, more on
ideological grounds, it seems, than on the basis of a careful and critical
analysis of what engineers and scientists do. But the historiography of
science and technology has also contributed to the continuation of this
idea. Although many historians of technology admit that the distinction
is problematic, it is nevertheless an integral part of most of the work done
in this field. This, of course, is a rather dangerous situation, for it may
lead to an uncritical adoption of the self-image of scientists and engineers
and its associated rhetoric.
As an example of how implicit assumptions with regard to the science-
technology dichotomy may influence the historiography of technology, it
is worthwhile to pause for a moment at the notion of a "Second Industrial
Revolution". This notion is increasingly becoming part of the standard
vocabulary of historians of technology and is usually associated with a
revolutionary phase in the development of technology around the turn of
the century due to the (systematic) technologieal exploitation of scientific
knowledge. It refers to the emergence of a science-based technology and
is related to the creation of industrial research laboratories and the
employment of scientists in industry. Ernst Homburg has carefully
analysed the origins of this notion and how it was introduced in the
8 PETER KROES AND MARTIJN BAKKER
history of technology." The notion was first used in the fifties of this
century, when it referred to the technological revolut ion due to
automation and nuclear technology and its social impact. It was employed
in a "politic-programmatic"" way to argue for or against social changes
related to the ongoing technological revolution. In the beginning of the
sixties, it was transformed into a historical concept by Geoffrey
Barraclough and David Landes. According to Homburg the original
concept of a Second Industrial Revolution was based upon two
presuppositions which fitted very well with the ideology of the fifties with
regard to the relation between science and technology, viz. the idea that it
is possible to draw a coherent distinction between science and technology
and the linear science-technology-innovation model. He argues that in
the transformation of the notion of a "Second Industrial Revolution"
into a historical concept part of the ideological connotation and its
presuppositions were carried along without being subjected to a critical
examination. That is why, in his opinion, the use of this notion is
problematic, especially because criticism of the last decades on the
conception of science and technology as monolithic activities and on the
linear science-technology-innovation model has not resulted in a
reinterpretation of this notion.
Gur intention is not to argue that, in view of the above considerations,
any distinction between science and technology is obsolete and that
therefore quest ions about how science and technology influence each
other are senseless. We agree with Homburg that " fruitful historical
research into the relation between science and technology is certainly
possible and also desirable" ,15 on condition that new conceptions of
science and technology are developed . We have merely tried to indicate
that research in this field appears to be guided by an outdated conception
of science. In order to pose the right kind of questions it is necessary to
abandon the Greek distinction between science and technology, and to
search for areinterpretation of these notions such that a more fruitful
framework arises for analyzing the interaction between science and
technology. Such areinterpretation has, in our opinion, to be based
primarily on an analysis of what engineers and scientist actually do
instead of on what they claim they do.
A new promising approach to the old science-technology issue may
already be developing. Staudenmaier's analysis of TC's papers addressing
the science-technology issue has led hirn to the conclusion that in these
papers historians of technology are not so much dealing with the science-
INTRODUCTION 9
March 1992
The Editors
NOTES
1 The proceedings were published by Nathan Reingold and Arthur Molella in a special issue
J lbid.
4 There are, for instance, no major international journals devoted to the history of science
and technology. A superficial bibliographical search resulted in the following list of journals
explicitly devoted to the history of science and technology : Historical Studies in lrish Science
and Technology, Centaurus, Scientia Canadensis and the Dutch Gewina .
s See Reingold and Molella (1976), p. 625.
• See in particular Otto Mayr's contribution to the proceedings .
7 According to Kohler the whole issue is basically ideological; Reingold and Molella (1976),
p.621.
• For more information about Hindsight and Traces, see Rosenberg [1982), pp. 207-216 .
9 Staudenmaier (1990), p. 718.
12 See A. Sarlemijn, ' Science and Technology - Present Relations in Historical Perspective',
'4 They do not give up the distinction between science and technology altogether, but reduce
it to a purely social construction. See, for instance , T.J . Pinch and W.E . Bijker, 'The Social
Construction of Facts and Artifacts: Or How the Sociology of Science and the Sociology of
Technology Might Benefit Each Other', in Bijker (1987), pp , 17-50, and Bijker (1990).
" For the following, see Aristotle, Nicomachean Ethics, 1139b-1I4I b ; l177 s_l177 b ;
Metaphysics, 982b ; 1025b-1026 s; Eudemian Ethics, 1216b •
I. These four causes are : the material. formal final and efficient cause.
14 PETER KROES AND MARTlJN BAKKER
17 In antiquity, mechan ics was considered to be an art; it was transformed into a science in
22 According to the British physicist Ernest Rutherford submarine detection was " a problem
2. In his review article Staudenmaier (1990), p. 718, qualifies Vincenti 's work as " the last
3 S Ju st as the philosophy of science did in the sixties and seventies for analyzing the way
38 The conference was held on 6-9 November 1990 at Eindhoven University of Technology
and attracted considerable attention; about seventy scholars attended the conference,
coming from The Netherlands (26), the United Kingdom (13), the USA (9), Germany (9), the
Scandinavian countries (6), France (4), Spain (3), Israel (I) and Australia (I). Nine "invited"
papers (published in this volume) and fort y "contributed" papers were presented . Most of
the participants (about 50) were historians of technology; it was in fact no suprise that the
philosophers of technology were in the minority, since the history of technology is much
better institutionalised as a discipline than the philosophy of technology.
INTRODUCTION 15
REFERENCES
Bijker, W.E. et al. (eds.): 1987, The Social Construction of Technological Systems,
Cambridge Mass .: MIT Press .
Bijker , W .E .: 1990, 'Do Not Despair : There Is Life after Constructivism', Kennis en
Methode, No . 4, pp . 324-345 .
Durbin, P .T . (ed.) : 1980, A Guide to the Culture of Science, Technology, and Medicine,
New York : Free Press .
Durbin, P .T . and Rapp, F. (eds.): 1983, Philosophy and Technology, Dordrecht : Reidel.
Homburg, E.: 1986, 'De 'Tweede Industriele Revolutie', Een problematisch historisch
concept', Theoretische Geschiedenis 13(3), pp . 367-385 .
Hughes , T .P .: 1989, American Genesis, Penguin Books .
Huning, A.: 1974, Das Schaffen des Ingenieurs; Beiträge zu einer Philosophie der Technik,
Düsseldorf: VDI -Verlag .
Lenk , H . and Moser, S. (eds.) : 1973, Techne, Technik, Technologie, Pullach bei München:
Verlag Dokumentation .
Mitcham, C. and Mackey, R. (eds.): 1972, Philosophy and Technology, New
York /London: The Free Press .
Rapp, F. (ed.): 1974, Contributions to a Philosophy of Technology, Dordrecht: Reidel.
Rapp , F.: 1981, Analytical Philosophy of Technology, Dordrecht: Reidel.
Reingold, N. and Molella, A. (eds.): 1976, 'The Interaction of Science and Technology in
the 1ndustrial Age' , Technology and Culture 17(4), pp . 621-742.
Rosenberg , N.: 1982, Inside the Black Box: Technology and Econom ics, Cambridge:
Cambridge UP .
Sarlemijn, A. and Kroes , P . (eds .): 1990, Between Science and Technology, Amst erdam :
North-Holland.
Staudenmaier, J .M .: 1985, Technology's Story tellers, Cambridge Mass .: MIT Press .
Staudenmaier, J .M .: 1990, 'Recent Trends in the History of Technology', American
H istoricol Review 95, pp. 715-725.
WALTER G . VINCENTI
PRELIMINARY REMARKS
the general shape and arrangement commonly agreed to best realize the
operational principle . It becomes arrived at and agreed upon (perhaps
implicitly) in the formative years of the device. Though less strictly
determinative than the operational principle, it too constitutes a given in
normal design. The normal configuration of the airplane over most of its
existence has been the engine-forward, tail-aft biplane (mainly) until the
1930s and monoplane thereafter. This arrangement was arrived at ,
primarily in France, before World War 1. Since then and until recently, it
has only rarely occurred to designers that an airplane could or should be
arranged differently.
The operational principle and normal configuration together define
the normal design of a device. They form the basis for the " accepted
tradition" that Constant spoke of in his characterization of normal
technology (though he did not allude to them specifically). What we may
call radical technology then involves change in either the device's accepted
configuration or its operational principle. In the latter case, the configur-
ation too must change - a new and appropriate normal configuration will
have to be arrived at once the new operational principle has been
established. Considerations like these underlie our ideas of invention and
innovation. The design component of radical technology I shall refer to as
radiea/ design.
Radical and normal design constitute, of course, two limits of a
spectrum more than they do a dichotomy. The degree of radicalness is
obviously greater when the operational principle is changed than when
only the normal configuration is altered. And either the principle or
configuration can be modified in some degree rat her than completely
replaced . As usual in these matters, the distinctions are sometimes
difficult to make . They do, however, provide a useful tool for analysis.
The operational principle and normal configuration afford clear
instances of engineering, as against scientific, knowledge. They may be
analyzed and in some cases even triggered by the findings of science; they
are in no way, however, contained in or dictated by those findings. As
stated by Polanyi, "The complete [by which he meant scientific]
knowledge of a machine as an object teils us nothing about it as a
machine ." The operational principle and normal configuration call for
added acts of insight and experiment, usually by inventors or engineers.
The foregoing statements accord with a perceptive distinction by
Herbert Simon in his The Scienees 01 the Artificial. 6 As stated by Simon,
the natural sciences deal with how things are. Engineering design, like all
22 WAL TER G. VINCENTI
design, deals with how things ought to be. The operational principle and
normal configuration of an airplane, for example, are in no sense
knowledge of how a flying machine innately is; they are knowledge of how
a particular kind of flying machine ought to be to serve certain purposes.
They provide instances par excel/ence of engineering (as distinct from
scientific) knowledge . As I point out in my book, all knowledge for
engineering design, including the necessary attendant knowledge of how
things are, serves finally to implement how things ought to be. That, in
fact, is the criterion for its usefulness and validity . (Similar remarks would
apply to engineering knowledge for production and operation.)
Still another idea, obvious also from everyday design, dominates the
epistemology of engineering . In design activity (as also in production and
operation), the practical use to which knowledge is put is paramount. The
necessity that the knowledge be useful both motivates the knowledge and
determines its nature. When need for its use disappears - as when , say, the
reciprocating steam engine was replaced by the steam turbine - the
knowledge becomes neglected and, for practical purposes, forgotten. In
engineering knowledge, practical use is of the essence. This requirement is
obvious, but in the preoccupation with scientific knowledge it tends at
times to be forgotten .
This is not to say that that use is unimportant in science; it is important,
but in a very different way. Science I take to be a search for understanding
of observable phenomena, such understanding itself constituting a form
of knowledge. This search for understanding or knowledge is open-
ended . In science, understanding hence finds use in generating more
understanding - or, equivalently, knowledge finds use in generating more
knowledge. As put somewhat differently by historian Hugh Aitken,
"Most of the informational output of science - the new knowledge
generated - is channelled back into science itself. "7
In the foregoing paragraphs, knowledge figures in engineering activity
as a means to a utilitarian end , in scientific activity as a means to more
knowledge (and hence in a sense as an end in itself). This difference
conforms with Simon's distinction between science as dealing with how
things are and engineering design with how things ought to be. Engineers,
of course , also use knowledge to generate more knowledge. Such use,
ENGINEERING KNOWLEDGE, DESIGN , AND HIERARCHY 23
however, is less requisite than its use in designing artifacts. The essential
point for present purposes is this : By employing knowledge in design as
weIlas in generating more knowledge - that is, for two uses instead of one
- activity in engineering is distinctively asymmetrie from activity in
science.
This asymmetry has an epistemologieally important result. In science,
where knowledge is used to generate more knowledge, the institutional
loci of generation and use are overwhelmingly one and the same - the
scientifie research laboratory. In engineering , generation and use may
also take place together in the engineering research-and-development
laboratory. As numerous studies testify, however, distinguishing between
scientific and engineering knowledge proves difficult if one examines
them, as commonly done, in the context of these similar and overlapping
research institutions. In engineering, where practical application is
paramount, however, the situation is in fact different. There knowledge
finds its main and defining use in a separate institution - the industrial
design office. Distinction between the two kinds of knowledge thus
becomes operationally possible . To identify the nature of engineering
knowledge in terms of its use - which, by definition, should be the starting
point of any epistemology of engineering - we can look to see what
knowledge is employed in the design office. For scientifie knowledge we
do the same in the scientific laboratory. Failure to notice this possibility
may be one effect of approaching the history and epistemology of
engineering with biases from the history and epistemology of science. An
outstanding historian of technology, whose work I have used and admire,
despaired of distinguishing between engineers and scientists when he
perhaps fell into this trap by seeing both groups as working "in
laboratories of like appearance" (emphasis added) ." Focusing on normal
rat her than radieal technology helps avoid this pitfall.
To visualize the foregoing ideas, I find Fig. I helpful. The key notion
here is the differentiation between knowledge used by scientists and
engineers (solid-line boxes) and knowledge generated by those
communities (solid-line band) . At the level of use, distinction between
knowledge in science and engineering can be made as explained above;
representation by separate boxes is therefore realistic . At the level of
generation, as likewise observed, such distinction is problematic;
representation by a spectrum is thus more credible, with knowledge
generated by scientists toward the left and by engineers toward the right.
The activities that produce this knowledge then appear as a corresponding
24 WAL TER G . VINCENTI
t t t t t
I
I Knowledge genereted Knowledge genersled
I by selenllsls by engineer.
I
I
I
I Knowledge used by
selenI1sts
(to generete
Knowledge used by
englneers
(to design srtlfsets end,
more knowledge ) sec:ondarily, generate
more knowledge)
_____ t- _
I I
I I
I Design setlvlty I
I I
L I
spectrum (dashed-Iine band), with purely scientific activity far to the left
and purely engineering activity correspondingly to the right . The activities
of hybrid individuals or of groups working together in laboratories then
fall in the dashed area in the center. It is this spectrum of activities that
generates and determines the spectrum of knowledge directly below.
Scientists and engineers for their respective use then draw on this
generated knowledge as needed from anywhere along the spectrum. The
same piece of knowledge may thus be found identically (or appropriately
modified) in both solid-line boxes. If so needed, an item mayaiso show up
in one box but not in the other. As indicated in the diagram, the
knowledge used by engineers finds employment primarily in design
(dashed-Iine box) and secondarily in the knowledge-generating activities
of engineers. Knowledge used by scientists goes preponderantly toward
the single use of generating more knowledge. The essential asymmetry
between engineering and science thus appears clearly in the diagram. In
ENGINEERING KNOWLEDGE, DESIGN, AND HIERARCHY 25
DESIGN HIERARCHY
EPISTEMOLOGICAL RESUL TS
Studying normal aetivity at the lower levels, I find, has its dividends.
Kinds of engineering knowledge present themselves that might otherwise
be overlooked. My intention here has been to clarify this underlying
historiographie and epistemological orientation of my work. I have spaee
here to mention only a few results, more to illustrate the nature of the
findings than to provide an adequate and eomplete exposition. The reader
may reeognize some of them from my published articles; they are
elaborated in my book . (Generalized eategories into which the examples
fall are italicized in the following numbered aeeount.)
(I) The operational principle and normal eonfiguration have already
been pointed to as basic concepts in normal design. (2) To help translate
these concepts into a concrete device, engineers devise criteria and
specifications; a sophisticated example here is the specifications used by
ENGINEERING KNOWLEDGE, DESIGN, AND HIERARCHY 27
system. The potential for social influence - even social determination or,
to use recent scholarly parlance, "social construction" - clearly exists at
all times. Technical constraints, however, have their base at least partly in
physical laws and requirements. To circumvent them is frequently
impracticable and in some cases physically impossible . When social
influences and technical constraints conflict, the constraints must often
prevaiI. Social influence would thus be expected to appear least where
elevation of our surface is highest, that is, in normal design at the
lowermost levels of hierarchy. It may be expected most at the lowest
elevations, to wit, in radical design at the upper hierarchical levels. I
submit that such theoretically argued outcome is by and large the fact.
As a case in point, consider again an aircraft and its landing-gear. At
the levels of project definition and conceptual design - that is, at the top
of the hierarchy - social factors obviously have wide scope for influence .
Even in normal design, the performance, size, and general layout of the
vehicle follow more or less directIy from the needs of the sometimes
contending social constituencies it serves. In radical design, where the
basic configuration or even principle of operation may be up for grabs ,
the connection is even closer. Across the full design-type spectrum, the
predominating social purpose and influence often leave their imprint in
the very appearance of the aircraft - a carrier-based fighter-bomber
visibly serves a different social purpose from a commercial four-engine
transport.
At the levels of component and subcomponent design, a different
situation obtains for both radical and normal design. As indicated earlier,
the problems of landing-gear layout and (especially) subcomponent
design are primarily technical problems amenable to and demanding
mainly technical solutions. The internal logic of the technical situation
takes over and leaves little room for direct social influence. The landing
gear and its subcomponents exhibit a great deal ofthe associated technical
purpose but very littIe, if any, social. (I do not regard the bare
requirement to get an airplane safelyon and off the ground as a
contextual influence, though it's obviously a social given.) The same
kinds of remarks can be made about other components and sub-
components of the airplane and, indeed, with regard to all complex,
hierarchically structured devices. Overall, the situation described in these
paragraphs reflects a crucial fact: At the upper levels of hierarchy, the
object of design functions typically as a component of a sociotechnical
system; at the lower levels it constitutes a component of a technical
ENGINEERING KNOWLEDGE , DESIGN , AND HIERARCHY 33
Stanford University
NOTES
I W.G. Vincenti, What Engineers Know and How They Know It: Analytical Stud ies from
4 E.W. Constant, The Origins 01 the Turbojet Revolution (Baltimore: Johns Hopk ins
1981), pp . 132-133.
7 H.G.J. Aitken, Syntony and Spark - The Origins 01 Radio (New York: Wiley
Interscience, 1976), p. 314.
• O . Mayr, 'The Science-Technology Relationship as a Historiographie Problem',
Technology and Culture 17(October, 1976), pp. 663-673, quotation from p, 677.
• Fuller discussion of a truncated version of this diagram appears in the book.
10 G. Ryle, The Concept 01 Mind (London: Hutchinson, 1949), pp . 27-32 .
SocioJogyofTechnology Might Benefit Each Other' , in W.E . Bijker, T.P . Hughes , and T.J.
Pinch (eds.), The Social Construction 01 Technological Systems (Cambridge , Mass.: MIT
Press, 1989), pp, 17-50 .
EDWIN T. LA YTON JR .
power ofthe head (or fall ofwater) and the square root ofthe horsepower.
For a centrifugal pump the specific speed varies inversely as the 3/4th
power of the head and the square root of the volume of flow.4 It is a
commonplace observation that seemingly minor changes in the shape of
an airfoil will change its performance (e.g. lift and drag) , and (less
obviously) that this performance is influenced by the shape of the flow
patterns in the surrounding fluid."
The change of observed behavior or properties of an artifact (e.g.
speed, strength, drag) with size are often called scale effects. Scale effects
were among the most important barriers to the development of a scientific
understanding of engineering artifacts. For most artifacts, engineers had
to understand elusive, often complex, dimensional relationships in order
to conduct model experiments or to develop accurate mathematical
theories. The struggle to und erstand these dimensional barriers to
scientific understanding played an important role in the historical
evolution of engineering. After centuries of struggles with elusive
dimensional problems, engineers and scientists by the 20th century were
led to the discovery (by the methods of "dimensional analysis") that
particular problems in engineering have an underlying dimensional
structure; once this structure was understood it enabled engineers and
others to gain a general scientific knowledge of artifacts similar in form
and substance to the scientific knowledge of nature developed by the
physical sciences. Metaphorically speaking, engineering knowledge was
freed from its jail of shape ."
The first recorded encounter between engineering and scale problems
arose after the invention of the torsion catapult in Classical Antiquity.
Starting from a well-constructed and successful catapult, a Hellenistic
engineer might very weIl have assumed that he could make an equally
good catapult with three times the power by simply building another
catapult three times the size of the first. For reasons that were not
obvious, attempts to scale-up catapults often failed; catapults built in
imitation of a model catapult but at a different scale often exhibited
unpredictable and undesirable changes in performance as compared with
the catapult used as a model. Since catapults were important new tools
of war, Hellenistic engineers sought to find a rule of design, such that
a catapult might be built at an arbitrary, predetermined size which would
demonstrate predictable performance. We are fortunate in having abrief
sketch of the early effort of Hellenistic engineers to und erstand and
control scale effects . It was written by a noted engineer, Philo of
38 EDWIN T . LAYTON JR .
Byzantium (c. 250 B.C .). According to Philo the HeIlenistic engineers:
. ..discovered that the diameter of the bore was the basic element, principle , and
measurement in the construction of artillery. But it was necessary to determine this diameter
not accidently or haphazardly, but by some definite method by which one could also
determine the proportionate measurement for all magnitudes on the instrument. But this
could not be done except by increasing or decreasing the diameter of the bore and testing the
results . And the ancients did not succeed in determining this magnitude by test, because their
trials were not conducted on the basis of many different types of performance, but merely
in connection with the required performance. But the engineers who came later , noting the
errors of their predecessors and the results of subsequent experiment s, reduced the principle
of construction to a single basic element, viz., the diameter of the circle that receives the
twisted skeins.'
Though the details are not completely c1ear, the ancient engineers
apparently used the method now caIled "parameter variation", an exper-
imental method long favored by engineers, though little used in the basic
sciences." This design effort, which amounted to a major engineering
development and research undertaking supported by HeIlenistic kings,
was successful in its immediate goal. 9
The HeIlenistic engineers understanding of dimensionality was limited,
however. They discovered no general relationships that would enable
engineers to predict the behavior of fuIl-sized prototypes by means of
experiments upon scale models. With the exception of the catapult, the
behavior of artifacts continued to change in unpredictable ways as they
were made larger or smaIler. The ultimate failure of engineers in the
ancient world to master the general problem of scale effects encountered
in experimentation with artifacts was expressed by Vitruvius, who wrote,
"In some machines the principles are of equal effect on a large and on a
smaIl scale; others cannot be judged of models." 10
These limitations persisted until weIl into the Nineteenth Century and
constituted a serious barrier to the development of a science or sciences of
engineering . Newtonian physics provided insights of enormous import-
ance, but the complexity of engineering reality defied even Newton's
reasoned simplifications.
enormous implications for engineering. But it took some time for the
technological promise of Newton 's science to be fully realized in
engineering practice. Eighteenth-century engineers, such as CharIes
Hutton (1737-1823), used Newton as a basis for engineering mechanics,
involving the ability to construct idealized mathematical models of
various artifacts. Newtonian science was cast into analytical form and
developed by French scientists such as Jean le Rond D'Alembert (1717-
1783), and this powerful analytical mechanics had a large influence on
engineering in France and Europe generally. It would be hard to
overemphasize the value to technology of having a vocabulary, a set of
correct scientific concepts, and symbolic means of representing techno-
logical systems . 11
Idealization was important in the interactions of engineering with
Newtonian physics . At one level, the fact that mathematical models based
upon Newtonian science were idealized was an advantage; it made for
simplicity and ease of calculation for approximate results . But at another
level, the idealizations built into Newtonian physics constituted a barrier
to an exact analysis of engineering artifacts. Thus attempts to use
Newtonian physics to create a theory for engineering ran into problems
that were not fully understood until the 20th century. To get a correct
theory, engineers had to take into account shape and other dimensional
properties left out of Newtonian science . It is notorious that Newton was
less than fully successful in dealing with certain other topics, notably fluid
resistance, and his successors left fluid viscosity and other things out of
their theories, though the associated phenomena (such a turbulence) were
inescapable parts of engineering practice.
Clearly, a consistent mathematical model of physical reality.must
incorporate rules to ensure independence of scale, since otherwise the
mathematical model derived from physical theory might apply at one
scale, but not at all scales. Such a science would lack general validity.
Newton did in fact create a physics that was independent of scale . That
is, his theories were successful in predicting the behavior of a great range
of objects regardless of scale, as long a certain !imitations were
understood . Newton analyzed the scale problem and concluded that his
physics was valid at all seales in Proposition 32 , of Book Two of his
Principia, He considered the case where there are two systems of bodies
composed of an equal number of particles, where nothing is postulated
about their relative scale relationships. He then showed that if these two
systems are similar in geometrie relations, in motions , and in forces and
40 EDWIN T . LAYTON JR.
if they met certain other conditions (e.g. they are not in contact with one
another), then their behavior, and hence the laws of Newton's physies
would be independent of scale. This critieal proposition reads in part:
Suppose two similar systerns of bodies consisting of an equal number of particles , and let the
correspondent particles be similar and proportional, each in one system to each in the other,
and have a like situation among themselves , and the same given ratio of density to each
other; and let them begin to move among themselves in proportional times , and with like
motions (that is, those in one system among one another, and those in the other among one
another). and if the particles that are in the same system do not touch one another, .. .I say
that the particles of those systems will continue to move among themselves with like mot ions
and in proportional times."
under the conditions of actual practice."21 This was , in fact, what was
done at the early textile center, LoweIl, Massachusetts, where James B.
Francis used only full scale turbines and other apparatus for his famous
hydraulic experiments conducted in the middle years of the 19th
century." Francis was in the forefront of engineers who were developing
experimental methodology and instrumentation which permitted an exact
comparison between theory and observation. Francis found that math-
ematical theories developed for engineering were not adequate guides to
engineering practice, a fact which he blamed upon idealization. For
example, Francis developed a weir formula to measure the flow of water
through reetangular openings . He was led to this successful investigation
by the inaccuracies of existing formulas . As he commented: "The result,
however, of these numerous labors, is far from satisfactory to the
practical engineer. On a careful review of all that has been done, he finds
that the rules given for his use , are founded on the single natural law
governing the velocity of fluids, known as the theorem of Torricelli;
omitting, in consequence of the extreme complexity of the subject, all
consideration of many other circumstances whieh it is weIl known,
materially affect the flow of water through orifices. "23
Francis found in the case of the turbine that mathematical theories
were no more accurate in dealing with the properties of these artifacts
than were experiments with scale models. The unreliability of theory led
Francis, to compile empirical tables giving, for turbines whose diameter
varied by one foot increments from 2 to 10 feet , the quantity of water
discharged, the horsepower and the rotary speed in revolutions per
minute, for turbines under heads varying from five to forty feet , at
increments of one foot. His data revealed serious errors in the available
theories, in rotor speed, in the rate of discharge, and in the power
developed. He commented that, "The turbine has been an object of deep
interest to many learned mathematicians, but up to this time, the results
of their investigations, so far as they have been published, have afforded
but !ittle aid to the Hydraulie engineer. "24
GRAPHICAL SOLUTIONS
The first, geometric similitude, was the first one to be fuIly achieved
through the descriptive geometry of Gaspard Monge (1746-1818), first
published in 1795. This was an important extension of Renaissance
projective methods and it was itself a fundamental engineering science
and also the starting point for many other important developments in
engineering science ." As it solved only one of the criteria for complete
similarity, scale effects remained and engineering was not freed from
dimensional constraints. Nevertheless geometric methods served
engineers amazingly weIl. Along with educated intuition and experience,
engineers managed to use graphics as a framework for profound scientific
advances in engineering. Graphics became a way ofthinking and knowing
as weIl as a means of communication, which has persisted until after the
middle of the Twentieth Century.
In the 19th century there was a strong tendency for the engineering
sciences to be expressed graphicaIly and mapped into descriptive
geometry, as, for example, in the cases of the graphical statics of Carl
Culmann, the circle diagrams of Otto Mohr, 27 the velocity triangles of
1ulius Weis bach and others for turbines, 28 and the indicator diagrams and
thermodynamic cycle diagrams for heat engines which began with the
work of lohn Southern and lames Watt. 29 Francis attempted to develop
a graphical method for the design of turbines; it was not completely
successful, but it revealed important facts that analytic theories failed to
reveal."
Indeed, engineers developed an entire sub -branch of graphics which
dealt with the graphical solution of equations which, in most cases , could
ha ve been solved equaIly weIl by standard mathematical methods.
American and also many British engineers came to feel comfortable with
geometry and to distrust analysis. lohn Perry, though a pupil of Lord
Kelvin, was an early mechanical engineering educator in Britain who
developed graphical methods for teaching the calculus to engineering
students." Methods of graphical solution of systems of equations are still
employed in engineering."
The success of graphical methods of analysis based on geometry was
probably due in part to the fact that geometry has similarity relations built
into its very fabric. We usuaIly think of the angles of triangles as measured
in degrees or radians. But if they are considered as the ratios of sides of
similar triangles, as in trigonometry, then they are also dimensionless
quantities. In general, a ratio of numbers with the same dimensions has
the effect of canceling out the dimensions. Since aIl mathematical
DIMENSIONALITY AND ENGINEERING SCIENCE 45
Though specific systems of units vary, as between the English and S.I. (or
International System) of units, for any given problem a set of
fundamental dimensions will be selected, such as mass, length, time, and
temperature, which are independent of the particular systems of units
employed. The number of fundamental dimensions is usually kept as
small as possible for greater simplicity. From the fundamental
dimensions, derived units can be constructed, such as force, energy, and
power.
To illustrate the principle of dimensional homogeneity, we can imagine
an equation as consisting of two or more terms that are to be added
together. Even though the entities which combine to form each product
may be different, the products to be added must have the same dimensions
raised to the same index. We can see velocity as the ratio of length and
time or [L] divided by [T]. For greater simplicity we avoid ratios and
express the dimensions as products raised to a certain power. Velocity is
therefore expressed the product [L] multiplied by [T-I ] . Since length,
time, and mass constitute an important set of fundamental dimensions,
they are often grouped together. In the above expression mass can be
expressed by capital M raised to the power zero.
We can express a hypothetical relation in which all of the products or
terms to be added have the same indices of the same fundamental
dimensions, for example, we find that the dimensions of all the terms can
be expressed by the dimensional product [MI] [U] [T-2 ] , where M is mass,
L is length, and T is time. These are, for example, the dimensions of
energy (and also of work and torque). If all the terms of an equation have
these same dimensions (after cancellation and simplification of the units
of dimension), the equation is dimensionally homogeneous.
If, however, we have an equation in which the terms to be added do not
have the same indices of the same dimensions, such an equation is said to
be inhomogeneous. We may suppose all but one of the terms in an
equation have the dimensions and their exponents just noted (for energy),
but that the term to be added to them has different indices of the
fundamental dimensions . For example, we can imagine a second term to
be added to the first in which the indices of dimensionality are different,
say those of velocity, [U] [MO] [T-I ] . In this case the equation is not
dimensionally homogeneous. Such an equation would be the symbolic
equivalent to adding kinetic energy to velocity, an operation having no
physical meaning.
Dimensional analysis was devised by Joseph Fourier as a byproduct of
DIMENSIONALITY AND ENGINEERING SCIENCE 47
The theory of heat will always attract the attention of mathematicians, by the rigorous
exactness of its elements and the analytical difficulties peculiar to it, and above all by the
extent and usefulness of its applications; for all its consequences concern at the same time
general physics, the operations of the art s, domestic uses and civil economy .3.
The specific elements which in every body determine the measurable effects of heat are three
in number, namely the conductibility relative to the atmospheric air, and the capacity fOT
heat. The numbers which express these quantities are , like the specific gravity, so man y
natural characters proper to different substances."
DIMENSIONALITY AND ENGINEERING SCIENCE 49
in 1959, gave separate lists of properties and terms used for heat-
exchanger calculations, for thermal radiation, and for mass transfer.
Whereas Fourier had three "specific elements", heat capacity and two
measures of heat conductibility, later elaborations of heat transfer.
particularly convective heat transfer, had to take many more specific
elements onto account. In Jakob's Heat Transfer published in two
volumes in 1949 and 1955, there are nearly eleven pages devoted to
nomenclature. The 287 definitions of terms include some 59 which specify
specific properties of materials, energy fluxes, or characteristic lengths. In
addition Jakob listed 12 dimensionless parameters used in heat transfer."
In Eckert and Orake's Heat and Mass Transfer of 1959 there are five
pages devoted to nomenclature. Eckert and Orake list 55 specific
properties, along with fourteen dimensionless parameters needed for heat
transfer. The empirical richness of the field is further suggested by Eckert
and Orake's tables of property values given in an "Appendix of Property
Values".45
Enumerations of specific properties are inherently subjective. Indeed,
a certain design subjectivity is built into every engineering science. Some
specific properties may be divided into several related terms because of the
different machines or structures to which the theory is applied. In other
words, the number of "specific properties" depends on human goals and
the state of the art as weil as upon nature.
As engineering sciences have become more general and abstract, the
linkage with specific contexts was, inevitably, less weil defined . The
classification schemes adopted by contemporary engineering scientists in
the field of heat transfer seldom make the specific design context
completely clear. Eckert and Orake's reference to heat exchangers is
exceptional. Eckert became interested in heat transfer as an offshoot of
his interest in power plant combustion. In particular, he faced problems
of radiant heat transfer in the design of steam boilers fueled with
pulverized coal. In the index to his book Eckert makes no reference to
steam boilers or to coal. Eckert does, however, make brief references to
specific contexts in the text. The burning of pulverized coal in steam
boilers raised specific issues, some centered on "Beer's Law". The actual
design context, while not very explicit in the book, was very clear in
interviews of Eckert conducted by Layton and Goldstein.:"
As an analysis of the textbook by Eckert and Orake makes clear, the
proper context for understanding the engineering sciences is that of the
design of artifacts, though considerations internal to science are also of
DIMENSIONALIT Y AND ENGINEERING SCI ENCE 51
In the first quarter of the Twentieth Century a few pioneers saw possibil-
ities for the unification and systematization of engineering sciences by
means of dimensionless parameters. The engineering theory of the
turbine, for example, remained unsatisfactory because certain key terms
could not be calculated accurately by the application of mechanics,
notably the rotor speed and the amount of water used. The first of these,
the "specificspeed", was discovered by Rudolph Camerer in 1905. 62 He
discovered the capacity coefficient and other dimensional fundamentals
which enabled him, with a number of other insights, to unify and
systematize the theory ofturbines by 1914. 6 3 On further investigation, the
specific speed turned out to be determined by the shape and configuration
of certain key components of the turbine. For example, In the Pelton
impulse turbine the specific speed is determined by the ratio of the
diameter of the rotor to the diameter of the jet of water that propels it; in
a Francis reaction turbine the geometry of the rotor vanes is a critical
determinant of specific speed. In this context it is clear why a purely
mechanical theory failed to make accurate predictions of rotor speed."
Wilhelm Nusselt was one of the most important of the unifiers of heat
transfer theory. Nusselt had earned his doctorate in mechanical
engineering at the Technische Hochschule in Munich under Oskar
Knoblauch . Knoblauch trained three of the foremost founders of
modern, engineering heat transfer: Nusselt, Ernst Schmidt, and Max
Jakob. Schmidt was in turn the teacher of Ernst R .G .Eckert. 6 5
The study of convective heat transfer was one of the most important
tasks undertaken by these pioneers in heat transfer. After Ludwig
Prandtl's founding of modern fluid mechanics and boundary layer
theory, it became possible to write the equations of convective heat
transfer, but they could not usually be solved. As John Lienhard has
shown, Nusselt was a pivotal figure in the development of the theory of
convective heat transfer. He was the first to apply dimensionless
parameters outside hydrodynamics, and in doing so he helped to
illuminate the correlations among groups of parameters and consequently
their larger role in the formulation of engineering theories . In aseries of
DIMENSIONALITY AND ENGINEERING SCIENCE 55
that expressed important truths about the world. It turns out also that
these are the sort of truths that are of great value for engineers attempting
to understand the detailed behavior of matter. H.A. Becker has called
Kline's approach the "Pythagorean method" and also, synonymously,
"the method of ratlos" . By whatever name, Kline's approach is
fundamental in that it combines dimensional analysis with the physieal
analysis of the problem to provide a richer and more powerful tool for
engineers. A notable feature of Kline's work, as extended by Becker is the
emphasis upon physieally significant ratios of forces, dimensions, fluxes,
and the like. Kline's work is a remarkable synthesis of much of the
concepts and procedures related to dimensional analysis, put into a more
meaningful physical context. 75
The result of the research by Kline and others building upon
dimensional analysis has been a much deeper understanding of the role of
dimensionality in engineering. In some early works the 7r terms were
referred to as "variables" . Kline has stressed the important point that the
terms whieh are combined to constitute dimensionless ratios are no longer
variables but parameters. This is one of the more significant results of the
investigations of dimensionality . These parameters reveal a fixed set of
dimensional relationships, expressed in the 7r terrns, which define the
dimensional structure of the problem (and indeed of a system of problems
sharing the same dimensional structure). Once this underlying
dimensional structure of the problem or system is properly expressed, it
does not vary; when it does vary one is going from one class of problems
with a partieular dimensional structure to another dass of problems with
a different dimensional structure."
H.A. Becker is an outstanding representative of those engineers and
others who have sought a deeper understanding of dimensionality by
means of the mathematieal theory of groups. This line of analysis led
Becker to examine related families of dimensionless parameters and to
classify them by the nature of the physieal entities whose ratios they
express: force, flux ratios, charge ratios, and geometrie ratios . Following
Kline, Becker arranged related clusters of these ratios on grids whieh
reveal many important physieal relationships and interconnections, in a
manner reminiscent of the periodie table of chemieal elements. It is
significant to note that this methodology places configuration in the
foreground. Becker wrote that, "the idea is akin to that of a system, but
with emphasis placed on the 'shape' of the system and the events that
occur within it. "77
58 EDWIN T. LAYTON JR .
University 0/ Minnesota
NOTES
contents of the Vaschy-Buckingharn 'Ir theorem, but a much bro ader defin ition is used here.
I ha ve been much influenced by Stephen J . Kline, Similitude and A pproximation Theory
DIMENSIONALITY AND ENGINEERING SCIENCE 61
(New York : McGraw-HiIl, 1965), who uses "dimensional analysis" in a broad sense. He
analyzes four distinct types of dimensional analysis (pp. 262-215). These are (I) the 11"
theorem (discussed below), (2) the use of governing equations, that is, normalizing the
governing differential equations to make them dimensionally homogeneous. This method
was pioneered by Lord Rayleigh; it is favored by physicists, and (3) the "method of
similitude" , (which is largely Kline's own creation). To this one should add the method of
"configurational analysis" which is built upon Kline's work, but which emphasizes the use
of group theory for the analysis of dimensional issues. (For configurational analysis, see
H .A. Becker, Dimensionless Parameters (New York : John Wiley, 1976), Vol. 3, pp. 10-11.
To these I add historical methods of analyzing dimensions such as that of Fourier, who
invented the first formalism of dimensional analysis. In addition, I follow Kline in including
the derivation of modellaws within dimensional analysis (pp . 2-7). Modellaws are often
called mies or principles of similitude (or similarity). Other works dealing with dimensional
analysis that I found particularly helpful include Percy W. Bridgman, Dimensional Analysis
(Carnbridge, Mass: Harvard University Press, 1921); W.J . Duncan, Physical Similarity and
Dimensional Analysis (London: Edward Arnold, 1953); H.E. Huntley, Dimensional
Analysis (London: McDonald, 1953); Henry L. Langhaar, Dimensional Analysis and
Theory 0/ Models (New York: John Wiley, 1951); and L.1. Sedov, Similarity and
Dimensional Methods in Mechanics (New York : Academic Press, 1959). On modelling
theory I found Langhaar especially useful. Several of these contain useful bibliographies,
notably Becker and Kline. Kline has comments on rnajor contributions to the literature
which can be found in the text . On the history of dimensional analysis I am deeply in the
debt of Enzo O. Macagno, 'Historico-Critical Review of Dimensional Analysis', J. FrankIin
lnst. 232(1971), pp. 393-394. Macagno 's study is a critical and informative history of
dimensional analysis; Macagno follows the more common practice of restricting the
meaning of "dimensional analysis" to the Vaschy-Buckingham theorem and closely related
matters.
7 Philo of Byzantium quoted in Morris R. Cohen and I.E. Drabkin, A Sourcebook in Greek
Technology and Culture 15 (January, 1974), p. 41; Waller G. Vincenti, What Engineers
Know and How they Know It, Analytical Studies from Aeronautical History (Baitimore:
Johns Hopkins, 1990).
' 0 Marcu s Vitru vius Polio, The Architecture 0/ Marcus Vitruvius Polio in Ten Books,
others, American edition edited by Robert Adrian based on 5th and 6th London editions,
1818). Volume 11 is an applied mechanic s for engineers which used a simplified Newtonian
notation. "Hutton" seems to have been very influential in the English speaking countries .
It went through a total of29 editions, of which nine were American, between 1798 and 1860.
(National Union Catalog 0/ pre-J956 Imprints 262, pp. 184-186 .) Though still used in
English speaking countries as late as 1860, Hutton's work was made obsolete by the French
tradition which employed Leibnitz 's notation and shifted mechanics from a geometric to an
analytical formalism . For a broad survey of mechanics which stresses the European and
French perspectives, see Rene Degas, Histoire de la Mecanique (Neuchatel: Editions du
Griffon, 1950).
12 Isaac Newton, Mathematical Principles 0/ Natural Philosophy and His System 0/ the
World, 2 vols., transl. Florian Cajori (Berkeley: University of California Pre ss, 1962),
Vol.Il , pp . 327-328 . See also Proposition 33, Theorem 26 including the several corollaries
(pp . 328-331) (hereafter: Newton, Principia).
IJ Newton, Principia, Book I, section 12 deal s with spherical bodies , section 13 extends the
analy sis to non-spherical bod ies, pp . 193-225. See particularly Proposition 88, Theorem 45,
" If the attractive forces of the equal particles of an y body be as the distance of the places
from the particles, the force of the whole bod y will tend to its center of gra vity; and will be
the same with the force of a globe, consist ing of similar and equal matter, and having its
center in the center of gravity." (I, p . 216).
'4 Newton's ontology also included his "geometrization of space" , and othe r th ings. These
issues have long been debated. Though now somewhat outdated, historically important
work s in understanding Newton 's ontology (or metaphysics) included Edwin A. Burtt, The
Metaphysical Foundations 0/ Modern Science: A Historical and Critical Essay (New York :
Harcourt Brace, 1925) and Alexandre Koyre, Etudes Galileennes (Paris: Hermann, 1939)
and the same author's , Metaphys ics and Measurements: Essays in Scientific Revolution
(London: Clapham and Hall, 1968).
.. Newton, Principia, I, pp. 328-336. Newton's awareness of the limits, and possibilities, of
his ontology is remarkable and evident in these propositions of fluid resistance . Newton did
deal with some of the effects of the mutual interactions of the particles of a fluid, in
Proposition 33, pp . 328-329 , where he shows that the centripetal and centri fugal forces by
which the particles of the system act on each other, can be reduced to simple considerations
of the for ce and matter as with a simple system by using the similarity principles developed in
Proposition 32. Thus Newton refuted Descartes , while not solving the problems of Fluid
Mechanics with which modern engineers must deal. The fact that Newton's physics was
associated with a particular ontology does not mean that it was invalid except with that
ontology. Thus it is possible (with later analytical tools) to expand Newton 's F = MA to get
the Navier-Stokes equation, and by considering the ratios of two such expression s for two
different systems to derive the key dimens ional concepts of dynamic and kinematic similarity.
Thi s is done, for example, by Max Jakob, Heat Trans/er, 2 vols. (New York : John Wiley,
DIMENSIONALITY AND ENGINEERING SCIENCE 63
Royale des Seiences (1767), pp . 495-503. Borda's tests were repeated and extended using a
towing tank (so that the model did not move through disturbed water) and the towing tank
became a significant advance in model testing methodology. Some other conflicts between
theory and experiment in the 18th century are summarized in Terry S. Reynolds, Stronger
Than a Hundred Men. A History 0/ the Vertical Water Wheel (Baltimore : Johns Hopkins,
1983), pp. 204-223, 237-241.
" Froude's relationship to Ferdinand Reech and the story of the law of similarity and the
dimensionless parameter named after Froude is complicated and will not be discussed in this
paper . As Rouse and Ince note, Froude did not originate the similarity rule (it was discovered
by Reech), and Froude never used the "Froude Number" . On the other hand he did anticipate
boundary layer analysis. See Rouse and Ince, History 0/ Hydraulics, pp. 154-155, 182-187.
I. John Smeaton, 'An Experimentallnquiry Concerning the Natural Powers of Water and
Wind to Turn MiIIs, and Other Machines, Depending upon Circular Motion ' , Phil. Trans.
Royal Soc. 51(1, 1759-1760), pp . 100-101. Smeaton's role in the experimental tradition in
engineering is discussed in Terry S. Reynolds, Stronger than a Hundred Men, pp. 218-233.
20 Reynolds, Stronger than a Hundred Men , pp. 223-226, 280-284.
which Francis measured turned out to be dependent upon dimension , as will be noted below
in discussing specific speed.
25 One can infer from Newton's Proposition 32 basic geometric, time, and force similarity
conditions, where land 2 stand for similar systems, and S, T and F are any lengths, time
intervals, and forces occurring in the similar systerns, the relation s S,/S2 = a, T ,/T2 = b , and
T ,/T2 = c, where a, b, and c are factors of proportionality. A common alternative for this
system of units is to consider length, time and force instead of mass.
2. Peter Jeffrey Booker, A History 0/ Engineering Drawing (London : Chatto & Windus,
1953), pp. 190-197,283-288. Both of these are systems of analyzing the stresses within a
framework by graphical means. Culmann, a mathematician by training , became convinced
tha t structural calculat ions could be best done graphically. See Carl Culmann, Die
graphische Statik (Zurich : Meyer und Zellers Verlag, 1869).
2. Julius Weisbach, A Manual 0/ the Mechanics 0/ Engineering (New York: John Wiley,
64 EDWIN T. LA YTON JR .
Trans. A. Jay Du Bois, 1880), II, pp. xliii-xlv, 222-226, 321-326, 355-358, 397-400.
2. D.S.L. Cardwell, From Watt to Clausius (Ithaca : Cornell University Press, 1971), pp.
80-83 .
30 Francis, Lowell Hydraulic Experiments , pp. 36-43 .
31 Larry Owens, ' Vannevar Bush and the Differential Analyzer: Text and Context of an
Early Computer' , Technology and Culture 27 (January, 1986), pp. 88-89. John Perry,
Calculusfor Engineers (London, 1897).
32 Philip Franklin, 'Oraphical Representation of Functions' , in Baumeister, Marks'
design in David Billington, Robert Maillart's Bridges. The Art of Engineering (Princeton :
Princeton University Press, 1979), pp. 5-7
34 As noted above, 1 am much indebted to Kline and follow his very broad definition of
Theory of Heat (New York: Dover reprint of the 1878 English edition, 1955).
36 lbid. , p. 6.
37 lbid. , 2. This is true for heat transfer where rates (that is time) are critical; but the later
(length), t (time), v (temperature), c (heat capacity) , and two heat transfer coefficients (in
Fourier's nomenclature) , "specific conductibility", K, and "surface conductibility", h (p.
130). See also Jakob, Heat Transfer, Vol. I, pp. 3-5 .
41 Macagno, 'Historico-Critical Review of Dimensional Analysis', pp. 391-402. See also
H.E. Huntley, Dimensional Analysis (London : McDonald, 1952), contains abrief history
(pp. 33-44) . For another historical study see Alton C. Chick, "The Principle of Similitude',
in John R. Freeman (ed.), Hydraulic Laboratory Practice (New York: the American Society
of Mechanical Engineers, 1929), pp. 796-797.
42 Max Jakob, Heat Transfer, Vol. I, pp. 3-5.
43 lb id., I, p. 5.
4. Eckert and Drake, Heat and Mass Transfer, pp. x-xiv, The append ices of properties are
DIMENSIONALITY AND ENGINEERING SCIENCE 65
•• Newton, Principia, Book Il. See also Rouse and Ince, History of Hydraulics, pp . 83-86.
•• George Gabriel Stokes, 'On the Effect of the Internal Friction of Fluids on the Motion of
Pendulums', in George Gabriel Stokes, Mathematicaf and Physicaf Papers, 5 vols .
(Cambridge: Cambridge University Press, 1880-1905), Vol. III , pp. 1-141, especially pp. 7-
17. The broader historical context of the Navier-Stokes equation is sketched by Rouse and
Ince , History of Hydraulics, pp. 193-200. Dominique Noir has argued that the Stokes law
of similitude for the drag of a sphere in a resisting fluid can be deri ved from Aristotle's
physics . See Dominique Noir , ' La prerniere loi de similitude de 1a rnecanique des fluides ',
Revue de l'Histoire des Seiences et de feurs Applications 25 (1972), pp. 271-274.
50 Hermann von HeImholtz, ' Über ein Theorem, geometrisch ähnliche Bewegungen
flüssiger Körper betreffend , nebst Anwendung auf das Problem, Luftballons zu lenken',
Monatsber. Kön. Preuss. Akad. Wiss. Berlin 188 (1873), pp . 501-514. This is a rem ar kable
for eshadowing of dimensional analysis in which he anticipated the later similarity
considerations asso ciated with the Froude, Mach, and Reynolds criteria; there is no
evidence, however, that he anticipated the existence or the role of dimensionless parameters.
For the more modern modelIing ofthe Helmholtz's problem see Edgar Buckingham, ' Model
Experiments and the Forms of Empirical Equations', Trans. Am. Soc. Mech. Engineers 37
(1915), pp. 273-277.
51 Lord Rayleigh (John William Strutt), ' T he Principle of Sirnilitude', Nature 95 (1915), pp .
66-68, 644 is a very remarkable virtuoso performance. Kline has a remarkably clear and
insightful chapter in which he analyzes the cornparatives ad vantages and drawbacks of
Rayleigh's method with that of Buckingham and others. (See Kline, Similitude and
Approximation Theory, pp . 262-215 .) For a succinct statement, see also Murdock.
'Mechanics of Fluids' in Baumeister, Marks' Standard Handbook for Mechanicaf
Engineers, Sec. 3, p. 50.
52 Rayleigh 's use of dimensional analysis and his use of the inverse Reynolds number is
Scientific Papers, 6 vols. (New York: Dover Reprint of 1869-1919, 1964), Vol. IV , p. 575.
5' Jakob, Heat Transfer, pp . 429-431 . Reynolds is not cornpletely clear on his reliance upon
mine Whether the Motion of Water Shall Be Direct or Sinuous and the Law of Resistance
in Parallel Channels', Papers on Mechanicaf and Physicaf Subjects , 3 vols . (Cambridge:
Cambridge Univ. Press, 1900-1903), Vol. Il, pp . 54-55 . See also Rouse and Ince, History
of Hydraulics, pp . 206-211, which puts Reynolds' work in perspective.
5. Reynolds , 'On the Dynamical Theory of Incompressible Viscous Fluids and the
character (p. 394). Though Reynolds was very irnportant, the concept of the dimensionless
parameter and its role in engineering was the work of many investigators, and only gradually
crystallized in the period 1890 to 1920.
•• Reynolds, ' On Certain Laws Relating to the Regime of Rivers and Estuaries and on the
Possibility of Experiments on a Small Scale' , in Papers, H, pp. 326-335. See also his three
committee reports on scale-modeling in hydraulics, ibid., pp. 351-380.
60 Reynolds, 'On the Theory of Lubricat ion and its Application to Mr. Beauchamp Tower's
the Criterion ' , and ' On the Extent and Action of the Heating Surface of Steam Boiler', ibid. ,
pp. 81-85 . This analogy linked heat with mass transfer; it is discussed in Ernst Eckert , 'Heat
Transfer' , in D.M. McDowell and J .D. Jackson (eds.), Osborne Reynolds and Engineering
Science Today (Manchester : Manchester University Press, 1970), pp. 160-175. McDowell
and Jackson's anthology gives a good idea of the scope of Reynold's work.
62 Rudolph Camerer , 'Klassificat ion von Turbinen', Z. Ver. Deutsch. Ing. 49 (4 March,
1905), p. 308.
63 Rudolph Camerer, Vorlesungen über Wasserkraftmaschinen (Leipzig: Wilhe1m
Englemann , 1914). Camerer was assisted by the fact that American turbine companies
produced stock turbines at standard sizesand these were tested at the Holyoke Testing flume
at constant head, so that entire families of geometrically similar turbine could be compared
in their speed and their "swallowing ability" (capacity) at constant head (p. 295). For the
capacity coefficient of turbines, see Denis G. Shepherd, Principles ofTurbomachinery (New
York: Macmillan, 1956), pp. 31-32.
64 Daugherty and Franzini , Fluid Mechanics with Engineering Applications, pp. 445-446,
468-469,490-493 .
.. John H. Lienhard, 'Ernst Kraft Wilhelm Nusselt', Dictionary of Scientiflc Biography,
s.v. See also John H . Lienhard , 'Notes on the Origins and Evolution of the Subject of Heat
Transfer', Mechanical Engineering 105 (June, 1983), pp. 20-27 . See also Elizabeth Jacob,
'Max Jakob; July 20, 1879-January 4, 1955', and Virginia Dawson, 'Frorn Braunschweig to
Ohio: Ernst Eckert and Government Heat Transfer Research', and Ernst R.G. Eckert,
'Ernst Schmidt - As 1 Remember Hirn', in Edwin T. Layton and John H. Lienhard, eds.
History of Heat Transfer (New York: American Society of Mechanical Engineers, 1988),
pp. 87-116 , 125-137, 139-146.
66 John H. Lienhard, ' Nusseit' , and ' Notes on the Origins and Evolution of Heat Transfer'.
pp. 25-28 and the same author's application of his principle in 'Sur les lois de similitude de
DIMENSIONALITY AND ENGINEERING SCIENCE 67
electricite', ibid., pp. 189-211. Macagno has useful remarks on Vaschy for which I arn,
again, in his debt; he points out that Vaschy did not make it completely explicit that the
terms he dealt with were dimensionless, though he did note that his terms were independent
of the system of units employed (a characteristic of dimension1ess numbers).
•• Aime Vaschy, 'Sur les lois de similitude en ph ysique", p. 345. Vaschy's statement is very
compressed; his statement of his theorem takes less than a page, and he moves on to
illustrations and applications at the bottom of the first page . At the outset of the theorem
he states tha t • "The most generallaw of similitude in mechanics and physics results in the
following theorern." He then states that a function of n physical quantities (a" ai ... •an ) can
be reduced to another of (n-p) parameters of the form fix" X2 •.., x n - p ) in which the
parameters x" X2 x n_p are monomial ("single term") functions of a" a2•.... Vaschy does not
state that this is the most general or fundamental form for physical equations, only that it
is possible . Vaschy , in his first statement, makes no statement suggesting that the x
parameters are dimensionless. Later on in the paper, however, he does say that these terms
ar e independent of the units in which these terms are expressed . This is today adefinition of
a dimensionless parameter, but perhaps in 1892 these useful entities were not so weil known
and an explicit statement might have helped some readers. In any case Vaschy and his fellow
French telegraph engineers do not appear to have made wide use of the theorem . One
problem appears to have been Vaschy's extreme brevity .
7. Buckingham, 'On Ph ysically Similar Systems ' , J. Wash. Acad. Sei. 4, pp . 345-350 (Sec.
2. pp . 345-346). As Langhaar has shown , the mere fact of dimensional hornogeneity leads
directly to the 11" theorem. (Langhaar, Dimensional Analysis, pp . 55-58). In some early
works the 11" terms were called variables , but the term parameter is more correct. The
dimensionless parameters are composed of terms which , in the non-dimensional form , are
correctly called variables. But the process of dimensional analysis does more than ju st group
together "variables" into groups in which the units of dimension cancel out.
11 In his paper 'Model Experiments and the Forms of Emp irical Equations', p. 265,
Buckingham presented a short version of his theorem which is strikingly similar to Vaschy's
earlier statement of the theorem.
" Buckingham's paper 'Model Experiments and the Forms of Empirical Equations", pp .
273-296, presented to the American Society of Mechanical Engineers in 1915 was a
particularly striking and influential example of Buckingham's concern to popularize his
work among engineering. See also Buckingham, 'Physically Similar Systems ' , pp . 347-353
and his 'T he Principle of Similitude', Nature 96 (December 9. 1915), pp . 396-397, and his
' Notes on the Dimensional theory of Wind Tunnel Experiments', Smithsonian
Miscellaneous Collections 62 (4, 1914). pp. 15-26 .
73 Langhaar, Dimensional Analysis and Theory 01 Models , p. 64. Buckingham discussed
model experimentat ion in his ' On Physically Similar Systems', pp . 369-372. and with many
practical examples in his ASME paper, ' Model Experiments and the Forms of Empirical
Equations' , pp . 273-296.
,. Kline, Similitude and Approximation Theory.
,. For remarkable examples of the power of Kline's methods see his Chapter 4, pp . 68-199.
7. Kline, Similitude and Approximation Theory, pp. 74-75 .
77 Becker, Dimensionless Parameters , p . 6.
" A. Sarlemijn and Pet er Kroes, 'Technological Analogies and Their Logical Nature'. in
P .T . Dublin (ed.), Philosophy and Technology 4: Technology and Contemporary Life
68 EDWIN T. LAYTON JR .
8\ Ernst R.G . Eckert and R.J. Goldstein (eds .), Measurement Techniques in Heat Trans/er
(London: The Advisory Group for Aerospace Research and Development, NATO, 1970).
This book does not include the Mach-Zehnder Interferometer, which , however, is discussed
by the editors in numerous publications .
81 On the Mach-Zehnder interferometer see W . Kinder, 'Theorie des Mach-Zehnder
(September, 1959), p . 55. The result of dimensional hornogeneity, as shown much later, was
that engineering theories were then expressed as functions of dimensionless terms.
8> Bridgrnan, Dimensional Analysis. Bridgman's proof of the 1f theorem is given on pp. 36-
47 . Bridgman set forth his " operatio nalism" and his refutation of the idea of absolute
"fundamental" dirnensions, pp . 10-25.
86 Quoted in EricTemple Bell, Men o/Mathematics(New York : Simon and Schuster, 1937),
p.338.
PETER KROES
logical function of an artifact is, however, irrelevant and not taken into
account; artifacts studied in science are not necessarily technologically
relevant, or are isolated from their technological context. Apparently,
this circumstance may lead to the development of different kinds of
theories within a technological and scientific setting .
THEORIE
OE LA
lh,"'l illt' prou"ll'r I'in u l clilud. d" metbode. e n uu r pout e.aluu 10:'
...sfc l. lIU le. proportio n, ..I" muhi"" ~ ".' ,", ur ; 0:1 j, Y , u b' til ut"·
11IIt leri. d o (ortft ul.. C1111"yli'1ue• • pto p"'•• de"'rm int1" I. " i l tu~
·r un e machin e de enee ,ou t IIno cbu co ee ne ue • •• charc c p.or u ne
l'i1eue üeee , " ' 3 potisuion ro ll t d ~. eIfeIl 'DUltl' I U. fcree en ehe .
'-3101 '. \011 t!Tel ut tle raUf 1111" ~o n ,o m m ' li o n een nue d'n u Cl de com -
IIuu ilJI1', 1.1 c.hmr r,e 0101 I ~ de len h' 'lu"il ("'1,1 1 lui do n ne r PO" t l ui (:li,.·
I'rooln i...· I nn m.H imum cI'elTe' uu le , ete • • ('tc. ;
n:n' l) 'c: ~
" .. 11" " 1 1 1-::11 1\ u , f' lIl1lr ' '' 'I YAt":l tl l. L' ,uni LLu l l l , "r; l "t "I · ""' '':
PARIS ,
IlACHELJEH, Il\IPRll\lEUH-LIBRAIRE
rou. LU U ....T lliKAT JQUU .
1839
THEORIE
DE LA
MACHINE A VAPEUR.
gO o .
CHAPITRE PRDlIER.
PREUYES 01:: L'I:iEXACTITUOE DES ~JtTllOOES
ORDlNAIRES OE CALCUL.
a paradigm of exact, critical engineering study : an exhaust ive investigation of the generat ion
of steam and its progress from the boiler through the cylinder to the condenser or
atmosphere. All types of engines were considered and the treatment was satisfactorily
analy tical.
Although one did believe that it was possible to compute in advance the performance of an
engine that was to be built, experience showed that one could be confident about the required
power only in those cases where the engine was a fair copy of an already existing machine, and
that an y deviation made the power to be expected uncertain.
According to Pambour, the existing theory for computing the net power of
steam engines, which he refers to as the Ccefficientenrechnung (p. 9) was
defective. Roughly, it proceeded in the foIlowing way. It assumed that the
pressure in the boiler was the same as the pressure in the cylinder. From this
pressure, the back pressure (or the atmospheric pressure) had to be
subtracted, resuIting in the effective pressure P . Then, given the velocity v
of the piston, the "theoretical" power W of the engine was computed; W
= Pav, with a the surface of the cylinder . In practice, the power delivered
PAMBOUR'S THEORY OF THE STEAM ENGINE 75
Two points are of importance for an engine : the resistanceit puts into motion, and the ve/ocity
with which this is done. First of all, therefore, the following two problems turn up .
First . When the velocity is determined for an already existing engine : to find the resistance
that the engine can move forward with th is velocity.
Second. When the resistance is determined for an already existing engine : to find the
ve/ocity with which the engine will put th is resistance into motion.
Here the expression "for an already existing engine" implies that the
evaporating capacity of the boiler is supposed to be known . These two
problems are directly related to the determination of the power of steam
engines, for the product of resistance and velocity is equal to the power
("(Nutz)-wirkung"). Pambour's theory also addresses another kind of
problem [ibidem) :22
76 PETER KROES
But there is still a third problem emerging from the two previous ones , namely :
Third, When the resistance and the velocity are determined : to find the dimensions of the
engine that will be able to overcome this resistance with the given velocity . In the case of steam
engines, what is being searched for here is the size ofthe boiler, or, ifyou like, the evaporation
capacity corresponding to the required power.
Consequently, the moment the steam from the boiler enters the cylinder, its pressure changes
and becomes equal to the resistance of the piston . This circumstance alone constitutes the
foundation of the theo ry of steam engines and explains all that is necessary.
PI = R (1)
The second principle allows the computation of the velocity v of the piston.
Let S be the amount of water evaporated per unit of time in the boiler, and
q the number by which S has to be multiplied in order to get the correspond-
ing volume of steam at pressure P in the boiler. Assuming that the
temperature ofthe steam does not change, Boyle's law can be applied to get
the volume of steam of pressure PI (in the cylinder) generated per unit of
time ; this volume is equal to
(2)
This must be equal to the volume swept through by the piston per unit of
time; that gives
qSP
av =-- (3)
PI
qSP
av =-- (4)
R
For this simple case, this equation solves the three problems Pambour
started out with . If the velocity is known, the resistance can be computed
or vice versa (assuming that S is known), and if the desired power is known
(vR) then S follows immediately!
The derivation of a similar equation starting from more realistic
assumptions turns out to be much more difficult. The most general
situation considered by Pambour is a steam engine with condensation
working expansively (p. 175). In such an engine, the temperature of the
steam is not constant. The pressure and the temperature change when the
steam leaves the boiler and enters the cylinder, and also during the
expansion phase. Thus, Boyle's law cannot be used to calculate the change
in pressure of the steam entering the cylinder or during expansion.
Nevertheless the relation between pressure and volume of the steam in the
cylinder has to be known, since the volume occupied by a certain amount
of steam at a given pressure is a main element in the calculation of the
78 PETER KROES
power of a steam engine (p. 84). So, the first problem Pambour has to solve
is how the changes in the temperature of the steam affect the relation
between the pressure and volume . He argues that steam during its passage
through the engine is always saturated, i.e. is always in astate of maximum
density (p. 82):26
In general, therefore, the steam will always keep the highesl density corresponding to its
temperature during its operation in the cylinder; as if it had not ceased to be in contact with
the water from which it was produced .
29254
I' = (engines with condensation)
1,784+p
30981
I' = (engines without condensation)
4,395+p
The first one is valid for low pressure engines, the second for high pressure
engines. Both relations are presented in the same form:
m
1'=--
n+p
with n and m empirical constants which are different for engines with or
without condensation.
Now it is easy to derive a formula for the relation between pressure and
volume for a fixed amount of saturated steam during the expansive phase
(p.83):
PAMBOUR'S THEORY OF THE STEAM ENGINE 79
with MI (PI) and M (P) the initial and final volume (pressure)." This
relation, together with the principles of dynamic equilibrium and of
conservation of steam, is sufficient for solving the three problems stated
earlier.
Pambour first introduces the following notation (p . 176):
7r : the pressure of the steam at a certain moment after cut off,
A : the totallength of the path of the piston,
AI: the length of the path of the piston until cut off,
x the length of the path of the piston at the moment that the pressure of
the steam is 7r,
C C times a is the "c1earance" or the dead volume in the cylinder filled
each stroke with steam through which the piston does not sweep.
During the expansive phase, the differential of the work done, dW, is
equal to 7raox. Integrating over the whole expansive phase, adding the
work done before expansion, and equating the total amount of work done
to the load times the path, Pambour derives the following formula (p . 177):
A+C AI
aRA = a(AI + c)(n + Pl)[log nat - , - - + -,- - ] -nds: (6)
"I +C "I +C
This he calls the first general equation. The second basic equation follows
from the equality of the amount of steam generated in the boiler and
consumed by the engine (p . 179):
mS
(7)
n+P1
mS A+C AI
V= [log nat - , - + -,-] (8)
a(R + n) "I + C "I + C
with the load r ; it is denoted by 4> + br with 4> the friction when there is no
load (r == 0); and (3) the back pressure p; for engines without conden-
sation, p is equal to the atmospheric pressure, and with condensation p is
the press ure in the condenser. 30
Putting
A+C AI
k == log nat - - + - - - (9)
AI+C AI+C
we get
mSk
v==--- (10)
a(n+R)
From this expression it also follows that the net power W, just as the
velocity v, does not depend on the pressure P in the boiler, but on the
evaporation rate S [ibid.] . Neither does W depend upon the pressure PI in
the cylinder. This is, of course, a quite remarkable conclusion, since the
pressure in the cylinder usually had been one of the key elements in
computations of the power of steam engines.
Pambour goes on to determine the velocity for which the net power
reaches a maximum given afixed cut off, When designing steam engines,
it is important to know the maximum net power, because " a steam engine
may never be built so that its highest possible power is equal to the power
it has to deliver regularly, for otherwise no power would be left for an
accidental increase in the resistance" (p. 186) .31 Thus, under standard
conditions steam engines have to operate below maximum performance.
The maximum net power occurs for the lowest possible velocity; in that
PAMBOUR'S THEORY OF THE STEAM ENGINE 81
(12)
Thus , the highest possible effect of the engine depends actually only on the evaporation Sand
the pressureP of the steam in the boiler ; that was to be expected, because the power of the
steam is determined only by these two . The cylinder and the motion of the piston, with their
dimensions, are just the means to transfer the power; they cannot increase or decrease it; and
also the speed of the motion ofthe piston cannot influence the highest possible effect, because
the speed can be increased and decreased at will only by the dimensions of the cylinder.
Thus , the construction of the boiler, not the construction of the engine,
determines the maximum net power!
Finally, Pambour discusses the "unbedingt-grössten Nutzwirkung" ,
i.e., the maximum net power for the most favourab/e cut off, This cut off
is given by:
(13)
For this value of the cut off, the net power is at a maximum. Pambour
rounds off his general theory of the steam engine by remarking that the
conditions for maximum net power differ from the conditions for the
maximum load that an engine can drive. In the following chapters he
elaborates this theory for different types of engines.
The count de Pambour is the inventor of the theory of steam engines, that is, the mies by
which, with sufficient accuracy for its operation, the effects of an already existing steam
engine can be calculated from the evaporating capacity of its flue and from the pressure ofthe
steam in the boiler, and by which, conversely, when the engine still has to be built, the lauer
two quantities can be calculated given the required power.
And in the Introduction Pambour writes that the shortcomings of the old
theory were the cause of "many deceptions in the use of engines and of the
ensuing controversies between the buyers and manufacturers" (pp . 6_7) .36
Engines often turned out to be either too powerful or too weak. In the first
case, engines could not operate under conditions of maximum efficiency,
in the second, the pressure in the boiler had to be increased, frequently
PAMBOUR'S THEORY OF THE STEAM ENGINE 83
The phenomenon ofthe production ofmotion by heat has not been treated from a sufficiently
general point of view. It has been treated almost exclusively with respect to engines whose
character and operation make it impossible for the full potential of the phenomenon to be
realized. In such engines, the production of motion is, as it were, curtailed and incomplete,
and it becomes difficult to perceive the principles underlying the process and to study its laws.
In order to grasp in a completely general way the principle governing the production of
mot ion by heat , it is necessary to consider the problem independently of any mechanism or
any particular working substance. Arguments have to be established that apply not only to
steam engines but also to any conceivable heat engine, whatever working substance is used
and whatever operations this working substance is made to perform.
It is easy to irnagine a host of engines suitable for developing the mot ive power of heat through
the use of elastic fluids . But, whatever approach is adopted, we must not lose sight of the
following principles :
I. The temperature of the fluid must first be raised as high as possible, in order to secure a
large fall of caloric and thereby the production of a great amount of motive power .
2. For the same reason, the cooling must be carried as far as possible.
3. We must see that the passage ofthe elastic fluid from the highest to the lowest temperature
is brought about by an increase of volume. In other words, we must see to it that the
cooling of the gas is a spontaneous consequence of rarefaction.
Whereas Carnot's theory about the transformation of heat into work may
be applied to any kind of heat engine, that is of any design whatsoever, the
PAMBOUR 'S THEORY OF THE STEAM ENGINE 85
Heat engines
Cornot
Pombour
turbines, for instance, or those using another working agent, such as air
engines, fall outside its scope." The object of Pambour's theory is thus a
dass of artifacts based upon a given design, in which a reciprocating piston
is a key element. This design stillleaves room for all kinds of variations,
such as piston operated steam engines working under high or low pressure,
with or without cutting off and condensation (p. 164). Within the dass of
piston operated steam engines Pambour distinguishes nine different types.
His theory is intended to be a general theory which covers all these cases (p.
166).
A particular dass of artifacts (and not the physical processes occurring
in these artifacts, nor the physical theories describing these processes)
constitutes the main topic of Pambour's theory. For solving his three
principal problems, Pambour uses Newtonian mechanics and analyzes in
great detail the behaviour of steam in the cylinder on the basis of gas
theory. These theories by themselves, however, do not delimit the domain
of application of his theory in a significant way. Of course, Pambour has
to assurne the validity of the theories he uses; this is a conditio sine qua non
for the validity of his theory. The domain of application of Pambour's
theory is, however, further restricted in an essential way by certain
technological design characteristics. Beside the obvious "scientific"
constraints, imposed by the use of scientific theories , technological
constraints also confine the domain of application. These technological
constraints are closely related to the validity of one of the two basic
principles of the theory, namely the principle of the conservation of steam.
CONC LUSION
This is not the way the principle of the conservation of steam, nor the two
basic equations are derived. In the case of the steam engine, the
technological design becomes from a physical point of view so compli-
cated that it cannot be translated into a set of boundary conditions which
can be fed into a physical theory. The design is no longer subordinated to
a physical theory by being simply its boundary conditions. Instead the
design itself becomes a crucial element in building a theory .
In order to avoid misunderstanding, it is necessary to set the above
distinction between scientific and technological theories briefly into a
broader perspective. The distinction put forward here does not imply that
a sharp boundary exists between scientific and technological knowledge.
As Mayr has remarked, any boundary between the two will be arbitrary;"
If we can make out boundaries at all between what we call science and technology, they are
usually arbitrary.(...)Trad itionally we regard physics as a science and the manufacture of
diesel engines as a technology, But what is thermodynamics, when textbooks are available
in all shades of emphasis, ranging from purely practical concerns to the most esoteric
theory?
NOTES
II Matschoss (1908).
Crelle of the French edition of 1844. Page numbers in the text refer to this edition.
" The invention of the Prony brake (around 1822) had to a great extent solved the problem
of directly measuring the power of a steam engine. The context of the invent ion of the " frein
dynarnometrique" is illustrative of the importance of finding reliable methods for
measuring the actual output of steam engines. Prony invented his method in connection with
a law suit about the efficiency of a Woolf engine for which he was asked to act as an expert
witness (see Payen (1976), p. 136).
" All quotations from Pambour's book are translated by the author; in each case, the
German text is given in a footnote . "Man glaubte zwar, im Voraus die Wirkungen einer zu
erbauende Maschine berechnen zu können, aber die Erfahrung zeigte, dass man der
verlangten Leistung immer nur dann gewiss sein konnt e, wenn die Maschine eine reine
94 PETER KROES
Nachbildung einer schon vorhandenen war, und dass die zu erwartende Wirkung durch jede
Abweichung ungewiss wurde ."
20 In spite of the great theoretical progress made during the nineteenth century in calculating
theoretically the power of steam engines, methods using empirically determined coefficients
of the kind rejected by Pambour were even being used in the twentieth century; see Ewing
(1926), p. 351.
21 " Es kommt auf Zweierlei bei einer Maschine an : auf den Widerstand, welcher sie in
Bewegung setzt, und auf die Geschwindigkeit, mit welcher dies geschieht. Zunächst also
ergeben sich folgende zwei Aufgaben.
Erstlich , Wenn für eine vorhandene Maschine die Geschwindigkeit bestimmt ist: den
Widerstand zu finden , den sie mit dieser Geschwindigkeit fortzutreiben vermag.
Zweitens. Wenn für eine vorhandene Maschine der Widerstand gegeben ist: die
Geschwindigkeit zu finden, mit welcher sie ihn in Bewegung setzen wird.
22 " Aber es gibt noch eine dritte Aufgabe, die aus den beiden vorigen von selbst hervorgeht,
nemlich:
Drittens. Wenn der Widerstand und die Geschwindigkeit gegeben sind : die Maasse der
Maschine zu finden, welche im Stande sein wird, diesen Widerstand mit der gegebenen
Geschwindigkeit zu überwinden. Bei den Dampfmaschinen ist, was hier gesucht wird, die
Grösse des Dampfkessels, oder , wenn man will, die der verlangten Wirkung gemässe
Verdampfungsfähigkeit.
23 In the German text the above quotation is immediately followed by aremark added by the
translator; it says: " Also, presumably, the dimensions of the cylinder and the other parts of
the engine" ("Auch wohl die Maasse der Dampfstiefel und der andern Theile der
Maschine.") For ereile the design problem therefore involves more than the capacity of the
boiler only . For Pambour, however, these other design problems, important as they might
be, are nevertheless subordinated to the most crucial design problem , namely finding the
capacity of the boiler given the desired outpur .
24 Pambour points out that the condition of the equality of motive force and resistance
concerns the mean value of both quantities during a whole stra ke, and not their
instantaneous values.
as "Folglich, so wie der Dampf aus dem Kessel in den Dampfstiefel tritt , ändert er seine
Spannung und nimmt diejenige an, welche dem Wiederstande des Kolbens gleich ist. Dieser
Umstand allein giebt der Theorie der Dampfmaschinen ihre Begründing und erklärt alles
N öthige."
2. "Im Allgemeinen wird also der Dampf während seiner Wirkung im Stiefel immer die
seiner Wärme entsprechende grösste Dichtigkeit behalten; eben so, als wenn er nicht
augehört hätte, mit dem ihn erzeugenden Wasser in Berührung zu sein."
This assumption was later abandoned by steam engine engineers; in the Rankine cycle " dry"
or superheated steam changes into "wet" or saturated steam dur ing expansion (see Kerker
(1960), pp . 266-267).
27 Boyle's law had been used for that purpose by Poncelet and Morin (Kerker (1960), p.
266).
28 The derivation of a single expression for the relation between specific volume and
pressure turns out to be rather complicated, partly for mathematical reasons (see p. 73).
29 Note that this expression is only valid when the amount of steam does not change .
30 For steam locomotives three other sources of resistance have to be taken into account,
PAMBOUR'S THEORY OF THE STEAM ENGINE 95
namely the resistance of the air , the force necessary to set the engine in motion and the force
necessary for blowing the furnace (p. 180).
3 1 " Eine Dampfmaschine darf nie so gebaut werden , dass Das, was sie regelmässig zu leisten
hat, ihre möglich-grösste Wirkung sei, weil ihr son st für eine zufällige Vergrösserung des
Widerstandes keine Kraft übrig bleiben würde."
3 2 "Also hangt die m öglich-gr össte Wirkung der Maschine eigentlich nur allein von der
Verdampfung S und der Spannung P des Dampfs im Kessel ab; was auch natürlich ist, weil
sich danach allein die Kraft des Dampfs richtet. Der Stiefel und der Kolbenlauf, mit ihren
Maassen, sind nur die Mittel, die Kraft zu übertragen, ohne sie vergrössern oder schwächen
zu können; und auch die Geschwindigkeit der Bewegung des Kolbens kann auf die möglich-
grösste Wirkung keinen Einfluss haben , weil sie durch die Maasse des Stiefels allein sich
beliebig vergrössern und verkleinern lässt."
3 3 Kerker [19601, p. 266.
l4 In the historical introduction to his famous A Manual of the Steam Engine and Other
Prime Movers, Rankine remarks [18781, p. XXXI): "The investigations of the Count de
Pambour on the theory of the steam engine , although not involving the discovery of any
principle in thermodynamics properly speaking, were conducive to the progress of that
science by pointing out the proper mode of applying mechanical principles to the expansive
action of an e1astic fluid ."
ae "Der Herr Graf von Pambour ist der Erfinder der Theorie der Dampfmaschinen, das
heisst, der Regeln, nach welchen sich mit zureichender Genauigkeit für die Ausübung die
Wirkungen einer vorhandenen Dampfmaschine aus der Verdampfungskraft ihrer Esse und
aus der Spannung des Dampfs im Kessel berechnen lässt, und umgekehrt dieses beides für
die verlangte Wirkung, wenn die Maschine erst gebaut werden soll. "
38 " vielen Täuschungen bei dem Gebrauche der Maschinen und der dann folgenden
3. When discussing the possibility of steam engines with three cylinders , Carnot remarks
[19861, pp. 107-108: "We shall say no more on this subject , since it is not our aim here to
enter into the con structional details of steam engines ."
3 9 According to Fox , the background of Carnot's work is the debate ab out the economy of
the Woolf engine (Carnot , [19861, p. 8) : "It is true that Carnot constructed his theory
without reference to any particular engine, but I am convinced, none the less, that his work
should be seen as a contribution to the lively debate, cent ring on the que stion of economy
and conducted almost exclusively in France, that was sparked off by the recognition o f the
remarkable qualities of the Woolf engine in about 1815."
40 • Krug [19811, p . 8.
4 1 " ein Modell des technischen Idealprozesses auf einer Abstraktionsebene, in der
konstruktive Parameter noch keine Rolle spielen."
42 According to Herivel, an orientation towards generality and abstractness together with a
concern for applicability and utility was characteristic of the French mechanical school to
which Carnot belonged (Herivel [19761, pp. 89-92).
43 See Payen [19761, p. 125.
44 Carnot (1986), p. 2 .
47 The last paragraph of Carnot's treatise c1early shows that he was aware that thermal
efficiency is only one factor alongside many others determining the overall efficiency of
steam engines (Carnot [1986)), p. 113.
4. Steam turbines were still at their infant stage at the time Pambour wrote his book; he
briefly mentions them in his survey of different kinds of steam engines (pp , 162 ff.).
49 Note that the principle of the conservation of steam has a counterpart in Carnot's theory;
Carnot assurnes that the amount of "elastic fluid" (air, for instance) does not change dur ing
a cycle. He analyzes a c/osed cycle (the same amount of working medium is used over and
over again) . This assumption, however, is not even explicitly stated as a principle, because
it is considered self evident in the context of Carnot's theory. Pambour, however, is
considering an open cycle in which in each cycle a fresh amount of working medium is
employed. For Pambour the principle of the conservation of steam is of central importance,
for it allows hirn to compute the speed of steam engines given the evaporating capacit y of the
boiler. Thus, both theories employ a similar conservation principle, but the status of these
principles is completely different.
50 "in Folge der eigenthümlichen Bauart einer Maschine" (p. 181).
[1926)), p. 80.
5 3 Within engineering circles simple rules, which are only approximately valid, are often
preferred for practical reasons over complicated mies which are supposed to be exactly
valid.
54 There exists, for instance, no "theory of pistons" analogous to gas theory in which piston
characteristics, such as piston mass, surface and length of stroke, are related to each other .
55 According to Kerker [19601 , pp . 267-268 Pambour was not interested in the problem of
57 As Channell remarks [1982], p. 39: "So long as science and technology each had their own
independent framework, there could be Iittle direct Interaction between the two, In fact,
such an interaction requires the creation of a new body of knowledge - engineering science
-the purpose of which is to transform the concepts and discoveries in one area so that they
can be incorporated into the other area."
5. Mayr [1982], pp. 157-158.
5 9 General relativity and elementary particle physics are two reference points at the scientific
end; bodies of practical knowledge gathered in traditional crafts Iie at the other end of the
spectrum . Vincenti [1984]discusses a technological innovation in American airplanes which
came about without science.
60 Kroes [1989], p. 378.
REFERENCES
INTROOUCTION
The Plumbicon has been so successful for broadcasting applications , that no strong
competitor has yet emerged to replace it.
Later designs , like the Japanese saticon, are not regarded as real
improvements; the Plumbicon can be considered as the successful closure
of a long sequence of pickup tube designs, going back into the late 19th
century:
L
"
Fig. 10 '0 The Nipkov disk ; S is the selenium cell and i is the current of this cell.
the current i in the cell also varies: it will be strong in the lighter
parts and weak in the darker parts of the image.
(3) Corresponding to the current a lamp will emit more or less light
when reproducing the image. If the same type of disk is used in
recording and reproducing, the image will reappear.
Yet the realization of Nipkov's idea was impossible at the time he
developed it. There were four problems ; for at least three of them the S-
and T-factors required to solve them were lacking in Nipkov's time:
Fig. 20 The image of two wornen, recorded and displayed with a Nipkov disk .
104 ANDRIES SARLEMIJN AND MARC DE VRIES
(1') In reproducing the image the brightness and contrast depend on the
speed with which a lamp can follow the variations in current. Due to
current variations there were indeed variations in light intensity, but
the reaction of the lamp was too slow: once heated, the filament kept
on emitting light for a relatively long time . In that sense television
presupposes a microtechnological development like electronics.
(2') The second problem is the slow reaction of the selenium cell. Im-
provement of this cell presupposes an increase of knowledge of the
relevant materials.
(3') In the third place no devices were then available for amplifying the
signal from the cell. This also presupposes electronics.
(4') A fourth problem was inherent in the mechanical approach. Along
the border of the Nipkov disk 625 holes (depending on the line system
chosen) must be made. One can calculate, that because of the size of
those holes, the disk should have a diameter of 5 meters for an image
size of 24 millimetres! The disk should turn at a speed of 25 rotations
per second . This rotation, in addition, must be free of vibration,
which is no easy design requirement to fulfil.
Therefore, television was both an ingenious and an impossible idea of a 24
year old German student. But at the age of 77 Nipkov could admire the
realization of his idea : he was present at the demonstration of television
at the Berliner Funkausstellung of 1937.
The idea of television matured with the emergence of electron theory and
electronics . Crucial developments in this respect were:
- the discovery of the electron in 1895 by Perrin and Thomson,
- the design of the cathode ray tube in 1897 by Braun, and
- the first design of a rectifier/detector (by Fleming in 1902) and that of
an amplifier (by Lee de Forest in 1906).
The first person to recognize the importance of these developments for
television was A.A. Campbell Swinton. His publications between 1909
and 1924gave an insight into the essential elements of television based on
electronics but he was unable to put his ideas into practice." For the first
applications of electronics we have to turn to the work of J.L. Baird in the
UK and of C.F. Jenkins in the USA.
APPLICA T10N-ORIENTED RESEARCH 105
" T ELEVISOR ..
ORDE R FORM .
7.
U. II:U T,"LL'I~'O" DC'·CI.O".,(.. r <.:0 .11'.. .. ' . Ll loIlTLI.l.
19Z8 IU. Lo "c. AC_ I
u.,...""... wC:
"1 0'<01.'l.Urr' l 10 mc Morlt' n ' 1OC'C .liCII In )_, l?lS Llll l. ,,=u,
.1 ,I.. "'I.' I'.. a .... l . •"tl /CCl ro J" .I ur-' lhe' CO~dl lo,)o1 " rn4or-e>.1
I... u ".
S., ,
I
--
-::.., r-n
1.',. ( /..Jl
PtHl,. 1 A-'J .CJI
Fig, 3. The front page of the Baird Company's brochure for television sets and the
order form for the sets.
The selenium cell had also been replaced by Jenkins in the USA; he used
an alkali metal photocell. In June 1925 Jenkins had demonstrated his
system with bevel-edged glass disks and transmitted the moving image of
a slowly rotating model windmill ." This, however, was still a far cry from
modern television . From the contemporary point of view we can never-
theless understand the enthusiasm with which these results were received
and the amazement they evoked. In the Philips Lab, too, a discussion
sprang up as to whether Philips ought to do something in this direction as
weIl. The result was that Philips did indeed enter the field.
In September 1927 M.J . Druyvesteyn was employed by the Philips Lab
106 ANDRIES SARLEMIJN AND MARC OE VRIES
and was given the assignment by its director Gilles Holst to "do something
on television"." Only a few weeks later, on 4 October, a patent was
registered in Holst's name ." In December of that year Druyvesteyn gave
his first demonstrations: the first time to the Dutch Physics Association
and then - for two weeks - to a broader audience. The receiver was just
a few rooms away from the transmitter. For broadcasting the short wave
was used (7.85 m according to Van der Pol). Druyvesteyn's pickup and
receiver devices were based on the use of a Nipkov disk, a photoceIl and
a neon incandescent lamp . At first a static image of a lantern was
transmitted; later a film was used."
Neither Druyvesteyn hirnself, nor the board of the Philips Lab
expected much success from work in this field. That is why it was
abandoned shortly afterwards (in 1929 or 1930). Some years later the
work was resumed, but by that time a number of new developments had
gathered speed abroad.
In the history of pickup tubes the work of the Russian emigrant Vladimir
Kosma Zworykin (1889-1982) was very influential: he replaced the
Nipkov disk by an electronic component, which was a crucial step in the
development. By substituting the slow mechanical scanning by the much
faster electronic scanning, it became possible to split up the image into
many more lines, while keeping the number of images per second
unchanged. The second reason for the importance of this step was the fact
that the image was now stored during the time that elapsed between two
scans, which produced a much higher tube sensitivity. In fact the whole of
this further development was aimed at making optimum use of this
intermediate time in order to improve the light sensitivity of the tube.
In 1925 Zworykin applied for a patent for this concept. He then
worked at the Westinghouse Electric Corporation (WEC) but he was
unable to convince his superiors of the benefits of television. In 1929 a
Radio Corporation of America (RCA) official visited the WEC Lab and
became acquainted with Zworykin's results. He became enthusiastic and
offered to Zworykin a position as director of the RCA Lab. Zworykin
accepted and in 1933 he could write that his design had been put into
practice and functioned weIl: the iconoscope."
APPLICATION-ORIENTED RESEARCH 107
V H
G E A
K
· ·················· ················v
F
'--_....:.-_---+-+-.........
The second important feature was the electrical storage of the image. The
scanning of the image was not done immediately, as in the Nipkov system,
but the image was first recorded by a device consisting of three layers (see
Fig .5):
- a mosaic layer, C, of cesium grains that are isolated from each other,
- an isolating mica layer, M, and
- a conducting metal signal plate, P .
Through the lens L the image falls on the mosaic of cesium grains.
Depending on the amount of light in the various parts of the original
image, a number of electrons are emitted by every grain and are captured
by the collector anode Ac . Therefore, the grains become positively
charged, depending on the light intensity. This positive charge is stored
because the grains perform a dual role:
- on the one hand they transform light rays into electric energy, just like
photocells,
108 ANDR1ES SARLEMIJN AND MARe DE VR1ES
- on the other hand each cesium grain, together with M and the con-
ductor P, behaves like a capacitor that stores the electric energy; the
grain remains positively charged until it is discharged by the electron
beam.
Next, the scanning beam discharges the grains one after the other. When
a grain is discharged this causes a small current. This discharge current
produces a difference in voltage over resistor R, which is sent to amplifier
Am. This is now called the video signal, which varies according to the
amount of light in the image.
The iconoscope had one serious disadvantage. When the target was hit
by the scanning beam , secondary electron emission occurred. Some of
/ I \ R
C M P
Ac C M P
I /
This, however, would mean that the maximum modulation frequency is doubled, which
should be avoided."
Cf
I I --------+-----~
1000 V
Cd=deflection coifs
4.a. The Emitron, the Eriscope, Farnsworth's Tube, and the Orthicon:
How S-, T- and PIl-factors Separated the USA and the UKfrom the
European Continent after the War
Cf
+1 Cd Cf
An initially more promising route led to the design of the orthicon. In the
USA this tube was developed by H. Iams and A. Rose in the RCA Lab . 2 2
It was developed further during the war, resulting in 1946 in the image
orthicon.
The word "orthicon" sterns from the Greek "orthos", which means
"perpendicular" . In contrast to the iconoscope in the orthicon the
electron beam hits the target perpendicularly. A double sided target was
used in this tube: in Fig. 8 the beam from the electron gun (cathode K and
anode A) scans the target F at the "back" side (seen from the image side).
Fis a semitransparent plate. The lens L projects the image on its "front"
side. F is scanned with a low velocity beam: the velocity with which the
112 ANDRIES SARLEMIJN AND MARe DE VRIES
electrons hit the target is low and secondary emission can thus be better
controlled. In the iconoscope there was a large potential difference
between the cathode ofthe electron gun and the target (about 1000volts);
this caused a high velocity scanning beam . The voltage of the anode was
about the same as the voltage of the target. The iconoscope was therefore
called "anode stabilised" .
In the orthicon, however, the gun cathode potential is the same as the
target potential (the orthicon is therefore called "cathode stabilised") and
there is a large potential difference between target and anode . This caused
the electrons coming out of the target to go to the collector anode Ac that
is located between electron gun and target (in the figure this is the
"returning beam" G) and then to fall on the collector Ac. Since the
invention of the orthicon low velocity beam scanning became the main
trend in the Anglo-Saxon pickup tube designs. Its potential for outdoor
recordings of acceptable quality was recognized . For indoor recordings,
too, less light was needed because of the better light sensitivity of the tube .
Attempts to further improve the light sensitivity of the orthicon were
made and these led to the image orthicon developed by the RCA Lab
during the war. The image orthicon is drawn schematically in Fig. 9. The
main difference from the orthicon is the added photocathode Pc and the
electron multiplier S. which we have already seen in Farnsworth's design.
A
c
~ t~·-~~---·
I - - - .' - -
F
Am
100 V
R
Like the orthicon, it had a scanning beam that went back and forth. On
its way back the beam passed through a hole in the anode and hit the
electron multiplier S. A double sided target was used, which was also
borrowed from the orthicon. Thus the image orthicon combined many of
the principles which were already available, such as low velocity beam
scanning, a double sided target, the combination of a photocathode and
a target, and the electron multiplier .
Cf Cd
II r..~::::=====;:::,
4.b. The Vidicon (1950), the Storting Point of the Work on the
Plumbicon as the Unification of the Different Design Traditions
For the vidicon, which is based on the use of photoconduction, S-, T-, M-,
and P/J-factors can be indicated, that were mainly present in the USA.
Especially there, both during and after the war, the technological
know-how with respect to semiconductors had increased because of the
work on radar and the transistor. This had caused a growing interest in
solid state and surface physics, which in turn influenced the development
of television. The vidicon design of the RCA Lab was based on a better
understanding of the principles of semiconductors and conductors. In
Europe there was skepsis concerning this design: it was doubted whether
semiconductors would be able to yield stable pickup tubes. 25
After the Second World War specific M-factors in the USA stimulated
the development of television: they were mainly related to the standard of
living and the possibilities for spending money on luxuries, such as
television sets. In postwar Europe, until the early 1950s, there was no
prospect for such markets."
The M-factors were governed by the P/ J-factors, of which the varieties
in line systems are an outstanding example. And these systems
corresponded to certain designs for pickup tubes, as we showed above .
Therefore the vidicon cannot be said to have originated from serendipity
or happy coincidence: it was based on the deliberate assumption that it
was possible to combine and exploit those factors.
That was the concept behind the work in the RCA Lab . P.K . Weimer,
S.V. Forgue and R.B. Goodrich worked on the vidicon and in 1950
presented their new invention in the March issue of Electronics and later
APPLICATlON -ORIENTED RESEARCH 115
Cd Cf
Pickup tubes are masterpieces of glassblowing skill and because of the mot ivation of very
skilled glassblowers the most unorthodox constructions have been realised. Mr . Scham pers
was the man with the golden hands , who began as a glassblower , but by self study rose to
the level of scientific assistant , and has contributed significantl y in all stages of Plumbicon
development. He invented the completely new vacuum melting connection between the bulb
and the foot of the tube - which most people thought was not possible - and realised it
together with the head of the glassblowing department, Haans, and developed it into a
method that could be used in the factory without the help of glassblowers . The choice of the
type of glass for both the image iconoscope and the Plumbicon was based on the
requirements that were imposed by these difficult constructions, in which tensions could not
be avoided, and on the optical requirement for a completely f1at screen."
Keeping the target material free from pollution was another problem.
These problems could only be solved by cooperation between researchers
and production workers .
The development of the Plumbicon shows the importance of a
piecemeal rationality: microtechnological designs like this one are the
result of intensive planning and coordination of R&D activities; each
design, however, can be considered as a new starting point for the
planning of activities that lead to further improvements in different fields:
in this case in electronics .
To gain more insight into the kind of coordination, required with
regard to organisation, steering and planning, we will deal with the
developments which led to a successful invention in the Philips Lab .
It would be wrong to look at these activities as a straightforward steering
towards a goal that was weIl defined and known from the outset.
118 ANDRIES SARLEMIJN AND MARC DE VRIES
The perception that a certain path will be a dead end sometimes creeps
up slowly after various analyses of the S-, T-, M- and P/J-factors.
In our historical-methodological study we did indeed come across such a
path, which turned out to be a dead end only after some time : the horne
cineac .
After 1936 Philips published hardly anything on pickup tubes. The only
person to publish articles in the 1930s was Rinia , who obtained many
patents for Philips. He was the man with the "golden fingers" : his designs
were excellent. When electronic pickup tubes had already been function-
ing for a long period (since the introduction of the iconoscope) he
developed a Nipkov disk that performed better than the best iconoscope
of that time." Apart from an article on this disk, nothing was published
by the Philips Research Lab in this area . Only in 1951, after a silence of
about 10 years, was a better version of the image iconoscope announced.
What was the reason for this gap in publications?
Here the myth of the both ingenious and strange Gilles Holst, who was
director of the Philips Lab between 1914 and 1946, seems to provide the
answer ." According to that myth, he seems to have had almost no confi -
dence in the development oftelevision. Some documents ofthe Philips Lab
archives seem to confirm this myth about Holst: he seems to have been such
a genius that he let people work on projects in which he did not believe
hirnself; individual persons would, more or less against their will, work on
television, whereas Holst hirnself seems to have been much more in favor of
the idea of ahorne cineac. This horne cineac would, every day, deliver the
news to subscribers by way of a film that they could watch whenever they
wanted. In this way people could escape from the tyranny of a dominant
television in their hornes. Fear of this tyranny would fit in with the per-
sonal convictions ofHolst, who held explicit ideas on humane technologies.
However attractive this image of Holst's ideas may be for explaining
television research at Philips, we think that other explanations are more
realistic. Holst did, in fact, very carefully watch the developments with
APPLICATION-ORIENTED RESEARCH 119
not only will the programs cost considerable amounts of money, but the limited range of the
transmitters will cause the need for a great number of transm itters. The expensive
broadcasting and the high purchasing costs will make it extremely difficult to build a
commercial exploitation. And yet ... television will come . It will just take a number of
years."
This was a realistic insight into the given factors of that time. In 1941 a
special committee within the Lab stated that television would become a
device that would bring together good music, horne amusement and the
newspaper. For the horne cineac, however, Philips would have to venture
onto a market with which it had no connection: the movie world .:" One
year later Philips started experimental television broadcasting in
Bindhoven.:"
Altogether these facts support the idea that Holst undertook the
necessary preparations for the research program which ultimately yielded
successes Iike the Plumbicon.
The fact that, in the early 1940s, no final decision was taken in the
horne cineac versus television issue seems to be sufficiently explained by
the lack of concrete perspectives on S-, T-, and M-factors. Those
responsible were unable to base any decision on certainty, so both projects
were kept going until further information became available.
5.b . Management Changes in the Philips Lab and the First Successes:
New Iconoscopes and the Use 01 Lead Oxide in Vidicons and
the Plumbicon (1962)
glass plate with a limited conduction. The charge was not emitted, but
leaked away through the target. The same idea had been used in the
vidicon and it reappeared in the Plumbicon.
This broad range of research activities around various types of
semiconductors hints at a managerial organisation that facilitated
concentration on the set of requirements to be formulated for a weIl
functioning tube.
The Plumbicon was revealed to the public in 1962. Although its
invention has always been ascribed to one of the Lab's staff members ,
E.F. de Haan, it was the acknowledged result of a weIl-planned, collective
effort. The amount of work involved and the organization of research in
the Lab rule out serendipity-like and unplanned discovery by one single
person . No doubt Oe Haan played a vital role in the R&O activities and
even more so in the technical applications of this tube, but in his
announcement he cited a 1954 article by Heijne and others on their work
on the pickup tube with a light sensitive layer of lead monoxide.
The advantage of lead monoxide is that it can be doped so that it
becomes a so-called PIN diode, consisting of three layers:
- a P-Iayer,
- an intrinsic semiconductor I-Iayer, and
- an N-Iayer.
The PIN-diode together with the tin layer can be seen as the resistancel
capacitance combination that we already saw in the vidicon. The electron
scanning beam has the function of charging the P-Iayer. The diode blocks
the current when no light hits the target. That way there can be no dark
current, which, of course, improves the quality of the image.
6 . CONCLUDING REMARKS
The Plumbicon combines small size, simple construction and easy operation with a low dark
cur rent , high sensiti vity , high speed of respon se, and good resolution . The Iife of the
APPLICA TlON-ORIENTED RESEARCH 123
"Plumbicon" is no shorter than that of other studio-quality tubes . It also produces striking
advantages particularly in colour television and in X-ray television set-ups ,
exaggerate their earlier influence. Braun's tube, the diode and the triode
were largely based on experience without an advanced theory of electrons.
These devices were sufficient for an empirical development of the pickup
tube. Only with the finishing touch, in the vidicon and the Plumbicon, did
solid state and surface physics come to play an important role in the
search for suitable materials.
According to oral tradition in the Philips Lab, Holst would not have
had a clear view of market developments. However, when we reconstruct
the technological possibilities and problems of the 1930s and the
management philosophy of the Philips corporation of that time, it
appears that his hesitation was justified: the S- and T-factors were
inadequate to meet the markets on which Philips wanted to operate.
The dilemma between push and pull, is at least in our case, a pseudo-
problem. The two concepts in no way relate to the reality of innovative
activity, and their simple dichotomy tends to obscure our understanding
of innovation. The ever changing relations between So, T-, M- and PI J
factors deserve our full attention when analyzing the "rationality" of
design.
NOTES
* The authors want to thank Dr L. Heijne and Dr T.S . te Velde (both have worked in the
Philips Research Laboratory on crucial projects of television technology) for their
comments on earlier versions of this article, and the employees of the Philips Concern
Archive for their cooperation in identifying resources for this study .
l Abbreviated to "Philips Lab".
1 Towler [1986], p. 67
J Efforts to raise the analysis of normalization processes to a specific science (the so-called
"Normierungskunde") did not succeed so far ; for the formulation of such an effort, see
Berg [1974].
4 We find this vision for instance in all publications by P . van Andel (one ofthem is [1989]).
, This opinion dates from the Sixties, when revolutionary students and their ideologists
thought that corporations would pursue their own interests and not society's interests. A
recent and more explicit version of this idea is found in the publications of the co-called
Starnberger School with as its most important representatives : Böhme, Van der Daele and
Krohn; for the publications the reader can consult the list of references. During a conference
in 1986 in Erlangen , Germany, a lot of "acadernics" appeared to hold these ideas (see
Sarlemijn, [1987]).
126 ANDRIES SARLEMIJN AND MARC DE VRIES
• The programrne, in which this concept fits, has been expla ined in Sarlernijn, (1990).
7 For more details of this tripartition , see Sarlemijn, (1984) and (1990).
8 After 1895-1905 the equations of classical physical theories became calculation methods
for the engineering sciences : the discussions about theoretical concepts (like the aether) had
by then been decided. This distinction between physical theories and engineering methods is
explained in the works of Casimir (1958), (1979) and [1983J. See also note 42.
9 Nipkov's German patent received number 30105 in 1884. Nipkov's method has been
explained extens ively by Zworykin and Morton [1954J and Csorda [1985J.
10 Figures land 2 have been taken from Holm [1956J. The recording of the two women is
probably Rinia's work, who experimented with the Nipkov disk (see Rin ia et al. [l937J,
[1938J and (1939) and also note 20).
11 See Na/ure 78 [1908J, p . 151, Journal of the Röntgen Society 8 [1912J, pp . 1-15 and
Wireless World 14 (1924), pp . 51-56, pp . 82-84, pp. 114-118. See also the judgement on
Allan A . Cambell Swinton's merits in Zworykin and Morton (1954), pp . 245-247 . In'
Schagen et al. [l951J, p. 71, Cambell's designs are mentioned as "strange drawing" .
12 For areport of Baird's experiments, see Nature (1925) 115, p. 505. Baird's contribution
642-645 .
14 This expression, which we find in Druyvesteyn's letter (PAD , (1967», has become a weil
known saying to indicate Holst's and Ca simir's management practice in the Philips Research
Lab .
15 Abramson [1974J, p. 7.
24 This conviction is expressed in Schagen, Bruining and Francken [1951J, p. 73) in the
following way: "According to these possibilities, pickup tubes are divided as folIows : (I)
pickup tubes with low electron velocities ("low velocity tubes"), in which the target is
stabilized on the cathode potential (that is why they are also called " cathode stabilized
tubes" , abbrev. CPS tubes), and (2) pickup tub es with high electron velocity ("high velocity
tubes") , in which the target is stabilized on the potential of the collector (e.g . 1000 V). To
the first group belongs the image orthicon which is mainly used in the USA, to the second
belongs the image iconoscope, which is preferred in Europe . One of the reasons for this
preference has to do with the great number of fines for which was decided on the Western
European continent (625, in France 819) . With high electron velocity one can better fulfil the
high requirements the definition of such a great number of lines poses to the focusing of the
scanning beam" (italics Sarlemijn and De Vries) . An earlier internal report (PCA (1948»,
probably written by Lewin , contained another vision: the 567 lines and 25 irnages per second
system is not technically different from the American system, but implies an economical
APPLICATION-ORIENTED RESEARCH 127
optirnum, because increasing the number of Iines would increase costs without appreciable
improvement of the image qual ity, and reducing the number of Iines would not save much
cost, but would seriously worsen the image quality , Again we see the problem of the number
of transmitters that Holst put forward time and again .
as see Albers [1953].
2. Although from 1948 onward all doubt in the Philips Lab disappeared (see Sarlemijn
[1990]), yet for a long time in the Fifties one was not confident about the speed of growth
and the final size of the market. This largely depended on the extent to which programmes
would become bette r. In The Netherlands the production of programme s was rather
primitive at that time (see Alders [1953]).
21 See Abramson [1974]. p. 7.
2. Plumbicon is the trademark for a certain type of vidicon. That is the reason why the name
- according to the intention of the patent application - should be written with a capital P .
2. Philips Archives. File Plantinga NL 661, first report.
JO L. Heijne in a letter to the authors, 14.01.199J.
JI This judgernent can be derived from the following publication s: Rinia [1938], Rinia and
Dorsman [1937]. Rinia and Leblanc [1939] . It has been confirmed in several interviews.
J 2 This myth forming has something paradoxical about it: it usually sterns from deep
resentment during management meetings on the television. That does not clash with our
thesis, which has three aspects: because of the So, T-. M- and P/J-factors Holst estimated
television to be not ripe for Philips' aspiration for the market. He kept watehing those
factors accurately without being led by speculative thoughts . He worked on the mobilization
of forces as soon as he believed a realization would become possible, putt ing the most
competent workers on those matters (e.g. the problem of the number of transm itters) from
which he expected the most problems.
J4 This is confirmed by nearly all PAD s that have been included in the list of references.
J 5 According 10 a rnessage of Philips Lamper (Oslo), dated September 19. 1929 (PAD), a
television set would cost about 700 Marks . Calculat ions, made in Eindhoven , resulted in an
amount of Dfl. 1.000.-.
J' PAD (1944) .
J 1 PAD [1938] .
gain the necessary experience as soon as possible, the Philips Research Laboratory put into
operation a 20 kilowatt television transmitter, that willtransmit programs and films for 4.5
hour s a week. In Eindhoven and surroundings a Iimited number of receivers will be
mounted, of a simple type with cathode tubes for direct reception as weil as projection -
receivers that give an image of about50 x 50 cm" . In PAD [1951]we read : "Already before
1939 by means of a moving caravan Philips gave television demon strations in most
European capital s. The results thus gained were very good and added to Philips ' prestige in
the area of television" .
41 We meet this concept of a "ripe idea" also in a theoretical context, namely in the works
of H .B.G. Casimir; he discusses theories that are ripe for fundamental developments in an
128 ANDRIES SARLEMIJN AND MARC DE VRIES
industriallab. In (1979) Casimir answered the question when quantum theor y was "ripe"
enough to serve as a starting point for industrial research as folIows: when the basic
equations had been found . Before that they had to be the subject of academic research.
Indeed in about 1930 in the Bell Lab as weil as in the Philips Lab quantum theory and
quantum mechanics became the object of intensive studies.
'2 This document is reproduced in Sarlemijn [1990).
REFERENCES
Abramson , A.: 1974, Electronic Motion Pictures, New York, etc.: Arno .
Abramson, A.: 1987, The History of Television, 1880 to 1941, Jefferson, etc.: McFarland .
Alders, J .c.: 1953, Opmars der electronen, Assen: Born.
Andel, P . van and Andreae, B.: 1989, Serendipiteit, de paradox van de ongezochte vondst,
Groningen: RuG.
Berg, S.: 1974, Normung, Forschung, Entwicklung (Normungskunde, Heft I, hrsg. v.
Deutschen Normenausschu sz). Berlin, usw.: Beuth.
Berthillier, M.: 1947, 'The Eriscope', Le Vide 2, pp. 355-359.
Böhme, G., van der Daele, W. and Krohn, W.: 1973, 'Die Finalisierung der Wissenschaft',
Zeitschrift für Soz iologie 2, pp. 128-144.
Böhme, G., van der Daele, W. and Krohn , W.: 1978, ' Finalisierung Revisited' , in Böhme,
G. (Hrsg), Die gesellschaftliche Orientierung des wissenschaftlichen Fortschriftts,
Frankfurt alM : Surkamp, pp. 195-250.
Burns, R.W.: 1986, British Television: The Formative Years, London : Peregrinus .
Campbell Swinton, A.A.: 1908, ' Distant Electron Vision', Nature 78, p. 151.
Casimir, H.B.G.: 1958, ' Certainty in the Exact Sciences', in Verslag van de plechtige viering
van het 150 j arig bestaan der KNA W . Amsterdam : Noordhollandsche, pp. 243-251 .
Casimir, H.B .G.: 1979, De kringloop van natuurkunde en techniek in de 20e eeuw,
Haarlem: Holl. Mij der Wetensch.
Casimir, H.B.G.: 1983, Haphazard Reality, Half a Century of Science, New York, etc.:
Harper and Row.
Csorda , I.P .: 1985, Image Tubes, Indianapolis : Sams.
De Haan, E.F .: 1962, 'The Plumbicon , a New Television Camera Tube' , PTR 24, pp.
57-58 .
De Haan, E.F ., van der Drift, A., and Schampers , P.P .M.: 1963/4, '''The Plumbicon", a
New Television Camera Tube' , PTR 25, pp. 133-161.
Doorn , A.G . van: 1966, '''The Plumbicon" Compared with Other Television Camera
Tubes', PTR 27, pp. 1-14.
APPLICATION -ORIENTED RESEARCH 129
Everson, G.: 1974, The Story 0/ Television: The Life 0/ Philo T. Farnsworth , New York:
Arno .
Farnsworth, P .T.: 1934, 'Television by Electron Image Scanning', J. Frankfin Inst. 218
(October), pp . 411-444.
Fink, D.G .: 1960, Television Engineering, New York: McGraw-HilI.
Francken , J .C. and Bruining, H.: 1953, 'New Developments in the Image Iconoscope'. PTR
14, pp . 327-335.
Heijne, L. 1960: Photoconductive Properties 0/ Lead-Oxide Layers (Thesis) Amsterdam :
Free University.
Heijne, L. : 1991, Letter to the authors in reaction to a draft version of this articie. (14
January).
Heijne, L., Schagen, P ., and Bruining, H.: 1954, ' An Experimental Pickup Tube for
Television with Application of Photoconduction' , PTR 16, pp. 43-45.
Holm , W.A.: 1956, Televisietechniek, Amsterdam : Meulenhoff.
larns, H. and Rose, A.: 1937, 'Television Pickup Tubes' , Proc. IRE 25(August), pp .
1048-1070.
Inst , of Electronic Engineers: 1986, International Conference on the History 0/ Television
- From Early Days to the Present. London .
Jenkins, C.F .: 1925, ' Vision by Radio Photographs', Rad. Phot. Nat . Cap. Press.
Kerkhoff, F. and Werner , W.: 1951, Televisie, inleiding tot de physische en technische
grondslagen van de televisietechniek met een uitgebreide beschrijving van
schakelschema 's, Amsterdam : Meulenhoff.
Mayo, B.J .: 1986, 'Early Development of Electron Guns and Cathode Ray Tubes for
Television', in International Conference on the History 0/ Television - From Early Days
to the Present. London: Inst. of Electron ical Engineers, pp . 69-72.
McIlwain, K.: 1939, 'Survey of Television Pickup Devices', J. Appl. Phys. 10(7), pp.
432-442.
(PAD) Philips Archive Documents:
- 1929, Message from the Norsk Akt ieselskap 'Philips Lamper' , Oslo, 19 September .
- 1937, 'Televisie express, de toekomst der televisie', ir. H.H . Heerema compares
advantages and disadvantages of the horne cinema and the television and also makes
price comparisons (PCA 811.3).
- 1938, ' Waarom geen televisie in Nederland ', analyses by ir. M. Leeuwin (PCA 811.3 +
814.3).
- 1939, Van der Pol's deliberations with respectto possible places for transmitters for TV
in the Netherland s, message to dr. A.F. Philips (PCA 911.3 + 814.3) .
- 1940 (28.8), Division radio research, message to those who work on television in the
Research Lab (PCA NL 608).
- 1940, Annual survey of the FM research in Eindhoven with appendices on the progress
in 1941 and 1942, by B. van der Pol (PCA 811.2) .
- 1941 (30.10), 'Celophane', report of the commission Department (NL 828).
- 1944 (30.6), Advice of J .F. Berger and ir. J .P . Bourdrez concerning the horne cineac.
- 1946, 'Possibilities for television', text of a radio broadcast by ir. M. Leeuwin on 14
August 1946.
- 1948 (8.1), 'Television in Europe', ir. M. Leeuwin compare s the advantages and
disadvantages of various line systems.
130 ANDRIES SARLEMIJN AND MARC DE VRIES
- 1948, 'Sorne ideas with repect to a Benelux coperation in the area of television' , by
Leeuwin (PCA 811.3 + 814.3).
- 1951, 'Philips and television ', Philips Press Service.
- 1967(12.12), letter of Prof. Dr. M.J . Druyvesteyn to Van Lokeren Campagne Public
Relations, N.V. Philips.
- Interview (tape 2) with dr. Haantjes by dr. Bruining and ir. StieItjes.
Philips' Oloeilampenfabriek: 1938, Philips Bedrijven, een korte beschrijving van .. . .
Eindho ven.
Puetz , J. : 1978, Television. S.l. : Verlaggesellschaft Schulfernsehen.
Rinia, H .: 1938, 'Television with Disc According to Nipkov and Interlinear Scanning', PTR
3, pp. 289-295.
Rinia, H. and Dorsrnan, C: 1937, 'Television Installation with Disc According to Nipkov",
PTR 2, pp. 72-76.
Rinia, H. and Leblanc, X.: 1939, ' Nipkov's Disc', PTR 4, pp. 46-52 .
Rose, A. and larns, H. : 1939, 'The Orthicon , a Television Pickup Tube ' , RCA Rev, 4, pp.
186-199.
Sarlernijn, A.: 1984, Histor isch gegroeide relaties tussen natuurwetenschap en techniek;
Eindhoven : TWIM .
Sarlernijn, A.: 1987: 'Analogy Analysis and Transistor Research' , Methodology and Sei.
20(3), pp. 40-61
Sarlernijn, A.: 1987, ' Konvergenz im Bezug auf Planung der Forschung' , in Burrichter, C.
and Lauterbach, O. (Hrsg), Wissenschaftsforschung im internationalen Vergleich,
Erlangen : DOF, pp . 227-281.
Sarlernijn, A. : 1990, Het ontwerp, spil van de techniekkultuur (inaugural lecture).
Eindhoven : University of Technology .
Sarnof', D.: 1939, 'Probable Influences ofTeievision on Society', J. Appl. Phys. 10(7), pp ,
426-431.
Schagen, P .: 1951 , On the Image Iconoscope, A Television Camera Tube (thesis).
Amsterdam : Free University.
Schagen, P ., Bruining, H., and Francken, J .C .: 1951 , 'The Image lconoscope, a Pickup
Tube for Television', PTR 13, pp. 71-85 .
Schagen, P :, Boerrnan, J.R ., Maartens, J.H.J ., and van Rijssel, T.: 1955, 'The
"Scenioscope", a New Pickup Tube for Television', PTR 17, pp. 173-182.
Smith, J .M. and Stewart, J.: 1986, 'From Vacuum Tubes to Solid State Imagers - the Past,
the Present, the Future', in International Conference on the History of Television - From
Early Days to the Present. London : Inst. of Electronica1 Engineers, pp. 73-76 .
Stern , R.H. : 1950, The Federal Communications Commission and Television; The
Regulatory Process in an Environment of Rapid Technical Innovation (thesis).
Cambr idge, Mass.: Harvard University
Teer, K.: 1986, ' Looking Back at Distant Vision: Television Technology from 1936-1986',
PTR 42, pp. 297-311.
Thiele, R. and Fix, H.: 1959, ' Ein vergleichende Betrachtung der heute verfügbaren
Fernseh-Bildaufnamerohren', Radio-Mentor 25, pp. 448-452.
Tiltman , R.F .: 1974, Baird of Television : The Life Story of John Logie Baird. London :
Seeley.
Towler, 0 .0. : 1986, 'Milestones in Television Pickup Performance', in: International
APPLICATION-ORIENTED RESEARCH 131
Conferenceon the History of Television - From EarlyDays to the Present. London : Inst.
of Electronical Engineers, pp. 64-69.
Van der Daele, W., Krohn, W., and Weingartner, P. (Hrsg.): 1979, Geplante Forschung.
Frankfurt alM : Surkamp.
Van der Mark, J.: 1936, 'An Experimental Television Transmitter and Receiver', PTR I,
pp. 16-22.
Van der Mark, J. : 1936a, 'Television' , PTR I , pp. 325-330.
Weimer, P.K., Forgue, S.V., and Goodrich, R.R .: 1950, 'The Vidicon Photoconductive
Television Pickup Tube' , Electronics 23, pp. 70-73 .
Zworykin, V.K.: 1935, 'The Iconoscope' , Proc. IRE 22(January), pp. 16-32.
Zworykin, V.K. and Morton, G.A.: 1954, Television, the Electronics of Image
Transmission in Colorand Monochrome. New York: Wiley, London : Chapman and Hall
(2nd ed.).
ROBERT FOX AND ANNA GUAGNINI
133
P. Kroes and M . Bakker (eds.), Technological Development and Science in the lndustriol
Age, 133-153.
© 1992 Kluwer Acodemic Publishers.
134 ROBERT FOX AND ANNA GUAGNINI
In turning to the particu1ar sector we have chosen for this paper - that of
electrical manufacturing and supply - we do not wish to challenge the
main statistical data on which current assessments of economic success
and failure are based. The implications of the figures given by such
authors as I.C .R . Byatt and Peter Hertner, which demonstrate the
capacity of German manufacturers to dominate markets outside as weil as
within the Empire, are clear." It is beyond question that weil before the
First World War, possibly even by 1890, Germany, along with the U.S.A.
and Switzerland, was firmly entrenched in what we term the fast lane of
electrical technology, while Britain and France drifted into a distinctly
slower lane from which escape became increasingly difficult as the gulf
between the leaders and the followers in the sector widened, as it did
inexorably, especially between 1900 and 1914. This slow lane was one in
which we also find Italy. For, despite the remarkable advance of the
economy of Italy from about 1890, technological dependence remained a
dominant characteristic of Italian industrial development until after the
First World War.
Our scepticism, then, concerns not the data themselves but rat her the
appeal to research as an explanation for them. In our view, the tenets
underlying such an appeal demand a conscious scrutiny which they have
all too seldom received. How justified is the assumption that improved
economic performance can only be achieved through a major engagement
in independent innovative activity? And is it invariably the case that
innovation of this kind emerges from the paradigmatic royal road of
research pursued, in the conventional late-twentieth-century manner, by
laboratory-based scientists?
Our answers to these questions are set in a very specific context. Our
concern is primarily with Western Europe and exclusively with the period
before the First World War, when electrical manufacturing and
installation emerged as a classic instance of a science-based industry. It is
symptomatic of the difficulty of formulating answers that even in the
seemingly innocent phrase "science-based industry" there lie teasing
historiographical difficulties, especially with regard to the period we are
discussing. For it is assumed, all too easily in our view, that the
description "science-based" is necessarily synonymous with "research-
based" and with activities geared to the quest for novelty pursued in the
laboratory. That assumption, which seems to draw such plausibility as it
RESEARCH AND ELECTRICAL ENGINEERING 135
concerned with the growth of research for innovation. He wished rat her
to see industry manned by a new breed of scientifically trained engineers,
and the national and industrial laboratories for whose establishment he
tirelessly campaigned were conceived first and foremost as settings for
quality control and testing and only incidentally as sources of
fundamentally new products or processes. 10
In Italy, a stance similar to Le Chatelier's was taken by Giuseppe
Colombo . Speaking from his powerful position at the head of the Istituto
Tecnico Superiore of Milan from 1897 to 1921, Colombo advanced a
programme for the modernization of Italian industry that drew on his
experience as a successful consuIting engineer, a member of parliament,
and the founder and first president of the Societä Italiana Edison. It was
a programme that gave pride of place to the introduction and diffusion of
existing technologies, rather than to the development of new ones .!'
Above all else, it required men who were able to assess both the
technological and economic merits of existing practices abroad and to
adapt those practices to the conditions prevailing in Italy. For this
purpose, engineers versed in the latest technologies were indispensable,
and under Colombo the Istituto Tecnico Superiore duly provided an array
of theoretical and practical courses oriented to the purveying of received
knowledge. The low priority that was accorded, by contrast, to the
fostering of fundamentally new departures was reflected in the virtual
absence of facilities for research."
It is important to reflect on the reasons why Le Chatelier and Colombo
chose to give innovation this secondary role and to conceive laboratories
and training in ways that may seem, in retrospect, unnecessarily
restricting. Were they simply being myopic? Were they failing to perceive
the key to industrial success that lay in innovative research and thereby
contributing to the perpetuation of the virtual imprisonment of their
countries in the slow lane? Both suggestions seem to us unconvincing. One
reason for this is that, although we readily admit that in electrical
engineering an indifference towards laboratory-based innovative research
undoubtedly existed in both France and Italy, we do not believe that
attitudes in countries in the fast lane were consistently different.
Certainly, the commitment to such research in, say, Germany or the
U.S.A. does not seem to have been strong enough for us to see it as a main
source of those countries' technological superiority. The roots of that
superiority lie in a far more varied mosaic of economic, political, and
technological factors which the literat ure on the history of industrial
RESEARCH AND ELECTRICAL ENGINEERING 137
that could and did take place in the workshop or in a factory or generating
station. Here, the innovation was typically of a far less spectacular kind
and of an adaptive rather than a fundamentally novel character. Hut it
provided another channel by which science could enter industrial practice
and often produce more immediately profitable results . Indeed, it is
tempting to suggest that the key to the technological success of countries
in the fast lane of the electrical industry may have less to do with the very
visible world of the laboratories than with the far obscurer mechanisms
for the transfer of scientific knowledge through experimentation
conducted on the shop floor or at an installation.
The force of this suggestion lies in the location of the cutting edge of
electrical technology in the quarter of a century or so before the First
World War. The location was unequivocally in the exploitation of high-
voltage, polyphase alternating current. It was a world of large generators,
long-distance transmission lines, electric traction, and the electrification
of industry. In all of these areas, there were major new departures, but
they were seldom ones that originated in the laboratory, at least in the
laboratory as it is traditionally conceived . The laboratory that mattered
for the pioneers of high-voltage A.C. before 1914 was the workshop or the
site . Here, the activity was scientific to the extent that it involved
experiment, theoretical insights, sophisticated mathematical techniques,
and the systematic analysis of production methods. The workers in this
"laboratory" , however, were engineers rather than reconverted
physicists. They constituted a very different breed who combined
scientific competence with experimental interests focused firmlyon the
machines and installations themselves . They were the products of courses
oriented explicitly to electrical technology, albeit to an electrical
technology that demanded an increasingly refined grasp of physical
theory.
The contributions made by these men were at times highly innovative,
as the cases of Marcel Deprez, Nikola Tesla, or Michael Dolivo-
Dobrowolsky illustrate. Hut, more frequently, they were of a less
spectacular kind, affecting quality control, reliability, or production
costs, or the harnessing of electric power in new industrial contexts. The
case, recently described by Ulrich Wengenroth, of the exceedingly
intricate tasks that were involved in adjusting polyphase A .C . motors to
the complex requirements of a textile mill is typical." As Wengenroth
shows, the gearing and coupling mechanisms that were required in order
to ensure a uniform speed irrespective of the number of machines in use
140 ROBERT FOX AND ANNA GUAGNINI
were not fundamentally novel. But the quest for a solution called for
precise measurements of efficiency and a degree of mechanical ingenuity
and flair that it is all too easy for the historian to undervalue by
comparison with the purely electrical side of the task.
The argument we want to develop from this discussion is that, in the fast
lane as much as in the slow lane, the contribution of science to industrial
practice cannot be fully understood without a c1ear appreciation of
changes of the kind we have described, some with their focus on such
tasks as trouble shooting and the adaptation of electrical technology to
customers' requirements, others yielding innovations in contexts other
than that of the conventional laboratory. If the argument is accepted, it
has significant implications for the continuing debate on the extent , and
the implications for industrial performance, of the technological gap that
opened between, say, Germany and the three slow-Iane countries -
Britain, France, and Italy - with which our own work has been chiefly
concerned. Britain , France, and Italy, after all, expanded their provision
for technical education to levels which ensured that about the turn of the
century the supply of manpower at least matched demand. In our view,
therefore, the basic technical knowledge and skills were not wanting, and,
on this score at least, we see no impediments to the development of a
technological strategy, founded on the assimilation and adaptation of
borrowed technologies, that offered every prospect of success.
The questions that emerge from this interpretation are of a somewhat
different kind from those conventionally asked in discussions of the
relations between scientific knowledge and industrial practice. They turn
less on the capacities of different countries to generate either an elaborate
provision for advanced technical education or independent programmes
of research and innovation than on their being in a position to exploit the
manpower and the facilities (including laboratories for such routine tasks
as testing and standardization) that were more generally available than
most of the current secondary literature would suggest.
Such questions are riddled with snares and deceptively cogent answers.
Quite apart from the temptation to seize on a low level of laboratory-
based innovative research as a satisfactory explanation of flagging
industrial competitiveness, it is all too easy to associate any signs of a lack
RESEARCH AND ELECTRICAL ENGINEERING 141
For the British, the most telling lessons were learned in the painful task of
establishing networks based on high-voltage alternating current. The
impediments, most conspieuously in London, were of a very diverse
character. As Thomas Hughes has shown, they included locallegislation
whieh was intended simply to avoid any risk of a monopoly but which led,
in practiee, to a damaging fragmentation of responsibility for the supply
of electricity to the capital. 19 In a structure that saw London provided with
no fewer than forty -nine distinct supply systems and with a baffling
diversity of voltages, frequencies , and techniques of transmission and
distribution, coherence was impossible , and the incentive to serious
investment was gravely undermined.
But there were other lessons of a more explicitly technologieal charac -
142 ROBERT FOX AND ANNA GUAGNINI
World War, in the Societä Italiana Edison as elsewhere, Italy was not
significantly different from France. Supply companies , in particular,
tended to be heavily financed by German or Swiss banks and, by virtue of
that allegiance, were unusually open to foreign technology, while the
manufacturers and installers of electrical machinery found themselves
responding to demands that encouraged ingenuity of an adaptive rather
than a fundamentally innovative kind . Again, it is important not to
underestimate the successes that could be achieved in these circumstances,
in Italy even more obviously than in France. Italian demography and
geography, for example, were particularly favourable to the growth ofthe
hydroelectric installations which became, from the start, the country's
most notable speciality in electrical technology . It is true that France, too,
had her successes in hydroelectricity, but the distance between the main
areas of generation, around Grenoble and in the Pyrenees, and the most
populated manufacturing region around Paris remained an impediment
until an improved transmission network was in place after the First World
War. In Italy, by contrast, the sources of hydroelectricity in the Alps and
the main body of industrial and (less crucially) domestic consumers in
Milan and Turin were far closer to one another.
In this context, the potential profits and hence the incentive to reduce
the costs of production and improve the techniques of transmission were
far greater than they were in France before 1914. And it is part and parcel
of the Italian success story in hydroelectricity that this financial incentive,
far from being ignored, stimulated a major technological response. When
the Paderno generating station, near Milan, was opened in 1898, its
alternators came (as they would have done anywhere in Italy or France)
from abroad, in this case from Brown, Boveri, while the transformers
came from Ganz in Budapest. 27 But the turbines, insulating mechanisms ,
and cables were of Italian origin . Even more importantly, so too was the
overall design of the plant, which was the work of Galileo Ferraris (then
professor of electrical technology at the Polytechnic of Turin) and the
company's technical director Guido Semenza. The result was impressive:
a technologically daring installation operating at 13 500 volts and with
transmission lines extending over 33 kilometres, which at the time of its
completion in 1900 was one of the largest hydroelectric power stations in
Europe.
In this as in many other achievements of Italian electrical technology,
the element of originality was considerable, but it owed virtually nothing
to the laboratory . The contribution of the Pirelli company, which was
148 ROBERT FOX AND ANNA GUAGNINI
responsible for the cabling, was typical, For at the time, and until after the
First World War, Pirelli maintained only a small company laboratory,
devoted almost entirely to testing and product contro!. Its technological
advances in the design of electric cables for the transmission of alternating
current at increasingly high voltages were accordingly the work not of
bench-based physicists but of gifted, weIl trained engineers whose world
was that of the workshop and the site." In a pattern that had analogies in
numerous other Italian companies, the unsung heroes of Pirelli's pre-war
success were the director of the cable department, Leopoldo Emanueli,
and the firm's chief electrical engineer, Emanuele Jona, both of them
highly educated graduates of Italian engineering schools.
Another success story even more remote from the world of laboratory
research was .that of Ercole Marelli. 29 After receiving virtually no formal
education, Marelli was apprenticed to one ofthe earliest electrical firms in
Italy, the largely unsuccessful Tecnomasio, before establishing his own
modest workshop for the production of small electrical appliances in
1891. Among these appliances were the electric fans of the type that was
just beginning to be imported, for Italian domestic use, from the United
States . By 1898, the range of Marelli's products had grown remarkably,
to embrace not only domestic fans but also large and technologically more
ambitious ventilating systerns for industrial plants. The progress of the
firm was and remained a source of intense pride . By 1911, the official
company history looked back on an expansion that had taken the work
force from forty in 1897 to its present 1500 and on a vigour that had
resulted in a flourishing export market and in a manufacturing plant
equipped with the most modern machinery. Yet the only source of
technical know-how to which the publication referred was the drawing-
office, with its staff of engineers working on problems of design related to
the immediate needs of production. Equally revealing was the last
chapter, devoted to what were seen as the constraints on the future
prosperity of the firm and, more generally, of the Italian electrical
industry as a whole . Significantly, most of the constraints were
commercial ones associated with the lack of governmental support and
protection. There was no sign that either the manning or the improvement
of technical pro cesses was seen as presenting any significant difficulty.
RESEARCH AND ELECTRICAL ENGINEERING 149
CONCLUSION
The case histories we have reviewed are intended to illustrate, rather than
prove, our contention that the presence or absence of research
laboratories before 1914 does not, on its own , provide a satisfactory
explanation for the divergence between the countries of the fast lane and
the slow lane in electrical technology. In making this claim , we do not
underestimate the effect that the flow of scientific knowledge undoubted-
ly had on the growth of the electrical industry. We simply feel that so long
as historians focus exclusively on the contribution to the flow made by
research laboratories, and hence on major new departures, their view
must, of necessity, be a partial one . As we have argued, in this period a
main channeI for the entry of knowledge into practice remained the
essentially "low-key" route via the shop floor or engineering workshop.
When that second path to innovation is given the prominence we believe
it deserves, the boundaries between the conditions prevailing in the
countries of the fast lane and the slow lane become far less c1earthan if we
consider laboratories alone, and some existing explanations of techno-
logical success and faiIure appear exceedingly fraiI.
Plainly, in a paper of this length , it is impossible to take the argument
much further. That said, if we venture into the realms of counterfactual
history, we are convinced that the establishment of laboratories and a
serious engagement in fundamental innovative research would not have
allowed Britain, France, or Italy to make good the technological ground
they lost inexorably from the later 1880s. For laboratories to be effective ,
the context had to be one that made novelty technologically and, wherever
possible, economically rewarding. In the slow lane, the context was quite
simply not there; and no amount of investment in facilities for research
could have made up for that lack. There was, in other words, no short cut.
This disarmingly simple point carries with it a chastening message for
the historian, since its elaboration would call for a daunting degree of
contextualization and the interweaving of causes at the expense of the
monocausal explanation, We have to grapple with the fact that, in one set
of circumstances, Ferranti's technological daring led to faiIure, whereas
the research of the General Electric Company on the metallic filament and
ofSiemens & Halske on products ranging from Werner BoIton's tantalum
lamp to the high-speed telegraphic apparatus that was marketed from
1912 spawned profitable breakthroughs. SimiIarly, though in a quite
different sector , we have already referred to the divergence between Italy
150 ROBERT FOX AND ANNA GUAGNINI
day. We fully accept that, for the last century, a research-based industrial
policy, pursued in the right context, has been capable of success. Hut,
equally, it has not been a panacaea to be adopted in all places and at all
times. At least that degree of uncertainty has to be acknowledged, if
historians are to debate the significance of history with contemporaries in
search of easy and obvious solutions to the difficulties of the flagging
economies of the late twentieth century.
University oj Oxford
NOTES
I John Joseph Beer, The Emergence of the German Dye Industry, IIlinois Studie s in the
Development in Western Europe from 1750to the Present (Cambridge, 1969), especially pp .
323-326, and David C . Mowery, 'Industrial Research, 1900-1950', in Bernard Elbaum and
William Lazonick (eds.), The Decline ofthe British Economy (Oxford, 1986), pp . 189-222 .
J Mowery, ibid., p. 189.
4 I.C.R. Byatt , The British Electrical Industry, 1875-1914 (Oxford, 1979), pp . 150 and 166,
and Peter Hertner, '11 capitale tedesco nell'industria italiana fino alla prima guerra
mondiale', in Bruno Bezza (ed.), Energla e sviluppo. L 'industria elettrica e la Societ äEdison
(Turin, 1986), pp . 213-256.
5 The point appears on p. 172 of the English edition of the Catalogue, published as
International Exposition . Paris 1900. Official Catalogue. Exhibition of the German Empire
(Berlin, 19(0) .
6 Luigi Gabba, ' L' insegnamento della chimica nelle universitä e negli istituti superiori' ,Atti
Report of the Departmental Committee on the Royal College of Science. Volume I: Final
Report with Appendix I (London, 1906). On the parallel campaign for the improvement of
scientific (as opposed to purel y technological) education and research, see Roy M.
MacLeod, 'The Support of Victorian Science: The Endowment of Research Movement in
Great Britain, 1868-1900', Minerva 4 (1971), pp. 197-230, and Peter Alter, The Reluctant
Patron. Science and the State in Britain, 1850-1920 (Oxford , Hamburg, and New York ,
1987), especially Chapters 2 and 3.
• Typical statements include : Albin Haller, 'L'industrie chimique a I'Exposition de
Chicago' , in Ministere du Commerce, de I 'Industrie, des Postes et des Telegraphes.
Exposition Internationale de Chicago 1893. Rapports publies sous la direction de M.
Camille Krantz . Comite 19. Produits chimiques et pharmaceutiques, materiet de lapeinture,
parfumerie, savonnerie (Paris, 1894), especially pp. 10- 17, and Henry Le Chatelier,
'Rapport sur les laboratoires nationaux de recherches scientifiques ', Comptes rendus
hebdomadaires des seances de l'Academie des Seiences 163 (1916), pp. 581-588 . Cf. the
152 ROBERT FOX AND ANNA GUAGNINI
views on the education of industrial seientists and engineers , see his preface to Leon GuiIIet,
L 'enseignement technique superieur a l 'apres-guerre (Paris, 1918), pp . 9-28.
11 Carlo Lacaita, 'Giuseppe Colombo e le origini deII'Italia industriale', in Carlo G. Lacaita
(ed.), Giuseppe Colombo. Industria e politica nella storia d 'ltalia. Scritt i scelti, 1861-1916
(Bari , 1985), pp. 5-86. For a typical statement of Colombo's views, see 'Le gallerie delle
macchine del lavoro e dei materiale ferroviario all'Esposizione Nazionale di Milano (1881)',
in Colombo, Scritti e discorsi scientiflci (2 vols., Milan , 1934), Vol. 2, pp, 1060-1087 .
., Anna Guagnini, 'Higher Education and the Engineering Profession in Italy , The Scuole
of Milan and Turin, 1859-1914 ', Minerva 26 (1988), pp . 512-548 .
IJ The thrust of Nathan Rosenberg's approach is expressed very clearly in the collection of
GE and Bell, 1876-1926 (Cambridge, 1985), espeeially Chapters 3 and 4. See also George
Wise, Willis R . Whitney, General Electric, and the Origins 0/ U.S. Industrial Research (New
York, 1985), Chapters 5-7 .
IS David Cahan, An Institute for an Empire . The Physikalisch-Technische Reichsanstalt
1871-1918 (Carnbridge, 1989); Crosbie Smith and M. Norton Wise, Energy and Empire. A
Biographical Study 0/ Lord Kelvin (Cambridge, 1989), pp . 684-698 and passim; M. Norton
Wise and Crosbie Smith, 'Measurement, Work , and Industry in Lord Kelvin' s Britain' ,
Historical Studies in the Physical and Biological Seiences 17 (1986), pp . 147-173 ; Graeme
Gooday, 'Preeision Measurement and the Genesis of Physics Teaching Laboratories in
Victorian Britain' , Brit , J. History 0/ Science 23 (1990), pp . 25-51.
16 On Regnault, see Robert Fox , The Caloric Theory 0/ Gases from Lavoisier to Regnault
existence in the mid-century, and on the development laboratories of Schuckert before the
amalgamation with Siemens in 1902, see Georg Siemens, Der Weg der Elektrotechnik.
Geschichte des Hauses Siemens (2 vols., Freiburg and Munich , 1961), Vol. 2, pp. 58-59. The
very different character of AEG 's policy is brought out in Ulrich Wengenroth's recent
biographical sketch ofEmil Rathenau; see Wengenroth, 'Emil Rathenau ' , in Wilhelm Treue
and Wolfgang König (eds.), Berlinische Lebensbilder - Techniker (Berlin, 1990), pp. 193-
209, espeeially pp . 203, 204, and 207.
RESEARCH AND ELECTRICAL ENGINEERING 153
I. Ulrich Wengenroth, 'L 'electriflcation dans I'industrie textile', in Fabienne Cardot (ed.),
Histoire et structure economique de l 'electrification (Paris, in press), and 'Shaping a
Technological Potential: R & D in Eleetrie Drives, 1890-1930', a paper read to the historical
group of the Institution of Eleetrical Bngineers, London, in Oetober 1991 and now awaiting
publieation .
" Thomas P . Hughes, Networks 0/ Power. Electrification in Western Society , 1880-1930
(Baltimore and London, 1983), pp. 238-247 . Other aeeounts of Ferranti's early innovations
and their teehnical and eeonomic implieations include Gertrude Ziani de Ferranti and
Richard Inee, The Life and Letters 0/ Sebastian Zian i de Ferranti (London, 1934), pp. 54-
68; R. H. Parsons, The Early Days 0/ the Power Station Industry (Cambridge, 1939), pp.
21-41 ; and W. L. RandelI, S. Z. de Ferranti (London, 1943).
20 Hughes, ibid ., pp. 443-460.
I'industrie electrique fran ..aise, 1880-1931. Causes et pratiques d'une dependance', Annales
ESe, 3ge annee (1984), pp. 1020-1043.
" See the brochure announeing the formation of the company : Compagnie Generale
d'Electricite. Societe anonyme en formation . Notice (Par is, 1898), pp. 3-4.
13 The policy of the compan y emerges very clearly from the annual reports and the rich
eollection of cutting s from the finaneia l press in the file on the CGE in the Archives
Nationales (65 AQ. G.16O). The case of the CGE is discussed in Broder, 'La
multinationalisation de l' industrie electrique fran..aise', op . cit. (note 21), pp . 1031-1032.
2. The point emerges briefly from the company history , La belle histoire de la CEM (Paris,
1950), and even more strongly from the annual reports and newspaper euttings in the
Archives Nationales (65 AQ. M.118).
2S The growing resentment at the dominant position of senior German staff in Freneh
industry on the eve of the First World War is reflected in Louis Bruneau , L 'A llemagne en
France. Enquetes economiques (Paris, 1914), p. viii and passim, and in the paper s
concerning the Compagnie Generale d'Electricite de Creil (not to be confu sed with the CGE)
in the Archives Nationales (65 AQ. G.207). Feelings ran so high at Creil that they prompted
the resignation of four of the Freneh members of the Conseil d'administration in March
1912.
26 Claudio Pavese, 'Le origini della Societ ä Edison e il suo sviluppo fino alla costituzione dei
" gruppo"' , in Bruno Bezza (ed.), Energia e sviluppo, op . eil. (note 4), pp. 65-167.
27 Guido Semenza, 'L'impianto di Paderno' , Atti della Associazione Elettrotecnica Italiana
1 (1897), p. 121, and Note sui nuovi impianti della Societä Generale Italiana Edison di
Elettr icita, 1895-1898 (Milan, 1899).
2. Alberto Pirelli, La Pirelli. Vita di una azlenda industriale (Milan, 1946), and Hector
Sacchetto , History and Development of the Oi/ Filled Cable (n.p ., n.d . (1969)), Chapter 2.
2. On this case, see Vent'anni di vita della Ditta Ettore Marelli & C., Mi/ano . 1891-1911
(Milan, 1911).
H. LINTSEN . G . VAN HOOFF AND G. VERBONG
INTRODUCTION
After 1805, the Rijkswaterstaat, and thereby the profession of the civil
engineer in public service, was moulded in the French and military fashion.
It rapidly acquired a formal structure, an elaborate system of work
distribution, an extensive hierarchy and strict discipline. The engineers,
who were officers in the Rijkswaterstaat, wore uniforms and attached
considerable significance to the honour of the corps. Furthermore,
training of these engineers was transferred to military schools. In these
institutes the trainees had to conform to numerous regulations concerning
behaviour and discipline, so that an esprit de corps was fostered as weIl as
a sense of hierarchy between the cadets and the teachers, and among the
cadets themselves. The system was considered to be complete as soon as
the cadets could be separated entirely from the outside world. This stage
was reached in 1829through the boarding school facilities provided by the
newly-founded Royal Military Academy in Breda . But it was not only the
setting of the engineering education that changed, its contents had also
become subject to revision since 1805. The training became a theoretical
one with a considerable emphasis on mathematics and science. These
militarily trained civil engineers, together with the technical officers of the
army, formed the Royal Institute of Engineers in 1847. They made a
significant contribution to the changes within the Rijkswaterstaat and the
development of the Dutch infrastructure.
In this article we analyze the complex relation between the rise of a
professional community, the transformation of technological knowledge
and their effects on industrialisation. As we have taken industrialisation
as a central theme in our research , it would not be appropriate to
concentrate upon the profession of civil engineers, whose main concern
was public works. We will focus on the profession of mechanical
engineering and its relation to the development of steam technology in
The Netherlands. By shifting our attention from the civil engineers with
their evident change from "shop" culture to "school" culture to the lesser
known group of mechanical engineers, we may also be able to qualify the
implicit idea that such a professional transition holds for all engineering
sciences. Our main questions are:
1. Was there a community of mechanical engineers and does it show
traces of a change from shop to school culture?
2. Is there a relationship between the transformation of the technical
communities (from "shop" to " school" ) and the transformation ofthe
"body of knowledge", from a craft into a technical science?
3. Have professional communitie s ("shop" or "school") in the above
158 H. LINTSEN. G. VAN HOOFF AND G. VERBONG
One problem which appears when describing the group of Dutch tech-
nicians in the engineering industry is that of a more specific definition of
the category, In Dutch, a variety of terms were used to denote those who,
in any way, worked on machines: werktuigkundige, werktuigbouwkundi-
ge, mecanicien, ingenieur. In fact, until the start of areal school for
mechanical engineers in Delft in 1842, these terms may designate self-
made tinkerers as well as fully trained civil engineers who had gone into
machine building. Hence, even by putting together all people called
werktuigkundige we cannot claim to cover the professional group which
we are looking at in this article. In the engineering industry there were
unskilled workers, craftsmen (such as blacksmiths, fitters and boiler-
makers), the foremen, the draughtsmen, the supervisors, designers, the
director, and so on. We confine ourselves to the category of higher
technical functions, generally described as managers, engineers, heads of
departments, designers, and mechanics. We have information on the
educational backgrounds of 420 people in this category in the nineteenth
century (see Table 1).4
The majority had received their training mainly or exclusively in
practice, and had not gone through any specific training institute. They
may have attended evening classes, or qualified by means of horne study
MECHANICAL ENGINEERING IN THE NETHERLANDS 159
TABLE I
Backgrounds of higher techn icians in the engineering industry during the whole of the
nineteenth century (N = 420)
Number 0/0
and the like. Only about a quarter of the higher technicians had been
trained at some institution or other. The nature of the formal training
could differ considerably, for we may find students of civil and military
schools, schools of an intermediary and higher level, Dutch and foreign
schools, schools specifically for mechanical engineers or for other
disciplines, like building or civil engineering . Hut the practical training is
not clearly defined, either. This will become clear by looking more closely
at the types of firms where these technicians were employed.
The group of technicians worked or had worked in 160 different
engineering firms and workshops. In total, there were more than 200
companies in The Netherlands in the nineteenth century engaged in
mechanical engineering. They were spread all over the country, with
concentrations in the coastal provinces of North and South Holland. As
in most other countries, the factories and workshops differed in a number
of respects:"
- Origins: some had started as forges, others immediately started as
engineering factories, a number of others developed from trading
companies, millwrights' shops, coppersmiths and the like.
- Size: from the 1830s, The Netherlands had some large machine building
firms: Feijenoord in Rotterdam and Van Vlissingen in Amsterdam,
each employing approximately 800 people in 1847. However, through-
out the nineteenth century the smaller firms dominated. Around the
turn of the century the average company in the industry had some 100
employees. Hut the majority of companies (some 600/0) employed fewer
than 25 men.
160 H. LlNTSEN, G . VAN HOOFF AND G . VERBONG
However, this does not mean that there are no signs whatsoever of a
professional community. From 1860 onwards there was a nucleus among
the mechanical engineers which, by the turn of the century, would become
a community with a "school" cuIture. This core was formed by higher
technicians employed by the major engineering companies, the navy,
government services (including the State Board of Controllers of Boilers)
and the workshops of railway companies. However , the influence of this
group on mechanical engineering in The Netherlands would only be feIt in
the twentieth century.
The lack of any professional association then raises the question: can this
kind of heterogeneous and fragmented profession have played any role of
importance in industrialisation (in particular in connection with the supply
of capital goods) and in the transformation of engineering into a more
technical science. In answering this quest ion we will concentrate on the
construction of steam engines, and elaborate two aspects. What contri-
bution did the Dutch machine building industry make to the construction
of steam engines? How were Dutch steam engines built, what quaIi-
fications and knowledge were required, and how were these obtained?
Where the application of steam technology was concerned, The
Netherlands was not wholly dependent on foreign countries. Even in the
first half of the century, when steam was used only in a limited number of
applications, approximately half of the installations were of Dutch
manufacture. The origins of 17I steam engines are known, of which
number 85 engines were built in The Netherlands, 36 in Belgium, and 46
in other countries. 7
After 1850, when the country changed over from mainly wind and
horse power to steam power, the Dutch machine industry continued to act
as a major supplier of steam engines. Around 1880, 500/0 of all steam
engines were of Dutch manufacture. Fifteen years later the Dutch share
had increased to 55% . Around 1880, some 56% of all boilers were made
in The Netherlands, by 1910 this percentage amounted to 64% .8 When
assessing whether this is relevant or not, one ought to consider that the
Dutch engineering factories and workshops had to compete on an open
market with Belgium, England and Germany, countries with a much
longer tradition in engineering .
162 H . LINTSEN , G . VAN HOOFF AND G . VERBONG
Quite independently of this assessment, one may ask how it was possible
for a country with such a fragmented profession in mechanical
engineering to be able to manufacture steam engines on such a large scale.
The answer is essentially very simple: the existence of a professional
community of mechanics or mechanical engineers was not an essential
precondition for the production of capital goods. This may become clear
by looking at the actual production of steam engines.
Building a steam engine involves a wide range of metal working
techniques:" casting, forging, machining, and joining . Cast iron and
wrought iron were the most important materials, together with other
materials such as brass and bronze, and, later in the nineteenth century,
steel. Initially timber was also used, even for some highly stressed
components, such as the crankshaft and the beam. However, in the course
of the nineteenth century the significance of wood decreased .
The process of casting involves pouring molten metal into moulds, so
that, after cooling, the required shape is obtained. Specialist skills
included the making of models, the making of moulds and the melting of
the various alloys according to precisely defined procedures to achieve the
right material properties. Apart from casting, some of the preparatory
treatment of the metal took place in the forge, in which wrought iron was
heated locally, without actually being molten. The required shapes were
produced using hammers and anvils, dies, mandreis and other tools.
Large forgings were shaped by steam and drop hammers in the second
half of the century and hydraulic forging presses by the end of the
nineteenth century .
In forging and casting, it was difficult or even impossible to give the
part the precise shape required . Especially in critical areas it was not
possible to guarantee adequate accuracy . After casting, a component
shrinks during the cooling process. Forgings were made by hand and
under visual control. No two forgings were wholly identical. Conse-
quently many components required further finishing. Many of these
operations in the nineteenth century were mainly performed using hand
tools. However, there was also an increasing use of machine tools . The
most important machines were lathes, drilling machines, planing
machines and grinding machines. Joining and finishing the various
components took place in the fitting shop which had the workbench as the
central point. Drilling, hammering, knocking, chamfering, bending,
MECHANICAL ENGINEERING IN THE NETHERLANDS 163
straightening and filing were therefore the most important skills the fitter
had to possess, next to a basic sense of accuracy.
The factory or workshop building a steam engine did not necessarily
perform all these activities itself. Often it did not have a foundry, and
castings were supplied by specialist foundries . Forging, machining and
joining operations were usually performed in the factory or workshop,
even though in some cases the emphasis was on fitting, most components
being bought in. For example, the making of the cylinders with minimum
tolerances was aprecision job. But for this special tools had been
developed, which were also available to the Dutch engineering industry.
A large part of the work consisted of skilled work which required years
of experience. The operatives received this practical training on the job. A
small number had received a more formal training in industrial and
drawing schools (after 1830), technical schools (after 1860), secondary
technical schools (after 1850) and, in some cases, company training
schools. The required "theoretical" skills, such as measuring and simple
drawing, the reading of drawings and the like were taught during the
training at such schools, or were acquired through horne study. Recruiting
qualified labour was done by in-house training, by taking over
experienced workers from other firms in the same field, or by taking on
independent craftsmen, mainly from the metal working sector . In the
early period of machine production, skilled workers from abroad
dominated. They played an important role far into the nineteenth
century, but at the end of the century the large majority of skilled workers
were Dutch.
Although the very first steam engines in The Netherlands probably all
came from England and Belgium, we saw that Dutch engineering firms
soon supplied quite a substantial proportion. Before 1850, at least sixteen
companies produced steam engines .!" However, the differences were
significant. One company, P. van Vlissingen & Dudok van Heel of
Amsterdam, built 28 or more engines during this period. Some others
produced from three to ten machines, whereas the majority only
produced one or two engines. This also included some firms - not
engineering companies as such - who constructed steam engines for
themselves, such as H . Bekker of Gemert, who, in 1829, built a 4 HP
steam engine for his own cotton spinning mill.
Another example of a machine shop with a small production was J .A .
Mercx of Tilburg, who was a watchmaker by trade. Having worked for
some time in an engineering company in Aachen (Germany) , he returned
164 H . LlNTSEN, G. VAN HOOFF AND G. VERBONG
to his native town in 1835. There he started with the installation of new
steam engines and the repair of existing ones. In 1840 his affairs had
grown to such an extent that he needed a steam engine to drive lathes and
drilling machines. He built it in his own workshops, and bought a second-
hand boiler from a neighbouring cloth manufacturer. The components of
the engine were in part made by other machine shops . During the
following ten years he only built one further steam engine. Then the
production increased. During the nineteenth century eighty engines were
produced in total, including a nu mber of 150 HP compound engines.
The construction of steam engines did not really pose a problem in The
Netherlands in the last century. Even a clever entrepreneur or technician
who was not involved in the engineering industry was able to rig up a
steam engine. However, this does not say anything about the quality of
steam engine engineering. This depended on the skills of the engineers , on
the workshop, and on the design of the steam engine. We now turn to the
latter, for it is the design of the installation which is at the heart of
engineering science.
During the whole of the nineteenth century - internationally speaking
- we see how the use of steam is brought to perfection, both technologi-
cally and economically. During this period there were numerous debates
and numerous discussions; '! should steam engines be vertical or
horizontal? In which way could one gain maximum advantage from the
expansion of steam? In what cases are condensers to be used, and of what
type? What is the ideal valve gear? Which design allows the minimum
steam consumption, etc.?
Such questions were the starting point for research, development and
design, and they were mainly the concern of technicians with a higher
education. What interests us here is the contribution of the Dutch
engineers and technicians to the evolution of steam technology and the
way in which they used their acquired knowledge in their designs. Is an
adequate reaction to the international technical development to be
expected from the fragmented group of higher technicians and engineers?
In view of the fact that steam engineering was developing on a broad
front, we shall confine ourselves to one aspect, namely the dimensioning
of the steam engine . First of all we will describe the international consider-
MECHANICAL ENGINEERING IN THE NETHERLANDS 165
-----
~
~ "" ""\
\
I \
f
I \
I \
I I
D I J
i =---------'~
\ I
\ I
\ I
\ I
\ /
" " -, /
10,000 x ~D2 X C
Ni X Pi
75 4
T
indicator
power
T
indicator
pressure
Talge
surtace velOCity of piston
Pj~
T,
C 2· n· s / 60
T.,
effeetive power
numberof
revolutions
stroke
ations and theories related to this question. Then we will return to The
Netherlands and investigate the Dutch position in more detail.
What is the dimensioning problem? It deals with the piston diameter,
the length of the stroke, the steam pressure and the speed of the steam
engine in order to achieve a specified power. The basic formulae for the
calculation of these dimensions are given in Fig. 1. The basis for this
calculation - the indicated capacity is proportional to the pressure in the
cylinder, the piston surface and the average piston speed - was already
166 H . LINTSEN, G. VAN HOOFF AND G . VERBONG
known in the eighteenth century. However, these formulae were not used
for design purposes until the beginning of the nineteenth century." This
was not necessary, since there were only a few types of engines,
Newcomen and Watt beam engines, for which there were tables with the
dimensions and their relation to capacity. These tables were based on
actual, working steam engines.
In the course of the last century, calculating methods became in-
Pi
Pmax. 1-----1
Pmin.
Pi = Pressure in cylinder
Pmax . = Max. pressure
Pmin. = Min. pressure
V = Volume
Vb = Volume at the beginning of expansion
V1 = Volume cylinder
Starting point
- Steam is an ideal gas
- Temperature does not change during expansion
- The law of Mariotte applies: P max Vb = p.V
Fig. 2. Calcu lation of Pi of a steam eng ine with expans ion (before Pambour).
MECHANICAL ENGINEERING IN THE NETHERLANDS 167
Starting-point
- Steam is saturated during expansion
- Temperature changes during expansion
- The law of Navier applies:
a
v=--- a , ß Practice-induced
ß+ P coefficients
P max. Vb +
Figure 2 illustrates the relation between steam press ure and cylinder
volume. During part of the stroke steam is supplied until volume Vb is
reached . The steam supply is then cut off, the steam expands during the
rest of the strake, and the pressure drops. The "classic" theory assurnes
that steam is an ideal gas, that the temperature does not drop during
expansion, and that Mariott's Law applies. The average pressure in the
cylinder could therefore be calculated by means of an integral (see Fig . 2).
Starting-point:
- Steam is not saturated during expansion
- Corrections on Navier and several other aspects of
steam during cyclus
- Corrections by Regnault, Rankine, Fairbairn, Zeuner, Clausius
The third category of books was intended for a higher level of technical
education and related functions, like chief engineers, directors,
supervisors, and army officers . It is typical of the situation in The
Netherlands, that only two Dutch authors in the nineteenth century wrote
books like this, namely Verdam at the beginning, and Delprat in the
middle of the century.
Verdam was a young and promising lecturer at the University of
Groningen for the application of chemistry and mechanical engineering to
the "useful arts". Although his lectures ceased in 1828 due to a lack of
interest, he remained active in mechanical engineering. Between 1829and
1837 he published a six volume work, Grondslagen der toegepaste
werktuigkunst [Foundations of Applied Mechanics), of which three
volumes deal with steam engineering." He was also an adviser to the
government in engineering matters and was responsible for the State
inspection of the safety of steam installations in The Netherlands.
Verdam had a thorough knowledge of and rich experience with steam
engineering. The completely new approach by Pambour is something that
he did not mention. He could not have known them in any case, as
Pambour did not publish his innovative ideas until 1838.
Delprat, an army officer , studied at the famous Ecole des Ponts et
Chaussees in Paris for some time . From 1821 he was a lecturer, first at the
Koninklijke Artillerie en Genieschool [Royal School of Artillery and
Engineering) in Delft . In 1829 he left this school for Breda where he
became lecturer at the Royal Military Academy, where he eventually was
appointed commander. His textbook Beginselen der werktuigkunde voor
de kadetten der artillerie, genie en van den waterstaat [Principles of
Mechanics for Cadets in the Corps of Artillery, Engineers and Public
Works), appeared in several editions, the first in 1842.11 There he
accurately followed the theoretical developments . He was weil acquainted
with the theories of Pambour, Carnot, Regnault, Clausius and others ."
It is remarkable that the one school for mechanical engineers, the
Polytechnic School which was set up in Delft in 1863, did not develop its
own textbooks on mechanical engineering . The emphasis in education in
mechanical engineering was on drawing, on reading and explaining
technical drawings and on the design of steam engines." The basis for the
calculations needed for design was dealt with in the section toegepaste
mechanica (applied mechanics). Between 1864 and 1876, Dr. L. Cohen
Stuart - civil engineer and mathematician and scientist, also director of
the Delft Polytechnic - dealt with it in a weekly two-hour lecture, for
172 H . LINTSEN, G . VAN HOOFF AND G . VERBONG
I Ni
ri=O,77 und - = 1,29(Taj V);-=43.I;Ni= 1,5*43,1 =64,7;
r. c
Pi = 0,46*4 - -1,024*0,21 = 1,62(Taj.III);
75 1
F= 10000 *43,1* 1,62 =0,1996;
n = 30c = 45.
s
However, this is not to say that the design of the steam engine is a simple
matter. We are mere1y indicating that authors of technical books
managed to make complex information easily accessible and fit for
immediate practical use.
Whether technicians actually used the new approaches - which was our
third question at the beginning of this section - will have to remain an
174 H . LINTSEN , G . VAN HOOFF AND G . VERBONG
open question. One might expect so, but it is hard to prove, since the
formal calculation procedures are rarely found in the archives. Frequently
there are drawings, but the preparatory stages leading up to them are
difficult to reconstruct accurately . Sometimes drafts are found but the
design process employed by a mechanical engineer in the last century is
hard to reproduce.
social relations are looking for recognition and power and which are
trying to create monopolies in certain areas of the labour market. This
certainly applies to the "school"-based communities and more specifi-
cally to the engineering professions, as Layton has described it for the
American engineer and Lundgreen for the engineers in Europe and
Lintsen and Disco for the Dutch engineers." So beware of the rhetoric of
the engineers. Do not believe its advocates if the professional community
considers itself to be indispensable to industrial and technical develop-
ment.
NOTES
Dom inance of School Culture and the Engineering Professions , Annals of Science 47(1990),
pp . 33-75.
3 H .W . Lintsen, Ingenieurs in Nederland in de negent iende eeuw. Een streven naar
9 Ibid. , Ch . 2.
(Rotterdam, 1860); N.A. Irnelman, Hel stoombedrijf. Beknopte handleiding bij de studie
van het geheeie stoomwezen voor machinisten en studeerenden (Deventer, 1921); see also:
A. Jongkees, Beginselen der stoomwerktuigkunde. Leidraad bij het onderwijs van
machinist-leeriingen (Hellevoetsluis, 1884) (2nd ed. 1891); D.J . Wagner, Beschrijving der
inrichting en werking van de stoomwerktuigen (Dordrecht, n.d.); J.C. Graue, Praktisch
rekenboek en handleiding voor het onderwijs in de stoomwerktuigkunde (Leiden, n.d.);
A.D.F.W . Lichtenbelt, Handleiding bij hel onderwijs in de beginselen der stoomwerktuig-
kunde (Rotterdam, 1905); J.A. Dittlof Tjassens, Leerboek der stoomwerktuigkunde
(Leiden, 1882) and Stoomwerktuigkunde, in four parts, with illustrations (Marine-
Machinistenschool, 1914-1915); Stoom , handleiding voor het stoombedrijf (Groningen/
Amsterdam , 1929); J .H. Harte, Volledig machinenboek ofhandboek 101 de theoretische en
praktische kennis van alle soorten van stoom en andere werktuigen en derzelver onderdeelen
(Gorinchem , 1852); E.F. Scholl, De gids voor machinisten, 4th ed, (Leiden, 1892).
• 6 G.J . Verdam, Grondslagen der toegepaste werktuigkunst, in six parts, with atlas
(Groningen, 1829-1837).
17 J.P. Delprat, Beginselen der werktuigkunde voor de kadetten der artillerie, genie en van
(Amsterdam, 1870). This textbook goes into detail about the theories of Pambour and
others . Other foreign publications translated into Dutch: Chr. Cremer, Hel stoomwerktuig
(Rotterdam, n.d.): simple handbook for direct technical training, G. Pipijn and A. van den
Steen, Leerboek der werktuigkunde, stoomketels & stoommachines met praktisch
onderricht o ver het vervaardigen en behandelen der gewone stoomwerktuigen (Gent, 1876);
textbook for direct teaching at technical colleges (The fourth ed, appeared in 1897); J.
Bourne, Leer- en handboek der stoomwerktuigkunde (Nieuwediep, 1858); practical
handbook for technical operatives and companies, Bernoulli's Vademecum , practisch
handboek voor berekeningen, dagelijks voorkomende in de bouw- en werktuigkunde,
originally translated by J.G. van Gendt jr ., revised by G.J. W. de Jongh (Amsterdam, 1884).
23 Des Ingenieurs Taschenbuch (Berlin, 1902), p. 822.
24 E. Layton, The Revolt of the Engineers; Social Responsibility and the American
INTRODUCTION
This model [of technology as applied science] ... assumes that science and technology
represent different functions per formed by the same community. But a fundamental fact is
that they constitute different communities, each with its own goals and systems of values .
They are , of course, similar in that both deal with matter and energy, But these similarities
should not be overstated. Each community has its own social controls - such as its reward
systern - which tend to focus the work of each on its own needs . These needs determine not
only the objects of concern, but the "language" in which they are discussed. These needs
may overlap; but it would be surprising if this were a very frequent occurrence. One would
expect that in the normal case science would beget more science, and technology would lead
to further technology. I
possibility that science and technology may not retain distinct boundaries
and separate identities. It suggests, indeed, that we should expect to find
a phenomenon of hybridization occurring, in which a merging or blending
of traditions, practices, knowledge, and values from the worlds of science
and technology takes place.
The present paper focuses on this phenomenon of hybridization at the
level of individual , hybrid careers. The idea of a hybrid career was hinted
at by Edwin Layton in his article, 'Mirror Image Twins: The Communities
of Science and Technology in Nineteenth Century America' . There, in a
brief paragraph, he made the comment that information exchanges
between the worlds of science and technology depend upon individuals
who straddle both worlds:
For information to pass from one community to the other often involves extensive
reformulation and an act of creat ive insight. This require s men who are in some sense
members of both communities. These intermediaries might be called " engineer-scientists"
or " scientist-engineers" , depending on whether their primary identification is with
engineering or science. Such men playa very important role as channels of communication
between the communities of science and technology."
The aim of this paper is to take this observation further, and to explore
hybrid careers as a means to learn more about how the domains of science
and technology are interconnected. The term itself, "hybrid career" is
intended to suggest two methodological criteria for such an analysis.
First, "career" is intended to emphasize the idea of a path over time; it
emphasizes that we need to look at how lifetime career paths wend their
way back and forth between the domains of science and technology. Such
an approach is important because the structure and trajectory of a career
path helps to shape the activities that are carried out in each domain, and
thus influences the way these domains are interlinked.
Second, the term "hybrid" is intended to suggest that an individual
need not retain a primary affiliation with either the scientific or
technological community over the course of a career (as Layton implies).
The term further cautions us against the common practice of designating
someone's entire career as "scientific" or "technological" based on a
segment of their career path . Too often such designations lead to the
assumption that any crossing of boundaries by an individual from one
domain to the other is a temporary aberration or sideline with little
bearing on the person's "principal" line of creative work . In contrast, the
idea of a hybrid career suggests that such crossing of boundaries may be
180 EDA KRANAKIS
Navier was trained as an engineer and served in the state engineering corps
throughout his career. As a student, he attended the Eco/e Po/ytechnique
and the Eco/e des Ponts et Chaussees . After graduating in 1807, he
became a memb er of the Corps des Ponts et Chaussees, stationed in Paris,
with the rank of ingenieur ordinaire. By 1822 he had attained the rank of
Ingenieur-en-chef:" Over the years Nav ier was involved in many practical
engineering projects, including the design and construction of several
THE INTERACTION OF SCIENCE AND TECHNOLOGY 181
As to the election, I forbid myself, as you can imagine, any remark about the person or the
work of my competitors.... In confining myself thus to general considerations, I cannot help
but hope that longer consideration will perhaps bring a person who has given great service
to rnathematics, on the belief that this species of knowledge is that which it is most irnportant
to encourage and to honour, and which will always const itute the most solid foundation for
the glory of the Academy. ' 9
problem. What he did was, first, to develop a very abstract and idealized
mathematical model of the technology he was investigating, and then to
add complexifying hypotheses, one at a time, so as to gain more accurate
information about selected features of the system's behaviour. To return
to the example of the suspension bridge, Navier first modeled it simply as
a massless, perfectly flexible, inextensible cable loaded uniformly along
the horizontal. He then progressively modified his analysis to account for
the influence of the following factors on the cable's equilibrium (each
factor being considered separately):
(1) the actual weight ofthe cable and ofthe hangers (which link the deck,
or roadway, to the cable);
(2) additional concentrated or distributed loads on the cable;
(3) the cable's actual elasticity (i.e. the elasticity of iron) ;
(4) the expansion or contraction of the cable due to temperature changes;
(5) the kind of tower-cable system adopted.
Navier's hybrid repertoire clearly influenced the kind of knowledge he
produced. For example, he contributed significantly to the growth and
development of the theory of elasticity and the theory of structures, but
hardly at all to the elaboration of experimental, design, or production
techniques in engineering (nor did he ever take out a patent.) Navier was
part of an engineering elite in France, and the interorganizational network
of which he was apart, while it brought hirn into close contact with French
scientists, tended to isolate hirn intellectually and socially from engineers
and entrepreneurs in the private sector . The kind of knowledge Navier
produced reflected this situation. He was not one to devise simple,
accessible rules or formulas that the technologists of his day could readily
apply. To read his work at all presupposed a knowledge of mathematics
and theoretical mechanics that was far beyond the ken of most of them .
Indeed, when considered relative to the standards of most technologists of
his time, the kind of knowledge Navier produced appeared esoteric and
not immediately relevant to practice.
According to the standard that Thomson had set up it is not sufficient to obtain an analytical
result, and to reduce it to numerical computation, every step in the process must be
associated with some [physical) intuition, the whole argument must be capable of being
conducted in concrete physical terrns."
Applied in the case of electricity, this standard led Thomson to the view
that the concreteness and proven efficacy of his telegraphy theory
(physically embodied in the Atlantic cable) was a valid justification for
accepting a theory of electricity grounded in the same model. In contrast,
he rejected Maxwell's theory of electromagnetism, in part because he
could not associate it with any concrete, physical intuition or model.
Another crucial element in Thomson's scientific repertoire, which also
reflected his technological experience, was a commitment to precision
measurement. He saw precision measurement taking on an increasingly
important role in technological development. Time and again, either
through his own experience or through his knowledge of the experiences
of his engineering colleagues, he saw thousands of dollars being saved or
lost through measurement or inattention to measurement; he saw
companies and technological projects succeed or fail, seemingly directly
in relation to their directors" commitment to precise measurement. To
give one example, Thomson's telegraph theory showed quantitatively the
retardation of the signal that would occur in underground and underwater
cables, relative to their length. The theory further showed how such
effects could be minimized or compensated for by increasing the diameter
of the wire and the thickness of the insulation in relation to the cable's
length . Yet Thomson realized that to be able to design cables in accord-
ance with this theory, precise measurements were required, for example
on the resistivity of copper and the specific inductive capacity of gutta-
190 EDA KRANAKIS
3 . VILHELM BJERKNES 3 •
Unlike Navier and Thomson, Bjerknes did not have a dual career in the
sense of working alternately as a scientist and engineer (or inventor).
Rather, as we will see, Bjerknes 's links to the world of technology came
more indirectly, through a career shift which brought hirn from theoretical
physics to meteorology, and thence to practical forecasting. As a student,
in the years 1889-1891, Bjerknes , studied mathematics and mathematical
physics in Paris and worked in the laboratory of Heinrich Hertz in Bonn .
He returned to his native Norway to complete his studies , and received his
Ph.D. there in 1892 for work on the application of mechanics to physics.
The following year Bjerknes took up a lectureship in mechanics at the
Stock holm Hogskola, where he remained untill907. He then accepted a
professorship in mechanics and mathematical physics at the Royal
Frederik University in Christiania (Oslo), where he remained untilI913 .
In the first phase of his career, Bjerknes pursued a research program
aimed at creating a mechanistic foundation for all of physics. More
specifically, he sought to unite electrodynamics and hydrodynamics, by
developing a mechanics of the ether. In this work, however, Bjerknes
found hirnself becoming isolated from the mainstream of interest in
physics. For whereas his research was rooted in a mechanical world view,
the dominant trend in physics was toward research programs rooted in
electromagnetic and energeticist world views.
In response to this situation, Bjerknes began to orient his research
toward meteorology. He had initially been introduced to problems in
meteorology by researchers in that field who saw relevance in his work on
192 EDA KRANAKIS
ether mechanics. And he found that he could have an impact and a degree
of recognition in meteorology that he could never hope to achieve in
theoretical physics, given the latter's changing focus. Bjerknes therefore
transformed his goal of developing a complete mechanics of the ether into
the goal of developing a precise mechanical physics of the atmosphere. In
1903 he formulated such a research program explicitly, and by 1906 it
became his principal, long-term research commitment.
On a professional level, Bjerknes's growing commitment to meteor-
ology was reflected in his acceptance, first, in 1913, of a position as
director of a newly established geophysics institute in Leipzig, and then,
in 1917, of a professorship in Bergen, Norway, attached to the
Geophysical Institute of the Bergen Museum. This commitment was also
reflected in his growing involvement with aeronautical and meteorologi-
cal associations. For example, he became a member of the Swedish Aero-
nautical Society, the Norwegian Ballooning Society, the Leipzig
Aeronautical Society, the Berlin Society for Aeronautics, and he became
closely involved with the International Commission for Scientific
Aeronautics, the major forum of the international meteorological and
aerological communities.
As Bjerknes shifted his interest to meteorology, he became increasingly
drawn into the realm of technology . For meteorology was at the inter-
section of a developing interorganizational network that linked scientists
interested in studying the physics of the atmosphere, like Bjerknes , with
military, business, and commercial groups whose activities required or
benefitted from more accurate weather forecasting : farmers, fishermen,
shippers, aviators, commercial airlines, etc. Representatives from many
of these groups participated actively in aeronautical and meteorological
organizations, such as the International Commission for Scientific
Aeronautics, and they lobbied (either through these organizations or
through the government) to have meteorological practice and research
defined around their needs.
World War I had the effect of considerably broadening and strength-
ening this interorganizational network. Perhaps most notably, weather
forecasting and the collection of weather data came to be seen as essential
for military planning, and for military and civilian aviation. In the words
of Bjerknes, WWI brought about "the militarization of meteorology" .37
A number of government weather bureaus, such as France 's, were placed
under military control during the war, and kept this organization even
after the war. In 1919, the British Meteorological Office, which had been
THE INTERACTION OF SCIENCE AND TECHNOLOGY 193
independent from the time of its organization in 1867, was placed under
the control of the Air Ministry and was given "for its primary and
immediate object the satisfaction of the requirements of aircraft" . 3 8
Bjerknes participated in extending and strengthening this inter-
organizational network. He served on national and international
committees concerned with Iinking meteorology to the needs of aviation.
These included a commission established in 1919 (und er the auspices of
the International Meteorological Committee) to investigate the
application of meteorology to aviation, and also the Norwegian
government's Commission on Aerial Transport. He wrote reports and
proposals for the military and for government outlining the advantages of
Iinking meteorology to military planning, farming, and the like. And
most importantly, he took steps to establish a practical weather
forecasting service that would meet the needs of these groups.
Bjerknes's decision to take up weather forecasting occurred shortly
after his return to Norway to take up the chair at the Bergen Museum. He
soon realized that the Bergen Geophysics Institute was not going to
receive the level of funding he had anticipated, and that his best chance to
obtain adequate resources to do research and train students was to orient
his work more closely toward immediate, practical forecasting needs. He
therefore lobbied in 1919 to obtain private and government funding to
establish a weather forecasting service at his institute. He had already
established such a service temporarily during the summer of 1918 with
funding from the Norwegian parliament. This temporary service was
intended mainly to provide weather forecasts for farmers, as a means to
help alleviate serious food shortages resulting from wartime trade
restrictions. In his 1919 proposal, Bjerknes pointed out the continuing
advantage to farmers of weather forecasts, and also the benefits it would
offer for commercial aviation, which was just being established in
Norway. (Two years earlier, in a secret report to the Norwegian minister
of defense, Bjerknes had also suggested the need to establish a field
weather service for the Norwegian military, and he suggested several ways
in which his geophysical institute could contribute to this effort.)
Bjerknes's proposal for a weather service was accepted, and thus he
established and became director of the West Norway Weather Bureau.
Since forecasting depended upon extensive data collection , Bjerknes and
his group became involved in establishing the necessary infrastructure for
the collection and transmission of weather data, e.g. in setting up obser-
vation stations and wireless telegraphy stations. Bjerknes also worked
194 EDA KRANAKIS
CONCLUSlONS
The model of hybrid careers set out in this paper depicts them as being
linked, on one side, to interorganizational networks, and on the other
side, to the emergence of hybrid repertoires of practice. I illustrated this
model and its two organizing concepts by looking at the careers of Navier,
Thomson, and Bjerknes. Now we need to go a step further, however, and
THE INTERACTION OF SCIENCE AND TECHNOLOGY 197
University 0/ Amsterdam
THE INTERACTION OF SCIENCE AND TECHNOLOGY 201
NOTES
emphasize the boundaries and cognitive differences between the worlds of science and
technology, Layton hirnself acknowledged that these two domains have become increasingly
intermeshed since the 19th century. He also hin ted at the need for a more integrated model of
science and technology: "In many modern social contexts physics and engineering have
become so intermixed that it is difficult if not impossible to sort them out into neat
pigeonholes . For example, in studying large, interdisciplinary research organizations, it is
often more helpful to think of physicists and engineers as part of a complex 'research system '.
Many men trained in physics 'do' technology, just as many men trained as engineers 'do'
science, including 'pure' or undirected research ." See Edwin T. Layton, 'Technology and
Science , or, Vive la Petite Difference' , Philosophy ofScience Association 11(1976), pp. 173-
183. The modell am proposing here is intended to allow for the possibility of the emergence
of this kind of integrated "research system",
4 In using the terms "science" and "technology" to refer to these realms of activity, I am in
effect adopting a convenient shorthand. I do not mean to suggest that the lauer are
historically unchanging.
5 My idea of science and technology as intersecting worlds owes a lot to the social worlds
excellent studies by Crosbie Smith and M. Norton Wise, on William Thomson, and by Robert
Mare Friedman on Vilhelm Bjerknes. See Smith and Wise, Energy and Empire: A
Biographical Study of Lord Kelvin (Cambridge: Cambridge University Press, 1989); and
Friedman, Appropriating the Weather: Vilhelm Bjerknes and the Construction ofa Modern
202 EDA KRANAKIS
Construction des Ponts, C.-L.-M .-H . Navier (ed.) (Paris : Firmin Didot, 1809). This work
contains a biography of Gauthey written by Navier.
11 On the Societe Philomatique, see Maurice Crosland, The Society 0/ Arcueil (Cambridge,
John Herivel, Joseph Fourier: The Man and the Physicist (Oxford: Clarendon Press, 1975),
pp . 128-129.
IJ A. Brunot and R. Coquand, Le Corps des Ponts et Chaussees (Paris: Editions du Cent re
National de la Recherche Scientifique, 1982), pp. 3-37; Charles Coulston Gillispie, Science
and Polity in France at the End 0/ the Old Regime (Princeton : Princeton University Press,
1980), pp. 479-498; Anne Querrien, 'Ecoles et Corps : Le Cas des Ponts et Chaussees, 1747-
1848,' unpublished MS, Bibliotheque de l'Ecole des Ponts et Chaussees; Coteile, Esquisse
Historique sur I'Institution des Ponts et Chaussees (Paris : Paul Dupont, 1849).
14 A. Fourcy, Histoire de i'Ecole Polytechnique, Jean Dhombres (ed.) (Paris : Belin, 1987),
passim ; Terry Shinn, Savoir Scientifique et Pouvoir Social: L 'Ecole Polytechnique, 1794-
1914 (Paris : Presses de la Fondation Nationale des Seiences Politiques, 1980), pp . 9-23 .
IS CharIes Dupin, Eloge de M. le Baron de Prony, Chambre des Pairs, 2 April 1840, pp. 22-
Pierre Costabel , and Pierre Dugac (eds.), Simeon-Denis Poisson et la Science de son Temps
(Palaiseau : Ecole Polytechnique, 1981), pp . 95-104; Stephen P . Timoshenko, History 0/
I.
Strength 0/ Materials (1953, rpt. New York : Dover, 1983), pp. 119-122.
C.-L.-M.-H . Navier, Rapport aM . Becquey et Memoire sur les Ponts Suspendus (Paris :
Irnprimerie Royale, 1823), 2nd ed. Paris: Carilian-Goeury, 1830. Navier's theory of
suspension bridges is examined more fully in Eda Kranakis, 'Navier 's Theory of Suspension
Bridges,' in J .L. Berggren and B.R . Goldstein (eds.), ' From Ancient Omens to Statistical
Mechanics: Essays on the Exact Seiences Presented to Asger Aaboe' , Acta Historical
Scientiarum Naturalium et Medicinalium 39(1987), pp. 247-258 . The history of Navier's
Invalides suspension bridge over the Seine is reviewed in Eda Kranakis, 'The Affair of the
Invalides Bridge,' Jaarboek voor de Geschiedenis van Bedrijfen Techniek 4(1987), pp. 106-
130.
I' Claude-L. -M.-H . Navier to Sylvestre F. Lacroix, 9 October 1823, Lacroix MSS 2396,
Bibliotheque de l'Institut de France .
'0 Navier discussed these techniques in his Memoire sur les Ponts Suspendus, 2nd ed. p. 11.
11 This analysis of Kelvin's career is based on Smith's and Wise's comprehensive biography
and on two additional articles: M. Norton Wise and Crosbie Smith, 'Measurement, Work
THE INTERACTION OF SCIENCE AND TECHNOLOGY 203
and Indu stry in Lord Kelvin's Britain' , Historical Studies in the Physical and Biological
Seiences 17(1)(1986), pp . 147-173; and M. Norton Wise, 'Mediating Maehines', Science in
Context 2(1)(1988), pp . 77-113 . Specifie citations will be given only for direet quotations.
" Thomson's links with Walter Crum were partieularly close sinee he also married Crum's
daughter, Margaret, in 1852.
23 To understand the establishment of the Glasgow eha ir of engineering as part of a broader
Harmony of Theory and Praetice: The Engineering Scienee of W.J .M. Rankine',
Technology and Cu/ture 23(1982), pp . 39-52.
" Quoted from Smith and Wise, p. 283.
26 Ibid. , p. 284.
27 These are Thomson's own words , quoted from Smith and Wise, 'Measurernent, Work,
29 Ibid ., p . 19I.
34 The idea oeeurred to Thomson du ring aperiod in which his brother James was writing to
hirn about methods for analyzing the effieiency and "mechanical effect" (i.e, work) of
steam engines and water wheels . Signifieantly, William Thomson used the same terrn,
"rnechanical effeet", in his analy sis ofthe eondueting sphere s. It was not a widely used term .
James Thomson took it up from Lewis Go rdon, who had proposed it as a translation of the
German, mechanische Wirkung.
J5 This point is further explained in Smith and Wise, 'Work, Measurement, and Industry in
Weather, and from Robert Mare Friedman, 'Constituting th e Polar Front, 1919-1920', Isis,
73(268)(September, 1982), pp. 343-363 . Specifie citations will only be given for direet
quotations,
3 1 Quoted from Friedman, Appropriating the Weather , p . 145.
38 Ibid.
39 Bjerknes's need for data was one ofthe rnajor factors that initially led hirn toward greater
hypothesis.
41 It must be emphasized that many empirieal studies, even carried out with in the context of
204 EDA KRANAKIS
these models, provide evidence for a broader range of transfers between the domains of
science and technology .
.. By research ideologies, I mean normative ideas about how research ought to be done ,
what methods should be used, and what relationships should exist among research rnethods.
Research is here taken to include the full range of exploratory activities that scientists and
technologists engage in: theorizing, inventing, designing, experimenting , tinkering, etc.
Research ideologies offer guidance in answering questions such as the foHowing: is it best to
derive theories from empirical practice, or viceversa? How should theory and experiment be
linked in a research prograrn? How should new technological artifacts be designed - in
relation to empirical experience or to formal theories?
4J It should be noted that in the case of dual careers, like Navier's or Thomson's, the
transfer of knowledge, practices, and values between the worlds of science and technology
may result in not just one but two (interrelated) hybrid repertoires . For example, Thomson's
scientific and technological repertoires were not precisely identical, but because each shaped
the other, they did come to share many features. Due to space limitations , I did not examine
this phenomenon in the examples. Thus, in the case of Navier, I examined only his
engineering repertoire, while in the case of Thomson, I examined only his scientific
repertoire.
•• The existing repertoire may be either the individual's own, or it may be understood as a
shared repertoire, comprising knowledge, practices, and values common within a discipline
or field of expertise at a particular time and place.
• s This is not to suggest, however , that a hybrid repertoire would be a spur to creativity in
aH contexts or situations .
•• In asense, the professionalization of engineering can be seen as part of this process.
Moving the locus of technological training from the workplace to the university and
changing it from an apprenticeship into a formal education complete with books, lectures,
and instruction in experimental method s necessarily implies a broad transformation of the
technologist 's repertoire of practice.
47 In fact, as I have shown in another article, this was indeed the case. See 'Social
INDUSTRY-I10USE ESTABLISHMENT,
FOIl 2000 I'RllSONS. OF A\,L AORS.
~ For th p r.'111:In:llirJIlI flf the lU""rr31 F i:;:u rr tl of t hi" Pur r.. see .. Out lint'
z
n
of .. w erk, ('ntitl,.,1 I · A I ~ I· F.R M AS Ao r .H E1'lT H lrROVr;o :", Upnth:am'. [Tl
" ' ork8, rnl. viii., 11. 3li9 to l'' 4:l9 . ;I>
Z
Th e Han~(". of IIrll..."I:l:rll ne.l C ,i l ~ :Ir,. r"'~PN' t i\', ·I,. 1lU tr roltf"rI t n r rrnm
o
)':nd to E nd of tll(' ,,,,Nnl W:alle. n ~ r ,l.il,itl'll in 11... f : N " I ' ~ l l I' I.I\s : tlll"J' e re -l
[Tl
here ffr rt'lt'nlr ,l In ('111 t hrnuJ.:h Il~- a I .int' II:tralll·1 In t1.f' ~i, 11' flf ll H' I'ul!~on : o
in the lkd ..gta~,.... ,,·1.31 ill rrl 'rt'1'l("nlr " :111 nHr" in thf' l . ra ll ~l.I . ill r rorolMl to :I:
be in IIt'O in Ihe U("8("riplion. Z
o
F!G. I . - E I,F:\ ' ATIO/'ö.
r"
o
Cl
0<
OB!
Fig. 1. Drawing of Bentham's Panopticon . From Panopticon; or, The lnspection-House: Containin g the Idea of a Ne w Principle
of Construction Applicable to Any Sort of Establishment, in Which Persons of Any Description Are to Be Kept Under
lnspection; and in Particular to Penitentiary Houses, Prisons, Poor-Ho uses, Lazarett os, Hou ses of Industry, Manufactories, 9
Hospitals, Work -Houses, Mad-Hou ses, and Schools by Jerem y Bentham, The Works of Jeremy Bentham ,
Val. IV (1838-43 edit ion ) facing page 39.
208 JOHN M . STAUDENMAlER, S.L
Late in the 18th century Samuel and Jeremy Bentham designed the
"Panopticon", a "12 sided polygon formed in iron and sheathed in glass
in order to create the effect of 'universal transparency"'. 9 Samuel
Bentham invented the plan as an ideal factory for peasant workers in
Catherine the Great's Russia but it was his more prominent brother,
utilitarian philosopher Jeremy, who promoted the idea with the
entrepreneurial fervor of a missionary. He recognized that the model
could be applied not merely to factories, but to any institution aimed at
regulating the behavior ofthe unruly - orphans and the insane in asylums,
students in school, workers in factories, or prisoners. In 1787 he began a
several decade campaign for Parliamentary funding of a model prison, to
be managed by himself at a profit, in which a marginal ring of transparent
cells held isolated prisoners under the twenty-four-hour-a-day scrutiny
from the opaque inspectors' tower."
Bentham saw the prison as a factory ("a mill for grinding rogues
honest and idle men industrious"!') and scrutiny drove the mechanism for
manufacturing reform.
... the more constantly the persons to be inspected are under the eyes of the persons who
should inspect thern, the more perfectly will the purpose of the establishment have been
attained. Ideal perfection, if that were the object, would require that each person should
actually be in that predicament, during every instant of time. This being impossible, the next
thing to be wished for is, that, at every instant, seeing reason to believe as much, and not
being able to satisfy himself to the contrary, he should conceive himself to be so."
Bentham laid down the principle that power should be visible and unverifiable. Visible: the
inmate will constantly have before his eyes the tall outline of the central tower from which
he is spied upon . Unverifiable: the inmate must never know whether he is being looked at at
any one moment ; but he must be sure that he may always be so ... in the peripherie ring, one
is totally seen, without ever seeing; in the central tower, one sees everything without ever
being seen. 13
Onee imbued with the self image of living under the serutiny of unseen
eyes, twenty four hours a day, the prisoner eould be released into society.
Bentham intuitively grasped the nature of a profound revolution in the
West's understanding of how the individual relates to the governing
struetures of society. His Panopticon plan abandons the earlier judicial
foeus on the eriminal aet (with its appropriate punishment) and
eoneentrates on the criminal (and his/her need to be eorreeted) . In the
older way of doing things, if I am judged guilty of a erime I beeome liable
to a specific punishment, but I am not defined as a defeetive human nor
am I excluded from the larger human eommunity in principle. As in the
Greek term for sin, hamartia ("missing the mark"), my guilt has to do
with the aet I eommitted. Onee my debt is paid, it is assumed that I have
a plaee within the human community. This even holds in eases of eapital
punishment for soeieties that provided explicit reeoneiliation prior to
exeeution. Onee absolved, the repentant sinner died reunited with the
eommunity of the Chureh which, nevertheless, exaeted the death penalty
that aeerued to the eriminal aet. Presumably, both exeeutioner and the
exeeuted would hold eommon status in the heavenly eommunity in which
both believed and on which the praetice was based.':'
Bentham's plan abandons the assumption that erimes ean be paid for
by punishment without eroding the ordinary humanity of the eriminal.
The Panopticon redefines eriminals as defeetive and needing eorreetion.
Foueault sees this shift from punishable aet to defeetive eharaeter
pervading nineteenth and twentieth eentury European penal praetice;
judges have "taken to judging something other than erimes, namely, the
'soul' of the eriminal". 15 Judging the interior of the eriminal and applying
eorreetive remedies implies some exterior standard against which one is
measured and that in turn implies some person or group who get to set the
standard.
The unstudied rhetorie of those who claim the elite status of therapist
frequently reveals the depth of eontempt for those deemed in need of
repair that is implicitly part of this social arrangement. Thus Karl
Menninger writes:
210 lOHN M . STAUDENMAlER, S .1.
We, the agents of society , must move to end the game of tit-for-tat and blow-for-blow in
which the offender has foolishly and futilely engaged hirnself and us. We are not driven, as
he is, to wild and impulsive actions, With knowledge comes power, and with power there is
no need for the frightened vengeance of the old penaology. In its place should go a quiet,
dignified, therapeutic program for the rehabilitation of the disorganized one, if possible, the
protection of society during the treatment period, and his guided return to useful citizenship,
as soon as this can be effected. "
One could ask, of course, how any group of human beings acquires the
right to set standards for the correction of the character of other human
beings. How do some human beings become "the" agents of society and
members of this imperial "we", while others remain locked in the third
person (i.e., "the disorganized one" who needs "our" guidance to be
returned to useful citizenship)? The device ofthe invisible tower preempts
such questions by masking the power relationships at work here. The
desideratum of Benthamite incarceration is not a community wherein
identifiable people negotiate the norms of acceptable human behavior.
Instead, the Panopticon inculcates the experience of invisible scrutinizers
who, indeed, need not be in the tower at all.
t'was the signal for the demolit ion of everything they eould lay their hands on .... The
drapery around the boxes was torn , the eushions in the pit ripped open , the seats broken, and
ehairs were flying in all direet ions .
The next night, the chastened company performed "the last note that ever
Rossini cornposed" . 20 Audience-performer interaction was the norm and
public performance sometimes approached the chaotic.
After the Civil War, however, a new discipline began to take hold .
Thus, noted conductor Theodore Thomas insisted on silent audiences,
sometimes turning to stare them into submission before continuing the
performance. He and others like hirn conducted stern lessons in
conformity and passivity. This increasingly powerful movement of
audience reform reverses the polarities of Bentham's prison. The watched
hold the position of power at center stage. Losing the power to interact,
the audience is marginalized in the act of watching."
The new style of audience conformity started to take hold just as
electronic media began to revolutionize the very structure of public
discourse. Before Samuel Morse's invention of a telegraphic code, and for
the most part before the organization of the national wire service about
1870, information rarely traveled faster than a horse could trot. "The
news" reported on a minuscule universe, one's village and a surrounding
countryside perhaps thirty miles across. Messages from beyond horse-
range arrived long after things happened and their rhetorical form
differed accordingly. Pre-electronic newspapers published "corre-
spondences" , leisurely essays for readers who needed subtle details to
understand the gradual unfolding of far away events. On the local scene,
however, news writers and news readers shared the same living space.
Thus, the reader could ordinarily supply a host of nuances that the printed
account only suggested and disagreements about interpretation could be
settled right in town. Like the boisterous theater and music hall audiences,
newspaper readers could actively intervene in public discourse.
Beginning about 1870 national telegraphic wire services began to
change all that. 22 In 1876, for example, if I lived in Chicago I would
probably have read in the morning paper about the "Molly Maguire"
trials hundreds of miles away in the eastern Pennsylvania coal region. I
would learn that "the Mollies" were anarchistic Irish miners conspiring to
violently destroy the lives and property of coal mine owners. I would have
read, the morning after it happened, that twenty four were convicted and
ten hanged. I would not have known, however, that most historians have
214 lOHN M . STAUDENMAlER, S .1.
since come to interpret the trial as a frame-up and that the key witness was
a Pinkerton detective in the employ of the owners. 13 The wire service news
crafters did not choose to include that perspective and I, at my Chicago
break fast table , lived too far off to know more than the wire service told
me. Live radio news beginning in the 1920s and more recent television
coverage took instantaneous news one step farther. Instead of reading
about news one day late, the audience could now "participate" in events
as they happened. Despite this dramatic intensification of audience
involvement, however, the listeners or viewers gained not one whit of
active power to shape public discourse .
Arecent exception demonstrates the rule. During the Reagan-Bush
administrations, while the New York Times and other mainstream media
accepted the administration interpretation of Central America (e.g.,
struggling democracies in Salvador and Guatemala vs vile dictatorship in
Nicaragua), a grass roots network communicated a distinctly different
reading of events and significantly influenced national policy as a result.
One has only to note, however, the enormous individual, group, and
church efforts involved to recognize the untypical character of the
example. Whether I favor or oppose the mainstream ideology is not at
issue here. Whatever my ideology vis avis the electronic media version at
any given time, I typically relate to electronic discourse as a passive,
isolated and powerless member of the audience , the very model of a
Benthamite prisoner in reverse, frozen in the act of watching. Even when
I shout at the TV because of a particularly odious ad or newscaster
remark, nothing public or civic happens . It wears me out to try to imagine
that I might change things.:"
In viewing the urban masses, advertiser s associated consumer lethargy as much with weak-
kneed conformity as with cultural backwardness.... Emotional appeals succeeded because
only by seeking this lowest common human denominator could the ad vertiser shake the
masses from their lethargy without taxing their Iimited intelligence."
Across the Atlantic and more than a century after Bentham, a second
factory design came to be seen as an icon of the spirit of modernity. This
factory, however, acquired the capital investment and physical reality
that eluded Bentham. Within a few years of its completion in 1923, the
Ford Motor Company's massive River Rouge plant achieved mythic
stature and world acclaim as the ultimate expression of "Fordismus", the
triumph of rational efficiency over nature and the burdens of life.
Tourists from around the world, by the hundreds of thousands made
pilgrimage to the Rouge in the late Twenties. German engineer, Otto
Moog was not unusual when he recorded his impressions in language that
combines an almost schizoid mixture of quasi-religious intimidation and
awe with an exultant sense of liberation.
No symphony , no Eroica, compares in depth, eontent , and power to the music that
threatened and hammered away at us as we wandered through Ford's workplaces ,
wanderers overwhelmed by a da ring expression of the human spirit. ,.
When construction began along the banks of the Rouge River in 1915,
"Henry Ford" was already a household name . More than half the autos
in the United States were Model Ts, arguably the most successful match
between a single technical design and its societal context in recorded
history . For weIlover a decade, Ford sold the ugly, durable vehicles as fast
as they could be manufactured. Ford had already begun to capture world
attention when, in 1914, his simultaneous completion of the moving
assembly line and doubling of wages ("The Five Dollar Day") stunned
competitors and intensified his image as industrial genius and working
man's friend .
Less visible was his fixation on contro!. Workers, for example, became
eligible for the $5 Dollar Day labor reform package only when they
submitted to and passed inspection about intimate details of their
personal lives. Members of the newly created Sociological Department
'~r, "'" ;''-' . .clj\.lI1_ aJ.• •
~,... 1
",,,,...1 .....:, 1oCl.1XlO _
.. 1_ ~) "U ..
..s.',.lIJ.ar.' ,,"I'l '.Lw.I.
cn
o
rn
Z
o
rrr
;l>
Z
tl
o-j
r:n
o
::r::
z
o
r-
o
o
><
Fig. 2. The Highland Park Plant opening on to Woodward Avenue . From the collections of Henry Ford Museum & Greenfield
Village (Neg. no . 0.3139) . Reproduced with permission.
IV
.....
-.l
218 lOHN M . STAUDENMAlER. S.1.
checked cleanliness, debt, drinking habits, etc. A failing grade meant that
the profit sharing bonus was put in escrow until the worker mended his
ways. Failure to comply eventually meant firing." Seeking to escape
dependence on outside suppliers, Ford used the Rouge river's access to the
Great Lakes and adjacent intersecting rail lines to create the logistical
heart of a mine-mouth-to-dealership empire. He purchased mines and
forests, a rubber plantation in Brazil, a rail line and a Great Lakes
shipping line so that by the mid 1920's Ford boats and trains carried Ford
iron ore, hardwood and other inputs to the largest industrial plant the
world had ever seen.
lronically, while the Rouge appeared to the world as the symbolic
capstone ofFord's world-class technological triumph, the design concepts
on which it was based reveal unmistakable signs of Ford's retreat into a
solipsistic world that excluded of those whose independent thinking
threatened hirn. Thus, in 1919, after major stockholders sued (and won)
because of Ford's practice of diverting potential dividends back into
company expansion, Ford conducted an elaborate and deceptive strategy
for buying them out. Almost simultaneously, three of "his ablest
lieutenants" (C. Harold Wills, John R. Lee, Norval Hawkins) resigned
under pressure. 30
During the twenties, Ford's reclusive tendencies deepened. Fairlane ,
the Ford mansion completed in 1916, stood on the banks of the Rouge
river in Dearborn miles from most of Detroit's elite society who lived
across town along Grosse Po inte's mansion row. After his national
humiliation during the 1919Chicago Tribune trial , he withdrew more and
more into the company which he now totally controlled and, as the decade
continued, to his personal playground, the Henry Ford Museum and
Greenfield Village. For some years after it became operational, the
Museum and the Village were not open to the public . The complex served
as a small trade-based boarding school. Private groups and individuals
were sometimes permitted to visit the collections but no policy of public
admission was implemented until June 22, 1933.3 1
One ofthe Rouge's design departures from its predecessor at Highland
Park would later become a world famous symbol of Ford's rejection of
debate and dissent. Just as Ford pursued integrated control of inputs
through his network of transportation lines converging on the Rouge, so
he sought even more control over workers than the house-to-house
inspections and in-factory spy networks of the 1914 labor reforms
provided. The earlier Highland Park plant opened directly onto
SCIENCE AND TECHNOLOGY 219
Fig. 3. The Battle of the Overpass . From the collections of Henr y Ford Museum
& Greenfield Village (Neg. no. 0.4951). Reproduced with permission .
Fig. 4. The Dynamo Showroom from the Highland Park Plant. From the collections of
Henry Ford Museum & Greenfield Village (Neg. no. 833.72). Reproduced with permission.
the Fair's Hall of Science were met in the foyer by the "Fountain of
Science" with Louise Lentz Woodruff's three-piece sculpture, "Science
Advancing Mankind" . Two life-sized figures, male and female, faced
forward with arms uplifted. Both were dwarfed by the massive figure of
a metallic robot twice their size. In the words of Lenox Lohr, general
manager of the exposition, the robot typified "the exactitude, force and
onward movement of science, with its hands at the backs of the figures of
a man and a woman, urging them on to the fuller Iife." The sculpture's
iconographic ideology was reinforced by the official Guidebook's
stunning, bold-faced thematic motto: SCIENCE FINDS, INDUSTRY APPLIES,
MAN CONFORMS. 34
The Chicago Fair's technocratic ideal is rooted in a violent disjunction
between the combined force of "Science" and "Industry" (itself a
conflation of business and technology) on the one side and "Man" on the
other. The role of "Man" in the modernist equation is not to "Critique"
fallible decisions made by scientists, business managers or engineers; it is
222 lOHN M . STAUDENMAlER, S .1.
All these stories abound with instances of the same paradox: Ford's
Model T provides geographical mobility for an entire generation of
people with modest means, giving them access to sophisticated
transportation technology, both cheap and easy to repair. Meanwhile the
much more sophisticated system that produces them encloses Ford
workers in prison factories hedged about with enforcers and impermeable
fences. Bentham's Panopticon would free the criminal from torture and
dungeon but offers re-instatement into society at the terrible cost of
programmed conformity to a mythic, omnipresent, and invisible
scrutinizer. Twentieth century citizens break free from the suffocation of
minuscule village perspectives through the mediation of global and
instantaneous electronic media. At the same time their capacity for active
participation in public discourse and the political order erodes.
It is a telling irony, one that reveals a great deal of the strengths and
liabilities of Western-style modernity, that both scientific and
technological practice themselves often mirror Bentham's tower and
Ford 's policed perimeter. Panopticonism and Fordism work to interdict
negotiation about system design and operation between scrutinizer and
prisoner, management and labor, the governing elite and the governed, in
short, they shield experts from the intervention of non-elites. In precisely
the same way, the scientific method's controlled variable environment
with its canon of replicability excludes non-experts from the workings of
the method. Science has its own aesthetic, a sanitized place cleansed of the
untrustworthy presences of bias, vested interest, emotion or tradition.
The controlled variable experiment, therefore, constitutes a revolutionary
new form of cultural space: an inaccessible interiority cuts off the working
of the experts' method from the turbulence of outside society.
This protected environment has, as we have seen, developed into a
multi-form and pervasive cultural force. Its most important influence
may weil be the radically increased social distance between decision
centers, where options are debated, taken, and eventually legitimated,
and the periphery, where non-decision-makers endure the results of the
SCIENCE AND TECHNOLOGY 223
The Bad News. Disempowering the local and indigenous, however, can
be powerfully seductive for elite design constituencies. Note that the
patterns of culture so briefly sketched above reveal a cultural context
wherein small elite power centers pair off against large arrays of
marginalized conformers. Whether in the Bentham panopticon, the
increasingly class-divided theatres, the twentieth century electronic media
audiences, or a Ford factory, these systems of cognition and discourse all
insure what Bentham intuitively insisted on: that those on the margin not
be able to see or actively influence those at the command central tower.
This tends to play itself out in Science with contempt for local
traditions of wisdom, whether pre-Industrial or non-Western, with the
conviction that non-experts have nothing to teach the design elite in
question . Managers of technological systems, for their part, frequently
invest in designs aimed at preventing workers and "the public" from
having access to managerial decision making processes. In other words,
elitist control sometimes masquerades as a technical requirement when it
merely enforces a hankering for working space that has been cleared of
outsider critique."
So we come to the question; do science and technology as ordinarily
practiced constitute a liberating force or an imprisoning one? Insofar as
scientific and technical elites find ways to render their value-Iaden
judgments about priorities accessible to outside critique and pluralistic
debate, their work will flourish as an essentially beneficent influence
within society at large. lnsofar, on the other hand, as such elites succumb
to the seduction of the Bentham Tower and the Ford Fence, they
ironically replicate the very evils they claim to overcome. Inside the tower
and behind the fence experts easily succumb to the fleshy temptations
inherent in any enclosed local environment - old boy networks and the
blind biases of unchallenged assumptions. The structure of the
protecting claim to dispassionate and value-free objectivity and the
rejection of human pluralism implied therein can easily mask the
question of power. 39
To summarize in necessarily over simple terms: the policing of the
method, once the value 01 its project has been agreed on, needs to be the
SCIENCE AND TECHNOLOGY 225
establishes the value of one design over another must include non-elite
outsiders.
This is very hard work and I do not intend to minimize its difficulty.
Indeed, the primary purpose of this paper is to call attention to the
cultural patterns which render it so difficult for Westerners to find the
patience and the nerve to engage in tough debates about scientific and
technical priorities, about the allocation of the increasingly scarce
resources available to our societies, and about the determination of the
questions: "Who wins, who loses, and how much does it matter?" The
cultural trends toward a Benthamized public conformity-where
individuals shrink from the responsibilities of adult citizenship in the
public arena-run deep . They are embedded in our affectivity and our self-
understanding and they shape our expectations about how much can be
hoped for from the public order. Citizens of Western societies must
reopen Bentham's tower and break through Ford's fences so that we can
risk a rebirth of the turbulent and unpredictable forms of genuine civil
discourse as the millennium in which Western culture achieved global
power draws to a close. Lacking the courage to do so would very likely
insure that the next millennium will not treat the West so kindly as the last.
NOTES
I For the full argument see my Technology's Story tellers: Reweaving the Human Fabric
1969). Landes eharaeatures all eognition that does not take a rneans-to-ends form as
"superstition and magic", p. 21. See the following pages for further examples of Landes '
disjunetion between rat ionalit y and all other modes of eonsciousness whieh are defined as
defeetive by Landes ' dismissive, and oeeasionally derisive, tone .
The most artieulate eritique of this definition of rationality that I have yet eneountered is by
226 lOHN M. STAUDEN MAlER, S.l.
Renato Rosaldo , Cu/ture and Truth: The Remaking 01 Socia/ Ana/ysis (Boston: Beacon
Press, 1989). Thus, for example, "The point is to break object ivism's monopoly on truth
claims, not to throw out the bab y with the bath water. ... When the workings of culture are
reduced to those of a control mechanism, such phenomena as passions, spontaneous fun,
and improvised activities tend to drop out of sight."(p. 102) or "In my view, optionality,
variability, and unpredictability produce positive qualities of social being rather than
negative zones of analytically ernpty randornness ." (p. 112)but see also the full extent of his
argument throughout Part One.
• See, for some recent discussions of contextualism, Robert C. Post and Steven H. Cutcliffe
(eds.), In Context: History and the History 01 Technology, Essays in Honor 01 Melvin
Kranzberg (Bethlehem, PA : Lehigh University Press , 1989), pp , 150-171.
s Scholars representing a broad range of interpretative perspectives have come to agree on
the importance of contextual factors for interpreting science. See, for example, Arnold
Thackray, 'History of Science', in Paul Durbin (ed.), A Guide to the Cu/ture 01 Science,
Technology, Medicine (New York: Free Press, 1979); Evelyn Fox Keller, Reflections on
Gender and Science (New Haven , CT: Yale University Press, 1985); Bruno Latour, Science
in Action: How to Follow Scientists and Engineers through Society (Milton Keynes, Bucks.:
Open University Press, Cambridge MA: Harvard University Press, 1987); and Steven
Toulmin, Cosmopolis (New York: MacMillan, 1990).
• The concept of pro fessionalism , that ernpowers the licensed practitioner while
delegitimizing outsiders is a central achievement in 19th century Western society, See, for
example, Burton l. Bledstein, The Cu/ture 01 Professionalism: The Midd/e C/ass and the
Development 01 Higher Education in America, esp. Ch. 3.
7 My thinking here origina tes with the stud y of technological style in the United States. Thus
(I) Cells " divided from one another ... secluded from all communication with each other,
by partitions";
(2) " Each cell has in the outward circumference a window, large enough, not only to light
the cell, but , through the cell, to afford light enough to the correspondent part of the
[inspector 'sl lodge. The inner circumference of the cell is formed by an iron grating, so
light as not to screen any part of the cell from the inspector's view" ;
(3) "To prevent thorough light, whereby ... the prisoners would see from the cells whether
or no any person was in the lodge, that apartment is divided into quarters, by
partitions" ;
(4) "Small lamps, in the outside of each window of the lodge, backed by a reflector, to
throw the light into the corresponding cells, would extend to the night the security of the
day." 'Panopticon' p. 40
It Works of leremy Bentham, Vol. 10, p. 226. Cited in Carolyn C. Cooper, 'The
Portsmouth System of Manufacture', Technology and Culture 25(2) (April, 1984), p. 193.
SCIENCE AND TECHNOLOGY 227
Retributive Idea ', in Jeffrie G. Murphy and Jean Harnpton, Forgiveness and Mercy (New
York : Cambridge University Press, 1988), p. 158.
15 Ibid., p. 19. Emphases mine
" 'Therapy, Not Punishment' , Harpers Magazine (August 1959), pp . 63-64 (emphases
mine). Menninger is quoted, together with Bertrand RusselI, B. F. Skinner, and Benjamin
Karpman to mueh the same effeet , in Herbert Morris, ' Persons and Punishment', The
Monist 52 (October, 1968), pp. 480-481. Morris ' essay has aequired the status of a classic
eritique of the therapeutic replacement of punishment. He roots his argument in the inherent
dehumanization of the person when guilt is replaced by treatment. " In th is [therapeutic]
world we are now to imagin e when an individual harms another his conduct is to be regarded
as a symptom of some pathological condition in the way a running nose is a symptom of a
cold." p. 480.
•, Harley Shaiken, Automation and Work in the Computer Age (New York : Holt, Reinhart
and Winston, 1983). Shoshanah Zuboff, In the Age 0/ the Smart Machine, passim .
•• See George F. Madaus, 'Curriculum Evaluation and Assessment', in P. Jackson (ed.),
Handbook 0/ Research on Curriculum (New York : Macmillan, forthcoming) . See also
Zuboff (especially Chs . 6-7 , 9-10) for discussion of the use of computerized algorithms in
decision-rnaking roles as precisely the same process of distancing the evaluated from the
point at which the evaluative judgment is rendered.
•• Erving Goffman, Asylums: Essays on the Social Situation 0/ Mental Patients and Other
Inmates (Garden City, NY, 1961). John F. Kasson, Rudeness and Civility: Manners in
Nineteenth Century America (New York: Hili and Wang, 1990). See Ch . I for his estimate
that an average of three new etiquette books annually before the Civil War rose to an average
of five or six per year from 1870 through World War I. See Chs . 4 and 5 on body control
advice .
20 See Lawrenee W. Levine, Highbrow Lowbrow: The Emergence 0/ Cultural Hierarchy in
Levine notes that the gradual disciplining of audience interaction was accompanied by the
creation of separate establishments for upper dass and lower dass patrons.
22 On the early history of the wire news services see Daniel J . Czitrorn, Media and the
American Mind: From Morse to McLuhan (Chapel Hili , 1982), especially Ch . I; Richard
Sehwarzlose, 'Harbor News Association: The Formal Origins of the AP ', Journalism
Quarterly 45 (Summer, 1968), pp. 253-260; Robert Luther Thompson, Wiring a Continent;
The History 0/ the Telegraph Industry in the United States, 1832-1866 (Princeton:
Princeton University Press, 1947) and Richard B. DuBoff, 'Business Demand and the
Development of the Telegraph in the United States, 1844- I860', Business History Review 54
(Winter, 1980), pp . 459-479 and 1983: 'The Telegraph and the Structure of Markets in the
United States, 1845-1890', Research in Economic History 8, pp . 253-277.
23 The Molly trial was probably the first "national media" event in U.S. labor history.
Joseph Rayback describes the immediate and long-terrn effeets of the tr ial as folIows . "The
evidenee against them, supplied by James McParlan, a Pinkerton [detective] , and
228 JOHN M . STAUDENMAlER, S .1.
corroborated by men who were granted immunity for their own crimes, was tortuous and
contradictory, but the net effect was damning. All twenty-four were convicted; ten were
executed. The trial temporarily destroyed the last vestiges of labor unionism in the anthracite
area. More important, it gave the public the impression that miners in general were inclined
to riot, sabotage, arson, pillage, assault, robbery, and murder; and that miners were by
nature criminal in character and were to be condemned and disciplined by the more
respectable element in society. The impression became the foundation for the antilabor
attitude held by a large portion of the nation to the present day." A History of American
Labor (New York: MacMillan, Free Press, 1966) p. 133.
24 In his recent critical assessment of Richard Rorty, Christopher Norris uses U.S . media
coverage of the Irac war to exemplify some of the philosophical and pragmatic problems
that reside in electronically mediated public discourse. ('The "End of Ideology" Revisited :
The Gulf War, Postmodernism and Realpolitic', Philosophy and Social Criticism 17(1)
(November), pp . 1-40.
" Roland Marchand, Advertising the American Dream: Making Wayfor Modernity, 1920-
1940 (Berkeley: Un iversity of California Press, 1985), pp. 68-69 and passim . See also Daniel
Pope, The Mak ing oJ Modern Advertising (New York : Basic Books, 1983); Leiss, Kline and
Jhally, Social Communication in Advertising, and T .J . Jackson Lears, ' From Salvation to
Self-Realization: Advertising and the Therapeutic Roots of the Consumer Culture, 1880-
1930', in Richard Wightman Fox and T .J . Jackson Lears (eds .), The Culture of
Consumption: Critical Essays in American History, 1880-1980 (New York: Pantheon
Books, 1983), pp. 1-38. I am also deeply indebted to many conversations with Pamela
Walker Lurito for my understanding of changing advertising trends .
2. Debates ab out the effectiveness of advertisements in programming consumer motivation
are commonplace in recent studies. Michael Schudson, Advertising, the Uneasy Persuasion:
Its Dubious Impact on American Society (New York: Basic Books, 1984) argues the case
against it. Leiss, Kline , and Jhally Social Communication in Advertising, Chs, 2, 3, discuss
the pros and cons of both sides , citing Schudson, Stuart Ewen, Christopher Laseh, and
others; their own position tends to favor Schudson's .
21 (Robert O . Derrick, architect hired 10 design the Henry Ford Museum, Oral
Reminiscences) p . 50 [emphases mine) . Geoffrey C. Upward, A Home for Our Heritage:
The Building and Growth of Greenfield Viilage and Henry Ford Museum, 1929-1979
(Dearborn: The Henry Ford Museum Press, 1979), p. 50.
2. Otto Moog , German Engineer, in Thomas P . Hughes, American Genesis, p. 291,
author's translation of Otto Moog, Drüben steht Amerika: Gedanken nach einer
Ingenieurreise durch die Vereinigten Staten (Braunschweig: G. Westermann, 1927), p . 72.
Hughes cites another German engineer, Franz Westermann, saying "the most powerful and
memorable experience of my Iife came from the visit to the Ford plants .. . " (p . 292).
2. See Stephen Meyer , The Five Dollar Day (Albany: State University of New York Press,
1981), especially Chs . 5-8.
JO On the stockholder buyout see Alan Nevins and Frank Ernest Hili, Ford: Expansion and
irritated by the presence of anyone in the company who might not work with hirn in complete
harrnony ." (ibid., p, 145).
31 See Geoffrey C. Upward (museum editor) , A Home for Our Heritage: The Building and
Growth of Greenfield Village and Henry Ford Museum , /929-/979 (Dearborn, MI: The
Henry Ford Museum Press , 1979) p. 76.
In 1919, Ford sued the Chicago Tribune for libel, was grilled on the stand with lines of
Questions demonstrating his flimsy educational background, and was awarded six cents in
damages. Robert Lacey, Ford: The Men and the Machine (Boston : Little, Brown and Co.,
1986), pp. 197-202.
J2 Greenfield Village was officially opened on October 21, 1929 when the aging Edison,
flanked by Ford and President Hoover re-enacted the invention by turning on a replica light
bulb . The " Festival of Light" was transmitted over a national radio network, one of the first
"live " media events in history. The only published account that treats the event in any detail
is Upward, Home for Our Heritage, Ch. 3.
33 Lacey, for example: " .. . there was only one beautiful room in the entire building: the
powerhouse. This was a spare , clean chamber which Henry had designed hirnself . .. . and he
created a very Ritz of power stat ions, all marble and gleaming brass dials and pipes. Around
the floor were set out little generators, raised on plinths like so many modern sculptur es ..."
pp. 149-50. See also, Collier and Horowitz, p. 71. See also Nevins, pp . 20-21.
The clean, uncluttered, "Ford" style that Charles Sheeler would make famous with his late
twenties photographs and paintings may represent the continuation of, and not a completely
fresh artistic reflection on, the Ford style. See Mary Jane Jacob, 'The Rouge in 1927:
Photographs and Paintings by Charle s Sheeler' , in The Rouge: The Image ofIndustry in the
Art of Charles Sheeler and Diego Rivera (funded by the Ford Motor Company Fund and
Founders ' Society Detroit Institute of Art s) and , more recently, Karen Lucic, Charles
Sheeler and the Cult of the Machine (Carnbridge, Harvard University Press , 1991).
3 4 Chicago Century of Progre ss International Exposition , Official Book of the Fair,
(Chicago: A Century of Progress , Inc., 1932), p. 11. I am indebted to Lowell Tozer , 'A
Century of Progress , 1833-1933: Technology's Triumph Over Man' , American Quarterly
4(l)(Spring, 1952), pp . 78-81 , for first calling my attention to the Exposition and to Cynthia
Read-Miller, curator of photographs and prints in the archives of the Henry Ford Museum
and Greenfield Village, for copies of the Official Book and photos of the iconography
referred to here.
For the Lohr quote see, Fair Management: The Story of a Century of Progress Expos ition
(Chicago: The Cuneo Press , Inc., 1952), p. 96.
3. Popular feelings about technocratic elitism were c1early mixed. Industrial unions
flourished in the Thirties as workers organ ized to contest managerial high-handedness . On
the other hand, even so shocking an episode as 1937's battle of the overpass evoked an
outpouring of fervent support for Ford's dictatoriallabor style in hundreds of handwritten
letters from ordinary citizens around the country . (Archives and Library Department,
Henry Ford Museum & Greenfield Village, Dearborn MI, Ace, 292 Box 43.) My cursory
inspection suggests that the vast majority of the letters, although not all, strongly favor
Ford's position .
3. Recently several sociologists of technology have developed the concept of "negotiation
spaces" to address much the same point. Thus : " ... we have shown how the proponents of
the project mobilized the actors in agiobai network and sought to create a relatively
230 lOHN M . STAUDENMAlER, S .1.
autonomous negotiation space where a local sociotechnical network might be designed and
brought into being without constant interference from outside." lohn Law and Michael
Callon, 'Engineering and Sociology in a Military Aircraft Project: A Network Analysis of
Technological Change', Social Problems 35(3) (June , 1988), p. 290.
37 Steven Toulmin argues that it is precisely this systematic rejection of the cognitive validity
of the local, the oral, the specific, and the timely in favor of the universal, the written, and
the timeless, that characterizes what came to be the central orthodoxy of "modernity". For
his argument that aversion to the specific and local sterns from Europe's loss of nerve (and
sense of humour) in the face of the Thirty Year's War with its carnage of competing religious
orthodoxies, see Cosmopolis, Ch . land 2. The passage about the local, oral, specific, etc,
begins on p. 32.
David Harvey's interpretation of changing capitalist social definitions of space and time and
their influence on contemporary society is the most helpful and sophisticated that I have
read . See his The Condition 0/ Postmodernity (Cambridge, MA: Basil Blackwell, 1989),
especially Part III.
3. Shoshanah Zuboff records repeated instances where managers react "irrationally" to the
democratizing influences of open-access computer data bases despite the evidence that such
open access, a process for which she coined the name "inforrnating", increased efficiency
and profitability, managers experienced disorientation and fear when their control over
subordinates was threatened. See, In the Age 0/ the Smart Machine, passim
3. Peter Sand man has developed a persuasive model explaining the increasing social cost
that comes due when non-elites are excluded from such prioritizing debates. See his 'Hazard
versus Outrage in the Public Perception of Risk', in Vincent T. Covello, David B.
McCallum, and Maria T . Pavlova (eds.), Effective Risk Communication (New York:
Plenum, 1989), pp. 45-49, on the importance of inclusion of non-elites within such
prioritizing debates. See also Parker Palmer, The Company 0/ Strangers: Christians and the
Renewal 0/ America's Public Life (New York : Crossroads, 1985), for a cornplementary
societal analysis .
ARIE RIP
I .INTRODUCTION
The scientists - for the benefit of the entire nat ion and almost always at its request and
expense - are seeking 10 multiply the knowledge which can serve the increase of industry,
234 ARIE RIP
wealth, and the beauty of life, the improvement of the political organization and the moral
development of the individual. Yet, not immed iate utility must be looked for, as is so often
done by the uninformed. Everything that informs us ab out the natural forces or the forces
of the human spirit is valuable and in time may pro ve useful , normally in a place where one
had least expected this .!'
these were very much "hunting licenses": " .. . everyone went through
much the same traumatic adolescence in developing product application
information." The main problem was that the new polymers fulfilled
their promise too weIl:
As soon as enough linear polyethylene became available to try to put it to work, it was found
that the "Chain Straighteners" had done their work too weil, for in polyethylene as in
society, absolute straightness turns out to be more virtue than is wanted for practical
purposes. The advantages of hardness, stiffness, strength and heat resistance in finished
products were gained at the cost of seriou s difficulties in the manufacture and use of those
products. Since Phillips polyethylene had the "purest", 100070 linear cha in structure, it
suffered these difficulties in the highest degree . {... )'6
In every one of the major developed markets for which linear poly-
ethylene seemed suited, the initial bright promise was tarnished by one or
other unanticipated "different" characteristic that proved to be a
significant disadvantage. For example, the established markets for
packaging film were not open to linear polyethylene because its high
crystallinity rendered it opaque and prone to splitting .
All these nightmarish problems came to light just as the first polyethylene plants were
coming on-stream, with all the trauma associated with plant start-ups, One after another,
each of the "rniracle progeny of plastics " began to look more Iike a retarded child . The
company managements that had fathered them were forced to contemplate infanticide and
were, in turn , threatened by their boards of directors with financial sterilisation. Heads were
greying, if not rolling, in more than one organisation, and PhilIips, for one, were very close
to closing their plant and aborting the whole enterprise when salvation showed up in the
form of a toy .
It would be gratifying to a chemist to be able to report that the research teams had leaped
to the rescue just in the nick of time with technical solut ions to all the technical problems.
As a matter of fact, they did eventually solve nearly all of them , but not quickly enough to
save the situation had it not been for the fortuitous burgeoning of an unprecedented,
unplanned, unforeseen market. The Wham-O Toy Company introduced the " Hula Hoop" ,
made from extruded polyethylene tubing , and the ensuing craze created a non-critical
demand that swept the glutted warehouses clean and put the plant s back on an around the
clock production schedule.
Th is fad, even though it faded rapidly, bought time for the polymer chemists to
determine that most of their woes arose from the super-crystallinity of their super-linear
polymer chains, and that by judicious adulteration with touches of a second monomer
(introduced in the polyrnerisation step) , they could put a very occasional kink in the
molecular chains and thus alloy the crystallinity of the polymer just enough to suppress most
of its deficiencies while preserv ing most of its sterling qualities. {... )17
236 ARIE RIP
~~ ~
V
• EXPLOITATION
(Semio)emPiricalQ
~~
- .....
findin~ '::- ': \. rauonalization
rec ogmzed as such ' \ "- .;:.>
~--
~~
\
~
.
~CIENTIFIC EXPLORATION
The history of the linear polyethylene case further illustrates this model; 19
another example is the rapid exploitation of Röntgen radiation after its
discovery in the laboratory. While the original starting point may weIl
derive from scientific work, as these two examples indicate, it clearly is
not a matter of applying a scientific finding. There appear to be three
types of connections between science and technology: a laboratory effect
SCIENCE AND TECHNOLOGY AS DANCING PARTNERS 237
and barely articulated, but with the increasing need to justify and
legitimate actions, e.g. of an R & D department with respect to its board
of directors, or technologists asking for government support, there is ad
hoc or even systematic articulation (including special assessments of
trends and potential of scientific and technological domains, produced by
consultants or expert committees).
So there are two processes at work: one is the articulation of
expectations and incorporation of new information and assessments; the
other is sedimentation, the emergence of a shared and stabilized repertoire
of expectations , on which actors can draw as aresource. As soon as that
happens, one can speak of a matrix of expectations, because decisions,
choices and further assessments will be embedded in it. Such matrices of
expectations are never static; expectations evolve, and there is negotiation
about their worth (their reliability and appropriateness). So any
description of a matrix of expectations is a snapshot, taken of an evolving
process. Still, it is important to intro duce the concept as standing for a
relatively stable phenomenon, because it allows one to understand, first,
how heuristics can spread and how commitments to them emerge, and,
second, how others than search actors can be involved: even if not actually
participating in problem solving, managers and third parties will base
decisions drawing on the cultural matrix of expectations.
Because such matrices of expectations play an intermediary role, it
becomes understandable in principle how the whole development is
structured through rules and institutions formed around expectations,
even before there is any actual external selection .
I shall use the example of synthetic dye chemistry in the 19th century
to briefly illustrate the dynamics, and prepare the ground for a discussion,
in the next sub-section, of socialloci of search processes. 18 The success of
aniline red (discovered/invented in 1859, after a variety of attempts
comparable to Perkin's "extraction" of aniline purple from aniline)
attracted other producers into the synthetic dye field, and the correspond-
ing, and rat her primitive heuristic was "tinker with aniline", a heuristic
that indeed produced a number of interesting dyestuffs. The aniline red
paradigm was rationalized by Hofmann on the basis of the chemical
theory of types, and although his type formula did not represent a
definitive solution as to the chemical constitution of the "mother
compound" rosaniline, it was quite effective in stimulating and directing
innovative activity. 19 The attempts at elucidation of the chemical
constitution, and the perceived success of such attempts in guiding search
240 ARIE RIP
for further dyestuffs, can be seen as the first step toward a new, more
general heuristic : "Look for the Muttersubstanz (that is, the skeleton of
the dyestuff's molecule), and if it is found, each derivative of the
Muttersubstanz may be an innovation" . This heuristic became the central
component of a new paradigm of purposeful synthesis, which could
already be discerned in rudimentary form in the elucidation of the
chemical constitution of alizarin and its subsequent synthesis in 1869.3 0
C1early the two-branched model of development of a finding is
applicab1e, but in addition, search heuristics of more general va1ue
emerge. In these examples, the heuristics cluster around an exemplar: a
product or process that is an exemplary achievement. Thomas Kuhn
(1970) redefined his "paradigm" as consisting of such an exemplar in
combination with a " disciplinary matrix" . In the same vein, we can now
speak of a technological paradigm as an exemplar (in this case, one
achieving technical and commercial effects), embedded in a cultural
matrix of expectations. Further work in the frame of such a paradigm
leads to a trajectory, a sequence of innovations that are related through
their use of a cluster of heuristics (which may weIl be modified along the
trajectory);" Note that the analogy with science remains : "normal
science" in the framework of a paradigm, or, alternatively, the pro-
gressive elaboration of a Lakatosian research program, are equiva1ent to
the technological trajectory. In fact, the chemistry in that period can be
analyzed with the same categories of search processes, heurist ics and
partial stabilization in paradigms.
In another example from synthetic dye chemistry, the emergence of
azo-dyestuffs, the dynamics of emergence of a technological paradigm
can be seen in detail. There was an exemplary process (a coupling
reaction, named after Griess, which produced coloured compounds that
were interesting to scientists); there was an exemplary product (Roussin's
naphthalene dyes, with very useful properties, found by trial and error in
an industrial workshop), both with their own search histories and kept
more or less secret. The actual matrix of expectations emerged in a
dramatic way, with Hofmann disclosing the secret of preparation of both
classes of compounds in 1877, and emphasizing that a vast domain for
synthetic dye chemistry was now opened, of which the limits were not yet
in sight.
The conditions outlined here for the emergence of a paradigm are
important for ongoing technologica1 development in general, even when
these do not fit the model of a paradigm plus trajectory."
SCIENCE AND TECHNOLOGY AS DANCING PARTNERS 241
animal husbandry), but they are necessary to maintain visibility for wider
publics. The whole business is precarious, as the decrease in stock market
value of the same firms shows, as weIl as the difficulties into which a
number of them have run .
More generaIly, one could say that in a world where innovation
competition becomes more important, early promises about innovations
will be valued as such, that is, as building blocks of the necessary matrix
of expectations . A speculative market of early promises emerges; a
process fuIly analogous to the emergence of stock exchanges, where
shares in firms (and before that, in merchant ventures) could be traded in
terms of the expected profits. The R & D firms that carry this market
started out as a specific nexus, but when this type of firm became
institutionalized, the market became relatively autonomous, and
academic research groups and other actors could also start offering their
early promises for "sale", i.e, to be taken up in exchange for resources.
Strategie mobilization of science and technology can then become a
regular phenomenon, rather than ad hoc and based on bilateral
negotiation.
Three points can be made to conclude this outline of a quasi-
evolutionary model of search processes in science and technology.
Scientific and technical communities used to be the locus for appraisal of
expectations and developments of heuristics as such, and while being
shaped by institutionallocations of the practitioners (as will be discussed
in the next section), could do so relatively autonomously. Many more
actors are now involved explicitly. Analytically, the point is that the
activities continue to be at the level of fields of science and technology
(including patterns of decontextualization and recontextualization, as I
noted in Section 2).42
A second point is that the model connects two aspects that are often
kept separate: the action and process level, where actual production of
technology, adoption and diffusion occur, and the level of cognitions and
legitimations, where heuristics, expectations, and rules are to be found . In
other words, to "producing a working artefact" it adds two other
important activities: "making a clever move" (this is getting more
attention in recent technology studies)? and "telling a good story". The
telling of stories, so that they have effects, is an important aspect of socio-
cognitive-dynamics, but has been neglected in general (even if there is
some interest from organization studies), but certainly in technology
studies."
244 ARIE RIP
large scale, and whole new areas of research were created in which
fundamental work occurred, but within an overall framework defined as
component or deviee search. Polymer science, radar and other electronies
areas, and atomie energy research, are well-known examples. New
speciaIties could emerge, and the training of such specialists was seen as a
responsibility of the universities (as the pressure - but only by the I960s
- for chairs in polymer science and heterogeneous catalysis shows).
By the 1980s, three strands have to be woven into the fabrie of the
cosmopolitan design network . First, the meta-modeling role of the
technieal sciences in academic settings remains important, and links up
with the mathematieal modeling approaches that have become more
sophistieated. Second, the design supply role has become more
important, and has extended to more sciences. One element of the recent
programs to promote strategie or innovation-oriented science is exactly
the mobilization of academie sciences for design supply, with the
programs on new materials (polymers, membranes, ceramies) as key
examples. Third, innovation competition between firms and strategie
positioning of states and blocks of states (e.g. in the TRIAD) has added
a new element to technieal innovation: not the actual outcome of the
innovation is the primary aim, but the coverage of a potentially important
area of innovation. One indieator is that firms are prepared to spend (a lot
of) money on the support of academie research in order to have a
"window on science" . The coupling with local design practiees is
becoming remote, also in firms themselves. Or, aIternatively, local design
work is shaped by the exigencies of the innovation race, as can be seen
cIearly in the unending quest for the "next generation" of chips, the very
large scale integrated circuits now containing more than a million
components. Expectations always shape technologieal development
(Section 3); there is now a quasi-autonomous dynamie of strategie
research, in whieh universities can take part (fully and often only very
partially, given the complexity and resource requirements of strategie
research in many areas).
If one thing has become cIear from the analysis in the preceding sections,
it is that there is a variety of relations between sciences and technologies,
252 ARIE RIP
that these relations and the patterns in them evolve, and that the evolution
is shaped by the socialloci of the search processes and the global scientific
and technologieal fields, with an important mediating role of the division
of labour in global design networks. By now , science and technology are
intertwined and form a complex - dancing partners that cannot be pried
apart.
The science-technology complex has been built up over decades, and its
shape shows the traces of its origins. I have focused on meso-level
developments , talking about institutions, technical domains and scientific
disciplines. These are, of course, embedded in overall societal
developments, and one should, ideally, write the history also at that level.
One attempt to do so is Tom Hughes's American Genesis, which
analyzes how the second industrial revolution is characterized by the way
the Americans built a technologieal world. This has created a
technologieal momentum: the systems, and their preferred character-
isties, take on an autonomous character and are difficult to direct or
control other than along an inertial projection. Since these mature
systems experience most of their social shaping in their early stages, they
bring out of their past, the solutions to past problems. Thus, the
momentum of the modern may be so great in the United States that the
next great technological and cultural change may occur among other
peoples in another nation."
The diagnosis of what happened is impressive, but there is a curious
neglect of the changes in science and technology that may be happening
now, in the Uni ted States as weIl as elsewhere. What is the new dance that
the dancing partners are developing? There are good reasons to think that
qualitative transformations are occurring.
One way to analyze the transformations is to conceive of an emerging
new techno-economie paradigm, as Freeman, Perez and others do. They
relate the transformation to the combination of micro-electronics and
telecommunieations becoming a so-called pervasive technology, having
repercussions in all sectors of society. 58 In doing so, their interest (as
economists) is primarily in the performances actually realized, not in the
search processes and strategie behavior that goes into the creation of these
performances. 59 Without pronouncing on the value of their conceptions,
I limit myself here to what is directly relevant to my theme of relations
between science and technology: the micro- and meso-levels of search
processes, the practices in which these are embedded, and the structures
that are emerging.
SCIENCE AND TECHNOLOGY AS DANCING PARTNERS 253
6 . CONCLUDING REMARKS
How far have I come in analyzing the dance of science and technology? In
two ways, I tried to overcome barriers to our understanding: I did not
start with science, its nature and its possibilities with regard to technology,
but as it were turned the tables, by looking at search processes in general,
and at technology being rationalized, becoming theoretieal, " scientific"
even, and thus offering opportunities for existing science to get involved.
Secondly, I emphasized the danger of following legitimatory usage of
science and technology as umbrella terms," and focused on ongoing
processes and institutionalized activities instead. This is not to say that
there is no such thing as "science" , or "technology", or that it makes no
sense to talk about the world of science and the world of technology, and
how these interact. But the notion of science as such is constructed (and
maintained) by actors in specific contexts, for example when in the 19th
century calls are made for engineering to become scientifie, or when
science is defended as being the backbone of Western civilization. The
force of science does not exist in a vacuum , but derives from these
contexts and what actors do with such constructions. SimiIarly, there are
times and piaces where worlds of science are created, and become
sufficiently institutionalized to continue as such. Engineers (and other
people not usually seen as scientists or scholars in the 19th century) can
then be said to follow hybrid careers when they also playa role in the
world of science." For the late 20th century, however, it may be
misleading to analyze in terms of a world of science, that is a direct
successor to the world of science as it existed, at least for physicists in the
I920s. 72 That is why I have emphasized that transformations occur: while
there is some continuity of actors, institutions and practiees, the overall
pattern shifts, and it now means something different to be in the world of
science.
In this approach, there is the unavoidable ambiguity of using common
sense notions of scientists and engineers, of science and technology and
their demarcations, to mobilize empirieal material and interpret it, and on
SCIENCE AND TECHNOLOGY AS DANCING PARTNERS 257
the other hand not wanting to be imprisoned by such categories which will
hide rather than disclose. I have taken a sociological detour through
search processes, heuristics and expectations, local design work and
cosmopolitan design networks with their division of labour, to be able to
reconceptualize what is happening in the science-technology complex.
Central to this reconceptualization is the relation between local search ,
design and anticipation activities, and repertoires, cognitive-technical
infrastructures and interactions and institutions at the field level. In these
terms, one can give an analytic definition of what science is, if one wishes
to do so: when search products are "consumed", that is, assessed and
used, by others like yourself, and a cognitive and evaluative repertoire has
emerged together with sociallinkages and mobility implying community
formation, it makes sense to speak of (a) science. In contrast, technology
has search processes which are directly or indirectly related to local design
and construction activities . (Note that this definition makes large parts of
social science and policy analysis fall under technology.) The dancing
partnership of science and technology now becomes a relation between
activities oriented to different reference points and groups, rat her than a
matter of combining different cognitive-technical repertoires.
Obviously, in concrete cases people draw on different domains and
specialties, some of which have dynamics like science, and others like
technology. At this level, there may weIl be problems of interaction, but
these relate to the general problem of combining insights and skills from
different sources . Similarly, institutional problems, like interaction
between university and industry, or new roles for academic research, need
not derive from any general issue of "science" versus "technology", but
from heterogeneity of institutions. As I have argued, such heterogeneous
social loci will, in fact , feed into new definitions and demarcations of
science and technology.
In a sense, I define the question of the relation between science and
technology away . In so far as it is to be discussed, it is in a battle of
legitimat ions . To analyze and understand interactions, one had better not
use the umbrella terms.
Looking back at the alternative analyses and conceptualizations I have
discussed in this paper, however, these do not completely exhaust the
original question. One can do away with "science" and " technology" ,
because these are rhetorical constructs rat her than actual dancing
partners, but one still has to face the question, at the micro-Ievel, of the
relation between knowledge and artifacts .
258 ARIE RIP
Although science and technology both involve cogn itive processes, thei r end result s are not
the same. The final product of innovative scientific activity is most Iikely a written
statement, the scientific paper, announcing an experimental finding or a new theoretical
position. By contrast, the final product of innovative technological activity is typically an
addition to the made world: a stone hammer, a clock , an electric motor.?
University 0/ Twente
SCIENCE AND TECHNOLOGY AS DANCING PARTNERS 259
NOTES
1 In Price [l984a) he adds that what keeps them linked is that both dance to the music of
necessity of working on the situations with a number of exarnples , including one from a
study by Marc Bloch: " In the late Middle Ages, the grinding stones, the gears, the wheels
and the rivers are good unexpected allies that , once tied together in one mill, makes a
formidable stronghold. But their efficiency stops there. A stronghold can be in the middle
of a battlefield , thus bearing on the issue of the battle , or away from the battlefield . If each
household goes on grinding corn by hand , the Prince, who holds the communal mill, will
hold not hing but wood , water and stones. The mill will become a stronghold only if the
Prince fetches the militia, enforces the King's ruling, the Church 's teaching and compels
every household to break their hand-grinders and to pass through the miller's stone ." (p. 26)
The moral of this story , as Latour makes c1ear, is tha t social power is necessary to ensure
technical power (as well as vice versa), In his harangues against technological determinism
(labeled " trajectories" in the 1990paper) he gets carried away by his own rhetoric, however,
and forgets that there may be patterns in the sociotechnical alliances that might usefully be
described as trajectories .
• See Wolfgang Krohn's introduction to Zilsel (1976), esp. pp . 23-29 .
, Homburg (1983).
• Casimir (1983) . Latour and Woolgar (1979).
, Rip (1985).
• Böhme, Van den Daele and Krohn (1978) .
• This idea draws on Disco, Rip, Van der Meulen (1992) and ongoing work of these authors.
See further Section 4.
10 Nelson and Winter (1977) , Dosi (1982) , Dosi et al.. (1988).
" See for example the way even Dosi (1982) essentially works with a sequence model, where
findings go downstream from science to technology, passing filters (that include socio-
economic criteria) .
13 Ziman [1976) , pp . 180-181.
I.
" McMiIlan [1979), pp. 90-91.
For our argument here, the starting finding may be treated as a point source . When one
looks more c1osely, it dissolves into contingent processes, and only with the benefit of
retrospection can one identify a source. Compare Latour's (1987) analysis of the Diesel
engine.
" McMiIlan (1979).
'0 Prakke [1988].
" In a study on knowledge transfer (Rip, Schoemaker and Meus [1983]), it was argued that
the "matching' of university research and societal needs requires cognitive and social
transformations, both the aggregation of research questions through sectoral organizations,
260 ARIE RIP
patients ' associations, trade unions ; and disaggregation of academic approaches, and that
one can observe such processes in actual successful knowledge transfer. Note further that
laboratory-based, in the main text, should be read as including method-based (e.g. social
science methods like surveys). The point is that scientific findings derive their "universality"
from being produced in restricted environments (see Rip (1982)); this point is the same as
Harre's, when he argues that "manipulation of matter under the control of a skill is prior
to a more conventional scientific activity' (Harre [1990], p. 35). In the complexities of the
real world, applicability is not assured by "universality", and requires either de-
universalization or concretisation (Böhme, Van den Daele and Krohn (1973)) or
transforming the world into something resembling the laboratory (Rip [1982], Latour
(1983)).
22 This is also argued by Latour, but then from apower perspective, when he focusses on the
Machiavellian engineering that goes on in all walks of life (Latour (1986), (1987)). The
search perspective is not blind to power, but takes the cognitive aspect seriously. In the
search perspective, there is the same attention to the need to find and enroll allies, both
human and non-human: microbes and electrons have to be kept under control in order to
have robust findings . But there is also attention to exploration of possibilities, following a
lead, elaborating a cognitive schema . This is not sufficient to explain the "success", i.e. the
robu stness of findings. But it is necessary to understand the direction that is taken .
2J This view of scienceand technology has been put forward by a number of authors by now,
including de Solla Price and Layton, and has been dubbed the "two cultures model" by
Barnes and Edge [1982, pp . 147-154, esp. p, 151]. See also Weingart [1978] for a
differentiation-scientification thesis which is quite compatible with the present anal ysis.
24 The term "bricolage" , in the sense used here, has been introduced by Levy-Strauss, who
distinguished between "esprit d'Ingenieur" (that is, the French graduates of the Grandes
Ecoles with their mathematical bent), and the "esprit de bricolage" (quoted after Barnes
(1974), p. 58, 146). I adapt the term further to describe those aspects of science and
technology where the researcher scouts for whatever he can use, and is prepared to make do
with what he finds .
25 Nelson and Winter's [1977, 1982) argument focusses on "routines" and " routines to
develop new routines " (in their terminology), where routines stand for ways of doing things,
while rout ines to develop new routines are heuristics of search processes.
2. Expectations about new scientific findings or technological possibilities are voiced, and
acted upon , continually. The early promise of biotechnology has attracted biologists, R&D
firms, and venture companies . The discovery of superconductivity in materials at liquid
nitrogen temperatures had a magnetic effect on physicists and chemists, as weil as funding
agencies all over the world. Expectations, however, are not limited to such highly visible
examples of early prom ises. They are at play already when a researcher chooses between
options to folIowand does so in terms of the expectation of success.
27 The notion was introduced by Van den Belt and Rip (in Bijker et al. [1987], p. 140: " ...
there have to be expectations ab out the success of continuing work within [a particular]
cluster of heuristics - expectations that must be embedded in the subculture of the technical
practitioners and others involved in the development." They add in a footnote: " Sometimes
the relevant expectations may be of a broad, even sociological nature . In the case of the
French VEL project, the EDF engineers working on this particular project acted on an
explicit vision of where French society was going - avision that was characterized by Callon
SCIENCE AND TECHNOLOGY AS DANCING PARTNERS 261
as a variant of Tourainian sociology . One can also think of the Japanese Fifth Generation
project, which seems to have been predicated on, inter alia , Daniel BeIl's ideas on the coming
of the post-industrial society. UnIike Touraine for the French EDF engineers, however, Bell
became an acknowledged ideologue for the Japanese "knowledge engineers [...[." See for
more extended discussion and empirical cases of the cultural matrix of expectations,
Vergragt, Mulder, Rip en Van Lente (1990).
,. Van den Belt and Rip [1984, 1987).
,. By contrast, the structural formula proposed by Kekule in 1866 (using a theory of organic
chemistry which was more correct according to our present day view) mere!y amounted to
the translation of Hofmann's formula into his own theory but did not indicate new
directions for innovation. Instead of immediately leading to technological applications, the
contribution of Kekule 's structure theory must be found in the way it affected the cultural
matrix of synthetic dye chemistry. It was only in 1880 that the constitution of rosaniline was
finally elucidated by Emil and Otto Fischer in terms of Kekule's theory. This elucidation was
the source ofa new sequenceof dye innovations. (Quoted from Van den Belt and Rip (1987),
p .I44.)
30 Van den Belt & Rip (1984) note that the Muttersubstanz heuristic was after a time used in
be distinguished from Nelson and Winter's "natural trajectories", which are, in our
terminology, stabilized heuristics. See their (1977), p, 56: "[ ... ] there are certain powerful
intra project heur istics that apply when a technology is advanced in a certai n direct ion, and
payoffs from advancing in that direction that exist under a wide range of demand
conditions. We call these directions "natural trajectories" . If natural trajectories exist,
following these may be a good strategy." Nelson and Winter's equ ivalent of our
technological paradigm is " technological regime" ,
32 Note that there is no impIication that technological developments always follow
trajectories. A trajectory is a contingent happening . This point underscores my difference
with Latour (compare notes 3 and 22). " Trajectory" does not imply any technological
determinism . It indicates that search processes are sometimes oriented in specific ways,
33 Constant [1984].
3. Homburg (1983).
35 This was recognized as such at the time, and led to problems with the patent law, which at
that time required inventiveness for a patent to be awarded , so did not allow products from
routinized research to be patented. The solution was to add an escape clause: if a commercial-
ly useful effect was found , the finding was still patentable (Van den Belt and Rip (1987)).
3. This observation is related to the debate, started by Latour [1987], whether one should
include non -human as weIl as human actors in the network dynamics. My position is that to
understand technological dynarnics, one should include expectations and the way these
anticipate and mobilize . Voicing expectations , and being swayed by them, are Iimited to
human actors. In fact , Callon (in Callon et al. [1986] and Callon (1987)), in his notion of
actor-world as a scenario created and pushed by a human or institutional actor (in his case,
Electricite de France) is using this notion already. Electrons can act, are actors (or agents) in
that sense , but cannot interact on the level of anticipations and legitimations,
262 ARIE RIP
3. A similar observation can be made about the Starnberg model of development of sciences
which has been used to der ive policy impl ications, for example the differential possibilities
of orientation towards social relevance in different phases of development (Schäfer (1983).
Rip (1981)).
40 Van den BeIt and Rip (1987); see also Schot (1991).
4. Genentech, the first of such firms to offer its shares on the open market, was introduced
to the New York Stock Exchange in October 1980. Its issue price was $35. but prices jumped
to $89 in just one hour, to stabilize at $75 by the end of the first day . This phenomenon has
been quoted often, while much less notice has been taken of the fact most of the increase
disappeared Inter on: by December 1980. Genentech stock stood at $45.
42 The field level is less visible in another locus for negotiation of expectations: inside firms.
It is difficuIt to make general pronouncements, because not many cases have been
documented yet. One example is the way the strong fibre Twaron was developed within
Akzo Company. At one moment, the board of directors decided to stop the developrnent ,
but the network of alliances and expectations lower in the firm had become so strong that
work continued (and ach ieved success in the end) . See Mulder and Vergragt (1988).
43 Cf. Bijker, Hughes and Pinch (1987).
44 The example of engineering design will indicate the possibilities of this approach. A
design or a blueprint is not just a set of instructions, a virtual artefact as it were, but also a
story ab out how functions can be fulfilled and specifications can be met. The story must be
convincing to move others into realizing the artefact (or an artefact more or less like the
design intended); in addition, chances for its eventual survival should be increased. Thus, a
good design has to take further requirements into account: the mobilization of varied
resources for its realization and survival. It is a matter of heterogeneous engineering to
develop working technology. Law (1987)
Engineering design is also the place where one should anticipate on future niches and future
expansion. If all this is taken up in the "story telling", technological development would
become reflexive. i.e . self-conscious about its dynamics. In practice, however , there are
severe limitations. For one thing, there is too much division of labour, which induces
modular optimization . Computer-aided design, an important tool , helps to experiment with
SCIENCE AND TECHNOLOGY AS DANCING PARTNERS 263
meeting specifications, but lets designers forget ab out heterogeneous engineering . The
danger is that so-called path-dependencies, unavoidable in quasi-evolutionary develop-
ments, will become counter-productive. Soete [1988J, p. 56, quotes Arthur and David for an
example of locked-in technological development due to path dependency: the QWERTY
keyboard of typewriters and almost all keyboards, including those of computers, which is
not the most efficient outlay, but which has resisted attempts at change . The introduction,
in the late 19th century, of this kind of outlay of letters was related to the requirement of
slowing down the speed of the typist. to avoid the hammers getting cluttered all the time. See
David [1985). In other words, rationalization of design should not just be optimalization of
particular bits and pieces, but take process rationalization as starting point.
•• This section draws heavily on Disco, Rip and Van der Meulen [l992J.
•• Compare Vincenti's discussion of normal design in relation to normal configuration of a
device, e.g. in Vincenti [1991J, and Vincenti, this volume.
•, Pambour's theory of the steam engine is a good example, also because it was widely used
to improve local design work (see Kroes, this volume). The attempts, for example by
Reuleaux in the 1860sand 1870s, to formulate a general theory of machines are almost too
abstract. In mechanical engineering curricula, however, one sees definite indications of such
a generalized approach .
•• Mathematization of design had already been prepared in France in the circles of the Corps
des Ponts et Chaussees and among Polytechniciens in general, for example in the elaborate
theory of suspension bridges published by Henri Claude Navier in 1823(and which won hirn
admission to the Academie in the following year) or the thermodynamic theory of steam
engines published by Sadi Carnot in the 1820s. For Navier and the social origins of French
theoretical mathematical modelling see Kranakis [1989J . Carnot's work itself is not
mathematical; later Clapeyron and Clausius elaborated actual mathematization of
thermodynamics.
•• A similar point can be made (and has been made) about the importance of teaching
practices, and the teaching function in general , for the emergence of theory in science, e.g.
by Janich [1978), and is prefigured in Böhme et 01. [l978J analysis of medieval cathedral
building. In Rip [1982J, I add to this argument by looking at technical dynamics (i.e. control
of conditions and effects) within science, and show that "Increasing restrictedness leads to
empirical generalizations and conceptual distinctions, i.e. "bottorn-up" theory formation .'
(p. 231) Generalization is important in science to spread one's know1edge claims more
widely; this becomes important in technology as soon as technical sciences have created a
domain for themselves, with colleague technical scientists rather than designers and
practitioners as "consumers' . The point in the main text is that generalization is already
stimulated by the need to spread one's trainees more widely.
s 0 In radio engineering, on the other hand, the utter novelty of the technology implied a
diffuse and uncertain theoretical base and the consequent dominance of practice over
theory . Technical models tended to be intuitive or at best highly speculative and, such as they
were, to be generated mainly within firms manufacturing wireless equipment and within
state agencies with mandates in this domain. Only after radio technology began to stabilize
in the mid-I920s did engineering schools become significant research sites for "meta-
modelling" (generally aimed at high-level theoretical models of waveforms, propagation,
etc.) and did reasonably elaborated cosmopolitan design networks emerge. Cf. Aitken
[1978J; Disco [199OJ discusses the cases of radio engineering and reinforced concrete .
264 ARIE RIP
51 Buchholz [1979]; Rip and Van der Es [1980]. Also Vermeij [1990] for a glimpse of the
networks was typical of the European continent and that other developmental patterns
prevailed elsewhere. In Britain and especially the United States, where engineering schools
were often tacked on to universities and where universities in any case had long experience
in interfacing with practical fields (e.g. agriculture, industry, and mining) academic research
developed an applied orientation early on in the game .
es The not ion of a design hierarchy is used to structure the division of labour in specific
design tasks, e.g. of an airplane (see Vincenti [1991], and this volume) . It has also been used
10 describe the overall pattern of design in an industry or sector (see Clark [1985]), where the
dominant design at the top functions like a paradigm . I introduce aseparate concept ,
technical hierarchy, to emphasize the stratified character of modern technology in general,
which allows for a generalized division and coordination of labour.
56 It is for this reason that in Pavitt's [1984] taxonomy of innovation patterns supplier-
dominated innovation comes out as one of the four main patterns. Suppliers can also take
initiatives, based on their generalized position, e.g. when in the 1960s the petrochemical
industry developed biodegradable fatty acids for use in synthetic detergents (to overcome
environmental problems) rat her than the detergent industry itself (Daey Ouwens et 01.
[1987], Ch . 4). Note that the research was of a problem-solving nature, and thus did not lead
to a lasting pattern of division of labour.
57 The quotes are from Hughes [1989], p . 2 and 471, while the description of technological
While they do not use a linear model for the relation between science and technology, they
seem to assurne such a pattern for the new electronic and communication technology, as it
spreads through society , Freeman, accordingly, has been emphasizing the need to invest in
the implementation of these technologies.
59 Rip [1992].
SCIENCE AND TECHNOLOGY AS DANCING PARTNERS 265
6. Kodama (1990).
6' Laredo and Callon (1990), pp. 166-167 .
62 Rip [I989b) refers to IEEE Spectrum [Oct . 1988), pp , 24-28, for the simulation link
between design and manufacturing of VLSI. He also suggests that the innovation race to
create ever larger chips could now be done completely on computers: war games instead of
actual war. Orson Scott Card's science fietion novel Ender's Game offers an interesting (and
moving) portrayal of the ambiguities of such a situation.
6J Laredo and Callon (1990).
6' There is (I think) complete equivalence with scenario building to explore futures and
articulate decision making. This would explain the similarity with IIASA type model
building (cf. energy, acid rain) and other mixed technical-social modeling, e.g . of transport
systems , The logistics of transport systems (including their embedding) become object of
technologieal development work .This is clear, for example, from the evolution of the
relevant departments in Delft Technologieal University. See P rins (1987).
" See Rip (1990) and (1991) for a more extensive discussion.
" The acronym PLACE was coined by John Ziman to indieate the Proprietary, Local,
Authoritarian, Commissioned and Expert nature of industrial research and a lot of present-
day scient ifie research, and contrast this with the Mertonian "CUDOS" norms:
Communalism, Un iversalism, Disinterestedness, Organized Scepticism. (Ziman (1983),
(1990)).
67 Professional engineering has evolved its own kind of " scientizing" over aperiod of a
century or so ; a similar argument can be made for more recent developments, e.g . in expert
advice on health, on environment and on global issues. Then, it is understanding relevant to
decision-rnaking that counts. So me (and sometimes most) of the work is not made public,
and if public, often in reports rather than "regular' scientifie publi cations (compare the
increasing prominence of grey literature). Still , there is recognition of performance, and
accompanying rewards in terms of resources and careers. So scientists can (and will) move
in this direction. Clearly, the traditionally organized R&D system is not sacrosanct, in an y
case not something for whieh many scienti sts will sacrifice career chances, and actually
possibilities of research. And one can in fact see some "new cosrnopolitanism" , becau se
hybrid institut ions emerge like mixed scientifi c-policy conferences, organizations like the
International Institute of Applied System s Analys is near Vienna that create career resources
and mobility. In this way, contributions of experts will still be visible and a fun ctioning
reward system can evolve again .
6. The term has been used , see the subtitle of Wittrock and Elzinga (1985), but without the
critieal assessment required.
6. For "technologie de base " see Laredo and Callon (1990), for " strategic science " Irvine
and Martin (1984), and Van Lente and Rip (1991). In fact , the word "technology" itself used
to be such a new category: At the Chicago World Fair of 1933, the linear model ideology was
still expressed as "Science Finds, Industry Applies, Man Conforrns" (my emphasis) . Use of
the term " technology" is related to legitimation strategies of engineers and to the
transformations during and after World Wa r II .
7. Cf. also Rip (1989).
11 Kranakis, this volume.
REFERENCES
Abernathy, William J ., Clark, Kim B., and Kantrow, Alan M.: 1983, Industrial Renaissance
- Producing a Competitive Future for America, New York : Basic Books.
Aitken, Hugh G .J .: 1978, 'Science, Technology, and Economics: The Invention ofRadio as
a Case Study,' in Krohn, Wolfgang, Layton, Edwin T . and Weingart, Peter (eds.) The
Dynamics 0/ Science and Technology, pp. 89-112.
Barnes, Barry: 1974, Scientific Knowledge and Sociological Theory , London: Routledge
and Kegan Paul.
Bames, Barry, and Edge, David (eds.): 1982, Science in Context, Milton Keynes: Open
University Press.
Bijker, Wiebe E., Hughes, Thomas P. , and Pinch, Trevor 1. (eds.): 1987, The Social
Construction 0/ Technological Systems. New Directions in the Sociology and History 0/
Technology, Cambridge, MA : MIT Press.
Böhme, Gernot , van den Daele, Wolfgang, and Krohn , Wolfgang: 1973, ' Die Finalisierung
der Wissenschaft' , Z. f Soziologie 2, 128-144.
Böhme, Gernot, van den Daele, Wolfgang, and Krohn, Wolfgang. 1978, 'T he
"Scientification" of Technology", in Krohn , Wolfgang et al. (eds .) The Dynamics 0/
Science and Technology, pp . 219-250.
Buchholz, Klaus: 1979, 'Die gezielte Förderung und Entwicklung der Biotechnologie", in
Van den Daele , Wolfgang et al. (eds.) Geplante Forschung, pp . 128-144.
Callon, Michel : 1987, 'Society in the Making: The Study of Technology as a Tool for
Sociological Analysis', in Bijker, Wiebe E . et al. (eds .) The Social Construction 0/
Technological Systems. New Direct ions in the So ciology and History of Technology, pp.
83-103 .
Callon, Michel, Law , John, and Rip, Arie: 1986, Mapping the Dynamics 0/ Science and
Technology, London: Macmillan .
Casimir, Hendrik: 1983, Haphazard Reality , A HalfCentury 0/ Science , New York: Harper
and Row .
Chubin, Daryl E . and Chu, Ellen W . (eds .) : 1989, Science 0// the Pedestal. Social
Perspectives on Science and Technology, Belmont, CA : Wadsworth Publ.
Clark, Kim B.: 1985, 'The Interaction of Design Hierarchies and Market Concepts in
Technological Evolution", Research Policy 14, pp . 235-251.
Cole, Gerald S.: 1988, 'T he Changing Relationships Between Original Equipment
Manufacturers and Their Suppliers' , Int. J. Technology Management 3, pp. 299-324.
Constant, Edward W. 11: 1984, 'Communities and Hierarchies: Structure in the Practice of
Science and Technology', in Laudan, Rachel (ed .), The Nature 0/ Technological
Knowledge. Are Models 0/ Scientific Change Relevant? Dordrecht: Reidel.
Cozzens, Susan E., Healey, Peter, Rip , Arie, and Ziman, John (eds.): 1990, The Research
System in Transition . Dordrecht: Kluwer Academic Publishers.
Daey Ouwens, Cees, van Hoogstraaten, Pieter, Jelsma, Jaap, Prakke, Frits, and Rip , Arie:
1987, Constructief Technologisch Aspectenonderzoek, Een verkenning. 's-Gravenhage:
NOTA.
Daniels, George H. : 1967, "The Pure-Science Ideal and Democratic Culture', Science 156,
pp. 1699-1705.
David, Paul A .: 1985, 'Clio and the Economics of QWERTY' , Amer. Econ . Rev, 75, pp .
332-337.
SCIENCE AND TECHNOLOGY AS DANCING PARTNERS 267
Disco, Cornelis, Rip, Arie, and van der Meulen , Barend: 1992, "Technical Innovation and the
Universities : Divisions of Labor in Cosmopolitan Technical Regimes', Social Science
Information 31, pp . 465-507.
Dosi, Giovanni: 1982, "Technological Paradigms and Technological Tra jectories: A
Suggested Interpretation of the Determinants ofTechnical Change', Research Policy 11,
pp. 147-162 .
Dosi, Giovanni, Freeman, Christopher, Nelson, Richard, Silverberg, Gerald, and Soete, Luc
(eds.): 1988, Technical Change and Economic Theo ry , London: Pinter.
Elias, Norbert, Martins, Herminio, and Whitley, Richard (eds.): 1982, Scientific
Establishmentsand Hierarch ies, Sociology of the Seiences Yearbook, Vol. VI, Dordrecht:
Reidel.
Elliot, Brian (ed.): 1988, Technology and Social Process , Edinburgh : Edinburgh University
Press .
Freeman, C.: 1974, The Economics of Industrial Innovation, Harmondsworth, Penguin .
Gremmen, Bart (ed.): 1991, The Interaction between Technology and Science , Wageningen :
Department of Applied Philosophy, Wageningen Agricultural University .
Harre, Rom : 1990, 'The Dependence of ' Hi-Tech' Science on Technology through the Role
of Experimental Technique', RISESST 1, pp. 35-42.
Homburg, Ernst: 1983, "The Influence of Demand on the Emergence of the Dye Industry.
The Roles of Chemists and Colourists', J. Soc. Dyers and Colourists 99, pp. 325-333.
Hughes, Thomas P . : 1983, Networks of Power. Electrijication in Western Societies, 1880-
1930, Baltimore, MD : Johns Hopkins University Press.
Hughes, Thomas P . : 1986, "The Seamless Web : Technology, Science, Etcetera, Etcetera',
Social Studies of Science 16, pp. 281-292 .
Hughes , Thomas P .: 1988, 'Inventing Controls for Large Technological Systems', in Mayer ,
Evelies (Hrsg .), Ordnung, Rationalisierung, Kontrolle, pp . 83-94, Darmstadt:
Technische Hochschule.
Hughes, Thomas P.: 1989, American Genesis: A Century of Invention and Technological
Enthusiasm, 1870-1970. New York, N.Y.: Viking .
Irvine, lohn and Martin, Ben R.: 1984, Foresight in Science. Pieking the Winners, London:
Frances Pinter.
Janich, Peter: 1978, ' Physics - Natural Science or Technology?' in Krohn , Wolfgang et al.
The Dynamics of Science and Technology, pp . 3-27.
Knorr-Cetina, Karin D. and Mulkay , Michael (eds.): 1983, ScienceObserved. Perspectiveson
the Social Studies of Science, London: Sage.
Kodama, Fumio: 1990, 'Japanese Innovation in Mechatronics Technology', in Sigurdson,
Jon (ed.) Measuring the Dynamics of Technological Change, pp. 39-53, London: Pinter
Publishers.
Kranakis , Eda : 1989, 'Social Determinants of Engineering Practice: A Comparative View of
France and American in the Nineteenth Century', Social Studies of Science , pp. 5-70.
Kroes, Peter: 1992, 'On the Role of Design in Engineer ing Theories; Pambour's Theory ofthe
Steam Engine ' , this volume .
Krohn, Wolfgang. Layton, Edwin T ., Jr ., and Weingart, Peter (eds.): 1978, The Dynamics
ofScienceand Technology, Sociologyofthe Seiences Yearbook, 1978, Dordrecht: Reidel.
Kuhn, Thomas S. : 1970, The Structure ofScientific Revolutions, 2nd enlarged ed ., Chicago :
University of Chicago Press.
268 ARIE RIP
Laredo, Philippe, and Callon , Michel: 1990, L 'imp act des programs communautaires sur le
tissue scientifique et technique franca ise, Paris : La Documentation Francalse.
Latour, Bruno : 1983, 'Give Me a Laboratory and I Will Raise the Wor1d' , in Knorr-Cetina
and Mulkay (eds.), Science Observed. Perspectives on the Social Studies 0/ Science, pp.
141-170.
Latour, Bruno: 1987, Science in Action , Mihon Keynes: Open University Press .
Latour, Bruno : 1988,' How to Write "The Prince" for Machines as weilas for Machinations',
in Elliot , Brian (ed.) Technology and Social Process, pp. 20-43 .
Latour, Bruno: 1990, ' The Prince of Machines (ltalian version)', RISESST I , pp. 15-34 .
Latour, Bruno , and Woo1gar, Steve: 1979, Laboratory Life. The Social Construction 0/
Scientiflc Facts, Beverly Hills and London: Sage.
Law, John: 1987, 'Techno1ogy and Heterogeneous Engineering: The Case of Portuguese
Expansion', in Bijker, Wiebe E. et al. (eds.) The Social Construction 0/ Technological
Systems. New Directions in the Sociology and H istory of Technology, pp . 111-134.
McCormmach, Russell: 1983, Night-Thoughts 0/ a Classical Physicist, New York: Avon
Books (orig. Harvard V.P . 1982).
McMilIan, Frank: 1979, The Chain Straighteners, London, Macmillan .
Mu1der, Kare! and Vergragt, Philip: 1988, 'Weaving the Super-Fibre Network' , paper, 4S-
EASST Conference , Amsterdam, 16-19 November 1988.
Nelson, Richard and Winter , Sidney: 1977, ' In Search of a Useful Theory of Innovation',
Research Policy 6, pp. 36-76.
Nelson, Richard and Winter , Sidney: 1982, An Evolutionary Theory 0/ Economic Change,
Cambridge MA: Belknap Press of Harvard University.
Pavitt, K.: 1984, 'Sectoral Patterns of Technical Change: Toward s a Taxonomy and a
Theory', Research Policy 13, pp . 343-373.
Pinch, Trevor J . and Bijker, Wiebe E.: 1984, 'The Social Construction of Facts and
Artefacts' , Social Studies 0/ Science 14, pp . 399-442 .
Prakke, F.: 1988, 'Manufacturing Innovation Between the State and the Market ', in
Gerstenfeld, A. and Warnecke , H.J . (eds.) Proceedings of Conference on Manufacturing
Research: An International Perspective, Amsterdam, Elsevier.
Price, Derek J . de Solla: 1965, 'Is Technology Historically Independent of Science? A Study
in Stat istical Historiography', Technology and Culture 6, pp . 553-568 .
Price, Derek J . de Solla: 1984a, 'Notes Towards a Philo sophy of the Science/Technology
Interaction,, in Laudan, Rache! (ed.), The Nature 0/ Technological Knowledge. Are
Models 0/ Scientific Change Relevant?, pp . 105-114, Dordrecht: Reidel.
Price, Derek J . de Solla: 1984b, 'The Science/Technology Relationship , the Craft of
Experimental Science, and Policy for the Improvement of High Technology Innovation',
Research Policy 13, pp. 3-20 .
Prins, G.: 1987, Een hele hijs. Afscheidsrede hoogleraar in de Transporttechniek, Delft:
Technische Universiteit .
Rip, Arie: 1981,'A Cognitive Approach to Science Policy', Research Policy 10, pp. 294-311 .
Rip, Arie: 1982, 'The Development of Restrictedness in the Sciences', in Elias, Norbert et al.
(eds.) Scientific Establishments and H ierarchies, pp. 219-238.
Rip, Arie: 1985, ' Interdisciplinarity without Disciplines? A View from the Sociology of
Science', in Mar, Brian W. , Newell, William T., and Saxberg, Borje O. (eds.) Managing
High Technology, pp. 185-192, Amsterdam: Elsevier Science Publishers.
SCIENCE AND TECHNOLOGY AS DANCING PARTNERS 269
Rip, Arie: 1986, 'The Mutual Dependence of Risk Research and Politiea1 Context', Science
and Technology Studies 4(3/4) (Fall/Winter), pp. 3-15.
Rip, Arie : 1989a, 'Opening the Black Box of Technological Development', in Chubin, Daryl
E. and Chu, Ellen W. (eds.) Science offthe Pedestal. Social Perspectiveson Science and
Technology, pp. 137-144 .
Rip, Arie: 1989b, 'Oe partiele rationaliteit van VLSI ontwikkeling en gebruik', T. Nedl.
Elektr. and Radio Genootsch. 54(2), pp . 55-60.
Rip, Arie: 1990, 'An Exercise in Foresight : The Research System in Transition - to What?',
in Cozzens, Susan E. et al. (eds.) The Research System in Transition, pp . 387-401.
Rip, Arie: 1991 , 'The R & 0 System in Transition: An Exercise in Foresight', in Science and
Social Priorities - Perspectives of Science Policy for the /990s, Proceedings of the
conference sponsored by International Council for Science Policy Studies, UNESCO, and
Institute of Theory and History of Science, Czechoslovak Academy of Sciences, Prague,
June 5-7 , 1990, Prague, pp. 43-55.
Rip, Arie : 1992, 'Cognitive Approaches to Technology Policy', forthcoming in the
Proceedings of the Third NISTEP International Conference on Science and Technology
Policy Research, Oiso (Japan), 9-11 March 1992.
Rip, Arie, Schoemaker, Nelleke, and Meus, Corrie : 1983, De Universiteit als " Transf er
Agent", Leiden: LISBON, University of Leiden.
Schäfer, Wolf (ed.): 1983, Finalization in Science. The Social Orientation of Scientific
Progress, Dordrecht : Reidel.
Schot, Johan: 1991, Technology Dynamics: An Inventory of Policy Implications for
Constructive Technology Assessment, 's Gravenhage: Netherlands Organization for
Technology Assessment, NOTA Werkdocument 35.
Van den Belt, Henk and Rip, Arie: 1984, Technologie-ontwikkeling: Het Nelson-
Winter/Dosi model, Leiden : DSWO, University of Leiden.
Van den Belt, Henk, and Rip, Arie : 1987, 'The Nelson-Winter/Dosi Model and Synthetic
Dye Chemistry', in Bijker, Wiebe E. et al. (eds.) The SocialConstruction ofTechnological
Systems. New Directions in the Sociology and History of Technology; pp. 135-158.
Van den Daele, Wolfgang: 1978, "The Ambivalent Legitimacy of the Pursuit of Knowledge',
in Boeker, Egbert, and Gibbons (eds.) Science, Society and Education, pp . 23-62, VU
Bookshop, Amsterdam.
Van den Daele, Wolfgang, Krohn, Wolfgang, and Weingart , Peter (Hrsg .): 1979, Geplante
Forschung, Frankfurt a/Main : Suhrkamp.
Van Lente, Harro, and Rip, Arie: 1991, 'Membrane Technology: An Example of the
Dynamics of Strategie Scientific-Technologieal Fields', revised version of paper presented
at the International Workshop on Network Approaches in Technology Studies,
University of Groningen 30-31 May 1990, Enschede: University of Twente, Department
of Philosophy of Science and Technology .
Vergragt, Philip, Mulder, Karei, Rip, Arie, and van Lente, Harro: 1990, De matrijs van
verwachtingen ingevuld voor de polymeren Tenax en Twaron, Groningen/Enschede,
University of Groningen/University of Twente, NOTA .
Vermeij, Erwin : 1990, Contextuele verschillen in de ontwikkeling van technische
toepassingen van methaangisting, bestudeerd m.b. v. een variatie-selectie model, TWIM-
studies No . 9, Eindhoven: Technieal University Eindhoven, Research Cent re for
Engineering Sciences, Innovation and Society.
270 ARIE RIP
Vincenti, WaIter G. : 1991 , 'The Scope for Social Impact in Engineering Outcomes: A
Diagrammatic Aid to Analysis', Social Studies of Science 21, pp. 761-767.
Weingart, Peter: 1978, 'The Relation between Science and Technology - A Sociological
Explanation', in Krohn et al. (eds.) The Dynamics ofScience and Technology; Sociology
of the Seiences Yearbook, pp. 251-286, Dordrecht: Kluwer Academ ic Publishers.
Wittrock , Bj örn and Elzinga , Aant (eds.): 1985, The Unlversity Research System . The Public
Policies of the Home of Scientists, Stockholm: Almq vist and Wiksell International.
Zilsel, Edgar: 1976, Die sozialen Ursprung der neuzeitlichen Wissenschaft, Krohn , Wolfgang
(H rsg .) , Frankfurt a/Main: Suhrkamp.
Zirnan, John: 1976, The Force of Knowledge. The Scientific Dimension of Society,
Cambridge: Cambridge U.P .
Ziman, John: 1983, 'The Collectivization of Science', Proc. Royal Soc.• B 219, pp . 1-19 .
Zirnan, John: 1990, 'Research as a Career' in Cozzens, Susan E. et al. (eds.) The Research
System in Transition, pp . 345-359 .
NAME INDEX
271
272 NAME INDEX
Also 0/ interest:
R.S. Cohen and M.W. Wartofsky (eds.): A Portrait 0/ Twenty-Five Years Boston
Colloquia for the Phi/osophy ofScience , 1960-1985. 1985 ISBN Pb 90-277-1971-3