The Competitiveness of Nations in a Global Knowledge-Based Economy


Michael Polanyi

Science and Technology

Personal Knowledge: Towards a Post-Critical Philosophy

Harper Torchbooks, NYC, 1962, 174-184

In the list of three kinds of learning of which animals are capable I have placed trick learning before sign learning, since motoricity is fully developed in the lower animals before they achieve the capacity for recording complex perceptions.  Yet the capacity to perform a useful action presupposes some purely intellectual control over the circumstances in which the action is to take place.  Technology always involves the application of some empirical knowledge and this knowledge may be part of natural science.  Our contriving always makes use of some anterior observing.

Putting it this way, we become aware of the incommensurability of the two things combined in a technical performance.  Suppose you hammer in a nail.  Before starting, you look at the hammer, the nail and the board into which you will drive it; the result is knowledge which you can put into words.  Then you hammer in the nail.  The result is a deed: something is now firmly nailed on.  Of this you can have knowledge, but it is not itself


knowledge.  It is a material change which counts as an achievement.  Knowledge can be true or false, while action can only be successful or unsuccessful, right or wrong.

It follows that an observing which prepares a contriving must seek knowledge that is not merely true, but also useful as a guide to a practical performance.  It must strive for applicable knowledge.

The conceptual framework of applicable knowledge is different from that of pure knowledge.  It is determined primarily in terms of the successful performances to which such knowledge is relevant.  Take hammering again.  This performance implies the conception of a hammer, which defines a class of objects that are (actual or potential) hammers.  It will include, apart from the usual tools of this kind, rifle butts, shoe heels and fat dictionaries, and establish at the same time a grading of these tools according to suitability.  The suitability of an object to serve as a hammer is an observable property, but it can be observed only within the framework defined by the performance it is supposed to serve.

There are three kinds of observable things which can be defined by their participation in practical performances: (1) materials, (2) tools, including all manner of installations, and (3) processes.  Timber, textiles, fuels, are technical materials; hammers, engines, houses, railways, are tools or installations; fermentation, cooking, smelting, are technical processes.  Many of these technical conceptions comprise a variety of otherwise disparate objects (for example, different kinds of textiles, from cotton and wool to nylon and glass fibres, and different means of lighting, from candles to discharge lamps), but all these objects are specially prepared, shaped or otherwise so contrived as to make them technically suitable.  To this extent these classes of objects or processes are known, and the individual objects or processes themselves are intelligible only within the framework of a useful performance which they successfully serve.  Pure knowledge, lacking this framework, and pure science in particular, ignore these classes and cannot understand these contrivances.  We cannot eliminate instrumentality from technical knowledge, any more than we can represent natural science in terms of practical procedure.

A gap is opening up here between two kinds of knowledge, both of which refer to material things: one derived from an acknowledged purpose, the other unrelated to any such purpose.  The disparity of science and technology which I am examining here will prove relevant later to the relation between the science of inanimate things, in which no purpose is apparent, and that of living beings which can be understood only in teleological terms.  We should keep this prospect in mind while proceeding to elucidate further the characteristic logical structure of technology.

Primitive technology may be regarded as a mere extension of bodily skills employed for the satisfaction of bodily appetites.  And even in highly complex and predominantly articulate branches of technology, like the manufacture of cloth or the production of steel, there is involved a


measure of unspecifiable know-how which is essential to the efficiency of labour and the quality of its product.  Manufacturing experience remains a valuable qualification to a technician, and its possession by the aggregate of a country’s technicians is a great national asset.  But even though the teachings of technical science can become effective only by their skilful execution, the foundation of modern man’s technical mastery lies in the explicit exposition of technology by textbooks, journals, patents, etc.

Technology teaches action.  This is made plain when it speaks in imperatives, as it often does in cookery books or directions for the use of machinery.  The symbol at the head of a prescription is an imperative prefacing an order to make up a medicine; crafts like weaving or welding are taught in imperatives.  All technology is equivalent to a conditional command, for it is not possible to define a technology without acknowledging, at least at second hand, the advantages which technical operations might reasonably pursue.  It is true, of course, that anything a man does or can conceivably do, could be described as the pursuit of an advantage if we imputed to him the purpose of achieving the consequences of this action; but a technology which would teach all such imputable purposes would be as meaningless as a science which would give a list of all observable facts.  A technology must therefore declare itself in favour of a definite set of advantages, and tell people what to do in order to secure them.

Technology teaches only actions to be undertaken for material advantages by the use of implements according to (more or less) specifiable rules. [1]  Such a rule is an operational principle. As implements are defined and understood in terms of an action which they serve, they are defined and understood likewise in terms of the operational principle which tells how to perform such an action. [2]

I have spoken before of the operational principles which we observe subsidiarily in the performance of a skill, and also of the operational principles applied - again for the most part subsidiarily - in achieving scientific knowledge.  I have shown symbolic operations carried out according to certain explicit rules and have noted that such operations require that symbols should be manageable, just as tools have to be serviceable.  Modern electronic devices used for the automatic control of technical processes show that some highly formalized operational principles of technology can be readily affiliated to mathematical operations.  The meaning of technical implements resembles that of mathematical symbols, in so far as they are both intended for use in a certain range of operations,

1. ‘Material advantages’ should exclude inter alia the achievement of symbolic expression or of human interactions.  Thus the construction of churches and prisons or the manufacture of handcuffs are tasks of technology, but the ultimate uses of these objects are not part of technology.  The word ‘implement’ is meant to designate all three classes of useful things: materials, devices and processes.  Action according to ‘specifiable rules’ excludes artistic performances.

2.  Operational principles will be taken to include here the constructional principles which tell us the way technical devices, like machines or houses, are to be built.


in the service of which they can be replaced by a whole class of equally serviceable, though otherwise disparate, entities.  This kinship can be pursued through the whole subsequent analysis of operational principles.

The difference between scientific knowledge and an operational principle of technology is recognized by patent law, which draws a sharp distinction between a discovery, which makes an addition to our knowledge of nature, and an invention, which establishes a new operational principle serving some acknowledged advantage.  New inventions rely as a rule on known facts of experience, but it may happen that a new invention involves a new discovery.  Yet the distinction between the two will still hold: only the invention will be granted protection by a patent, and not the discovery as such.

The reason is obvious.  A patent has two functions, namely, publicly to disclose its subject matter, and to grant a monopoly in respect to its use.  If applied to new knowledge its first function would preclude the second: once such knowledge is publicly disclosed it can no longer be anyone’s monopoly.  But the patent can grant and enforce a monopoly for the practice of any new operational principle; it can restrain unauthorized persons from using the new invention which it makes generally known. [1]

Invention has it in common with discovery that it can claim to be what it is only if it is surprising.  It must be separated from its antecedents by a considerable logical gap.  I have mentioned already that in case of doubt the courts undertake to assess whether this logical gap is wide enough to warrant the acknowledgment of an invention.  This width measures the ingenuity of the invention.

But a new operational principle may be acknowledged by patent law, and yet not be an invention in the technological sense.  A new ingenious process for extracting tap-water from champagne may be an invention in the sense of patent law, but it would not be acknowledged as such by technology.  For in addition to the disclosure of a new operational principle, technology requires that an invention should be economic and thus achieve a material advantage.

Hence any invention can be rendered worthless, and indeed farcical, by a radical change in the values of the means used up and the ends produced by it.  If the price of all fuels went up one hundred-fold, all steam engines, gas turbines, motor-cars and aeroplanes, would have to be thrown on the junk heap.  A brilliant invention is often rendered nonsensical overnight by a better invention: tram-cars are as absurd today as the horse-drawn buses which they once displaced.  By contrast to this, the validity of a scientific observation cannot be affected by changes in the value of goods.  If diamonds became as cheap as salt is today, and salt as

1. The law could try to grant a monopoly for the future practical applications of a new discovery; but no patent law does this, for it is impracticable.  The law endorses thereby once more the sharp distinction between a knowledge of the facts of nature (achieved by discovery) and the knowledge of an operational principle (achieved by invention).


precious as diamonds are now, this would not invalidate any part of the physics and chemistry of diamonds or of salt.  If either of the two minerals became so rare as to be practically inaccessible, this might affect the interest attached to their study, but it would leave unimpaired the validity of its results.  Nor is there any true parallel in science to the extinction of an invention by the emergence of a more profitable way of achieving the same advantage.

The beauty of an invention differs accordingly from the beauty of a scientific discovery.  Originality is appreciated in both, but in science originality lies in the power of seeing more deeply than others into the nature of things, while in technology it consists in the ingenuity of the artificer in turning known facts to a surprising advantage.  The heuristic passion of the technician centres therefore on his own distinctive focus.  He follows the intimations, not of a natural order, but of a possibility for making things work in a new way for an acceptable purpose, and cheaply enough to show a profit.  In feeling his way towards new problems, in collecting clues and pondering perspectives, the technologist must keep in mind a whole panorama of advantages and disadvantages which the scientist ignores.  He must be keenly susceptible to people’s wants and able to assess the price at which they would be prepared to satisfy them.  A passionate interest in such momentary constellations is foreign to the scientist, whose eye is fixed on the inner law of nature.

Hence there arises a conflict of values which makes it difficult to mix the two occupations.  From his experience of developing atomic weapons in Los Alamos during the Second World War, J. R. Oppenheimer wrote: “The scientist is irritated by the practical preoccupations of the man concerned with development, and the man concerned with development thinks that the scientist is lazy and of no account and is not doing a real job anyway.  Therefore the laboratory very soon gets to be all one thing or all the other.” [1]

This sharp division between science and technology is entirely compatible with the existence of domains which in one respect or another form a transition between them.  The older crafts which still form the

1. J. R. Oppenheimer, ‘Functions of the International Agency in Research and Development’, Atomic Scientific Bulletin, 1947, p. 173.  See also V. B. Wigglesworth, ‘The Contribution of Pure Science to Applied Biology’, The Annals of Applied Biology, 42 (1955), pp. 34-44.  Speaking of pure scientists working on practical war-time problems, Wiggles-worth writes: ‘In the pure science to which they were accustomed, if they were unable to solve problem A they could turn to problem B, and while studying this with perhaps small prospect of success they might suddenly come across a clue to the solution of problem C.  But now they must find a solution to problem A, and problem A alone, and there was no escape.  Furthermore, there proved to be tiresome and unexpected rules which made the game unnecessarily difficult: some solutions were barred because there was not enough of the raw material available: others were barred because the materials required were too costly; and yet others were excluded because they might constitute a danger to human life or health.  In short, they made the discovery that applied biology is not “biology for the less intelligent”, it is a totally different subject requiring a totally different attitude of mind’ (p. 34).


majority of modern industries were invented by mere trial and error, without the aid of science.  By contrast, electrotechnics and much of chemical technology are derived from the application of pure science to practical problems.  Hence there is the following interrelation between science and technology.  To the extent to which a technical process is an application of scientific knowledge it contributes nothing to science, while empirical technology, which is itself unscientific, may well offer - for this very reason - important material for scientific study.’

We have, correspondingly, two forms of enquiry that lie between science and technology.  Technologies founded on an application of science may form a scientific system of their own.  Electrotechnics and the theory of aerodynamics are examples of systematic technology which can be cultivated in the same way as pure science.  Yet their technological character is apparent in the fact that they might lose all interest and fall into oblivion, if a radical change of economic relationships were to destroy their practical usefulness.  On the other hand, it may happen that some parts of pure science offer such exceptionally ample sources of technically useful information that they are thought worth cultivation for this reason, though they would otherwise lack sufficient interest.  The scientific study of coal, metals, wool, cotton, etc., are branches of such technically justified science.

Systematic technology and technically justified science are two fields of study lying between pure science and pure technology.  But the two fields may overlap completely.  The discovery of insulin as a cure for diabetes was an important contribution to science, owing to the intrinsic interest of its subject matter; it was also the invention of an operational principle serving to cure diabetes.  The same quality applies over large parts of pharmacology.  It holds, indeed, wherever a process inherent in nature is interesting to science owing to the importance of its outcome, while at the same time it can also be operated at will for achieving this desirable outcome.  Such coincidences between science and technology are fully accounted for by the same principles which define them in general as completely disparate domains. [2]

1. On the range of undisclosed knowledge buried in empirical technology, see p. 52 above.

2. In the address by Wigglesworth just cited (p. 178, n. 1), the author describes the varying relationships which obtain between pure and applied science in the biological field.  These two ‘totally different subjects’ may contribute to each other’s good in a number of ways.  E.g. for the pure scientist ‘one of the most efficient correctives to the dangers of over-specialization is provided by the stimulus of contact with practice’ (p. 36).  On the other hand, applied biology may turn to pure science for the systematic explanation of its practical discoveries (p. 38); and of course the applied biologist ‘in thinking about any practical problem... is continually making use of the whole range of scientific knowledge that exists about all its component parts’ (p. 40).  Yet the authorities are warned that this mutual advantage depends in the last analysis on the independence of pure biology from the narrower demands of the applied subject: ‘.... the D.S.I.R. makes grants for any research proposals which are of exceptional “timeliness and promise”.  The difficulty is that the most original ideas are at the outset both unpromising and untimely.  Only research which is totally unfettered can advance into the most unpromising fields... I very much doubt whether it would have been reasonable for the [A.R.C. to have supported, for example, Darwin’s experiments on the curving of bean shoots or the early experiments of the Wents on the growth of the oat coleoptile - because no one could have foreseen the impact that these observations were going to have on the agriculture of the future... But at least the Research Councils should take great care not to impede the advance of pure science...  Knowledge is a delicate plant, and... it is an undesirable practice to keep pulling plants up to see how the roots are getting on’ (pp. 42-3).]

HHC: [bracketed] displayed on page 180 of original.


Nothing could have appeared more obvious until recently than this difference between pure science and technology.  It is unquestioningly embodied in the general framework of higher education, as shown by its division into universities and colleges of technology; it is expressed in the current distinctions between pure and applied chemistry, pure and applied physics, pure and applied mathematics, etc., in the description of university chairs, journals and international congresses; it determines the conditions of employment of scientists in universities on the one hand and industrial laboratories on the other; it underlies the operation of the patent law.

This framework survives practically unchanged in the countries not subject to Marxism and has not been altogether abandoned in the Soviet Union either.  But since the rise around 1930 of the Neo-Marxian theory of science, which became within the subsequent decade the official doctrine of the U.S.S.R. and gained widespread influence outside it, the distinction between science and technology, even where still upheld in practice by the continued operation of these institutions, is violently challenged in principle.

This is part of the drive, described earlier on, for subordinating cultural values to a radically utilitarian conception of the public good: a materialistic outlook paradoxically imbued by inordinate moral aspirations.  Such an attack is of course double-edged.  It denies the effectiveness of pure intellectual passions in guiding scientific discovery, by affirming that every important step in the progress of science occurs in response to a specific practical interest; while it also denounces the pursuit of science for its own sake as irresponsible, selfish, immoral.  Taken literally, the two attacks are mutually incompatible, for something that does not really happen cannot be denounced as morally wrong.  But the materialistic interpretation of culture is a disguised imperative: it both declares that culture really is, and decrees that it ought to be, the servant of welfare.  This is part of the Laplacean system in which morality must seek the sanction of science by representing itself in terms of scientific predictions. [1]

I am not much concerned here with the question how serious this menace to science may prove in practice.  While the official repudiation by Stalinist orthodoxy of science pursued for its own sake led to the persecution and death in 1942 of Russia’s most distinguished biologist, N. I. Vaviov, and had resulted by 1948 in the suppression or serious distortion of various branches of biology, it seems otherwise to have imposed on natural scientists little more restraint than the obligation falsely to declare their work

1. The mechanism of this transformation will be examined in the next chapter.


to be guided by its practical usefulness.  And this may be all.  People may perhaps continue indefinitely to cultivate pure science, while professing a theory of science which exposes this occupation as a pretence or condemns it as an abuse.  Yet the spread of this doctrine among scientists in countries where they are not compelled to subscribe to it, does raise the question which is relevant for us here, whether the distinctive passions which animate the cultivation of science may be superseded one day by other passions, or may even simply fade away for lack of response to them.

I have answered the last question in the positive sense, when warning that science may be once more discredited, as it was by St. Augustine, if it cannot avoid denaturing our conception of man. [1]  The appreciation of natural science is of recent origin and its tradition is rooted in a limited area.  It is a single shoot of one civilization among many others of equal antiquity and richness.  The Greeks never developed a systematic natural science, nor did Byzantium or China, despite their technological achievements. [2]  Today we can speak confidently of sixteenth- and seventeenth-century science only because, with modern hindsight, we can easily separate the genuine works of science from unscientific admixtures.  Kepler’s Harmonics, published in 1619, was imbued with astrology, and it is typical in this respect of much subsequent writing among scientists for the following two or three generations.  I have mentioned already that Glanvill, one of the founders of the Royal Society in 1660, argued persistently for the acknowledgment of witchcraft.  Another founding fellow, John Aubrey. published little else than a treatise on occult phenomena. [3]  The Cartesian spirit dominating France at that time was a-prioristic rather than experimental.  Newton himself still occasionally used religious arguments in science; for example when he suggested that God gave the world an atomic structure, as most conducive to his purpose.  The great controversies of the nineteenth and twentieth centuries show that the struggle against intrusion of extraneous points of view into science have never ceased and that grave differences continue to persist in respect to these issues between a dominant majority and various dubiously established minorities of scientists.  Yet we may acknowledge that by the time Newton’s influence became prevalent, and particularly through his Optics, the method of observational science became effectively consolidated.  Since then, in spite of such uncertainties and vagaries as I have described in the section on Scientific Controversies, we may recognize a coherent body of men, standing in the same scientific tradition, moved by the proper temper and true appreciation of science.  Arago acclaiming Leverrier’s discovery of Neptune in 1846 as ‘one of the noblest titles of his country to the gratitude and

1. P. 141 above.

2. Stephen Runciman, Byzantine Civilization, London, 1936, ch. IX, and Joseph Needham, Science and Civilization in China, 2, Cambridge, 1956, pp. 26-9, 84.

3. Lytton Strachey, Portraits in Miniature, London, 1931, p. 23.


admiration of posterity’ [1] expressed this in clear accents.  No contribution to knowledge could be more useless than was the discovery of this remote new planet.

Actually, up to that time natural science had made no major contribution to technology.  The industrial revolution had been achieved without scientific aid.  Except for the Morse telegraph, the great London Exhibition of 1851 contained no important industrial devices or products based on the scientific progress of the previous fifty years.  The appreciation of science was still almost free from utilitarian motives.

But these sentiments were held within a very small area and were shared at no time by more than a minority of the local population.  The migration of science overseas and into Asian and African countries occurred slowly at a later period, when the medical, industrial and military value of science had greatly increased and could serve to recommend the reception of science to industrially less developed countries.  These auspices did not favour a true appreciation of science.  In all parts of the world where science is just beginning to be cultivated, it suffers from a lack of response to its true values.  Consequently, the authorities grant insufficient time for research; politics play havoc with appointments; businessmen deflect interest from science by subsidizing only practical projects.  However rich the fund of local genius may be, such an environment will fail to bring it to fruition.  In the early phase in question, New Zealand loses its Rutherford, Australia its Alexander and its Bragg, and such losses retard further the growth of science in a new country.  Rarely, if ever, was the final acclimatization of science outside Europe achieved, until the government of a country succeeded in inducing a few scientists from some traditional centre to settle down in their territory and to develop there a new home for scientific life, moulded on their own traditional standards. [2]

Encircled today between the crude utilitarianism of the philistine and the ideological utilitarianism of the modern revolutionary movement, the love of pure science may falter and die.  And if this sentiment were lost, the cultivation of science would lose the only driving force which can guide it towards the achievement of true scientific value.  The opinion is widespread that the cultivation of science would always be continued for the sake of its practical advantages.  It was expected, for example, that Lysenko’s theories, if false, would be soon abandoned by the Soviet Government because they could produce no useful results.  This expectation overlooked the fact that such questions cannot be decided in practice.  Lysenko’s theories are actually the theoretical conclusions which Michurin in Russia and

1. See W. M. Smart, ‘John Couch Adams and the Discovery of Neptune’, Nature, 158, (1946), pp. 648-52.  Or listen to Ball commenting on the fact that Lalande would have discovered Neptune in 1795 if only he had believed what he saw on the 8th and 10th of May in that year.  ‘But had he done so, how lamentable would have been the loss to science.  The discovery of Neptune would then merely have been an accidental reward to a laborious worker, instead of being one of the most glorious achievements in the loftiest department of human reason’ (Sir R. S. Ball, The Story of the Heavens, London, 1891, p. 288).  

2. On tradition, see also p. 53 above.


Burbank in the U.S. derived from their substantial successes as plant-breeders. [1]  Almost every major systematic error which has deluded men for thousands of years relied on practical experience.  Horoscopes, incantations, oracles, magic, witchcraft, the cures of witch doctors and of medical practitioners before the advent of modern medicine, were all firmly established through the centuries in the eyes of the public by their supposed practical successes.  The scientific method was devised precisely for the purpose of elucidating the nature of things under more carefully controlled conditions and by more rigorous criteria than are present in the situations created by practical problems.  These conditions and criteria can be discovered only by taking a purely scientific interest in the matter, which again can exist only in minds educated in the appreciation of scientific value.  Such sensibility cannot be switched on at will for purposes alien to its inherent passion.  No important discovery can be made in science by anyone who does not believe that science is important - indeed supremely important - in itself. [2]

In saying this, I have acknowledged that values which I deem to be transcendent may be known only transiently to a small minority of mankind.  There is no contradiction in this: it correctly reflects the fact that universal validity is not an observed fact.  When we say that a statement is generally accepted or that no sane person would deny it, etc., we are saying something about the attitude of people towards the statement, which accredits the statement only if we accredit those people’s judgment of it.  But there is no general warrant to do this: the maxim ‘quod semper, ubique, ab omnibus’ has often proved erroneous.  The standards by which we observe or appraise can never be derived from statistical surveys.

Indeed, we cannot look at our standards in the process of using them, for we cannot attend focally to elements that are used subsidiarily for the purpose of shaping the present focus of our attention.  We attribute absoluteness to our standards, because by using them as part of ourselves we rely on them in the ultimate resort, even while recognizing that they are actually neither part of ourselves nor made by ourselves, but external to ourselves.  Yet this reliance can take place only in some momentary circumstance, at some particular place and time, and our standards will

1. See T. Dobzhansky, ‘The Fate of Biological Science in Russia’, Proceedings of the Hamburg Congress on Science and Freedom, London, 1955, p. 216.  The attempt to define science in terms of its practical success has already been shown to be logically untenable (see p. 169 above).

2. Some parallels from remoter fields may throw light on the principle involved here.  Suppose it were decided by psychiatrists that a general increase in psychoneurotic ailments could only be checked by a restoration of religious faith; this would not make us all believe in God.  In fact no ulterior advantage can make us believe in God, while if we do believe in God no consequent disadvantage can make us lose our faith.  Or suppose that the people of the United States came clearly to the conclusion from a study of British experience that they would live together more intimately if their common affections were attached to a King and a Royal Family.  This would not in itself produce such affections, or establish a monarchy in the U.S.  No genuine affections can ever be produced by ulterior motives; they must discover and uphold their satisfaction in themselves.


be granted absoluteness within this historical context.  So I could properly profess that the scientific values upheld by the tradition of modern science are eternal, even though I feared that they might soon be lost for ever.  This duality will be stabilized later within the concept of commitment.