12

Machines

The Renaissance view of nature as a machine … is based on the human experience of designing and constructing machines. The Greeks and Romans were not machine-users, except to a very small extent: their catapults and water-clocks were not a prominent enough feature of their life to affect the way in which they conceived the relation between themselves and the world. But by the sixteenth century the Industrial Revolution was well on the way. The printing-press and the windmill, the lever, the pump, and the pulley, the clock and the wheel-barrow, and a host of machines in use among miners and engineers were established features of daily life. Everyone understood the nature of a machine, and the experience of making and using such things had become part of the general consciousness of European man. It was an easy step to the proposition: as a clockmaker or millwright is to a clock or mill, so is God to Nature.

– R. G. Collingwood, The Idea of Nature (1945)¹

§ 1

The great philosopher and archaeologist R. G. Collingwood proposed a fairly straightforward technological determinism: new machines encourage new ways of thinking. There are two problems with his argument. The first is that the only machine that he lists that was new in the Renaissance was the printing press. The Middle Ages, it has long been maintained, saw a technological revolution, with the invention of the clock, the widespread diffusion of the water mill and the wheelbarrow, and the development of the various hoists and windlasses required to build the cathedrals; the view of nature as a machine should have come not in the sixteenth century but the fourteenth.² The second problem is even more fundamental: although Collingwood devotes a long chapter to the Renaissance idea of nature, and returns to the claim that the Renaissance viewed nature as a machine, he never gives a single example of someone describing nature as being like a machine. Collingwood was so sure that the Renaissance thought of nature as a machine that he failed to notice that he had produced no evidence to support his claim.

With this cautionary tale in mind, let us begin by asking a question which seems too obvious to need asking but is in fact an essential preliminary: What is a machine? First, in conceptual terms at least, there are the ‘simple machines’. Archimedes studied three elementary tools that could be used to move weights: the lever, the pulley and the screw. Hero of Alexandria (10–70 CE) added to these the windlass and the wedge, and in the late sixteenth century Simon Stevin included the inclined plane. All these simple machines provide a mechanical advantage in moving a load. The modern science of mechanics was crystallized in Galileo’s On Mechanics (1600 in manuscript; first published by Mersenne in 1634).³ Galileo was the first to demonstrate that the work performed by a machine could never be greater than the work put into it, so that machines can never trick nature into doing something that breaks its normal rules. (Thus a lever enables a light weight to lift a heavier weight, but the light weight travels further than the heavier weight, so the work done on each side of the fulcrum is the same.) Galileo thus established a new equivalence between natural processes and artificial ones. Because Galileo thought about machines in this narrow, technical way, he never says that the universe is a machine or that all natural processes can be understood in mechanical terms; neither does he ever compare the universe to a clock, which he certainly could have done had he wanted to.

What Galileo did discuss was atomism. The atomism of Democritus, Epicurus and Lucretius implied that the universe is made up of building blocks that function through their size, shape and solidity. As Democritus put it, ‘By convention sweet, by convention bitter, by convention hot, by convention cold, by convention colour: but in reality atoms and void.’⁴ In a world of atoms and the void all natural processes result from the ways in which atoms jostle together. In 1618, as a result of a conversation with Isaac Beeckman, the young Descartes came up with an alternative to ancient atomism: where the ancients had thought of atoms bumping into each other in empty space, Descartes rejected the possibility of empty space and thought in terms of corpuscles filling all the available space, as water fills the ocean. The next year Descartes formulated his famous doctrine cogito ergo sum, ‘I think therefore I am’; consequently, there is something, one thing, I know for certain. On this secure foundation he set out to build a new philosophy to replace that of Aristotle, and he began to publish elements of his new system in 1637. Since the publication of a book-length article by Marie Boas in 1952 it has become customary to refer to these two alternatives to the Aristotelian theory of forms and qualities – the atomic philosophy of the ancients (revived by Galileo, Gassendi and others) and the corpuscular philosophy of Descartes – together, as ‘the mechanical philosophy’.⁵ The term was certainly widely used in the late seventeenth century, but it is more misleading than helpful.

Descartes, who was taken to be the founder of the mechanical philosophy, never described himself in print as a mechanical philosopher; he says that all mechanical laws are physical or natural laws (which Galileo had shown), but not that all laws of nature are mechanical laws: he does not describe nature as a mechanical system. He does use the term ‘mechanical philosophy’ once in a letter (in 1637), where he refers to ‘the rather greasy and mechanical philosophy’ – in other words, the sort of philosophy a cart-maker would have. He was replying to a critic who had described his philosophy as ‘coarse and rather greasy’ and ‘excessively gross and mechanical’ – that is, too physical (as we might put it) to count as a philosophy at all. He wrote, ‘If my philosophy seems to him excessively gross because it considers shapes, sizes, and motions, as happens in mechanics, he is condemning what I think deserves praise above all else, and in which I take particular pride.’⁶ (Leonardo, too, had decided that being a mechanical philosopher, in this sense, should be a matter of pride.)⁷

The term ‘mechanical philosophy’ was coined by Henry More (a Cambridge don and life-long admirer of Plato) in 1659, after Descartes’ death, in the course of an attack on Cartesianism, which he had once enthusiastically supported.⁸ More wanted to defend the idea that spirit and purpose are active in nature and reject the Cartesian claim that natural processes are soulless, that matter is passive and that everything that happens (leaving aside the free choices of God, angels and men) happens of necessity. Outside England the term was slow to catch on: the first reference to it in Latin is in Samuel Parker’s Disputationes of 1678, and in French in Pierre Bayle’s Nouvelles de la république des lettres (1687).⁹ In English there was an alternative: Robert Boyle invented the term ‘the corpuscularian philosophy’ in 1662 to cover both ancient atomism and Descartes’ new corpuscular theory.¹⁰ ‘The corpuscular philosophy’ and ‘the mechanical philosophy’ are thus two competing terms for exactly the same thing: indeed, the first occurrence of either term in French is in a reference to la philosophie mécanique ou corpusculaire (1687; when Boyle was translated two years later the phrase used was ‘la philosophie des corpuscules’).¹¹

This is how Walter Charleton, writing in 1654, summarized what would soon be called the mechanical philosophy. Everything he says could have been said by Descartes:

Consider we, that the General Laws of Nature, whereby she produceth All Effects, by the Action of one and Passion of another thing, as may be collected from sundry of our praecedent Discertations, are these: (1.) That every Effect must have its Cause; (2) That no Cause can act but by Motion; (3) That Nothing can act upon a Distant subject, or upon such whereunto it is not actually Praesent, either by it self, or by some instrument, and that either Conjunct, or Transmitted; and consequently, that no body can move another, but by contact Mediate, or Immediate, i.e. by the mediation of some continued Organ, and that a Corporeal one too, or by it self alone.

Having set up his definition, Charleton goes on to attack the traditional concepts of sympathy and antipathy, and to argue that they must be reconceptualized in mechanical terms:

Which considered, it will be very hard not to allowe it necessary, that when two things are said either to Attract and Embrace one the other by mutual Sympathy, or to Repell and Avoid one the other, by mutual Antipathy; this is performed by the same wayes and means, whereby we observe one Body to Attract and hold fast another, or one Body to Repell and Avoid conjunction with another, in all Sensible and Mechanique Operations. This small Difference only allowed, that in Gross and Mechanique operations, the Attraction, or Repulsion is performed by Sensible Instruments: but, in those finer performances of Nature, called Sympathies and Antipathies, the Attraction or Repulsion is made by Subtle and Insensible.

This means that he now knows in principle how sympathy and antipathy work:

The means used in every common and Sensible Attraction and Complection of one Bodie by another, every man observes to be Hooks, Lines, or some such intermediate Instrument continued from the Attrahent to the Attracted; and in every Repulsion or Disjunction of one Bodie from another, there is used some Pole, Lever, or other Organ intercedent, or somewhat exploded or discharged from the Impellent to the Impulsed. Why therefore should we not conceive, that in every Curious and Insensible Attraction of one bodie by another, Nature makes use of certain slender Hooks, Lines, Chains, or the like intercedent Instruments, continued from the Attrahent to the Attracted, and likewise that in every Secret Repulsion or Sejunction [pushing apart], she useth certain small Goads, Poles, Levers, or the like protruding Instruments, continued from the Repellent to the Repulsed bodie? Because, albeit those Her Instruments be invisible and imperceptible; yet are we not therefore to conclude, that there are none such at all.¹²

This mechanical philosophy, as described by Charleton, would have made perfect sense to Lucretius, but he would have found the label itself deeply puzzling, because the Romans did not think of poles and hooks as machines, any more than we do – these are the simple machines of the mathematicians. But the Roman idea of a machine was neither Charleton’s nor ours. The key source for knowledge of Roman machinery is Vitruvius’s On Architecture, which describes machines used in construction and for warfare. When Vitruvius wrote about a machine (Latin: machina) he used the word to means something quite different from what we mean. Scaffolding is a machine. A scaling ladder is a machine. A tower on wheels built to enable you to approach the enemy’s walls and scale them is a machine. A platform on which spectators stand is a machine. Roman machines do not necessarily do work in the sense of moving things, nor do they necessarily have moving parts. Their common characteristic is that they are substantial structures designed to be stable. Thus a hoist with a pulley is a machine, but what makes it a machine seems to be the fact that it is solidly supported. A trebuchet is a machine, but what makes it a machine is not that it throws large rocks but that it is made out of great solid baulks of timbers lashed together. The nearest synonym to machina was fabrica, a word that can often be translated as ‘structure’. When Lucretius talks about the machine of the world (machina mundi) he does so in the context of discussing the dissolution of our universe. When our world ends its structure will come apart. The machine of the world is thus the stable structure of our universe: the heavens, the earth and the four elements. All these will disappear when the universe comes to an end and a new one is born.¹³

The phrase machina mundi is echoed by Tertullian (160–225) and Augustine (354–430), and thus appears throughout medieval philosophy (in Sacrobosco, for example), even though the text of Lucretius had been lost and was not rediscovered until 1417,ii but it does not imply an interlocking set of moving parts, a geared system or a power train. To translate it as ‘machine of the world’ is to mistranslate. The best translation into English is perhaps a phrase of John Wilkins’s of 1675 which is surely meant to be the English equivalent: ‘this visible frame which we call the World’.¹⁴

§ 2

With time, of course, the original meaning of Lucretius’s phrase was lost; as machines changed, the meaning of Lucretius’s phrase changed with them. Crucial here is the clock. One of the primary purposes of the first clocks was to model the movement of the heavens, not just to tell the time. Thus in 1364 (about sixty years after the invention of the escapement mechanism which made the mechanical clock possible) Giovanni de’ Dondi constructed in Padua an astrarium (a star-arium, or planetarium) which showed the time, the movements of the sun, moon and other planets, and the religious feast days. Part of his purpose was to prove that the Ptolemaic system was an accurate representation of how the heavens really work and not just a mathematical model.¹⁵ It was natural therefore to claim that, since a clock models the heavens, the heavens are like a clock. So far as we know, this claim was first made by Oresme in 1377, seven years after the erection of a clock on the Palais Royal in Paris: the movement of the spheres, he said, was perhaps like ‘a man making a clock and letting it go and be moved by itself’.¹⁶ Implicitly, he was saying that God might be rather like a clockmaker. Collingwood is right: ‘[I]t was an easy step to the proposition: as a clockmaker … is to a clock … so is God to Nature.’ But no medieval author compared the universe to anything so coarse as a mill; and Oresme’s comparison was very carefully limited: he was comparing the circular movement of the heavens to the turning wheels of a clock, not the whole universe to a clock; he did not think of clocks as machines and he does not use the clock metaphor to prove the existence of God. Oresme had no intention of expounding a mechanical philosophy, for he lived in a world of Platonic and Aristotelian forms; indeed, he ended up accepting the conventional view that the heavenly spheres were governed by spiritual intelligences.

Around 1550, however, commentators on Vitruvius (writing in Latin) began to express dissatisfaction with his account of what a machine is.¹⁷ They wanted to include water-wheels and clocks among the machines (for the first time) and give prominence to powered machinery. (The Greeks and Romans had very few water-wheels and no clocks, hence their lack of interest in powered machinery.) Thus the modern idea of the machine was born by giving a new meaning to the Latin term machina. This new understanding of what a machine is meant that automata – that is to say, devices that move of themselves (including clocks) – were now for the first time classified as machines.

Clocks had long had little statues that came out of the clock or moved to mark the time: the bell indicating the hour was often struck by the figure of a man with a hammer: a jacquemart, or, in English a ‘jack’. Sometimes statues of the Virgin Mary and child appeared and the three kings paraded by them; or mechanical heralds emerged and blew on trumpets. The clock inside Strasbourg Cathedral (first built in 1352–4) was the most famous example of such an elaborated clock. On its top, for example, stood a gilded cock that flapped its wings, opened its mouth, stuck its tongue out and crowed to mark midday.¹⁸ The modern cuckoo clock is a simplified version of these ‘automatic’ mechanisms. One of the most sophisticated was the repeater mechanism invented in 1676: if you pulled on a string the clock would chime the last set of hours and quarters: in other words, the clock would answer you if you asked it the time (a capacity that was primarily intended for use in the dark).

A particularly important example of the new concept of the machine is provided by The Relations of Moving Forces by the French Protestant Salomon de Caus (1615).¹⁹ De Caus was interested only in moving mechanisms, whether they were driven by air pressure (he devised a primitive steam engine), by flowing water or by descending weights. He invented, for example, a player organ: an organ which, like a player piano, automatically played music according to the information conveyed by pins on a turning drum. He constructed elaborate fountains, with grottos containing mechanical singing birds. But he also described machines for pumping water, sawing wood and performing other industrial tasks. In making ornate fountains and singing birds he was following classical precedents; but the Greeks and Romans had nothing to say about powered hammers and saws, about automata that performed mechanical tasks beyond the strength of a human being.

De Caus is important because when Descartes wrote about machines it is particularly his machines that he had in mind, and it is de Caus who transmits the new terminology from Latin into French, and through Descartes into English.²⁰ Before they read Descartes in 1637 the English had called complex machines ‘engines’, not ‘machines’.²¹ (The word ‘engine’ comes from the Latin ingenium, meaning intelligence, from which we derive ‘ingenious’; in English the word meant ‘cunning’ before it came to be used for a cunning device.) So when More invented the term ‘mechanical philosophy’ he was playing his part in importing the new terminology into English.²² ‘Engine’ and ‘machine’ still have overlapping meanings as a consequence of Descartes’ influence on the English language.

Descartes himself designed, and perhaps built, automata: he designed an improvement to the mechanism of clocks but also a tightrope walker, powered by magnets, and an installation in which a dog springs at a partridge, which flies away. There was even a story that he had made himself a woman; convinced that this living machine must be inhabited by a devil, she was thrown overboard by a ship’s captain when his ship, on which Descartes was a passenger, was caught in a storm.²³

Descartes’ remarkable and novel claim, first stated in the Discourse on Method, was that animals are automata, that is, complex, self-moving machines. They appear to have some extra quality that we call ‘life’, or ‘intelligence’ but are in fact simply performing predetermined routines, like the cock on the Strasbourg Cathedral clock. The soul, Descartes claimed, is unique to rational human beings; animals have no soul and no capacity to reason. (The Aristotelians had distinguished three different types of soul – vegetative, animal and rational – and so had no problem in acknowledging that animals had a sort of soul.) When Descartes describes something in nature as a machine it is biological entities which he always has in mind. He denied that animals were designed; but they do move (like de Caus’s machines), and they do reproduce themselves, so, once they have come into existence, their complex structures do not have to be put together again from scratch.

If animals are machines and nothing but machines, then for Cartesians it must follow that the human body, which is evidently similar to the body of an ape, works like a machine, and Cartesian doctors were eager to study human anatomy as an example not of a geared mechanical system but of a hydraulic system of the sort that powered de Caus’s fountains and player organs. If the human body is a machine it must have a power source, which is perhaps why Descartes wanted to think of the heart as a heat engine rather than as a pump (de Caus had powered fountains by heating water by shining sunlight through lenses); to describe it as a pump simply begs the question of what powers the pump. But, of course, once animals have been claimed to be machines, it is a small step to arguing that human beings are also machines, and so to the adoption of a systematic materialism of a sort which would have been anathema to Gassendi, Descartes, Boyle and Newton. Julien Offray de La Mettrie’s Man the Machine (1748) is a logical development of this sort of uncompromising mechanistic thinking.²⁴ The challenge set by Descartes was, of course, that of building an automaton that could behave like an animal. A hundred years later Jacques de Vaucanson (1709–82) made a mechanical duck which could walk, quack, eat and defecate.²⁵

Descartes does not think of the universe as being like a clock because in his view outer space is filled not with the crystal spheres of Ptolemaic astronomy, nor with the gears and levers of de Caus’s machines, but with liquid vortices which carry the planets in their orbits around the stars.²⁶ However, he does say that understanding the universe is comparable to the problem of understanding a clock. If you look at a clock or a watch from the outside you can tell that there is a mechanism which turns the hands. You may conclude that the hands are driven around by a descending weight. But they might equally be driven by a spring (or be regulated by a pendulum – but Descartes died before the invention of the pendulum clock). You can only tell exactly what is going on if you can take the clock apart.²⁷ Descartes thinks that much of our understanding of nature is like this: we can come up with a convincing account of how things might work, but we cannot know for sure that that is how they really work because the mechanism is invisible to us; hidden not because it is hidden inside a box but because it is too small to see. The initial hope was that the microscope might make the invisible visible, and it did, for example, when it showed how a fly can walk up a pane of glass. But it could not show the mechanism which causes light to reflect or refract, or the particles that cause smells.²⁸ Descartes thought that sometimes experiments could be constructed to enable a secure choice between possibilities (thus you can experiment with a spherical bottle filled with water and show how a rainbow is produced), but this might not always be possible: in Descartes’ view, Pascal’s vacuum experiments did not serve to eliminate the possibility of a plenum. The clock metaphor is thus used by Descartes to make an epistemological argument about the limits of our understanding, rather than as an analogy as to how the universe actually works.iiii

An undated eighteenth-century handbill announcing an exhibition of three of Vaucanson’s automata: the flute player, the drummer and the digesting duck. The inner workings of the duck are unknown, despite various attempts to construct replicas.

Once a clock has started, it performs its tasks automatically, but it cannot regulate itself. If it is running fast, it cannot slow itself down, nor catch up if it is running slow. One of de Caus’s machines, however, has a sophisticated feedback mechanism.²⁹ The machine is designed to use the weight of water to lift half a tank of water while the other half balances it by going downwards. It consists of three tanks on different levels. In order to work the machine two valves have to be closed when the lower tank is full; this is done by having an overflow which fills a hopper; when the hopper empties, a weight closes the valves. Self-regulating machines were very rare in the seventeenth century (a few years later Cornelis Drebbel designed an incubator for chicken’s eggs with a thermostat to regulate the temperature) and de Caus’s may have been the first to be invented since Hero of Alexandria designed the float regulator, which we still use in water tanks and toilet cisterns.³⁰ They become common only in the late eighteenth century: methods for automatically turning valves on and off at the right moment would later prove crucial to the workings of the steam engine (the fan-wheel which turns a windmill into the wind is another simple example). The self-regulating machine is not just a crucial step towards a whole series of more advanced technologies. It is the founding concept of modern social science: David Hume’s theory of the balance of trade and Adam Smith’s conception of the market depend on the concept of a feedback mechanism.³¹ There is an old debate about why the Greeks and the Romans failed to develop a general theory of economic behaviour; one good answer is that they had no machines with feedback mechanisms, so they lacked an essential tool for thinking about social processes.³²

§ 3

So classical atomism, as revised and reconstructed in the seventeenth century by Gassendi, Descartes and others, explained nature in terms of interacting particles. After 1659 the English usually called this ‘the mechanical philosophy’, although Boyle was soon to introduce the much less confusing terms ‘the corpuscularian philosophy’ and ‘the corpuscular philosophy’. (Eventually, the term ‘mechanical philosophy’ was exported into French, spreading confusion even among Cartesians.) Atoms and corpuscles are not machines, but their interaction is determined by size, shape and hardness, just as the interaction of the parts of a clock are. The disciples of Gassendi (such as Charleton in England) and of Descartes (such as Henry Power) disagreed about lots of things, but they agreed that one should privilege corpuscular explanations for natural processes. Boyle and Newton followed them in this, although they did not insist that everything was susceptible to corpuscular explanation (indeed, Newton’s theory of gravity proved to be the great exception, destroying in the end the corpuscular philosophy out of which it had been born).

De Caus’s self-regulating machine for raising water, from La Raison des forces mouvantes (1615).

A quite separate argument was the claim that the universe has been designed to serve a purpose, like a clock or any other complex machine, and that it therefore demonstrates the existence of God. Descartes does not use this argument. His universe can barely be described as ‘designed’, being the result of allowing very basic laws to work themselves out in practice, and it is not so much mechanical as fluid: it involves vortices and other liquid flows, not cogs and gears. It has been claimed that the modern argument from design first appears in John Wilkins, one of the founders of the Royal Society, in Of the Principles and Duties of Natural Religion, published posthumously in 1675.³³ According to that argument, we can tell that the universe has a creator because only an external agent could have designed and constructed it so that its different parts serve their specific functions. An Aristotelian could never argue like this: Aristotle himself did not believe the universe had a creator, and his medieval successors thought that purposiveness was part of the very fabric of nature.iiiiii Wilkins, however, is not the first to use this argument, for it can be found in Henry More in 1668, and before More in an early-seventeenth-century Dutch Jesuit with an interest in probability, Leonardus Lessius, in 1631.³⁴

Tracing the argument back through these texts, it becomes apparent that it is a variation on a much older argument, which is sometimes called the infinite monkey theorem – the claim that a monkey randomly tapping keys on a typewriter for all eternity would eventually produce the works of Shakespeare. A refutation of this argument, in essence, can be found in Cicero, although of course without either typewriters or Shakespeare. Cicero rejected the claim of the atomists that the universe is the product of chance, saying:

I cannot understand why he who considers it possible for this to have occurred should not also think that, if a countless number of copies of the one-and-twenty letters of the alphabet, made of gold or what you will, were thrown together into some receptacle and then shaken out on to the ground, it would be possible that they should produce the Annals of Ennius, all ready for the reader. I doubt whether chance could possibly succeed in producing even a single verse!³⁵

Similarly, Lessius and his successors say that if you throw a quantity of bricks down on the ground you will never get a palace. A book requires an author; a palace requires an architect; a clock requires a clockmaker; and the universe requires a creator. We have already seen Kepler suggesting that even a salad requires a cook.³⁶ (This argument will, of course, later be disputed by Hume and by Darwin, but for a long time it seemed almost irresistible.)

Boyle has his own version of this argument, which he uses not against the atomists but against the scholastics. They, he maintains, do not have an adequate conception of an omnipotent god. Indeed, he may well have had Aquinas in mind:

[T]he difference betwixt their Opinion of God’s Agency in the World, and that which I would propose, may be somewhat adumbrated by saying, That they seem to imagine the World to be after the nature of a Puppet, whose Contrivance indeed may be very Artificial, but yet is such, that almost every particular motion the Artificer is fain (by drawing sometimes one Wire or String, sometimes another) to guide, and oftentimes over-rule, the Actions of the Engine; whereas, according to us, ’tis like a rare Clock, such as may be that at Strasbourg, where all things are so skillfully contriv’d, that the Engine being once set a Moving, all things proceed, according to the Artificers first design, and the Motions of the little Statues, that at such hours perform these or those things, do not require, like those of Puppets, the peculiar interposing of the Artificer, or any Intelligent Agent employed by him, but perform their functions upon particular occasions, by vertue of the General and Primitive Contrivance of the whole Engine.³⁷

You cannot substitute a text or a palace for the clock in Boyle’s version of the argument (as you can in Wilkins’s) because texts and palaces are static while clocks move. His argument is that it is astonishing that in our universe everything carries on according to the same general laws without intervention being necessary to correct faults in the machinery, without an archer directing the arrow. To mount this argument he needs not just a complex contrivance but one which is in continuous movement. Only a clock will do: indeed, only a pendulum clock, since earlier clocks were constantly in need of adjustment. In that respect, Boyle’s argument is a new and distinctively mechanical one.

But the legacy of the mechanical philosophy was not simply modern versions of the argument from design, which are still widely defended in the form of Intelligent Design. The future lay not only with the new philosophies, whether mechanical or Newtonian; it also lay with the new machines. De Caus in 1615 played with a very simple steam engine, and working steam engines require simple feedback mechanisms. Vaucanson not only made a mechanical duck, he also devised a machine that would automatically weave brocades. Friedrich von Knauss (1724–89) invented a mechanical hand that wrote on a piece of paper just like a living hand; he also constructed the first typewriter.³⁸ The Industrial Revolution would depend on the skills of such men, skills that would have been familiar to the craftsmen who built the first Strasbourg Cathedral clock. The Scientific Revolution started as a revolution of the mathematicians; it would eventually turn into a revolution of the mechanics. There is a direct line of descent from the Strasbourg clock to the spinning jenny.

This brings us back to the problem with which we began. The Strasbourg clock was built in the middle of the fourteenth century – but the mechanical philosophy was invented three centuries later. Machines did not change much in the meantime, but philosophers did. Once Lucretius was available (he was rediscovered in 1417), his concept of the machina mundi could be turned into a quite new idea, the idea of a clockwork universe. In order for this to happen, however, the text of Lucretius was not enough. What was needed was not just new machines but also a new language for discussing machinery. Before this new language, clocks could be used to understand the heavens but not terrestrial physics or biology. It was engineers such as de Caus who, by generalizing the concept of a moving mechanism, made the clockwork universe and the mechanical man possible.

Geography had been remade at the beginning of the sixteenth century by mariners; the philosophy of nature was remade in the seventeenth by the ‘mathematicians and engineers’.³⁹ Natural philosophy was no longer an enterprise to be conducted simply with pen and paper. Boyle’s air pump and Huygens’ pendulum clock were philosophical machines – machines made by philosophers (with the assistance, of course, of technicians) – the first to tackle a scientific problem, and the second to embody a scientific theory. They helped transform the way in which philosophers thought about machinery, just as Descartes’ obsession with automata resulted in a new mechanical philosophy. Already in the seventeenth century the mathematicians’ revolution was becoming indistinguishable from a mechanical revolution. Collingwood’s claim that the Industrial Revolution was ‘well on the way’ by the sixteenth century seems to me misconceived, for no new power sources had been brought to bear, but in Chapter 14 I will argue that it was indeed well under way by the end of the seventeenth century, thanks to the appearance of a new type of expert, the engineer-scientist.

§ 4

It will now be apparent that Descartes and Boyle both have what we may term mechanical philosophies, but that they are very different. Of the core three arguments we have distinguished – the corpuscular philosophy, animals as automatons and the clockwork universe – they agree on the first, but each picks one and only one of the other two. Animal automata lead to atheism if humans are held to be little different from animals, but not if one can prove (as Descartes thought he could) the existence of an immaterial mind. The corpuscular philosophy leads to atheism if it is combined with the claim that the universe arises from chance, but not if this further step is blocked, as Boyle sought to block it, by the argument from design. Descartes and Boyle are confident they can protect themselves from atheism, the first by distinguishing mind from matter, and the second by regarding the natural world as providing proof of God’s design.⁴⁰ Boyle’s argument was to prove rather robust, and he was to inspire a long tradition of Christian theologians, such as William Paley (1743–1805); until Darwin, there was no good answer to it, though Hume did his best to blunt it in his posthumous Dialogues Concerning Natural Religion (1779). Descartes’ argument did not fare so well; even Locke thought that there might be such a thing as thinking matter.⁴¹ Newton proved to be right to ask:

If we say with Descartes that extension is body, do we not manifestly offer a path to Atheism, both because extension is not created but has existed eternally, and because we have an absolute idea of it without any relationship to God, and so in some circumstances it would be possible for us to conceive of extension while imagining the non-existence of God? Nor is the distinction between mind and body in this [Cartesian] philosophy intelligible, unless at the same time we say that mind has no extension at all, and so is not substantially present in any extension, that is, exists nowhere; which seems the same as denying the existence of mind … And hence it is not surprising that Atheists arise ascribing that to corporeal substances which solely belongs to the divine.⁴²

Many eighteenth-century atheists, such as d’Holbach and Diderot, were to take their inspiration from Descartes’ mechanism and turn it into a systematic materialism with no room for God.

The corpuscular philosophy was absolutely crucial to the Scientific Revolution in that it provided an alternative to the Aristotelian doctrine of forms or substances, of immaterial essences; consequently, it excluded teleology from the nature of things.⁴³ It was helpful for the generation of the theoretical models which explained air pressure and the vacuum, even if those models were unacceptable to Descartes. But the Newtonian revolution (which we will come to in Chapter 13) ensured that it quickly ceased to be a core part of science as that concept was inherited in the eighteenth and nineteenth centuries. Modern physics, chemistry and biology do not emerge out of the corpuscular philosophy, but out of its collapse. The corpuscular philosophy is, in the end, a parenthesis between scholasticism and Newtonianism.

If Newtonianism destroyed the corpuscular philosophy, it greatly strengthened the argument from design. Only an omnipotent God, creating laws of nature and ensuring that they constantly took effect in the world, could explain the working of gravity, since gravity could not be explained in either Aristotelian or corpuscular terms. Newton’s God did not direct individual arrows to their targets; he established the laws that determined the flight of each and every arrow. Newtonianism was thus only conceivable within a culture which had elaborated and become dependent upon the argument from design. That culture was peculiarly English for, as we have seen, Descartes studiously avoided appeals to design. In this respect, Newton is Boyle’s heir, and exporting Newtonianism to the Continent depended not just on persuading scientists abroad to accept the possibility of action over a distance but also on persuading them to adopt the design argument.

Voltaire said in 1733 that the English and the French lived in two different worlds, the world of Newton and the world of Descartes:

A Frenchman, who arrives in London, will find Philosophy, like every Thing else, very much chang’d there. He had left the World a plenum, and he now finds it a vacuum. At Paris the Universe is seen, compos’d of Vortices of subtile Matter; but nothing like it is seen in London. In France ’tis the Pressure of the Moon that causes the Tides; but in England ’tis the sea that Gravitates towards the Moon … The very Essence of Things is totally chang’d. You neither are agreed upon the Definition of the Soul, nor on that of Matter … How furiously contradictory are these Opinions!⁴⁴

Voltaire serves to remind us that there was more than one world in which scientists could live.

But in both the world of Descartes and the world of Newton the laws of nature were inexorable, and human beings inhabited a universe which, far from reflecting back to them an image of themselves, macrocosm to microcosm, appeared utterly indifferent to their existence. For both, the sun was just one star among an uncountable multitude, and the universe was, if not infinite, at least without any known limit. ‘When I consider the short duration of my life, swallowed up in the eternity before and after, the little space which I fill, and even can see, engulfed in the infinite immensity of spaces of which I am ignorant, and which know me not, I am frightened … The eternal silence of these infinite spaces frightens me,’ wrote Pascal, who had helped bring this new world into being.⁴⁵ Meanings inscribed by God in the forms of things, the great chain of being, sympathy and antipathy, natural magic, had been replaced by blind mechanisms and inexorable laws. Even animals, in Descartes’ view, were just automata. Blake painted Newton playing at God, measuring out the universe; preoccupied with the simplified, mathematical language of the laws of nature, he can no longer see the complexity and variety that surround him; quality has been reduced to mere quantity. This was the beginning of what Weber called ‘the disenchantment of the world’.⁴⁶