Early modern philosophy and science are often said to have put a “mechanical” or “mechanistic” conception of nature at the center of Western thought.  Robert Boyle referred to it as “the mechanical philosophy.”  Historian of science E. J. Dijksterhuis characterized it as a “mechanization of the world picture.”  Tim Crane calls it “the mechanical world picture.”  But what does a mechanical or mechanistic conception of the world amount to? 

Dijksterhuis’s book The Mechanization of the World Picture surveys the history of the period during which this conception rose to hegemony, and in the Epilogue he considers several possible interpretations that the survey suggests.  First, it is commonly said that a mechanistic conception of the natural world is one which sees it as a kind of machine, analogous to a clock.  And such metaphors are, he says, indeed frequent in writers of the period.

However, Dijksterhuis does not think this actually captures what is essential to the mechanical world picture.  The reason is that the notion of a machine implies teleology.  For example, what makes a clock a clock is that it serves the function of telling time.  It is true that this suggests a model of teleology that differs from the one associated with Aristotelianism, according to which natural teleology is intrinsic to a substance – in contrast to the teleology possessed by human artifacts, which is imposed from outside by their designers and users.  (I have discussed the various conceptions of teleology in many places, such as this article.)  All the same, the early modern mechanical philosophy was inspired by ancient atomism, which eschewed teleological explanation.  And as the mechanical world picture developed, teleological explanations came to be seen as scientifically deficient.  So, Dijksterhuis argues, the machine model does not really capture what is essential to it.

A second possible interpretation of a mechanical or mechanistic mode of explanation sees it as essentially concerned with discovering the hidden mechanisms underlying natural phenomena.  Think of the way that we come to understand how an automobile engine works when we see how the burning of fuel creates small explosions that move pistons, which in turn move the crankshaft, and so on.  Of course, such an engine is, like a clock, a human artifact with its own distinctive teleology.  But the teleology is not what is doing the explanatory work on this second interpretation of what a mechanical explanation amounts to.  Rather, the idea is that what illuminates our understanding is coming to see how the behavior of the whole results from the arrangement and interaction of the parts as they push and pull against one another.

The problem with this interpretation, Dijksterhuis points out, is that not all explanations that came to be regarded as mechanistic actually work this way.  For example, Newton was never able to identify a mechanism by which gravity worked.  And while some at the time were critical of his account of gravity precisely for that reason, eventually it came to be regarded as a paradigm of successful mechanistic explanation. 

A third possible interpretation considered by Dijksterhuis holds that a mechanistic conception of nature is essentially “anti-animistic” in character, in the sense that it rejects any explanation of a thing’s behavior in terms of some principle internal to it.  Contrast this with Aristotle’s view in the Physics that what is natural to a thing is precisely what does follow from an internal principle.  A mechanistic explanation, on this interpretation, is one that explains a thing’s behavior in terms of external factors (whether something pushing or pulling on it, the laws that govern it, or what have you). 

The trouble with this interpretation, in Dijksterhuis’s view, is that there are mechanistic explanations that do not operate this way, and Aristotelian explanations that do operate this way.  For example, inertial motion seems to be motion that springs from an internal principle, whereas Aristotelian theories of projectile motion appealed to an external principle.  Hence, this cannot be the key to what sets mechanistic explanations apart from the Aristotelian explanations they were meant to replace.

A fourth interpretation takes mechanistic explanation to be explanation modeled on mechanics, in the sense in which that term came to be used in modern physics.  And mechanics in that sense had to do with explaining local motion in terms of a mathematical description of the relations between objects and the laws that govern them.  Newtonian mechanics became the paradigm, and twentieth-century physics took the mathematical approach further still.  This is the interpretation Dijksterhuis endorses, so that he takes the mechanization of the world picture to amount ultimately to a mathematization of the world picture.

In his article on “Mechanistic explanation” in The Cambridge Dictionary of Philosophy, philosopher of science David Hull rejects this sort of interpretation, on the grounds that many physical phenomena cannot be explained in terms of mechanics in this sense, but are still thought to be in some sense explicable mechanistically.  Hull writes: “Historically, explanations were designated as mechanistic to indicate that they included no reference to final causes or vital forces.  In this weak sense, all present-day scientific explanations are mechanistic” (p. 476 of the first edition).

Crane’s account of the mechanical world picture in his book The Mechanical Mind emphasizes both the rejection of final causes and explanation in terms of mathematically formulated laws of nature.  But he also says that pre-modern philosophy and science were committed to an “organic” conception of nature, whereas mechanical explanations reject this.  What he means is more or less what Dijksterhuis describes in the third, “anti-animistic” interpretation of the mechanical philosophy.  To common sense, a living thing seems to operate according to an inner principle, which directs it toward a natural end or final cause.  Pre-modern Aristotelian thought modeled inorganic substances too on these aspects of living things.  By contrast, the mechanical world picture starts with inorganic things, understood as devoid of any such internal principles or final causes.  It then seeks to model all of nature, including living things, on this way of conceiving of the inorganic.

What should we think of all this?  On the one hand, all of these suggestions do indeed capture themes that have historically been associated with the mechanical world picture, even if some of them are more central to it than others are.  On the other hand, it is not surprising that it should prove difficult to find a precise and generally accepted definition.  The reason, I would suggest, is that from the beginning, the aim and content of the mechanical world picture has been more negative than positive.  It is more about what it is against than what it is for.  And what it is against is Aristotelianism.

In particular, I would suggest, the idea of giving mechanistic explanations of natural phenomena was always fundamentally about finding ways to make them intelligible without reference to anything like substantial forms, intrinsic teleology, and related ideas in Aristotelian philosophy of nature.  The point was to replace this with an approach that was simpler and allowed for the prediction, control, and technological exploitation of natural phenomena.  The other themes described above were natural consequences of this. 

For example, machines have only accidental forms rather than substantial forms, and their teleology or final causes are imposed from outside rather than intrinsic.  Their behavior is also typically predictable and controllable, and, naturally, they exist in order to serve the interests for which we develop technology.  If you want a non-Aristotelian conception of natural objects that will be useful for the practical purposes that interested early modern scientists and philosophers, the machine is a natural model.  Into the bargain, it provided an alternative way to think of the world as dependent on God – hence Aristotle’s notion of a prime unmoved mover gets replaced by the notion of a divine machinist – which was important to thinkers who, for all their hostility to Scholasticism, were still mostly theists (but became less important as atheism became more common among the intelligentsia).

Since we understand how machines work by determining what sorts of parts they are made of and how those parts are arranged and interact with one another, it was only natural that the mechanical world picture would initially not only think of natural phenomena on the model of machines, but also think of physical explanation as a matter of identifying hidden parts and mechanisms.  Since the machines with which early modern thinkers were familiar typically operated by means of parts pushing and pulling on one another, it was natural that this crude sort of causation would for a time also be a standard part of the mechanical mode of explanation.  Since the teleology of machines and the arrangements of their parts come from outside them – from the designers and users of machines – it is only natural that the mechanical world picture would initially think of natural phenomena as operating according to externally imposed influences (motion imparted from outside, laws of nature, or what have you) rather than something intrinsic to them.

Furthermore, since the most impressive technologies are those that work with mathematical precision (such as a well-made clock), it was also only natural that the mechanical world picture would come to favor mathematical models of mechanisms and mechanical processes.  And this approach would indeed go on to have spectacular success in allowing for the prediction, control, and technological exploitation of natural processes.

In short, the initial motivation for the mechanical world picture was to develop a non-Aristotelian conception of nature that would facilitate certain practical goals; the machine model was the most promising candidate; that model in turn suggested several further themes, some of which did not last; but the theme of mathematization proved the most fruitful and thus survived and became the model for how to carry out mechanistic explanation.  For this reason, as the mechanistic conception of the world developed, it essentially came to be about eschewing anything that smacked of final causality and substantial forms in favor of a mathematical mode of explanation that allowed for strict prediction, control, and technological application – with push-pull causation, the identification of hidden mechanisms, exclusive appeal to external principles of operation, and even the machine analogy itself eventually dropping away as inessential to the mechanical world picture.