Peter Anstey writes …
What is the precise relation between experimental philosophy and mechanical philosophy in the seventeenth century? In my last post I showed how neither Henry More nor Henry Stubbe were particularly clear about this. In this post I examine the view of Robert Boyle.
Boyle is sometimes credited with coining the English term ‘mechanical philosophy’* and he was certainly the first person to use the term ‘experimental philosophy’ in a book title. In 1663 he published Of the Usefulness of Experimental Philosophy which was soon followed by Henry Power’s Experimental Philosophy of 1664.
If we look at frequencies of use in Boyle’s writings, it turns out that he used the term ‘experimental philosophy’ roughly twice as often as ‘mechanical philosophy’ or ‘mechanical hypothesis’. This raw fact is in itself rather telling for those recent historiographical debates over the nature and status of mechanical philosophy in early modern philosophy that almost entirely ignore experimental philosophy. However, the key question is: Were the terms synonyms for Boyle or did they denote two different things?
The best early statement of Boyle’s view of the content of experimental philosophy is in the ‘Proemial Essay’ to Certain Physiological Essays first published in 1661. He starts with a criticism of previous natural philosophers such as Aristotle and Campanella:
they have too hastily, and either upon a few Observations, or at least without a competent number of Experiments, presum’d to establish Principles, and deliver Axioms. (Works of Robert Boyle, 1999–2000, 2: 13)
What experimental philosophers should do instead is:
set themselves diligently and industriously to make Experiments and collect Observations, without being over-forward to establish Principles and Axioms, believing it uneasie to erect such Theories as are capable to explicate all the Phaenomena of Nature, before they have been able to take notice of the tenth part of those Phaenomena that are to be explicated. (Works of Robert Boyle, 2: 14)
This clearly has to do with the role of observation and experiment in relation to theory in the acquisition of knowledge about nature. Now let’s see how Boyle defines the mechanical philosophy. In The Excellency and Grounds of the Mechanical Hypothesis (aka the mechanical or corpuscular philosophy) Boyle states the kernel of the view as follows:
the Universe being once fram’d by God, and the Laws of Motion being setled and all upheld by His incessant concourse and general Providence; the Phænomena of the World thus constituted, are Physically produc’d by the Mechanical affections of the parts of Matter, and what they operate upon one another according to Mechanical Laws. (Boyle Works, 8: 104)
The mechanical affections referred to here are the shape, size, motion and texture of corporeal bodies.
Now this is really quite different from experimental philosophy. For, it is the sort of theory that one should arrive at as a result of practising experimental philosophy. This is why Boyle’s book The Origin of Forms and Qualities has a ‘speculative part’, which outlines the theoretical content of the mechanical philosophy, and a ‘historical (or experimental) part’, which provides experimental support for the speculative theory. Here is how he describes the relation between the two parts:
it was very much wish’d, that the Doctrines of the new Philosophy (as tis call’d) [i.e. mechanical philosophy] were back’d by particular Experiments; the want of which I have endeavour’d to supply, by annexing some, whose Nature and Novelty I am made believe will render them as well Acceptable as Instructive.
Thus, for Boyle, experimental philosophy and mechanical philosophy are entirely distinct: the former provides the evidential grounds of the latter. This is why, as Dmitri Levitin has shown, Boyle prefers Democritus to Epicurus. In Boyle’s view, the former based his atomism on experimental philosophy, the latter on speculative philosophy. (Levitin, ‘The experimentalist as humanist: Robert Boyle on the history of philosophy’, Annals of Science, 71, 2014, 149–82).
It may be that some philosophers and even natural philosophers conflated experimental philosophy with mechanical philosophy, but in Boyle’s mind they were distinct.
* Actually, the question turns out to be slightly more complicated than it looks because Henry More used the term ‘mechanical hypothesis’ in 1653 (An Antidote against Atheism, 44) and when Boyle first introduces the term in 1661 in Certain Physiological Essays, he uses ‘Mechanical Hypothesis or Philosophy’ (Boyle Works, 2: 87).
Peter Anstey writes …
The mechanical philosophy, at least since the work of Marie Boas Hall and E. J. Dijksterhuis, has played a prominent role in the historiography of early modern natural philosophy. By contrast, experimental philosophy has been largely absent. Take, for example, Richard Westfall’s The Construction of Modern Science (Wiley, 1971). It has a whole chapter dedicated to the mechanical philosophy whereas the term ‘experimental philosophy’ appears only once in the entire book –– on the penultimate page –– and this is in a quote from Newton’s ‘General Scholium’ in a discussion of Newton’s concept of force and the term’s presence is irrelevant to Westfall’s narrative. It is also rather telling that the term ‘experimental philosophy’ does not even appear in the 662 pages of Floris Cohen’s The Scientific Revolution: A Historiographical Inquiry (Chicago, 1994).
It is interesting to reflect, therefore, that the English terms ‘experimental philosophy’ and ‘mechanical philosophy’ came into common use around about the same time, in the late 1650s. Moreover, when the new experimental philosophy emerged in England in the 1660s it was frequently associated with and even conflated with the mechanical philosophy. (Experimental philosophy was also commonly identified with corpuscular philosophy, though this is not our concern here.) Robert Hooke famously spoke of ‘the real, the mechanical, the experimental Philosophy’ in the Preface to Micrographia of 1665. By the end of the seventeenth century, however, the two had come to be fairly clearly demarcated. The Newtonian John Keill, for instance, lists four ‘sects’ of his day, two of which are the experimental philosophers and the mechanical philosophers (Introductio ad verum physicam, Oxford, 1702, p. 2).
The process by which this ‘decoupling’ occurred is quite convoluted and this is the first in a series of posts that will attempt to set out some points of reference from which we can understand how experimental philosophy and mechanical philosophy came to be clearly demarcated.
Let us begin with two Henrys, Henry More and Henry Stubbe. More was not a practitioner of experimental philosophy: in fact, he was not a natural philosopher at all. He was, however, a Fellow of the Royal Society. Stubbe was a physician and critic of the Royal Society and experimental philosophy. Now when Henry Stubbe attacked the Royal Society, and in particular its apologist Joseph Glanvill, he claimed that Henry More had given up his association with the Society because of the Society’s commitment to the mechanical philosophy which tended to atheism (Stubbe, Legends no Histories, London, 1670, p. 173).
More responded to Stubbe’s claims in a letter to Glanvill (c. 1671):
he [Stubbe] looks upon that Mechanick Philosophy which I oppose, to be the Philosophy the Royal Society doth profess, or would support. But the Philosophy which they aim at, is a more perfect Philosophy, as yet to be raised out of faithful and skilful Experiments in Nature, which is so far from tending to Atheism, that I am confident, it will utterly rout it and the Mechanical Philosophy at once, in that sense which I oppose, namely, as it signifies a Philosophy that professeth, That Matter having such a Quantity of Motion as it has, would contribute it self into all those Phaenomena we see in Nature. (Glanvill, A Praefatory Answer to Mr Henry Stubbe, p. 155)
More opposes a mechanical philosophy that is competent to explain everything and leaves no place for a deity. But this does not mean that he opposes mechanical explanations tout court. Alluding to a passage that Stubbe quotes from Thomas Sprat’s History of the Royal-Society of London he says:
I believe indeed most of us, I am sure my self does conceive, that Generation, Corruption, Alteration and all the Vicissitudes of corporeal Nature are nothing else but Unions and Dissolutions … of little Bodies or Particles of differing Figures, Magnitudes, and Velocities. But this thus bounded is not the Mechanical Philosophy, but part of the old Pythagorick or Mosaick Philosophy … (p. 156)
More is happy to acquiesce in corpuscular explanations, so long as their limitations are recognised. He goes on:
I think it is plain, what Mechanical Philosophy that is, that may incline Men to Atheism, and that is not the experimental Philosophy, which the Royal Society professes. (p. 157)
Clearly More accepts both the corpuscular explanations of a mitigated form of mechanism and experimental philosophy. Just how he conceives the relation between the two, however, is not clear from this letter.
When we turn to Stubbe we find a similar lack of differentiation. For example, earlier in Legends no Histories, Stubbe claims that no prince has ever been called great because he used ‘any knick-knacks of Experimental or Mechanical Philosophy alone’ (p. 4).
What these passages show is that for some writers the relation between mechanical philosophy and experimental philosophy was not clearly defined. They also illustrate how tempting it would be for those scholars who view the emergence of modern science through the lens of mechanism to reduce experimental philosophy to mechanical philosophy.
It may even be that part of the explanation of the relative neglect of experimental philosophy in the historiography of early modern natural philosophy is the tendency to conflate it with mechanical philosophy. In my next post I shall examine Robert Boyle’s view of the relation between mechanical and experimental philosophy.
Kirsten Walsh writes…
- the apparent continuity between Newton’s usage [of the term ‘experimental philosophy’] and that of the early Royal Society is, however, largely an illusion.
To support this claim, Shapiro argues that, whereas ‘experimental philosophy’ was used as a synonym for ‘mechanical philosophy’ by the early Royal Society, for Newton, the two terms had different meanings. This is demonstrated by the fact that Newton adopted the experimental philosophy, but not the mechanical philosophy.
Shapiro explains that the mechanical philosophy is characterised by adherence to some or all of the following theses:
- the world and its components behave like a machine; or, more strongly, the world can be described solely by the mathematical laws of mechanics; all causation is by contact action so that the immaterial, spiritual agents are banished; matter is composed of invisible corpuscles; and hypotheses about the properties and motions of these invisible corpuscles may be formulated to explain visible effects.
Here Shapiro is conflating mechanism and corpuscularianism. However, Peter Anstey explains in his recent book, John Locke and Natural Philosophy, that these are distinct (but related) philosophies. The leading idea of the mechanical philosophy is that natural phenomena should be explained by analogy with the functioning of machines. The corpuscularian philosophy is primarily a philosophy about the underlying nature of matter, whereby explanations of natural phenomena are constrained by appeal to the invisible corpuscles which constitute all material bodies. Thus, the former is a theory of explanation; the latter, a theory of matter. There is a significant amount of overlap between the mechanical and corpuscularian philosophies, for example the focus on shape, size, motion and texture. But, they are not interchangeable. For example, Anstey points out that it wasn’t the case that everyone who held a corpuscularian theory of matter was a mechanical philosopher.
In contrast, the experimental philosophy emphasises that we can only acquire knowledge of nature by first accumulating observations and experiments and then turning to theory and hypotheses. Thus, the experimental philosophy is a theory of method, which can be viewed as placing epistemic constraints on philosophical endeavours, as opposed to the explanatory constraints of the mechanical philosophy, or the ontological constraints of the corpuscularian philosophy. So, at least notionally, these are three distinct philosophical positions.
Shapiro argues that, in practice, the early Royal Society didn’t distinguish between these philosophical positions. As evidence, he cites a passage from the preface to Robert Hooke’s Micrographia in which Hooke runs together “the real, the mechanical, the experimental philosophy”. But if we look at Hooke’s other work for uses of the term ‘mechanical’, we find that he can and does distinguish the mechanical from the experimental.
When Hooke explicitly discusses experimental philosophy, he emphasises the importance of constructing natural histories. For example, in his ‘General Scheme’, where he sets out his “Method of Improving Natural Philosophy”, Hooke explains that the best way to proceed is according to the Baconian method of natural history. He says there are three “ways of discovering the Properties and Powers [of bodies]”:
- I. By the Help of the Naked Senses.
- II. By the Senses assisted with Instruments, and arm’d with Engines.
- III. By Induction, or comparing the collected Observations, by the two preceding Helps, and ratiocinating from them.
When he discusses III, Hooke explains that an understanding of mathematics and mechanics “will most assist the Mind in making, examining, and ratiocinating from Experiments”:
- Mechanicks also being partly Physical, and partly Mathematical, do bring the Mind more closely to the business it designs, and shews it a Pattern of Demonstration, in Physical Operations, manifests the possible Ways, how Powers may act in the moving resisting Bodies: Gives a Scheme of the Laws and Rules of Motion, and as it were enters the Mind into a Method of accurate and demonstrative Inquiry and Examination of Physical Operations. For though the Operations of Nature are more secret and abstruse; and hid from our discerning, or discovering of them, than those more gross and obvious ones of Engines, yet it seems most probable, by the Effects and Circumstances; that most of them may be as capable of Demonstration and Reduction to a certain Rule, as the Operations of Mechanicks or Arts.
Later in the same discussion, Hooke enumerates the different kinds of observations one should make when constructing natural histories:
- 25ly, To enquire and try how many Mechanical Ways there may be of working on, or altering the Proprieties of several Bodies; such as hammering, pounding, grinding, rowling, steeping, soaking, dissolving, heating, burning, freezing, melting, &c.
Hooke is using the term ‘mechanical’ in (at least) two different senses. In the first sense, the term describes the processes of machines; in the second sense, the term describes manual work. But he conflates neither of these with the experimental philosophy. They are distinct, albeit related, philosophies.
Previously on this blog we have claimed that some features of Newton’s early methodology, for example his early use of queries, suggest that he was influenced by the new experimental philosophy of the early Royal Society. I do not claim that Newton’s experimental philosophy is continuous with the experimental philosophy of the early Royal Society, so I do not take issue with Shapiro’s main claim. But I do take issue with his claim that the ‘mechanical philosophy’ and ‘experimental philosophy’ were considered by the early Royal Society to be synonymous.