Defining early modern experimental philosophy (3): Some clarifications
Alberto Vanzo writes…
This is the third post in a series on defining early modern philosophy. In my earlier posts (1, 2), I claimed that
[E] To endorse the method of (early modern) experimental philosophy is to believe that one should only firmly commit to those substantive claims and theories that are warranted by experiments and observations.
In this post, I will clarify three related points in order to address some misunderstandings concerning the nature and the usefulness of the notion of experimental philosophy. I will make three claims. First, experimental philosophers did not need to eschew any theories altogether. Second, this does not entail that even Descartes, Leibniz, or some Scholastics were experimental philosophers. Third, the difficulty in classifying certain authors as experimental philosophers is not as worrying as it is sometimes portrayed.
Experiments and theories
As it should be clear from [E], one need not take on a fully a-theoretical attitude in order to endorse the method of experimental philosophy. The Proem of the Saggi di naturali esperienze, alongside other texts by experimental philosophers, states that the sole purpose of the Accademia del Cimento is “experimenting and narrating”, while eschewing any “hint of anything speculative”. Nevertheless, experimental philosophers were not required to completely avoid any theories. Those who endorse the method of experimental philosophy can consistently entertain theories, put them forward as hypotheses to be tested experimentally, reject theories that are incompatible with experimental evidence, tentatively or provisionally endorse the theories that they take to be more probable than their competitors, and even firmly endorse certain theories (such as corpuscularism), insofar they are warranted by experience. In the light of this, Boyle, Hooke, Locke and Newton can be said to have endorsed the method of experimental philosophy, even though they did not endorse the fully a-theoretical stance that is often associated with the movement.
Who wasn’t an experimental philosopher?
One may fear that, once we grant that experimental philosophers could be engaged in some forms of theorizing, the notion of experimental philosophy becomes too inclusive, so that Descartes, Leibniz, and even certain Aristotelians can be classed as experimental philosophers. This is not the case. The fact that an experimental philosopher can endorse theories and substantive natural-philosophical claims does not entail that there is no real distinction between those who endorsed and those who did not endorse the method of experimental philosophy.
Descartes and Leibniz too engaged in experiments. Leibniz even claimed to be “strongly in favour of the experimental philosophy”. However, insofar as Descartes thought that he had firmly established the highly speculative cosmogonical theories of the Principles of Philosophy, in spite of the scant empirical evidence with which he backed them up, he was hardly following the methodological prescription spelt out in [E]. As for Leibniz, although he stressed the importance of experience for natural philosophy, he also held that some basic propositions of natural philosophy (like the principle of uniformity of nature) could be established only a priori. This view is incompatible with [E], we should not take Leibniz to be an experimental philosopher even though he stressed the importance of experience.
Problems of classification
Although there is a real distinction between those who endorsed the method of experimental philosophy and those who did not, establishing whether an early modern author really was an experimental philosopher is sometimes difficult. There can be a disconnect between rhetoric and methodology, or methodology (understood in the etymological sense of method-talk) and actual, practised method.
Some philosophers called themselves experimental philosophers, or endorsed the rhetoric of the movement (e.g. using “hypothesis” and “speculation” as a term of abuse), but they thought they were entitled to endorse certain natural-philosophical claims a priori. Others claimed that they were following the method of experimental philosophy, but they endorsed theories that outstripped the empirical evidence. Leibniz, as we have seen, is an example of the disconnect between rhetoric and methodology. Another example is provided by those Jesuits, like Daniello Bartoli, who rehearsed the experimentalist rhetoric, but attempted to take the sting out of experimentalism and to combine it with some Aristotelian doctrines that were hardly warranted by experience. In other cases, there was a disconnect between methodology and actual method. An example is provided by the account of vision of the late seventeenth-century Italian natural philosopher Francesco Bianchini. Not only his rhetoric, but also his methodological pronouncements were in line with [E]. However, his account of vision was far more speculative than those pronouncements allowed.
The disconnect between rhetoric, methodology, and practised method determines a difficulty in establishing whether certain authors were experimental philosophers. This would be worrying if a primary aim of the study of early modern experimental philosophy were providing a handy classification of early modern authors. However, the point of studying early modern experimental philosophy is not pigeonholing early modern authors, but understanding their philosophical views and practices, even when the former were not in line with the latter. Finding these discrepancies should not be surprising. People sometimes fail to practice what they preach. Other times, they use a rhetoric that is at least partly out of step with their actual views. Early modern philosophers were no exception.
Butler and Clarke on the infinity of God
Juan Gomez writes…
In my three previous posts I have been commenting on the experimental methodology in religion and the contrast between the preferred methods of Samuel Clarke and Joseph Butler. We saw bishop Halifax’s general description of the a priori and a posteriori methods, and then we examined Butler’s preference for the latter and Clarke’s adoption of the former. Today I want to conclude this series of posts on religion by focusing on one specific application of the a priori method and Butler’s criticism of it.
In his A Demonstration of the being and attributes of God (1704) Clarke provides an argument for the infinity of God. As we have already mentioned in previous posts, Clarke prefers the a priori method because it provides ‘demonstrable proof’ of the attributes of God, while the a posteriori method can only provide probable knowledge. Clarke appeals to a set of premises from which he concludes that God, i.e. the only self-existing being, must be infinite. The following is Jonathan Bennett’s rephrasing of Clarke’s argument:
- ‘x is self-existent’ means that it’s a contradiction to suppose that x doesn’t exist
- ‘x is finite’ implies that there are places at which x doesn’t exist.
Therefore:
- It is a flat-out contradiction to suppose that something is both self-existent and finite.
So if you accept that God is self-existent, then you must also accept that he possesses the attribute of infinity. However, Butler is not convinced by this argument. In particular, Butler thinks that Clarke cannot say that if a being can be absent, without contradiction, from one place, then it can also be absent, without contradiction, from all places, which is exactly what Clarke claims is absurd. This is the passage that Butler criticises:
- “To suppose a Finite Being, to be Self-Existent; is to say that it is a Contradiction for that Being not to exist, the Absence of which may yet be conceived without a Contradiction: which is the greatest Absurdity in the World: For if a Being can without Contradiction be absent from one place, it may without a Contradiction be absent likewise from another Place, and from all Places…”
Butler points out that a being, without contradiction can be absent from another and all places as long as it is at different times, but it is most certainly an absurdity to claim that a finite being can be absent from all places at the same time, since this would entail that it ceases to exist. Clarke replies that Butler is mistaken here, since it is indeed possible for a finite being to be absent of all places and all times, since such being is not necessary. Clarke appropriately switches back the conversation to talk of necessary beings, but this still is not enough to satisfy Butler.
In a follow up letter to Clarke’s reply, Butler now objects to the definition of self-existence. Butler argues that from the claim that a being necessarily exists it does not follow that it exists everywhere; it only follows that such being must exist somewhere, but there is no contradiction in supposing it is absent from other places at the same time.
It seems that the key to the whole discussion is the idea of necessity. If the self-existent being is necessary, then this must be the case everywhere, so it would be a contradiction to think that a necessary being is absent at one place. Butler seems to be missing this point in his exchange with Clarke. However, to be fair to Butler, at the time the correspondence took place, Butler was still very young (21 years old), and he eventually changed his mind and accepted Clarke’s argument.
However interesting the discussion between Butler and Clarke, I only wanted to provide an example of the way the debate between a priori and a posteriori methods took place within a religious context. Butler might have eventually accepted Clarke’s argument for the infinity of God, but 20 years after his correspondence with Clarke he was going to prefer the a posteriori method in his Analogy (1736), holding on to his belief that in matters of faith probable knowledge is all we need.
Why did Butler end up preferring the a posteriori method even though he accepted Clarke’s argument? The answer lies in the limits of knowledge that resulted from a commitment to experimental philosophy. Figures like Butler and Turnbull applied the experimental method to subjects that went beyond the observation of the natural world. This meant that the application of the experimental method could not be as straightforward as it was in natural philosophy, for the objects under consideration (‘moral objects’) are unobservable in the sense that they cannot be experienced via the five external senses. Knowledge of the existence of a future state and the attributes of God is not directly accessible to us, given our human nature. Anything we can possibly now about such things must be by analogy to the knowledge we gain from the observation of the natural world. This is why our knowledge of these ‘moral objects’ can never be demonstrable, but only probable. From our observation of the natural world we can safely conclude that it is probable that there is a future state, and highly probable that God is infinite (omnipresent), but we can never construct a demonstrable proof of these conclusions.
In one of the final exchanges on this topic Butler confesses that he insisted on his objection because he wanted a demonstrable proof of the infinity of God, and this misled him:
- “…your argument for the omnipresence of God seemed always to me very probable. But being very desirous to have it appear demonstrably conclusive, I was sometimes forced to say what was not altogether my opinion.”
Clarke suggests in his reply that this is a fault many are mislead into, and he blames one of the preferred targets of advocates of experimental philosophy, René Descartes:
- “…the universal prevalency of Cartes’s absurd notions (teaching that matter is necessarily infinite and necessarily eternal, and ascribing all things to mere mechanic laws of motion, exclusive of final causes, and of all will and intelligence and divine Providence from the government of the world) hath incredibly blinded the eyes of common reason, and prevented men from discerning him in whom they live and move and have their being…”
The upshot of applying the experimental method to moral topics was that it could only provide probable (not demonstrable) knowledge. However, recognizing this limitation was far more reasonable than those chimeras constructed by speculative philosophers.
