Skip to Navigation Skip to Content Skip to Search Skip to Site Map

Monthly Archives: October 2010

Baconian versus Newtonian experimental philosophy

Peter Anstey writes…

Eric Schliesser’s comments about the utility of the experimental/speculative distinction, provide an opportunity for me to lay out a distinction that is absolutely central to our project. But let’s hear from Eric first: I quote from his blog post on It’s Only a Theory:

    It ignores at least one other group of philosophers, namely those that believed in (mathematical) theory mediated measurement. I am thinking of Galileo, Huygens, and Newton, among the best known. These are not best described as experimental, although all were accomplished experimentalists (and Newton’s Optics is often assimilated to experimental traditions), but their work has very different character from say, Bacon or Boyle. (They are also not best described as speculative, because all three practiced a self-restraint on published speculation.) Certainly after the Principia this approach created standing challenge to all other forms of philosophizing. So the Otago framework will run into big trouble in 18th century.

We’ve already shown that, in fact, the terminology of the experimental philosophy is very prevalent in the 18th century and, moreover, that the experimental philosophy was extended beyond natural philosophy into moral philosophy and even aesthetics. See, for example, the works of George Turnbull which are a good example of experimental moral philosophy.

But the important issue Eric raises has to do with those who practised ‘theory mediated measurement’ such as Galileo, Huygens and Newton. What our research has shown is that the experimental philosophy was practised in two quite different ways. Up until the 1690s, Boyle, Hooke and the early Royal Society practised experimental philosophy according to the method of Baconian natural history. However, from the last decade of the seventeenth century Newton’s new mathematical natural philosophical method came to be seen as the preferred method of experimental philosophy. The Baconian natural history program started to run out of steam in the 1690s and it soon came to be replaced by the Newtonian method. This is, in fact, the explanation of Newton’s common refrain ‘Natural philosophy is not natural history’. And Newton himself had a large hand in the demise of the Baconian approach to experimental philosophy both through criticism and through his own positive alternative. Far from providing an exception to our framework, Newton, the self-confessed experimental philosopher, is one of the central players!

Does Newton feign an hypothesis?

Kirsten Walsh writes…

Newton’s famous pronouncement, Hypotheses non fingo, first appeared in 1713, but Newton’s anti-hypothetical stance is present as early as 1672, in his first papers on optics.  In his first publication, he introduces his notion of certainty, and insists that his doctine of colours is a theory; not an hypothesis:

    For what I shall tell concerning [colours] is not an Hypothesis but most rigid consequence… evinced by ye mediation of experiments concluding directly & without any suspicion of doubt.

Despite these clear anti-hypothetical themes, a corpuscular hypothesis lies beneath Newton’s theory of light and colours.  What are we to make of this?  Is Newton guilty of feigning an hypothesis? Is Wolff correct when he says that Newton “indulges in hypotheses in those very areas in which they think he abstained from employing them“?

To begin, what does Newton mean by Hypotheses non fingo?  ‘Fingo’ has been variously translated as ‘frame’, ‘make’, ‘imagine’ and ‘devise’. Experts argue that ‘feign’ is the most appropriate translation.  While it has a variety of meanings, such as to form, to invent, to forge, or to suppose erroneously, the word ‘feign’ also carries the nuance of pretence, counterfeit, or sham.  Thus, they argue that while Newton indeed conceived or framed hypotheses, he did not attach any special epistemic status to them.  He maintained a clear demarcation between theories that were supported by experimental results and hypotheses that were merely unsupported speculations.

Now let’s take a closer look at Newton’s early optical papers.  Newton claims that his doctrine of colours is a theory, not an hypothesis, for three reasons:

  1. It is certainly true, because it is supported by (or deduced from) experiment;
  2. It concerns the physical properties of light, rather than the nature of light; and
  3. It has testable consequences.

These are the three key aspects of Newton’s early methodology.  He refers to them again and again throughout the debate that followed the publication of his first optical paper.

Newton explicates his corpuscularian view in his first optical paper and describes light rays as substantial bodies.  But when his opponents accuse him of hypothesising, Newton argues that he is not guilty.  Firstly he argues that this hypothesis is not necessary for his explanation of colours.  Secondly he argues that he attaches no special epistemic merit to his hypothesis because:

  1. It is not supported by experiment;
  2. It concerns the nature of light; and
  3. It has no testable consequences.

While Newton never gives up his corpuscularian view, he attempts to explicate and promote his theory without referring to it.  He argues that he doesn’t need to provide any hypothesis on the nature of light – his theory on the properties of light is sufficient on its own.

I claim that Newton isn’t guilty of violating his anti-hypothetical stance.  He demonstrates that he can distinguish between theory and hypothesis, giving the former higher epistemic status than the latter.  He does not pretend to have empirical support for his corpuscular hypothesis, nor does he try to ‘prop it up’ on other grounds.  Perhaps he regrets having ever opened the proverbial can of worms, for the next time he explicates his theory of light and colours, he does so without any reference to the corpuscular hypothesis or the nature of light.

That Newton can tell the difference between good scientific explanations and speculations is further supported by his use of queries in these early optical papers, but more on this next time.  To conclude,  I think Newton is not guilty of feigning an hypothesis.  What do you think?

Experiment and Hypothesis, Theory and Observation: Wolff vs Newton

Alberto Vanzo writes…

Looking for sources for knowledge of experimental philosophy in eighteenth century Germany, I found some interesting texts by relatively unknown authors (at least beyond the circle of specialists). Christian Wolff is one of them. He was the most famous German philosopher in the first half of the eighteenth century. His philosophy was taught in many universities and his works were very popular. For instance, his German Logic knew no less than 14 editions during Wolff’s life.

Wolff knew several British experimental philosophers. He cited works by Robert Boyle and Robert Hooke, he was the Locke reviewer for an important journal (the Acta eruditorum), and he polemized with the Newtonian John Keill on the existence of the vacuum. He is a good example of the fact that German thinkers were acquainted with the works and the methodological views of British experimental philosophers in the first half of the eighteenth century.

A number of Wolff’s statements might make us think that he was himself an adherent to the early modern version of x-phi. Like British experimental philosophers, Wolff criticizes Descartes’ attempt to explain a great variety of natural phenomena in the light of few general principles that he established a priori. Like Hume and Hutcheson, Wolff is eager to extend the dominion of experimental philosophy beyond the boundaries of physics. He projects the disciplines of experimental cosmology, experimental teleology, experimental theology, experimental politics, and even experimental ontology. For his philosophical system as a whole, he chooses the name of “universal experimental philosophy” (philosophia experimentalis universalis). How could Wolff have been a more enthusiastic adherent to the program of experimental philosophy?

Yet contrary to the appearances, Wolff’s views were quite different from those of his British counterparts. This can be seen by comparing him with Newton. Newton, like virtually every other early modern experimental philosopher, claimed that he did not feign any hypothesis (his famous hypotheses non fingo). Wolff rebuts that Newton

    indulges in hypotheses in those very areas in which they think he abstained from employing them […] In fact, what else is universal attraction or gravity, which is represented by a measure of attraction, if not a hypothesis which is assumed because of certain phenomena and then is extended to all matter?

According to Wolff, not only did Newton feign hypotheses, but he did well to do so. This is because natural philosophers must proceed like astronomers:

Christian Wolff

Christian Wolff

From some present events, they infer what they have to assume, in order for [the events] to follow [from it], and they posit that their hypothesis applies to all [similar] events […] To determine whether they did well to assume the hypothesis, they infer what follows from it on the basis of a correct reasoning, in order to compare it with the remaining events that they have either observed, or that they derive from observations. [They do this] in order to see whether what has been observed agrees with the hypothesis. If they find that [observations and hypothesis] are in contrast with one another, then they improve the hypothesis, and in this way they constantly move closer to the truth.

Wolff holds that there is a circular relationship between observation or experiment on the one hand, and theory on the other hand. He stresses

    how much theory owes to observations and how much, on the other hand, observations owe to theory, since observations perfect theory and theory in turn continuously perfects observations. He who is ignorant of any theory and does not have much ability to use the faculty of knowing will only discover obvious and mostly imprecise [truths] on the basis of observations. There would not be much progress, unless one could presuppose some theory; and the more [a theory] is developed, the more discoveries one will make by means of observation[s].

Unlike Newton, Wolff was no great scientist. However, the quotes above suggest that his methodology of science is worth a serious reading. His acknowledgments of the interaction between theory and observation sound modern. They sketch a version of the hypothetico-deductive method that might provide an interesting alternative to Newton’s strict inductivism.

In summary, Wolff is a good example of the Germans’ knowledge of British experimental philosophy in the first half of the eighteenth century. His views are also interesting in their own right. So are Johann Nicolaus Tetens’ comments on observational vs speculative philosophy or Johann Heinrich Lambert’s distinction between theory-testing experiments and experiments that have a life of their own – two hundred years before Ian Hacking. More on this another time.

In the next post, Kirsten will discuss Newton’s method, in particular his rejection of hypotheses and his use of queries. See you next Monday!

Monthly Update: Events, CFPs, and Readings

Hello, readers!

Below is a list of upcoming events, call for papers, recent posts and a journal article of some relevance to early modern experimental philosophy.

Upcoming Events:

Calls for Papers with deadlines in October-November:

  • Annual Conference of the New Zealand Division of the Australasian Association of Philosophy. University of Waikato (New Zealand), 5-9 December 2010. Deadline: 29 October.
  • Hume after 300 Years: The 38th International Hume Society Conference. Edinburgh, 18-23 July 2011. Deadline: 1 November.
  • New York City Workshop in Early Modern Philosophy. New York, 25-27 February 2011. Deadline: 15 November.


Brett Fulkerson-Smith discusses Kant’s “experiment of pure reason” in the last issue of the Kantian Review.

Two reviews relating to the early Royal Society have recently been published: a review of John Gribbin’s book The Fellowship at Some Beans and a review of an exposition on John Aubrey and the roots of the Royal Society in the Times Online.

Here on our blog, we introduced ourselves and our project. We wrote about our research on Newton’s mathematical method and on the experimental method in British moral philosophy, in particular in George Turnbull. We discussed the difference between contemporary and early modern experimental philosophy. We argued that it is better to interpret the history of early modern philosophy in the light of the distinction between experimental vs speculative philosophy, rather than rationalism vs empiricism. In the comments, Neil Rickert, Benny Goldberg and Gary Banham provided valuable suggestions. Thanks!

We did not quite manage to fully convince Eric Schliesser. He discussed our ideas over at It’s Only A Theory.

Have we missed some event, call for paper, or reading? Would you like us to include your writings or events in the next monthly update? Do let us know! Also, you can subscribe to our mailing list or RSS feed if you would like to be notified of new posts. For more frequent updates, follow us on Twitter.

See you next Monday with a post on Christian Wolff vs Isaac Newton on the experimental method.

Turnbull and the ‘spirit’ of the experimental method

Juan Gomez writes…

You will probably recognize the following phrase: ‘An attempt to introduce the experimental Method of Reasoning into moral Subjects.’ It is the subtitle of David Hume’s Treatise of Human Nature of 1739-40, and the first explicit mention of the application of the experimental method in moral topics. Many scholars have pointed to it, and claimed that Hume was the first one to go forward with this attempt. However, others (Tom Beauchamp, Alexander Broadie) have also noticed that this idea did not originate with Hume. I will show here that the ‘spirit’ of the experimental method was very much alive at least 20 years before the publication of Hume’s Treatise. In fact, contrary to the most commonly held view, Hume should not be the reference point when studying the emergence of the “science of man”. Rather, we should look at the Aberdeen philosophers, in particular at George Turnbull and his lectures at Marischal College in the 1720’s.

I will make a prima facie case for this claim with only a few quotes (available in this document), but please do contact me if you are interested in the topic, since there is more than enough evidence that I would be happy to discuss with you.

To begin with, Hume was not the first to allude to the application of the experimental method in moral philosophy. Francis Hutcheson had already done this in his 1725 Inquiry. The subtitle of this work explains that it contains the following:

Turnbull's Principles of Moral Philosophy(1740)

Turnbull's Principles of Moral Philosophy(1740)

    the Principles of the late Earl of Shaftesbury are explained and defended against the Fable of the Bees; and the Ideas of Moral Good and Evil are established, according to the Sentiments of the Ancient Moralists, with an attempt to introduce a Mathematical Calculation on subjects of Morality. (emphasis added)

Hutcheson doesn’t use the words ‘experimental method’, but saying that he will give a ‘mathematical calculation on subjects of morality’ is perfectly in line with the spirit of the experimental method (specifically with the Newtonian method). To be fair to Hume, he does recognize Hutcheson as one of the philosophers who has “begun to put the science of man on a new footing.” (Treatise (1739), Introduction, p. 6-7) So Hume might have recognized that he was not the first, but a number of modern scholars have not.

Moving on to George Turnbull, whom I believe is mistakenly underrated as a figure of the Scottish Enlightenment, published his Principles of Moral Philosophy in 1740, the same year Hume published the third volume of his Treatise, which is on Morals. This would at least lead us to think that both Hume and Turnbull were working on the application of the experimental method in morality at the same time. But as Turnbull mentions in the introduction to his Principles, his book is based on the lectures he gave at Marischal College between 1721 and 1726, around the time when Hume was a student at the University of Edinburgh. Besides the numerous remarks in the Principles that show Turnbull’s devotion to the experimental method, there is a key document that shows that he was teaching the young Aberdeen students the moral philosophy he explains in the book he published 17 years later.  The document is the 1723 graduation thesis, which the graduating students (Thomas Reid among them) had to defend, titled De scientiae naturalis cum philosophia morali conjunctione (On the unity of natural science and moral philosophy).

I am currently working on the 1723 thesis, and at this moment I can let you know that it is strengthening my belief in the importance of Turnbull in the development of the ‘science of man.’ For now I’ll leave you with enough quotes from the Principles that show that if we want to study the development of the science of morals, we should start focusing more on Turnbull and Aberdeen, and less on Hume and Edinburgh.