Isaac Newton wrote little on the foundations of mathematics, as now understood. By comparison, Isaac Barrow, who preceded him in the Lucasian Chair of Mathematics at Cambridge, wrote quite extensively about the nature of mathematics and its usefulness to learning, despite his relatively short career as a mathematician. Partly, of course, that should be attributed to Newton's spartan style. In natural philosophy he rejected the ‘method of hypotheses’ of Hooke and the Cartesians in favour of a ‘deduction from the phenomena’, according to which mathematical relationships would be evinced directly from the empirical evidence prior to any philosophical speculation. But this itself constitutes a philosophy of science, and a very important one historically, given its influence on the logical positivists. In the same way, then, Newton's emphasis on mathematical certainty, and on the superiority of the geometrical approach to the analytic methods of the Cartesians and Leibnizians, likewise constitutes a philosophy of mathematics. Yet, despite his undisputed place in the pantheon of mathematical greats, despite the tremendous notoriety of his controversy with Leibniz over the invention of the calculus, and the huge importance to Newton himself of his opposition to analysis (not to mention its effect on subsequent British mathematics), up until now there has been no systematic study of Newton's philosophy of mathematics.
In his latest book Niccolò Guicciardini sets out to remedy this state of affairs, and does so with remarkable lucidity and thoroughness. The first part of the book is devoted to a comparison of Newton's methods with Descartes's, the second to Newton's writings from the 1670s, where a tension between his mathematical practice and his views on method first emerges. Part III is a study of Newton's recasting of his method of series and fluxions in synthetic form, while Part IV concerns the mathematical methods Newton deploys in his Principia. In Part V Guicciardini examines ‘perhaps the most philosophically freighted texts that Newton wrote’ (p. xviii), those of the 1690s and early 1710s, where he portrays himself as the heir to the mathematics of Euclid and Apollonius. Finally, in Part VI, Guicciardini examines Newton's polemical strategy in his dispute with Leibniz, which occupied him especially after 1710, with particular attention to his ‘scribal policy’ of circulating manuscripts rather than publishing them.
The comparison of Newton's methods with Descartes's is especially important, not just because Newton cut his teeth on Van Schooten's Latin translation and edition of Descartes's Géométrie only to become its fiercest critic (p. 57), but because one can hardly understand Newton's rejection of Descartes's methods without knowing what was meant by key terms of the mathematics of the day. The term ‘analysis’ is particularly troublesome. For although Descartes is rightly credited with founding analysis by the reduction of geometry to algebraic equations, in fact, as Guicciardini notes, ‘the Géométrie displays few elements of what is nowadays identified as analytic geometry’ (p. 32). There ‘analysis’, in a usage originating with Pappus, means the resolution of the problem given, a first step prior to the finding of a solution by a geometrical construction of curves. Descartes's originality lay in devising an algebraic method for resolving problems into equations, where letters represent known and unknown line segments, and operations on them are closed, so that the product of two line segments is another finite line segment, and not an area. For this only finite algebraic equations could be deployed.
Here, of course, Newton went far beyond Descartes, and in two short years had achieved results which, had they been published, would have established his reputation as one of the foremost practitioners of the new art. This lay largely in his contribution to the so-called ‘new analysis’, which allowed a resolution of a finite quantity, such as an area, into an infinity of infinitesimal elements. For by 1666, building on the work of Barrow and Wallis in particular, he had invented the analytic method of fluxions; proved what we call the fundamental theorem of the calculus; discovered the binomial series; and, by means of infinite series expansions of terms, made comprehensive tables of analytic quadratures (integrals, in Leibnizian terminology). His subsequent rejection of this ‘new analysis’ in favour of the synthetic methods of the Principia has been a source of puzzlement to historians, as has his strategy in his dispute with Leibniz, where he attributes the direct method of fluxions (Leibniz's differentiation) to de Sluse and Barrow, and lays claim rather to have discovered a battery of techniques for the solving of inverse problems.
Guicciardini finds the key to unlocking these interpretive problems in what Newton has to say on method and certainty, with careful attention to the shifting meanings of the key terms. Newton's increasing disdain for algebra is rooted in the commitment to the Pappusian method he shares with Descartes, where finding an equation does not constitute finding a solution. For this a synthesis is required, and this means devising a geometrical construction, for which neither the Cartesian common analysis nor the new analysis to which he had contributed suffices. Moreover, Newton disagreed profoundly with Descartes about what curves were allowable in such constructions. Descartes would not allow curves traced by mechanical motions whose two components had velocities in the ratio of straight to curved lines – such as, for example, the cycloid – since this violated the restriction to finite algebra. Against this Newton scoffed that it was absurd to exclude the cycloid, a curve which could be used to cut an angle in any given ratio, when it is ‘exceedingly well known and very easily described through the motion of a wheel or a circle’ (p. 67). This, Newton came to believe, was the correct way to understand mathematical quantities. They should not be conceived as ‘consisting of least possible parts’, as by the Leibnizians, ‘but as described by a continuous motion’, in conformity with the method of the ancients as well as the reality of physical nature (p. 226). ‘Synthesis’ thus becomes for him a term connoting not just the providing of a geometrical construction, but the genesis of figures by continuous motions. Hence the expenditure of effort in formulating a properly synthetic method of fluxions. This understanding of mathematics is famously encapsulated in Newton's preface for the Principia where he describes geometry as ‘part of universal mechanics’ and ‘founded on mechanical practice’.
There is far more to be found in this rich and learned study than can be sketched in a short review. The analysis is informed throughout by a sympathetic reading of the work of other Newtonian scholars, and Guicciardini skilfully weaves their insights into his finely wrought argument. The resulting book is an invaluable aid to our understanding of the complex mind of Newton and at the same time a significant contribution to the history of mathematics.
