You are on page 1of 3

Available online at www.sciencedirect.

com

ScienceDirect
Historia Mathematica 44 (2017) 280–282
www.elsevier.com/locate/yhmat

Letter to the Editor

Letter to the editors of the journal Historia Mathematica

Dear Editors,
There is a long, unpleasant tradition of denigrating Leibniz’s mathematical achievements. It might suffice
to mention names like Nicolas Fatio de Duillier, John Keill, and James Marc Child. Recently Viktor Blåsjö
has continued this tradition with his paper entitled “On what has been called Leibniz’s rigorous foundation
of infinitesimal geometry by means of Riemannian sums,” Historia Mathematica, 44. As is usual in this
scholarship, he cites Child. Blåsjö’s title refers to the title of my paper published in the journal Synthese in
2002. For the sake of Leibniz, I feel obliged to reject and to refute the author’s wrong, misleading statements
and arguments, and use of contrived citations. This is, indeed, a completely unacceptable strategy. Two
examples are given below.
What is at stake here is Leibniz’s work De quadratura arithmetica circuli ellipseos et hyperbolae cujus
corollarium est trigonometria sine tabulis. In 1993, I published the last version, elaborated in Paris in 1676,
for the first time. Leibniz added a list of the “more notable” – that is, of the first seven – theorems. Theorem 6
is characterized as follows: “Theorem 6 is most thorny. Therein it is overly carefully demonstrated that the
procedure of constructing certain rectilinear step spaces (spatia rectilinea gradiformia) and in equal fashion
polygons can be continued to such a degree that they differ from each other or from curves by a quantity
which is smaller than any given quantity. . .It serves, however, (servit tamen) to lay the foundations of the
whole method of indivisibles in the soundest way possible (firmissime).” And a bit later: “It brings about
only this that two spaces of which one passes into the other if we progress infinitely, approach each other
to a difference which is smaller than any arbitrary assigned difference.”
Leibniz thus described his aim and result clearly: Theorem 6 laid the rigorous foundation of infinitesimal
geometry, especially of a general integration theory. “Indivisibles” are infinitely small, that is, according to
Leibniz’s definition smaller than any given quantity. Leibniz retained the notion of indivisible, but changed
its original meaning. The subject of “servit tamen” must be “propositio 6”. Latin grammar does not leave
any other choice: It is impossible that “it” means “the strategy of proving that the approximation involved
can be made arbitrarily good” as Blåsjö maintains (p. 137).
In fact, Blåsjö tries to demonstrate two assertions: (1) Theorem 6 is about one specific integration for-
mula, not integrability in general, and (2) Leibniz did not think of it as a foundational innovation, but as
a rather pedantic, and basically routine, way of applying what is essentially the ancient Greek method of
exhaustion. Both assertions are wrong and contradict Leibniz’s own explicit statements.
As to the first assertion, the title of Blåsjö’s paper correctly speaks about “infinitesimal geometry”. But
immediately following, the abstract speaks about “infinitesimal calculus”. But Leibniz’s treatise represents
a comprehensive discussion and rigorous foundation of infinitesimal geometry, not of the infinitesimal
calculus. The replacement or fabricated citation is repeated on page 135. On page 141 a general foundation
for the calculus is denied. Yet, though there is a close relationship between geometry and calculus they do
http://dx.doi.org/10.1016/j.hm.2017.07.001
0315-0860/© 2017 Elsevier Inc. All rights reserved.
Letter to the Editor 281

not coincide. The first notion must not be replaced by the second. It is exactly this difference between the
geometrical approach of the treatise and the more efficient calculus that caused Leibniz in 1691 to renounce
the publication of the treatise, not because it was worthless1 : “But the subject increased during the working
out and the time was lacking to polish the treatise for a publication . . . especially because it did not seem
to be sufficiently worthwhile to explain in the usual way at some length what our new analysis expounds
without much ado.”
The main idea of the differential calculus (linearization) is emphasized in the scholium after theorem
232 : “They (the readers) will notice what a large field of discovery is available once they have correctly
understood only this that every curvilinear figure is nothing but a polygon with infinitely many, infinitely
small sides. If Cavalieri, even Descartes himself had sufficiently observed that, they would have achieved
or hoped for more.” In 1686 Leibniz called his differential and integral calculus “analysis of the indivisibles
and the infinites”.
The general theorem 6 is used in order to most rigorously (summum rigorem) demonstrate theorem 7
– but not only theorem 7 as Blåsjö maintains. Theorem 6 concerns only the curve passing through the
points D of figure 3, not the curve passing through the points C. In contrast to Blåsjö’s statement there is
no problem in this respect.
The quadrature – or in modern terms the integration – exclusively deals with the curve passing through
the points D which has to be continuous and monotonously increasing or decreasing. The whole demonstra-
tion consisting of eight steps shows that the difference between the areas of the mixtilinear figure limited
by this curve and of the step figure composed of rectangles can be made smaller than any given quantity or
– in order to use Leibniz’s terminology – infinitely small.
For example, theorem 6 also provides the justification for the “absolutely geometrical quadrature”
(quadratura absolute geometrica) of the logarithmic curve (theorem 46). This reliance of theorem 46 on
theorem 6, in itself, falsifies Blåsjö’s assertion. That is, in theorem 46 Leibniz replaces mixtilinear areas by
rectangles and vice versa – which is allowed because of theorem 6. Otherwise these replacements would
be arbitrary. In 1993 I have provided a detailed analysis of this geometrical, complicated quadrature that
implies the application of theorem 6.3
As to the second assertion Leibniz emphasized again and again the rigor of the proof of theorem 6
and the innovation4 of his handling infinitely small and infinite quantities: See especially the scholia after
the theorems 22 and 23. Therein Leibniz explains that one has to be cautious whenever one would like
to calculate with the infinite and to use the method of indivisibles. Conclusions regarding the infinite are
dangerous if they are not guided by the thread of a proof5 .
He emphasizes the great light that is turned on in the whole method of indivisibles and the generality
of his integration method. Inscriptions and circumscriptions are no longer necessary but rather something

1 “Sed quod materia sub manibus crescente limare ad editionem non vacavit . . . praesertim cum nunc prolixius exponere vul-
gari more, quae Analysis nostra nova paucis exhibet, non satis pretium operae videtur”. (G.W. Leibniz, Quadratura arithmetica
communis sectionum conicarum, quae centrum habent etc. In: Acta Eruditorum April 1691, pp. 178–182. I cite the reprint in:
G.W. Leibniz, Mathematische Schriften, ed. by C.I. Gerhardt. Vol. V. Halle 1858, pp. 128–132, here p. 128).
2 “Sentient (sc. lectores) autem quantus inveniendi campus pateat, ubi hoc unum recte perceperint, figuram curvilineam omnem
nihil aliud quam polygonum laterum numero infinitorum, magnitudine infinite parvorum esse. Quod, si Cavalerius, imo ipse Carte-
sius satis considerassent, majora dedissent aut sperassent.“ (G.W. Leibniz, Sämtliche Schriften und Briefe, series VII, volume 6.
Berlin 2012, p. 586 = LSB VII, 6, p. 586.)
3 Les courbes analytiques simples chez Leibniz. In: Sciences et techniques en perspective 26 (1993), 74–96.
4 “Quae de infinitis atque infinite parvis huc usque diximus, obscura quibusdem videbuntur, ut omnia nova.” (LSB VII, 6, p. 585.)
5 “Libenter hanc contemplationem persecutus sum, quia specimen exhibet cautionis circa ratiocinia de infinitis, et methodum
indivisibilium . . . Eleganti argumento quam lubrica sit ratiocinatio circa infinita, nisi demonstrationis filo regatur.” (LSB VII, 6,
p. 583f.)
282 Letter to the Editor

in between, in modern terms “Riemannian sums”.6 Blåsjö speaks about “Riemannian” proof (p. 145) thus
replacing my “sums” by “proof”. This is another example of his fabricated citations. I never wrote such
nonsense. It is needless to say that “sums” do not coincide with “proofs”.
Blåsjö wonders why Leibniz did not submit the treatise to the French Academy of Sciences as he had
intended to do, and doubts its significance (p. 136). The correspondence between Leibniz and Huygens,
however, makes it clear that Leibniz hoped to become an honorary or corresponding member of this
academy by means of the treatise, though he would have to be absent, as he wrote to Huygens on the
eighth of September 1679: “J’ay laissé à Paris mon manuscript de la quadrature arithmétique, à fin de l’y
faire imprimer un jour.” A bit later, on the twentieth of October 1679 he wrote: “Car sans parler de quelques
decouvertes mathematiques de mon crû (particulierement de ma quadrature dont j’ay achevé la demonstra-
tion dans les formes, avec quantité d’autres propositions considerables y comprises, et qui pourroit estre
adopté de l’Académie) je suis. . .”
In other words, Leibniz did want to publish it. Yet, he had to leave Paris some months after he had
finished the treatise. He spoke about “numerous considerable theorems” that can be found therein, that is,
originally he rated it very high. The publication could not be realized in Paris. The publishable copy left in
Paris got lost when it was sent from Paris to Hannover. This loss made the situation even more complicated.
Finally Leibniz gave up a publication for the reasons mentioned above. This matter of fact does not imply
any evaluation of its true significance. Very many of Leibniz’s important mathematical ideas and theorems
remained unpublished during his lifetime, for example his determinant and elimination theory.
Moreover he published some results in 1682 and in 1691, respectively, in particular his famous series
for π4 , in other words the arithmetical quadrature of the circle. But the treatise comprehends far more than
this special, wonderful result that decorates the title page of the first volume of the Miscellanea Berolinensia
published in 1710. For good reasons Leibniz was very proud of his discovery, which was especially praised
by Huygens. He repeatedly emphasized the fruitfulness of his method of dealing with the infinitely small
and the infinite that caused him to elaborate this long treatise. It discusses such crucial notions as that of
demonstration, rigor, method, curves, infinitely small, infinite etc. and proves the convergence criterion
named after Leibniz for alternating, infinite series (theorem 49). How is it possible to doubt the significance
of this treatise?
I would like to conclude with a personal remark. Up to now I did not believe that I would have to write
such a letter. And I regret very much that a young scholar chose a polemical confrontation, instead of
cooperation with somebody who has been studying and editing Leibniz’s manuscripts for several decades.
With best regards, Eberhard Knobloch

Eberhard Knobloch
Berlin-Brandenburg Academy of Sciences and Humanities

Available online 24 July 2017

6 “. . . ne semper inscriptis vel circumscriptis uti, et ad absurdum ducere, et errorem assignabili quovis minorem ostendere necesse
sit . . . quanta hic lux accendatur in tota methodo indivisibilium.” (LSB VII, 6, p. 585f.)

You might also like