60 of 60 people found the following review helpful
V. N. Dvornychenko
- Published on Amazon.com
Most of us got our first glimpse of the fascinating history behind the calculus in first-year calculus. That is, we did if we were lucky -- for the fast pace in acquiring basic calculus skills leaves little extra time. Perhaps we managed to learn that Newton and Leibnitz are regarded co-discoverers of the calculus, but that their splendid contributions were marred by a bitter - at times positively ugly - rivalry. We may also have learned something about their precursors, for example Descartes, Fermat and Cavalieri.
If these glimpses left a taste for more, Boyer's "The History of the Calculus and Its Conceptual Development" is just the book. Boyer begins by tracing the calculus roots back to Ancient Greece. During this period two figures emerge preeminent: Eudoxus and Archimedes. Archimedes was a pioneer whom many consider the "grandfather" of calculus. But lacking modern notation he was limited in how far he could go.
The role played by Eudoxus is more ambiguous. He represents that vein of mathematics which treats "infinity" with the greatest caution - if not abhorrence. Although magnitudes are allowed to become arbitrarily large, they can never actually become infinite. This has given rise to two schools of thought: 1) those that consider a circle to be a polygon of infinite number of sides (completed infinity), and 2) those that allow that a circle can be approximated arbitrarily closely by means of polygons, but disallow this process ever being completed (incomplete infinity or "exhaustion" method). Both schools remain with us to the present.
Their relevance to calculus is this: the first gave rise to "infinitesimals" (infinitely small quantities); the second to the "limit" or "epsilon-delta" approach.
In chapters II and IV Boyer discusses the contributions of the precursors of Newton and Leibnitz. These include Occam, Oresme, Stevin, Kepler, Galileo, Cavalieri, Torricelli, Roberval, Pascal, Fermat, Descartes, Wallis, and Barrow. The tremendous contributions of Descartes are well known. Fermat came very close to anticipating Newton and Leibnitz. Barrow is important in that he was the mentor of Newton.
Chapter V deals with the works of Newton and Leibnitz, as well as their monumental feud. During this feud Newton often exhibited a cruel and vindictive streak. (There are those who think this aspect of his personality was a source of his power. Others, following Freud, attribute his powers to sexual sublimation. He never married.)
Chapter VI deals with the period of rapid development which followed after the methods of Newton and Leibnitz became widely known. As Newton was the more secretive, the methods and notation of Leibnitz gained the upper hand. The great luminaries of this period were the Bernoullis, Euler, Lagrange and Laplace. Benjamin Robins carried on the work of Newton in his home country, using Newton's notation and methods. However, this increasingly became a rearguard action. During this phase technique progressed at a tremendous rate, but the logical foundations of the calculus remained shaky. Many of these pioneers thought in term of infinitesimals (a type of completed infinity).
Chapter VII deals with the revolution that took place from approximately 1820 to 1870. During this time the foundations of the calculus were completely recast and put on a rigorous basis. The principal names associated with this phase are Cauchy, Riemann and Weierstrass. The results of this revolution were that "infinitesimals" were discarded. These were replaced by the now-familiar epsilon-delta methodology (limits) - a complete triumph for the followers of Eudoxus!
In chapter VIII Boyer seems to express the opinion that with the triumph of the epsilon-delta method the evolution of calculus has been completed. One cannot help but harbor a suspicion that this triumph is ephemeral. There are several reasons for this. Most beginning calculus student instinctively dislike the epsilon-delta formulation as something artificial. Maybe they are right. Just as the method of Eudoxus in geometry was largely made irrelevant by the discovery of irrational numbers, so one feels there may be something "lurking out there" which will "blow away" the deltas and epsilons. In fact, recent research in "non-standard analysis" seems to have rehabilitated infinitesimals so some degree. Finally, it is of great interest that the maximum rate of progress was during the period when infinitesimals (completed infinity) were allowed. Using apparently fallacious methods these pioneers obtained profound results - and rarely made mistakes!
In a lighter vein, an apparently serious problem with infinitesimals is that there appears to be a need for an unending chain of these: first-order infinitesimals, second-order infinitesimals, etc. Between every two "ordinary" numbers (finite magnitudes) lie infinitely many first-order infinitesimals. But, between any two of these lies an infinity of second-order infinitesimals, and so on. This endless chain brings to mind the following jingle: Big fleas have little fleas/ Upon their back to bite 'em /And little fleas have lesser fleas / And so ad infinitum. / Ogden Nash