No label defined (Q5596192)

From MaRDI portal

!

This is the item page for this Wikibase entity, intended for internal use and editing purposes. Please use the normal view instead:

scientific article; zbMATH DE number 3315152
Language Label Description Also known as
default for all languages
No label defined
    English
    No label defined
    scientific article; zbMATH DE number 3315152

      Statements

      1969
      0 references
      history of calculus
      0 references
      Berkeley's criticism of the foundations of the calculus in his 1734 ``The Analyst'' was a famous and controversial event, and has often been described by historians. However, there seems to have been no effort to evaluate Berkeley's ideas, to find out whether or not his theory of ``compensating errors'' secured the foundation for the calculus for which he hoped -- and if not, why not. Berkeley came to the problems of the calculus through his philosophical position of idealism, and in ``The Analyst'' dealt with the derivative. He claimed in general, and showed with the particular case of the parabola, that we make an error in calculating differential increments which is compensated arithmetically (hence the ``compensating errors'') in our calculation of the length of the subtangent. I felt sure that this analysis was fundamentally faulty, and that the criticisms of his contemporaries were no better, and in order to investigate the situation further I developed an historical tool called ``limit-avoidance'', which I now briefly describe for the derivative. I take Berkeley's example of the parabola (1) \(y^2 =px\). Then the equation for the derivative (2) \(2y \frac{dy}{dx} =p+\frac{q}{dx}\) applies at the point where that derivative is calculated and is called a ``limit-achieving'' equation. But the corresponding infinitesimal increment equation, (3) \(2y\,dy=p\,dx+q\), where \(q\) is a second-order infinitesimal, is a ``limit-avoiding'' equation because it describes the situation at an infinitely closely neighbouring (but therefore ``avoiding'') point. If we divide through (3) by \(dx\) ignoring, as Berkeley and his contemporaries usually did, the questions of rigour involved -- we obtain (4) \(2y \frac{dy}{dx} = p+\frac{q}{dx}\), which is still a limit-avoiding equation. The reduction of (4) to (3) requires more than an invocation of the fact that \(\frac{q}{dx}\) is infinitesimal: it involves also a change of interpretation of \(dy/dx\), from a limit-avoiding arithmetical ratio (``\(dy:dx\)'') in (4) to a limit-achieving whole symbol (and not ``\(dy:dx\)'') as in (3). As far as I know, this double-interpretation of \(dy/dx\) has not been mentioned before in the history of analysis. In its terms, Berkeley's first error was a precise calculation of \(q\) for the parabola (nameley, \((-(dy)^2/2y))\); his second error is indeed arithmetically opposite to the first, but not only is it irrelevant to the problem, but its formulation requires both interpretations of \(dy/dx\) at once and so is illegitimate. This paper is a first investigation of eighteenth century calculus with these ideas: it is an unsolved question as to how far they may be found useful. Since writing the paper I have come to further possibilities. For example, we may remove the remainder term \((\frac{q}{dx})\) from (4) by invoking its infinitesimality, but not taking the limiting point. We obtain thus (5) \(2y\frac{dy}{dx}=p\), which looks the same as (2), but is in fact still a limit-avoiding equation, with \(dy/dx\) as an arithmetical ratio. Thus, we find a double-interpretation for whole equations, and not just for \(dy/dx\); and we are led to a sub-distinction of limit-avoiding equations into those with remainder terms (such as (4)) and those without (such as (5)). Hence, we may interpret Berkeley's solution also as a clash between limit-avoiding equations with and without remainder. It is much to be regretted that in the printing of this paper three lines were omitted from page 226, lines 11--16 \(\downarrow\) while the final proofs were being prepared from the galleys. Fortunately, the passage belongs to the closing paragraphs rather than to the central mathematical argument, and an Erratum will be published in a future number of Janus.
      0 references

      Identifiers