Unit I: Number and Magnitude — Scholarly Notes¶
Philological Collation, Symbolic Realization & Cross-Repository Commentary
1. Philological Collation: The Kronecker vs. Dedekind Textual Variant¶
Collating the 1902 Flammarion French text (Poincaré 1902, pp. 31–40) against Greenstreet’s 1905 English translation (Greenstreet 1905, pp. 19–27) uncovers one of the most instructive textual strata in La Science et l’Hypothèse:
In Poincaré’s original January 1893 essay “Le continu mathématique” (Revue de Métaphysique et de Morale, Vol. 1, pp. 26–34), Poincaré inadvertently wrote Kronecker where he described Richard Dedekind’s 1872 Stetigkeit und irrationale Zahlen definition of irrational numbers by cuts (Schnitte / coupures) of $\mathbb{Q}$.
When revising the essay for Ernest Flammarion, Poincaré corrected the attribution on French pp. 32–33 to M. Dedekind (« Voyons donc quelle est la définition de M. Dedekind » and « au point de vue de M. Dedekind, le nombre incommensurable $\sqrt{2}$… »), while overlooking a single residual occurrence on p. 40 (« Nous reconnaissons là la conception de Kronecker »).
Greenstreet’s 1905 translation—working from an early impression retaining the 1893 reading—prints “Kronecker” at all three places (EN pp. 20, 21, 26), whereas Halsted’s 1905 American edition (Halsted 1905) follows the corrected Flammarion text (Dedekind on pp. 32–33). Both witnesses are preserved verbatim in their respective tracks here so the editorial evolution remains transparent.
2. Poincaré’s Synthetic A Priori Recurrence vs. Russell & Hilbert¶
Chapter I forms the opening salvo in Poincaré’s two-front foundational debate documented across the GEOMETOR Scriptorium (Concordance PC.01):
Against Logicism (``geometor/bertrand-russell``): In The Principles of Mathematics (1903; cf. Russell 1897 and Russell 1899), Bertrand Russell (following Frege and Peano) defined the natural numbers as the intersection of all inductive classes, claiming to reduce mathematical induction to an analytic definition. Poincaré countered (later expanded in Science et Méthode, 1908) that verifying the consistency and non-emptiness of that impredicative definition covertly presupposes the very synthetic a priori intuition of finite iteration it purports to eliminate (petitio principii).
Against Pure Formalism (``geometor/hilbert`` & ``geometor/hoenen``): In his Special Preface to Halsted’s translation above, Poincaré remarks on Hilbert’s Grundlagen der Geometrie (Hilbert 1899) that teaching geometry purely axiomatically “is surely this time to burn all bridges of reliance upon sensory intuition.” Crucially, Hilbert’s own metamathematical consistency proofs rely at every step on finite combinatorial induction—confirming Poincaré’s thesis (echoed from a Thomistic standpoint by P. H. J. Hoenen in De Noetica Geometriae, Caput I § 6) that formal axiomatics presupposes intuitive arithmetic recurrence.
3. Exact Symbolic Realization in geometor.poincare (Fig. 1)¶
In geometor.poincare.disk.construct_dedekind_continuum_cut(), Chapter II’s crucial example (FR p. 39 / EN p. 26)—that over the first-order continuum $\mathbb{Q}^2$ a circle and a square’s diagonal miss one another without Dedekind cuts—is modeled with zero floating-point approximations:
Over $\mathbb{Q}$, the perceptual threshold points $Q_1 = (2/3, 0)$ and $Q_2 = (3/4, 0)$ satisfy $Q(O, Q_1) = 4/9 < 1/2 < 9/16 = Q(O, Q_2)$, illustrating Fechner’s indistinguishability band around the cut.
In
geometor.model, intersecting the exact diagonal line $x - y = 0$ through $O(0,0)$ and $S(1,1)$ with the unit horizon circle $\Omega: x^2 + y^2 = 1$ creates the quadratic extension $\mathbb{Q}(\sqrt{2})$ and yields the exact cut point $D = (\frac{\sqrt{2}}{2}, \frac{\sqrt{2}}{2})$ with rational quadrance $Q(O, D_x) = \frac{1}{2}$ verified symbolically viasp.simplify.