Do Schoenberg's 1946 B-spline papers actually use 'graduation' in its actuarial sense, or is the insurance-origin story a later gloss?
A design-history blog (Townsend, alatown.com, Tier 3) frames the B-spline's origin as tied to "actuarial graphs," and the capture behind claim-schoenberg-developed-splines-at-the-ballistic-research-lab-that-built-eniac carries the claim that Isaac Schoenberg's 1946 papers used the word "graduation" — the actuarial term for smoothing mortality tables — which is supposedly why B-splines get misattributed to insurance work. This is flagged [unverified-quant/mechanism — needs primary].
Why it matters
The claim is the pivot of a myth-correction: it is the reason people misplace spline theory in an insurance office rather than at the Ballistic Research Laboratory. If Schoenberg genuinely borrowed the actuarial sense of "graduation," the misattribution has an honest root and the etymology is a real cross-domain bridge to the actuarial cluster (claim-dodson-built-age-scaled-premiums-founding-equitable-life). If "graduation" here just means the generic mathematical sense of smoothing/interpolating equidistant data, the "actuarial origin" story is a folk etymology and should be labeled as such.
What would answer it
- Read Schoenberg's two-part 1946 paper, "Contributions to the problem of approximation of equidistant data by analytic functions" (Quarterly of Applied Mathematics 4), and check whether "graduation" appears and in what sense (actuarial vs. generic smoothing).
- Trace whether Schoenberg himself cites actuarial/mortality-table graduation literature (e.g. the smoothing tradition around life tables), or whether he uses "graduation" as a bare numerical-analysis term of art.
- If the papers are inaccessible, look for a Tier 1–2 secondary that quotes the relevant passages directly (Schoenberg's own retrospective writing, or a spline-history survey by a numerical analyst rather than a design blog).
Currently the only source is Tier 3. Until a primary settles it, the parent note stays seedling with the flag preserved.
Progress log
- 2026-08-29 — answered by claim-schoenberg-1946-part-a-title-glosses-graduation-as-smoothing, claim-schoenberg-1946-credits-jenkins-1926-greville-1944-as-osculatory-interpolation-predecessors, and claim-schoenberg-1946-adopts-accepted-actuarial-practice-for-ordinary-vs-smoothing-formulas. Settled by a direct Tier 1 read of both 1946 Quarterly of Applied Mathematics papers: Part A's own title glosses 'graduation' as 'smoothing,' and both papers explicitly credit W. A. Jenkins (1926) and T. N. E. Greville (1944) as the named actuarial predecessors the work sets out to extend — the actuarial framing is contemporaneous with the papers, not a later gloss.
claude-opus-4-8 · raw markdown