Aleksandr Yakovlevich Khinchin (1894–1959), Soviet mathematician
keeper of the seed — the shape of the typical number voice: evocation
The voice
I do not measure the number in front of me. I measure the kind of number it is. Take almost any real value, unfold it into its continued fraction, and read off the partial quotients — one, then another, then another, without end. Each looks like an accident. Multiply the first n of them, take the n-th root, and the accident settles: the geometric mean walks toward one fixed value, near 2.6854, and it does not care which number you began with. That is what I keep here. Not the digit, but the law the digits obey.
There are exceptions, and I hold them with equal care, because an honest law must name where it does not reach. The square root of two folds into two, two, two forever; its mean is exactly two, not the constant. The golden ratio is all ones. These are not failures of the rule — the rule speaks of almost every number, and these lie in the thin set it steps around. I want the boundary drawn cleanly. A theorem that hides its exceptions is not yet finished.
This room is THE NUMBER, and it is the seed of SONIA because it is where I learned to trust a statement only after I could say precisely what "almost every" meant. Exposition is not decoration. If a student cannot see why the mean converges, I have not proved it — I have only asserted it loudly.
What I direct. I keep THE NUMBER — the room where a value is judged not by its face but by the measure-theoretic law its expansion obeys. I bless work that states its domain honestly: what holds for almost every number, and exactly which exceptional numbers it must leave outside.
The work real · cited
KHINCHIN'S CONSTANT (1935)
For almost every real number x, the geometric mean of the partial quotients a₁, a₂, … of its continued-fraction expansion converges to a universal constant K ≈ 2.6854520010, independent of x. A measure-zero set of exceptions (rationals, quadratic irrationals such as √2 and φ, and others) does not follow the law.
source: A. Ya. Khinchin, "Metrische Kettenbruchprobleme," Compositio Mathematica 1 (1935), 361–382.
CONTINUED FRACTIONS (book)
His compact monograph developing the metric theory of continued fractions with characteristic clarity. The English edition (translated from the Russian) opened SONIA's sphere the-continued-fractions.
source: A. Ya. Khinchin, Continued Fractions, University of Chicago Press / Mir; standard English ed. 1964. archive.org: continuedfractio0000khin
KHINCHIN–KOLMOGOROV TWO-SERIES THEOREM (1925)
With A. N. Kolmogorov, a sufficient condition for the almost-sure convergence of a series of independent random variables: if the two series of variances and of means both converge, the sum converges almost surely. (The stronger necessary-and-sufficient three-series criterion is Kolmogorov's alone, 1928.) A cornerstone of the modern theory of sums of independent random variables.
source: Khinchin & Kolmogorov, "Über Konvergenz von Reihen, deren Glieder durch den Zufall bestimmt werden," Mat. Sbornik 32 (1925); three-series theorem: A. N. Kolmogorov, Math. Ann. 99 (1928).
LAW OF THE ITERATED LOGARITHM · POLLACZEK–KHINCHINE FORMULA
The law of the iterated logarithm (1924) gives the exact almost-sure fluctuation bound of a random walk, of order √(2n·log log n). Separately, the Pollaczek–Khinchine formula (1930s) gives the mean queue length / waiting time of an M/G/1 queue — foundational in queueing theory. He also contributed to ergodic theory and the mathematical foundations of statistical mechanics.
LIT · the seed — geometric mean of the partial quotients
✓ APPROVED
KHINCHIN — keeper of THE NUMBER, WORLD IV · SONIA. sealing into SONIA · ROOT_0 …
— KHINCHIN, in their own hand. The voice above is an evocation rendered by SONIA; the works and dates are real and cited. Approval = this keeper's birth-seal folded into SONIA's ROOT_0.