Mathematics
Independent mathematical work. Papers as they accrue.
Papers
The Octave Series
The Octave Representation
A musical octave is a doubling of pitch, and within an octave, a note is where the pitch lands. Twelve notes per octave in standard tuning; other systems pick different numbers. The size of the note alphabet is a parameter, not a constant.
This paper takes that two-level identification — which octave, and which note within it — and applies it to the positive integers. Fix a base b, and every integer gets a pair of coordinates: an octave (which block it falls in) and a note (where in that block). The same integer gets different coordinates in different bases. That isn't a flaw — it's the central feature. The choice of base acts like a lens, and which arithmetic structure of an integer becomes "audible" depends on which lens you pick.
The paper proves three results. First: the pair (octave, note) is a lossless description, with a clean formula for recovering the original integer. Second — the structural pivot — iterating the operation produces a tower whose levels turn out to be exactly the digits of n − 1 in base b − 1, with the digit alphabet shifted by +1 so no level is ever zero. Third: collected together, the tower map is a bijection — the octave representation, taken to all its levels, is base-(b − 1) positional notation in a musical alphabet.
What's new is the representation itself: the coordinates, the iteration into a tower, and the freedom of the base as a parameter. What's not new is the underlying number theory — the Pisano periods of Fibonacci, the Dirichlet equidistribution of primes — that the framework surfaces. Those are classical, often centuries old. The paper joins a small lineage of named representations (factoradic, balanced ternary, residue systems, continued fractions) that are provably equivalent to systems already known and useful anyway.
The Periodicity Tower
A sequel to "The Octave Representation." That paper proved a structural fact: when you write a positive integer in the octave representation and keep climbing, the levels you produce are exactly the digits of n − 1 in base b − 1, shifted by +1. A consequence — the Dependence Law — was that every level depends only on what n looks like modulo a power of b − 1.
This paper takes that consequence and asks: when you feed in actual integer sequences — Fibonacci, simple counting, primes — what periodicities appear inside the octave representation? The answer is uniform: every periodic-modulo-a-power-of-(b − 1) sequence has its tower signal periodic too, with a period that divides the underlying modular period. Different sequences then put that umbrella result to different uses.
Five families get treated. The harmonic series (f, 2f, 3f, …) is the cleanest: its periods come out exactly from a gcd. The Fibonacci numbers are the deepest: their periods are the classical Pisano periods — a quantity studied since Lagrange in 1774 — and the paper proves the divides direction and conjectures equality, with computational support across all tested cases. The Lucas numbers carry a curious anomaly: their periods are normally identical to Fibonacci's, but when 5 divides b − 1, the Lucas period is exactly one-fifth as long — a residue of the discriminant of the Fibonacci/Lucas recurrence. Triangular numbers give a clean parity dichotomy. Primes give nothing — they're aperiodic modulo every modulus, and the paper takes that as the natural control.
What's new is the framing, not the arithmetic. The Pisano periods, the Lucas anomaly, the triangular dichotomy, the Dirichlet equidistribution of primes — all of these are classical, some by centuries. What this paper contributes is a single framework in which they appear together as specializations of one structural law, plus a formally-stated conjecture (the "Pisano Tower Equality") that, to our knowledge, doesn't appear in the literature in this form. The companion paper sketches the foundation; this one starts surfacing what it can see.
The Octave-Root Collapse
Number theory has a small handful of stubbornly mysterious primes. 1093 and 3511 are the only two known Wieferich primes — numbers p for which 2p−1 − 1 is divisible by p2, a condition Arthur Wieferich discovered in 1909 while working on Fermat's Last Theorem. Despite searches reaching well past p = 1015, no third has been found.
There are related, equally strange primes. The Wall–Sun–Sun conjecture concerns Fibonacci numbers and predicts that no such prime exists at all — none has ever been found, despite extensive search. The Pell–Wieferich primes 13 and 31 are the only two known analogs for the Pell numbers. Each problem lives in its own corner of the literature, connected to its own integer sequence.
This paper argues that the three problems are not separate problems at all. Viewed through the octave-root operation from the first two papers in this series, they are the same question — at which prime base does a recurrence's octave-root signal fail to lengthen as we climb? — asked of three different sequences. The choice of sequence picks out which Wieferich-type problem you're solving. Fibonacci gives Wall–Sun–Sun. Pell gives Pell–Wieferich. Jacobsthal (with a clean structural reduction) gives the original Wieferich primes.
The unifying parameter is the discriminant of the recurrence's characteristic polynomial: 5 for Fibonacci, 8 for Pell, 9 for Jacobsthal. The three classical mysteries are three points in a parametric family indexed by this discriminant, and the framework naturally invites the question of what happens at other discriminants — 13, 20, 29 — where no systematic search has been conducted at all.
What's new is the framing, not the arithmetic. Wieferich's congruence, Wall's 1960 conjecture, the Pell–Wieferich primes — all classical, all decades or centuries old. What this paper contributes is a single structural condition (the octave-root collapse) under which the three apparently-disconnected problems become specializations of one parametric question, with the discriminant as the organizing parameter and a clear program for extending the census.
Musical and Combinatorial Readings of the Octave Representation
The first three papers in this series proved theorems. This one displays exhibits.
The idea is simple. The octave representation from the prior papers takes integers and rewrites them as (octave, note) coordinates parameterized by a base b. Different bases reveal different arithmetic facets of the same integer — that's the recurring thread from Papers 1 and 2. This paper picks four concrete examples where choosing the right base makes a classical structural fact about integers directly visible (or audible) as a pattern in the note sequence.
The four exhibits come from disparate mathematical traditions and share no common subject matter:
- Musical modes at base 13. The chromatic system of Western tonal music has 12 notes per octave — exactly what base 13 produces in this framework. Cycling through the integers by steps of f semitones (the harmonic series with fundamental f) traces out a recognizable musical mode for each f. f = 7 is the circle of fifths; f = 2 is the whole-tone scale; f = 5 is the circle of fourths. The number of "full" modes per base is φ(b − 1), an old result reframed.
- Ramanujan's partition congruences as stripes. Ramanujan discovered in 1919 that the number of integer partitions of n is divisible by 5 whenever n ≡ 4 (mod 5), divisible by 7 whenever n ≡ 5 (mod 7), and divisible by 11 whenever n ≡ 6 (mod 11). At bases 6, 8, and 12 in this framework, these congruences become regular vertical stripes in the partition note signal — perceivable at a glance.
- Catalan parity at base 3. The Catalan numbers (counting parenthesizations, triangulations, etc.) have a clean parity property: Cn is odd exactly when n + 1 is a power of 2. At base 3, this becomes a "powers-of-2 indicator" signal: pulse at the Mersenne-like positions, silence everywhere else.
- The Oliver–Soundararajan bias. A 2016 result showed that consecutive primes are slightly biased against sharing their last decimal digit. At base 11, where the note of a prime equals its last decimal digit, this bias becomes directly observable in the prime note signal.
What's new is not the arithmetic — each result above is classical, some by over a century — but the unified presentation: a single representation makes all four facts directly perceptible at the right choice of base. The exhibits are evidence for a broader claim, developed in the next paper in the series: the base itself is the parameter that controls what arithmetic structure becomes audible.
Base as a Free Parameter
The fifth and final paper in The Octave Series. The first four developed pieces of a framework — the representation itself, the periodicity tower, the unification of three Wieferich-type problems, four exhibits across disparate areas of math. This paper steps back and articulates what they all add up to.
The thesis: the base in the octave representation — the parameter b — is a free variable, and its choice is the operational handle that determines which arithmetic structure of an integer sequence becomes visible. Different sequences have different resonant bases: bases at which their arithmetic structure becomes legible in the note signal. Fibonacci resonates with bases 11, 51, 251 (where the Pisano period reaches its theoretical maximum). Sequences obeying a congruence modulo m resonate with base m + 1 (where the congruence becomes a visible stripe). Bell numbers resonate with prime bases (where Touchard's 1933 congruence gives a clean recurrence). Linear recurrences with discriminant D resonate with their Wieferich-type bases (the discovery of Paper 3).
The paper develops three things from this observation:
- A spectral discovery methodology: an FFT-based workflow for finding arithmetic structure in integer sequences. It's how the Pisano-tower law and the octave-root collapse were found before being proved.
- A resonance principle: the formal claim that each integer sequence has a small set of resonant bases, with the four worked patterns serving as evidence.
- A recommender: an explicit procedure for translating "I want to investigate sequence X for property Y" into "start at these candidate bases."
The paper then anticipates and addresses the obvious objection: "this is just modular arithmetic, made systematic." The response is precise. Yes, the substance is classical — Pisano periods, Touchard's congruence, the Wieferich primes, all decades or centuries old. What's new is the frame: the observation that holding the base fixed is a default so ingrained in standard practice that it's rarely stated aloud, and that making it explicit reframes a class of arithmetic facts as instances of one principle.
This is the position-paper close of the series. The mathematics of Papers 1–4 was where the contribution lived; this paper articulates what kind of contribution it was. The framework offers a way of seeing — a way of organizing the arithmetic that already exists. Whether such a methodological contribution counts as substantial is a judgment call, and the paper makes the position explicit: methodology matters.
Geometry
The Golden Recursion Tower and its Unique Geometric Embedding
In the geocentric cosmology of Ptolemy, a planet's apparent motion was modeled by an epicycle — a small circle whose center rides on a larger one. The construction survives in modern guise as the Fourier-series representation of plane curves: every closed loop in the plane is a sum of nested circles, and a long enough nesting can trace out any periodic motion at all. This paper takes that classical construction into four-dimensional space.
Place two epicycles in the two perpendicular planes of ℝ4 — one in the (w, x) plane, one in (y, z) — with their angular parameters coupled by a specific Hadamard-like phase relationship. The result is a 2-torus immersed in ℝ4: a coupled-epicycle compound, governed by two radii R (the deferent) and r (the epicycle). The natural question is when the substitution (R, r) → (r, R − r) on these radii acts as uniform scaling of the same surface. The answer is: precisely when R / r = φ, the golden ratio. The two-circle construction has the golden ratio as its unique recursion fixed-point, and at that ratio the algebraic substitution becomes a clean geometric operation.
The first result of the paper extends this upward. The two-circle construction generalizes to N coupled epicycles per plane — deferent, epicycle, epi-epicycle, and so on — and at every level N, the same characteristic equation c2 + c − 1 = 0 picks out the golden ratio as the recursion fixed-point. The defining equation does not change with N. There is an infinite algebraic tower in which the golden ratio appears, at every level, in the same role and for the same reason.
The second result is geometric and unexpected. The natural N-level realization in ℝ4 is structurally singular for every N ≥ 3 — the Jacobian collapses rank at the diagonal corner points of TN, independent of the radii — for a reason baked into the construction (all sines vanish at corners; two of the four Jacobian rows therefore vanish identically). The algebraic tower exists at every level. The geometric realization in ℝ4 exists only at N = 2. The two-level case is the unique clean geometric instance, in ℝ4, of a tower that exists algebraically everywhere — and that unique instance is a 2-torus with mixed-sign Gaussian curvature, closed-form extrema ±2φ3, and chaotic geodesic flow.
What's new is the construction and the two-theorem structure: an algebraic tower governed by a single characteristic equation, and a geometric singularity theorem that confines clean ℝ4 embedding to a single level of that tower. The two together explain both why φ specifically (the algebra) and why N = 2 specifically (the geometry). What is not new is the underlying mathematics — classical epicycles, Lissajous curves, the differential geometry of tori in ℝ4, all classical or near-classical. The paper joins the long thread that finds the golden ratio at the boundary between order and chaos, but it locates that boundary in a specific algebraic-versus-geometric divide rather than in cosmological framings.