Traditional Methods & Sources¶
This is the canonical record of what Stellium’s traditional (Hellenistic / Perso-Arabic) techniques actually compute, and on whose authority. For each technique it states the default, the key citations, the genuinely contested forks (and which Stellium ships as default and why), and any simplifications — with a link to the full source-verification report behind it.
It sits between three neighbours and overlaps none of them:
Doc area |
Answers |
|---|---|
How do I use this technique as a practitioner? (teaching) |
|
|
What does Stellium implement, and from which sources? (scholarship) |
How is it built in the code? (architecture) |
Design specs in docs/development/specs/ capture a
change before it’s built; this doc is the durable sourcing that outlives them.
A principle throughout: where the tradition genuinely disagrees, Stellium
surfaces the fork as a toggle or preset rather than silently choosing —
and documents which default it ships and why. The raw investigations (with
verbatim quotes and full bibliographies) live in research/.
Planetary years¶
The shared primitive (stellium.core.planetary_years) behind ZR, Firdaria, and
length of life. Full report: research/planetary-years.md.
Least / minor years (Sun 19, Moon 25, Mercury 20, Venus 8, Mars 15, Jupiter 12, Saturn 30; sum 129) — canonical and near-universal. Valens, Anthology III/IV; Firmicus, Mathesis II.29. What ZR consumes.
Greater years (Sun 120, Moon 108, Saturn 57, Jupiter 79, Mars 66, Venus 82, Mercury 76) — the five non-luminary values are the sum of each planet’s Egyptian terms across the zodiac (they total 360). Ptolemy, Tetrabiblos I.20; Houlding (Annotated Lilly, fn 154). Stellium verifies this against its own term tables in the test suite.
Mean years — derived as
(least + greater) / 2(half-integer luminaries Sun 69.5, Moon 66.5), never hand-typed. Skyscript/Houlding; the alcocoden Table of Years.Greatest years — genuinely contested; stored as attributed variants with no default. De Vore vs. the astronomical reconstruction (Neugebauer/ Houlding); the Moon is the flashpoint (25 / 320 / 420 / 520 across Antiochus-Rhetorius / Lilly / Bonatti / later-Arabic).
Firdaria periods are a separate table (see below) — not the least years.
Correction shipped: the popular claim that Capricorn’s ZR period of 27 =
“¼ of the Moon’s greater years (108/4)” is folk-etymology, not in Valens —
the arithmetic is coincidental and no source connects it to the sign-periods.
Stellium’s code comment was corrected accordingly. (The sourced symbolic note,
if wanted, is Manwaring’s “27 lunar mansions.”) Code: engines/releasing.py.
Dignity-data fix it surfaced: the term-sum cross-check caught a transposed
pair of Egyptian terms in Sagittarius (Saturn/Mars at 21–30°), now corrected
(Ptolemy I.20; Lilly). Code: engines/dignities.py.
Zodiacal Releasing¶
Full report: research/zodiacal-releasing.md.
Single source: Vettius Valens, Anthology Book IV (chs. 4–10, Riley numbering). Named “Zodiacal Releasing” and reconstructed by Robert Schmidt (Project Hindsight, 1996); popularized by Chris Brennan (Hellenistic Astrology, 2017). Also Demetra George (Vol. II), Manwaring (Delphic Oracle).
Stellium default = the “Brennan / Modern Standard”: release from the Lot (Spirit for career, Fortune for body), whole-sign tropical, 360-day years, Capricorn 27 / Aquarius 30, sect-reversed lots, loosing of the bond to the opposite sign after a full 12-sign circuit, truncation at parent boundaries, peaks measured from Fortune, same-sign Spirit rule.
Contested forks & our choice:
Fork |
Positions |
Stellium |
|---|---|---|
Reverse Fortune/Spirit by sect |
Valens/Schmidt/Brennan: yes · Ptolemy/Lilly: no |
yes (default; toggle) |
Period → date year length |
Valens/Brennan: 360-day · minority: 365.25 |
360 ( |
Capricorn period |
27 (standard) · 30 (minority) |
27 ( |
Loosing target |
opposite (Valens) · trine (some contemporaries) |
opposite ( |
Code: engines/releasing.py (ZodiacalReleasingEngine), chart.zodiacal_releasing().
Firdaria¶
Full report: research/firdaria.md.
Period values (Sun 10, Venus 8, Mercury 13, Moon 9, Saturn 11, Jupiter 12, Mars 7, Head 3, Tail 2; sum 75) and the sect-ordered Chaldean sequence are near-universal. Abu Ma’shar (Dykes, Persian Nativities); al-Biruni; al-Qabisi; Bonatti; Lilly.
Sub-periods: seven equal parts in Chaldean order from the major ruler; node majors don’t subdivide and nodes aren’t sub-rulers (Abu Ma’shar).
Stellium default = the “Abu Ma’shar / Persian” preset: nodes at the end in both sects, 7 equal sub-periods, repeat past 75, real 365.2425-day year (the 360-day year is mundane-only, not natal firdaria).
The one real fork — nocturnal node placement, shipped as presets:
Preset |
Nocturnal nodes |
Basis |
|---|---|---|
|
at the end (both sects) |
Abu Ma’shar / al-Biruni / Hand / Birchfield |
|
after Mars (~ages 39–44) |
Bonatti / Zoller (from an ambiguous al-Qabisi paraphrase) |
|
at the end (alias of default) |
— |
|
none (70-year cycle) |
practitioners who reject the node periods |
Deferred (noted, not built): the AB Method (Subramanyan) — proportional
sub-periods with the nodes participating. Code: engines/firdaria.py,
chart.firdaria().
Length of life (hyleg / alcocoden)¶
Full report: research/planetary-years.md (alcocoden section). A computed traditional indicator, not a prediction of actual lifespan — the result is fully itemized so the reasoning is auditable.
Method = Perso-Arabic years table,
method="lilly"default. The years-table apparatus is Perso-Arabic (Māshā’allāh, Sahl, Abū Ma’shar, al-Qabisi) synthesized into Bonatti and Lilly (via Houlding’s annotated edition); Ptolemy’s rival directional method (Tetrabiblos III.10–11) has no years table and is reserved but not implemented (method="ptolemy"raises).Hyleg: sect-ordered candidates (sect light, other light, Lot of Fortune, prenatal Syzygy, Ascendant backstop); first in a hylegiacal place wins.
Alcocoden: the almuten of the hyleg’s degree that beholds it; grants years by angularity — angular → greater, succedent → mean, cadent → least.
Modifiers (Lilly): benefics add / malefics subtract their least years, the lights add by soft aspect / subtract by hard, fall or retrograde halves, combustion → months.
Documented simplifications (v1):
Hylegiacal places use whole-house membership (the degree-based Ptolemaic boundaries, e.g. 5° above the Asc, are a noted future refinement).
The Full-Moon syzygy uses the Moon’s degree (the “luminary above the horizon” refinement isn’t applied).
Year modifiers use real moiety-orb aspects (not whole-sign, which over-counts) and the total is floored at 0.
Lilly-only; a Bonatti preset (differing modifier list) is a future addition.
Known behaviour: the years table tends to overshoot for a strong (angular) alcocoden — this is the technique, not a defect.
Code: engines/length_of_life.py, chart.length_of_life() / chart.hyleg().
Almuten of a degree¶
The essential-dignity victor over a longitude (domicile 5, exaltation 4,
triplicity 3, term 2, face 1). Ibn Ezra / Lilly weighting. Triplicity is
sect-dependent; node exaltations are never scored. Distinct from
get_strongest_planet() (almuten among placed planets). Code:
engines/almuten.py.
Recurring sources¶
Primary texts and the translations/editions Stellium’s implementations lean on:
Vettius Valens, Anthology — Mark Riley (CSU Sacramento PDF; print ed. Brennan, 2022); Robert Schmidt (Project Hindsight, 1996).
Ptolemy, Tetrabiblos I.20–21 (terms/greater years), III.10–11 (length of life).
Firmicus Maternus, Mathesis II.29.
Abū Ma’shar, On the Revolutions of the Years of Nativities — trans. Benjamin Dykes, Persian Nativities.
al-Bīrūnī, Book of Instruction — trans. R. Ramsay Wright (1934).
al-Qabīsī, Introduction to Astrology — trans. Dykes.
Guido Bonatti, Liber Astronomiae — trans. Dykes.
William Lilly, Christian Astrology — Deborah Houlding’s annotated edition (Skyscript) for the greater/mean/least tables and the term-sum derivation.
Modern: Chris Brennan (Hellenistic Astrology, 2017), Demetra George (Ancient Astrology in Theory and Practice), Robert Hand (ARHAT), Curtis Manwaring (Delphic Oracle; the “27 lunar mansions” note), Steven Birchfield.
Full per-parameter attribution, verbatim quotes, and the contested-point
evidence are in the three reports under research/.