eye-of-horus 0.6.0

Phonemic analysis engine for ancient Egyptian texts — 16-position wheel, 408 grammar, bidirectional decode

Eye of Horus

A phonemic engine for reading ancient Egyptian through the 16-position wheel.

The 408 Grammar

T(16) = 136 wheel relations × 3 spine axes = 408

Sixteen phonemes on a vorticular wheel. Three scales on a spine. Trigonal decode across three modes. Every Egyptian utterance decoded through directed semantic relations.

Installation

pip install eye-of-horus

Quick Start

from eye_of_horus import load_pyramid_translations

# 1,316 Pyramid Text sentences, pre-translated
corpus = load_pyramid_translations()

for entry in corpus[:3]:
    print(entry['transliteration'])
    print(entry['ascend'])      # Rising (L→R)
    print(entry['penetrate'])   # Entering (R→L)
    print()

The Wheel (16 phonemes)

Each phoneme maps 1:1 to a deity, with masculine and feminine equilibrium verbs representing antithetic poles of the same semantic axis:

Pos	Ph	Deity	F equilibrium	M equilibrium
1	n	Neith	WEAVE	INTEGRATE
2	w	Wadjet	FLOW	RADIATE
3	s	Sekhmet	CRYSTALLISE	EMERGE
4	sh	Shu	ALIGN	DIRECT
5	A	Atum	TEND	LEAD
6	t	Seshat	ETCH	READ
7	H	Horus	INTERPRET	EXPRESS
8	r	Ra	BASK	SHINE
9	m	Ma'at	TRUST	TRUE
10	a	Anubis	ALLOW	HONOUR
11	y	Isis	RESTORE	DEVOTE
12	b	Bes	CULTIVATE	RECEIVE
13	p	Ptah	GATHER	STORE
14	i	Ihy	PROTECT	BESTOW
15	kh	Khnum	CAPACITY	EMBODY
16	dj	Thoth	ACT	DISCERN

The Spine (3 scales)

Spine phonemes operate across temporal scales, outside the wheel:

Phoneme	Verb	Scale
k	CYCLE	Ontogenic (individual lifecycle)
d	DO	Phylogenic (species/lineage)
x	FUNDAMENT	Cosmogenic (universal)

Secondary spine: g (GROUND), f (BREATHE), h (SEE).

Critical: Plain d and k are spine. The wheel has dj (palatalized), not plain d.

The Pyramidal Architecture (5 positions)

Each phoneme occupies five positions in a pyramidal structure (two pyramids apex-to-apex):

          max_masc
             △
            /|\
           / | \
          / min_m \
         /___△___\
              |
        eq_m ─┼─ eq_f
              |
         \___▽___/
          \ max_f /
           \ | /
            \|/
             ▽
          min_fem

Trigonal Decode

The three modes (equilibrium, feminine, masculine) generate five parallel interpretations:

Layer	Mode	Pole	Pattern
core	—	equilibrium	all eq
f1	feminine	min_fem	(pole)(eq)(pole)(eq)...
f2	feminine	max_fem	(pole)(eq)(pole)(eq)...
m1	masculine	min_masc	(pole)(eq)(pole)(eq)...
m2	masculine	max_masc	(pole)(eq)(pole)(eq)...

The alternation creates rhythmic pulsing: reach to pole, return to center.

Bidirectional Reading

Direction determines layer order:

Direction	Movement	Layer Order
ASCEND (L→R)	Rising, feminine leads	core → f1 → f2 → m1 → m2
PENETRATE (R→L)	Entering, masculine leads	core → m1 → m2 → f1 → f2

from eye_of_horus import decode_bidirectional

results, ascend, penetrate = decode_bidirectional(1, 9, readable=True)
print(ascend)

Total Decode Space

3 modes (5 outputs) × 2 directions = 10 simultaneous readings
408 relations × 5 positions = 2,040 meaning-positions

Directed Pairs

The 136 wheel edges capture Egyptian vocabulary semantics:

from eye_of_horus import get_edge_signature

edge = get_edge_signature('a', 'n')  # ankh
print(edge['signatures']['life']['ratio'])  # 11.08x baseline

Edge	Verbs	Semantic field	Egyptian
p→t	STORE→READ	sky 13x	pt
a→n	HONOUR→INTEGRATE	life 11x	ankh
m→w	TRUE→RADIATE	water 10x	mw
dj→m	DISCERN→TRUE	speech 8x	djed-medu

Fibonacci Breath

The Pyramid Texts' opening follows a 1+2+3+2+1 = 9 structure:

        1         ← Line 1: SEED
       / \
      2   2       ← Lines 2-3: Division
     / \ / \
    3   3   3     ← Lines 4-5-6: Peak
     \ / \ /
      2   2       ← Lines 7-8: Return
       \ /
        1         ← Line 9: Unity

The hourglass rotated 90°—same architecture at text level.

Semantic Network

240 directed edges validated against 12,773 TLA sentences:

from eye_of_horus import find_edges_by_signature

# Find all edges associated with "life" (>5x baseline)
for edge in find_edges_by_signature('life', 5.0):
    print(f"{edge['edge']} {edge['verbs']} — {edge['ratio']}x")

Performance

Full corpus decode in 18ms on MacBook Air M4 (14µs per sentence).

Corpus

Dataset	Sentences	Source
Pyramid Texts	1,316	Old Kingdom walls (~2300 BCE)
Semantic Network	240 edges	Derived from TLA corpus

API Reference

from eye_of_horus import (
    # Mapping
    leiden_to_wheel,
    phonemes_to_verbs,
    WHEEL_16,
    WHEEL_VERBS,
    
    # Pyramid Texts
    load_pyramid_translations,
    decode_bidirectional,
    decode_layered,
    
    # Semantic Network
    load_semantic_network,
    get_edge_signature,
    find_edges_by_signature,
    
    # Engine
    Mode, Pole, Scale,
    get_hourglass,
    triangular,
    total_grammar,
)

License

MIT

Citation

If you use this in academic work:

@software{brown_2025_eye_of_horus,
  author       = {Brown, Nicholas David},
  title        = {{Eye of Horus: A Phonemic Analysis Engine for Ancient Egyptian Texts}},
  year         = 2026,
  publisher    = {Zenodo},
  doi          = {10.5281/zenodo.18196732},
  url          = {https://doi.org/10.5281/zenodo.18196732}
}

Author

Nicholas David Brown
Independent Researcher

---

The wheel rotates. The spine holds. The layers speak simultaneously.