The datacentral project - research

Papers on social chaos, hermeneutics, computational culture, and personal knowledge ecosystems.

Navigation:
Research | Datacentral | History

Section 1 - What are ideas

Introduction

Ideas are often treated as abstractions—passive contents of thought or linguistic units. Yet across disciplines, from evolutionary theory to hermeneutics and cognitive science, a different picture emerges: ideas behave like dynamic, adaptive structures that persist, mutate, and organise themselves across minds. This section synthesises several perspectives to develop a working model of “living ideas,” which will serve as the foundation for the rest of the essay.

1.1 The selfish gene

Richard Dawkins’ The Selfish Gene (1976) reframed evolution by shifting the locus of selection from organisms to genes. Genes, in this view, are replicators—entities that persist through time by making copies of themselves. Organisms are “survival machines” constructed by genes to ensure their continued replication.

Dawkins extends this logic to culture through the concept of the meme: a unit of cultural transmission that spreads through imitation. Memes do not replicate perfectly; each mind interprets and reshapes them. What persists is not the exact form of an idea, but a pattern stable enough to survive repeated reinterpretation. If ideas behave like replicators—transmitted, transformed, and preserved across minds—then they can be studied as dynamic participants in cultural evolution, competing for limited cognitive and social resources. Additionally, similar to genes memes evolve in complexes—clusters of mutually reinforcing ideas that stabilise one another. Throughout this essay, I argue that just as genes create organisms as survival machines, memes create their own higher‑order memetic structures.

1.2 Interpretation as the Mechanism of Replication

The study of interpretation—hermeneutics—offers a framework for understanding how ideas replicate. A central concept is the hermeneutic circle: understanding moves between parts and wholes, shaped by the interpreter’s prior assumptions. These assumptions, or “prejudices” in Gadamer’s sense, guide interpretation but are also revised by it. Meaning is therefore not static; it is continually reconstructed.

Neuroscience provides a complementary model. The Thousand Brains Hypothesis suggests that the neocortex consists of many parallel columns of neurons (cortical columns), each learning partial models of the world. Ideas correspond to specific configurations of these cortical columns, and new experiences update both the columns themselves and the patterns of interaction between them.

Ideas replicate by inducing new neural configurations in another mind. Interpretation is therefore the mechanism through which ideas mutate, stabilise, and propagate.

1.3 A Simple Model of Intelligence

Chaos theory describes systems that are highly sensitive to initial conditions, where small changes in input can produce large changes in output.Intelligent behaviour can be understood as an emergent property of such systems—patterns of stability that arise from complex, nonlinear interactions rather than from the behaviour of individual components. For the purposes of this essay intelligence is defined as the capacity of a chaotic system to stabilise patterns against disruption by responding adaptively to internal and external changes.This might seem arbitrary however this definition captures the essence of a system that internalises external patterns—a primitive form of learning. Additionally, replicators can be seen as primitive implementations of this stabilising intelligence. If memes are replicators, then memes can be understood as a form of proto‑intelligence or proto‑life: structures that maintain themselves by shaping the systems that host them.

Section 2 - the history of the idea of living ideas

Introduction

Long before memetics or cognitive science, human cultures imagined ideas, symbols, and deities as possessing life and agency. Ancient religions, esoteric traditions, and philosophical systems all described ideas as forces that act upon the world. This section traces those traditions and shows how they anticipated, prefigured, or parallel the framework developed in Section 1.

2.1 Historical background: Rosicrucianism and gnosticism

Early Christianity was a diverse landscape of competing interpretations. Among these were Gnostic traditions, which described a cosmology of emanations from a transcendent source (the Monad), culminating in a lesser creator—the Demiurge—responsible for the material world. Though suppressed by emerging orthodoxy, Gnostic themes resurfaced in later esoteric movements.

Rosicrucianism, emerging in the early modern period, blended the traditions of Christian mysticism, Neoplatonism, Hermeticism, and Kabbalah. While not a direct continuation of Gnosticism, it inherited similar symbolic structures: layered realities, emanations, and the transformative power of hidden knowledge. These traditions expanded the notion of a “spiritual plane” and introduced conceptual layers—such as the “mental plane”—that framed ideas as active forces within a metaphysical hierarchy.

These esoteric systems treated ideas not merely as static abstractions but as dynamic agents—entities capable of influencing behaviour, shaping perception, and interacting with human consciousness. This anticipates the memetic notion of ideas as active participants in cultural evolution.

2.2 Egregores

Origins in the Book of Enoch

The term egregore originates in the Book of Enoch, where the Greek egregoroi refers to “the watchers”—angels tasked with observing humanity.In this text, some watchers become corrupted, presenting themselves as gods and teaching forbidden knowledge, including magical practices.This story shaped early Christian attitudes toward pagan deities and magic, associating egregores with powerful, idea‑like entities.

Transformation in Western Esotericism

In later esoteric traditions, the egregore evolved into a symbol of collective psychic energy—an entity formed by the shared beliefs and emotions of a group. Detached from its mythological origins, the egregore became a metaphor for how ideas can acquire autonomy and exert influence independent of any single individual.

In this essay, the egregore is abstracted further: the egregore is treated as a memetic structure: a network of mutually reinforcing memes that behaves like a living system.

Thoughtforms, Godforms, and Tulpas

Theosophy introduced related concepts:

These concepts differ in “geometry”—the structure of relationships between the memes that constitute them.

2.3 Carl Jung: Archetypes and the Inner Life of Ideas

Carl Jung offers a psychological counterpart to the esoteric and memetic frameworks discussed earlier. In The Red Book and his later theoretical writings, Jung explores the idea that the human psyche contains recurring symbolic patterns he called archetypes. These archetypes are inherited forms—structural tendencies that shape how humans imagine, narrate, and interpret the world. They manifest in myths, dreams, religious symbols, and cultural narratives, giving rise to recurring motifs such as the Hero, the Shadow, the Mother, and the Trickster.

For Jung, these archetypes arise from what he termed the collective unconscious: a shared layer of the psyche that transcends individual experience. Although the collective unconscious is not a literal shared mind, it reflects the deep structural similarities in how human brains organise meaning.Archetypes can therefore be understood as cognitive attractors—stable patterns in the architecture of the mind that guide interpretation and behaviour. Within the model developed in Section 1.3, archetypes function as the deep patterns that shape how minds stabilise meaning.

This perspective aligns naturally with the memetic framework. If memes are units of cultural transmission, archetypes are the deep templates that shape which memes resonate, replicate, and endure. Mythology becomes, in Jung’s view, the outward expression of these internal attractors. Thus, Jung provides a model in which living symbols are not supernatural beings but emergent structures arising from the dynamics of the psyche. They behave like autonomous agents because they influence thought, emotion, and action in ways that appear independent of conscious intention.

Jung therefore offers a way to understand memetic entities—egregores, thoughtforms, gods—not as literal beings but as expressions of deep psychological structures: the inner life of ideas.

2.4 Hegel: History as the Outer Life of Ideas

Hegel provides a crucial philosophical foundation for understanding ideas as dynamic, evolving structures at the collective level. In the Phenomenology of Spirit, he argues that human consciousness develops through a historical process in which ideas, practices, and social forms unfold over time. Hegel’s concept of Geist—often translated as “Spirit”—describes this collective dimension of thought. Like Jung’s collective unconscious, Spirit is not a supernatural entity but a structural reality: the totality of a society’s concepts, values, institutions, and symbolic forms.

For Hegel, history is the story of Spirit coming to understand itself. Ideas do not remain static; they evolve through a dialectical process in which contradictions generate new forms of understanding. Each historical epoch embodies a particular configuration of ideas that eventually encounters internal tensions, giving rise to new configurations

A central moment in Hegel’s account is the emergence of self‑consciousness through recognition. Individuals become aware of themselves not merely through introspection but through their relations with others. This relational model parallels how ideas exist not within a single mind but across many minds, sustained through shared recognition and reinterpretation. Just as self‑consciousness arises through mutual recognition, cultural ideas persist through continual communal reinterpretation.

Hegel’s view of history as a self‑developing process anticipates modern theories of cultural evolution. If Spirit is understood as the collective structure of ideas within a society, then historical change becomes the transformation of large‑scale memetic complexes driven by dialectical tension. These tensions mirror the internal psychological dynamics described by Jung.

2.5 Language

The Sapir‑Whorf hypothesis suggests that words shape how people conceptualise the world. Within the framework developed in Section 1, language becomes the medium through which specific configurations of cortical columns are induced in another mind, enabling the replication of ideas.

Section 3 - Modelling: A deep dive evaluation

Introduction

In this section the conceptual framework developed so far—memetic replication, interpretation, archetypes, and historical dynamics—is synthesised into a formal model which we can then implement as a program and use to test our hypothesis.

3.1 Representing Ideas: The Geometry of Memetic Structures

To model these memetic structures mathematically text is treated as the primary medium through which ideas manifest and replicate. The properties we need to model are:

A simple but powerful starting point is to treat a text, concept, or cultural artefact as a sequence of tokens, and to analyse the relationships between these tokens.

Reference Index Representation

The reference index representation is a simple model I have created to obtain a sequence of tokens from a medium of text. Instead of focusing on the words themselves, we assign each unique token a numerical index based on its first appearance. For example:

“I have an apple. The apple is red.”

becomes:

1, 2, 3, 4, 5, 6, 4, 7, 8

with a dictionary mapping:

{“I”:1, “have”:2, “an”:3, “apple”:4, “.”:5, “The”:6, “is”:7, “red”:8}

This dictionary is called the “reference index”

The meaning of the numbers is irrelevant; what matters is the pattern of recurrence and adjacency. We can represent the sequence as a directed graph in which each token “induces” the next:

1 → 2 → 3 → 4 → 5 → 6 → 4 → 7 → 8.

Thoughtforms as Cycles

A thoughtform can be defined as a token that participates in a cycle within the sequence graph.
In the example above:

Cycles represent self‑reinforcing ideas: concepts that return to themselves through interpretation. This aligns with intuition: “apple” is the thematic centre of the sentence.

Egregores as Networks of Memes

In this model, an egregore is represented as a set of interconnected memes.

An “egregore set” is defined as a set of memes organised around at least one thoughtform.

So to summarise:

3.2 Quantifying Strength: Probabilities and Attractors

To measure the strength of a thoughtform, we consider:

A simple measure is:

[ \text{strength} = \frac{\text{frequency}}{\text{sequence length} \cdot \text{average return distance}} ]

Which gives us:

For example in 1→2→3→4→5→6→4→7→8 we could calculate the strength of the thoughtform 4 as strength = 2/(9*3) = 2/27 which is approximately 0.07

To measure how correlated a thoughtform and a meme is we can apply the same logic however now we would treat a “cycle” as the sequence from the thoughtform to the meme. For example with our sentence “I have an apple. The apple is red.” it might make sense to calculate the correlation between “apple” and “red”. To do this we would take the section 4→7→8 which we would treat as a cycle, effectively replacing 8 for 4 or vice versa. We would then say strength = 2/(3*2) = ⅓. This is a relatively high strength which we would expect considering the words “apple” and “red” are semantically linked by “The apple is red”.

An egregore set can therefore be defined as a set of memes such that at least one is a thoughtform and all others have correlation strengths exceeding the thoughtform’s baseline strength.

3.3 Interpretation

Interpretation can be modelled as the interaction between two egregore sets, E1 and E2:

  1. If the anchoring thoughtform of E1 is not a member of E2(and vice versa)

    interpretation fails; the sets remain distinct.

3.4 Tensor Representation

Each egregore is represented as a tensor anchored on its thoughtform.

This allows egregores from different media to be compared in a shared coordinate system.

3.5 Reference Index Unification

To compare egregores across files, all tensors must use the same token ordering.

Procedure:

  1. Convert each reference index into an ordered list of symbols.
  2. Merge all lists into a unified index, removing duplicates.
  3. Remap all tensors from their original indices into the unified index.

This ensures consistent tensor alignment.

3.6. Comparing Media

To compare two files:

  1. Extract their egregore tensors.
  2. Normalise them to the unified reference index.
  3. Compute similarity using the dot product of corresponding tensors.
  4. Rank files by similarity.

3.7 The experiment

I created a direct implementation of the method above in python including a function that allows you to input two pieces of text (and the file name of each piece of text) and it will create the references of the texts, tokenise the texts, extract the thoughtforms, create the thoughtform sets, create the thoughtform set reference indexes,extract the egregore tensors unify the reference indices and thoughtform set indices updating the egregore tensors into a common format and give the strength of the two text files as a decimal value. I can now directly test this method by uploading text files and seeing if these strength values actually reflect how similar the text files are.

To perform this test I selected 3 sets of pairs text for each category:

Findings

High similarity
Medium similarity
Low similarity

Text pairing Reason Score
The Raven vs Annabel Lee Same Stoic concepts 0.012450540585190168
Meditations book 2 vs Enchridion Same psychological motifs 0.00024798518482772023
The tell tale heart vs The black cat Similar ethics, different metaphysics 0.0007056744017995863
Tao Te Cheng versus 1-10 vs Meditations Book 1 Both books have similar themes however they approach them differently 1.3893001838147057e-05
The Yellow Wallpaper vs The Tell‑Tale Heart Both books explore similar themes but from different perspectives 0.00028391642516193273
Aesops fables vs the book of proverbs Both books are similar thematically but come from different contexts 4.439103935979719e-05
Ozymandias vs On the origin of species These texts are conceptually unrelated 7.380805499250904e-07
Tao vs the black cat These texts are both poetry but vastly different 0.0001305644612593834
The black cat vs ozymandias These texts are both completely different 2.7565450814343112e-05

This reflects that the system is very sensitive to thematic overlap and is working successfully.

3.8 My python implementation

import json
import math
import os
import numpy as np
from functools import lru_cache
from collections import Counter

# ------------------------------
# Cached file loading
# ------------------------------
@lru_cache(maxsize=None)
def load_json(path):
with open(path, "r") as f:
return json.load(f)

# ------------------------------
# Token utilities
# ------------------------------
def parse_tokens(tokens):
return list(map(int, tokens.split(",")))

def extract_thoughtforms(tokens):
seq = parse_tokens(tokens)
counts = Counter(seq)
return [tok for tok, c in counts.items() if c >= 2]

# ------------------------------
# Upload utilities
# ------------------------------
def get_uploads(files):
return [load_json(f"uploads/{file.split('.')[0]}.json") for file in files]

def create_unified_reference_index(files):
seen = set()
unified = []

for index in get_uploads(files):
# index is dict token → pos
for token in sorted(index, key=index.get):
if token not in seen:
seen.add(token)
unified.append(token)

return {tok: i for i, tok in enumerate(unified)}

# ------------------------------
# Strength calculations
# ------------------------------
def calculate_strengths(tokens, file):
seq = parse_tokens(tokens)
counts = Counter(seq)

thoughtforms = extract_thoughtforms(tokens)
file_output = f"uploads/{file.split('.')[0]}_thoughtforms.json"

if not os.path.exists(file_output):
tf_index = {i: tf for i, tf in enumerate(thoughtforms)}
with open(file_output, "w") as f:
json.dump(tf_index, f)

total = len(seq)
return [counts[tf] / total for tf in thoughtforms]

# ------------------------------
# Egregore extraction
# ------------------------------
def extract_egregore_sets(tokens, file):
strengths = calculate_strengths(tokens, file)
seq = parse_tokens(tokens)
n = len(seq)

distances = [n * s for s in strengths]

positions = load_json(f"uploads/{file.split('.')[0]}_thoughtforms.json")

eg_sets = []
for i, dist in enumerate(distances):
pos = positions[str(i)] # correct mapping: thoughtform index → position
lo = max(0, math.floor(pos - dist))
hi = min(n, math.ceil(pos + dist))
eg_sets.append(seq[lo:hi])

return eg_sets

# ------------------------------
# Tensor remapping
# ------------------------------
def remap_tensor(tensor, old_index, unified_index):
new = [0] * len(unified_index)

# Reverse lookup once
pos_to_token = {pos: tok for tok, pos in old_index.items()}

for old_pos, strength in enumerate(tensor):
tok = pos_to_token.get(old_pos)
if tok in unified_index:
new[unified_index[tok]] = strength

return new

# ------------------------------
# Egregore tensor calculation
# ------------------------------
def calculate_egregore_tensors(tokens, file):
eg_sets = extract_egregore_sets(tokens, file)
seq = parse_tokens(tokens)

tf_path = f"uploads/{file.split('.')[0]}_thoughtforms.json"
thoughtforms = load_json(tf_path)
tf_list = [thoughtforms[str(i)] for i in range(len(eg_sets))]

# Ensure anchor is included
for i, eg in enumerate(eg_sets):
if tf_list[i] not in eg:
eg.insert(0, tf_list[i])

strengths = []
for i, eg in enumerate(eg_sets):
anchor = tf_list[i]
anchor_pos = eg.index(anchor)

   `precursors = [abs(anchor_pos - j) * len(seq) for j in range(len(eg))]`

   `# Compute strengths`  
   `s = []`  
   `for j, p in enumerate(precursors):`  
       `if p == 0:`  
           `s.append(0)`  
       `else:`  
           `s.append(seq.count(eg[j]) / p)`

   `strengths.append(s)`

return strengths

# ------------------------------
# Main similarity function
# ------------------------------
def calculate_egregore_strength(tokens1, file1, tokens2, file2):
E1 = calculate_egregore_tensors(tokens1, file1)
E2 = calculate_egregore_tensors(tokens2, file2)

idx1 = load_json(f"uploads/{file1.split('.')[0]}.json")
idx2 = load_json(f"uploads/{file2.split('.')[0]}.json")

TF1 = load_json(f"uploads/{file1.split('.')[0]}_thoughtforms.json")
TF2 = load_json(f"uploads/{file2.split('.')[0]}_thoughtforms.json")

TF1_list = [TF1[str(i)] for i in range(len(TF1))]
TF2_list = [TF2[str(i)] for i in range(len(TF2))]

E1_dict = dict(zip(TF1_list, E1))
E2_dict = dict(zip(TF2_list, E2))

shared = set(E1_dict) & set(E2_dict)
if not shared:
return 0

unified = create_unified_reference_index([file1, file2])

total = 0
for tf in shared:
t1 = remap_tensor(E1_dict[tf], idx1, unified)
t2 = remap_tensor(E2_dict[tf], idx2, unified)
total += np.dot(t1, t2)

return total

# ------------------------------
# Text utilities
# ------------------------------
def extract_reference_index(text, file_name):
words = list(dict.fromkeys(text.split()))
mapping = {w: i for i, w in enumerate(words)}

with open(f"uploads/{file_name.split('.')[0]}.json", "w") as f:
json.dump(mapping, f)

def extract_tokens(text, file_name):
path = f"uploads/{file_name.split('.')[0]}.json"
if not os.path.exists(path):
extract_reference_index(text, file_name)

ref = load_json(path)
return ",".join(str(ref[w]) for w in text.split())

def compare_texts(text1, file1, text2, file2):
return calculate_egregore_strength(
extract_tokens(text1, file1),
file1,
extract_tokens(text2, file2),
file2,
)

##

Conclusion

Across biology, psychology, philosophy, and esoteric tradition, a striking pattern emerges: ideas behave less like inert abstractions and more like dynamic, evolving entities. Dawkins’ memes replicate through minds; hermeneutics shows how interpretation continually reshapes them; neuroscience reveals the physical substrate through which they take form. Esoteric systems personified these dynamics as egregores and thoughtforms, while Jung framed them as archetypal attractors structuring the psyche. Hegel extended this logic outward, showing how ideas organise entire epochs of history through dialectical transformation.

When these perspectives are synthesised, a coherent picture appears. Ideas persist, mutate, compete, and organise themselves into higher‑order structures. They shape the minds that host them and the societies those minds build. In this sense, ideas exhibit the essential properties of life: self‑maintenance, adaptation, and reproduction. They are not alive biologically, but they are alive structurally — as patterns of organisation that endure across minds and generations.

The modelling framework developed in this essay makes this claim concrete. By representing ideas as recurrent patterns within sequences of tokens, we can identify thoughtforms as cycles, egregores as networks of cycles, and archetypes as stable attractors across many texts. Interpretation becomes the mechanism through which these structures interact, merge, or transform. History, in turn, becomes the record of these memetic structures evolving through time.

We have also shown experimentally that this model is accurate, to at least some degree.

Thus, the question “Are ideas alive?” can be answered: ideas are alive to the extent that they behave like intelligent systems — systems that persist through replication, adapt through interpretation, and shape the environments that sustain them. Human civilisation is, in this view, the ecology in which ideas grow. And just as organisms evolve to fit their niches, ideas evolve to fit the minds and cultures that host them. To understand history is therefore to understand the life of ideas — the invisible yet powerful forces that govern how societies think, act, and become.