This is a study guide, annotated bibliography, and book recommender covering Western political, social, economic, and philosophical thought — with extensions into literature, psychology, religion, art, architecture, and film where those fields carry political or philosophical weight. It was built using Claude Code, Anthropic's command-line coding assistant, at the direction of Peter Keane. Peter set the scope, approved or rejected additions, and directed revisions; Claude Code produced the entries, the cross-links, the build scripts, and the site.
The graph is meant to be a useful first step for someone exploring these areas — not a definitive reference. The hope is that a reader browsing the entries and following the cross-links will notice connections, patterns, and lines of influence that suggest where to read next.
The core is the liberal, Marxist, conservative, and religious traditions in political and social thought, roughly from the 18th century to the present. From there it extends into moral philosophy, the American civil-rights and feminist traditions, non-Western political and religious thought, literature, psychology and psychoanalysis, visual art and film, and architecture and urbanism. Coverage is selective and reflects the interests that directed its construction.
Thinkers are the main entry type — short profiles with key ideas, key works, and recommended secondary sources, cross-linked by influence and association. Concepts, Schools, and Events provide definitional context. Topics are different: they are exploratory pages organized as annotated bibliographies, covering areas that are still contested, still being excavated, or that cut across multiple thinkers in ways the individual entries don't capture.
The selection is curated, not comprehensive. It leans American, post-1800, left-of-center (in the American sense), and Christian (on the religious side). Conservative thinkers of substantial weight are included, but the center of gravity is to the left. These biases are acknowledged rather than defended — they reflect the interests that motivated the project.
Within that curated selection, the entries themselves try to present uncontested facts and relationships rather than editorial argument. The descriptions of what thinkers thought, wrote, and influenced are drawn from standard scholarly and critical consensus. But the selection of which thinkers to include, and the density of coverage in different areas, is a curatorial choice.
The content was generated by Claude, Anthropic's large language model (LLM). An LLM is a statistical system trained on a large corpus of text — books, articles, websites, and other written material. During training, the model learns patterns in how words, sentences, and ideas relate to one another. When it generates text, it is doing pattern matching against that training data: predicting what text is most likely to follow from a given prompt, based on the statistical relationships it learned during training.
This means the entries reflect the consensus of the model's training data, with all the biases and gaps that implies. The model does not "know" things in the way a person does — it identifies patterns in text. Where its training data is rich (canonical Western philosophy, major American writers), the entries tend to be reliable. Where the training data is thinner (non-Western traditions, living thinkers, contested recent history), the entries are more likely to contain errors or omissions.
No entry has been independently fact-checked. Readers should treat entries as starting points, not authorities, and should verify particulars before citing them.
Errors and omissions are expected. Corrections can be submitted to the GitHub repository at github.com/pkeane/kg.