Free. No account, no email, no sign-up.
What they are, how yours gets named, and what the chained labels mean.
People and ideas should not be placed in boxes, yet it’s just so fun to do.
Everything below is true of this instrument, and was true before you opened it.
Use accordingly: a mirror that distorts in legible ways.
Where you align, where you stand apart, who's closest, and who's most opposed.
Don't see your favorite thinker? Adjust the filters above to include more, or exclude more.
Toggle thinkers to overlay their positions on each axis alongside yours (gold dot).
Pick any two axes to plot thinkers on a plane. Click thinkers below to add them, or jump to a preset combination.
Place yourself on a plane whose four poles are philosophers rather than abstract axes. Each axis runs between two rival thinkers; your marker shows how close you are to one pole versus the other. Plot any other thinkers as reference points between the same rivals.
Only thinkers with at least a 70% overall match to you are eligible to be plotted (community pool members included when they qualify). The four endpoints can be anyone.
Drag any belief and watch your closest matches re-sort live (by axis similarity); see exactly which positions drive who you align with.
Which philosophical traditions are closest to the test-taker's worldview?
Markers colored by similarity to the test-taker's profile.
Every thinker placed on a horizontal time axis, dot color by tradition. Drag to pan, scroll to zoom, click any dot for the full profile.
How deep? Each tier adds axes and matching foundational questions.
Where you align, where you stand apart, who's closest, and who's most opposed.
Don't see your favorite thinker? Adjust the filters above to include more, or exclude more.
Toggle thinkers to overlay their positions on each axis alongside yours (gold dot).
Pick any two axes to plot thinkers on a plane. Click thinkers below to add them, or jump to a preset combination.
Place yourself on a plane whose four poles are philosophers rather than abstract axes. Each axis runs between two rival thinkers; your marker shows how close you are to one pole versus the other. Plot any other thinkers as reference points between the same rivals.
Only thinkers with at least a 70% overall match to you are eligible to be plotted (community pool members included when they qualify). The four endpoints can be anyone.
Drag any belief and watch your closest matches re-sort live (by axis similarity); see exactly which positions drive who you align with.
Which philosophical traditions are closest to your worldview? Sorted by average similarity, with internal diversity (standard deviation) shown.
Markers colored by similarity to your profile. Click markers for details.
Every thinker placed on a horizontal time axis, dot color by tradition. Drag to pan, scroll to zoom, click any dot for the full profile.
A deeper workshop for the 400 canonical thinkers in the corpus. Beyond the axis-based similarity view above, the Crypt hosts research-grounded encyclopedia entries, simulated debates between thinkers, foundational-answer filters, cross-thinker statistics, pole-to-pole correlations, a regression builder, and two absurdity / critique-susceptibility rankings; tools for reading the corpus sideways rather than just against your own profile.
Type “exit” or click the ✦ above to leave.
Alphabetical. Each card: short research-based analysis, key unique concepts, and real historical/conceptual connections to other thinkers in the corpus.
Click any name to expand. Each thinker has a key idea (central thesis). Sortable and searchable.
Pick a foundational question to see how each philosopher sorts into the answers. Thinkers with blended answers appear under every option they selected.
Corpus-wide statistical analysis across all 400 canonical thinkers and 20 axes.
Every canonical thinker ranked by internal tension across the curated tension library: Hard (a logical clash between two commitments), Soft (a standing tension), and a Combined score with Hard double-weighted. Click any thinker to unfold each tension.
The axis pairs with the strongest signed correlation. Click any row to reveal a scatter plot. Click ⇄ to flip a pole.
Build a custom multiple linear regression over the 400 canonical corpus. Pick a response variable (an axis score, or a categorical-question answer as a binary target), then mix and match predictors drawn from the 20 axis scores, the 20 foundational question answers (one-hot), and school/tradition affiliation (one-hot). Fits via QR decomposition with listwise deletion for nulls; reports OLS coefficients with Student-t standard errors, adjusted R², and F-statistic p-values.
Pairs of thinkers from schools that are far apart on average who nevertheless converge individually. Each school is reduced to a centroid (mean position across its members) and only pairs from schools below 60% centroid-similarity qualify; so the remaining surprises aren't adjacent-tradition coincidences. Pairs must also share at least 5 non-null axes and 5 non-null foundational answers, so similarity is computed from a substantive shared base. Click any pair to open a side-by-side compare modal.
Average axis scores by historical era, drawn from the full 400 canonical corpus. Each bar shows how the era's thinkers sit on that axis on average; the darker tick shows the corpus-wide mean for reference.
All 400 canonical thinkers × 20 foundational questions, colored by option. Hover a cell for details. Click a thinker to open their profile. Sort by name, era, or a specific foundation to see clusters.
Here, what matters isn’t which side of each question you land on; it’s how strongly you hold it.
The spectrum bends into a horseshoe: the two extremes curl up to the near‑touching tips and moderates sink to the bottom. So two thinkers count as kindred when they share the same conviction; even when they’re convinced of opposite things. Opposites coincide.
Each arm fills on its own: how firmly you commit to a pole’s statements drives that side from the base up to the tip. Answer strongly in one direction and that arm fills all the way: no matter how you answer the other side. Hold both poles with conviction and both arms fill at once.
Every match %, the closest / farthest lists, and the axis gauges re‑rank around shared conviction instead of shared position.
Turn it off any time to return to the normal left↔right view.
Here, every pole stands on its own. Each axis rises as two independent pillars, its left pole and its right pole, and each bar climbs only as high as your conviction for that one side.
Unlike the horseshoe, opposites do not coincide. The 40 poles across the 20 axes become 40 separate dimensions, kept paired under the axis they belong to, so holding the far‑left pole is scored as a wholly different commitment from holding the far‑right one.
Two thinkers count as kindred only when they raise the same pillars; when they share an actual pole, not merely a matching intensity. Commit hard to opposite poles and you read as far apart, even though you’re both decisive.
Every match %, the closest / farthest lists, and the axis bars re‑rank around shared poles; stricter than Default’s position, and unlike Horseshoe’s shared conviction.
Switch back to Default or Horseshoe any time from the Axis view menu.
The other three views apply one lens to all twenty axes. Adaptive picks the lens per axis, because the axes are not alike: some have poles that genuinely exclude each other, some have poles most people hold together, and some bend round so the two extremes end up resembling each other.
Default where the poles compete for the same ground and the middle is a real intermediate. Pillars where both poles can be held at once, so each is scored as its own commitment. Horseshoe where maximal commitment to either end resembles the other end more than it resembles the middle, so what gets measured is conviction rather than direction.
Each axis was decided from what its two poles actually assert — not from where thinkers currently sit, since that would let a scoring artefact choose its own lens. Most axes turn out to be a single dial with two directions; a few bundle two separable questions, and two bend far enough that the extremes meet.
Every match %, closest / farthest list and axis card is computed with each axis under its own treatment, inside a single similarity score.
Switch back to Default, Horseshoe or Pillars any time from the Axis view menu.