Atlas / Glossary
Glossary
Every term used on this site, defined plainly — including the ones borrowed from instruments built by other people, with what they measure and who built them.
Anchor · Anti-pluralism index · Chapel Hill Expert Survey · Coding · Composition · Confidence rating · Contested flag · Distance · Drift · Equal weighting · GAL–TAN · Horseshoe theory · Inter-coder reliability · Lineage · Manifesto Project · Measurement failure · Signature pair · Subsidiarity · Substitution · Trajectory · V-Dem / V-Party
Anchor
A written definition fixing what a particular score means on a particular axis, with real cases that sit there. Anchors are set before any coding happens; without them two people using the same scale will not mean the same thing by +7.
Anti-pluralism index
A measure from the V-Party project built from items on whether a party rejects the legitimacy of its opponents, encourages violence, or disrespects fundamental rights and the electoral process. Close to a direct measurement of this atlas's organisation-of-power axis, and produced by a much larger team.
Chapel Hill Expert Survey
A survey asking political scientists who specialise in a country to place its parties on a series of scales. Running since 1999, strongest on European parties, and the standard reference for party positioning in comparative politics. Reports disagreement between experts, which is what makes it trustworthy.
Coding
Assigning a position on a defined scale, using written anchors and documented evidence. A coding is a judgement with its reasoning attached — not a measurement in the physical sense, and not an opinion either.
Composition
Which axes carry the gap between two organisations, as distinct from the size of the gap. Two pairs can sit the same distance apart for entirely different reasons, and the reason is the content of any claim about them.
Confidence rating
A declared level of certainty attached to every entry. Low means the organisation is recent and moving, that independent reporting is thin, or that qualified specialists disagree about the classification itself.
Contested flag
Marks an entry where specialists agree about the facts and disagree about what they amount to. That is a disagreement about the concept rather than the evidence, and more research does not resolve it. Both readings are given.
Distance
The straight-line separation between two organisations across all six axes at once. Each axis runs from −10 to +10, giving a theoretical maximum separation of about 49.
Drift
When the centre of political opinion moves and an organisation stands still, producing the appearance of movement. One of the five failure modes, and the reason claims about a party “becoming extreme” so often survive scrutiny badly.
Equal weighting
The assumption, built into every distance figure here, that a ten-point gap on economics matters as much as a ten-point gap on method. Nothing justifies it except convenience, and the interactive atlas lets you vary it.
GAL–TAN
Green–alternative–libertarian against traditional–authoritarian–nationalist: the second dimension the Chapel Hill surveys run alongside economic left–right, because one line stopped fitting. Adjacent to this atlas's social-and-cultural-order axis.
Horseshoe theory
The claim that the far left and far right curve toward each other. True about a real convergence on the axes of form — authority, scale and method — and false as a description of political space, which is why anarcho-communism and anarcho-capitalism sit nineteen points apart on economics and at the same point on authority.
Inter-coder reliability
The standard evidence that a coding scheme works: hand two independent coders the same case and the same codebook and measure how far apart their scores land. This atlas has one coder and therefore cannot report it, which is its principal limitation.
Lineage
A documented organisational relationship — continuation, split, doctrinal derivation, succession or suppression. Descent is a fact about origin. It is weak evidence about current position and becomes weaker with every decade.
Manifesto Project
A long-running project that codes the actual text of party programmes sentence by sentence into fixed categories. Reproducible in a way expert judgement is not, and limited by the gap between what a party promises and what it does.
Measurement failure
A disagreement in which both parties assert something true, the assertions concern different properties, and the language conceals the difference. Distinct from a genuine value conflict, which no measurement resolves.
Signature pair
Two organisations that a simpler model would place in the same neighbourhood and that a given axis separates. Each of the six axes here has one, and they are the actual argument for the axis existing.
Subsidiarity
The principle that decisions should be taken at the smallest competent level. A position on the scale-of-decision axis, with roots in Catholic social thought and a central place in European constitutional argument.
Substitution
Establishing something true about one axis and stating a conclusion about another, with no argument connecting them. The most common failure in political argument, and almost never a lie — the premise is usually correct and the reach of the evidence is not.
Trajectory
An organisation's positions recorded at dated anchor points, each tied to a documented moment. Nothing is interpolated between anchors, because nothing is known between them.
V-Dem / V-Party
Varieties of Democracy: an instrument measuring the character of political regimes across hundreds of indicators and several centuries, using multiple coders per country-year with explicit uncertainty estimates. Its party-level companion produces the anti-pluralism index.
A term used here that is not on this page? That is a defect. Say so and it gets added — vocabulary that only its author understands is not a vocabulary.
The codebookLearning modulesOne or two emails a month, or none at all
The update carries codings that changed, new claim checks, and at least one finding that went against the atlas. Nothing else, and one click to stop.
Site owner: set FORMSPREE_ID in offers.py to activate this form.
Cite this page
Scores are revised on evidence, so a citation should carry the version it refers to. Version 4.3.1 is current; the revision log records every change.
Free to quote and reproduce with attribution. This is a synthesis rather than a peer-reviewed dataset — where an authoritative figure is needed, cite the Chapel Hill Expert Survey, the Manifesto Project or V-Dem.