About

Methodology

Onomastikon is a digital reference work created by Shriram Rajagopal that catalogs the endonyms and exonyms of ancient entities, including civilizations, cities, regions, and geographic features. This includes what they called themselves and what their neighbors called them. Each entry records the original script, transliteration, reconstructed pronunciation, literal meaning, and primary sources, with attention to uncertainty.

153 entities
87 complete
39 languages
ancient scripts rendered
958 name entries
two-paragraph treatments
1976 citations
sources on every entry

Editorial scope

The project covers entities of the ancient world, including civilizations, cities, regions, and geographic features. The temporal scope runs from the Bronze Age to the fall of Constantinople in 1453. Each entry focuses on its subject as a primarily linguistic but also historical record. Readers seeking detailed historical context should consult the standard reference works cited in each entry.

One page per name-thread

Entities are split by name-tradition, not by polity. Egypt spans thirty dynasties on one page because Kemet is one continuous thread; Persia carries Pārsa from the Achaemenids through the Sasanians, while the Sasanian state-name Ērānšahr belongs to the separate Iran entry. A polity earns its own page only where it introduced a genuinely new endonym thread, never one page per dynasty.

Regions and identification across languages

Regional names attested in different languages do not always pick out identical geographic units. Greek Mesopotamía, Akkadian Māt birīt nārim, and Hebrew Aram-Naharaim all refer transparently to the land between the Tigris and the Euphrates, and are treated as variant names of the same entity. By contrast, a single root can fan out into several distinct entities when it comes to name different kinds of unit. Sanskrit Sindhu, the great river, gave its name through Old Persian Hinduš to a whole subcontinent; the river itself (the Indus), the lower-river region (Sindh), and the civilization (India, whose own endonym is Bhārata) are recorded as three separate entities, since a river, a region, and a civilization are not interchangeable even where their names share a root. Where two names overlap partially or are framed differently, the entry's prose explains the distinction.

Entry status

Each entity entry carries a status label. Developing indicates that an entry is still being expanded with additional languages and sources. Complete indicates an entry with the full range of languages with attested names for that entity.

Name families and transmission

Each entity page groups its names into families by shared root and draws a transmission map of how those names traveled, and the two views encode different relationships. In a family tree, an indented form is a borrowing, one name taken from the one above it (Latin Aegyptus from Greek Aígyptos); forms at the same level are cognates, parallel reflexes of a common ancestor that no language took from any other (the Semitic names for Egypt, each inheriting the root mṣr in its own line). The map draws only borrowings as arcs: cognate forms share a family and a homeland cluster but are not joined by a line, because no name passed between them. A page can therefore show a whole family with no connecting arcs at all, where every form is a parallel inheritance rather than a borrowing. A borrowing is recorded only where an entry's own prose asserts it without hedging.

Source tiers

Entries draw on sources in roughly the following order of priority, with higher tiers preferred wherever they offer adequate coverage.

Standard reference works in the relevant philological tradition: the Oxford Classical Dictionary (4th ed.), the Cambridge Ancient History, the Reallexikon der Assyriologie, the Lexikon der Ägyptologie, and the Encyclopaedia Iranica.
Specialist lexica and grammars: the Chicago Assyrian Dictionary (CAD), Liddell–Scott–Jones, Beekes' Etymological Dictionary of Greek, the Hebrew and Aramaic Lexicon of the Old Testament (HALOT), and comparable works for each language tradition.
Primary text corpora: the Cuneiform Digital Library Initiative (CDLI), the Thesaurus Linguae Aegyptiae, the Perseus Digital Library, the Thesaurus Linguae Graecae, and ORACC.
Peer-reviewed monographs and journal articles for contested cases or recent reassessments.

Wikipedia and other tertiary aggregators are used as starting points for navigation but never as citations.

Reconstruction and uncertainty

Reconstructed pronunciations follow the standard linguistic convention of a leading asterisk (e.g., *kuːmat). Where multiple reconstructions are current, the entry follows the source cited and notes the alternative(s). All dates are approximate, as denoted by "c." (circa). Date ranges represent attested periods of time in which names were used. For instance, Latin existed for roughly a millennium before Aegyptus appears in Latin sources as a name for Egypt. So the entry dates reflect the narrower window of the name's use rather than the full lifespan of the language to which it belongs. Where a name's first attestation is securely dated to a specific inscription or text, the entry notes this and specifies the date of the source.

Every name entry carries one of three confidence labels:

Attested: The form is documented in primary sources, even where its pronunciation must be reconstructed (as is typical for Egyptian and cuneiform corpora).
Reconstructed: The form itself, not merely its pronunciation, is inferred, typically through comparative reconstruction or projection from a later attested form.
Disputed: Scholars disagree substantively about the form, its meaning, or its identification with the entity in question.

Transliteration conventions

Each language follows the transliteration standard of its established philological tradition: the Leiden Unified Transliteration for Ancient Egyptian; the Chicago system for Akkadian (CAD); the Library of Congress ALA-LC romanization for Sanskrit; standard scholarly conventions for Greek, Latin, Old Persian, and Biblical Hebrew. Where multiple systems coexist, the entry notes the choice.

Limitations

Onomastikon is the work of a student in active study, not a credentialed philologist. Entries are checked against the cited sources and corrected when error is identified. If you spot an error or have a suggestion, the feedback page is the place to say so; each entry's source Markdown can also be viewed, cited, and improved by pull request on the project's GitHub repository.

Source and data

The full source of the project, including the underlying data for every entry, is available at github.com/shriram-rajagopal/onomastikon. The complete dataset (every entity, language, and name record with its sources) is also published machine-readable at onomastikon.org/onomastikon.json. Code is released under the MIT license; content and data under Creative Commons Attribution-ShareAlike 4.0.

Each entity and language page carries a "Cite this page" block with a formatted citation and a BibTeX record.