names and Indic structure coexist with separate later roads; contact layers, rather than names alone, identify shaping environments.

A social designation can spread or survive language shift. It must be compared with grammar and broader vocabulary rather than turned into everyone’s itinerary.

The Council of Europe’s [“From India to Europe” factsheet](https://rm.coe.int/from-india-to-europe-factsheets-on-romani-history/16808b18ed) discusses Ferdowsi’s *Shahnameh* story of an Indian ruler sending musicians to Persia’s Bahram V. Written centuries after that reign, it cannot identify the ancestors of all Roma. As with the *Aeneid*, a narrative may preserve meaningful regional relationships without verifying its genealogy: compatibility with Indian–Persian contact is weaker than identification of Romani’s transmitters.

Manchester’s [group names](https://www.kratylos.org/~raphael/romani/whatis/language/names.html) include *Arli* or *Erli*, from Turkish *yerli*, settled or local. Settlement does not contradict earlier migration. Past movement does not oblige descendants to maintain one permanent way of life.

## 15 What the combined evidence supports

| Claim | Assessment |
|---|---|
| Romani is an Indo-Aryan language | Strongly supported by inherited vocabulary, regular sound correspondences and grammatical relationships |
| The language is historically Egyptian | Not supported by its comparative linguistic structure; the medieval designation has a separate documentary history |
| Iranian and Armenian languages influenced Romani | Supported by shared early loanwords, without requiring every contact to fall into one consecutive itinerary |
| Greek contact was substantial | Supported by ordinary vocabulary, numerals and grammatical influence |
| Greek craft words prove that the ancestors first learned those crafts in the Greek world | Invalid inference; lexical replacement does not establish the first acquisition of a practice |
| Regional European languages shaped later Romani varieties | Supported, with different contact layers and histories of movement |
| Every Roma community had the same legal status | Contradicted by the different regional institutions documented in European history |
| All communities historically called “Gypsy” have the same language genealogy | Invalid inference from a shared external label |
| Romani cultural memory exists only in oral tradition | Incorrect; written literature, recorded performance and scholarly documentation coexist |
| Population evidence independently supports an Indian ancestral component | Supported by the sampled genome-wide studies, alongside later regional admixture |
| Shared ancestry or shared vocabulary establishes every family’s route | Invalid inference; particular movements require particular evidence |
| Europe’s exonyms identify the people’s origin | Invalid; “Gypsy” reflects an “Egyptian” attribution and *Tsigan*/*Zigeuner* a Byzantine sect name, while the language is Indo-Aryan |
| Heavy Iranian and Armenian borrowing implies a large Middle Eastern genetic contribution | Invalid inference; genome-wide data show limited such ancestry despite the linguistic layers |
| Romani and Domari are one language from one migration | Not established; the languages are separate, and their later contact layers differ |
| Genetic dates fix the departure from India | Invalid; estimates are model-dependent and some carry error margins of centuries |
| A modern anthology is a surviving medieval migration chronicle | Incorrect; its literary and historical value concerns a different kind of record |

The Indian–west Asian–European outline describes inheritance and contact: robust Indian linguistic ties, substantial Greek influence and varied regional histories. It cannot settle every itinerary, departure motive or household biography.

Matras supplies comparative questions, Hancock a testable interpretation and the anthology several communities’ literary voices. None needs to function as a lost founding scripture. Words, documents, genomes and literature illuminate different parts of