From Golaavehi to Foalhavahi: Phonetic Evidence and the Lost Toponymy of Folhlhavai (Chagos)

Publication Note

This article distils a longer Pen for Rights research monograph, Fragments of a Lost Nomenclature: Attestation, Transmission, and Reconstruction in the Historical Toponymy of Folhlhavai, which examines Maldivian chronicles, historical linguistics, comparative toponymy, Portuguese textual evidence, European cartography, and a comparative Seychelles–Mascarene gazetteer. The complete evidentiary record, source audit, supporting tables, and individual reconstruction trials are available in that monograph. Detailed provenance for the twenty-two-name comparative control corpus and the individual cartographic objects discussed below is likewise provided in the monograph and its appendices. What follows sets out the central argument and findings for a general scholarly readership.

The Claim That Required Testing

On 25 February 2026, Nigel Farage told the House of Commons that “all the French did was rename the islands from the Maldivian language” (House of Commons 2026). Farage did not invent the proposition. In compressed political form, he was repeating an argument developed by the late Mohamed Haleem, widely known as Badi Alibege Haleem, President of the Maritime and Underwater Heritage Society of Maldives. In two articles contributed to Foalhavahi Fureyfu, Haleem placed French and British toponyms recorded for Folhlhavai alongside proposed earlier Dhivehi equivalents, drawing on oral tradition, comparative reading, and his own etymological interpretation (Haleem 2022).

The proposition matters because it is testable. If a substantial part of the colonial nomenclature could be shown to derive from earlier Maldivian names, the consequence would reach well beyond vocabulary. It would touch the history of geographical knowledge, the practice of navigation, and the way European powers remapped the central Indian Ocean. Pen for Rights treated Haleem’s compilation as a corpus of hypotheses awaiting independent test, not as either proof or fabrication in advance of that test.

The audit changed the shape of the inquiry almost immediately. Haleem’s (2022) two lists advance seventy proposed correspondences. Fifteen have independent lexical verification, meaning the Dhivehi root invoked can be shown to exist apart from Haleem’s own compilation, while none of the seventy island-specific correspondences has yet been independently verified against a source predating that compilation.   At the level of the wider gazetteer, twenty-five of the thirty-nine entries examined in the compatibility analysis, 64.1 per cent, depend on Haleem for their island-specific attribution, and none of these has independent historical confirmation. The maximal claim, that the French and British gazetteer can simply be restored as a lost Dhivehi original, does not survive this evidence.

That, however, is only half the finding. Beneath the unverified bulk of Haleem’s corpus, a smaller naming tradition survives on its own terms, independent of his compilation. This narrower substrate is the true subject of what follows.

The fairest assessment of Haleem’s contribution lies between uncritical acceptance and outright dismissal. His compilation preserved a body of correspondences that deserved systematic testing, and without it many of the questions examined here might never have been assembled into a testable corpus at all. Haleem’s greatest contribution may therefore prove not to have been that every proposed correspondence was correct, but that he preserved a corpus sufficiently concrete to be tested.

How the Names Were Tested

Historical toponymy invites confirmation bias because resemblance between names costs so little to produce. A French navigator and a Maldivian navigator might independently name the same island for its shape, its position against a reef, or a bird or fish observed there. A later map may simply copy an earlier one without adding fresh observation. A researcher working backward from a French name, translating it into Dhivehi, risks mistaking the product of that exercise for recovered history. A workable method must ask whether a proposed relationship between two names survives the ordinary, non-historical explanations for their resemblance.

Pen for Rights developed for this purpose a nine-stage Hierarchical Toponymic Reconstruction Model, the HTRM. Its architecture draws on Maniku and Lutfy’s studies of Maldivian atoll nomenclature and on Ragupathy’s Etymological Dictionary of Maldivian Island Names, which supplies the comparative corpus of Dhivehi place-name structures, geographical generics, and phonological interchanges against which every correspondence in this study has been tested.

The model examines the type of geographical unit being named, the attested morphology of the proposed Dhivehi form, its occurrence within internal Maldivian onomastics, its phonological admissibility, its chronological plausibility, its cartographic transmission history where relevant, and the strength of competing explanations. Each proposal receives one of six classifications: historically attested, strongly reconstructed, model-supported, provisional, unresolved, or rejected. The model can also identify a name attached to the wrong geographical unit rather than forcing a simple choice between acceptance and rejection, a distinction that proves useful at Egmont Atoll, where the geographical-unit test retains the proposed name Meedhaafushi at its correct, sub-island scale while identifying Varinmadhoo as the compatible candidate for the distinct island the colonial configuration actually requires.

The Seychelles–Mascarene Control

A further control proved necessary, because even a close semantic match between a colonial and a Dhivehi name can mislead. The study constructed an exploratory twenty-two-name comparative corpus drawn from Seychelles and Mascarene colonial nomenclature, coded under the same thematic scheme applied to the Folhlhavai material.

Its purpose was to test how readily ordinary European colonial naming, working without any Dhivehi source, produces the same categories of semantic names that might otherwise be mistaken for translations from an indigenous language. Fauna names account for six of the forty-three Folhlhavai entries, 14.0 per cent, against a closely similar share of the comparative corpus of twenty-two colonial toponyms.   Ordinary descriptive naming is a weak discriminator between colonial invention and genuine Dhivehi transmission, whatever the surface neatness of a particular pairing might suggest.   The comparison for personal names is more marked: 18.6 per cent against 31.8 per cent under the full denominators, and 25.8 per cent against 46.7 per cent when restricted to categories shared by both datasets. Because the control sample remains small, this gap is recorded as suggestive rather than statistically conclusive.

The finding bears directly on the proposed correspondence between Kukulhudhoo and Île Poule. Dhivehi kukulhu denotes a domestic hen, and Kukulhudhoo is an orthodox Maldivian island-name formation attested elsewhere in the national corpus. Île Poule, “Chicken Island,” belongs nonetheless to a semantic category occurring at a comparable frequency in the control corpus, so a French Namer encountering the same island might well have arrived at an identical designation with no debt to any Dhivehi name.   The HTRM classifies Kukulhudhoo as model-supported, but of low discriminatory value: linguistic naturalness of this kind is not the same thing as demonstrated historical transmission.

Golaavehi and the Foalhavahi Question

The most consequential indigenous result concerns the recurrent royal designation Golaavehi. Raadhavalhi, the official Maldivian royal chronicle, records the form across five dated attestations spanning roughly 1388–1549: Golaavehi Kanbulo, Golaavehi Nasuraddeen, Golaavehi Sanfaa Rani, Golaavehi Ranikilege, and Golaavehi Aisha Rani Kilege, spanning both male and female bearers and several distinct titulary constructions (Raadhavalhi 1979, 19, 20, 23, 28). Golaavehi was no modern invention, no isolated nickname, and no scribal accident. It belonged to established Maldivian onomastic vocabulary across roughly one hundred and sixty-one years.

A second, wholly independent tradition strengthens the case. Qadi Hassan Taj al-Din’s Arabic chronicle, the Tārīkh Islām Dībā Mahal, records a Sultan Nasir al-Din al-Qalaawahi, identifiable with confidence as the same ruler the Raadhavalhi records as Golaavehi Nasiruddeen (Taj al-Din 1982).   Two distinct Maldivian historical traditions, composed in different scripts and transmitted through separate documentary systems, preserve related onomastic forms attached to a single identifiable ruler.

Ragupathy’s Table 4 of phonological interchanges permits the correspondences G~K, H~K, and H~F (Ragupathy 2008), so the composite pathway required to relate Golaavehi, Kolaavahi, Hoalhavahi, and Foalhavahi to one another is phonologically admissible within an independently established framework, not one invented to serve this argument.   The HTRM classifies the proposed Golaavehi/al-Qalaawahi relationship with Foalhavahi as model-supported.   The indigenous form-family is historically attested, and the phonological correspondence required to relate it to Foalhavahi is independently admissible. What remains missing is referential rather than phonological. No primary source presently located states explicitly that Golaavehi or al-Qalaawahi denoted the place here called Foalhavahi.

If Golaavehi ultimately proves geographical, its recurrence among royal persons would raise a historically consequential possibility: that members of the Hilaaly or Hulule’(dynasty) political order carried an affiliation connected with Foalhavahi or its wider maritime region. The present evidence does not yet prove that proposition, but it makes the question legitimate in a way that mere phonetic resemblance never could.

Kandhoo, Candú, and a Documented Maldivian Naming Mechanism

The case of Kandhoo shows why the structure of Maldivian place naming matters here. Maniku and Lutfy document a recurrent convention by which an atoll derives its name from a constituent island, most clearly in Ke-Huvadhoo’s relationship to Huvadhu Atoll, with comparable patterns in Felidhoo’s relationship to Felidhe Atoll and Nilandhoo’s to Nilandhe Atoll. Deriving a larger geographical unit from a constituent island’s name is not a mechanism invented to explain Folhlhavai; it exists independently within documented Maldivian nomenclature.

João de Barros supplies a separate sixteenth-century anchor. His Década III da Ásia, published in 1563, records both Candú and Adú within the Maldivian island chain he describes, placing that chain as extending to seven degrees south and naming Maldiva as the seat of a king styled lord of the whole archipelago (Barros 1563, fols. 72–73). Candú therefore has independent sixteenth-century textual attestation, anchoring the identification with Kandhoo Atoll in a documented source rather than in inference alone. The proposed relationship between a separate, constituent island-level Kandhoo and the colonial name Île du Coin deserves attention on this basis: if an island called Kandhoo supplied the name of the larger atoll, the mechanism would be structurally orthodox within documented Maldivian practice.   Candú stands historically attested at atoll level. The island-level Kandhoo remains model-supported rather than independently confirmed.

Sultan Uthman al-Fahandawi and the Fehendhoo Problem

Taj al-Din’s (1982) chronicle preserves the relational form al-Fahandawi, demonstrating that a Fahand-/Fehend- onomastic element existed in Maldivian royal usage apart from any modern Folhlhavai reconstruction. This lends genuine historical interest to proposed forms such as Fehendhippolhu. A real island called Fehendhoo exists in Baa Atoll, in the northern Maldives.   There is therefore independent evidence both for the historical al-Fahandawi relational form and for Fehendhoo as a genuine Maldivian geographical name, but not yet evidence sufficient to identify Sultan Uthman al-Fahandawi’s referent with Folhlhavai rather than Baa Atoll. The HTRM accordingly returns a verdict of unresolved for Fehendhippolhu. What would settle the matter is a manuscript, genealogy, or itinerary capable of locating the Sultan’s affiliation with precision.

From Foalhavahi to Folhlhavai

The proposed morphological analysis is conceptually simple. Folhlha– is analysed as the Foalhavahi name-family, while vai is proposed as a reflex of Dhivehi bai, “part, portion, division.”   Folhlhavai would therefore signify a Foalhavahi division or geographical segment, an archipelago-level name built from an island-level one.

The proposition rests on more than visual resemblance. Ragupathy’s (2008) comparative framework independently permits the B~V correspondence the derivation requires. Maldivian nomenclature also furnishes a structural analogue in the Huvadhu/Suvaidhoo name-family: Lutfy (1987a; 1987b) records Ke-Huvadhoo as the source of Huvadhu Atoll, while separately documenting a wider Suvaidhoo conception encompassing Huvadhu, Fuvahmulah, and Addu together. The same geographical name-family is therefore documented across island, atoll, and supra-atoll scales in Maldivian nomenclature, and an island-level Foalhavahi coexisting with an archipelago-level Folhlhavai fits that documented architecture.

Of the five conditions the HTRM specifies for this derivation, only one, general B~V phonological admissibility, is presently confirmed.   Four remain outstanding: bai’s independent lexical sense in the relevant environment, its behaviour as a compound-final constituent, the compound-specific development of bai into vai, and Folhlha-’s independent connection to Foalhavahi. The confirmation of B~V admissibility and the outstanding bai→vai development are not the same proposition, and the derivation now stands as a falsifiable linguistic hypothesis with clearly stated evidentiary requirements rather than an assertion resting on resemblance alone.

An Evidentiary Dictionary of the Principal Names

 

Form	Analysis	Independent evidentiary basis	Status
Golaavehi	Recurrent royal/onomastic designation; possible Foalhavahi relationship	Five Raadhavalhi attestations, c. 1388–1549 (Raadhavalhi 1979, 19, 20, 23, 28); al-Qalaawahi parallel in Taj al-Din (1982)	Form attested; Foalhavahi identification model-supported
Foalhavahi	Indigenous name identified with the Pollouoys/Pollowois/Polvara exonymic series	Independent European textual and cartographic transmission	Strongly reconstructed under the HTRM as the proposed indigenous referent underlying the Pollouoys/Pollowois/Polvara exonymic series
Folhlhavai	Folhlha- + bai/vai, “Foalhavahi division/part”	B~V independently admissible (Ragupathy 2008); bai→vai development and remaining morphological conditions outstanding	Unresolved
Kandhoo / Candú	Constituent island name underlying atoll-level Candú	Candú textually attested by Barros (1563, fols. 72–73); Maniku-Lutfy eponymic structure independently documented	Atoll historically attested; island-level reconstruction model-supported
Kudarah	kuda compared through a secondary semantic route with Île Parasol	Dhivehi lexical basis independently attested; provenance of the specific correspondence remains unresolved	Provisional
Raiymasfushi	raiymas “red snapper” within a regular island-name formation	Lexical elements independently attested; island-specific attribution remains Haleem-dependent	Unresolved
Kukulhudhoo	kukulhu, “hen/chicken,” compared with Île Poule	Comparable frequency in Seychelles–Mascarene control corpus	Model-supported, low discriminatory value
Fehunudhoo / Varinmadhoo	Directional reconstruction within Egmont Atoll	Directional vocabulary independently supported; opposing reef-extremity configuration; Meedhaafushi reassigned to its correct sub-island scale	Model-supported
Fehendhippolhu	Fehen-/Fehun- atoll-level construction	al-Fahandawi root historically attested (Taj al-Din 1982); referent disputed (Baa Atoll vs. Folhlhavai)	Unresolved
Fehenaavari	Proposed sector/bank construction	vari independently attested as a Maldivian geographical element (Ragupathy 2008)	Unresolved (HTRM-indeterminate)
Hoaravahi	Proposed alternative reading of the Golaavehi/Foalhavahi chain	Independently explicable through hoara, “great frigatebird”	Rejected as a necessary stage in the chain
Holhimadulu	Proposed alternative name for Kandhoo Atoll	Conflicts with the stronger, independently attested Candú record	Rejected

 

The table exposes a distinction that runs through the study as a whole. A proposed name can be entirely natural on linguistic grounds without being historically verified. A form such as Candú, conversely, can stand textually attested in the historical record while the deeper history of its name, and its relationship to the wider Folhlhavai naming tradition, remains genuinely open.

The Cartographic Problem: One Geography, Two Names

A separate Maldivian source bears on this same maritime region at a different geographical scale. Sultan Hassan IX’s 1560–1561 titulature to the Portuguese crown refers to “the seven islands of Pullobay,” a collective designation that Albert Gray and H. C. P. Bell, in their Hakluyt Society edition of François Pyrard’s account, identify with the Chagos Islands (Gray and Bell 1887–1890).   That identification is Gray and Bell’s own editorial judgement rather than an independent primary attestation, and it is treated here as a separate documentary stream from the individual European cartographic representations discussed below, not as corroboration of them merely because both concern the same maritime space.

Among the stranger findings of this research is one concerning Foalhavahi itself. A transmitted European series comprising the forms Pollouoys, Pollowois, Polvara, Polvera, Poluerera, and Polverera appears consistently across the cartographic record, and can appear on the same or related maps quite separately from another feature labelled Diego Garcia or Diego Gratiosa.   If the identification of this series with Foalhavahi proposed here is correct, European cartography placed two distinct names for a single geographical referent onto the map as though they denoted two separate places, a pattern the monograph terms cartographic co-reference failure.   The possibility is entirely plausible historically: early modern cartography routinely accumulated information from disparate pilots, textual traditions, and successive map lineages, and a cartographer had no need to observe a feature personally before reproducing an inherited name for it.

The twelve cartographic objects examined in the study, spanning 1626–1851, should not be mistaken for twelve independent witnesses. The pre-1817 subset, comprising Speed, Tavernier, Jansson, Visscher, De Ram, Homann in two states, Jaillot, and Homann’s heirs, displays identifiable copying and transmission relationships, while Thomson (1817), Wyld (1837), and Tallis (1851) document the subsequent destabilization of that inherited geographical register.   Thomson’s map separates the Maldive Islands, terminating at Suadiva, from Adu and Candy farther south, marking the latter pair with explicit uncertainty, a divergence that predates Robert Moresby’s celebrated survey of the Maldives by seventeen years. Whatever generated Thomson’s uncertainty cannot simply be attributed to Moresby’s later hydrographic work. Nineteenth-century British hydrography did not abruptly erase an otherwise stable earlier geography at a single moment; the transformation was more gradual, and considerably more untidy documentarily, than that.

Why This Matters Beyond Etymology

Toponymy records how communities organise geographical space through language, and it preserves traces of navigation, settlement, lineage, political geography, and resource use long after the institutions that generated the names have disappeared from the record.   Folhlhavai raises the question of whether the colonial archive begins the geography of the region it describes, or merely preserves one later, partial layer superimposed on an older geography that existed independently of it.

Candú and Adú have a genuine sixteenth-century textual anchor in Barros (1563). A substantial European transmission series preserves forms this study identifies with Foalhavahi. The Golaavehi/al-Qalaawahi form-family exists independently within Maldivian historical writing across two distinct documentary traditions (Raadhavalhi 1979; Taj al-Din 1982).   Several proposed micro toponyms conform to independently documented Maldivian morphological and onomastic structures rather than merely resembling them by coincidence.   The Seychelles–Mascarene control demonstrates, with equal force, why such correspondences must never be overclaimed: European descriptive naming practice can independently produce the categories of names that might otherwise appear to be translations from Dhivehi.

The historical argument must be kept apart from the legal one. A map label is not a title deed; an etymology is certainly not one. Toponymy cannot substitute for the distinct body of legal evidence governing title, administration, recognition, effectivités, decolonisation, or self-determination. But the absence of dispositive legal force does not make toponymy historically irrelevant. It bears directly on prior geographical knowledge, established nomenclature, patterns of navigation, and the conceptual geography through which Maldivians understood the southern Indian Ocean before the colonial cartographic order became dominant. Surveying, classification, and naming were never merely clerical exercises; they numbered among the principal instruments by which older, indigenous geographical worlds were translated into colonial systems of knowledge, often with consequences that outlasted the colonial administrations that first imposed them.

What Must Be Found Next

Three clusters of archival work bear directly on the principal questions this study leaves open.   The first concerns Golaavehi and Foalhavahi: whether a primary source explicitly connects the two, and whether genealogical or titulary records explain Golaavehi’s recurrence among Hilaaly-period royal persons. The second concerns al-Fahandawi and Fehendhoo: which Fehendhoo the Sultan’s designation actually names, and whether an independent source preserves Fehendhippolhu outside Haleem’s compilation.   The third concerns Folhlhavai’s morphology: whether the proposed bai/vai construction can be demonstrated in other Dhivehi compounds, and whether navigational notebooks, manuscript maps, or oral-geographical materials preserve Folhlhavai microtoponyms independently of Haleem.

A single primary document bearing directly on any one of these questions could materially alter the evidentiary position more than a hundred further speculative correspondences built on resemblance alone.

Conclusion

The research rejects the maximal Farage/Haleem proposition that an entire colonial gazetteer can presently be recovered by translation, while establishing an independently evidenced Maldivian nomenclatural substrate.   Within that substrate, the Golaavehi/al-Qalaawahi identification with Foalhavahi is model-supported, with the remaining problem principally referential rather than phonological; the study also provides a reproducible method for distinguishing historical evidence from attractive reconstruction.

The question is no longer whether Golaavehi existed. It did.   Nor is its proposed relationship with Foalhavahi merely a resemblance without linguistic support: under the HTRM it is model-supported.   The principal unanswered question is whether surviving primary evidence can connect the historically attested Golaavehi/al-Qalaawahi designation specifically to Foalhavahi as a geographical referent.   Finding that evidence is now an archival problem, not an exercise in speculative etymology.

Selected References

Barros, João de. 1563. Década III da Ásia. Lisbon. Livro Terceiro, Capítulo VII, “Em que se descreue o sitio das ilhas de Maldiua,” fols. 72–73.

Gray, Albert, and H. C. P. Bell, trans. and eds. 1887–1890. The Voyage of François Pyrard of Laval to the East Indies, the Maldives, the Moluccas, and Brazil. 3 vols. London: Hakluyt Society.

Haleem, Mohamed. 2022. “A Comprehensive Look at the Chagos Islands.” Foalhavahi Fureyfu, 27 October 2022, 27–30. Chagos WhatsApp Group.

House of Commons. 2026. “Diego Garcia and British Indian Ocean Territory.” Hansard, 25 February 2026.

Lutfy, Mohamed Ibrahim. 1987a. “The Atolls of the Maldives and Their Names” (Part I). Faththoora 9, no. 3 (June). Malé: National Centre for Linguistic and Historical Research.

Lutfy, Mohamed Ibrahim. 1987b. “The Atolls of the Maldives and Their Names” (Part II). Faiythoora 6, no. 5 (August). Malé: National Centre for Linguistic and Historical Research.

Maniku, Hassan Ahmed. 1980. “The Atolls of the Maldives and Their Names.” Faiythoora 1, no. 12 (1 March). Malé: Office of the President.

Raadhavalhi. 1979. Official chronicle of Maldivian royal succession. Official edition/reproduction. Malé: President’s Office, through the National Council for Historical and Linguistic Research.

Ragupathy, Ponnampalam, assisted by Naseema Mohamed. 2008. An Etymological Dictionary of Maldivian Island Names. Malé: National Centre for Linguistic and Historical Research.

Taj al-Din, Hassan. 1982. Tārīkh Islām Dībā Mahal, Part I, with continuation by Qadi Muhammad Muhibb al-Din and Qadi Ibrahim Siraj al-Din. Critical edition by Hikoichi Yajima. Tokyo.