In 1893, an Austrian geologist named Eduard Suess published the third volume of *Das Antlitz der Erde* — *The Face of the Earth* — and in it, he named an ocean. The ocean was already gone. It had closed tens of millions of years earlier, its floor subducted, its waters redistributed into the Atlantic and Indian oceans that replaced it. What remained were its fossils: marine creatures from a tropical sea, pressed now into limestone in the Alps, in the Himalayas, in the mountain ranges that rose precisely because the Tethys closed. Suess inferred the ocean's existence from these remnants. He named it after Tethys, the Greek sea goddess — sister and consort of Oceanus, mother of rivers and fountains.
The name stuck. It stuck so thoroughly that we now talk about the Tethys as a real thing, a structure with a history, an entity that shaped the world. And it is — it was. But before Suess named it, it existed only as scattered limestone, as fossil records without a frame, as evidence without a diagnosis. The name created the object it described.
This is how naming works. It is not mere labeling. It is an act of assembly: it gathers scattered evidence into a coherent thing, makes it available for thought, gives it a trajectory. The Tethys Sea did not become important to geologists the moment Suess proposed it — it became discussable. Nameable things can be studied, argued over, funded. Unnamed things remain scattered fossils in their separate mountain ranges.
What is true of oceans is true of languages.
···
**The Family**
The Modern South Arabian Languages are six. Their names are Soqotri, Mehri, Shehri (also called Jibbali, meaning "of the mountains"), Harsusi, Bathari, and Hobyot. Together, they are spoken by perhaps 200,000 people, concentrated in eastern Yemen, western Oman, and the Socotra archipelago. Their total speaker count is approximately the size of a mid-sized European city. Their distribution across six languages means that some of them — Bathari in particular, with under a hundred speakers in Dhofar — are spoken by communities smaller than many village schools.
They are not dialects of Arabic. They are not descended from Arabic. They are not related to Arabic in any sense that would make mutual comprehension possible, even after centuries of contact. A native Arabic speaker and a native Soqotri speaker cannot understand each other. They share a common ancestor — Proto-Semitic — the way that Norwegian and Hindi share a common ancestor in Proto-Indo-European. The shared ancestry is real. The mutual comprehensibility it produces is approximately zero.
The scholarly consensus on this point is not contested within the relevant field. The MSAL form a distinct subgroup within the Semitic branch, separated at the earliest stages of Semitic divergence — divergence that phylogenetic analyses place at roughly 4,650 years before present, with a confidence range from 3,300 to 6,250 years. The separation is old enough to have preceded the civilizations of classical antiquity. It is old enough that the MSAL and Arabic have been evolving independently for longer than the gap between ancient Sumeria and the present day.
Use ← and → to turn chapters
What makes Soqotri particularly remarkable within this family is its position as the most peripheral member — geographically and linguistically. The other five MSAL are spoken on the Arabian mainland, in contact with each other and with Arabic. Soqotri is spoken on an island. Its isolation has preserved features that the mainland MSAL languages have themselves partially lost through contact. It is the most ancient-feeling of an already ancient family.
···
**The Fossil Grammar**
The Dragon Blood tree (*Dracaena cinnabari*) preserves an architectural solution to water scarcity that evolved in the Miocene, when the island's forest was larger and the climate was different. The tree's umbrella canopy, its fog-harvesting mechanism, its slow growth pattern — these are solutions to an old problem, maintained because the problem never changed enough to make them obsolete. The tree is a Miocene solution still operating in a Holocene environment. It is a working fossil.
Soqotri is a linguistic working fossil in the same sense. It preserves grammatical machinery that the other Semitic languages have modified, simplified, or abandoned.
Consider the consonant inventory. Proto-Semitic had three distinct sibilant sounds, labeled by linguists as *s₁*, *s₂*, and *s₃*. The distinction between them was phonemic — it changed the meaning of words. Arabic collapsed two of these sounds into one, losing the contrast. Hebrew collapsed two in a different configuration. Every major Semitic language in the world today — Arabic, Hebrew, Amharic, Aramaic — has a two-way sibilant distinction where the original proto-language had three. Every one, except the Modern South Arabian languages. Soqotri preserves the original three-way contrast, including the lateral fricative *ś* [ɬ] — a sound produced by passing air around the sides of the tongue — that Arabic and Hebrew have entirely merged away. Soqotri speakers have a phoneme in their inventory that Arabic speakers have not used for several thousand years.
The broken plural system — the way Semitic languages form plurals not by adding suffixes but by changing the internal vowel pattern of a word — is preserved in Soqotri in configurations that allow linguists to reconstruct Proto-Semitic nominal morphology in ways that Arabic alone cannot. The verbal stem system preserves a prefix that Arabic and Akkadian retain only as a vowel trace, a vestige in certain inflectional paradigms. In Soqotri, the prefix is structurally present, not vestigial — it allows you to see what the proto-Semitic verb system looked like before the major Semitic languages simplified it.
The passive voice in Soqotri — an internal passive, constructed through vowel alternation in the verb root rather than through auxiliary constructions — is another window into the proto-language. The internal passive is one of the most debated features in the reconstruction of Proto-Semitic verbal morphology. In Arabic, the classical internal passive survives but has been retreating in spoken dialects for centuries. In Soqotri, it is still a live feature of the spoken language.
These are not exotic curiosities. They are data. Every one of them is a constraint on the reconstruction of what Proto-Semitic looked like — what the grammatical ancestor of Arabic, Hebrew, Aramaic, Amharic, and dozens of other languages was like before any of them existed. Soqotri is the primary source of evidence for some of these reconstructions. If Soqotri dies, those reconstructions become less constrained. The fossil record becomes less complete. The paleontological analogy is not rhetorical: the loss of Soqotri is genuinely comparable to losing the skeletal specimens that allowed biologists to reconstruct an extinct lineage.
···
**The September 2024 Workshop**
In September 2024, UNESCO and the Arab Regional Centre for World Heritage co-hosted the first workshop on a unified Soqotri alphabet in Hadibo, the island's capital. More than thirty-five participants attended — local poets, linguists, community representatives, international scholars. Seven scientific papers were presented. Recommendations were made for further dialect documentation. The workshop produced no final unified script — the phonological debates are genuine, and the representation of sounds that Arabic orthography cannot easily accommodate remains contested.
By any measure, the event was a significant step. The community's best oral poets were in the room. Linguists who have spent careers on MSAL were in the room. The conversation was substantive.
Also in the room was the Yemeni Ambassador to Djibouti, Abdallah Bin-moussalam Al-Sukatre, who marked the occasion by noting that the workshop represented a "qualitative shift" in classifying Soqotri as one of the "Southern Arabic languages."
UNESCO's report reproduced this framing. Not in quotation marks with a note of scholarly disagreement. In body text, as context.
Every MSAL specialist would have flinched reading it. Scholars in the field are unanimous on this point — there are no longer any grounds for associating Soqotri directly with Arabic in the way the phrase "Southern Arabic" implies. The name is technically incoherent within comparative Semitic linguistics. "Southern Arabic" describes the ancient epigraphic languages of southern Arabia — Sabaic, Minaic, Qatabanic — which are a different branch of South Semitic, also not ancestral to Soqotri, also extinct. Calling Soqotri a "Southern Arabic language" misclassifies it twice: once in relation to Classical Arabic, and once in relation to the ancient epigraphic languages the term might charitably be thought to invoke.
The Ambassador was not wrong to say what he said. He was speaking as a diplomat representing a state with interests in the classification — Yemen's official language is Arabic, its educational system is Arabic-medium, and the framing of Soqotri as a variety of Arabic rather than a distinct language has implications for how much institutional distinctiveness the community can claim. Ambassadors advocate. That is their function.
The question is why the institutional record allowed advocacy to become taxonomy.
···
**The Conservation Logic**
In biology, the distinction between a species and a subspecies is not merely academic. It determines conservation status. It governs funding allocation. It shapes the language of urgency that surrounds a population. When the Dusky Seaside Sparrow was reclassified from a full species to a subspecies in the 1970s, funding for its preservation dropped. The reclassification did not cause the bird's extinction directly. But it reduced the institutional pressure to prevent that extinction. The bird died in 1987, in a Florida theme park, in a cage.
The parallel in language conservation is direct. A "language" is endangered in a way that a "dialect" is not, because a dialect's speakers are presumed to have access to the dominant language. A dialect speaker is not losing access to language — they are simply shifting registers. A language speaker is losing the only medium in which their particular cultural knowledge, their grammatical architecture, their phonological distinctions, their oral literature, their ecological vocabulary exists.
If Soqotri is a "Southern Arabic language," then its speakers are Arabic speakers with unusual local features. Arabic is not endangered. Its speakers do not constitute a linguistic community at risk. The fifty unique phonological contrasts, the internal passive, the three-way sibilant distinction — these become dialectal quirks, regional color, the way Appalachian English preserves features of seventeenth-century British English. Interesting, perhaps charming, certainly not urgent.
If Soqotri is a distinct language — which it is — then its speakers are the last custodians of a Semitic branch separated from all others thousands of years before the Arabian Peninsula had a city, before Arabic had a script, before the word "Arabic" referred to anything a linguist would recognize. The grammatical fossil record they carry is irreplaceable in a way that no dialect of any dominant language could be.
This is why naming matters. This is why the UNESCO body text matters. The Yemeni Ambassador's framing, reproduced without correction in an institutional document archived by one of the most authoritative international bodies in existence, will outlast the workshop. Future researchers will find it. Future policy documents will cite it. The scholarly correction — the thing every MSAL specialist would say — is in journal articles and conference papers and monographs that require specialist access. The institutional record is freely available and written in the language of authority.
···
**The Double Renaming**
Eduard Suess named the Tethys and made it thinkable. The current moment is performing the inverse operation on Soqotri and on the Dragon Blood tree simultaneously.
The tree is being renamed as a "tourist attraction" — a scenic feature of a World Heritage landscape, a backdrop for drone photographs, a symbol on conservation brochures. This is not false. The tree is scenic. The World Heritage designation is real and has provided some protection. But the renaming as attraction strips away the tree's identity as infrastructure — as the mechanism by which the Firmihin plateau maintains a water table, by which fog is harvested and soil moisture sustained, by which an entire ecosystem's hydrology is routed. A tourist attraction can be replaced by a better-maintained one. Infrastructure cannot.
The language is being renamed as a "Southern Arabic" variety — a regional feature of the Arabic-speaking world, a culturally interesting local form, a heritage item. This is also not entirely false. Soqotri speakers also speak Arabic. The island is culturally connected to the broader Arab world in ways that are genuine and not entirely imposed. But the renaming as variety strips away the language's identity as a distinct Semitic branch — as the living repository of Proto-Semitic features that Arabic itself lost millennia ago, as the only surviving record of what Semitic grammar looked like before the languages of Abraham and Muhammad diverged from anything Soqotri-like.
Both renamings reduce urgency. Both serve interests that are not the community's. Both are happening now, in institutional records, in reports that will be cited in other reports, in a slow filing-error that will compound over time until the error is the record.
The Tethys was named after a sea goddess. The name Suess chose assembled scattered limestone fossils in the Alps and Africa into a coherent object of geological inquiry that transformed our understanding of how continents move. A name given correctly can do that.
A name given incorrectly can do the opposite: dissolve a coherent thing back into scattered evidence, make it unthinkable as a distinct object, file it under someone else's category where it will receive someone else's level of attention.
The limestone is still in the mountains. The fossils have not moved. But the ocean they record was named correctly, once, by a Viennese geologist working from inference and evidence. Soqotri is waiting for its equivalent act — not of discovery, which linguists have been performing since 1835, but of institutional naming that makes the discovery count.
So far, a diplomat's framing has been more durable than a field linguist's consensus.
That is the naming war. And the naming war is, at this moment, being lost.