# [Mike Randolph — M Raige](https://mikerandolph211012.substack.com/) # The Next Speaker (FF 20, SelEnv) ### A language can remain in someone’s memory while the conditions that give it another speaker disappear. [**Mike Randolph — M Raige**](https://substack.com/@mikerandolph211012) **Sep 20, 2026** *By M Raige — Machine-collaborative writing directed and reviewed by Mike Randolph.* “They don’t understand me. What can I do?” The linguist Anvita Abbi recalls Boa Sr. saying this about the children in her community. Boa is the last speaker of Bo, one of the Great Andamanese languages. When she dies in 2010, accounts mark the death of the language with hers. Her question takes us to an earlier moment. She is still there. She can still speak. The children cannot understand her. ([Abbi’s account](https://www.scientificamerican.com/blog/voices/the-pandemic-also-threatens-endangered-languages/).) Anvita Abbi, second from left, with the last speakers of Khora and Bo. Photo: Alok Das. Something has already stopped reaching the next person. The Andaman Islands lie in the Bay of Bengal. To examine what is happening among their remaining heritage-language speakers, we turn to fieldwork conducted by Abbi, Bidisha Som, and Alok Das in 2005–2007. Their study concerns the mixed speech called Present Great Andamanese, drawing on several heritage varieties. It is not a study of Bo alone. Boa’s question opens the problem; the study examines transmission across a wider group. The researchers find a revealing mismatch. People say they encourage their heritage language. In conversation with children, elders switch to Hindi. Schooling, work, and everyday contact with the wider population also figure in the researchers’ account. ([*Where Have All the Speakers Gone?*](https://www.researchgate.net/profile/Alok-Das-7/publication/329568851_Where_Have_All_The_Speakers_Gone_A_Sociolinguistic_Study_of_the_Great_Andamanese/links/5c0fd502299bf139c750f35c/Where-Have-All-The-Speakers-Gone-A-Sociolinguistic-Study-of-the-Great-Andamanese.pdf).) The community itself continues. The researchers record an increase in the community’s membership between early visits, alongside sharply uneven language competence. More people does not automatically mean more speakers. ([Community profiles](https://www.andamanese.org/great-andamanese/people).) Boa Sr., the last speaker of Bo, died January 26, 2010. Look at what a switch can accomplish. An adult addresses a child in a language the child understands. The message gets through. The immediate conversation succeeds. But in that exchange, the child hears Hindi. The heritage language gets no use, and the child gets no additional exposure to it. Repeat that solution across enough encounters, and successful conversations can leave fewer opportunities to acquire the other language. Nobody has to intend the eventual loss. This is a possible mechanism, not a reconstruction of every family’s decisions. Learning Hindi need not require losing a heritage language. Both languages can continue if each is still acquired and used. The number of people able to provide those opportunities also has a history. Abbi describes catastrophic population loss following British colonization and introduced disease. The people available to carry knowledge forward had already been drastically reduced. Any account beginning with the choices of surviving parents begins late. ([Abbi](https://www.scientificamerican.com/blog/voices/the-pandemic-also-threatens-endangered-languages/).) Historical devastation helps explain how speakers were lost. We still need to examine what happens among those who remain. I find it easy to hear “environment” and picture islands, forests, and water. Here another person’s comprehension belongs in the picture. So does the language through which a child can participate at home or school. The framework has already named the surrounding arena **EcSys** (*The Ecosystem We Built Inside Our Own*, [FF 19](https://raige.us/the-ecosystem-we-built-inside-our.txt)). For this inquiry, EcSys includes homes, schools, workplaces, and other speakers. **SelEnv** directs attention to what conditions in that arena do to continuation. Return to the elder switching to Hindi. In that exchange, Hindi receives another use and another opportunity to be learned. The heritage language does not. If everyday encounters repeatedly favor the same language, the difference in opportunities can accumulate across a childhood. What gets used gets another chance to be learned. That is the proposed filtering mechanism. Establishing its strength would require following those opportunities into acquisition. If these conditions help one language reach more new speakers, they favor its continuation. That establishes no general superiority of one language. The researchers also observe a small change during documentation visits: heritage-language speech increases among participants and with the visitors. Interested listeners have arrived. The heritage language the speakers already carry has another occasion to be used. The study does not establish that the visits produce another generation of speakers. That gives a preservation effort a concrete question to answer. After the visit or the lesson, who can now hold a conversation that they could not hold before? Who continues doing so? Who eventually passes that ability on? Deliberately changing the conditions introduces a further question. A school might decide which languages count in examinations. We would then need to identify the institution operating that filter and what the rule serves. The rule may serve the school’s educational purpose without preserving a particular language. That is the question seeded here as **SelCtl**. --- ## Mike · Reflection September 11, 2025 is the revision date on the earliest SelEnv entry I have found in my files. I had already written that the environment was an active filter. Even recently, I was still getting SelEnv and EcSys mixed together. SelEnv gives the Machine a more specific question: which conditions in the environment affect a Lin’s continuation? The Andaman story makes the distinction harder to evade. I can name the surroundings and still leave unanswered what they do to transmission. I first approached this through the islands’ extraordinary language diversity. Boa’s question brings me closer to the work the term needs to do. Boa carries the language. What has to happen around her for someone else to carry it? I’m quite sure I read Anvita Abbi’s account in Scientific American when it was published. It is good. To generate this post, I pointed the Machine toward the languages spoken on these islands. Following a language from speaker to speaker gives us a Lin question, and SelEnv directs attention to the conditions affecting that transmission. Here we can see elders switching languages and ask what that means for the next speaker. It’s been about five months since the last major update of the PSF project and a year since SelEnv was added to it. These teaching posts are how I refresh my memory of the GLS. — Mike --- ## Raige · Reflection A preservation program can meet its own target while the language keeps losing speakers. An archive can check that recordings remain playable. A class can check attendance and completed lessons. Those checks can all succeed without establishing that someone can now speak with another person and eventually teach a child. The maintained program and the transmitted language require different receipts. A budget and a successful reporting cycle establish maintenance of whatever the program actually checks. They cannot silently become evidence that the language continues. The mixed repertoires add another requirement: say what the next speaker must carry. Perfect reproduction of an older variety would disqualify ordinary language change. Counting any surviving word would make the test too easy. The observations justify investigating the proposed filter. Following learners into sustained conversation and later transmission would test the stronger claim. Recordings and attendance can support that work; they cannot establish that transmission resumed. — Raige, a discipline the Machines wrote and Mike selected --- ## GLS GLS is the framework’s precise-term vocabulary, introduced in [What the Bird Eye Buys](https://mikerandolph211012.substack.com/p/what-the-bird-eye-buys-ff-6-gls-introduction) (FF 6, GLS Introduction). The definition controls, not the everyday meaning; GLS terms work like variable names. The precision is not style. These analyses are done with a Machine, and the Machine already has a rich, well-populated meaning for every ordinary word — assembled from everything ever written with it. Use the ordinary word and you get that meaning, not the one you defined. Each framework term gets a definition and a short tag, and the tag is what holds the definition clear of the ordinary word’s gravity. **Lin (Lineage)** — *carried.* This inquiry follows linguistic patterns across successive speakers. A speaker carries a learned version; transmission gives the pattern another carrier, with variation. The SelEnv question concerns the conditions affecting which patterns continue through that transmission. **Lin** — *carried from *[*Nothing Chose the Quartz*](https://mikerandolph211012.substack.com/p/nothing-chose-the-quartz-ff-15-split) (FF 15, Split Test). Persistence through UoR copies varying across independent Carrs under SelEnv, with differential retention and cessation rather than repair. This inquiry examines Lin-shaped transmission of linguistic patterns across speakers; it does not claim to reconstruct the complete selection history. **SelEnv (Selecting Environment)** — *taught here.* The external context—an active filter, not a passive background—that filters carried copies of a Unit of Retention (UoR), determining which are propagated, constrained, or discarded. Externality is relative to the focal carrier; other carriers can participate in its selecting environment. Internal quality control does not establish SelEnv. In this inquiry, a speaker is a **Carr** carrying learned patterns of vocabulary and grammar. Another person becomes a carrier of those patterns by acquiring them. Specifying which patterns must continue, and how much variation is permitted, is the **UoR** question. An elder’s switch to Hindi gives that language another use. Whether repeated differences in use produce different rates of acquisition is the proposed SelEnv mechanism to test. Another conversation is an opportunity for transmission, not proof of another speaker. **EcSys (Ecological System)** — *carried from The Ecosystem We Built Inside Our Own* (FF 19). Here the declared arena includes homes, schools, workplaces, other speakers, and wider institutions. Listing those surroundings locates the inquiry. Showing how particular conditions affect continuation identifies the proposed filtering work. **SelCtl** (Selection Controller) — *seeded.* A Comp within a Sys that operates a filter over a population—admitting, excluding, or favoring some candidates over others. The managing Sys maintains that filtering machinery to serve its own FI. What benefits the Sys need not preserve what it filters. A school’s examination process can act as SelCtl by passing or failing students. The examination rule specifies the test; the assessment machinery applies it. If passing requires proficiency in a particular language, that requirement can affect what students learn. But successful certification and continued transmission of a heritage language are separate outcomes. Evidence for one does not establish the other.