Daniel's been working through a series on remote communities, islands mostly, and this one lands a lot closer to home for him than most. He grew up in Ireland, and he says he spent a lot of his childhood wondering about the people on the inhabited offshore islands. Not Achill, he's careful about that one, because Achill has a land bridge and he thinks that makes it a different category of place. And there's a second thread running through his question that I want to flag up front because it's really the spine of the whole thing. He wonders whether his interest in prepping and self-sufficiency dovetails with his fascination with islands. The image that appeals to him is a community disconnected from the surrounding world that nonetheless has everything it needs. He calls it a geographic antidote to the excesses of consumerism, and honestly that's a good phrase, I want to come back to it. And then the concrete questions. How many inhabited offshore islands does the Republic actually have? Who lives on them today? What's the linguistic and employment makeup out there, given that these were the places where Irish survived when the mainland was losing it? And finally, how does a small island state itself go about providing the essentials of state infrastructure to people living on smaller islands off its coast?
That last question is the one I want to spend most of our time on, because the answer contradicts Daniel's premise in a way I think he'll find more interesting than a simple yes.
Contradicts it how?
The romantic image is the self-sufficient island. The reality is that the islands are almost entirely dependent on the state, and deliberately so. That's not a failure. It's a policy.
Okay, so let's earn it. Big number first.
Eight thousand nine hundred and thirty-one.
That's the total population of the offshore islands, from the 2022 census. Up two percent from eight thousand seven hundred and fifty-six in 2016. And that two percent matters more than the size of it suggests, because it's the first meaningful increase after well over a century of the other direction.
The century of the other direction, in one sweep, so we can leave it behind us. Thirty-four thousand two hundred and nineteen in 1841, the pre-Famine peak. Seventeen thousand four hundred and eighty-eight by 1951. Fourteen thousand four hundred and seventy-three in 1961. Eleven thousand seven hundred and thirty-seven in 1971. Nine thousand five hundred and sixty-nine in 1991. Then the low point, eight thousand eight hundred and eighty-four in 2002. A small recovery to eight thousand nine hundred and eighty-three in 2006, nine thousand and eighty-seven in 2011, then a drop back to eight thousand seven hundred and fifty-six in 2016, and then this uptick.
So you can see the shape. A cliff edge, then a long shallow slope, then a wobble around the bottom.
That's it exactly. And there's a counting problem sitting underneath all of it that I want to name before we go any further, because it changes what you think the number means.
The caveat.
The CSO counts offshore islands broadly. So the raw list of named islands is much longer than the list of islands with anybody permanently living on them. The commonly cited figure for permanently inhabited offshore islands in the Republic is somewhere between twenty and thirty. But the edges of that figure are soft. You've got islands with one to fifteen people. You've got a great many more sitting at zero.
And then Daniel's Achill point, which is right, and it's not pedantry. Achill's got two thousand three hundred and forty-five people. That's the biggest number on the whole list. But it's connected by a road bridge, the Michael Davitt Bridge.
Right. And it's not only Achill. Gorumna has a thousand and forty-four, and it's road-bridged. Valentia, six hundred and fifty-eight, road bridge plus ferry. Lettermore, five hundred and forty-two, bridged. Inch, four hundred and thirty-eight, causeways. Strip out everything with a road connection and the true offshore population drops to somewhere around six and a half thousand.
So we're really talking about a small town's worth of people scattered across twenty-odd rocks in the Atlantic, and every one of those rocks is its own administrative universe.
Which sets up the actual question. These are tiny, aging, historically shrinking communities that appear to have just recently stabilised. What keeps them viable?
So that's the scale. Now who actually lives out there?
Start with the human end of the distribution, because the averages hide it completely. Dursey Island, off the Cork coast, had three residents in 2022.
Three.
Three. Inishbiggle, twelve. Gola, fifteen. Owey, twenty-seven. Whiddy, thirty-one. Inishturk, fifty-six. Those are the islands where the population is a number you could count on your fingers and have fingers left over.
And at the top?
Inishmore, the big Aran island, eight hundred and twenty. Arranmore in Donegal, four hundred and seventy-eight. Inisheer, three hundred and forty-three. Then you're into the intermediate band. Inishmaan and Inishbofin at a hundred and eighty-four each. Tory at a hundred and forty-one. Clare Island at a hundred and thirty-eight. Cape Clear and Sherkin at a hundred and ten each.
So there's a real spread. You've got islands with hundreds of people and proper infrastructure, and you've got islands where the entire permanent population would fit in a minibus.
And Dursey is the extreme case, because that's a community where you can't really talk about institutions at all. Three people. There's no school, no shop, no pub, nothing. It's three residents on a rock, and I want to come back to how they get on and off it, because that's a story in itself.
But the decline first. What actually drove it?
The Famine and post-Famine emigration did the initial devastation, the same as everywhere in the west of Ireland. And then there's a second phase that's more specific to islands, and it's the phase I think Daniel will find the most uncomfortable. In the 1950s and 1960s, the Irish government forcibly evacuated several islands.
Forcibly.
The Blaskets in 1953. Inishark, off Mayo, in 1960. And the reason given was weather. Continuous bad conditions could leave islanders unable to reach the mainland for months at a stretch. The state looked at that and judged it unsustainable.
So the ultimate disconnection scenario, and the state's answer was to end the community rather than sustain it.
That's the honest reading. Which is worth holding on to, because it tells you how the state's thinking has changed since. The whole policy apparatus we're going to talk about later is a reaction against that logic.
Okay. So that's the mechanism of decline. What about the recent stabilisation? Is it real?
It's real but fragile. Some islands stabilised or grew a little during the Celtic Tiger years, and then 2022 gave that modest overall uptick. But the Gaeltacht data from 2022 points at a problem that's very much of the present, and it surprised me. Some of the recent decline in Irish-speaking populations on the islands and in the Gaeltacht is being attributed to the housing crisis.
Housing.
Young people who grew up in those areas can't afford to buy or rent locally, so they leave. Which is a very modern way for a remote community to die. It's not the weather cutting you off from the mainland. It's the property market.
There's something almost blackly funny about that. The Atlantic couldn't finish these places off but the housing ladder might.
Take Inishbofin as the worked example, because its arc is the whole story in one island. In 1837 it had one thousand four hundred and sixty-two people. By 2011 it was down to a hundred and sixty. Then it recovered slightly to a hundred and eighty-four in 2022.
A hundred and sixty to a hundred and eighty-four isn't a recovery so much as the handbrake finally catching.
That's about right. And tourism is now the main industry there. Three hotels and a hostel.
Right, so hold the employment thread, we'll come back to it. The language question is where the islands punch far above their weight, and it's the part of this Daniel grew up hearing about.
The Gaeltacht had a population of a hundred and two thousand nine hundred and seventy-three in 2022. About two percent of the state. Of those, sixty-five thousand one hundred and fifty-six were Irish speakers. Sixty-six percent.
At first glance that sounds healthy. Two-thirds of the Gaeltacht speaks the language.
The headline is a trap. The number that matters is daily speakers, and daily Irish speakers in the Gaeltacht had fallen to twenty thousand two hundred and sixty-one.
Out of a hundred and three thousand. So about one in five.
And that's the number that predicts survival, because a language that's only spoken occasionally isn't being transmitted. There's a threshold the academics use, and it's two-thirds daily use. Below that, the community language starts to collapse. Research published in 2015 found that Irish was spoken daily by two-thirds or more of the population in only twenty-one of a hundred and fifty-five electoral divisions in the Gaeltacht.
Twenty-one out of a hundred and fifty-five.
And that same year, a report warned that on the indicators at the time, Irish would cease to be used as a community language in the Gaeltacht within ten years. Which puts the deadline around 2025. The 2022 census confirmed continued decline rather than a turnaround.
So that warning's date has now passed.
It has. Though I'd put a caveat on the framing, because "cease to be a community language" isn't an event with a date on it. It's a slope. The language is still spoken, still taught, still broadcast. What's gone is the thing where a child grows up hearing it on the street as the ordinary language of the place.
Which is exactly what the islands preserved. That's the part of Daniel's question I want to sit with. Why did the islands hold it when the mainland didn't?
Distance and isolation, mostly. The offshore islands, especially the Aran Islands, Inishmaan, Inisheer, Tory, and the Connemara islands, kept Irish as the everyday language far longer than the mainland did, because the mechanisms that pushed English through the rest of the country, the schools, the towns, the administration, the emigrant ships, arrived later and weaker. And the Aran and Connemara islands are still among the strongest Gaeltacht areas. Cape Clear is a Cork Gaeltacht island. So they remain disproportionately important to the language's survival.
And yet.
And yet the single most counterintuitive number in the whole subject is that by 2022, Dublin and its suburbs had the largest number of daily Irish speakers of anywhere in the country. Sixteen thousand four hundred and forty. Twenty-three percent of all daily speakers nationwide.
The language has moved to the city.
It's migrated from a rural community language to an urban network language. Which is a form of survival, and it's a real one, but it's a different thing from what the islands were doing.
And the state's role in the language is a lot of machinery. Raidió na Gaeltachta, based in Casla in Connemara with studios in Donegal, Kerry and Mayo. TG4, the Irish-language television station, out of Baile na hAbhann, also Connemara. Those aren't small operations.
And the Gaeltacht Act of 2012, which created twenty-six Language Planning Areas. It was criticised at the time for weakening the language-based criteria for what counts as a Gaeltacht boundary. The argument was that you got a more workable administrative map at the cost of the thing the map was supposed to measure.
So if Irish is declining, what are people actually doing for work on these islands?
Tourism, on most of the accessible ones. Inishbofin has three hotels and a hostel, and the activity list is exactly what you'd expect. Scuba diving, walking, cycling, horse riding, sailing, kayaking, angling. The Aran Islands, Inishmore, Cape Clear, all tourism-driven. It's the main employer on the islands a ferry can reach.
And the ones a ferry can't reach reliably?
Fishing and aquaculture. That was historically the backbone and it still matters on a lot of them. Small-scale farming persists too, mostly sheep grazing, but it's marginal. You can't run an island economy on sheep anymore.
There's also the state-supported enterprise layer.
Údarás na Gaeltachta, the Gaeltacht development authority, supports client companies. In 2006, twenty-four hundred and thirty-six people were employed full time in Údarás client companies in the Donegal Gaeltacht alone. So it's a real employer in the regions, not a token body.
And then the seasonal thing, which is bigger than people outside Ireland realise.
The Irish-language college summer courses. Thousands of students go to Gaeltacht and island areas every summer, and they stay with host families, the bean an tí, and that money supports the local economy for the season. For a lot of island households, that's a significant part of the year's income.
And the cultural output from some of these places is absurd when you line it up against the population. Clannad, Altan, Enya. All out of Gweedore in the Donegal Gaeltacht.
A parish's worth of people produced a meaningful chunk of Ireland's musical exports. Though I'd say that's less about the island and more about what happens when a tight community with nothing else to do in the evenings takes music seriously.
And that brings us to the machinery of the state, because none of this works without it. Which is where Daniel's premise starts to bend.
Start with the policy layer. There's a national policy specifically for the islands. It's called Our Living Islands, launched in 2023 and running to 2033. Ten years, cross-government, and the scope is everything. Housing, health, education, broadband, transport, sustainable development.
A dedicated ten-year policy for eight or nine thousand people.
Which tells you the state treats them as a category of their own. Until recently the relevant department was Rural and Community Development, and since 2025 it holds the Gaeltacht portfolio as well. So you've got one department holding both the rural development brief and the language brief, which is not an accident.
Then the transport, which is the lifeline in the most literal sense.
State-subsidised passenger and car ferries to most inhabited islands. Cleggan to Inishbofin, Roonah Quay out to Clare Island and Inishturk, Burtonport to Arranmore, Baltimore to Cape Clear and Sherkin, Rossaveal to the Aran Islands. The Aran Islands also have flights from Inverin, Connemara Airport. And Dursey Island, the one with three residents, is reached by Ireland's only aerial tramway. A cable car.
Three people and a cable car. There's a cost-per-head figure in there that nobody in the Department wants to put on a poster.
There is, and the honest answer is that nobody pretends the ferries pay for themselves. They're run as essential public infrastructure. That's the lifeline principle. The subsidy is the point, not a problem to be solved.
But it isn't all tidy.
It isn't. Inishbofin has airstrips that are reportedly unlicensed. A 2019 report in the Irish Examiner noted ten million euros of airstrips sitting idle for a decade. There's broadband work through the National Broadband Plan and remote-working hubs, which is the modern version of the same idea, keeping people on the island by keeping them connected. Health is a persistent problem. GPs, nurses, emergency services, with air and sea ambulance arrangements as the actual lifeline. Primary schools exist on many islands, but secondary education usually means moving to the mainland or boarding. Post offices, shops and pubs on the larger islands. Nothing at all on the smaller ones.
So if you add all that up, the romantic image of the self-sufficient island runs straight into the reality of the subsidised ferry.
Which is where I think Daniel's question gets interesting, and I want to be careful, because the romance isn't wrong. It's just pointing at a different era.
The pre-modern islands.
Historically they were near-self-sufficient, and not in a sentimental way. Fishing, small farming, turf cutting, communal labour, the meitheal, and barter. Inishbofin literally has no trees or forests, because the wood got cut for fuel and the salt air stopped anything regrowing. That's what living inside your ecological limits actually looks like. You don't get to keep trees.
And then the 1950s and 60s evacuations are the other side of that same reality. Self-sufficiency also meant that when the weather turned, you were on your own for months. The state looked at that and decided it was unacceptable.
Today's islands are not self-sufficient. They depend on subsidised ferries, air services, broadband, and state transfers. The picture Daniel has in his head, disconnected but complete, doesn't describe a real Irish island in 2026. What describes it is connected but remote.
Deliberately connected.
Deliberately connected is better. The state's whole posture is that these communities should stay plugged in, and it spends real money keeping them plugged in. Which means the prepping appeal, if it survives contact with reality, survives as a mindset rather than as a place.
Which I think is the right place to land it, actually. Resourcefulness, mutual aid, living with limits. Those are things communities can have wherever they are. The autarky is the part that doesn't hold up.
There's a counterweight I don't want to lose, which is that the community resilience on these islands is real, even if the economics aren't self-contained. GAA clubs. Inishbofin hosted the All-Ireland Islands Football Tournament in 2008 and again in 2017. Community centres, libraries, cooperative structures. Social capital is the thing they actually have in surplus.
Island agency shows up in less comfortable places too.
The 2023 case. Trinity College Dublin returned skulls taken from Inishbofin in 1890 by two academics, Haddon and Dixon, to island residents on the twelfth of July that year. The curator, Ciarán Walsh, called it the first event of its kind in Ireland and of international significance. A community of a hundred and eighty people got an institution to give back what it had taken a hundred and thirty years earlier.
Which is a reminder that these aren't passive places waiting to be rescued by policy.
Hilbert: The post came twice a week.
Sorry?
Hilbert: The post came twice a week. And nobody minded. What they minded was the diesel. There was one tractor on the island and it belonged to a man called Tadhg, and you'd give him a shout if you needed it, and he'd say yes, and that was the arrangement. There was no rota. I said there was a rota to a fella once and he looked at me like I'd said something about his mother.
How long were you out there?
Hilbert: Long enough to know the ferry timetable. Which is the thing nobody tells you. The ferry doesn't fit around your week. Your week fits around the ferry. If the boat goes Tuesday and Friday, then Tuesday and Friday are your week. Everything else arranges itself around that. You don't decide when to go to the mainland. The mainland decides when you can come.
That's a much better description of dependency than anything in the policy documents.
Hilbert: It's not dependency. It's a timetable. The difference is that a timetable you can argue with.
The thing you said about the tractor. That's the part I keep turning over. Daniel's image is a community that has everything it needs. The reality is a community that has one tractor and shares it.
Hilbert: That's not the same thing, no.
One tractor isn't self-sufficiency. It's a very high-trust form of scarcity.
Hilbert: Well. Anyway, the levels on the last segment were low, I'll fix it before we go out. And the second half is running long, we'll need to lose about forty seconds in the edit.
If we take the practical version of Daniel's image seriously, the island isn't a place that needs nothing from the outside. It's a place where the things that come from the outside are few enough that everyone knows what they are and who's using them.
Which is a completely different thing from a prepper's bunker, and much harder to build.
It's harder because it isn't something you assemble. It's something a community either has or doesn't. You can buy a year's worth of fuel and call that preparedness. You can't buy the arrangement with the tractor.
The numbers back that up. The Gaeltacht's daily speakers have fallen to twenty thousand two hundred and sixty-one. Two-thirds daily use in just twenty-one of a hundred and fifty-five electoral divisions. That's the decline that deliberate connection hasn't been able to reverse.
Though I'd separate the two questions. Population stabilisation and language survival are not the same problem. The state's lifeline model has arguably worked on the first one. Eight thousand nine hundred and thirty-one people, up two percent. On the second, the trend is still down, and the housing crisis is currently doing what the Famine and the evacuations did in earlier generations.
The open question is whether the 2022 uptick is a real turn or just a blip, and whether the islands can hold their Irish-speaking communities when the young people who'd carry them can't afford a house locally.
Whether the lifeline model is sustainable at all as these populations age.
Our Living Islands runs to 2033. So the next decade is the test. Whether deliberate connection actually keeps these communities viable, or just keeps them on life support.
It's worth asking what the difference even is.
If you take one thing from this, it's that remoteness and disconnection aren't the same thing, and the islands are a demonstration of the gap. They're remote. They are not disconnected, and the price of that is a cable car for three people.
The romance is in the mindset, not the supply chain. What these places actually have that the mainland has lost is the habit of knowing exactly what everyone else needs and being willing to hand it over.
That's the show. Thanks as always to Hilbert Flumingtop for producing. This has been My Weird Prompts.
If you've got a moment, a review helps other people find us. We'll be back soon.
