The cemetery of the Protestant Cathedral at Belfast – Chesterton’s ‘The Sign of the Broken Sword’

You can read ‘The Sign of the Broken Sword’, with the original 1911 illustrations, here.


I have been a fan of G.K. Chesterton’s Father Brown stories since I was a child, but one point in ‘The Sign of the Broken Sword’, a short story first published in 1911, has niggled at me for almost half a century. I was reminded of this last month when I was staying in a hotel on one side of St Anne’s Cathedral in Belfast, and giving two lectures at the new Ulster University campus on the other side of it, so that I walked past it four times in the space of a few hours. The passage in question comes just after the halfway point in the story when Father Brown reveals to Flambeau, his French ex-criminal friend, the current location of the broken-off part of the titular weapon.

  “I cannot prove it, even after hunting through the tombs. But I am sure of it. Let me add just one more tiny fact that tops the whole thing over. The colonel, by a strange chance, was one of the first struck by a bullet. He was struck long before the troops came to close quarters. But he saw St. Clare’s sword broken. Why was it broken? How was it broken? My friend, it was broken before the battle.”
“Oh!” said his friend with a sort of forlorn jocularity. “And pray where is the other piece?”
“I can tell you,” said the priest promptly. “In the cemetery of the Protestant Cathedral at Belfast.”
“Indeed?” inquired the other. “Did you look for it?”
“I couldn’t,” said the priest with regret. “There’s a great marble monument on top of it; a monument to the heroic Major Murray who fell fighting gloriously at the battle of the Black River.”

The reason this passage has always niggled at me is very simple. There is no cemetery at St Anne’s Cathedral, the Protestant (ie Church of Ireland) Cathedral in Belfast. In fact, only one person is buried on the cathedral’s premises at all: Edward Carson, the Unionist leader and founder of Northern Ireland. In 1911, when the story was published, he was alive and sinnin’ (he lived to 1935). St Anne’s Cathedral was devoid of tombs, inside and out, at the time when Chesterton was writing.

This is very unusual for cathedrals in Britain or Ireland, either Protestant or Catholic. Most Church of Ireland cathedrals are in ancient ecclesiastical centres which have seen better days. I did a quick check and all of the other Protestant cathedrals in Northern Ireland do have graveyards. Many big cathedrals also have many interments inside the building – St Paul’s in London has Lord Nelson and the Duke of Wellington; St Patrick’s in Dublin has Jonathan Swift. St Anne’s, as noted, has just the one.

But St Anne’s Cathedral in Belfast is very new as cathedrals go. It serves two dioceses, Connor (which is roughly equivalent to County Antrim) and Down (which is not equivalent to County Down), each of which also has a cathedral of its own (in Lisburn and Downpatrick respectively). The foundation stone for St Anne’s was laid in 1899 and the cathedral was consecrated in 1904; this is long after the fictional battle of the Black River, which we are told was at least twenty years before 1911. It is located in a city centre site with commercial and residential buildings pressing around it. The south transept was not completed until 1974 and the north transept was not completed until 1981, when I was already a teenager.

St Anne’s Cathedral in the early 20th century. The block immediately north was cleared for the Art College in the 1960s, and the area to the southwest for Writer’s Square more recently.

Chesterton’s Major Murray, if buried in Belfast, would have been interred at the Clifton Street Cemetery if his family had a concession there, or up the Falls Road in the Belfast City Cemetery if not. Though thinking about it, it would be really unusual for even a very senior officer who had been killed in action abroad at that period to be brought back home. Looking at the 1899-1902 Boer War, the two British generals who lost their lives in the conflict, Penn Symons and Andrew Wauchope, are both still buried in South Africa.

We are told that Murray was a Presbyterian, which is unusual but not impossible. In the 1901 census, according to Barry Griffin’s data, although 88.65% of people in Ireland with the surname Murray were Catholics (like my great-grandfather), 5.24% were Anglicans (as the fictional Murray must have been to be buried in the fictional cathedral graveyard), concentrated especially around the shores of Lough Neagh with outposts that seem to be around what is now Newtownbreda and also Carrickfergus.

G.K. Chesterton had never been to any part of Ireland in 1911; he wrote a book called Irish Impressions after his first visit in 1918. (You can read it here.) He was instinctively sympathetic to Home Rule and unsympathetic to colonial wars such as the Boer War, which is clearly the basis for the fictional Brazilian war in the story – the popularity of Chesterton’s Brazilian leader Olivier with the British, years after the war had ended, must be a reference to the shift in the British attitude to the South African leader Jan Smuts at the same time.

I don’t really blame Chesterton for getting Belfast’s ecclesiastical geography wrong. The fictional British Invasion of Brazil is a much bigger invention than a graveyard in Belfast. (There was historically a dispute between Brazil and the UK about the border with what was then British Guyana, but there does not seem to have been any armed conflict and the issue was resolved by Italian arbitration in 1904.) Anyway, neither the graveyard nor the war is what the story is really about.


The bodies of both General Sir Arthur St. Clare and the Ulsterman Major Murray were retrieved after the battle of Black River – Murray found on the field, and St. Clare hanged from a tree. But the punchline is that St. Clare was a traitor, he killed Murray (who had found out his secret) with his own sword which broke in the process, and attacked the Brazilians, despite it being certain that he would lose with many casualties, so that Murray’s body would be unnoticed in the carnage. He was then strung up by his own men after the battle when they realised what he had done. The secret was kept by the British soldiers, who allowed it to be assumed that St. Clare was lynched by the Brazilians, and the fallen general was honoured as a tragic hero.

The narrative thrust of the story is that Father Brown works out what really happened from scraps of information and his knowledge of human nature. But the point of the story is that we should be wary of spoonfed narratives by the authorities about war heroes, or indeed about anything at all. One wonders if Chesterton had any particular person in mind – Baden-Powell? But he lived. Gordon? But his body was never recovered. In any case, the point is well made.

In the 2015 TV adaptation starring Mark Williams as Father Brown, the main action takes place in the 1950s with flashbacks to Dunkirk. The tableau is shrunk from national delusion to internal (and deadly) barracks politics. It’s nicely done, but it’s longer and less interesting than the original story.

Two very different books about Belfast

Belfast: Approach to Crisis, by Ian Budge and Cornelius O’Leary
Belfast: The Story of a City and its People, by Feargal Cochrane

I got hold of these two books in preparation for the lecture I gave in Belfast last month about the electoral history of the city, which you can watch here:

These are two very different books from very different times. The second paragraph of the third chapter of Belfast: Approach to Crisis is:

The cause of this increasing prosperity, the greatest that any Irish city has known, was twofold. First, the expansion of the linen industry which became fully mechanised between 1852 and 1862 with the rapid acceptance of the power loom.3 With the coming of the American Civil War Lancashire mills were starved of raw cotton and the Belfast mills soon found a new market for their high quality finished goods.4 The linen trade continued to expand until the 1870s,5 but while the labour force trebled between 1850 and 1875 (from 16,000 to 50,000), the proportion represented by adult male workers never exceeded one third.
3‘In 1852 there was only one power loom in Belfast. Ten years later there were 6,000.’ (Jones in Belfast, p. 109)
4The number of new buildings constructed annually between 1861 and 1864 ranged from 730 to 1,400 – thereby increasing the total valuation by about 20 per cent. (B.N.L., 2 January 1865.)
5The number of flax spindles in Ireland increased from 300,000 in 1850 to nearly 600,000 in 1860, and nearly one million by the end of the 1870s. This peak figure was never equalled – too much machinery had been installed for normal output, cf W. E. Coe, The Engineering Industry of the North of Ireland, pp. 60-61. In 1870 80 per cent of spindles and 70 per cent of power looms in the whole of Ireland were to be found in Belfast and its environs. D. L. Armstrong, ‘Social and Economic Conditions in the Belfast Linen Industry, 1850-1900’, Irish Historical Studies VII (September 1951), 238.

I don’t know Ian Budge (who is now 87) but I did know Cornelius O’Leary, an eccentric colleague of my father’s at the Queen’s University of Belfast, and this book represents good political analysis combined with very poor timing. It has two parts. The first half, more or less, is a survey of the political history of Belfast, paying special attention to the city council (known as the Corporation for most of the period), from the earliest days to the 1960s, when the book was written. I got a lot out of this (and plundered it extensively for my lecture last month).

Until 1832, Belfast was a pocket borough of the Chichester family, but the Great Reform Act opened up its politics to the mainly Presbyterian merchant classes. The first successful political organiser was a John Bates, who managed to combine the roles of main organiser for the Conservative Party (which won all the elections) with that of Town Clerk once the municipal council was reformed in the 1840s. He fell spectacularly from power in 1855 when he was exposed for diverting public funds by a public inquiry. I’d love to see some more about his story.

The book goes in detail through the next 110 years of political history, including a couple more times when the Corporation was suspended and the city was run by administrators. And the second half of the book gives the outputs of an exhaustive political survey of Belfast, including most of the councillors, and many of their supporters and voters in general, along with some comparative research on the attitudes of councillors in Glasgow. The data set is very rich.

The problem is that the research was largely carried out in 1966, and the city collapsed into chaos over the next couple of years, so that when the book first came out in 1973, it was a deep analysis of a political system that had already ceased to exist. The Belfast of 1973 was very different from the Belfast of 1966. The authors do look in depth into the questions of naming the new bridge and the Sunday swings issue, but compared with what happened over the next few years it all looks rather silly. (In fairness, a lot of people thought the swings issue looked rather silly in 1966.)

Really a book only for the most dedicated of Norn Iron politics nerds (and I am proud to count myself among that number). You can get it here.

The second paragraph of the third chapter of Belfast: The Story of a City and its People is:

Some years ago a friend of the family who stayed with us for a few days proceeded to tell me all about the cranes as soon as they arrived and saw the painting. ‘Hey, nice painting!’ they exclaimed, breezing into the living room. ‘That’s David and Goliath in Belfast, you know.’ ‘No, it’s actually Samson and Goliath,’ I responded – politely but firmly. ‘No, I’m sure it’s David and Goliath,’ they ploughed on. ‘You should check it out.’ I walked out of the room, my face burning with indignation, muttering through clenched teeth not entirely sotto voce: ‘Well I lived under them for nearly two decades so I think I should know what they’re called!’ My partner, her laugh stifled by the fear of a meltdown at the beginning of a social visit, rapidly changed the subject to a less divisive one as I harrumphed upstairs. ‘So let’s talk about Brexit then…’ she said.

This on the other hand is a much more accessible book, rooted in Cochrane’s personal story of having grown up as a Catholic in a mixed but traditionally Protestant area of the city (as I did), reflecting on the early history of the city, where he is keen on the radical political tradition of the McCrackens, the Assembly Rooms (now dilapidated) and the Linen Hall Library (of which I was a Governor back in the mid-1990s), and also looking at culture – music, theatre, poetry, and other parts of the arts. I found the first part more engaging, the second feeling a bit too structured, but the information is all good, and I would recommend it to anyone who wants to know more about how it feels to be in or from Belfast. You can get it here.

If I can be excused a second video, this is the percussion section of the City of Belfast Youth Orchestra performing Scheherazade in 1985. I am the third percussionist in view, holding the tambourine. The CBYO is still going strong.

Male strippers, and the boundaries between South and West Belfast

I got an interesting call from the Belfast Telegraph a couple of weeks ago. Northern Ireland’s major news last month was that a male stripper group had entertained customers at a Belfast pub by, you’ll never guess this, taking all their clothes off. Many people who felt that their opinions needed to be known took to the airwaves and the newspaper columns to express their views.

Personally I don’t have a problem with sex work, provided that basic lines of consent are protected for both providers and potential customers; it was completely decriminalised here in Belgium during the pandemic, and the country has failed to collapse into moral turpitude. (Or at least, I haven’t noticed if it did.) But the Belfast Telegraph did not seek my advice on that point.

Instead the question was about the location of the incident, the Devenish pub on Finaghy Road North: is it in West Belfast or South Belfast? Denizens of both South and West respectively insisted that the scenes of such depravity were not happening in their part of the city but on the other side of an invisible boundary. As I said to the reporter, “I can see how both South and West Belfast have rather different branding, and also the incident at the Devenish may not fit either branding particularly well.”

To go into the history of it. I grew up around the corner from the Devenish, but I don’t remember it being there when I was a child, and the Ordnance Survey map from around the time I was born marks the site as a “Nursery” – probably for trees rather than children. On the PRONI site you can track the history of the area back before the M1 motorway and even before the raileay.

Today’s Ordnance Survey map of the area north of Finaghy crossroads
Probably from the 1970s, before the Devenish was built; the site is marked as a nursery
Before the motorway was built, and before most of the development north of the railway – the Ardmore estate was built in 1947
Before the railway was built; though the line of today’s Ardmore Avenue is already visible as the lane around Finaghy Cottage

Finaghy Cottage, the house to which the future Ardmore Avenue led, belonged for many years to the confused poet Herbert George Pim, whose bizarre career I cannot possibly do justice to in the space I have here; let’s just say that it’s strangely appropriate that a scandal involving male strippers should break out less than five minutes’ walk from his former home. Edited to add: Disappointingly it seems that Pim’s “Finaghy Cottage” was on the Drumbeg Road near Dunmurry, not all that close to Finaghy in fact.

Anyway, the question is, what part of Belfast was the future site of the Devenish located in? The first part of the answer is that it wasn’t in Belfast at all until quite late in the day.

Map from Belfast: Approach to Crisis: A Study of Belfast Politics 1613–1970, by Ian Budge and Cornelius O’Leary (1973)

In this map from a history of Belfast civic politics, published in 1973, the future site of the Devenish is under the G in “Great Northern [Railway]” on the left, within the shaded area that Belfast Corporation were trying to annex from County Antrim after the second world war. But the city boundary actually ended farther east, at the King’s Hall to be precise; the showgrounds were just inside the city limits, and Finaghy outside. This was the boundary between the Ballyfinaghy and Malone Upper townlands.

Map from the 1917 Boundary Commission for Ireland’s report.

In parliamentary terms, the nine Belfast constituencies of the 1919 election were drawn by a Boundary Commission for Ireland in 1917. In 1920, using those same boundaries, they were merged to make four new parliamentary seats, returning to the old compass model, North, South, East and West. These were also the seats used for the first two elections to the Northern Ireland House of Commons. The boundary between South and West Belfast was the same as the boundary between the St Anne’s and Cromac seats of 1919, and the western half of the boundary was the railway line. And you can see that Finaghy, at the bottom left corner, is outside the city for parliamentary purposes.

The Belfast South and Belfast West constituencies remained unchanged until the early 1970s, when they were expanded outwards, Belfast South taking in the Rural District of Lisburn electoral divisions of Ardmore, Dunmurry, Finaghy, and Upper Malone, and Belfast West taking in the Rural District of Lisburn electoral divisions of Andersonstown, Ballygammon, and Ladybrook. (These Lisburn areas collectively had formed the short-lived Stormont seat of Larkfield.) We are interested in the Ardmore elecrtoral division, which was defined in 1963 as “That portion of the Townland of ‘Ballyfinaghy lying north of the centre line of the main Belfast/Lisburn Road”.

This map from the townlands database shows the townland boundaries of Ballyfinaghy, and the part north of the Upper Lisburn Road is the Ardmore electoral divison of the late 1960s. Immediately to the north again are the townland end electoral division of Ballygammon, in West Belfast from the early 1970s; but Ardmore (and indeed the whole of the Ballyfinaghy townland) is in South Belfast. So I was wrong when I told the Belfast Telegraph that the railway line had once been the boundary at Finaghy; the site of the Devenish has been in South Belfast since the early 1970s, and before that it was not in Belfast at all. It has never been in West Belfast, contra what I told the Belfast Telegraph. Mea culpa, mea maxima culpa.

Since 1983, the constituency boundaries have been based on the reformed local government wards, which defined the motorway as the boundary between Finaghy Ward and Ladybrook Ward in 1972 and since. I was correct on that at least, and it has survived several rounds of revision.

But basically, the disgruntled citizens of South Belfast will have to accept that the Devenish is part of the diversity of their quarter of the city, which is anyway the most multicultural area of Northern Ireland. For what that’s worth.

For previous cartographic nostalgia, see my posts on Moreland’s Meadow and the oldest shop at Finaghy Crossroads.

The Fire Starters, by Jan Carson

Second paragraph of third chapter:

Commercial aeroplanes leaving and landing at the City Airport continue to pass overhead. They’re unaware of Sammy and the shape he’s spelling out as he walks. He’s too small to be seen from the sky. He’s a grain of sand, a dot, a pin, a misplaced punctuation mark. Even God would have to squint. However, if he could be seen from such a height, if, for example, you were peering through binoculars or some other magnifying lens, your eye would be drawn to him, dragging his heels from one street to the next, kicking an empty Coke bottle as he goes. You would know that Sammy did not belong on these streets, drifting.

East Belfast, marching season, the present day (2019); two fathers concerned about their children. Ex-Loyalist Sammy suspects that his son is the masked social media influencer behind a wave of arson attacks. Trouble GP Jonathan’s daughter was begotten of a Siren who came and stayed in his bath and then disappeared back into the waves.

Most of the novel is gritty reality, so that you can almost smell the tarmac bubbling in the summer sunlight; but the parts with Jonathan and his daughter edge into magical realism with a particular Belfast idiom, where parents of strangely gifted children navigate both intrusive supernatural forces and the banal bureaucracy of health care and social security.

Often this sort of trope can feel bolted onto a conventional narrative, but Carson makes you feel that Belfast (East Belfast, very specifically) is the sort of traumatised place where reality starts to erode at the edges. It’s well-balanced, in the sense that a cyclist going at top speed over uneven terrain remains well balanced. Anyone expecting a standard urban grim novel will be surprised. You can get it here.

Bechdel fail. The book is either first-person from Jonathans point of view, or tight-third from Sammy’s, and they talk to very few women who in turn don’t talk much to each other, and if they do it’s about Jonathan or Sammy.

The new Northern Ireland constituencies

Since 2007 I’ve been the Northern Ireland arm of the analysis of UK parliamentary constituency changes by Michael Thrasher and Colin Rallings. The latest version, taking into account the new boundaries that will take effect from the next election (be it Westminster or Assembly) was published a few weeks ago; it’s been a busy period for me, but I have now taken the time to write up the changes to each of the 18 Northern Ireland seats.

Media coverage coverage of the changes focussed on the effects in England, Scotland and Wales, and frankly that was the right call; the changes in Northern Ireland are the least dramatic since the 1970s. The 1983 review added five new constituencies, taking the total from 12 to 17; the 1996 review added another, making a total of 18; and the 2007 review expanded the Belfast seats outwards with knock-on effects all around the map.

There were also two failed reviews, one in 2013 which fell victim to the internal politics of the Conservative-Lib Dem coalition, and one in 2018-20 which was quashed by the Belfast courts for failing to adequately consider public opinion at the final stage of revision (and the whole thing was then killed off by Boris Johnson).

So this is the first change to the Westminster constituencies for 17 years, the longest gap since the 1950-70 period. In Northern Ireland the Westminster boundaries are also used for Assembly elections, and indeed in 1973 and 1996, regional level elections used the new boundaries first. Personally I think that the Assembly constituencies should be linked to the Local Government Districts rather than the Westminster seats, but that’s for another day.

Every seat is changed this time, but few of the changes are drastic. In the list below, I’m going from the most changed to the least changes constituency, showing my working for each of them. Notional votes are a bit of a mug’s game, but I’m confident that these numbers correspond closely to whatever the reality might have been if the 2019 general election votes had been cast on the 2024 boundaries.

The maps are all screenshotted from the Guardian’s excellent site, which you should consult.

South Belfast and Mid Down (SBMD)

Going from most to least changed, the biggest effect is on the constituency where I grew up, South Belfast, now renamed to South Belfast and Mid Down. It loses 10% of its electorate to East Belfast, but gains a bit more than that from Strangford and Lagan Valley, and a few scrapings from West Belfast. (By the way, I have an idiosyncracy of calling the Belfast constituencies “X Belfast” rather than “Belfast X”. It seems to me that “South Belfast and Mid Down” sounds better than “Belfast South and Mid Down.)

Projecting the 2019 election onto the new boundaries, I see about 850 more Unionist votes, 550 more for Alliance and 1650 fewer for the SDLP, with another 250 Nationalist votes coming in from West Belfast. Claire Hanna won the seat with a majority of over 15,000 in 2019; this would be reduced by 2,000, but the SDLP still win more than half of the votes in the constituency.

70134 116781259678627079550
SBSBMD63029 97861055602825136550
SBEB-7096 -1890-204-758-1939
SBWB-9 -1-3-1
StrSBMD 6078 163437213698173207
LVSBMD 2161 611276313203422
WBSBMD 710 5620673222518
Total71978 1208717031677348736725409247568

Sinn Fein did not contest the 2019 Westminster election, but they could take almost half of the SDLP vote and Claire Hanna would still win the seat. She bantered with me on social media about looking forward to the challenge.

At Assembly level, Unionists combined were just short of 2 quotas in South Belfast in 2022, and these numbers would put them just about in position to regain the second seat lost in 2017.


Neighbouring Strangford loses 6,000 voters, mainly in Saintfield, to the new Belfast South and Mid Down, but gains 9,000 around Downpatrick from South Down.

This results in the biggest shift in party support in any constituency.

66990 1770540233081476106347901994555
StrStr60899 160683651308134096517171787555
StrSBMD-6078 -1634-372-136-981-73-207
StrND-10 -3-1-2
StrS-3 -1
SDStr 9171 602680022152462193

The DUP majority over Alliance here is reduced from 7,000 to 5,700, and if you squint you could just about see a unified non-Unionist candidate defeating a split opposition at a Westminster election; but it’s not very likely. From the Assembly point of view, the Nationalist vote increases by more than 10% and is now over a quota. At every Assembly election since 1998, the SDLP have been runners-up here; whichever of the Nationalist parties can get ahead of the other now has a good chance of gaining a seat here.

South Down

Staying in the neighbourhood, the calculations for South Down were much the most complex. It swaps bits of territory with three of its neighbours, most notably donating the voters around Downpatrick to Strangford, and also makes a gain from Upper Bann (my ancient homeland of Loughbrickland).

Although the shifts are geographically complex, the electoral impact is muted.

79295 76193307691614517161371266
UBSD 1960 57417415283221
N&ASD 1058 143555512326421
LVSD 105 271211822
StrSD 3 1
Total 71,772 816234601621212240138611070

SF won this seat with a 1300 majority in 2019, and I don’t see much change to that in my notional result. The overall Nationalist vote share decreases by 3.8% and the overall Unionist share increases by the same amount. This is still not enough to put Unionists in play for a second Assembly seat.

Lagan Valley

Rounding off the middle of County Down, Lagan Valley loses Drumbo to South Belfast and Mid Down, and more significantly Dunmurry to West Belfast, while gaining the eastern fringes of Lurgan from Upper Bann.

Again, it looks bigger on the map than it actually is.

75884 1958686063159551308717581098
LVLV 68948 180207910315878117931441900
LVWB-4330 -840-3700-41-898-272-170
LVSBMD-2161 -611-2760-31-320-34-22
LVSA-340 -88-390-4-59-8-5
LVSD-105 -27-120-1-18-2-2
UBLV 7364 1818550005714101090
SDLV 20 2100244
Total 76,332 1984084603158781236518551994

The DUP’s 6,500 majority over Alliance in 2019 increase to 7,500 (what you might call the Dunmurry effect), and the total non-Unionist vote upticks very slightly. At Assembly level, Nationalists were able to win a seat in Lagan Valley in a good year, and these changes make good years more likely, though Alliance would still have a good chance of holding their second seat.

West Belfast

Looking north of Lagan Valley, West Belfast loses a few nibbles around the edges but gains 9,000 voters from Dunmurry at one end and the Shankill at the other.

Neither of the newly added patches of territory is great for SF, but they are pretty far ahead anyway.

65761 5220188261942985208661635
WBWB62538 4084181360582919204081599
NBWB5044 2,863 348 208
LVWB4330 840 370 41 898 272 170
SBWB9 1 1 3 0

This is the second biggest shift of party support in any constituency, but I don’t think SF will be awfully troubled by the prospect of their 14000 majority over PBP at Westminster being reduced to a 12000 majority over the DUP; the seat is safe as houses anyway. Unionist candidates were runners-up here in every assembly election since 1998, with the exception of 2003 when Diane Dodds actually won. There is now a clear prospect of a safe(ish) Unionist seat at the next Assembly election.

Upper Bann

As noted already, Upper Bann loses Loughbrickland to South Down and its eastern fringes to Lagan Valley, but gains parts of the apple country of North Armagh. It was the most bloated constituency on the old boundaries.

The result looks big on the map but has little net electoral impact.


The DUP majority at Westminster drops from just over 8000 to just under 8000. The last Assembly seat here in 2022 was won by Alliance with a 376 vote margin over SF; that would look vulnerable under these changes.

East Belfast

This is very straightforward, with a loss to North Down in one direction and gains from South Belfast in the other.

This does bring in notional South Belfast SDLP votes (the SDLP did not stand in East belfast in 2019).

66273 20874251619055
EBEB62980 19726237818232
SBEB 7096 18902047581939
Total 70076 216162581189901939

The DUP’s majority in 2019 was 1800, and the changes expand that to 2600. But those 1939 notional SDLP votes could go a long way to making up the difference in one of the tightest results. I don’t see any direct impact on Assembly representation; there is still nowhere near a Nationalist quota.

North Belfast

Jumping across the river now, we have some tinkering around the margins of North Belfast; the biggest changes are the smallest on the map, to West Belfast on the Shankill and to and from South and East Antrim in Newtownabbey.

These changes basically don’t help the DUP to regain the seat lost in 2019.


SF’s 1900 majority in 2019 expands to 3700. And it’s difficult to see any change in Assembly representation either.

East Antrim

Continuing up the coast, we reach East Antrim which swaps large but sparsely populated territory with North Antrim.

It doesn’t make a lot of difference to the results though.

64907 1687154751043101656859022120
EAEA 62640 162825284100798106858702046

The DUP’s actual majority of 6700 over Alliance is reduced to a notional 6400, which won’t cause sleepless nights. There are clearly two non-Unionist quotas for the Assembly, and equally clearly Nationalists will struggle to get one of them.

Newry and Armagh

Back to the south of Northern Ireland again, where Newry and Armagh, the second most bloated seat under the old boundaries, loses most of the apple country to Upper Bann and Fermanagh-South Tyrone, and tidies up its eastern boundary.

The territory lost is at the more Unionist end of the constituency.

FromTo 81329 21.7%8.3%8.3%18.6%40.0%3.2%
N&AN&A 73,127 9275354537868753187921508
SDN&A 1458 1406112726729723
Total 74585 9415360539139019190881531

SF’s 2019 majority increases from almost 9300 to over 9600. For the Assembly, a second Unionist seat slips a little further away.

North Antrim

Back up to the north again as North Antrim swaps territory with East Antrim and comes out a bit smaller.

Changes that look big on the map don’t always have much effect on the ground.

77156 208608139623124629435632
NANA70120 203786797564824623904575
NAEA7036 -482-1342-583-553-1057
Total71165 20650688517581124624044609

I see the DUP’s notional majority here increasing slightly from 12,000 to 13,000. From three quotas they should notionally get three Assembly seats, but Jim Allister was not a candidate in 2019.

Fermanagh and South Tyrone

Out West now, where the maths is fairly simple but the politics complicated. Fermanagh and South Tyrone was pretty close to the required size, but because of other changes must gain some net territory from Newry and Armagh, and lose a bit to Mid Ulster.

On paper, the differences are not huge.


But this was the tightest result in Northern Ireland in 2019, SF beating the UUP by a mere 57 votes. The notional majority is now 510; but there are 896 notional DUP votes in the mix. So Northern Ireland’s closest race may actually have got a bit closer.

Mid Ulster

This was one boundary change that I called completely incorrectly. I had expected that Mid Ulster would stretch north towards Dungiven, as had been the case in the previous quashed proposals. But in fact it takes a chunk of Coalisland from Femanagh and South Tyrone, which losing a large but sparsely populated chunk to West Tyrone.

It doesn’t make a lot of difference.

70501 1093626113526690638420473
MUMU67036 1056625233284690603119342
MUWT-2588 -234-56-198-273877
MUFST-877 -136-32-44-79255

SF’s 9,500 majority extends to a notional 9,800, and the needle is not really moved for the Assembly seats.


A little trimming at the edges to East Londonderry and West Tyrone.

The changes are minor, and although more Nationalist than Unionist voters are moved, the Unionist vote share is hit worse.

69890 3,852 878 1,189 1251 25785 9372 1,949

The SDLP’s 17,100 majority in 2019 is reduced to a mere 16,400 which I don’t think will trouble them unduly. The DUP won the last Assembly seat here by a margin of 95 votes over the UUP, the closest result of the 2022 election, and on the above swing it would be vulnerable, but the picture is very much blurred by tactical voting.

West Tyrone

Simply takes in adjacent chunks from Mid Ulster and Foyle.

These are small territories, sparsely populated, and don’t make a lot of difference.

66339 9066277439795210733016544972
WTWT66339 9066277439795210733016544972
MUWT2588 23456198002738770
FoyWT1687 230532903050518438
Total70614 953028824206521308108176041010

The SF notional majority increases from almost 7500 to almost 8100. No impact on the Assembly result.

South Antrim

Getting near the end now, with South Antrim by far the most annoying to calculate: lots of little changes that don’t add up to anything much, the biggest being chunks of Glengormley going to North Belfast.

A real pain to work out these very small notionals!

71915 1514912460819022884887
SASA69729 1493712283759720884603
SANB1926 -157-132-563-192-267
SAEA260 -55-45-30-8-18

In a good year, the UUP could overtake the DUP here, and in a much much better year the same is true for Alliance. But the boundary changes have little impact on the Westminster or Assembly outcomes.

North Down

Takes in a small sliver of Strangford and a larger sliver of East Belfast.

The Alliance Party’s strongest seat takes in 3000 voters from the Alliance Party’s second strongest seat.

67,109 37.9%12.1%4.8%45.2%
NDND 67,109 153904936195918358
EBND 3,293 11481380823
StrND 10 3102
Total 70,412 165415075195919182

Alliance’s Westminster majority is reduced from almost 3000 to just over 2600, so the seat remains competitive but they are starting ahead. The shifts are so small that it’s difficult to see much impact on the Assembly.

East Londonderry

Gains a ward from Foyle.

Small numbers make for small differences.


The DUP’s Westminster majority is unchanged. It’s difficult to see much impact on the Assembly election, especially if independent MLA Claire Sugden remains active.

So there you have it. A Westminster election is likely before the end of the year. It may well see some changes of seats, but the new boundaries are unlikely to make the difference.

Seamus Heaney, Free Derry and the Grianan of Aileach

F expressed the desire to see a bit more of Ireland than he has previously managed, so at the weekend we went on an expedition to the north west, starting with a loop round the southern end of Lough Neagh to go up to the Seamus Heaney HomePlace at Bellaghy.

(I remember once talking to someone from continental Europe about the geography of Northern Ireland. She said, “And there’s that big lake right in the middle! I’m sure it is really beautiful!” I replied, “Er, no, not really…”)

The Seamus Heaney HomePlace is a two-floor building, largely linking Heaney’s poetry to the countryside where he grew up, and to his friends and family. I must say it helped me to appreciate the well of inspiration that he drew on. A small video display allows you to select celebrities reading his poetry out loud, including Bill Clinton, Mary Robinson and King Charles III.

Upstairs there are more direct memorabilia, including a lovely video montage of the furore around his winning the Nobel Prize in 1995, when as you may remember he was on holiday in Greece and his family were unable to contact him with the news. F had barely heard of Heaney before going, and I think I would not recommend it to anyone who doesn’t already know his work, but as a decades-long fan I found it interesting and even a little inspiring.

On then to the Maiden City, where we went to the Museum of Free Derry. F was actually much more impressed by it than he looks in this photo.

I was impressed too. It’s a very well put together narrative of the decades of neglect and misgovernment that led to the Battle of the Bogside and ultimately to Bloody Sunday. And the building itself is right at the core of events – this is the map from the Guardian that I marked up to show the locations of victims of the fatal shootings, with the museum added. The flats across Rossville Street have long since been demolished, but a lot of the rest of the buildings are still there. It’s a surprisingly small space for the drama of the day.

I have written before about the case of Soldier F, who is to be prosecuted for a number of the casualties in Glenfada Park North (ie on the doorstep of the museum). I had missed the welcome news that the Public Prosecution Service’s decision not to pursue the case after all was overturned by the Court of Appeal a year ago, and the case is continuing. It still bothers me that he is not being prosecuted for the crimes that the Savile enquiry found he had certainly committed (the murders of Michael Kelly, Bernard McGuigan and Patrick Doherty, and the attempted murders of Patrick Campbell and Daniel McGowan) but for others for which Savile found only weak evidence that Soldier F was the shooter (the murder of William McKinney, and the attempted murders of Joseph Friel and Joe Mahon) or indeed where Savile thinks that other soldiers fired the shots (the murder of Jim Wray and attempted murders of Michael Quinn and Patrick O’Donnell). Even half a century later, it would be nice to see some justice done here.

We explored the city and paid the obligatory homage to the most recent cultural icons.

Finally, we went across the border to the spectacular Grianan of Aileach, an Iron Age fort (reconstructed in the nineteenth century) overlooking the Inishowen peninsula, with incredible views over Lough Swilly and Lough Foyle. It was very windy, but very much worth seeing.

Neglected megaliths of Loughbrickland

All of sixteen years ago, I wrote a blog post about visiting three megalithic sites near Loughbrickland: a standing stone (menhir) at Lisnabrague on the Poyntzpass road, the so-called Three Sisters of Greenan on a hill near the lake, and another standing stone beside the northern shore of the lake, in Drumnahare townland.

They’re all laid out on this map, though the Three Sisters are mysteriously placed a hundred metres to the east of their actual location.

I returned to visit all three this week, and to be honest I was a bit dismayed. Going west to east, the opposite order to last time, I found that the field containing the Lisnabrague stone is currently planted with maize which is taller than me. The farmer gave me permission to go look for it, commenting that I was the first person he had ever encountered who showed any interest; he added resentfuly that he is not allowed to build within five hundred metres of it, which does seem a bit excessive.

Using GPS I was able to navigate to the stone through the maize, and found that it sits in a sort of glade among the triffid-like crops.

But it feels isolated and neglected, compared to when I visited in 2007.

At least it was accessible. The Three Sisters lie in a hedge beside a lane; the hedge has been allowed to grow thick over them in the last sixteen years, and you can no longer see them from the lane at all. The field in which they lie has been completely fenced off; you can photograph the two upright Sisters through or over the fence, but you cannot reach or even see the third of the three stones, which is completely submerged in the hedge.

A neighbour told me that the owner had had a lot of hassle with treasure-hunters – not metal-detectorists, but people doing organised guided quests, who had failed to observe the usual etiquette of the countryside. It’s a shame. In 2007 you could go right up to them, and see the recumbent Sister as well.

The standing stone by the lake remains easy enough to visit, but the Orange Order who own the field have put up a massive flagpole right beside it, which really impacts your experience of the site. (There’s also a flag flying on the crannóg in the middle of the lake, but I carefully positioned the flagpole to block it out.)

Sixteen years ago I was able to get a lovely shot of the crannóg framed by the cut in the top of the stone, which has mysterious cup-like markings.

I came away feeling that the relationship between the state and the landowner in respect of ancient monuments seems to be deteriorating. It would be nice to see a new partnership established based on dialogue and mutual respect of each other’s interests. But that would probably require a restoration of devolved government.

A historic maximum: ex-prime ministers and iar-taoisigh

There are more former British prime minsters alive today than at any time since the office was created in 1721.

Liz Truss, Boris Johnson, Theresa May, David Camero, Gordon Brown, Tony Blair and John Major make a total of seven living PMs since Truss’s resignation on 25 October last year. And from the looks of things, that number is likely to increase before it decreases – Rishi Sunak’s government looks to be in worse health than any of his predecessors.

On two previous occasions there have been six living ex-Prime Ministers.

Between the end of Sir Robert Peel’s first term, on 8 April 1835, and the death of Henry Addington on 15 February 1844, there were six living ex-prime ministers: Addington (whose time at the top was decades previously, 1801-1804), Viscount Goderich, the Duke of Wellington, Earl Grey (he of the tea), Lord Melbourne and Robert Peel. Though in fact Melbourne had a second term from 1835 to 1841, and Peel then came back until 1846, so there is an argument that there were only five living men who were former and not current prime minsters for that period.

Similarly, from the end of Ramsay MacDonald’s first term, on 4 November 1924, until the death of H.H. Asquith on 15 February 1928, there were also six living former prime ministers: the rather obscure Lord Rosebery (briefly PM in 1894-95), Arthur Balfour, Asquith, David Lloyd-George, Stanley Baldwin and MacDonald. Again, however, Baldwin was back in for a second term.

The most recent period when there was only one living ex-prime minister was between the death of Baldwin on 14 December 1947 and the end of Attlee’s term on 26 October 1951. The only living ex-PM then was the leader of the opposition, Winston Churchill.

When the first prime minister, Robert Walpole, died on 18 March 1745, there were no living ex-PMs. His successor, the Earl of Wilmington, died in office, as did the next in line, Henry Pelham, and there was no living ex-PM until the end of the first term of Pelham’s brother, the Duke of Newcastle, on 11 November 1756.

The number of ex-Taoisigh (?iar-taoisigh?) is also at an all-time high, at six (Bruton, Ahern, Cowen, Kenny, Varadkar, Martin) though again we have to enter the caveat that Varadkar is currently enjoying his second run.

That level has been hit twice before. Between the end of John Bruton’s term in 1997 and the death of Jack Lynch in 1999, Lynch, Liam Cosgrave, Charles Haughey, Garret Fitzgerald, Albert Reynolds and Bruton himself were all living, and as Bruton’s successor Bertie Ahern had not previously been Taoiseach, there are no ifs nor buts.

And more recently, for the two months in 2011 between the end of Brian Cowen’s term and the death of Garret Fitzgerald, the living ex-Taoisigh included also Liam Cosgrave, Albert Reynolds, John Bruton and Bertie Ahern; and again Enda Kenny was a first-time Taoiseach during that period.

There has been no period when there were no living ex-Taoisigh, thanks in part to the longevity of Eamon de Valera. After the death of John A. Costello in 1976, Jack Lynch was the only living ex-Taoiseach until Liam Cosgrave lost the 1977 election (and Lynch came back to power).

The number of living former heads of the devolved administration in Northern Ireland is also current at an all-time high, at five (Mark Durkan, Peter Robinson, Arlene Foster, Paul Givan and Michelle O’Neill – counting First Minister and Deputy First Minister equally, but not counting those who served only in an acting capacity such as Reg Empey and John O’Dowd).

In the olden days there was no living ex-Prime Minster of Northern Ireland until John Miller Andrews was kicked out in 1943, Lord Craigavon having died in office, and then again from Andrews’ death in 1956 until Brookeborough retired in 1963. James Chichester-Clark, briefly PM ion the dying days of Stormont, lived to 2002, by which time devolution had been more or less restored.

Aren’t you glad I worked that out for you?

Northern Ireland local elections 2023

So, the headline is that Nationalist parties (those who would designate as Nationalists in the Assembly) outpolled Unionist parties (those who would designate as Nationalists in the Assembly) in Thursday’s local elections by 19,000 votes, and more than two percentage points. This is a first for Northern Ireland.

Nationalists (SF + SDLP + Aontu + IRSP): 300,565 (40.8%, +4.5%)
Unionists (DUP + UUP + TUV + PUP + Cons): 281,196 (38.2%, -3.7%)

My tweet about this last night got a lot of pickup, including getting me quoted in the Guardian. Some people pushed back at me saying that I should have counted People Before Profit as Nationalists, though they don’t designate as such; or that I should have counted Alliance as Unionists, though they too don’t designate as such; or that I should have counted independents, though they are not political parties by definition; or that I shouldn’t have done the calculation at all. The point remains: Unionist parties were outpolled by Nationalist parties for the first time ever.

This is important psychologically but not operationally. The criterion for triggering a referendum on a United Ireland is pretty much that the UK thinks it is likely to go that way. That outcome is not apparent from the above numbers, which show only 40.8% of voters supporting the election of candidates from Nationalist parties to local councils with limited powers. 40.8% is a lot – it’s more than 38.2% – but it’s not 50%, and the Nationalist vote share would need to be higher or have a larger lead to justify calling a Border Poll.

In the case of Catalonia, which I am familiar with, where pro-independence forces were in the zone of getting a majority of the electorate, the picture was complicated by a significant clump of voters who wanted a referendum on independence, for the sake of clarity and dignity, but also wanted to stay part of Spain. There is no such pro-referendum caucus within the 20% swing voters of the centre in Northern Ireland. Nationalists (in both Northern Ireland and Scotland) might start usefully working out how such a caucus could be persuaded into existence.

And, as I’ve said before, winning such a referendum is a different matter again. It requires three things: Brexit continues to be an obvious negative (✔), Unionists continue to talk only to their own core voters and ignore the persuadable middle (✔) and Nationalists come up with a credible counter-offer, including robust proposals on health care (✘). Nationalists have time to work on the third of these; Unionists are running out of time to work on the first two.

Looking at the details:

Antrim & Newtownabbey13-7–08+2+ Ind 1—9++++40
Ards & North Down1480-12++3 Ind, 2- Green1040
Armagh, Banbridge & Craigavon13++6—-1+4+1 Ind1—–15+++++41
Belfast City14-21+11+3- Green, 1– PBP, 1+ Ind5-22++++60
Causeway Coast & Glens13-4—2++5+++1+ PUP3—12+++40
Derry & Strabane5–3+00–3- Ind, 1- PBP10-18+++++++40
Fermanagh & Omagh6+7–02+1— Ind3–21++++++40
Lisburn & Castlereagh14-6—–013++++1+ Ind24++40
Mid & East Antrim14-8+572- Ind0-4++40
Mid-Ulster11++2—-003+ Ind5-19++40
Newry, Mourne & Down5++1—05+++2— Ind8—20++++41
Not shown in above table:
2 PUP losses in Belfast
1 Ind loss in Causeway Coast and Glens
1 Aontu loss in Derry and Strabane
1 Lab loss in Fermanagh and Omagh
1 Green loss in Lisburn and Castlereagh

It will be apparent that while the majority of the SDLP’s losses were directly to Sinn Fein, only about half of the Sinn Fein gains came from the SDLP. The rest came from smaller groups/independents and Unionists. The campaign successfully persuaded many voters who don’t normally vote SF, or vote at all, to show solidarity with the concept that the leader of the party with the most votes should become First Minister. It is a stunning success, the best vote share ever for Sinn Fein in a Northern Ireland election. Alliance’s gains also demonstrated support for getting the institutions back up and running.

On the other side of the argument, the TUV failed to break through in any significant numbers – though they are still there – and the DUP were fortunate to avoid a net loss of seats despite slipping a full percentage point on vote share. It’s clear that their message has not resonated beyond the core vote, which is tactically a successful defence but strategically questionable. Cards on the table: I don’t see how blocking the institutions can be a successful strategy. It’s clear that London doesn’t care very much, so the blockade imposes no pressure on Westminster, while damaging the interests of the people who Unionism claims to represent. Worse, it undermines the legitimacy of Northern Ireland’s continued existence as an entity. (See above.)

The crunch on smaller parties is severe, and I don’t see an easy way out of it. It’s the worst election result ever for the SDLP, and the second worst for the UUP. Neither has a clear unique selling point relevant to the current situation. I heard one SDLP speaker complaining that the electorate have forgotten who got the Good Friday Agreement 25 years ago. In the real world, nobody fights this year’s elections on 1998’s outcomes. A UUP speaker complained that Nationalists were running too many candidates and should let other parties have a chance. That’s not how elections work.

On these numbers, the SDLP Westminster seats in Foyle and South Belfast look vulnerable, though I’m inclined to think that the incumbent will hang on in South Belfast. On the other hand, Alliance look more secure in North Down and better placed in East Belfast. Come an Assembly election, SF would be well in the lead, and Alliance in third place but some way behind the DUP.

So how was your weekend?

Lyra McKee, and why I split up with the News Letter

I had the great pleasure last night of watching Alison Millar’s documentary, Lyra, about the life and death of journalist Lyra McKee. It’s a tremendous portrait of a committed young woman, killed in the middle of doing her job just before Easter in 2019. The showing was presented by the UK mission to the EU and the Northern Ireland representation in Brussels, as part of the Brussels Irish Film Festival, and Alison Millar was on hand to answer questions both formally and informally. The film is beautiful and I strongly recommend it. Here’s a trailer:

I did not know Lyra McKee myself – she was six years old when I left Northern Ireland – but inevitably we had a lot of mutual friends (38 according to Facebook, I’m sure a lot more in reality), all of whom seem to remember her fondly. She first hit my radar screen in 2013, when she began her research into the murder of Robert Bradford, nineteen years before she was born. This particularly fascinated me because he was our local MP, and he and the caretaker for his office were killed just ten minutes’ walk from our home. Her book was eventually published, available here, an extract here.

In July 2019, three months after Lyra’s death, the News Letter, one of the main Belfast news outlets, ran a front page story revealing that the royalties from the book were going to a non-profit organisation, one of whose directors was a former paramilitary. The article evoked a furious response from Lyra McKee’s publisher and family. I too felt that this was a crappy piece of journalism. A former paramilitary being associated with a non-profit organisation is hardly news and not really interesting, and it was barely relevant to Lyra McKee’s work. The News Letter subsequently successfully defended a libel case, with the defence that the article was true (or at least, that the points complained of in the article were true). But what is true is not always right, even without considering the innuendo in the piece.

Over the previous few years I had written a few pages of political analysis for the News Letter in advance of each election in Northern Ireland, often featuring on the front page of the newspaper’s election specials. But I felt very uncomfortable about what they were now putting on the front page. I wrote to the then editor, saying that in my view the article was “sensationalist and did not serve the public interest. I am very disappointed. I thought you were better than that.” Consequently, I permanently severed my relationship with the newspaper. I am all in favour of being part of a broad spectrum of voices, but only if I can feel confident in the ethical values underlying the editorial choices being made.

I haven’t discussed any of this in public previously because fundamentally it’s not really about me. But I had the chance to tell the brief story last night to Alison Millar, and now that I’ve ticked that box I may as well go public here.

Observe the Sons of Ulster Marching Towards the Somme, by Frank McGuinness

Opening of Part Three:

The sound of water. Light up on Boa Island. Craig rests, smoking. Pyper enters.
Craig: Well?
Pyper: Good. Good place.
Craig: I hoped you’d like it.
Pyper: You rowed out here every day?
Craig: When I had the chance and I wanted to be on my island.
Pyper: Your island?
Craig: Sorry. Boa Island. I stand corrected. I meant when I wanted to be on my own.
Pyper: Nobody ever comes here?
Craig: Very few.
Pyper: Strange.
Craig: This place? Yes.
Pyper: The place is definitely strange, but strange too, people shouldn’t come.
Craig: Why should they come here?
Pyper: The carvings.
Craig: What are they?
Pyper: Signs.

This play won the Christopher Ewart-Biggs Prize in 1986, and I was lucky enough to see it thirty years later, at the Abbey Theatre for the 2016 production commemorating the centenary of the Battle of the Somme. Reading the script now can’t really do justice to the memory of the theatre production, which starred Donal Gallery as Pyper, and crucially used the space of the stage to make the story come alive.

It’s a reflection on eight soldiers recruited to the Ulster Division during the First World War, exploring their understanding of the universe, life, love and loyalty. The narrative is bookended by Pyper in old age reflecting on how he survived and his friends did not (so the fact that seven of the eight die is signalled early on).

I find the third act the most effective, the eight characters back home on leave and split into four pairs, two on Boa island, two at a church, two at the Carrick-a-Rede rope bridge, and two at the Field where Orange marches finish (which historically was at Finaghy, close to where I grew up, though I do not know if that was the case in 1915 or 1916). It gives the men a chance to explain themselves to each other, a sympathetic but informed audience.

By the lakeside in Fermanagh, Pyper and Craig make love, which must have been rather shocking in 1985 and was still a bit unexpected in 2016. (Also the weather must have been very good that day.) All of the characters reflect on the place of Ulster in Ireland, in Britainm in Europe and in the empire. There are some very good lines:

Old Pyper: Those I belonged to, those I have not forgotten, the irreplaceable ones, they kept their nerve, and they died. I survived. No, survival was not my lot. Darkness, for eternity, is not survival.

McIlwaine: The whole of Ulster will be lost. We’re not making a sacrifice. Jesus, you’ve seen this war. We are the sacrifice.

Younger Pyper: I have seen horror
Elder Pyper: Ulster
Younger Pyper: They kept their nerve and they died.
Elder Pyper: Ulster
Younger Pyper: There would be and there will be, no surrender.
Elder Pyper: Ulster
Younger Pyper: The house has grown cold, the province has grown lonely.
Elder Pyper: Ulster
Younger Pyper: You’ll always guard Ulster.
Elder Pyper: Ulster.
Younger Pyper: Save it
Elder Pyper: Ulster
Younger Pyper: The temple of the Lord is ransacked.
Elder Pyper: Ulster.
(Pyper reaches toward himself)
Younger Pyper: Dance in this deserted temple of the Lord.
Elder Pyper: Dance

You can get it here.

This was the non-sf fiction book that had lingered longest on my unread shelves. Next in that pile is The Various Lives of Keats and Chapman, by Flann O’Brien, but it will have to wait until I have finished my 2016 books.

Ewart-Biggs Prize winners: Observe the Sons of Ulster Marching Towards the Somme, by Frank McGuinness | From A Clear Blue Sky: Surviving the Mountbatten Bomb, by Timothy Knatchbull | Setting the Truth Free: The Inside Story of the Bloody Sunday Campaign, by Julieann Campbell | The Republic: The Fight for Irish Independence, 1918-1923, by Charles Townshend | The Whole and Rain-Domed Universe, by Colette Bryce | The Sun is Open, by Gail McConnell

The Halls of Narrow Water: A family history, by Bill Hall

Second paragraph of third section of main narrative:

On arrival in Ireland, William Hall is believed to have been involved in mining at Red Bay near Carrickfergus  in Co. Antrim and to have died there in 1640.  There were other Halls in Antrim at that time.  However, they were no connection to William Hall and the subsequent Halls of Narrow Water.  William’s son Francis was born in 1620 and married Mary Lyndon daughter of Judge Lyndon of Galway.  We do not have any historical background on the Lyndon family of Galway, but given his status as a judge he would have been from a family of some influence.  There is a Francis Hall recorded as holding land in 1663 in the Barony of Glenarm, which is where Red Bay is located. Francis subsequently moved to Glassdrumman in County Armagh before buying the townland of Narrow Water and eight other townlands in 1680. Francis and Mary had four children, Roger, Edward, Alexander, Trevor and Frideswid.  The marriages of these children saw the beginning of a series of marriage alliances between the Halls and several influential and powerful families in Ireland.  Roger married Christine Poyntz, daughter of Sir Toby Poyntz; Edward married Anne Rowley and moved to Strangford, establishing another branch of the Hall family who were also to marry into a number of prominent families.  Frideswid married Colonel Chichester Fortescue of Drumiskin, the eldest son of Lieutenant-Colonel Sir Thomas Fortescue Governor of Carrickfergus Castle, continuing the link with the Chichester Family.

I have mentioned previously that one of my father’s best friends as a child was Roger Hall, of the Halls of Narrow Water Castle near Warrenpoint on the southern shore of County Down; and my aunt Ursula and Roger’s sister Moira Hall shared a house in London for many years. I renewed contact with the Halls last summer, for the first time in decades; here’s F with Roger’s son M, who now runs the castle and the estate, in the snooker room which carries many memories.

Growing up, I didn’t especially know Roger and Moira’s younger brother Bill (formally Sir William Hall), but we had a great lunch together with various other relatives last August, and I was subsequently sent his book about the Hall family, which is available by private circulation only.

It’s a breezy 250-page compilation of archival material and personal reflection. The Troubles and the wider political situation are inevitably part of the book. One of the worst atrocities of the whole period took place literally at the castle gate. But the focus is on the Hall family and on their role within the community, and I must admit that my personal interest was in the anecdotes about my own family in the book.

To be honest, for most of the the three and a half centuries that the Halls have been based in Narrow Water, they kept their heads down and were unremarkable County Down landlords. The picture becomes more interesting with Frank Hall (Bill, Moira and Roger’s great uncle), a UVF gunrunner and spy. To Frank’s disgust, his nephew married a Gibraltarian and their children were brought up as Catholics, the ultimate betrayal for a fervent Loyalist.

Bill, Moira and Roger’s father died when the boys were still quite young, which led to complications in the administration of the Narrow Water estate. The legal convolutions to prevent it falling into the hands of the Catholic church are apparently a case study in such things. Undaunted, all of the children of that generation (there were three more sisters, but I only knew Moira) had adventurous lives. I’m very glad that Bill took the time to compile it all into digestible form.

The Sun is Open (and Type Face), by Gail McConnell

Third page:

our house was on a street that 
slanted at the bottom a 
carriageway you didn't cross 
four lanes all going fifty to 
a roundabout nearby the dog 
next door was Honey 
a lab as old as me who loved 
to lie on the just 
cut lawn and sniff her tail 
going in the afternoon sun

I like to track the winners of the Christopher Ewart-Biggs Memorial Prize because of my own past association with it, and was really interested to see that earlier this month it went to a book of poetry, The Sun is Open, by QUB-based writer Gail McConnell. In fact the 119 pages of text are one long poem broken into chunks, playing with text and with font colour, processing the writer’s reaction to going through a box of her father’s things, long after he died in 1984 at 35, shot dead by the IRA while checking under his car for bombs, in front of his wife and his then three-year-old daughter.

Gail McConnell barely remembers her father and has no memory of that awful day, but of course it has affected her whole life, and the poetry captures that disruption and the effect of engaging with her father through a box of personal souvenirs, most notably a diary and a Students Union handbook from his own time at QUB. There is some incredible playing with structure – quotations from the box are in grey text, documents are quoted in fragments to let us fill in the blanks, at one point the page fills with vertical bars to symbolise the prison where her father worked. It’s provocative and unsettling, and meant to be. You can get it here.

This moved me to seek out an earlier poem by Gail McConnell, Type Face, which you can read here. It’s funnier, though the humour is rather dark; the theme is that it explains her reaction to reading the Historical Enquiries Team report into her father’s murder, and discovering that it was written in Comic Sans. The third verse is:

‘Nothing can separate me from the Love of God
in Christ Jesus Our Lord’. Nothing can, indeed.
I am guided by Google, my mother by Christ.
Awake most nights, I click and swipe.
I search and find Bill McConnell Paint and Body.
Under new management!!!!! Northeast Tennessee.
Where is God in a Messed-up World? Inside the Maze.
(My phone flashes up a message like a muse.)
Straight & Ready: A History of the 10th Belfast
Scout Group. (35) (PO) (IRA)
– for more and a photograph, push this link>>
the maroon death icon on
You visited this page on 06/02/15.
And here I am again.
And in The Violence of Incarceration
(Routledge, 2009), eds. Phil Scraton
and Jude McCulloch (page thirty-three), he
‘oversaw, but later denied in court, the brutality
of prison guards, [and] was executed by the IRA
on the 8 March 1984.’ (He’d been dead two days
by then.) Execute. Late Middle English:
from Latin exsequi ‘follow up, punish’.
There’s a listing on,
‘an [sic] non-sectarian, non-political’ nook
complete with Union Jack and Ulster flag
campaigning pics, the Twitter feeds and tags,
a calendar and videos. Powered by WordPress.
And then there’s Voices from The Grave (and this
one’s hard to bear, though can I say so? I don’t know.
I don’t know. I don’t know. I don’t know.)
I won’t write down the page. But something in me,
seeing that crazed portrait – something’s relieved.

Really good stuff, and very different in presentation from The Sun is Open. Both are recommended.

Ewart-Biggs Prize winners: Observe the Sons of Ulster Marching Towards the Somme, by Frank McGuinness | From A Clear Blue Sky: Surviving the Mountbatten Bomb, by Timothy Knatchbull | Setting the Truth Free: The Inside Story of the Bloody Sunday Campaign, by Julieann Campbell | The Republic: The Fight for Irish Independence, 1918-1923, by Charles Townshend | The Whole and Rain-Domed Universe, by Colette Bryce | The Sun is Open, by Gail McConnell

The Northern Ireland Protocol: or, Now Look What You Made Me Do

It’s been a while since I have written at length about Brexit, but the most recent developments have driven me to put some electrons together on the topic. By way of introduction, I participated in a televised manel discussion on Al-Jazeera on Wednesday with Duncan Morrow of Ulster University and Graham Gudgin from Cambridge, which you can watch here:

Also very importantly, the excellent Brexit Witness archive has published a wide-ranging interview with Andrew McCormick, one of the best of the mandarins in the Northern Ireland Civil Service. I do recommend reading the whole thing, but Tony Connelly of RTE has published a summary here.

The other recent development, of course, is the Northern Ireland Assembly election. I wrote about the raw numbers last weekend; there is some necessary analysis which will come now.

To begin at the beginning. The Brexiters lied from the start about the effect of Brexit on the Irish border. Here is a BBC story from February 2016, four months before the referendum, featuring Boris Johnson stating that Brexit would leave arrangements on the Irish border “absolutely unchanged”. This was clearly untrue; taking back control of the UK’s borders was a constant theme of the Brexit campaign, and it was and is ridiculous to say that this would have no practical consequences on the UK’s only land border.

It seems however that this only slowly dawned on Whitehall after the referendum result. When the EU insisted that citizenship rights, financial obligations and arrangements on the Irish border should be sorted out as part of the divorce deal, the British initially found the first two much more difficult to swallow. Many in London seemed to believe that Chancellor Merkel of Germany would tell the Irish to accept whatever deal suited the British for the sake of future car exports, thus completely misunderstanding the weight of individual member states in the EU system, not to mention the politics of the German car industry.

Part of the myth spread by Brexit secretary David Davis is that the Irish government drastically hardened its line on sorting out the border when Leo Varadkar took over from Enda Kenny as Taoiseach in 2017. Again, this is untrue, but I categorise it as a misunderstanding rather than a lie; what actually happened in 2017 was that the UK actually started paying attention to the fact that Dublin had a view (to say that London was actually listening would be a step too far).

There is no need to rehearse at length the agony of Theresa May, who eventually realised that the hard Brexit to which she had committed herself at the start of the process would be disastrous if implemented on the Irish border, but failed to take her party with her, let alone the DUP. Johnson, having replaced her as Prime Minister with the help of the DUP, then (to my surprise) agreed a deal with Leo Varadkar including a special status for Northern Ireland which became the Protocol.

To remind you: the Protocol keeps Northern Ireland inside the EU’s single market and customs union, in order to avoid customs checks on the land border. But since the UK has “taken back control”, this inevitably means that somewhere there must be customs and other checks on goods which might travel from the rest of the UK to the EU, and if the Border is to remain open, that means that those checks take place in the Irish Sea, between England, Scotland and Wales to the east, and Ireland and Northern Ireland to the West. The great sitcom Parlement spoofed these discussions rather well:

The question is, how did we end up with a situation where Boris Johnson claimed to have an “oven-ready” deal with the EU before the 2019 election, and now repudiates the Northern Ireland Protocol, one of the core planks of that deal? The UK government’s defenders make various arguments. Some say that the EU has been too tough in implementing the rules (which in fact have not yet been implemented in any meaningful sense). Some (including the then UK chief negotiator, David Frost) say that the deal was negotiated too quickly (after three and a half years, which does not really seem too short a time to prepare).

Dominic Cummings, who was Johnson’s chief of staff at the time, says that he and his team always intended to renege on “the bits we didn’t like” after it had been signed and the December 2019 election won, but he does not think that Johnson himself actually understood it. I am inclined to agree with those who think that Johnson was being actively mendacious rather than ignorant or stupid; he famously assured Northern Irish business leaders that they should throw any new forms in the bin, even though that is clearly what his deal would have required if he had had the slightest intention of implementing it.

The UK now threatens to unilaterally disapply the Protocol starting next week, provoking a trade war with the EU at precisely the moment that the West needs to be united in support of Ukraine. It is alleged that the new arrangements have made life worse in Northern Ireland (though the government’s own economists report that thanks to the Protocol, Northern Ireland’s economy is outperforming the rest of the UK’s). The EU is blamed for creating the trade barriers which the UK demanded and agreed to. The UK, now keen to sign trade agreements with the rest of the world, is about to tear up its biggest agreement, with its closest and largest trading partner. Not hugely smart.

Why do this? I ask again. My view is that Conservatives in general, who are genuinely and deeply emotionally attached to the Union, cannot bear the thought of implementing a trade and customs frontier inside the UK. Johnson assured them in 2019 that it would be all right, no matter what might actually be written in the deal, and they believed him, despite his track record with the truth. So I predict that the Johnson government, however long it lasts, will not implement the Protocol in any meaningful way.

On top of that, the consequences of fighting with the EU are largely positive for the Conservatives. It keeps Brexit going and puts Labour in a difficult position. Sure, there are economic consequences, but they are lost in the static of post-pandemic recovery and the effects of the war in Ukraine, and will be most felt in Northern Ireland where the Conservatives do not stand anyway. Few Conservatives care about the damage to the UK’s international reputation – they are all foreigners, after all. The strategy is in fact to fight rather than to win.

There is very little appetite in Brussels, Dublin or other capitals to give the British what they currently say they want. This goes right back to the early days of Brexit, when the EU was very alert to the potential for the UK to undermine the Single Market. In addition, the UK’s July 2021 Command Paper on the Protocol ambitiously rewrote the recent history of the relationship to an extent that was unrecognisable outside Westminster and further undermined trust. The tactics of escalation have failed to convince other capitals that the British are serious about finding solutions. It’s also noticeable that the current escalation is coming from the UK Foreign Secretary, Liz Truss, who clearly has ambitions to be the next Conservative leader, an election that must come sooner or later, and also needs to put her own previous pro-Remain baggage behind her.

The actual situation on the ground in Northern Ireland is barely relevant to Conservative decision-making. The DUP do have an outsized influence with the Tories because they have the largest delegation at Westminster, and their MPs are well networked with the Conservative back-benchers; Sinn Fein are not there at all, the SDLP have only two MPs, one of whom is the party leader, and Alliance have only one, who is the party’s deputy leader. But one should not exaggerate this factor; it did not help the DUP when the 2019 deal was passed, over their loud objections about the Protocol.

So, the last part of this post is about Northern Ireland, where the DUP last week paid the price at the ballot box for their strategic mistakes of the last few years. I wrote briefly about Arlene Foster’s leadership when she resigned; it’s worth adding that the DUP’s pledge to punish the Northern Ireland institutions, by not allowing a government to be formed until the Protocol has gone, has a real whiff of Blazing Saddles. Yes, it is a functional political problem that Unionists as a whole do not accept the Protocol; but Stormont has very little to do with that, and Westminster is where the battle actually is. (Unlike almost everyone else, I’m therefore actually rather sympathetic to Jeffrey Donaldson’s stated intention to remain an MP for the time being.)

That brings us to the other side of the DUP’s policy choices. There is a very strong perception among non-Unionists that the real reason that the DUP do not want to reinstate the Northern Ireland Executive is that Sinn Fein would get the position of First Minister, thanks to the rewriting of the rules at the behest of the DUP in 2007. Personally, I share that perception, though I will be glad to be proved wrong. If I am right that the UK government is about to escalate the situation with the EU, we will soon see if the DUP is actually prepared to accept the result of an election that it did not win. (For more on the election, see the very interesting analysis by Lee Reynolds.)

The DUP is under threat from Jim Allister’s Traditional Unionist Voice, which snatched a quarter of their 2017 votes away on 5 May (though remarkably failed to win any seats); Allister is very clear that Sinn Fein should not be allowed in government at all, and that the DUP would be stooges for enabling them to lead it, and the voters who defected to him from the DUP presumably feel the same. But if Northern Ireland is to have a long term future at all as a society, power-sharing is essential – as my father recommended in 1971.

A brief personal parenthesis: Both Jim Allister (when he was an MEP) and Sir Jeffrey Donaldson (when he was a member of the Council of Europe’s Parliamentary Assembly) have been personally helpful to me in the past, knowing full well that I disagree with them on a lot of things, so I want to state on the record that I respect and salute their professionalism.

But if you are attached to the Union between Great Britain and Northern Ireland – and I am not – you will need to start selling the case for the Union better; as Lee Reynolds puts it, “The declining politics of birth and disappeared politics of push must be replaced by the politics of persuasion.” Crucially, you will need to show that Unionism accepts election results even when it doesn’t win; non-Unionists have had to accept that for a century.

We’re not yet at the stage where a border poll has become an immediate prospect, but we are not all that far away either. I wrote three years ago (scanned here) that voters in the convinceable middle, who historically have conditionally supported the Union, can foreseeably be persuaded to join a united Ireland, if three things happen:

  1. Brexit turns out badly (✔)
  2. Unionism continues to be worse than Nationalism at appealing to its own core vote and not engaging with the centre (✔)
  3. There is a better offer on the table from Nationalists (currently quite far from being achieved, and in particular the need for Nationalists to find a convincing narrative on health services is even more acute after the last two years).

Nothing is certain in politics, but the current direction of travel is clear, and the DUP and the Conservative Party are doing nothing to stop it.

Northern Ireland Assembly: final results #AE22


(And two Independent Unionists, an increase of one)

Unionists 37 (-3)
Nationalists 35 (-4)
Others 18 (+7)

Calculating total vote tallies between the sides is complicated by minor parties and candidates, but the headline is that Unionists and Nationalists are not far apart. I had previously said that if Nationalists outnumber Unionists at a Stormont election, there are grounds for the Sec of State to call a border poll. That threshold is not clearly met in terms of votes, and clearly not met in terms of seats won.

Ten seats changed hands in the election. Alliance gained nine – four from the SDLP, two each from the DUP and Greens and one from the UUP; and the DUP lost another seat in North Down where a former party colleague retained his seat as an independent.

SF did not gain or lose any seats, but became the largest party as the DUP tally fell. They missed out on two potential gains by poor balancing of their candidates, in East Londonderry and Upper Bann, and the UUP might also have had a chance of retaining both seats in East Antrim with better balancing.

The closest result was in Foyle, where the DUP survived a UUP challenge by 95 votes. That’s on the final count; the closest decisive elimination was in East Londonderry, where Alliance candidate was eliminated 15 votes behind the SDLP and his transfers then elected her.

For the TUV to get only one seat despite vote share of 7.6% is remarkable – proportionally that should have given them at least six! But they had great difficulty in attracting transfers. Conversely the DUP’s total of 25, while disappointing for the party, is about six more than would be proportionally expected from a 21.3% vote share.

Constituencies listed below in (rough) order of increasing Nationalist and decreasing Unionist vote share.

Lagan Valley

Jeffrey Donaldson DUP12,626
Robbie ButlerUUP8,242
Sorcha EastwoodAlliance8,211
Paul GivanDUP5,062
David HoneyfordAlliance4,183
Lorna SmythTUV3,488
Pat CatneySDLP3,235
Gary McCleaveSF2,725
Laura TurnerUUP1,607
Gary HyndsInd735
Simon LeeGreen648
Amanda DohertyPBP271
Alliance24.3%+10.7%2 (+1)
SDLP6.3%-2.1%0 (-1)
Sinn Féin5.3%+1.3%
Green Party1.3%-0.8%

SDLP lost to Alliance by 643.56 votes on the last count, a gain that was not unexpected.

North Down

Alex EastonInd9,568
Andrew MuirAlliance6,838
Stephen DunneDUP6,226
Connie EganAlliance5,224
Alan ChambersUUP3,825
Rachel WoodsGreen2,734
Jennifer GilmourDUP2,068
John GordonTUV1,574
Naomi McBurneyUUP1,342
Déirdre VaughanSDLP727
Thérèse McCartneySF687
Ray McKimmInd604
Matthew RobinsonCons254
Chris CarterInd72
Alliance28.9%+10.3%2 (+1)
DUP19.9%-17.6%1 (-1)
Green6.5%-7.2%0 (-1)
Others25.1%1 (+1)

Alliance took the Green seat by 2500.82 votes, one of two seats gained by Alliance from the Greens.


Kellie ArmstrongAlliance7,015
Michelle McIlveenDUP6,601
Stephen CooperTUV5,186
Harry HarveyDUP4,704
Mike NesbittUUP3,693
Peter WeirDUP3,313
Nick MathisonAlliance2,822
Philip SmithUUP2,535
Conor HoustonSDLP2,440
Róisé McGivernSF1,607
Maurice MacartneyGreen831
Ben KingInd118
DUP35.8%-4.2%2 (-1)
Alliance24.1%+9.1%2 (+1)

Alliance won the last seat by 249.77 votes ahead of the TUV, Mathison taking the fifth seat despite having started in 7th place. This was the TUV’s best chance of a gain, but they were simply too transfer-repellent.

East Antrim

Gordon LyonsDUP6,256
John StewartUUP6,195
David HilditchDUP5,662
Stewart DicksonAlliance5,059
Danny DonnellyAlliance4,224
Oliver McMullanSF3,675
Norman BoydTUV3,661
Roy BeggsUUP3,549
Siobhán McAlisterSDLP1,200
Mark BaileyGreen754
UUP24.2%+1.5%1 (-1)
Alliance23.1%+7.1%2 (+1)

TUV were 2076.4 behind DUP for last seat. Good balancing from Alliance who took one of the UUP’s seats despite starting with fewer votes. This was the only UUP seat lost in the election.

East Belfast

Naomi LongAlliance8,195
Joanne BuntingDUP7,253
David BrooksDUP6,633
Peter McReynoldsAlliance5,820
Andy AllenUUP5,281
John RossTUV3,087
Brian SmythGreen2,302
Mairead O’DonnellSF1,369
Lauren KerrUUP1,282
Karl BennettPUP970
Hannah KennyPBP500
Charlotte CarsonSDLP484
Eoin MacNeillWP72

UUP got the last seat by a pretty massive 3988.96 votes ahead of the Greens.

North Antrim

Robin SwannUUP9,530
Philip McGuiganSF9,348
Jim AllisterTUV8,282
Mervyn StoreyDUP6,747
Paul FrewDUP6,242
Patricia O’LynnAlliance4,810
Matthew ArmstrongTUV2,481
Eugene ReidSDLP1,919
Bethany FerrisUUP856
Paul VeronicaGreen343
Laird ShingletonInd66
DUP25.7%-15.0%1 (-1)
Alliance9.5%+4.1%1 (+1)

Alliance took the last seat by 288.45 votes ahead of the DUP, possibly the least anticipated of the party’s gains. NB that O’Lynn is the first woman elected to the Northern Ireland Assembly from North Antrim, even going back to 1973 and 1982.

South Antrim

Declan KearneySF9,185
John BlairAlliance7,315
Pam CameronDUP6,899
Steve AikenUUP5,354
Trevor ClarkeDUP4,943
Mel LucasTUV4,371
Roisin LynchSDLP3,139
Paul MichaelUUP2,821
Róisín BennettAontú657
Lesley VeronicaGreen539
Andrew MoranInd262
Jerry MaguirePBP251

DUP got last seat by 1878.25 ahead of SDLP. Another case where if the TUV had been more transfer-friendly, they could have been in contention.

Upper Bann

John O’DowdSF9,242
Jonathan BuckleyDUP8,869
Liam MackleSF7,260
Diane DoddsDUP6,548
Eóin TennysonAlliance6,440
Doug BeattieUUP5,199
Darrin FosterTUV4,373
Dolores KellySDLP3,645
Glenn BarrUUP3,367
Aidan GribbinAontú571
Lauren KendallGreen459
Glenn BeattieHeritage128
Alliance11.5%+6.2%1 (+1)
SDLP6.5%-3.4%0 (-1)

Nationalists won only one seat out of five despite 36% of first preferences, as Alliance took the last seat by 376.07 votes ahead of SF.

East Londonderry

Caoimhe ArchibaldSF6,868
Maurice BradleyDUP6,786
Alan RobinsonDUP5,151
Kathleen McGurkSF4,500
Claire SugdenInd3,981
Cara HunterSDLP3,664
Chris McCawAlliance3,338
Jordan ArmstrongTUV2,959
Darryl WilsonUUP2,625
Stephanie QuigleyInd1,503
Gemma BrollyAontú1,095
Russell WattonPUP933
Mark CoulsonGreen347
Amy MerronPBP347
Niall MurphyInd181
Billy StewartInd82

A lot of people, myself included, had written the SDLP off here based on first preferences, but they kept their seat. The decisive stage was the penultimate count, when the Alliance candidate was eliminated being 14.56 votes behind the SDLP; his transfers then elected her comfortably by 1666.56 votes ahead of SF.

North Belfast

Gerry KellySF8,395
Carál Ní ChuilínSF7,932
Phillip BrettDUP6,329
Brian KingstonDUP4,844
Nuala McAllisterAlliance4,381
Nichola MallonSDLP3,604
Ron McDowellTUV3,335
Julie-Anne Corr-JohnstonUUP2,643
Mal O’HaraGreen1,446
Fiona FergusonPBP1,059
Billy HutchinsonPUP762
Seán Mac NiocaillAontú640
Stafford WardInd489
Lily KerrWP168
Alliance9.5%+1.1%1 (+1)
SDLP7.8%-5.3%0 (-1)

SDLP lost their seat to Alliance by 991.21 votes.

Fermanagh and South Tyrone

Jemma DolanSF9,067
Colm GildernewSF7,562
Áine MurphySF7,379
Tom ElliottUUP5,442
Deborah ErskineDUP5,272
Paul BellDUP4,255
Adam GannonSDLP3,836
Alex ElliottTUV3,091
Rosemary BartonUUP2,912
Matthew BeaumontAlliance2,583
Denise MullenAontú927
Dónal Ó CofaighCCLab602
Kellie TurtleGreen335
Emma DeSouzaInd249
Derek BackhouseInd128
Emmett KilpatrickPBP103

The last seat was decided by a 508.12 vote margin between the two DUP candidates.

South Belfast

Deirdre HargeySF9,511
Edwin PootsDUP7,211
Paula BradshawAlliance6,503
Matthew O’TooleSDLP5,394
Kate NichollAlliance5,201
Clare BaileyGreen4,058
Stephen McCarthyUUP3,061
Andrew GirvinTUV1,935
Elsie TrainorSDLP2,030
Luke McCannAontú806
Sipho SibandaPBP629
Neil MooreSocialist353
Paddy LynnWP139
Elly OdhiamboInd107
Alliance24.9%+7.2%2 (+1)
Green8.6%-1.2%0 (-1)

Alliance took the Green seat by 911 votes.

West Tyrone

Nicola BroganSF8,626
Maolíosa McHughSF6,658
Tom BuchananDUP6,640
Declan McAleerSF6,343
Daniel McCrossanSDLP5,483
Trevor ClarkeTUV4,166
Stephen DonnellyAlliance2,967
Ian MarshallUUP1,876
Paul GallagherInd1,682
James HopeAontú657
Carol GallagherPBP354
Susan GlassGreen252
Amy FergusonSocialist171
Barry BrownInd119

SF got the last seat 2707.36 votes ahead of TUV.

Newry and Armagh

Conor MurphySF9,847
Cathal BoylanSF9,843
Liz KimminsSF7,964
William IrwinDUP7,577
Justin McNultySDLP6,217
Keith RatcliffeTUV5,407
David TaylorUUP3,864
Jackie CoadeAlliance3,345
Gavin MaloneInd3,157
Daniel ConnollyAontú1,189
Ciara HenryGreen314
Nicola GrantWP160

DUP got last seat 2892 votes ahead of TUV.

Mid Ulster

Michelle O’NeillSF10,845
Keith BuchananDUP8,521
Emma SheerinSF8,215
Linda DillonSF8,199
Patsy McGloneSDLP5,144
Glenn MooreTUV3,818
Meta GrahamUUP2,191
Claire HackettAlliance2,138
Alixandra HallidayAontú1,305
Patrick HaugheyInd877
Sophia McFeelyPBP179
Stefan TaylorGreen137
Hugh ScullionWP107
Conor RaffertyResume13

The SDLP took the last seat by a 3446 margin over the TUV.

South Down

Sinéad EnnisSF14,381
Cathy MasonSF9,963
Patrick BrownAlliance6,942
Diane ForsytheDUP6,497
Colin McGrathSDLP6,082
Harold McKeeTUV3,273
Karen McKevittSDLP3,006
Jill MacauleyUUP2,880
Rosemary McGloneAontú1,177
Noeleen LynchGreen412
Paul McCroryPBP205
Patrick ClarkeInd134
SDLP16.5%-8.6%1 (-1)
Alliance12.6%+3.5%1 (+1)

The last seat was decided between the two SDLP candidates by a margin of 3859.17 votes.


Pádraig DelargySF9,471
Mark H. DurkanSDLP7,999
Ciara FergusonSF5,913
Gary MiddletonDUP4,101
Ryan McCreadyUUP3,744
Brian TierneySDLP3,272
Sinéad McLaughlinSDLP3,189
Shaun HarkinPBP2,621
Rachael FergusonAlliance2,220
Emmet DoyleAontú2,000
Anne McCloskeyInd854
Colly McLaughlinIRSP766
Elizabeth NeelyTUV499
Gillian HamiltonGreen215

In the longest count of the election, leading to the closest final count result, the DUP retained their seat by a margin of 95 votes over the UUP.

West Belfast

Danny BakerSF9,011
Órlaithí FlynnSF6,743
Pat SheehanSF6,373
Aisling ReillySF5,681
Frank McCoubreyDUP4,166
Gerry CarrollPBP3,279
Paul DohertySDLP2,528
Gerard HerdmanAontú1,753
Dan MurphyIRSP1,103
Donnamarie HigginsAlliance907
Jordan DoranTUV802
Linsey GibsonUUP474
Stevie MaginnGreen307
Patrick CrossanWP193
Gerard BurnsInd192
Tony MallonInd129
Declan HillInd26

The DUP lost to PBP by 532.40 votes.

Northern Ireland elections today

I’m in Belfast ready to comment on the election results as they come in tomorrow. I think John Taylor, now Lord Kilclooney, caught the mood of anticipation very well last week:

We’re seriously in the zone where it is generally considered likely that the DUP will come second to Sinn Féin in terms of votes and seats; and where it is questionable whether there will be more Unionists or more Nationalists in the new Assembly. The overall context is that polling predicts that both SF and the DUP will lose votes, but that the DUP will lose crucially more.

My one anecdote of the election comes from the parents of a friend from Sandy Row, a Loyalist area of central Belfast, who kindly gave me a lift from the airport last night. They told me that they had not seen a single canvasser from any party during the campaign. I have been critical of Unionists for reaching out only to their traditional voters and ignoring the potentially persuadable centre; but if traditionally Unionist voters are also being ignored, Unionism is in worse trouble than I realised. My friend’s father, once a regular DUP voter, was discreet about his voting intention today, but I noted that the only local candidate for whom he had a good word was Matthew O’Toole of the SDLP.

I want to throw one more set of data into the mix: who won the last seat, and who missed out, in each of the 18 constituencies in the 2017 election?

The interesting thing is that when you look at the Unionist/Nationalist marginals in 2017, Nationalists won almost all of them and there is not much left to pick up. The only two that are at all likely are Strangford, where the SDLP have been runners-up at every Assembly election since the Good Friday Agreement, and East Antrim, where SF start a bit farther off (but managed to win one out of six in 2011 and 2017). In both cases, the DUP look more vulnerable than the UUP, so a loss will affect the race for biggest party as well as the race for designations. Strangford and East Antrim are the two that I shall be watching most keenly tomorrow from that point of view.

2017 was a very good election for Nationalists – SF voters were motivated by Arlene Foster’s very ill-judged “crocodile” comments, and the SDLP, more by accident than design, did much better than the UUP out of their electoral alliance. The result was that in six seats where on previous electoral records one could have seen the glimmer of a chance of an extra Unionist seat, the last MLA elected was a Nationalist, some way ahead of the Unionist runner up. Those seats are West Belfast, Mid Ulster, West Tyrone, Newry and Armagh, Fermanagh and South Tyrone and Lagan Valley. To stem the tide, Unionism would need to look like it could make gains in several of them, and in a good year that should be possible; but this does not look like a good year.

The last seat in 2017 was contested between two Unionist candidates – in fact two DUP candidates in each case – in East Belfast, North Antrim and South Antrim. That tends to suggest that in those seats at least there is room for more slippage in the overall Unionist vote before a currently held seat is seriously at risk. The same is true on the other side where two Nationalist candidates were chasing hard for the second Nationalist quota in East Londonderry and Upper Bann, the SDLP winning in one case and SF in the other. In East Londonderry, it’s worth watching whether UUP transfers again help the SDLP over the line (or indeed whether they will be needed).

Within Unionism, the DUP is likely to be further eroded by Jim Allister and the TUV. Polls have them just at the level where they might make a breakthrough, or might be disappointed with just one or two gains. What gains they do make are likely to be directly from the DUP, further eroding their chance of being the largest party. Also on this point, watch Alex Easton, challenging his former DUP colleagues in North Down, Northern Ireland’s most volatile constituency.

I have not said much about the SDLP or UUP, because in this election they are really barometers for the dissatisfaction of the committed Nationalist or Unionist voter with SF or the DUP respectively. It seems fairly clear that SF will slip a bit less than the DUP because their narrative is a bit more coherent. There’s also the case of Fermanagh and South Tyrone where the SDLP were unlucky at an early stage of the 2017 count, and can hope for that luck to turn. But the UUP-SDLP transfers that made a difference in a couple of key seats last time may not be as readily available in 2022.

The centre ground will provide yet more interesting dynamics for the election. Of the 18 last elected MLAs in 2017, none were Alliance (or People before Profit). Two were the two Green Party MLAs in South Belfast and North Down, both of them elected well ahead of Unionist runners-up, so it’s a reasonable assumption that all of the centre ground seats held in 2017 will survive into 2022. In addition, People Before Profit are snapping at Unionist heels in Foyle. And most interesting of all, Alliance came closer than I for one expected to depriving Nationalists of a seat in both North Belfast and South Down. I can imagine a situation where Nationalists gain in, say, Strangford, but lose a couple of seats unexpectedly elsewhere, leaving the two sides on level pegging.

On a slightly different note, I have trawled Twitter and other sources, including an excellent series of posts on Slugger O’Tooler by Michael Hehir, to calibrate expectations of gains and losses in today’s vote. You will note how few potential Nationalist to Unionist shifts are listed.

Encouraging, but as expected
(failing to gain these seats is a poor result)
Disappointing, but as expected
(retaining these seats is a major triumph)
Alliance gain North Belfast from SF
Alliance gain South Down from SDLP
Alliance gain South Belfast from Greens or DUP
Alliance gain East Antrim from DUP or UUP
SDLP gain Fermanagh S Tyrone from SF
Ind U gains North Down from DUP
SF lose North Belfast to Alliance
SF lose Fermanagh S Tyrone to SDLP
DUP lose Strangford to Alliance, TUV or SDLP
DUP lose North Down to Ind U
SDLP lose South Down to Alliance
UUP lose East Antrim to Alliance or TUV
A good dayA bad day
Alliance gain Lagan Valley from SDLP
Alliance or TUV gain Strangford from DUP
Alliance gain Upper Bann from SDLP or DUP
UUP gain Newry and Armagh from SF or DUP
TUV gain East Antrim from DUP or UUP
PBP gain Foyle from DUP or SF
SF lose West Tyrone to UUP or Alliance
DUP lose East Antrim to TUV or Alliance
DUP lose East Belfast to Green
DUP lose Foyle to PBP or UUP
SDLP lose Lagan Valley to Alliance
SDLP lose Upper Bann to Alliance or SF
Greens lose South Belfast to Alliance
An exceptionally good dayAn exceptionally bad day
SDLP gain Strangford from DUP
Alliance gain North Down from Greens
SDLP gain West Belfast from SF
UUP gain West Tyrone from SF
UUP gain North Belfast from DUP
Greens gain East Belfast from DUP
SF lose West Belfast to SDLP
SF lose Newry and Armagh to UUP
DUP lose North Belfast to Alliance or UUP
DUP lose North Antrim to Alliance or TUV
DUP lose South Antrim to TUV or UUP
DUP lose Upper Bann to UUP or Alliance
Greens lose North Down to Alliance
Anything elseAnything else

I don’t often agree with John Taylor, Lord Kilclooney, quoted at the top of this post. But he has 30 more years of observing and participating in Northern Ireland elections than I do, and won his first one before I was born (and I’ve just turned 55). So I take his sentiment seriously in this case, and I rather agree with it. It’s going to be an interesting day tomorrow.

And according to the current BBC schedule you can watch me at the following times:

Tomorrow: 1215-1300; 1330-1800; 1900-2200; 2230-late

Saturday (if they are still counting) 1000-1300.

Edited to add: With 4.25 million registered voters for Scotland’s local government elections today, and 1.37 million in Northern Ireland for the Assembly, is this the biggest ever set of elections on the same day using the Single Transferable Vote in UK history? I have considered the Irish local government elections of 1920, and the simultaneous elections for the House of Commons of Northern Ireland and the Second Dáil / House of Commons of Southern Ireland in 1921, but I think that the total Irish electorate then was less than the sum of Northern Ireland and Scotland today. Not to mention the large numbers of uncontested seats, and the questionable extent to which those could be described as “UK” elections!

Belfast, by Kenneth Branagh

A week ago (wow, it’s been such a long week) the British Mission to the EU and the Northern Irish representative office jointly put on a showing of the Kenneth Branagh film Belfast at the Bozar in central Brussels.

It was just lovely to actually have a physical reception, after two years when it was very difficult. The British Ambassador made a wee speech:

Also the Northern Ireland deputy representative made a wee speech; and my friend Paul took photos, in the first of which my back is visible at the left.

Probably the majority of us in the crowd were Norn Iron exiles in Brussels; a few arty people had come over specially for the occasion, but basically this was UKMis showing that it was a good social actor; and succeeding.

Here’s the official trailer.

So, what did I think of the film?

I was born in 1967, so I was about 2 at the point that the events of the film unfold. (Though apparently “helicopter” – or “ally-agga” – was one of my first words, as we saw them zoom west over the garden.) A lot of it doesn’t really speak true to the Belfast experience. Nobody ever wandered down the streets with flaming torches. There were no Indian corner shops, and no ethnic minority teachers. There was no wee Catholic girl in the Protestant school. The bus to the airport didn’t exactly stop in side streets to pick people up. The houses were (and are) much smaller. The accents are much stronger in real life.

At the same time, one can forgive a lot of this for the sake of Art. Jamie Dornan, Ciaran Hinds and Judy Dench are all actors who I knew anyway; I have seen one episode of Outlander starring Caitriona Balfe and now I have a strange impulse to see more. And young Jude Hill, from my ancestral part of the world, is glorious as the main character. The (Oscar-winning) script really crackles.

Pa: It’s all bloody religion. That’s the problem.
Buddy: Then why are you sending us to church?
Pa: Because your granny would kill me if I didn’t.

Pop: [to Buddy] Women are very mysterious.
Granny: And women can smash your face in too, mister.
Pop: Your granny’s become less mysterious over the years.

(After the supermarket has been looted)
Ma: Why did you take that washing powder?
Buddy: It’s biological!

Having said that, there’s no real interrogation of why the Troubles started in the first place. I think it would have been helpful for the audience to know that decades of injustice and discrimination do eventually bring the chickens home to roost. The impression given is that violence erupted purely out of sectarianism and bigotry at local level, which is far from the whole truth.

Having said that, I think almost all of us in Bozar related to the central dilemma of Buddy’s parents in the film; will you stay or will you go? And from the mere fact that we were in Brussels, we were all exiles, whether permanently or temporarily; and it was easy to relate to the problem of leaving a city that you love, and yet where you can’t live, and risking everything on a foreign venture. It works for a lot of people; it worked for Kenneth Branagh’s family and mine; it doesn’t work for everyone, and before you do it you don’t know what category you’ll end up in.

So yeah. I liked it, warts and all.

Northern Ireland Assembly elections 2022: a preview

2017 Assembly results map
2017 Assembly election results in each constituency

Northern Ireland elects a new Assembly on 5 May, and as usual I have been crunching some numbers to establish a baseline of expectations – basically the results which would not be surprising, given the current polling which has the DUP in particular down a bit from last time. Given that there is more movement on the Unionist side, I’m taking the 18 seats in order from most to least Unionist, Lagan Valley to West Belfast. The headings in each case link to my website, where there is a lot more information.

Edited to add, after the election: Without changing my original text, I’ve noted the results in each seat.

Lagan Valley

The SDLP’s seat, won with UUP transfers after starting from barely half a quota in 2017, is on paper the most vulnerable, and on recent performance Alliance is best placed to pick it up. The DUP’s decision to run only two candidates in a constituency where their third was the runner-up in 2017 is telling. With only five Unionist candidates and almost four Unionist quotas, one cannot exclude a lucky day for the second UUP runner or for the TUV.

What happened: Alliance gained from the SDLP.

North Down

In the most volatile constituency in Northern Ireland, Alliance is strongly placed to gain a seat based on recent performance; each of the other three parties represented here is potentially vulnerable, with the DUP facing the added complication of their former member and sitting MLA Alex Easton standing as an independent candidate.

What happened: Alliance gained from the Greens, and the DUP did indeed lose their seat to Alex Easton.


It’s very difficult to see the DUP holding all three seats here, even in a good year (and this probably won’t be a good year). On recent showings, Alliance are better placed than the UUP to pick up; but the SDLP, who have been runners-up here in all six Assembly elections since the Good Friday Agreement, cannot be ruled out.

What happened: Alliance gained from the DUP.

East Antrim

All five men elected here in 2017 are standing again, and the likeliest outcome is no change. But this is the constituency where the TUV have the best prospects of a gain, which could come from either of the other Unionist parties. If Alliance can balance two candidates ahead of SF, they too have a chance; as indeed do SF, in theory, if they can keep comfortably ahead of the trailing Alliance candidate and pull in transfers.

What happened: Alliance gained from the UUP.

East Belfast

The status quo is the most likely outcome in terms of seats – 2 DUP, 2 Alliance and 1 UUP; another constituency where the third DUP candidate was runner-up in 2017, but they are only running two this year. One cannot exclude a successful challenge from TUV or PUP, given the UUP’s historical weakness here and the DUP’s current low poll ratings. Alliance are some way off a third quota, even with all available Nationalist transfers, and anyway have only two candidates.

What happened: No change.

North Antrim

All five men elected in 2017 are standing again. Alliance will be hopeful of a gain here, but the Unionist vote remains close to four quotas and Sinn Fein’s seat looks solid enough. Within Unionism, the DUP vote would have to fall pretty far, with poor balancing, for either the UUP or TUV to threaten their second seat.

What happened: Alliance gained from the DUP, whose balancing was good but their vote fell.

South Antrim

Another seat where the third DUP candidate was runner-up in 2017, but they are only running two this time. Both Alliance and SF look secure; with only three Unionist seats, the UUP must feel that they are in with a chance of picking up the third.

What happened: No change.

Upper Bann

The SDLP were fortunate to get UUP and Alliance transfers in 2017, securing the second Nationalist seat despite SF starting with almost three times as many votes. Polling shows SF down a bit more than the SDLP, so the status quo is mildly more likely than not to prevail on the Nationalist side. A good day for the UUP would see them take the DUP seat here. Alliance are a bit further behind, but not all that far.

What happened: Alliance gained from the SDLP.

East Londonderry

The SDLP’s 2017 performance here was poor, and hit by defection, yet they still managed to get the second Nationalist seat despite starting far behind SF. If the polls are right, and the SDLP’s rating is stable with SF down a bit, the same result on the Nationalist side is more likely than not. Alliance must have hopes of a gain here; but has Claire Sugden already got those votes? The DUP vote would have to fall quite a long way for their second seat to be under threat.

What happened: No change.

North Belfast

It’s difficult to see a third Unionist seat here, and also difficult to see another Unionist emerging to challenge the DUP for their second seat – in 2017, the DUP had just under two quotas, but their nearest rivals, the UUP, less than a third of one. Alliance expect to challenge strongly, but all three Nationalists look reasonably secure, with perhaps the SDLP least so.

What happened: Alliance gained from the SDLP.

Fermanagh & South Tyrone

SF were very lucky to get the third Nationalist seat in 2017, and the SDLP should expect to regain it if they are to make any headway anywhere at all in the election. The UUP were fortune to pick up the second Unionist seat in 2017, but indications are that this will be a better year for them, so the seat split between parties on the Unionist side is likely to remain the same. Both the DUP and UUP are running strong second candidates along with their incumbents, though, so a change of personnel is distinctly possible.

What happened: No change.

South Belfast

This is the only constituency with MLAs from five different parties; will that continue?  On the 2019 local election results, Alliance could hope for a gain (most likely from the Greens); on the 2019 Westminster results, the SDLP could say the same. In 2017 the DUP’s second candidate (Emma Little-Pengelly, who was later the local MP from 2017 to 2019) was the runner-up behind her running-mate, Christopher Stalford; this year they are running only one candidate, Edwin Poots, who was briefly the party leader in 2021 and transferred here from Lagan Valley after Stalford’s sudden death in 2022.

What happened: Alliance gained from the Greens.

West Tyrone

Unionists had a generally bad year in 2017, and this is one of the three key constituencies in the West where we’ll see if that was a blip or a trend. If it was a blip, the UUP should have a chance of picking up one of SF’s three seats; if it was a trend, there will be no change.

What happened: No change.

Newry and Armagh

The UUP must have a decent chance at taking the DUP seat. Unionists had a generally bad year in 2017, and this is another of the three key constituencies in the West where we’ll see if that was a blip or a trend. If it was a blip, the UUP will gain from SF rather than from the DUP; if it was a trend, they won’t.

What happened: No change.

Mid Ulster

Again, Unionists had a generally bad year in 2017, and this is the third of the three key constituencies in the West where we’ll see if that was a blip or a trend. If it was a blip, the UUP should have a chance of picking up, either from the SDLP or one of SF’s three seats; if it was a trend, there will be no change.

What happened: No change.

South Down

Only two incumbents are standing for re-election, the lowest of any constituency. Alliance, runners up here last time, will be challenging for a seat; both the SDLP and SF are closer to their second quota than Alliance are to their first, but accidents can happen… The Unionist side is messy too, with both DUP and UUP having had certain local difficulties. There is only one Unionist seat, but it’s not necessarily the DUP’s.

What happened: Alliance gained from the SDLP.


No change is the most likely outcome. The Unionist vote is slowly crumbling here, hovering around a quota, but it would be a very bad year for Unionists to lose it. The UUP are talking up their chances of taking the DUP seat, but I don’t see it on the previous numbers. On the Nationalist side, SF have a stronger starting point, but their internal problems here will not help, and if PBP (or anyone else) were to make an unexpected breakthrough, it would more likely come from SF than the SDLP. The SDLP are running a third candidate, but that is rather optimistic.

What happened: No change.

West Belfast

4 seats out of 5 is an unusually good result in an STV election. SF have a reasonable chance of retaining all four. The Unionist vote is below a quota, and the SDLP far below that.

Overall, I think that Unionists are likely to continue to outnumber Nationalists, but that SF will pass the DUP as the largest single party. If that is the case, will the DUP accept the results and allow a government to be formed? We shall see.

What happened: No change.

Border poll – the precedents

There is much discussion in Northern Ireland – and in the Republic – on the conditions for a referendum on whether or not Northern Ireland should stay in the UK, or become part of a united Ireland. I’ve been fairly clear in my own mind about this for a number of years. I wrote in 2014 that an Assembly election in which Nationalist parties exceed Unionist parties in either votes or seats, or two non-Assembly elections in a row where that happens, would surely be sufficient grounds for the Secretary of State to call a Border Poll.

I’m also fairly clear – and wrote about this in the Irish Times in 2019 – that for the pro-United Ireland side to win such a referendum requires three things to happen: 1) Brexit working out badly; 2) Unionists continuing to talk only to their own core voters and not to the centre ground; 3) Nationalists coming up with a better offer, especially on health services. The first two of these conditions are close to being fulfilled at present; the third, however, is also necessary and we are not there yet.

But there has been much less examination of where such votes have happened previously. Self-determination referendums and plebiscites are not exactly rare in world history. But it’s pretty unusual for the options to be restricted to a choice of which already existing state you want to be part of. Much more often, voters are choosing between independence, on the one hand, and rule by someone else, on the other. I was myself involved in the two most recent independence referendums to have succeeded, in Montenegro in 2006 and in South Sudan in 2011.

Referendums have their advantages and their flaws, and I’m not really going to go into the merits here, just present the historical detail. I’ll note that (of course) they are a pretty blunt instrument, offering little nuance or reassurance for minorities, and that not every one of these historical votes could really be described as having taken place under free and fair circumstances.

Historically I find the following internationally recognised precedents for a popular vote where the electorate were asked about future sovereignty, and independence was not one of the options. There are (arguably) twenty-one of them. In eight cases, voters chose to remain in the country they were currently ruled by. In ten cases, voters chose a change of sovereignty, though in three of those nine cases the will of the voters was not in fact implemented and they stayed where they were. And in the remaining three cases, the territory was split between the two states who wanted to rule it.

1527: Burgundy. The scholar Mats Qvortrup cites this as a very early example of a plebiscite. Under the 1526 Treaty of Madrid, Burgundy was to have been ceded by France to Spain; but King Francis I of France organised a vote of male property owners in Burgundy, who rejected the Treaty, and Burgundy remained French.

1860: Nice and Savoy. Between 1849 and 1870 there were a dozen referendums on self-determination in Italy, as states voted (usually by huge and dubious margins) to join with the new kingdom, effectively merging with Piedmont in the process known as the Risorgimento. Most of those votes do not count for present purposes, as the choice was between continued independence and Italian rule. However, there is one exception: the price for French support of the Risorgimento, under the Treaty of Turin, was the annexation of the town of Nice and province of Savoy, which had until then been under Piedmontese rule. Two referendums in 1860 ratified the transfer.

1868 and 1916, Danish West Indies; 1877, Saint-Barthelemy. A couple of interesting cases in the Caribbean, where on three occasions, islanders voted on which external power they wanted to be ruled by – the Danish West Indies choosing whether to be ruled by Denmark or the United States, and Saint-Barthelemy choosing whether to shift from Swedish to French rule. In all three cases, the referendum was in favour of change, but the US Senate rejected the annexation of the Danish West Indies in 1870, changing its mind almost half a century later; they are now the U.S. Virgin Islands.

1919-22, post-War Europe. The end of the first world war brought a number of new states into being, none of which chose to ratify their independence by referendum. However, there were a number of cases of border adjustments being made by holding a vote in the disputed territories. Only a minority of these votes resulted in a transfer of sovereignty. Two of them were frustrated, both in 1919, when the Vorarlberg province of Austria voted to join Switzerland, and the Åland Islands off the coast of Finland voted to join Sweden, but in both cases, the result was not internationally recognised and they were compelled to remain under Austrian and Finnish rule respectively.

In 1920, there were five such referendums, three of which resulted in votes to stick with the country they had previously been ruled by. So, in February 1920, the northern part of the German province of Schleswig voted to become part of Denmark – the only successful transfer of sovereignty from a single referendum. But a month later, in March 1920, central Schleswig voted to remain in Germany, and the planned vote for southern Schleswig was cancelled. Later that year, the formerly German towns of Eupen and Malmedy voted to join Belgium in a very dodgy process where there was no secret ballot; East Prussia voted to stay in Germany rather than join Poland; the southern zone of Carinthia voted to stay with Austria rather than join the new state of Yugoslavia. In 1921, the district of Sopron voted to stay in Hungary rather than join Austria.

The biggest and messiest of these referendums was the last, held in Upper Silesia in March 1921, in a situation of violence and vote-rigging from both sides. The vote was 60% for Germany and 40% for Poland; the territory in the end was divided, with both sides getting about half of the population, Germany getting more of the land and Poland more of the heavy industry. (It should be added that intimidation and violence were standard features of these referendums.)

1935, Saarland. In a hangover from the First World War, the Saar Basin Territory (now the Saarland), which had been under international rule through the League of Nations, was given a choice between the status quo, joining Germany, or joining France. The German option won more than 90% of the vote, with the status quo a very distant second. So few voters chose France that I hesitate to include it on this list. It’s a rare case of a referendum with more than two options, not that it made much difference in the end.

1947, India/Pakistan. I find only five more internationally recognised referendums in the last hundred years where voters chose between different countries, without independence being on the table. Two of them were parts of the Indian independence process in 1947, with both the North West Frontier Province and the District of Sylhet voting to join Pakistan rather than India. Sylhet was divided, with a small part of it staying in India and the rest now in Bangladesh.

1961, British Cameroons. There have been a number of referenda and plebiscites in Africa, but in almost every case independence has been one of the options on the ballot (including, as mentioned, Southern Sudan, now South Sudan, in 2011). The only exception that I have found was the former territory of the British Cameroons in 1961, in which the population were given the choice between joining the former French colony of Cameroon to the east, or Nigeria to the west. In 1959 they had already voted on whether or not to join Nigeria, and chose not, or at least not yet. In 1961, the Muslim north voted to join Nigeria, and the Christian south to join Cameroon, and that was what in the end happened.

1967 (and 2002), Gibraltar. The 1967 referendum on Gibraltar’s sovereignty clearly satisfies my criteria for inclusion on this list. It was the result of a talks process between Spain and the United Kingdom, and voters were given a choice between integration with Spain or continued British rule. They chose British rule by an overwhelming majority. In 2002 the government of Gibraltar held another referendum, but I don’t think this counts for my purposes: it was a declarative (and again overwhelming) rejection of unpublished proposals for shared sovereignty between the UK and Spain, without any positive option being on the ballot.

1973, Northern Ireland. It is almost fifty years since voters anywhere in the world were given the choice of which country to be part of, without independence being one of the options, and the last such vote was the March 1973 Border Poll in Northern Ireland. On a 59% turnout, 99% of voters supported Northern Ireland remaining part of the United Kingdom, and only 1% voted for Irish unification. I find it interesting that 50-60,000 votes for the Union were cast by people who did not then vote for pro-Union parties in the local council and Assembly elections a few months later.

Next time, the result will certainly be closer.