Cold Iron & Rowan-Wood

May 30, 2009

Place names and a sense of history

Filed under: essay — Tags: , , , , , , , — Sam @ 2:27 pm

Reading Rush-That-Speaks’ livejournal post about MammothFail, I finally codified one of the principal issues I have with a great deal of (particularly American) fantasy, and why I instinctively class it as “fluff” or “not serious” in comparison to other examples.

There’s no sense of history, or of change. The names are all instantly legible – Oaktown, Kingswood, or Greywood, for instance. And I’ve heard Americans asserting that this makes them “sound English”. The thing is, though, that in Britain that’s a marker of newness, not of antiquity – if a place has a name that any English speaker can instantly understand, it’s not been around for very long at all.

The three examples I cited are all places in Britain, but in translation – Acton, for instance, the town in the oaks. Coed-y-Brenin, near where I grew up in Gwynedd, is Welsh – it translates as “the King’s wood”. Lytchett, in Dorset, and Llwydcoed near Aberdare both mean “grey wood”.

Names tend to stay the same, or at least the same at their root, while languages change around them. The River Avon, for instance – afon is the Welsh word for “river”, and in Irish & Scots Gaelic it’s abhainn, so what that means is that some dim Anglo-Saxon came along, said “‘ere, whatcha call that thing?”, the Celt he asked said “‘s a river, innit mate”, and the Anglo-Saxon put it down on his map as the River River.

Sometimes, though, two almost-parallel terms can survive alongside each other. For instance, the Welsh names for a lot of towns & cities begin with Caer (as in Cair Paravel – but pronounced more like “kyre”) and the English versions will usually end in -caster, -cester, or -chester. Chester itself is referred to on Welsh maps as Caer, and Gloucester is Caerloyw (“shining fortress”). But the two words, caer and castrum, aren’t from the same place at all – the Welsh just means an enclosed place, more or less the same as the hay component in southwest English placenames, while the English term is from Latin military terminology.

Actual castles in Wales (most of which were built by the English as instruments of subjugation) get referred to as Castell – Castell Harlech in Snowdonia, for instance.

“Snowdonia”, of course, is another example of the same linguistic layering and obfuscation. Any English speaker will vaguely recognise that the -ia suffix means “place of” or “around that sort of general area”, but “Snowdon” is the Saxon name for the highest mountain, meaning “Snow hill”. And in Welsh it’s Yr Wyddfa (though I don’t know the etymology) while the area is Eryri. It’s tempting to think that that means “eyrie” (since eryr means “eagle”), but it’s more likely just “highlands”. Of course, this isn’t just English nationalism (though that plays a large role) – Welsh place names are notoriously difficult for the English anyone else to get right.

Which name you use for a place can be highly politicised, too – mention in the wrong pub that you’re thinking of a trip to Derry, or to Londonderry, and you may well be In Trouble.

Tolkien, unsurprisingly, is very good on this. Fornost Erain became Norbury of the Kings, and Amon Sul became Weathertop, while the Tower of the Sun became the Tower of Guard.

Robert Jordan has instances of interestingness, too – Mafal Dadaranell became Fal Dara, and Al’cair’rahienallen became Cairhien. Of course, since we learn this from the Ent expy, it’s an obvious homage to Treebeard’s comment that the Land of the Valley of Singing Gold has become the Dreamflower, but there’s nothing wrong with that.

Juliet McKenna’s Einarinn books have a couple of instances of the same thing – Kel’Ar’Ayen (the new continent) becomes Kellarin over time. Though, oddly, there’s no sign of anything similar happening to the original continent of Tren’Ar’Dryen, and the name just falls out of use.

On the other side, we have David Eddings (yes, yes, cheap target, I know). In the world of the Belgariad, almost all countries have uniform naming schemes. The capital of Tolnedra is Tol Honeth, and the other cities are all Tol Something; the capitals of Arendia are Vo Mimbre, Vo Astur (ruined) and Vo Wacune (ruined and genocided). Everything in Gar og Nadrak starts with Yar, and everything in Cthol Murgos with Rak. Of course, there’s an in-universe explanation for this, in that the Gods really did just dump people down into a wide-open uninhabited land, but again that’s an in-universe explanation. We don’t see it except from characters in the narrative, so we’re entitled to treat it with Suspicion… especially considering that marginal savage demon-worshipping peoples survive in the icy or jungle-covered parts nobody else wants. They even wear feathers.

Raking through the shelf of books I might want to read again someday, but probably not, I found an even better example – Jane Lindskold’s Through Wolf’s Eyes. Flipping to the front of the guidebook for the map, I see New Kelvin and Dragon’s Breath by the Sword of Kelvin mountains. The White Water River runs down to the sea at Port Haven, passing by Stilled, Gateway to Enchantment, Plum Orchard, and (oddly) Zodara. Scattered across the rest of the map, we see Eagle’s Nest Castle, Rock Fort (by Broadview), Revelation Point Castle, and Good Crossing.

This is clearly a colonialist land, though we can hold out some hope for Zodara. Flipping through it – since I haven’t a clue what it’s like after so long – I see kings, queens, Grand Duchesses, both “societies” and noble houses named after animals, but no mention of where the colonists come from (except a tantalizing note at the top of the obligatory genealogical chart full of Adjectivenoun Names that some dates are in the “Gildcrest Colonial Calendar”) and no mention of any indigenous population. Not even any fairy mounds.

Seriously, this makes Eddings look good.

2 Comments »

  1. Not the main thrust of the essay, I know, but the example you cite for linguistic survival in place names, the British name for the river Avon, is actually the exception rather than the rule. ‘Celtic’ placenames are pretty rare in England, though obviously far more common in Wales, Ireland and Scotland. The frequency of names with p-Celtic elements increases rather as you move further North and there seem to be a few clusters in the North-East, possibly associated with monastic settlements (recent scholarship is making claims for a few more: http://www.mun.ca/mst/heroicage/issues/10/fox.html), but it’s still a fairly small number.

    Comment by Karen — June 17, 2009 @ 1:14 pm

  2. I agree, but it’s still a pretty prominent exception, and I think that it still makes my case – even after the speakers receded, we still see linguistic nunatuks like those. I did have a couple of papers lying around that I was looking at before writing this, but annoyingly I seem to have lost them now.

    Comment by Sam — June 17, 2009 @ 2:03 pm

RSS feed for comments on this post. TrackBack URL

Leave a comment

Powered by WordPress