Friday, August 10, 2012

Esgaroth — what’s in a name?

This is a name I’ve been thinking about for a long time. The bulk of the notes on which I will be drawing for this post are already over a year old, and some of these thoughts are a lot older than that. The name has come up again and again in recent years — in Mark Hooker’s Tolkienian Mathomium, in John Rateliff’s History of The Hobbit, in Parma Eldalamberon 17, in Mark Hooker’s new book, Tolkien and Welsh, and so on — and each time I’ve been prompted to ponder the name a little bit more. With the impending Peter Jackson film adaptation of The Hobbit, the name will be at the tip of everyone’s tongues again soon enough. I think I finally have a gloss I like, but first, let’s review the state of the name.

Esgaroth does not appear in the “Nomenclature” Tolkien prepared for translators of The Lord of the Rings. This is not terribly surprising since the name is associated with The Hobbit, and occurs only a few times in The Lord of the Rings, mainly in connection with the earlier tale. Robert Foster did not attempt to gloss the name in his Complete Guide (1971, rev. 1978). Ditto J.E.A. Tyler in his Tolkien Companion (1976, rev. 2004). Neither Foster nor Tyler even guesses at the language, though it has usually been assumed to be Sindarin. Jim Allan doesn’t have very much more in his Introduction to Elvish (1978) — “[c]alled ‘Lake-town’ in Common Speech, which may be a translation” — though he does commit to identifying the name as Sindarin. In The Languages of Tolkien’s Middle-earth (1974, rev. 1980), Ruth Noel says the name means “Hiding Foam” (Sindarin esgal “hiding” + roth < ros “foam”). David Salo, surprisingly, omits Esgaroth from his Gateway to Sindarin (2004) entirely.

We learned something of Tolkien’s thinking about the name when the Eldarin Etymologies were published as part of The Lost Road (1987). Under a root √ESEK, Tolkien glosses Esgaroth as “Reedlake, because of reed-banks in the west”. Uh, what reed-banks in the west? Actually, this wasn’t merely an afterthought. In one of their songs, the Dwarves recall that “the reeds were rattling”, and the Elves likewise sing that the barrels of Lake-town will go “Past the rushes, past the reeds, / Past the marsh’s waving weeds”. So, although it is never really pointed out, there must be reeds in and around the Long Lake. Okay, moving on.

John Rateliff comments on Esgaroth and its etymology in The History of The Hobbit (2007, rev. 2011). John has the benefit of Tolkien’s explanation in the published Etymologies, which so many earlier thinkers did not, but he wants to reject it. He doesn’t like the fact that the name should apply to the body of water, the Long Lake, but actually applies to the town. He proposes an alternative etymology, again Sindarin: “city standing in or rising up out of the water, perhaps with a suggestion of pilings like reeds”. John’s instinct that the gloss in the published Etymologies isn’t altogether reliable may have some support from an unexpected quarter: Tolkien himself. In the “Addenda and Corrigenda to the Etymologies” (Part Two), published in Vinyar Tengwar No. 46 (July 2004), Tolkien himself has offered up a completely different gloss. Here we have pieces of Tolkien’s etymological thinking about names of Eldarin origin which were omitted from the version published as part of The History of Middle-earth. Under the root √SKAR², Tolkien explicitly glosses esgar as “shore” and esgaroth as “?strand-burg”. The question mark identifies cases where the editors had particular difficulty reading Tolkien’s handwriting. And editors Carl Hostetter and Patrick Wynne note that this is also a “[h]astily written entry not included in the published text”. John does not mention this additional gloss in his book (including the revised edition published last year). Neither does Mark Hooker.

A year before John’s book was published, but two years after the addenda in Vinyar Tengwar, Mark floated — no pun intended — an entirely new theory in his Tolkienian Mathomium (2006). Mark seeks glosses for Tolkien’s words and names from outside Middle-earth, as indeed I often like to do, and as indeed I will do again very shortly. Mark’s view of the name is that it is really of Celtic origin and means something like “an enclosed or guarded encampment on the water” (cf. Celtic elements es, ys, is, etc. “water” + gardd < garthan “enclosed encampment”).

In the summer of 2007, a year after Mark’s book and in the same year as John’s, we got Parma Eldalamberon No. 17, “Words, Phrases and Passages in The Lord of the Rings”. Although it is easy to miss, there is a peculiar reference to Esgaroth in this work. “Galion and Esgaroth are not Sindarin (though perhaps ‘Sindarized’ in shape) or are not recorded in Sindarin” (p. 54). Well indeed! Now, untangling exactly when Tolkien thought what about which root element is a very tedious exercise, and one, moreover, that is likely to remain inconclusive anyway. Nor is it of major importance here. The point I’d like to make is that Tolkien was clearly not sure about Esgaroth. It seems it was one of those words which had sprung up in his imagination without an etymology, for purely phonaesthetic reasons, and which he had some difficulty fitting into the development of his Elvish languages. Instead, he offers two totally different etymologies, plus a statement that it might not even be Sindarin at all! There are plenty of other examples of this elsewhere in the legendarium (e.g., see “The Problem of Ros”, published in The Peoples of Middle-earth).

When Mark Hooker returned to this word in his new book, Tolkien and Welsh (2012), he expanded on his view that the name was a hydronym of Celtic origin. I read that book in draft and commented to Mark on Esgaroth at the time (almost a year ago now). I told him that while I felt he had a reasonable, perhaps defensible Celtic gloss for the word, I still had a nagging feeling about it. Why Celtic, when everything else in the region is Norse? I felt that Celtic names might make sense in Bree and the Shire (in the west), but not in Dale and its environs (in the northeast). Tolkien was very clear, at many points in his notes and essays, that the words and names of the northeast had a Norse character — seen from outside Middle-earth, of course. Just as Dale, Bard, Smaug, and the names of all the Dwarves, plus Gandalf, are obviously Norse in form, why not Esgaroth?

Pursuing this line of reasoning, I have always thought there ought to be a clear Norse gloss for Esgaroth, but developing a theory I could defend has taken a while. Mark replied that most of the Norse words with the sound envelope that he could think of had meanings to do with oak trees. Mark thought this was a long-shot to explain a word he took to be a hydronym. I’d seen the same words myself, and some others, and it has taken some ruminating, but now I think I can share some new ideas. I meant to post this in August 2011, not August 2012, but, well, the days and nights got away from me!

I agree that Norse readings relating to eik “oak” might seem a bit improbable — at first. However, there is the fact that Lake-town is built up on wooden piers, so why not start there? There is also the conspicuous name Oakenshield, taken directly from the Norse eikenskjaldi. And don’t forget that the northeastern part of Mirkwood consisted in large part of oak trees. It was a giant oak that Bilbo climbed when the Dwarves hoisted him up to attempt to determine whether they were any nearer the end of their journey through the dark wood. Though they didn’t know it, they were. The Elvenking carried “a carven staff of oak” too. And the trapdoors out of which the Dwarves and Bilbo escaped the Elvenking’s realm were made of oak. Oak is obviously big in this part of Middle-earth! We aren’t told of what type of wood the piers of Lake-town were fashioned, but why not oak? It’s a good choice, and abundant in the right part of Mirkwood. It could have been pine — there were pine trees in that part of the country as well — but mighty Norse eikr suit the name of the town well.

For the second element, there are some “water words” that pop up in the Norse lexicon — e.g., sker “a rock in the sea, a skerry” or skári “a young sea-mew” — but compounds of any of these really strain credulity. But there is skorða “to prop, support by shores”. Aha! So eik + skorða would mean “to prop up with oak”. The k and s could easily swap spots (metathesis is one of the most common linguistic processes; cf. Old English áscian, ácsian “to ask”). This could give us *Eiskorða, which is very close indeed to Esgaroth. Close enough to satisfy me, at any rate, though if anyone can think of an objection, do let me hear from you.

Another word that might inform the toponym is the Old Norse verb, eiskra “to roar, rage”. This could be a reference to water, perhaps the waterfalls on the edge of the Long Lake, or just as likely a reference to the dragon. Moreover, there is auðr, with two compelling meanings: as a noun, it’s “riches, wealth”; as an adjective, “empty, void, desolate”. The compound *Eiskrauðr might therefore imply “roaring desolation” or “raging riches” or something similar. Naturally, from a point of view inside the history of Middle-earth, Esgaroth wouldn’t have gotten its name because of the dragon, but Tolkien could have bestowed such a name on it from the outside, perhaps unconsciously. There is even an echo in the Noldorin asgar, ascar “violent, rushing, impetuous”.

But this strikes me as not particularly likely. It might just be a secondary echo in this case, albeit a fortuitous one. Given the options, I think the real solution is *Eiskorða, meaning something like “a city propped up on oaken piers”. What do you think?


  1. Well, I think its Secondary World etymology is likely to be Sindarized Silvan Elvish, and that's probably all we'll ever find out about it. An ON form might have suggested this to JRRT in the Primary World, but we don't normally see Sindarin words that are modified ON, or any Primary World language, except very remotely in their roots.

    There are of course occasional breaks in the frame: Saruman is a Mercian OE name, but the internal evidence tells us that the original Westron/Rohirric name, whatever it was, also began with /s/, or Saruman would not have used the S-rune for his troops.

  2. Thanks for the comments, John, as always.

    To your first point, we might reasonably ask why the name of a town of Men should have an Elvish name in the first place. Perhaps Tolkien’s comment in “Words, Phrases and Passages” is on the mark after all: that the name is not Sindarin, even if “Sindarized” in form because of the broader Elvish influence in the region.

    As for Saruman’s s-rune, couldn’t the translator have merely substituted s here for whatever the original rune might have been? :)

  3. Lewis and Currie write in _The Unchartered Realms of Tolkien_, “Furthermore, on p356 H5 the word Esgaroth is said to be derived from a word root ‘Esek’, meaning sedge or reed which comes not from Noldorin as one might expect, but from a more obscure language of Tolkien’s: Ilkorin, the language of Kor, the city of the elves which is developed in ‘Book of Lost Tales’, i.e., this word goes back to some very early conceptions which were pre-‘Silmarillion’ and all that world (i.e. Arda).” I'm not sure what to make of this, nor do I know enough about this topic to refute or substantiate this. Tolkien does make one reference to "Esgaroth" in a letter (14 December 1937), though the letter does not explain the meaning of the word: "I don’t much approve of The Hobbit myself, preferring my own mythology (which is just touched on) with its consistent nomenclature – Elrond, Gondolin, and Esgaroth have escaped out of it – and organized history, to this rabble of Eddaic-named dwarves out of Voluspa, newfangled hobbits and gollums (invented in an idle hour) and Anglo-Saxon runes" (quoted in _Return of the Shadow_ Foreword). According to Tolkien, "Esgaroth" is part of his consistent nomenclature. I'm not sure if this helps or hurts your argument or is irrelevant.
    A couple other thoughts--if it could be determined that oak is a better wood than pine or any other type(s) of wood near Esgaroth for building a dock-like structure, this might help your case. Tolkien did compose two illustrations of Esgaroth. Are there any details in these illustrations to support your case? One minor supporting point is that Tolkien did use the word "oak" etymologically for at least two character names--the minor character Noakes in _LotR_ and the antagonist Nokes in _Smith of Wootton Major_. Also, as Shippey points out, in Tolkien's own private symbolism, the literature side of the English school was represented by the oak (Old English "A" is for ac, the "A-scheme"). All this to say that the oak was important to Tolkien's names.

  4. Very interesting! I had no idea this was such an elusive word. I always learn so much from your blog, Jase!

  5. Well, the towns and cities of the Dúnedain have Elvish names. Granted that they spoke Sindarin and most other Men did not, we might still suppose that Esgaroth was a "learned" name used by Dunedain mapmakers, rather than the name used by the inhabitants themselves.

    As for the S-rune, here's the relevant text from Book 3, Chapter 1:

         Upon [the Orcs'] shields they bore a strange device: a small white hand in the centre of a black field; on the front of their iron helms was set an S-rune, wrought of some white metal.

         "I have not seen these tokens before," said Aragorn. "What do they mean?"

         "S is for Sauron," said Gimli. "That is easy to read."

         "Nay!" said Legolas. "Sauron does not use the Elf-runes."

         "Neither does he use his right name, nor permit it to be spelt or spoken," said Aragorn. "And he does not use white. The Orcs in the service of Barad-dûr use the sign of the Red Eye." He stood for a moment in thought. "S is for Saruman, I guess," he said at length.

    Now we know that Sauron is a Quenya name and unchanged from the Red Book text to the L.R., so the rune must really have been an S-rune, or Gimli's guess would make no sense. Therefore, we also know that in actual Rohirric, Saruman's name must have begun with an S-rune as well.

    The biggest frame-break in the book, as far as I know, is this one from Book 3 Chapter 8: "This was Orthanc, the citadel of Saruman, the name of which had (by design or chance) a twofold meaning; for in the Elvish speech orthanc signifies Mount Fang, but in the language of the Mark of old the Cunning Mind." Now orþanc/orþonc 'cunning, skillful' is an OE word, so our only choice (if indeed this idea appeared in the original Red Book at all) is to suppose that orthanc had essentially the same meaning in OE and actual Rohirric, which would be quite a coincidence. I think the true explanation must be that Tolkien simply lost track of his conceit of translating the Red Book at this point.

  6. John, yes, the comment about Orthanc is the biggest slip of the translator’s veil, but the s-rune could easily be another one. Why do you accept one but not the other? Admittedly it’s easier to fit in a word of a different language starting with the same letter than a word of a different language with the same meaning, but I think both are slips in the conceit. And there are others too. For example, the pages from the Book of Mazarbul that Tolkien created, ostensibly to represent genuine artifacts of the Third Age, are merely transcribed English. Some fans like to argue that Tolkien-the-translator copied the original pages of the Book of Mazarbul, imitating the various scribal hands and reproducing the damage, etc., and translating them from their original language(s) into English, but of course this is silly. Why would he? The real answer is that the translation conceit can only be carried so far, and we sometimes see underneath it.

  7. There are two things going on here, and I think it’s easy to lose sight of which is which. There is the view from inside Middle-earth, where Old Norse is irrelevant, and in fact, does not even exist. Here, the name may be of Ilkorin origin, perhaps subsequently Sindarized, although it is hard to know for sure, since Tolkien has made contradictory statements about the name.

    Then there is the view from outside Middle-earth, where we can wonder whether any real-world languages might have influenced the name. Unless we have direct evidence in the form of a gloss by Tolkien, then we can only theorize. The reason I lay this out in this way is to point out that a statement like this one — “we might still suppose that Esgaroth was a ‘learned’ name used by Dunedain mapmakers, rather than the name used by the inhabitants themselves” — need not preclude the possibility of a Norse influence on the word. The two have little to do with one another. Either or both could be correct; either or both could be incorrect. Just because Tolkien says that Esgaroth “escaped” out of his earlier mythologizing doesn’t mean the name cannot have a Norse inspiration. Consider the name Varda, which I have argued also has a Norse origin.

    I think it’s also important to point out that I don’t pretend to have “solved” the name here. I’ve come up with a possible gloss which seems consistent with sense and setting and with Tolkien’s interests, knowledge, and statements about the region. But it’s hardly the only theory. Speaking of which, Mark Hooker has provided some new feedback on my ideas as well, and he’s given me the go-ahead to edit them into a comment which I’ll be posting here on his behalf when I get the chance. So, look forward to that. :)

  8. Esgaroth < *Eiskorða -- very ingenious!

    Here is another possiblility: given that _Dorwinion_ could have been inspired in part by the word for "wine" which is ultimately of some sort of Mediterranean origin; and given that snails are known to be part of the local fauna in the region of the Lonely Mountain; perhaps some of the wine-merchants of Lake-town developed a taste for eating this delicacy, which might well have led to a reputation in the region, whereby they were known as the "snail-folk." The inspiration for the Sindarized form Esgaroth might be the Provençal _escaragol_, with softening of the velar consonants and incorporating the ethnic ending -(h)oth. Of course the actual reference, since these are lake-dwellers, might be to a less outlandish sort of "shellfish"; still it is the echo of later French _escargot_ 'snail' that is probably most suggested to an English reader by the sound of Esgaroth.


  9. Hmm... This is an interesting idea, but I'm a little sceptical, for a couple of reasons. On the idea of it being Old Norse, the beginning sequence esg- is so reminiscent of Welsh ysg- that the Celtic inspirational etymologies seem more attractive.

    Another reason is that I'm not sure what would be going on here. Metathesis is certainly a common linguistic process, but why would Tolkien have made a Norse compound, added a bit of metathesis, and then further obscured the word (writing -sg- for -sk-, etc.)? He has enough Norse words in TH, that if it was Norse, I'd expect it to appear as such, not with this odd distortion - unless the word just came into his head as was, rather than being an elaborately 'Celticized' Norse word.

    The third thing, and this is an issue I've got with Mark Hooker's idea too, is the postulating of origins that have final -rth- for -roth-. Like the -sg- for -sk- thing, why would Tolkien have changed about the ending like this? Did he have some reason to hid his etymology for this word (in contrast with, say, Smaug or Gandalf)? This is like the second objection, but also applies to the Celtic etymologies.

    I think you do give the best explanation for the word: 'It seems it was one of those words which had sprung up in his imagination without an etymology, for purely phonaesthetic reasons, and which he had some difficulty fitting into the development of his Elvish languages.' Only I'd say it not only doesn't fit into Elvish, but into any other style either. It's a name that popped into his head, with a bit of story following along after. The rhythm of the word, as well as the -sg- combination, point to Welsh as the phonological frame from which the word was born, at least to me, but I don't see why we should expect any etymology at all (unless it happens to have a very precise match, in which case sure).

    Still a very interesting post, and I think this is a valiant attempt to get an etymology out of Esgaroth - but I'm going to stay a sceptic here :)

  10. Also, the Celtic associations (on a phonological rather than etymological level) are reinforce for me by the probable link of Celtic lake-towns in archaeological works and Esgaroth (as made in the Annotated Hobbit). This would be one reason to explain a very non-Norse word in association with that locale.

    The Bree connection might actually not be so far off. I think the Annotated Hobbit (I don't have it with me right now) points out that the Laketown scenes do a great job, with the remnants of the older town visible, of creating a sense of a modern mercantile town against a backdrop of something more ancient. A 'queer' and vaguely Celtic name like Esgaroth against the modern and intelligible Lake-town. An idea of this sort might have been in Tolkien's mind when he suggested an Elvish, but non-Noldorin/Sindarin name, giving the name a rather primeval pedigree.

  11. I love this kind of word-games — regardless of any reference the the question of whether there is any objective "truth" to the idea that this in some way influenced Tolkien :-)

    The reference to Esgaroth escaping out of the consistent nomenclature of the Silmarillion material in the December 1937 letter (cited by Josh above) is, I think, puzzling. I know of no reference to Esgaroth prior to The Hobbit. "Esgaroth" enters the text in what Rateliff identifies as the 'second phase', and I don't know when it enters the pictures, as I cannot find a dating for the relevant pictures (particularly nos. 65, 79 and 88 in The Art of the Hobbit by J.R.R. Tolkien).

    IF there is no earlier use of the word, the letter would seem to imply that Esgaroth should have a consistent etymology in the elven tongues of the time, but I have yet to see a satisfying etymology based on elements in existence prior to the Etymologies (late 1937 - early 1938). However, it would seem that Tolkien, at least about the time he wrote The Etymologies, did think of the name as Elvish.

    Without thinking it through very carefully, my instinct is to favour a purely phonetic mode of inspiration, unless it be from one of the real world languages that inspired the phonetics of an Elvish tongue (as Finnish did Qenya) . . . (but here I also think that I am going beyond my rather limited knowledge of Tolkienian linguistics).

  12. "David Salo, surprisingly, omits Esgaroth from his Gateway to Sindarin (2004) entirely."

    Not *entirely* -- on p. 339 I wrote "In the Third Age the names Amroth, Esgaroth [et al. ...] may be of the same sort, of Silvan (Nandorin) origin."

  13. David, I missed that! Thanks for pointing it out. And thanks to everyone for the interesting comments and feedback. As always, much appreciated. Nice to know one isn’t just hollering down a well. :)

  14. It seems to be Celtic, with perhaps a little influence from Norse:

    Esgaroth < *Eskaroth < *Iscaroth or *Iscaràth (or other variant), “water-fort” or “-settlement”, from some variant of Celtic *iscā > e.g., Old Irish esc “water” > Irish easc, Gaelic easg, “bog, marsh” (N.B., some variant of *iscā is a proposed origin for the name of the River Esk in Norse-settled Cumberland) + some variant of Celtic *rātis > e.g., Old-Middle Irish ráth, Gaelic ràth, Manx raath, Proto-Welsh *rath, Pritonic *roth, “fort”, “enclosure”, “settlement”.

    For more information about *iscā and *rātis, see the Scottish Place-Name Society’s website here and here respectively.

  15. Interesting discussion. While Tolkien was inspired by Welsh to invent the elvish language, many of the names for things come from Old English. I have a proposal that the first element 'es' may be a collapsed dipthong for 'aesc' or 'ash'. The oak was important to the Germanic peoples but it was Thor's tree. The ash was Odin's tree and also was Yggdrasil (Ygg's horse), the World Tree that help up middle earth. The second element 'gar' would then be 'spear' and the suffix '-oth' is problematic. Is it the Hebrew '-oth' or the OE -aþ or -oþ? The Hebrew word is for a feminine noun and, since 'gar is masculine and Mitchell Robinson says it is used for masculine nouns, I will go with the OE suffix. However Mitchell Robinson also states that the suffix is used to form masculine nouns from verbs and since 'gar' is definitely not a verb, it is not definite but I think it means 'inhabitants of' or 'realm of' and makes sense when you look at a map and see how Esgaroth looks like the head of a spear with the River Running as the shaft. So Realm of the Ash Spear which may be a kenning for something else. I hate kennings. Hats off to those who can figure them out.