bjarnasonr
New Member
I'm a catch-up listener and have only made it up to episode 133 at the moment, so if this is aged / subsequently revisited, I'm sorry! I'm refraining from joining any live discussion until I catch all the way up, so I have no idea what has been discussed since. I'll get there soon, and I'm looking forward to joining y'all!
I've just listened through the discussion of the metre of Boromir's prophecy poem (seek for the sword that was broken...), and I'm finding myself quite at odds with Corey's conclusions! Insisting that Isildur should be pronounced with the stress on the first syllable is tantamount to asking me to accept a poem wherein the word America is made of two trochees (i.e. (ˌame)(ˈrica) or something, using IPA stress notation). You cannot simply disregard the stress rules of a language without ending up with unnatural nonsense!† This is all the more an issue when the line containing Isildur works perfectly well without any restressing. As I understand it, while the prototypical iambic foot is an unstressed syllable and then a stressed syllable (xˈx), it can also be not only a single stressed syllable (as we see many times in this poem) but also two unstressed and then a stressed. In fact, the very first line requires both of these kinds of feet:
(ˈSeek) (for the ˈsword) (that was ˈbro)ken
(IIRC there's traditionally a special name for an (xxˈx) foot, but I'm a linguist, and in linguistics all we have are iambs and trochees - we just let them have somewhat flexible composition, with the exact rules on a per-language basis.)
Thus there is absolutely zero need to restress Isildur - the first foot is a normal two-unstressed-one-stressed iambic foot:
(for Iˈsil)(dur's ˈBane) (shall ˈwa)ken
It might be a bit odd to start a line with one of these (I'm not sure how often it happens in English poetry), but even if you disallow (xxˈx) iambs line-initially, you can just treat the first syllable as extrametrical just like the last syllable is in half these lines (a technical status that gets at the idea of a 'grace note' some commenters were mentioning): for (Iˈsil)(dur's ˈBane) (shall ˈwa)ken. Yes, you lose the parallelism with the other odd-numbered lines (starting with a stressed syllable out of the gate), but 1) that's a small price to pay for avoiding having to pronounce Isildur in a way that makes it not a phonologically valid Sindarin word, and 2) it now is parallel with the last line in a way that groups the last pair of lines off separately. This parallelism works pretty well with the meaning: the first pair is a command, the middle two pairs are a description of what happens at the place the dreamer is commanded to go, and the last pair gives the non-dream signs that accompany the prophecy as additional proof of its validity.
The line with Imladris, though, is not solvable just by using an (xxˈx) foot; we end up with a perfectly valid sequence of iambic feet, but only two of them:
*(in Imˈlad)(ris it ˈdwells)
I don't have a surefire solution to this problem, but I do have a possible solution (independently suggested by this page and argued for in more detail there; I'll summarise here). Tolkien's descriptions of Sindarin stress imply that Sindarin syllabifies VCCV sequences as VC.CV regardless of what consonants are involved, and thus VCCV(C)# sequences (where # is a word boundary) should be uniformly stressed on the penultimate syllable as ˈVC.CV(C)#. However, Sindarin stress is pretty much an exact copy of Latin in every other respect, and Latin doesn't treat all VCCV sequences identically: if the first C is a stop and the second C is one of /l r/, you get V.CCV instead, and the first syllable is light. This means that words like tenebrae 'darkness' are stressed as ˈtenebrae, since the middle syllable is light (te.ne.brae, not *te.neb.rae). If we assume Sindarin does the same thing, it resolves the entire problem: (in ˈIm)(laˌdris) (it ˈdwells).
Apparently this happens elsewhere in Tolkien; Gimli's poem in Moria has the line (of ˈmigh)(ty ˈkings) (in ˈNar}(goˌthrond), which requires the same altered syllabification rule (though we have to expand the list of C(l/r) sequences that can be syllabified this way). I'd be curious to hear if there's examples where this syllabification rule breaks a line in a poem (i.e. if Tolkien hadn't decided on including it or not), or if this is just a rule that Sindarin always follows but Tolkien forgot to specify in the appendices. (Thangorodrim is the one case that page raises where the rule doesn't seem to apply; but it involves the collective plural -rim, so the morpheme break might mess with the syllabification rules. That page also mentions silivren, but that word could be disallowed by positing that /θr/ is the only valid onset cluster that isn't a stop-l/r sequence: note that /θr/ is valid word-initially as well (e.g. in Thranduil), but e.g. silivren's /vr/ isn't.)
To summarise, rather than concluding that this is a 'Gondorian / Westron-influenced non-native pronunciation of Sindarin' where the stress gets moved to the initial syllable, we have two (separate) quite reasonable solutions for the apparent metrical problems in Boromir's poem. First, we take Sindarin stress to be even more an exact copy of Latin's, and syllabify Imladris as ˈIm.la.ˌdris rather than *Im.ˈlad.ris - a rule we maybe should apply in general in Sindarin, not just in this poem. Second, we render the line with Isildur as the perfectly valid (for Iˈsil)(dur's ˈBane) (shall ˈwa)ken. Those interpretations together lead to a perfectly normal and regular metre:
(ˈSeek) (for the ˈSword) (that was ˈbro)ken:
(In ˈIm)(laˌdris) (it ˈdwells);
(ˈThere) (shall be ˈcoun)(sels ˈta)ken
(ˈStron)(ger than ˈMor)(gul-ˌspells).
(ˈThere) (shall be ˈshown) (a ˈto)ken
(That ˈDoom) (is ˈnear) (at ˈhand),
(For Iˈsil)(dur's ˈBane) (shall ˈwa)ken,
(And the ˈHalf)(ling ˈforth) (shall ˈstand).
†I think there are a few languages in the world that have stress assignment rules but stress isn't a big deal, and so it doesn't sound as bad to apply stress in unexpected places. (I don't remember any examples off the top of my head.) I don't think we have any reason to believe this is true of Sindarin, though; certainly Sindarin otherwise copies Latin stress exactly, and I've never heard of any cases in Latin poetry where metre just outright overwrites word stress.
I've just listened through the discussion of the metre of Boromir's prophecy poem (seek for the sword that was broken...), and I'm finding myself quite at odds with Corey's conclusions! Insisting that Isildur should be pronounced with the stress on the first syllable is tantamount to asking me to accept a poem wherein the word America is made of two trochees (i.e. (ˌame)(ˈrica) or something, using IPA stress notation). You cannot simply disregard the stress rules of a language without ending up with unnatural nonsense!† This is all the more an issue when the line containing Isildur works perfectly well without any restressing. As I understand it, while the prototypical iambic foot is an unstressed syllable and then a stressed syllable (xˈx), it can also be not only a single stressed syllable (as we see many times in this poem) but also two unstressed and then a stressed. In fact, the very first line requires both of these kinds of feet:
(ˈSeek) (for the ˈsword) (that was ˈbro)ken
(IIRC there's traditionally a special name for an (xxˈx) foot, but I'm a linguist, and in linguistics all we have are iambs and trochees - we just let them have somewhat flexible composition, with the exact rules on a per-language basis.)
Thus there is absolutely zero need to restress Isildur - the first foot is a normal two-unstressed-one-stressed iambic foot:
(for Iˈsil)(dur's ˈBane) (shall ˈwa)ken
It might be a bit odd to start a line with one of these (I'm not sure how often it happens in English poetry), but even if you disallow (xxˈx) iambs line-initially, you can just treat the first syllable as extrametrical just like the last syllable is in half these lines (a technical status that gets at the idea of a 'grace note' some commenters were mentioning): for (Iˈsil)(dur's ˈBane) (shall ˈwa)ken. Yes, you lose the parallelism with the other odd-numbered lines (starting with a stressed syllable out of the gate), but 1) that's a small price to pay for avoiding having to pronounce Isildur in a way that makes it not a phonologically valid Sindarin word, and 2) it now is parallel with the last line in a way that groups the last pair of lines off separately. This parallelism works pretty well with the meaning: the first pair is a command, the middle two pairs are a description of what happens at the place the dreamer is commanded to go, and the last pair gives the non-dream signs that accompany the prophecy as additional proof of its validity.
The line with Imladris, though, is not solvable just by using an (xxˈx) foot; we end up with a perfectly valid sequence of iambic feet, but only two of them:
*(in Imˈlad)(ris it ˈdwells)
I don't have a surefire solution to this problem, but I do have a possible solution (independently suggested by this page and argued for in more detail there; I'll summarise here). Tolkien's descriptions of Sindarin stress imply that Sindarin syllabifies VCCV sequences as VC.CV regardless of what consonants are involved, and thus VCCV(C)# sequences (where # is a word boundary) should be uniformly stressed on the penultimate syllable as ˈVC.CV(C)#. However, Sindarin stress is pretty much an exact copy of Latin in every other respect, and Latin doesn't treat all VCCV sequences identically: if the first C is a stop and the second C is one of /l r/, you get V.CCV instead, and the first syllable is light. This means that words like tenebrae 'darkness' are stressed as ˈtenebrae, since the middle syllable is light (te.ne.brae, not *te.neb.rae). If we assume Sindarin does the same thing, it resolves the entire problem: (in ˈIm)(laˌdris) (it ˈdwells).
Apparently this happens elsewhere in Tolkien; Gimli's poem in Moria has the line (of ˈmigh)(ty ˈkings) (in ˈNar}(goˌthrond), which requires the same altered syllabification rule (though we have to expand the list of C(l/r) sequences that can be syllabified this way). I'd be curious to hear if there's examples where this syllabification rule breaks a line in a poem (i.e. if Tolkien hadn't decided on including it or not), or if this is just a rule that Sindarin always follows but Tolkien forgot to specify in the appendices. (Thangorodrim is the one case that page raises where the rule doesn't seem to apply; but it involves the collective plural -rim, so the morpheme break might mess with the syllabification rules. That page also mentions silivren, but that word could be disallowed by positing that /θr/ is the only valid onset cluster that isn't a stop-l/r sequence: note that /θr/ is valid word-initially as well (e.g. in Thranduil), but e.g. silivren's /vr/ isn't.)
To summarise, rather than concluding that this is a 'Gondorian / Westron-influenced non-native pronunciation of Sindarin' where the stress gets moved to the initial syllable, we have two (separate) quite reasonable solutions for the apparent metrical problems in Boromir's poem. First, we take Sindarin stress to be even more an exact copy of Latin's, and syllabify Imladris as ˈIm.la.ˌdris rather than *Im.ˈlad.ris - a rule we maybe should apply in general in Sindarin, not just in this poem. Second, we render the line with Isildur as the perfectly valid (for Iˈsil)(dur's ˈBane) (shall ˈwa)ken. Those interpretations together lead to a perfectly normal and regular metre:
(ˈSeek) (for the ˈSword) (that was ˈbro)ken:
(In ˈIm)(laˌdris) (it ˈdwells);
(ˈThere) (shall be ˈcoun)(sels ˈta)ken
(ˈStron)(ger than ˈMor)(gul-ˌspells).
(ˈThere) (shall be ˈshown) (a ˈto)ken
(That ˈDoom) (is ˈnear) (at ˈhand),
(For Iˈsil)(dur's ˈBane) (shall ˈwa)ken,
(And the ˈHalf)(ling ˈforth) (shall ˈstand).
†I think there are a few languages in the world that have stress assignment rules but stress isn't a big deal, and so it doesn't sound as bad to apply stress in unexpected places. (I don't remember any examples off the top of my head.) I don't think we have any reason to believe this is true of Sindarin, though; certainly Sindarin otherwise copies Latin stress exactly, and I've never heard of any cases in Latin poetry where metre just outright overwrites word stress.
Last edited: