Tolkien Studies

The Silmarillion is perhaps the linguistically most refined work of J.R.R. Tolkien. Polished for a lifetime, it is not surprising that it is written in a most remarkable and memorable of styles. In fact it has more than one style (as it is more than one text). Several distinct styles can be found in the variants of the Silmarillion tradition (now available in the volumes of The History of Middle-earth), which David Bratman distinguishes as the Annalistic, Antique, and the Appendical (71-75). But in the published Silmarillion, one has the feeling the categorization which Bratman suggests for the contents of the History volumes does not fit perfectly: styles change within units of the text, and the three categories seem somewhat loose and vague anyway. The movements of style and the resulting disunity in the 1977 text produce a fitting effect: Tolkien succeeds in implying, merely by the stylistic differences, that the Silmarillion is indeed a compendious volume, "made long afterwards from sources of great diversity" (S 8). Taking into account that it is in fact an editorial text, selected and made consistent from the numerous versions, according to (with some remarkable exceptions, I believe to its advantage) ¹ the latest intentions, by Christopher Tolkien, its style definitely signals how truly compendious it is—it suggests a history for the text, an evolution, in which the cryptic and compressed narrative of the 1926 "Sketch of the Mythology" or the Quenta Noldorinwa (written in the 1930s) became expanded to the majestic story (and language) in the Silmarillion.

Part of the fiction (and the point) of the Silmarillion is, however, that inside the textual world it is not a unified text either, but a compendium, a collection of texts. As such, it surely has a history there too; the different versions of the presentation frame (from the Lost Tales to the "latest intentions") hint that Tolkien imagined it to be a sort of comprehensive manuscript of a (narrative or historical) tradition.² There are a number of conclusions to be drawn from viewing the Silmarillion text so (of which I hope to make a more extensive study in the future); in this paper I will examine one of the aspects which bears closely on both the stylistic refinement and its implications, and on the history of the texts. It is clear that we are meant to view the Silmarillion thus, and in the manuscript analogue its being an editorial text diminishes in problematic value.

The text of the 1977 Silmarillion as we now have it includes and preserves many traditions—that is what its compendious nature means. But the curious duplication of the text (the supposition that it is, just as it [End Page 21] stands, a text inside the textual world too) makes this actually a double claim. The text has a history in "primary philology" (= Tolkien philology, as texts by J. R. R. Tolkien), and another one within, for "secondary philology" (as I elsewhere called this level).³ The two provenances may partly be parallel; but the differences in style in the 1977 Silmarillion text do more than suggest history and leave it at that. They also suggest different things for primary and secondary philology; and the difference is significant and critically meaningful.

Doubtless many readers have noticed that amidst the surges of style in the Silmarillion (from the elevated "mythological prose" of the Ainulindalë to the drier "descriptive prose" of, say, "On Beleriand and Its Realms") there are passages, short strings of sentences, individual sentences, or even single clauses which read as if they were poetry adapted to prose. In primary philology, this feeling is sometimes justified when we look up the variants and sources in the history of the text—but only in the stories which Tolkien wrote in verse, and the "adapted passages" are found much more widely than that. In secondary philology, however (up to a certain level parallel to this, since the verse works are also "duplicated texts," the "adapted sources"), the case is more complex. The "adapted" passages do not indicate "lost Tolkien texts": they indicate poetic works in the textual world (a number of which are mentioned but never written).⁴ In the manuscript context and the provenance of texts, this is in no way unusual: verse adapted to prose (and vice versa) is frequently found in medieval manuscripts and is equally easy to pick out. Malory's "Tale of King Arthur and the Emperor Lucius" betrays its source, the alliterative Morte Arthure in a similar way. Had this vanished (as it nearly has, except for the single remaining manuscript), Malory's text would be an indication that such a poetic work had once existed; it would point to it, even though the alliterative Morte was not extant. The adapted texts in the Silmarillion also indicate poetic tradition in the textual world, both deepening the breadth of cultural implications in the text (and enriching the world it creates) and offering us fragments of the actual text of these lost poetic works.

Like philologists writing on the lost poetic sources of an extant prose text, we will have to take a textual approach to be able to determine the significance of this phenomenon in the Silmarillion. We will have to consider the textual features that make a passage "adapted": its syntactic and rhetorical structures, rhythm, and euphony devices; these will have to stand out in marked contrast to the context. It is only then that we can go on to the interpretation of the implications, the suggestions about the poetic traditions and the texts themselves which become meaningful when integrated into the whole system of the Silmarillion. There is a well-marked contrast between the styles of the Ainulindalë and most of the [End Page 22] Quenta Silmarillion proper; our task will be to examine when the stylistic differences within one text lend support to a theory of "adaptedness," and further, what this theory means in the interpretation of the whole.

Many of the passages that stand out from the stylistic register of their context do so by virtue of their rhetorics, their stricter syntactic patterns. This is in some cases underlined by the specific use of rhythm, alliteration, and rhymes. Let us look at some examples:⁵

1. and they built lands and Melkor destroyed them;
   valleys they delved and Melkor raised them up;
   mountains they carved and Melkor threw themdown;
   seas they hollowed and Melkor spilled them;
   and nought might have peace or come to lasting growth,
   for as surely as the Valar began a labour so would Melkor undo it
           or corrupt it.
   (S 22)
2. Then he looked upon their glory and their bliss,
   and envy was in his heart;
   he looked upon the Children of Ilúvatar that sat at the feet of the
           Mighty,
   and hatred filled him;
   he looked upon the wealth of bright gems,
   and he lusted for them;
   but he hid his thoughts, and postponed his vengeance.
   (S 65)
3. and they came to the Enchanted Isles
   and escaped their enchantment;
   and they came into the Shadowy Seas
   and passed their shadows,
   and they looked upon Tol Eressëa the Lonely Isle,
   but tarried not;
   and at the last they cast anchor in the Bay of Eldamar.
   (S 248)

One might argue that these are not necessarily adapted from poetry; they simply show a conscious use of the syntactic structure of parataxis and balanced clauses—their authors were good rhetoricians at any rate. They all share a repeated pattern: two clauses connected with a simple conjunctive "and" (in boldface; occasionally with "but"), the first and second clauses bearing a structural and thematic similarity to each other (in italics; e.g., in example 1: "they built lands," "valleys they delved," "mountains they carved," and "seas they hollowed," where even the inversion adds a further stylistic overtone to the parallel); a clever utilization of polysyndeton and parallel syntactical structures. Further devices are to be observed in other examples:

4. and the House of Fëanor hastened before them along the [End Page 23] coast of Elendë:
   not once did they turn their eyes back to Tírion on the green hill of Túna.
   . . .
   but at the rear went Finarfin and Finrod, and many of the noblest and wisest of the Noldor;
   and often they looked behind them to see their fair city.
   (S 85)
5. I would not have any say that Túrin was driven forth unjustly into the wild,
   and gladly would I welcome him back;
   for I loved him well.
   . . .
   I will seek Túrin until I find him,
   and I will bring him back to Menegroth, if I can;
   for I love him also.
   (S 200)

In these, parallel structures are not confined to clauses only. Example 4 uses antithesis in parallel structures (very appropriate for the contrast between the attitudes of the different Noldorin houses); while in 5 (from a dialogue of Thingol and Beleg) we see tripartite parallel sentences (except for Beleg's short conditional clause), concluding on the same thematic note (and nearly the same words). The structures in examples 1-5 are all syntactically grounded stylistic devices, making use of parallel clauses, parataxis, and repetition.

Other traits which might indicate stylistic differences from the context are more specifically poetic devices. The following passages will be sufficient to demonstrate them:

6. But now upon the mountain-top dark Ungoliant lay;
   and she made a ladder of woven ropes and cast it down,
   and Melkor climbed upon it
   and came to that high place,
   and stood beside her, looking down upon the Guarded Realm.
   (S 74)
7. Then Finrod was filled with wonder
   at the strength and majesty of Menegroth,
   its treasuries and armouries
   and its [many-pillared halls of stone];
   and it came into his heart
   that he would build wide halls
   behind ever-guarded gates
   in some deep and secret place
   beneath the hills.
   (S 114) [End Page 24]
8. The light of the drawing of the swords of the Noldor
   was like a fire in a field of reeds.
   (S 191)

Example 6 uses paratactic structures very similar to those seen in 1-3, while also occasionally taking up alliteration (boldface) and starting with a line whose rhythm definitely stands out from the context.⁶ Example 7 makes much more of alliteration, and further introduces rhymes (italicized; both internal, as in line 3, and end-rhyme, as in lines 7-8). The whole of the passage is strongly rhythmical (the last line can be a shorter "coda," a closure to the "stanza"; and line 4 is in fact a verse line from one of Tolkien's other poems⁷ ). Finally, example 8 again exhibits a very interesting regularity of beat. This is partly again syntactical, since the pattern of multiple genitives determines the first line, but it goes on to the second line, where the alliteration underscores the effect. It is surprising how well these passages sound (and even scan, especially example 8) when read out loud; but that still does not make them "adapted."

Research into the primary history of these passages shows that they are definitely not "adapted" from verse (with the exception of line 4 of example 7; see n. 7 above). They come from the "prose Silmarillion tradition," and are nearly all present in comparable form in some earlier version (except example 1, for which I have found no source,⁸ and example 5, which emerged in the more or less independent "Túrin tradition"⁹ ). They evolved sometimes suddenly,¹⁰ sometimes by slow steps of refinement,¹¹ sometimes obviously by editorial action.¹² Slow shifts of structure and wording produced these texts (occasionally with the discarding of versions which would perhaps have done better¹³ ), and a detailed collation can reveal much about how Tolkien reformulated his sentences and worked step by step in shaping the language of his text. In fact, the text had really become a "tradition," where with time the work incorporated many layers and changes that are preserved or discarded according to the needs of the actual version worked on.

The secondary history, however, is more suggestive. These passages, as the final result of the painstaking stylistic development, stand out from their context by their marked language use: in the textual world, they can be indicative of conventions of style in certain narrative situations. All the passages mark central scenes, climaxes, or privileged points in the narrative; their author was obviously aware of how such climaxes and centers should be handled stylistically. As examples 1-5 show, certain rhetorical and syntactical structures were held to be appropriate; the noted affinity of "high material" to parataxis is definitely borne out in these parts. Thus, on the one hand, these examples of the adapted texts can give us an insight into the ways stylistic conventions function in the minds of fictitious authors in the textual world (which at the same time implicitly argues that the Silmarillion constructs numerous such authors [End Page 25] and author roles—again, as obviously part of its "meaning"); but on the other hand, the contrast between these examples and their contexts can suggest conclusions about the transmission of the texts. It is equally possible that these passages stand out because of a certain "stylistic leveling," carried out by the scribes or adapters (or perhaps editors) transmitting the texts. Marked style in central scenes and climactic parts is always more likely to be preserved in redaction than in cases where the redactor does not sense the scene to be central or important; and while authors can easily be supposed to be conscious of stylistic conventions, scribes and redactors perhaps cannot. The implications of these parts, in terms of the secondary history of the texts, thus yield conclusions both about the origin of the text (its author and its conventional context) and its provenance, both its production and reception.

One could reply that the examples are still not necessarily poetry adapted to prose; they are simply "poetic prose," which is not quite the same thing. The style of the Silmarillion, of course, is generally poetic anyway, except for the stretches of "descriptive narrative" (interpolations in a thematically compiled narrative manuscript?); it merely rises higher sometimes, as sometimes it descends lower. But even to judge the whole text "poetic," and suppose that the passages cited are "more poetic" (still not amounting to "poetry proper") is to evaluate by distinct criteria of style—in other words, to suggest poetic qualities a text has to stand in a context of poetic conventions, and it is exactly this that I am arguing the adapted texts indicate. They do not have to be actually adapted from verse (as most of them are not, and we would not know anyway if they were in the textual world); the point is that they indicate conventions. In another example, we can easily observe how the devices seen above sometimes appear in such density that the reader can hardly avoid the conclusion of adaptedness:

9. and even as the Noldor set foot upon the strand
   their cries were taken up into the hills and multiplied,
   so that a clamour as of countless mighty voices
   filled all the coasts of the North;
   and the noise of the burning of the ships at Losgar
   went down the winds of the sea as a tumult of great wrath
   and far away [all who heard that sound were filled with wonder].
   (S 106)

Here we have all: parataxis (though somewhat looser), alliterative patterns bridging the lines, the rhythm of the genitives (their quick pace even suggesting the crackling of the fire?) and of the final line (the part of which I enclosed in square brackets is incidentally a perfect blank verse line). Tolkien once started an alliterative poem on "The Flight of the [End Page 26] Noldoli" (given in Lays), though it never reached the actual crossing; one is tempted to speculate that this short fragment might have come from its textual world counterpart.

Further conclusions are possible about the implied poetic conventions from the evidence of the adapted texts. In the primary history of the texts, these passages sometimes really find a foundation in a real poetic tradition: that of Tolkien's own poetic texts. The cases in point are those when the passage in question is not from a story that is treated in those poetic works, but just a short textual bit that agrees with the poetic corpus (sometimes even against the prose). One such case is "he piled the thunderous towers of Thangorodrim" (S 118). This occurs in the Quenta Silmarillion (§105; late 1930s); but before that, the prose tradition invariably spoke only of the "towers of Thangorodrim," never supplemented with the alliterating epithet. The "thunderous towers," however, can be seen in more than one place in both the verse Túrin (ll. 714 and 951) and the Lay of Leithian (ll. 2051 and 3281).¹⁴ Both poems preceded the writing of the Quenta in the late 1930s, and I think it is evident that the epithet came to the prose tradition from the poetic one—a corroborated case of adaptation where the fragment of poetry embedded in the prose points justifiably to the poetic use. Another such detail is the curious imagery which accompanies Lúthien's remaining at the gates of Menegroth during the Hunting of the Wolf: "[a] dark shadow fell upon her and it seemed to her that the sun had sickened and turned black" (S 185). An irregular (because of the alliteration in the second half-line) alliterative line might be hiding in this (if it is, it is unattested); but at least the image of the "sickening sun" is paralleled in the verse Túrin,¹⁵ again pointing to a use of the image in the poetic tradition. These details suggest, firstly, that a tradition does exist. Epithets are apparently used in it;¹⁶ alliteration is a structural as well as a euphonic device (otherwise alliterating phrases would not travel together).

If such indications are supported by the material of primary philology, there is no reason to disregard the implications in secondary philology. Perhaps the most easily accessible part of the implied poetic convention is its forms; we have already seen that alliteration and rhyme, certain rhythms (iambic?) are among the formal features (see Wynne and Hostetter 116, 118-20, 122). Some other examples will show what else we can recover. Another phrase belonging here is "wild and wary as a beast" (S 165), which appears in the prose Túrin tradition (UT, 110), but has a parallel in the Lay of Leithian.¹⁷ "[A] dark lord upon a dark throne in the North" (S 205) recalls the Ring Poem's repeated formula for Sauron, while

10. and far and wide in Beleriand the whisper went,
    under wood and over stream
    and through the passes of the hills.
    (S 205) [End Page 27]

seems to point to the similar structures in Bilbo's first "poem" in The Hobbit (chapter 19, 359-60 and see below n. 7) and is marked by a characteristic rhythm. Two other instances of adapted verse lines are "and waited while the long years lengthened" (S 44) and "and none were safe in field or wild" (S 195): both exhibit easily scanning rhythm and rather conventional-looking phrasing, in the first case underscored by the use of alliteration and the figura etymologica. Alliteration and rhythm are beautifully seen together in

11. But there was a deep way under the mountains
    delved in the darkness of the world
    by the waters that flowed out
    to join the streams of Sirion.
    (S 125)

Finally, in two further examples rhythm and alliteration work closely together to produce a most remarkable effect:

12. Wisdom was in the words of the Elven-king,
    and the heart grew wiser that hearkened to him.
    (S 140-41)
13. Little foresight could there be
    [for those who dared to take so dark a road].
    (S 84)

These pleasantly scanning lines (again, a faultless blank verse line is found in example 13) reinforce our conclusions about formal characteristics in the implied poetic traditions. Many of these conclusions are in fact corroborated by what we know about Elvish poetic modes (Wynne and Hostetter);¹⁸ Tolkien (and his editor, Christopher Tolkien) here builds into his text parts which not only create cultural practices as mere facts or frameworks but also give some of their content.

We have already encountered cases where this reference to a poetic tradition was (in a general way) supported; in some longer stretches and numerous smaller examples adaptation from verse is a fact in the primary history of the texts. These examples, in stories which had been treated by Tolkien in poetic form, allow a useful glimpse of the process of what Tolkien does when he is really adapting from his own verse. Not only do these instances show the actual passage of the text from poem to prose; in the secondary layer, they also suggest cultural practices which integrate into the contexts seen in the Silmarillion, practices that seem to appear as "narrated," described. Most of such actually adapted texts come from the verse Túrin, but the Lay of Leithian is also a source. Only a few of them longer than a phrase or a few lines, they corroborate the theory of poetic style and its conventions inside the textual world (like "wild and wary as a beast," already quoted, or "guard him and guide him" [S 209], a favorite phrase in the Túrin story¹⁹ ). [End Page 28]

A relatively large number of lines go back explicitly to the verse Túrin. In one certainly adapted line, "bearing a burden heavier than their bonds" (S 208)²⁰ we can actually see how the adapter straightens out the syntax of the line (and makes a perfectly regular clause out of exactly the same phrases—which is still betrayed by its rhythm and alliterations). Other such lines are "that grief was graven on the face of Túrin and never faded" (S 208),²¹ and "he walked as one without wish or purpose" (S 209).²²

Lines whose origin I cannot clearly establish also fall into the pattern, and look very much like the actually adapted texts in their poetic devices. "[H]e listened to his lore and the tale of his life" (S 204), "in the dim dusk of the winter's day" (S 204), "in a mirror mishapen by malice" (S 214), and the couplet "she fled as in a madness of fear, / swifter than a deer" (S 219) are all lines where the beat, alliterations, and rhymes strongly suggest a poetic source. The verse Túrin never advanced to the later part of the story where these are found; but the lines do not appear to my knowledge in the Unfinished Tales 'Narn' or the Quentas either.

There are also lines which derive from the prose texts, with conclusions very similar to what we have already seen. "[S]et a doom upon them of darkness and sorrow" (S 197), although it has parallels in the verse Túrin,²³ is really closer to the wording of the Quenta Silmarillion (ch. 16 §22), as is "during that time his grief grew less" (S 199)²⁴ to Quenta Silmarillion ch. 17 §34.²⁵ Saeros's taunt with its firm three-line structure also comes ultimately from there (ch. 17 §39 = UT 80), while the line "he was senseless in a sleep of great weariness" (S 207) from the Quenta Noldorinwa (§12, with minor variation). The dying Glaurung's slandering of Túrin to Nienor (S 223), with its perceivable line structure and alliterations, has a source in the Unfinished Tales "Narn" (138), and Túrin's plea to his sword to "slay me swiftly" (S 225) also derives from there (UT 145).²⁶ These fragments of the poetic tradition testify for the formal features we have already deduced, and fill the bare suggestions of conventional poetic forms with content—here we have scraps from the poetic handling of the Túrin story, both primary and secondary, both extant and lost.

Secondary philology is all the more central here, because the Túrin part explicitly claims a poetic source, the Narn i Hîn Húrin. Elsewhere I have examined the critical importance of this suggestion in detail (see my "Great Chain"); what is significant in the present context is that such source references, coupled with lines adapted from poetry (both in fact, as can be shown by collation, and in fiction, as in the cases where no primary poetic source can be found but formal features of the texts in question place them in close connection with corroborated instances) form a special set within the corpus of the adapted texts. Perhaps a source reference in the vicinity or thematic sphere of such texts can be [End Page 29] used as complementary evidence of the text's integrity with the poetic tradition (as in the case of the fragment from the Noldor poem [example 9, perhaps also 8]—the Noldolantë, mentioned earlier [S 98], could then be equated with the source of this fragment); such instances would be of especial importance in the study of the implied poetic conventions and practices.

Further grounding is available in this class of instances for the interaction of prose and verse traditions, the stylistic conventions for central/climactic scenes, and I believe that even something about the compositional principles and methods, some of the implied cultural context of the poetry can be recovered. The first such important passage from the Túrin story is:

14. Then Túrin stood stone still and silent, staring
    on that dreadful death, knowing what he had done. (S 208)

The image itself is part of the prose tradition;²⁷ but that, in turn, and much of the actual wording of the text as well, goes back to the verse Túrin (ll. 1273-74):

stone-faced he stood       standing frozen
on that dreadful death       his deed knowing

Nearly all the alliterating words, together with the alliteration pattern itself, doubtless derive from the poem; the imagery and to some extent the very phrasing of this very moving central scene traveled between the versions virtually unchanged. One is, then, tempted to see an analogue in the following passage:

15. tall and terrible on that day looked Túrin,
    and the heart of the host was upheld
    as he rode on the right hand of Orodreth.
    (S 212)

Though I have found no source, either in the prose or in the verse traditions, for these lines (the verse Túrin never reached this point in the story), the parallels with the previous example strongly suggest to me a similar evolution: style and the formal devices of the poetic convention are equally well preserved in this other (though minor) climactic scene.

Perhaps the best example for the processes of adaptation, also indicative of compositional methods, is seen in the scene of Fingolfin's duel with Morgoth. This is recounted in the Lay of Leithian, prompted by the description of Anfauglith and the gates of Angband (ll. 3478-3634); otherwise it is not related to the story of the Geste. The very fact that it is inserted points to a compositional method which is characterized by its "situationalism": mention of a place, or name, or event, can set the author at any point to present other stories or episodes which have a [End Page 30] connection to the "cue," even if they do not advance the action, fragment it, or impede upon its forward thrust. This is a feature reminiscent of the methods of oral composition; the oral poet's control on the store of formulae and narrative "chunks" he has in his head is fundamentally different from the relationship of his literate colleague to his materials in being much more determined by his mnemotechnics and mechanisms of preservation (Havelock 175). Parallels can be found in Homer, for example: the catalogue of ships in Book 2 of the Iliad grows out of Helen's showing the Greek leaders to the Trojans from the walls. This is a link backwards, and at once a historical contextualizing; but it surely does not advance the plot, nor is it a necessary requisite (Havelock 177-79). There are similar "situational" anticipations and recountings in the Silmarillion (most notably in "Of Thingol and Melian," S 55). The oral poet has to repeat, tell the stories again and again, in order to keep them known and remembered, and to remember them himself (Havelock, 91-93; see further 145-64); this is why he goes into them as soon as they are "cued." C. S. Lewis's commentary on the Lay of Leithian in one place picks out something similar (when Thingol's minstrels recall the stories of Fëanor and the Silmarils, lines 1132-61) as "expanded by the late redactors who found their audiences sometimes very ignorant of the myths" (Lays 391). This is partly the same case (though this passage is more of a digression, modeled on those in Beowulf, than a "situational episode" launched by a "cue" in the narrative), since both later interpolation and original narrative aim at the preservation of stories. The difference is again in the authorial vs. redactorial layers, which thus are clearly visible (at least to Lewis) in the poetic work itself. But as author and redactor in this case stand in two readily distinguishable cultural contexts, orality and textuality as the source of texts are brought up for critical consideration. The passage from the Silmarillion that can be counted as adapted runs as follows:

16. for the rocks rang with the shrill music of Fingolfin's horn,
    and his voice came keen and clear
    down into the depths of Angband;
    . . .
    Therefore Morgoth came, climbing slowly
    from his subterranean throne,
    and the rumour of his feet was like thunder underground.
(S 153)

The passage, as example 14 above, is extant in the prose tradition,²⁸ but very close similarities exist between it and a part in the Lay of Leithian (lines 3545-47 and 3558-62):

while endless fastnesses of stone [End Page 31]
engulfed the thin clear ringing keen
of silver horn on baldric green.
. . .
Then Morgoth came. For the last time
in these great wars he dared to climb
from subterranean throne profound
the rumour of his feet a sound
of rumbling earthquake underground.

On the one hand, we see again that this tragic climax keeps its poetic form in the prose redaction; and on the other, we again get a glimpse of the actual poetic texts which might serve as sources in other places in the narrative. The actually adapted texts do more than imply an underlying tradition in verse: they in fact preserve it, and had Christopher Tolkien not decided to publish The Lays of Beleriand, these instances would be the only traces of it left (compare the case of Malory and the nearly vanished Morte Arthure). It is not surprising, in textual-world terms, that there are so many adapted lines and passages in this part of the Silmarillion—these are culturally central stories in the textual world ("the human-stories of the elves," as Tom Shippey applied Tolkien's phrase for them). We have here further suggestions about the cultural use of the poetic tradition in the textual world, to create heroic narrative poems of these central stories. It is then perfectly natural that prose adaptations from these high-prestige poetic compositions stay closer to the texts of the poems: again, not only origin but also transmission details are implied.

I have saved for the end three examples which could easily have fitted elsewhere, but which I consider to be most representative and suggestive, worthy of individual scrutiny. We have seen in the primary history that the adapted texts were either derived from the prose tradition, reached by small steps of refinement from early versions sometimes not at all outstanding (indeed often not very easily distinguishable); or were derived from the poetic works; or appeared all of a sudden in one or other version. The following examples are, I assume, no exceptions, though for one of them I have not been able to find any comparable parallels. These are also the most interesting cases of the adapted text, in terms of secondary philology: beautiful and perfectly crafted lines of great style and poetry, tight structure, a very high standard of refinement. Great poetry is implied to be behind them.

The first of these is a part of the Quenta Silmarillion proper, Ch. 1, that to my knowledge has no extant parallels; its use of nearly all the devices we have found to belong to the implied poetic tradition places it with the adapted texts. [End Page 32]

16. Green things fell sick and rotted,
    and rivers were choked in weeds and slime,
    and fens were made, rank and poisonous,
    the breeding place of flies;
    and forests grew dark and perilous,
    the haunts of fear;
    and beasts became monsters of horn and ivory
    and dyed the earth with blood.
(S 36)

The paratactic and repetitive structure at once draws attention to this passage. Made up of three two-line units, each telling about the corruption of a certain sphere, plus an initial 1+1 pair, its parallels are thematic as well as structural. The rhythm of the lines is broken twice with the shorter fourth and sixth lines, while the assonance of "slime" and "flies," and the fuller rhymes "poisonous" and "perilous" (italicized), strengthened by the alliteration, keep the composition together. Alliteration further links (very appropriately) "fens" and "flies," "forests" and "fear." This "stanza' is highly reminiscent of the style of the Ainulindalë, both formally (as seen in example 1) and thematically, and suggests a lost poetic tradition of the cosmogony, in fragments and style preserved in the prose redaction.²⁹ And this is, I believe, unmistakably poetry.

Another gem of stylistic polishing has a long textual history, but it is no less outstanding and suggestive for that. It is in the Quenta Silmarillion proper, Ch. 7, and appeared for the first time in comparable form in the Quenta Noldorinwa (§3), then went through the Quenta Silmarillion (§46) and both versions of the "Later Quenta" (§49b); but it was longer and less concentrated. About Fëanor's work, we are told:

17. Then he began a long and secret labour,
    and he summoned all his lore, and his power, and his subtle skill;
    and at the end of all he made the Silmarils.
(S 67)

The longer and earlier versions of these three lines had none of this crystallized compactness, welded together by the rhythm (both of the paratactic structure and of the enlisting of Fëanor's "tools": lore, power, and skill) and the pattern of alliteration. It also did not have the effective closure of the last iambic line (in this case not a blank verse line, since it is one foot longer than that), nor the assonance of the last couplet. The alliteration pattern took long to establish: up to the "Later Quenta," Fëanor's labour was said to be "long and marvellous," even though the verb "summon," which ultimately, I believe, influenced the insertion of "secret," had been present since the Quenta Silmarillion version. Here again we can see in the history a change (in the end discarded) which was for the worse. While the third "tool" on the list had always been "subtle"("subtle magic" in the Quenta Noldorinwa, "skill" in the Quenta Silmarillion), it was [End Page 33] changed to "subtle craft" in the "Later Quenta," but then evidently back again to fit the pattern—I assume authorially, since the versions known to me require an unjustifiable amount of editorial change to produce the reading quoted. What secondary history this fragment comes from we cannot tell; yet it points surely toward a poetic text in the textual world.

The last and most complex of the adapted texts I would like to draw attention to is also in the Quenta Silmarillion proper, in the "interpolated" descriptive chapter "Of the Sindar." The style of this chapter is generally eclectic (it would probably be classified to belong either to the Annalistic or the Appendical style), due to its "compilation nature"; it tells of many things,³⁰ and one has the feeling it is heavily compressed.³¹ But at one point, as if in summary, the following three sentences are inserted:

21. In Beleriand
    in those days
    the Elves walked,
    and the rivers flowed,
    and the stars shone,
    and the night-flowers gave out their scents;
    and the beauty of Melian was as the noon,
    and the beauty of Lúthien was as the dawn in spring.
    
    In Beleriand
    King Thingol upon his throne
    was as the lords of the Maiar,
    whose power is at rest,
    whose joy is as an air
    that they breathe in all their days,
    whose thought flows out in a tide
    untroubled from the heights to the depths.
    
    In Beleriand
    still at times
    rode Oromë the great,
    passing like a wind over the mountains,
    and the sound of his horn came down the leagues of the starlight,
    and the Elves feared him for the splendour of his countenance
    and the great noise of the onrush of Nahar;
    but when the Valaróma echoed in the hills,
    they knew well that all evil things were fled far away.
(S 95)

This is already present in the "Grey Annals," with two variant readings.³² The three long sentences, each starting with "In Beleriand" and each describing a different aspect of one-time Beleriand, are truly remarkable [End Page 34] for their tight but graceful structure, touching imagery, and majestic style. Though they exhibit no use of rhythm or alliteration comparable to previous examples, I have classed them as definitely adapted text, since the very conscious use of syntactical structures itself creates a rhythm which is genuinely poetic. In addition to the "In Beleriand" starting, shared by all "stanzas," two of them (the first and the third) also have time clauses at the start, in the same position as second line. It is these two stanzas that make more use of parataxis, thus "framing" the second stanza which is exclusively hypotactic. The three stanzas are unified by their imagery and their use of comparisons and similes: the pattern is given in the first stanza, after a four-line quick-paced setting which supplies the tone of the imagery as well. Images of nature, more and more complex, dominate the stanzas. They do not only become more complex linguistically (with more and more adverbials in the subordinate clauses) but also in conception: by the time we reach the second stanza, the similes of Thingol and the Maiar decidedly go into abstractions, leading on to the description of Oromë, very appropriately. The Oromë stanza, by the way, is less distinctively stylized than the rest; one is tempted to suspect an impatient fictitious redactor who did not clearly understand the purpose of the natural images and was not very well-versed in theology. The third stanza, as C. S. Lewis would have put it, is perhaps corrupt; but the beauty of the original can still be seen, since the original design is discoverable. The "In Beleriand" stanzas are a striking instance of what differences style can produce even within a single text—which itself is interpolated as "different" into the narrative thrust of the "tale of the Noldor." They illustrate perfectly that these differences are anything but mere ornament. They are there for specific reasons; either because the author or the editor put them there, or let them stand there. But both ways, they point to something prior to the work of the author and the editor. They imply texts and forms, conventions and traditions, which stand behind any text in a compendium; and, perhaps more importantly and most clearly perceivably in these stanzas, they sometimes supply the actual words that stand in the background.

Tolkien's texts work in a variety of ways to produce depth behind themselves. This feeling of depth can be illusory, or it can be real: I hope to have shown that at least in some instances in the Silmarillion, the "poetic depth" created by the adapted texts is very real. For, returning to the earlier counterpoint, even if we say that these are not "adapted poetic texts" but simply "poetic prose," we have presupposed poetic style and poetic convention already. Like Old English "rhythmical prose," which is very hard to differentiate from Old English "alliterative verse proper," the adapted texts cannot be proclaimed non-poetic and their poetic suggestions denied. If we did not know alliterative verse, [End Page 35] its patterns and beats, we would never be able to detect "rhythmical prose": we judge poetic prose in terms of (and in forms of) poetry, and this finally shows that poetry is the reference point. It is a fact of cultural history that narratives are composed first in verse (which offers better mnemotechnical opportunities) and only then in prose: Tolkien's text and Tolkien's world follow this rule.

In striving for verisimilitude and authenticity, Tolkien apparently repeats cultural history. One cannot write a mythology, primarily because myths are not written; what is great about Tolkien is that he manages to write not only texts but traditions. He goes even further: he supplies the background of his narratives with poetic traditions which are not there—but the very supposition uncovering this fact is based on pieces which are there, actual fragments from fictitious poetic traditions. This congenial device makes use of the painstaking stylistic refinement, and again shows up how important textual transmission is to the interpretation of Tolkien—indeed, how very crucial textuality is in Tolkien's mythopoesis.

In terms of primary interpretation, this is significant and is perfectly integrated to the system of the Silmarillion. This work is not only about telling stories that go with other stories (like the Lord of the Rings or the Hobbit): it is about the story of stories, both in a historical and a metafictional sense. Tolkien shows us how narratives are preserved; yet not only narratives are his concern but also language, the actual words that tell the tale. The preservation of style together with matter is a well-known phenomenon, as is the editor's and redactor's leveling of style. The Silmarillion discusses how stories come to be told in exactly these words: either the author (origin) or the editor/redactor (transmission) is conscious of the stylistic conventions. Both ways, the point is the existence and content of the conventions; Tolkien manages to have it both ways, and say something both about the nature of the poetic narrative sources (the cultural contexts, contents, and use) and the implied manuscript context (transmission). The Silmarillion, exactly as it stands in the 1977 text, is a profound work: an anatomy of story.

I said earlier that Tolkien's texts have subtle ways to create depth behind themselves, and I have examined in detail one of these ways; but it has in this inquiry, I hope, become clear that Tolkien has even subtler ways to fill this depth. The Silmarillion text, being a compilation of traditions and an editorial text, both in the primary and the textual worlds, works very much like an actual manuscript, holding in itself traces not only of the traditions that went into its making, but very often of the actual texts. This is no lost poetry of Tolkien, however; this is Tolkien's prose, paradoxically, one might say, giving us a glimpse of the lost poetry of Beleriand. [End Page 36]