In Peter Jackson’s films, there are numerous scenes of first Bilbo, then Frodo, hunched over and scribbling into a thick, leather-bound book. These scenes express an important idea from both The Hobbit and The Lord of the Rings: that these texts were authored not only by Hobbits but by specific Hobbits. Their point of view is evident in reading the books, which would become very different if authored by, say, Galadriel or Boromir.

But what about The Silmarillion? Who authored that book? This question is less easily answered, yes, but it can be answered, and just as the point of view undeniably shapes The Hobbit and The Lord of the Rings, so do the narrators of The Silmarillion provide a perspective that, once acknowledged, deepens and enriches the text.

The Wise: Rúmil and Pengolodh

Through most of the history of the Silmarillion, there were two narrators: Rúmil of Tirion and Pengolodh of Gondolin. In terms of textual history, Rúmil is the oldest, first emerging as early as 1918 in The Book of Lost Tales as the loremaster who recited the creation myth The Music of the Ainur to the mortal seafarer Eriol. As Christopher Tolkien remarks, poetically, in his commentary on this text: “Even after Eriol … had fallen away, Rúmil remained, the great Noldorin sage of Tirion … and The Music of the Ainur continued to be ascribed to him, though invested with the gravity of a remote time he moved far away from the garrulous and whimsical philologist of Kortirion” (45). Throughout the textual history of the Silmarillion, Rúmil was assigned the material concerning Aman, with the assumption that he did not follow the Noldorin exiles to Middle-earth, leaving a historiographic lacuna that would be filled by Pengolodh.

Also Noldorin, Pengolodh was born in Middle-earth, in Nevrast, and spent most of the First Age sequestered in the hidden city of Gondolin, according to a brief biography he is given as part of the 1960ish text Quendi and Eldar (War 396). To his credit are the chapters that transpire in Beleriand, as well as the Akallabêth and a few philological texts. Pengolodh’s isolation makes him an interesting if problematic choice as the narrator for the bulk of the Silmarillion because one wonders how comprehensive a history is possible for a historian to whom, according to The Silmarillion, “tidings of the lands beyond [Gondolin] came … faint and far” (288). Pengolodh first emerged in Tolkien’s writings slightly before 1930. Tolkien would continue attributing texts to him for the next three decades.

It’s important to emphasize that Rúmil and Pengolodh are not mere adornments on the text, simply names appended to create an illusion of historicity. As Tolkien worked sporadically on the Silmarillion across his lifetime, he constantly framed and reframed the role of each narrator in the text. Initially, for example, he assigned the entirety of The Earliest Annals of Valinor to Pengolodh. Later, he revised this text to assign the pre-exilic history to Rúmil. A third revision was undertaken to reassign some of the writings given to Rúmil back to Pengolodh because, if Rúmil did not venture to Middle-earth, he could not have known that material. In other instances, Tolkien adds parenthetical asides to Rúmil’s material from Pengolodh of material that Rúmil could not have known, and in one text, allows Pengolodh to contradict Rúmil’s account outright. This ongoing negotiation of attributions shows that not only did Tolkien think deeply about his narrators—who they were, what they would have known, what they could not have known—but that the narrators mattered. It is significant that, in the course of making a mythology that would become the basis of one of the 20th century’s most important literary works, Tolkien dithered over which imaginary loremaster wrote which imagined bit of history.

And this dithering, taken to its extreme, would create complications as Christopher Tolkien posthumously assembled the Silmarillion his father never completed. Here, an important, if oft-forgotten, peculiarity of The Silmarillion becomes key: The published Silmarillion is not entirely Tolkien’s. Not only unfinished when he died, it was left in messy, contradictory, and at times completely incomprehensible form, to be fashioned into a coherent, publishable whole by Christopher Tolkien with the assistance of the fantasy author Guy Kay. The state of his father’s work left Christopher with several significant decisions to make that would shape the text we would come to know as The Silmarillion. One of the most important, I would argue, concerned the narrators.

A "Devastating Surgery": Transplanting a Númenórean Narrator?

For, as anyone knows who has delved into the posthumous texts beyond The Silmarillion, to dither, to meddle, to rewrite and change was Tolkien’s normal state of being. And toward the end of his life, he exceeded even his own prodigious capacity for revision, beginning to contemplate bone-deep changes to the body of myth underlying the legendarium. The best known of these was his rejection of a flat-earth mythology and much of the cosmology that formed the foundation of the legendarium. I won’t claim that his emendations concerning the narrators would have been as significant as changing the actual universe, but they are close, and they are almost entirely overlooked. Christopher Tolkien considered them, of course, calling the question of narrators “a fundamental problem” (Morgoth’s 205), but even as he decided to ignore his father’s late cosmological revisions, he accepted revisions to the narrator that were made around the same time, despite what I believe is significant evidence to the contrary.

On a scrap of paper that Christopher dates no earlier than 1958, Tolkien wrote of the narrators: "It is now clear to me that in any case the Mythology must actually be a 'Mannish' affair. … What we have in the Silmarillion etc. are traditions … handed on by Men in Númenor and later in Middle-earth" (Morgoth's 370, emphasis in the original). He repeats this assertion again, in another late 1950s text. In less than one hundred words, Rúmil and Pengolodh, who had been in existence for more than three decades in the case of Pengolodh and more than four in the case of Rúmil, were seemingly unwritten.

This is why Rúmil and Pengolodh are the most important characters you’ve probably never heard of. Rúmil is mentioned twice in the published Silmarillion, both times referring to his alphabet, the Sarati, that Fëanor improved in making the Tengwar. Pengolodh is not mentioned at all. Had Tolkien not written those two notes in the late ‘50s, it is very likely that your Silmarillion would look different, at least in the sense that the narrators would be present in the text.

Christopher Tolkien, however, took his father at his word that the Númenóreans were the authors of The Silmarillion and removed the attributions to Rúmil and Pengolodh familiar to anyone who has worked with the early drafts of the Silmarillion. He later conceded this choice as “somewhat excessively pursued” (Peoples 142). Regardless, the result is a Silmarillion where the point of view gives an illusion of omniscience and that lacks the human grounding, present in The Hobbit and The Lord of the Rings, that a discernible narrator provides. As I will show, The Silmarillion is far from omniscient, but first, I want to briefly address Tolkien’s assertion of a “Mannish” narrator.

As I believe Christopher Tolkien came to accept, there is little textual evidence that Tolkien undertook revisions toward a Númenórean narrator. In one late version of the Silmarillion dated around the same time as the scrap of paper that unwrote Rúmil and Pengolodh, Tolkien stripped mentions of Pengolodh from the text. This proves it, right? Why else would he remove Pengolodh from the text? There is, however, what I can only describe as a textual tangle, all dated to 1958: There is the note that unwrote Pengolodh. There is The Later Quenta Silmarillion II, where Tolkien removed all mention of Pengolodh. Then there is The Annals of Aman, where he also started to remove Pengolodh, only to change his mind and add him back. The order in which these three texts were produced is unknown, and having a chronology would perhaps shed light on Pengolodh’s mysterious disappearing/reappearing act. Regardless, there is no evidence to definitively suggest that Tolkien in 1958 was poised to begin revisions to remove the Noldorin and introduce Númenórean narrators.

Commenting on his father’s experiments with a revised cosmology, Christopher wrote that “the old structure was too comprehensive, too interlocked in all its parts, indeed its roots too deep, to withstand such a devastating surgery” (Morgoth’s 383). I believe that this conclusion applies equally to the matter of the narrators and that Tolkien likewise realized it. However, when the workings of the universe itself are at stake, consideration of the person telling the story can easily fall by the wayside. Certainly, looking at the Later Quenta Silmarillion II, the text where Tolkien removed attribution to Pengolodh, there is no evidence of a change in narrator; in fact, it is possible to argue the opposite. This text contains the Statute of Finwë and Míriel, a text heavily eschatological in nature. Given the Númenóreans’ perilous fascination with death, it is well-nigh impossible to imagine a Númenórean text concerned with Elven eschatology that doesn’t take the opportunity to comment on Mortal eschatology (or at least contain some juicy marginalia ranting about the unfairness of it all). Furthermore, the same text does contain a section written by a Mortal narrator—Laws and Customs among the Eldar, attributed to the mortal man Ælfwinë—that is clearly from a Mortal point of view, casting no doubt on the fact that the rest of the Later Quenta Silmarillion II is from an Elven point of view. As with the cosmology, the roots originating from Rúmil and Pengolodh as narrator run deep. Revision would not be a surface endeavor but would likewise be a “devastating surgery” that would radically change the text.

Given this, I will proceed under the assumption that the Silmarillion is not only Noldorin but specifically authored by Rúmil and Pengolodh. And, as I noted earlier, these characters are not merely appended to the text; rather, reading the text with the narrators in mind reveals that Tolkien did write the Silmarillion from the distinct, if subtle, points of view of Rúmil and Pengolodh. This is evident in the bias toward and against different places and characters. It shows up subtly in the formulaic language the narrators use to report the history of different character groups. And finally, the narrators are revealed in the different ways they view the exiled Noldor, specifically the Fëanorians.

Bias in The Silmarillion

Two years ago at this conference, I spoke about historical bias in The Silmarillion. That year’s paper and presentation is available on my website, so I will be brief now. Historical bias in The Silmarillion was first observed in the scholarship by Alex Lewis in a 1992 presentation at the Tolkien Centenary Conference. When I joined the Tolkien fandom in the mid-2000s, bias was frequently observed by Silmarillion fans and a jumping-off point for fanfiction and other fanworks about The Silmarillion. This is how I became interested in the question and sought to quantitatively document what, if any, historical bias is present in the text.

As I noted earlier, Pengolodh was born in Middle-earth and was among the early settlers of Gondolin, a realm known for—even defined by—its isolation from the rest of Beleriand (War 396). Turgon, Gondolin’s king, felt understandable antipathy toward Fëanor and his sons, blaming them for his wife’s death while crossing the Grinding Ice into Beleriand (Peoples 345). This anti-Fëanorian bias is reflected in Pengolodh’s chapters in The Silmarillion. Along the same lines, Pengolodh’s work reflects a favorable assessment of the realms of Doriath, Nargothrond, and of course, Gondolin.

The graph above shows the number of words spent describing each of the realms of Beleriand in The Silmarillion. A clear preference for Gondolin, Doriath, and Nargothrond is evident and would fit a character with Pengolodh’s backstory. He would have been able to personally describe Gondolin and had access to sources that could describe Doriath and Nargothrond, which is not the case for most other realms in the book. That Nevrast is the fourth most mentioned realm is also important: It is where Pengolodh was born and, again, would have been one of the few places he was able to describe from experience. The data for characters, shown to the right, most often mentioned in the text looks similar. Again, except for Fëanor and Maedhros, all of the characters in the top twenty have some affiliation with Gondolin, Doriath, or Nargothrond. The same trend appears again when looking at battles: The battles involving characters from those three realms are described using more words than battles involving other characters, so much so that some of the battles that were key ground gained by the Noldor against Morgoth are not only barely described but aren’t even named. (Hence the Not-Great Battle and Battle 4.5, both of which were victories, even though Pengolodh takes pains to identify the former as “not reckoned among the great battles” in the same breath as he admits “thereafter there was peace for many years, and no open assault from Angband” [Silmarillion 133].)

This data, shown above, suggests that Tolkien wrote the latter half of the Quenta Silmarillion from the point of view of a character whose sources came mostly from Gondolin, Doriath, and Nargothrond. This has the effect of sidelining characters who weren’t from those areas but whose roles likely would have been larger than The Silmarillion itself suggests. However, Pengolodh’s point of view does more than simply spotlight some characters over others. He also shows a bias against certain characters. The Not-Great Battle is an example of this. Fingon is an unexpected target of Pengolodh’s disregard. Although Turgon’s brother, Fingon is also a close ally of the hated Fëanorians, and he can’t catch a break. He leads three key victories, all of them brushed past in the text or, as noted earlier, deliberately downplayed.

In the past two years, I’ve undertaken two other case studies of Pengolodh’s bias. The first compared the death scenes of Fëanor and Fingolfin. They are remarkably parallel. There is an emotional provocation that causes each character to seek single-handed combat with Morgoth. Powered by passion, they almost win—but don’t. There is a grievous turn in the enemy’s favor, followed by a spectacular death: Fëanor turning to ash, consumed by the power of his own spirit, and Fingolfin being crushed down three times but popping up again each time like an epic game of Whack-a-Mole. The difference in these two scenes is how Pengolodh depicts the characters: Fingolfin’s irrational, suicidal act is cast as heroic, almost godlike, elevated by comparisons to the Valar. Fëanor, on the other hand, is driven by vengeance and irrationality, his courage tainted by a whiff of blame for running too far ahead of the rest of his host. The description of Fingolfin’s fight scintillates on the page, glittering with references to light, and emphasizes the size of Morgoth compared to Fingolfin. Fëanor’s fight, in comparison, is muted, lacking in such grand, heroic language. Finally, the overwhelming grief at Fingolfin’s death contrasts with the complete lack of reaction from Fëanor’s sons and followers when Fëanor dies, reminding us that, while Fingolfin was loved, the Fëanorians were emotionally blank. These scenes are key because, being single combat and thus unwitnessed, Pengolodh had a choice of how to depict each character, and his choices reflect his biases. This is his chance to sermonize a bit, and he takes it.

The second case study considers Caranthir. Called “the dark,” Caranthir’s anger and temper are emphasized by Pengolodh. The actual evidence for this in the text, however, is thin. We see Caranthir most often forming alliances with other character groups in a way that is unusual for the Noldor. The Silmarillion describes Caranthir forming relationships with Dwarves, Edain, Easterlings, and Green-elves, most of them productive and mutually beneficial. Only one other Noldo displays the same degree of cosmopolitan interest: Finrod Felagund, the best friend of Turgon, Pengolodh’s king. Again, comparing these two characters and how they are treated is illustrative. Finrod’s alliances are celebrated and held up as indicative of his personal virtue. Unable to deny similar alliances for Caranthir, Pengolodh subtly undercuts them, much as he does to Fingon in describing the Not-Great Battle. When Caranthir aids the people of Haleth—a benevolent, heroic act no matter how you spin it—Pengolodh takes care to note that Caranthir had ignored them until they needed help and is remarkably silent about the “great honour” Haleth gave him, although one can only imagine that, had Finrod been her rescuer, this would have been described in resplendent detail. Yes, Caranthir formed a productive trade relationship with the Dwarves, but he called them ugly, and neither party liked the other much.

Pseudohistorical Language and the Narrator's Knowledge

Written as it is as a pseudohistorical text, The Silmarillion employs language that signals its purpose as a work of history. Looking closer at this language, furthermore, hints at the narrator. The narrator of the Silmarillion attributes sources both written and oral throughout the text. In some cases, like the Noldolantë or the Lay of Leithian, these sources are titled texts. In other instances, the narrator credits specific individuals. Thorondor delivers the news of Nargothrond’s fall and Fingolfin’s death; a harper named Glirhuin provides the prophecy of Túrin’s gravestone. Less specific are attributions to groups of characters, such as the Vanyarin Elves or the vague “the wise.”

Finally, there are attributions that are so vague that they can only imply the sort of untraceable oral tradition that has achieved the status of what might be termed a Middle-earth urban legend. These attributions are most interesting in pinpointing the identity of our narrator. They occupy a sort of gray area where the information is not common knowledge and requires a nod to a source, yet the information lacks a credible, named source. In short, these attributions signal the knowledge about which the narrator is least certain: where he has neither a specific source nor personal experience: it is said, it is told, or it is sung.

Looking at how these three formulas are used across The Silmarillion reveals that they are used for situations where there is no specific, credible eyewitness, including the narrator himself. In some instances, this is because no one could serve as a witness to those events. For instance, the formula is used to report Finrod Felagund's belief that he will not produce an heir: "it is said that not until that hour had such cold thoughts ruled him" (Silmarillion 151, emphasis mine). This thought it obviously out of reach to all except Finrod, who is not our narrator. In other cases, the formula reports on the history of peoples for whom a credible source is presumably out of reach of the narrator.

I was interested to see if the narrator used these formulas more often to discuss some groups of characters than others, so I compiled data about how often the formulas were used for each character groups, shown to the left. The results are quite illustrative. Keeping in mind that this formula signals uncertainty, the narrator seems least certain when reporting on the Ainur and on Mortal humans. The Ainur are understandable; indeed their distance is an inherent characteristic of this group. That Mortals are the group about which the narrator shows the second greatest amount of certainty underscores that this narrator is likely not Mortal. He would have had access to more sources about Mortals than Elves, a group about which he shows relatively little uncertainty. Thinking about the broader context of The Silmarillion underscores this point still further. Most of the Quenta Silmarillion concern Elves; there is relatively little material about Mortals. Yet within that reduced material, the narrator nonetheless shows a relatively high degree of uncertainty.

I also looked at which chapters in the Quenta Silmarillion used the three formulas most often. Again, the results are illustrative. That data is shown below, to the right. In the Quenta Silmarillion, one of these formulas is used a median average of one time, yet there are chapters with many more uses than this. Once again, they follow the pattern described above: Chapters involving character groups unfamiliar to the narrator have more uses of the formula, as do chapters focusing on historical episodes to which Pengolodh could not have borne witness, such as Fingolfin's fight with Morgoth and Eärendil's journey into the West.

Here, I want to take a step back and revisit the broad implication of this otherwise granular set of data. I must again emphasize that the narrators are not simply adornments tacked onto the text, and data sets like this offer evidence to that point. Tolkien established and maintained a distinctive point of view. He appears to have kept in mind what his narrator would have known and where the gaps in his knowledge would have been. He tooled not only the story he was telling but the precise language in which he was telling it to that information. I want to again return to Christopher Tolkien's commentary on his father's revised cosmology, in which he notes, “the old structure was too comprehensive, too interlocked in all its parts, indeed its roots too deep, to withstand such a devastating surgery” (Morgoth’s 383). He is, of course, speaking of the mythology of the Two Trees, their destruction, and the subsequent making of the Sun and Moon, but that passage applies equally well when considering the narrators and their point of view. It is not a simple matter of changing a name on a page but a deep excision of the reported histories, biases, and even language used in the process of creating a believable pseudohistory.

Turning again to the use of the formulas by chapter, there are not only chapters with well over the median uses of the formulas but also a trough: an early span of chapters where the formulas are used no more than once per chapter. This serves as evidence of another important point concerning the narrators, namely that there are two of them, and Tolkien does seem to have distinguished between them. The chapters in the trough were written by Rúmil, who unlike Pengolodh, would have personally known many—even all—of the characters about whom he wrote and likely stood as an eyewitness to many of the events he wrote about. At the very least, he had easier access to eyewitnesses than Pengolodh—locked away in Gondolin as he was—would have.

Rúmil vs. Pengolodh: Detectable Differences

Rúmil's chapters in The Silmarillion differ stylistically from Pengolodh's. They are more active, written in the vibrant style of an eyewitness account. Pengolodh's chapters have a distance to them, in part because of his use of devices, like the it is said/told/sung formula, that signal a historian writing about events experienced by other people—his sources—rather than himself. Pengolodh expresses more uncertainty, which creates the sense that he is pulling back in the narrative, careful to temper himself rather than permitting the unrestrained flight of a storyteller trying to bring a moment he himself experienced to life for a listener or reader.

Returning to bias, there are also evident differences in how Fëanor is characterized in the chapters assigned to Rúmil and those assigned to Pengolodh. Fëanor is one of the few characters shown by both narrators, although Rúmil shows him far more, while Pengolodh mostly covers the deeds of his seven sons. The chapters authored by Rúmil show a complex character, a consummate craftsman, linguist, and orator. Rúmil hardly holds Fëanor blameless for his slow fall into ruin; however, he also acknowledges other factors at play in that downfall. When Fëanor’s pride and arrogance are mentioned, Rúmil is quick to soften what could be seen as describing a fundamental character flaw. When Fëanor’s rebellious words come to the attention of the Valar, he is described by Rúmil as “eminent in self-will and arrogance,” a description that Rúmil quickly softens by pointing out that “all the Noldor had become proud” (73).

Rúmil heavily emphasizes Fëanor’s loving relationships with other characters: his wife Nerdanel, Aulë, and especially Finwë. Finwë’s adoration of him is mentioned multiple times, and when Finwë is murdered by Morgoth, Rúmil credits Fëanor’s grief for his rash, imprudent decisions. Pursuit of the Silmarils—which will loom large in Pengolodh’s writings—is a side note. Likewise for the sons of Fëanor: Celegorm frequents the House of Oromë, Maedhros is friends with Fingon, and in early drafts of the Silmarillion, Celegorm and Curufin were such good friends with Angrod and Aegnor that they brought them over the sea on their stolen ships. Fëanor and his sons, as shown by Rúmil, love and are worth being loved by others.

In contrast, Pengolodh’s accounts of Fëanor and his sons are almost entirely devoid of familial affection. As I mentioned earlier, Fëanor’s dramatic death scene is an emotional blank. His sons arrive and save him, begin to haul his body back to their camp, then stop at his command and watch him die. If they feel anything for the loss of their father, Pengolodh does not record it.

Rúmil, however, would have had a different perspective. As a linguist himself, he almost certainly knew Fëanor personally, and he certainly seems to have admired him. When Fëanor speaks first to his people and then to the emissary of Manwë after the Darkening of Valinor, Rúmil emphasizes the power of his words. He seems impressed, even awed, by his skill. While there is no evidence to suggest that Rúmil and Fëanor were friends, Rúmil’s account of Fëanor’s deeds contains a good deal of empathy and acknowledgement of the complexity of the situation.

Pengolodh, as noted earlier, as a subject of Turgon, likely shared his king’s enmity toward the Fëanorians. Furthermore, he could not have known Fëanor personally—he was born after Fëanor’s death—and his sole contact with the sons of Fëanor likely came when they attacked the settlement at Sirion. His account emphasizes the Fëanorians’ personality flaws, their lust for the Silmarils, and the dark inertia of their oath as the motives behind their actions. There is little acknowledgement of humanity or complexity of character. If they feel anything for the death of their father, their brothers, and their cousins, it is not acknowledged.

Conclusion

The role of point of view and, specifically, the unreliable narrator is a cornerstone of any discussion of literature. I discuss point of view with my middle school students, and they will hardly be rid of it as long as literature study is part of their educations. The unfortunate effect of removing the narrators from The Silmarillion is that a key component goes missing when discussing the book as literature. Instead, the point of view becomes blank, faceless, acquiring a false sense of omniscience that lulls us into reading the text without evaluating it in the same way that we evaluate the points of view of other works of literature.

After years now of thinking and writing about the narrators of The Silmarillion and their biases and the subtle ways Tolkien creates the sense of historicity, though, I am increasingly convinced that—even if the point of view is easy for a reader to overlook or forget—Tolkien certainly did not. Instead, evidence points to the fact that he not only had narrators in mind for different sections of The Silmarillion but manipulated the text in subtle ways depending on who was telling the story. Tolkien was an Anglo-Saxonist—a historiographically complex field—making it impossible to believe that he created those different narrators without wanting his readers to not only notice them but to think about how their points of view affect the history they are recounting.