The Silmarillion: Who Speaks? by Dawn Felagund

Dialogue does not occur evenly across The Silmarillion. While a little over 5% of the words in The Silmarillion as a whole are used in dialogue, this is very unevenly distributed across the chapters, with some chapters about half dialogue and six chapters containing no dialogue at all.

There is a lot more work to be done to tease out trends and patterns that might have some meaning, but just glancing at the graph above, some of those patterns do begin to emerge. First, dialogue increases as The Silmarillion progresses. In the second half of the book (calculated by chapter, not page or word count), only two chapters have no dialogue and only four chapters (inclusive of those two without dialogue) fall below the median of 5.3% dialogue. Put another way:

• In the first half of chapters, 71% of chapters are below the median.
• In the second half of chapters, 29% of chapters are below the median.

Why is this? My tentative theory is that we see the book moving from the realm of the mythic—from events that are passed down through the oral tradition and ancient written traditions—and into the historical, where the narrator has a greater array of sources, including eyewitness testimony, and begins to write with greater immediacy rather than the arm's-length style of myth and ancient history.

What I am curious about: As I dig deeper into these data, will I see this theory bear out in which episodes or characters/groups are granted actual dialogue? In other words, will characters and peoples lost to the mists of time speak less, as I would expect? Or will the type of dialogue (e.g., a formal speech that may have been preserved vs. an extempore conversation that would not) vary based on narrative distance? I have documented in the past that the narrator of The Silmarillion uses the "it is said/told/sung" construction more with characters who are less accessible, so there is evidence that Tolkien manipulated writing style based on what his narrators' access to various sources. Does he use dialogue similarly to communicate that "mythic distance"?

There are also chapters that are more expository in purpose (Valaquenta, "Of Beleriand and Its Realms") that do not contain dialogue. Without digging deeper into the chapters themselves, most of those without dialogue that aren't similarly expository are chapters where the material would be less accessible to Pengolodh as a narrator. Whether this bears added scrutiny remains to be seen!

Finally, in discussing these data on the SWG's Discord, polutropos noticed something interesting, which is that the chapter with the most dialogue—"Of Aulë and Yavanna," where almost 57% of the words of the chapter are given over to dialogue—was not in fact written by Tolkien. As document by Douglas Charles Kane in his book Arda Reconstructed, "This chapter is completely manufactured by Christopher, though using his father's own writings" (page 54). Where Kane usually includes a chart pointing to the source for each bit of The Silmarillion, his chapter on "Of Aulë and Yavanna" contains no such chart because, while he is able to document where ideas came from, Christopher actually wrote the chapter.

Interestingly, "Of the Noldor in Beleriand" is the chapter with the second most dialogue and, according to Kane, "The changes made in this chapter are among the smallest anywhere in the published text" (page 154). So Tolkien does sometimes write dialogue-heavy chapters—though without data to back me up (yet! it's coming!), most of that dialogue appears to come in the form of lengthier speeches, not necessarily the debate/conversation format of Of Aulë and Yavanna."

The biggest impact of the dialogue-heavy "Of Aulë and Yavanna," I suspect, will emerge as I dig more into the data on gender and who speak in The Silmarillion. Yavanna is one of the women who speaks the most in The Silmarillion, but almost all of her dialogue occurs in this chapter. If this chapter is constructed by Christopher, how does that impact the amount of speech women are permitted by Tolkien? Polutropos' observation spurred me to plan to document the source of the various dialogue sections: Are they original to Tolkien's writings or added? Kane, interestingly, is critical of Christopher Tolkien in Arda Reconstructed for what he perceives as Christopher removing women characters from the text. In this instance, we see a significant example of the opposite: a woman's role is not only expanded, but she is given an opportunity to speak.

The whole reason I decided years ago that I wanted to count instances and words of dialogue is because who gets to speak matters. Who gets to tell their own story in their own words?

(And, for the record, in this data set, "dialogue" is measured in instances of dialogue, not in word count.)

Of course, in The Silmarillion, it is more complicated than that because The Silmarillion is a pseudohistorical text, so we have to constantly question whether what the narrator is telling us happened (or is said) was in fact what happened (or what was said). The prevalence of group dialogue in The Silmarillion—when specific, quoted speech is attributed to a group of characters rather than an individual—attests to the inexact science that is dialogue in the text. There are twenty-six instances of group dialogue across the book.

So it would perhaps be more accurate to say that dialogue matters because it indicates who the narrator wants to allow to tell their story in their own words with the authority that comes from being important enough to quote.

Which groups of characters get to speak aren't surprising, but this does still tell us something important about the perspective we are given in The Silmarillion. Elves speak the most, but The Silmarillion is an Elven history, so we'd expect that. Within the category of "Elves," though, speech is entirely dominated by the Noldor and Sindar. One Teler (Olwë) gets a single instance of dialogue, and the Green-elves get to speak once as a group.

Why are these characters' perspectives absent? Does this simply reflect a limitation of the narrator, or is the narrator foregrounding the experiences and perspectives of Noldor and Sindar as more valuable—worth quoting?

Mortal Humans and Ainur speak almost the same amount. I find this interesting because, as I noted in the graphic, the Ainur made a big deal about flouncing from Middle-earth after the Noldorin rebellion. Yet they sure have a lot to say about things. In contrast, after they arrive in Beleriand, Mortal Humans are always at the hub of the action. Someone is always getting shot through the eye or something. Yet their stories, in their own words, are told only a bit more than the intoning, cursing, and speechifying that the Ainur get up to.

When Mortal Humans do speak, they are always "Men of the West": Edain, Númenóreans, or Dúnedain. We do not hear once from other groups of Mortal Humans, such as Easterlings, even speaking as a group. This certainly calls into question the absence of those perspectives from the story. Think about it: the Silmarillion narrator finds God a more accessible, quotable source than an Easterling soldier.

All of this corroborates other data and observations about Mortal Humans that I've made over the years (for example, the death data). Our Silmarillion narrator often seems to include Mortals almost grudgingly because, yes, they did important stuff in the story but tends to see them as ephemeral or expendable and really only likes to talk about them when they are doing cool stuff with Elves.

These data do add an interesting (to me anyway) perspective as well on the question of the Silmarillion historical tradition. In a nutshell, in the late 1950s, Tolkien considered that the historical tradition must be Númenórean, not Elven. Even though there was no strong evidence that he made significant revisions with this change in mind, it was enough for Christopher Tolkien to scrap mentioning a narrator in The Silmarillion at all, and several scholars have followed suit in asserting The Silmarillion is a "Mannish" history.

I have been a fiction writer much longer than a Tolkien scholar, and one does not simply walk into changing point of view. I've made the case for years now that Tolkien realized the depth of revisions that would be required and either changed his mind or just never got started on them. Regardless, the text we have is Elvish. These data support that: the predominance of Elven and Ainurian perspectives attest to a narrator whose main sources are Elves and Ainur and who subtly but nonetheless devalues the perspectives of Mortal Humans, even though they are the grist in the mill of the war against Morgoth.

Finally, there are Dwarves. The only named Dwarf who speaks is Mîm, in the "Of Túrin Turambar" chapter, which is an outlier of a chapter in its use of dialogue (it also has a different narrator) and which will probably gets its own analysis someday. Otherwise, Dwarves speak in groups. As with Mortal Humans, these data align with other data and observations I've made over the years of the status of Dwarves to the Silmarillion narrator. In this case, they are valued but often inaccessible—the opposite of Mortal Humans.

Methodology Notes

As already stated, all data are instances of dialogue, not word or sentence count.

Classifying Elves into subgroups is remarkably challenging. Here's how I did it for this project:

• "Noldor" includes characters who have Telerin or Sindarin ancestry in addition to Noldorin.
• "Teleri" includes only characters with only Telerin ancestry who live in Aman.
• "Sindar" includes only characters with only Sindarin ancestry who live in Middle-earth; mixed Noldorin/Sindarin is not included; Lúthien is included.

Finally, the "Other" group includes animals, dragons, Orcs, objects, and speakers whose identity is not stated even enough to determine what group they belong to.