Elvish Roots – Approaches to Neo-Elvish

I’ve discussed in other posts how Neo-Elvish reconstructions (Neo-Quenya and Neo-Sindarin) are useful tools for learning and using Tolkien’s languages. Many such reconstruction are widely accepted and not especially controversial. For example, we know there is a Quenya verb har- “to sit, stay” (PE17/162; UT/305). We don’t know its past tense but it is reasonable to assume it is *harne “sat” by comparison with the similar past tense carne “did” of car- “to do” (PE17/74; PE22/152). We also know there is a Sindarin word aglar “glory” (LotR/953), and we can reasonably guess its plural is *eglair by analogy with erain plural of aran “king” (PE17/40, 111). Such reconstructions are so widely accepted that most writers wouldn’t bother marking them with a “*” to indicate that they do not appear in Tolkien’s writings.

Other reconstructions are less straightforward, however, and there isn’t a complete consensus on the best approach for building on what we know of Tolkien’s languages. Newly created words that are extrapolations from Tolkien’s writings are generally called “neologisms” in the Elvish linguistics community, and there are vigorous debates on which techniques are legitimate methods for creating neologisms (and even what should or shouldn’t be considered a neologism). Here are a few statements that I think are relatively uncontroversial among Neo-Elvish writers:

• Words created by Tolkien are preferred over words created by fans (neologisms).
• Where multiple words with the same meaning were created by Tolkien, those created later in his life are preferred over words he created earlier on.
• As much as possible, we should adhere to the linguistic structures and aesthetics as Tolkien imagined them during and after the publication of The Lord of the Rings.

Even within these three seemingly simple statements there lurk quite a few complexities, however. For example, what if the only word available for a particular concept appears in Tolkien’s earliest writing. Is using such a word legitimate? What if an early word does not fit with the linguistic structures of Tolkien’s later ideas: is it legitimate to alter its form to be more consistent with the phonology of the later languages? When considering whether one words replaces another, what does it mean for two words to have the “same” meaning? Must they having identical translations, or merely similar? When it comes to adhering to Tolkien’s later ideas, what is more important: preserving Tolkien’s words themselves, or preserving the conceptual foundations and phonetic character of Tolkien’s later languages?

By way of illustration, let’s look at a specific example: words having to do with marriage, including the words for husband and wife. Over his life, Tolkien’s seems to have considered at least three distinct linguistic paradigms for these words:

A) In the 1910s (what I call the “Early Period”), Tolkien had a root ᴱ√BEÐE with the Early Quenya noun ᴱQ. veru(ner) for “husband”, the Early Quenya nouns ᴱQ. †ver(un)i (archaic) as well as ᴱQ. vesse or vestin for “wife”, and the verb ᴱQ. vesta- “to marry” (QL/101). In the 1910s, the conceptual precursor to Sindarin was called Gnomish, and it had equivalent words G. †benn (archaic) and G. bedhron for “husband”, G. †bedhril (archaic) and G. bess for “wife”, along with the verb G. benna- “to wed”, all of which were also derived from the same root √BEÐE (GL/22).

B) In the 1930s (what I call the “Middle Period”), Tolkien had a root ᴹ√BES with (Middle) Quenya nouns ᴹQ. venno “husband” and ᴹQ. vesse “wife”, along with the same verb ᴹQ. vesta- “to marry” as he used in the 1910s. There is also an interesting dual form ᴹQ. veru meaning “married couple”. The conceptual precursor to Sindarin from the 1930s was called Noldorin, and the analogous Noldorin nouns were N. benn and N. bess, but these no longer meant “husband” and “wife”, having shifted in meaning to “man” and “woman” instead. The normal Noldorin words for “husband” and “wife” were N. hervenn and N. hervess. There is no attested Noldorin verb for “to marry” (Ety/BES).

C) In the 1960s (in what I call the “Late Period”), Tolkien had a root √BER with Quenya nouns Q. veru “husband” and Q. veri “wife”, a partial restoration of the Early Quenya paradigm. He had the verbs Q. verya- “to get married (intransitive)” and verta- “to marry someone” (transitive). In this paradigm he did not provide Sindarin equivalents to the Quenya words (VT49/45).

As an additional note, the Sindarin word bess appeared in the unpublished epilogue to The Lord of the Rings written around the 1950 (the cusp between the Middle and Late Periods), but in this epilogue the word bess had the translation “wife” rather “woman” (SD/129).

Given all of the above, what words should we use for “husband”, “wife” and “marry” in Sindarin and Quenya? The answer depends on which elements of Tolkien’s languages you consider most important. I don’t know of anyone who would seriously propose using the forms from the 1910s (except insofar as they reappear later), but I’ve seen a variety of approaches for using the other words.

Quenya-only solution: If you only care about Quenya, put primacy on Tolkien’s own words and always want to use the latest attested forms, the logical choice is Q. veru “husband”, Q. veri “wife”, and either Q. verya- or Q. verta- for “to marry”, depending on whether you are a participant in the marriage (verya-) or you are the one officiating (verta-). This is probably the most common solution I’ve seen among people who primarily write Neo-Quenya.

Sindarin-only solution: If you only care about Sindarin, put primacy on Tolkien’s own words and always want to use the latest attested forms, the logical choice is N. hervenn “husband” and N. hervess “wife”, adapting the middle-period Noldorin words into Sindarin. An argument might be made for using S. bess as “wife”, but since this word was also used for “woman”, its meaning is more ambiguous, so using N. hervess makes it more likely your intent would be understood. A verb for “to marry” is trickier, but you can coin a neologism ᴺS. *besta- for the middle-period root ᴹ√BES, the equivalent of middle-period Quenya ᴹQ. vesta-. This is probably the most common solution I’ve seen among people who primarily write Neo-Sindarin.

The problem with the “Quenya-only” and “Sindarin-only” solutions are that they are not mutually compatible. They are based on two distinct paradigms, one from the 1930s (ᴹ√BES) and another from the 1960s (√BER). People that care about both Neo-Quenya and Neo-Sindarin therefore cannot simply adopt these two paradigms, because they are internally inconsistent. I’ve seen a variety of approaches to solving this kind of quandary.

Late-forms only: If you put primacy on late linguistic structures over all else, you might adopt the root √BER as the basis for both the Quenya and Sindarin paradigm. You could use the Quenya words Q. veru and Q. veri unchanged, and coin Neo-Sindarin equivalents: ᴺS. *bêr “husband” and ᴺS. *bîr “wife”. For “marry” you can coin neologisms ᴺS. *beria- and ᴺS. *bertha- “to marry” (intransitive and transitive respectively). This approach has several advantages and disadvantages:

• Advantages:
  • This is probably the most faithful representation of Tolkien’s later linguistic aesthetic.
  • It is a paradigm that Tolkien himself might have used, and simply failed to write down.
• Disadvantages:
  • You may be forced to replace attested words with neologisms; a Neo-Sindarin reader or writer who is unfamiliar with your approach may question why you are not using N. hervenn and N. hervess.
  • You risk introducing unplanned homonyms into the languages. For example, both N. beria- and N. bertha- already appear in Tolkien’s writings as Noldorin verbs with the glosses “to protect” and “to dare” (see below).

Mix-and-match: If you put primacy on words invented by Tolkien himself, you might look for a way to reorganize his words into an internally consistent paradigm. The hypothetical primitive forms *besū and *besī would also produce the Quenya words Q. veru and Q. veri, which means Quenya veru/veri and the adapted Noldorin words hervenn/hervess could part of the same paradigm if they were all derived from the middle-period root ᴹ√BES. The same is not true for the late Quenya verb forms, however: Q. verta- cannot be derived from ᴹ√BES. For consistency, we would be forced to use the older verb ᴹQ. vesta- “to marry” from the 1930s. In this approach:

• Advantages:
  • You are mainly using words created by Tolkien himself.
  • People who follow the “Quenya-only” and “Sindarin-only” approach would generally be comfortable with the words you are using.
  • You can avoid introducing homonyms to the languages by being somewhat selective in your word choices.
• Disadvantages:
  • You are inventing a paradigm that Tolkien himself never considered valid.
  • The collection of words you use are less mutually compatible than those originally imagined by Tolkien.
  • You may be forced to adopt earlier words over later words to maintain internal consistency: for example choosing ᴹQ. vesta- from the 1930s over Q. verya-/verta- from the 1960s in the example above.

Broad sweep: Another solution I’ve seen to this quandary is to allow the use of any words created by Tolkien that are not directly incompatible. Even in broad this approach, the very earliest words which are phonetically incompatible with Tolkien’s later ideas would not be used, but words from the 1930s like Quenya venno/vesse and Noldorin hervenn/hervess (used as Sindarin) could all happily coexist with words from the 1960s like veru/veri. Where multiple words have identical meanings, they might be distinguished from each other via slightly different connotations.

• Advantages:
  • You allow the use of the broadest collection of Tolkien’s own words.
  • This broader word base allows for more subtle distinctions in meaning.
  • People who use a narrower approach will frequently find their words remain valid in this generalized paradigm.
• Disadvantages:
  • You allow contradictory words to coexist in a way that Tolkien himself rarely did.
  • You may inadvertently introduce homonyms into the languages by allowing the use of words with the same forms from different periods of Tolkien’s life. For example, if you allow both Q. veru and ᴹQ. venno for “husband”, what does that mean for ᴹQ. veru “married couple”?

As a final consideration, your word-choices can have ramifications outside the set of words immediately under consideration. For example, there is a middle-period root ᴹ√BER “valiant” with derivatives like Quenya ᴹQ. verya- “to dare” and Noldorin N. bertha- of the same meaning. If we adopt the root √BER for “marry”, does that mean the middle-period words derived from ᴹ√BER “valiant” become invalid? What about words that reappear in later writings but don’t have a new etymology, such as the name S. Beren “Bold”? Accepting √BER “marry” doesn’t necessarily mean rejecting ᴹ√BER “valiant”, but accepting both means there are more words in the Elvish languages with ambiguous meanings.

Given all these complexities, what is the “right” choice to make for these and other similar Neo-Elvish constructions? The simple fact of the matter is that there isn’t one approach that is manifestly better than the others. Different people in the community advocate different strategies (full disclosure: I personally tend to follow the mix-and-match approach, so this article is likely biased in that direction, though I’ve tried to be fair to the other approaches). More realistically, the set of “approaches” listed here isn’t comprehensive, and different linguistic challenges within the Elvish languages can be more easily addressed using a variety of strategies. Few people are absolute purists in their approach, and can adopt different techniques in different circumstances.

That said, the choice of approach has ramifications on the kind of Neo-Elvish writing you produce:

Single-language-only: Focusing on a single language (Quenya or Sindarin) reduces the problems you need to deal with, because many of your choices become easier (you often don’t even have a choice). But doing so limits your ability to draw on the larger Elvish-language family and still maintain an internally consistent paradigm.

Late-forms-only: Using primarily late period words and linguistic structures is probably the purist representation of the Elvish languages as Tolkien imagined them when The Lord of the Rings was written. But making this choice often means you are more constrained in the words and concepts you have available, and you may be forced to invent to new words to fill gaps that could otherwise be filled by earlier words created by Tolkien himself.

Mix-and-match: This approach allows you to cherry-pick from Tolkien’s words in various periods while maintaining internal consistency, giving you are broad selection of linguistic concepts. But doing so means using Tolkien’s linguistic structures in a way that doesn’t represent of his decisions at any point in time. Furthermore, some words are mutually exclusive and simply cannot put into a coherent linguistic paradigm. Therefore, you will sometimes be forced to abandon words that Tolkien invented in the name of consistency.

Broad Sweep: This approach lets you use the broadest collection of words possible, given you the most options, which is especially useful when writing poetry or expressing subtleties of meaning. The price is that the internal consistency of the languages are eroded in a way that Tolkien himself rarely allowed. In some respects this is the easiest and most useful approach (in terms of word options) while being the least representative of Tolkien’s linguistic aesthetic.

Which approach you take depends on your own knowledge and preferences, and which aspects of Tolkien’s languages you consider the most important and which you are willing to diverge from.