Elvish Roots – In Defense of Hannon Le

This article is a bit of a departure from my usual writings in this series. It explores a single concept: the basis for “thank you” in Quenya and Sindarin, but veers off into a general discussion on the overall viability of words within Neo-Quenya and Neo-Sindarin. This article only reflects my own opinions (Paul Strack) and not necessarily those of the owner of this site (Dreamingfifi/Fiona Jallings).

All languages have various polite phrases, a common one being an expression of gratitude such as “thank you” in English, “domo” or “arigato” in Japanese or “köszönöm” (kossonom) in Hungarian. The (Neo) Quenya verb for “to thank, give thanks” has long been assumed to be hanta-, based on the festival name Eruhantalë “Thanksgiving of Eru” (UT/214), which might be decomposed into the elements Eru-hanta-lë “Eru-thank-(abstract noun)”. There is no similar phrase in Sindarin, but David Salo coined a (Neo) Sindarin verb hanna- “to thank” as a counterpart to Neo-Quenya hanta-, and this Neo-Sindarin verb was used in Peter Jackson’s Lord of the Ring movies in the phrase hannon le “thank you”. Many people use phrases like Neo-Quenya hantanyel or hantan len, as well as Neo-Sindarin hannon le or le channon, as way of thanking people in Elvish.

There have been objections raised to David Salo’s invented word, most notably by Carl Hostetter in his essay “Elvish as She Is Spoke”. Hostetter’s objection has to do with the probable origin of the (Neo) Quenya verb hanta-. After David Salo coined his word, an article on Elvish prayers was published in the Vinyar Tengwar academic journal in 2002, which mentioned a primitive root √HAN “add to, increase, enhance, honour (espec. by gift)” taken from notes written by Tolkien around 1970. The authors of that article (Patrick Wynn, Arden Smith and Carl Hostetter) suggested that this root might be related to the noun Eruhantalë “Thanksgiving of Eru” (VT43/14). If so, it is also the most probable basis for Q. hanta-, perhaps derived from Primitive Elvish *hantā- “to give honour” or something similar. If that were so, then the analogous Sindarin verb would be anna-, not hanna-, since a primitive initial h- was lost in Sindarin. This is problematic because anna- already appears as the Sindarin verb “to give”.

If we accept the logic above, the use of the Sindarin verb hanna- “to thank” seems to be wrong. This Sindarin phrase hannon le has become something of a flashpoint in the Elvish language community, especially among its more academically oriented members. This potential problem has been used as part of various arguments on how best to approach Tolkien’s language, and is sometimes used as a litmus test on the quality of a particular Sindarin texts. Texts that use this “wrong phrase” can be seen (by some) as inferior and lacking a deeper understanding Tolkien’s languages.

Recently, my friend Elaran attempted to resolve this issue by coining another (Neo) Sindarin phrase for “thank you”: annon allen (see his post here: https://www.elfdict.com/phrases/1-sindarin/56-how_to_thank_in_sindarin). To summarize his reasoning, there is some evidence that late in his life, Tolkien may have updated the Sindarin verb “to give” from anna- to an- based on a gerundal form aned (“giving”) that appears in notes from the late 1960s (PE22/163). If this is the case, then perhaps the verb anna- (from primitive *hantā-) can be used as a verb for “to thank” without introducing any conflicts. This new phrase annon allen has also gained some traction in the Elvish fan community as a replacement for hannon le.

Before examining Elaran’s new phrase, I’d like to first question the “wrongness” of hannon le. In particular, I’d like to explore what it even means for something to be right or wrong in the context of a fictional language like Elvish.

Let’s start off with what we know for certain about this question from Tolkien’s perspective. The answer is … not much. We know that Tolkien coined a name Q. Eruhantale “Thankgiving to Eru”. This means is pretty probable that Q. hantale = “thanksgiving”. It is speculation that Q. hanta- is a verb, however, since we don’t know if that element of the word can be used independently. It is even more speculative that hanta- is derived from the root √HAN, because the word and the root appear in two different texts written at two different times. It is more speculative still that Sindarin would have used the same root for its verb for “to thank”. No other Elvish word for expressing appreciation or gratitude appears in Tolkien’s published writings.

One possible standard for correctness in Elvish is “use only words that we know Tolkien would have used”. I personally find this standard to be unworkable. If this is our standard, not only will we have no Sindarin phrase for “thank you”, we won’t even have a Quenya phrase for “thank you”, since the extrapolation of Q. hanta- from Eruhantale involves a number of educated guesses. This is, in fact, the ultimate point of Carl Hostetter’s “Elvish as She is Spoke” article: we don’t know what Elvish words Tolkien would have used for “thank you”, and likely never will (and the same is true for many other words). If “only what Tolkien used” is our standard for correctness, we can’t construct anything resembling a functional Elvish language.

If we want a word for “thank you”, we have to move beyond what we know of Tolkien’s ideas for Elvish and into the more speculative realm of extrapolations and extensions of Tolkien’s languages, commonly referred to as Neo-Quenya and Neo-Sindarin (this is why I was throwing “Neo” in front of everything at the beginning of this article). In that realm, we generally have a more relaxed standard for viable Elvish words: not just “words we know Tolkien used” but also “words that Tolkien might have used”. In particular, we try to justify the introduction of the new Elvish words based on what we already know about Tolkien’s languages.

By itself, the standard of “what Tolkien might have used” is too broad, however. The space of possible Elvish words is infinite, and if we allow any conceivable word, it is still impossible to construct a functional Elvish language: because it would give us too many words rather than too few. There are almost as many possible approaches to extending Tolkien’s languages as there are people interested in Elvish, and not every member of the community agrees on which approach is best (or even whether this kind of extrapolation is legitimate at all). While I have my own opinions on the most effective methods for extending Elvish, I’m going to sidestep that question for now.

Consider instead what it means for a word to be right or wrong in a real-world language. From a purely utilitarian perspective, a word is correct in a language if both the speaker and the listener can agree on its meaning and it can thus be used for communication. There is no a priori definition for the English word “thank” outside the context of English speech; “thank” is correct in its meaning simply because English speakers agree that it is correct. This definition of correctness works for English because there is a large body of English speakers to enforce the norms of the language and educate new speakers on how it should be used (though this education does not have 100% fidelity, which is why languages slowly evolve over time).

If Neo-Sindarin and Neo-Quenya are to function as languages, this is ultimately the only standard that will work: a word in Neo-Elvish can only be considered correct if both the speaker and listener (or more frequently, writer and reader) can agree on the meaning of a word. Based on this, I think the most workable standard for the validity of a Neo-Elvish word is that (a) it is a word that may be plausibly derived what we know of Tolkien’s Elvish languages and (b) it has a degree of acceptance and use within the Neo-Elvish community. In the second respect, the Neo-Sindarin verb hanna- does quite well: a large number of Neo-Sindarin speakers use it for “thank” based on the influence of Peter Jackson’s movies. The main question is whether it can be plausibly derived from what we know of Tolkien’s ElvisI happen to agree that the most probably origin for (Neo) Quenya hanta- “to thank” is Primitive Elvish *hantā- from the √HAN. I also agree that, were the Sindarin word derived from the same primitive form, it would be anna-. However, direct derivation from a primitive form is not the only way that (Neo) Sindarin hanna- could have entered the language. It could for example, have been a loan word for Quenya, a phonetic adaptation from Q. hanta-, modified because the combination nt does not appear in the interior of Sindarin words. The question is whether such a Quenya loan word is plausible.

It so happens that there is a precedence for such loans between Quenya and Sindarin. In particular, in notes written by Tolkien in the song book A Road Goes Ever on, Tolkien said that S. le, the polite or reverential form of “you”, was such a loan word from Quenya (RGEO/65). In other notes written in the late 1960s, Tolkien clarified that that S. le entered Sindarin due to the influence of the Noldor in Beleriand, and became widely used everywhere outside of Doriath (VT49/51). Its lack of use in Doriath was almost certainly the result of the hostility of that nation to the Noldor because of the kinslaying in Aqualonde.

Outside of Doriath, however, the Noldor generally became the leaders of the Elves, and their Quenya language became something of a prestige language among the Elves. It is not unusual for certain phrases to migrate between a prestigious language spoken by the leaders of a people and the language of the common folk. For example, a large number of Old French words were adopted into English after the Norman invasion of 1066, when for several centuries the nobility of England primarily spoke French. Apparently Tolkien imagined le entering Sindarin through similar means as an adaptation of polite Q. lye, and (Neo) Sindarin hanna- “thank” could have been a similar loan word.

Using this scenario, I personally find it plausible that the verb hanna- could have been used in Sindarin. I make no claims on whether Tolkien himself would have found it plausible. I do, however, think this scenario is at least compatible with what we know about the Elvish languages and how they developed. Using the reasoning above, I think that hannon le or le channon could be perfectly acceptable Neo-Sindarin phrases.

At this point I’d like to return the Neo-Sindarin phrase coined by Elaran: annon allen “thank you, (lit.) I give thanks to you”. I have to admit that when I first heard of this phrase, I didn’t like it. It wasn’t the phrase itself I had a problem with; my issue was that its derivation assumes that the Sindarin verb anna- “to give” was revised to an-. This assumption strikes me as problematic: anna- is quite well-established as the Sindarin verb for “give” and revising it would be a major source of confusion. Most Sindarin readers easily recognize the present-tense forms anna and annon as “give” and “I give”. On the other hand, the revised forms of its present-tense, ân or enin, would probably not be understood by most Sindarin readers.

Nevertheless, annon allen is still a reasonably plausible phrase for “thank you”, and like hannon le it has gained a degree of acceptance in the Elvish speaking fan community. If I am going to play by my own rules, I should at least consider its viability before saying it is “wrong”. My general philosophy is that if two seemingly contradictory ideas are widely used in Neo-Sindarin or Neo-Quenya, we should first see if we can find some way to make them compatible before discarding them completely.

Continuing my proposed scenario above, if the verb hanna- entered the Sindarin language from Noldorin influence, annon allen might still have remained a phrase for “thank you” in the one place where the Noldor had no influence: Doriath. Even if we retain the Sindarin verb anna- “to give” (as I think we should), annon allen might survive in Doriath as a fossilized phrase, perhaps originally derived from Primitive Elvish or Ancient Sindarin *hantā-ni an-len(a). Even if the verb (h)anna- “thank” fell out of use because it conflicted with S. anna- “give”, its basic sense could have survived in this one phrase, though probably it was assumed by later Sindar that its actual meaning was “I give to you [thanks]”, with the word for “thanks” omitted.

In this scenario, the phrase annon allen could have reentered common Sindarin usage after the fall of Doriath, when the surviving Elves overwhelmed by Morgoth were forced together on the west coast of Beleriand. At this stage, the various Sindarin dialects (North Sindarin, West Sindarin and Doriathric) began to blend into what would ultimately become the common Sindarin dialect used throughout Middle Earth. Furthermore, by this point the Noldor had lost their position of prestige and leadership, so that perhaps the more archaic Doriathrin phrase annon allen “thank you” was seen as more formal and prestigious than hannon le.

If we accept this scenario, it is possible that both hannon le and annon allen could coexist in (Neo) Sindarin, with hannon le being more casual (“thanks”) and annon allen being more formal (“I give thanks to you”), somewhat like the distinction of Japanese “domo” versus “arigato”. This happens to reflect the use of the two phrases among Neo-Sindarin speakers, with hannon le being used by newer students of Sindarin who are only familiar with older and more casual Sindarin courses, and annon allen being gradually adopted by those who are more formal and academically-oriented in their use of the language. This is a scenario that allows everyone to be at least somewhat correct in their use of the language.

Now, I happen to consider the likelihood that Tolkien himself considered either of these phrases (much less both) as being valid Sindarin to be very small. I also don’t think the scenario proposed above reflects the thoughts of either David Salo or Elaran as the originators of these phrases. At this level of specificity, questions about whether something is “right or wrong” has reached such a degree of ambiguity that they become quite meaningless. The “rightness” of the phrases depends on the context which you look at them: whether they represent Tolkien’s own thoughts (almost certainly “no” in both cases), whether they reflect the thoughts of David Salo or Elaran (“yes” for one phrase but “no” for the other), or whether or not the phrase would be understood by a Sindarin reader (mostly “yes” for hannon le, even among those who object to its use, and often but not always “no” for annon allen given its newness, though this will change over time as its acceptance grows).

Given the fragmentary and sometimes contradictory nature of the source material, it is nearly impossible to say whether something is “correct” in Tolkien’s Quenya and Sindarin for the vast majority of topics. Neither Neo-Quenya or Neo-Sindarin has matured to the point where there is a concrete way to determine whether something is correct is those fan-based extensions either. At this time, I don’t think “right/wrong” or “correct/incorrect” are even useful concepts when applied to Neo-Quenya or Neo-Sindarin, since they imply some kind of normative framework for evaluating the languages which simply doesn’t exist. In the near term, I think its better to use a variety of other criteria for evaluating Neo-Elvish words and phrases, such as “comprehensibility”, “consistency”, “usefulness” or “plausibility”.

By “comprehensibility” I mean “how likely is it that a listener/reader would understand the meaning of an Elvish word or phrase”. In this respect, it important to consider what kinds of information is available to the typical student of Elvish, even if their knowledge is based on dated or imperfect sources. No two Elvish courses are in complete agreement, but the areas in which they do agree should be given extra weight, since it increases the chances that an arbitrary reader would understand the intent of a message. Furthermore, adding too many elements to Neo-Elvish can reduce comprehensibility. Too many homonyms or synonyms make it more difficult to understand a writer’s intent, and if we layer too much semantic meaning on an individual word, then that word becomes too muddled in meaning. Some degree of ambiguity is acceptable (even unavoidable) but too much makes the languages overly difficult to understand.

“Consistency” has a variety of dimensions. First is consistency with what Tolkien wrote, although this isn’t entirely clear cut, since a lot of what Tolkien wrote is (a) complex and (b) has inconsistencies of its own. Another factor is the internal consistency of Neo-Elvish. Since we can’t be consistent with every single thing Tolkien wrote, we should at least make sure that the Neo-Elvish we use it not contradictory within itself, and that it only includes elements of Tolkien’s writing that are consistent with each other. Finally, there is the aspect of consistency with what has been previously established about Neo-Elvish within the community. We should avoid adding new elements to the languages if some solution to a particular linguistic question already exists and is widely accepted. We should avoid rejecting established conventions unless (a) some major flaw is discovered or (b) we are lucky enough to have more of Tolkien’s own thoughts published that shine new light on a particular problem.

“Usefulness” is a criteria that can sometimes justify the inclusion of more dubious material. In some cases, the only examples we have of Tolkien’s thinking on a particular topic are from rejected material or from the earliest days of his writing. Sometimes we have no good alternatives, and our only choices are to either use early or iffy sources, or to invent something from whole cloth. Other times, Tolkien might introduce new ideas in his later writings that invalidate some of his earlier notions, but incorporating those new ideas requires throwing out big chunks of established Neo-Elvish while adding little new to the languages themselves. Tolkien sometimes worked in small speculative circles, where he would briefly introduce a new concept and experiment with it for time, only to discard it later. However, we shouldn’t give so much weight to usefulness that we throw out consistency and plausibility.

All three of the above criteria are subjective, and the fourth criteria, “plausibility”, is the most subjective at all. By “plausability” I mean “how likely is it that a given word could be part of Elvish as imagined by Tolkien”. I think “plausibility” is as close to “correctness” as we can get to at this time. This standard is the most difficult to work with, however, because there are so many ways to approach Tolkien’s languages. What might seem plausible to one person could seem quite implausible to another. For example, some people weight Tolkien’s later writings very heavily when it comes to evaluating plausibility, while others are more willing to draw ideas from Tolkien’s early writings. The factors one might consider for the plausibility of a given Elvish word is a complex topic deserving an entire essay of its own. It’s an essay I’m going to defer to another day, since I’ve already gone so far afield in this current article.