This is a repost of an article I originally published on (soon-to-be defunct) Google+.
This word analysis has to do with negations. It is a problem that has been discussed repeatedly. Bill Welden’s wrote up a thorough analysis of the conceptual development of negations in Tolkien’s Elvish languages in VT42, and I have little to add to that. Rather, this is my attempt to reconcile the various negative-paradigms Tolkien used throughout his life into a form that is internally consistent without requiring us to discard a large numbers of attested words.
To summarize the basics, by the end of Tolkien’s life he used essential three distinct roots for negations BĀ, LĀ and Ū.
BĀ is the easiest to work with. This root has to do with willful negation, refusing or forbidding. It requires a sentient entity performing the act of negation, and can be used this way in both Quenya and Sindarin in the derived particles: Q. vá, S. baw, as well as verbs for refusal or forbiddance: Q. and S. ava-. This verb was fossilized in Sindarin, nearly reduced to a negative adverb. In Quenya it could also be used as a negative prefix, but only for negatives that involved some intentional action as in Q. avalerya- “to restrain, to deprive of liberty” or avaquétima “unspeakable” in the sense of “that which must not be said”, as well as Q. váquet- “to say ‘I will not or do not’; to refuse, forbid, prohibit”.
LĀ was Tolkien’s basic negative root in the 1930s and 40s, and probably remains the most popular form of negation in Neo-Quenya. It was used for a “simple negative”, absence of fact. Q. láquet- simply means “to say ‘it is not’, to deny a fact”. As a prefix, especially in Quenya, it seems to represent a simple absence of an attribute or entity, without any addition connotations. It could be used as an adverb, prefix and (semi-)verb in Quenya at least. In Noldorin/Sindarin there are few examples of it, almost exclusively in the form of the prefix al- “no, not”.
In the 1930s there was another “strong negative” prefix ú- derived from UGU/GŪ. It came to have an “unpleasant” connotation (at least in Quenya) in association with roots like UMU/MŪ which had derivatives like ᴹQ. úmea “evil”. As an example, at one point Tolkien defined Q. úquétima “unspeakable” in the sense of “impossible to say or put into words, unpronounceable”. This LĀ vs. Ū weak/strong paradigm was also reflected in Early Elvish negative elements u vs (syllabic) ḷ, except that in the earliest period the meaning of the two were reversed.
In the 1950s, LĀ as a negative fell out of favor with Tolkien, in part because he increasingly used the root ALA with the meaning “good, blessed, fortunate” (this root also appeared in the Etymologies, but was barely used). The prefix al(a)- came to mean “well, happily” in both Quenya and Sindarin, and was thus no longer suitable as a negative prefix. In this period he abandoned LĀ in favor of using Ū the basic negative prefix. The only “canonical” appearance of negative LĀ in LotR was in S. alfirin as the name of a flower; it was unglossed in the text, but as indicated elsewhere it was clearly intended to mean “immortal”.
In the 1960s he changed his mind yet again (VT42/33; PE22/153, 160). This time Ū fell out of favor and he restored LĀ as the basic negative. The sense of Ū was modified to mean “difficult” or (as strengthened in Sindarin) “impossible”, becoming more of a pseudo-negative. Then in typical Tolkienesque fashion he vacillated at least one more time: flipping away from LĀ and back to Ū (VT44/4, 38). Tolkien changes his mind so many times and wrote so many confusing notes that it’s nearly impossible to figure out what his “final” decision was (if in fact there was one).
Many Neo-Quenya authors tend to favor either Ū over LĀ or vice versa. However, I think the best option is to restore the weak/strong negative paradigm of the 1930s, using LĀ for a “simple negative” and Ū for a “strong negative”, at least in Quenya. Unfortunately this probably means the root ALA “good, blessed, fortunate” needs to be abandoned; it conflicts too much with negative LĀ. I’d replace it with derivatives of MAY “excellent”: NQ. mai- and NS. mae-.
For Ū, I think it should only be used for a “strong” negative. It can have senses like “impossible”, “anti-, opposite”, “hard, difficult” or “bad, unpleasant”. As such, it really only makes sense when used to negate something that can have an opposite: **úcarne “anti-red” is nonsensical. It is also generally used to negate positive things, so that Q. úmare “anti-good” = “evil” and úvanima “anti-beautiful” = “ugly, hideous”, but **únorto “anti-horror” (= “?calm”) would feel strange and ungrammatical.
Thus, in Quenya, the three common negative particles are:
- lá “generic no, not”
- ui “emphatic no, not”
- vá “volitional no, not”
In answer to the question Ma tuluval enar? “Will you come tomorrow?”:
- lá “no” (I’m simply not coming)
- ui “definitely no; on the contrary” (in fact, I’m going away to Gondor)
- vá “no; I refuse” (you can’t make me come)
All three can be used as “semi-verbs”, inflected for person before another verb:
- lán tuluva “I will not come”
- uin tuluva “I will definitely not come”
- ván tuluva “I will refuse to come”
The subordinate verb generally has the inflection for tense (future in the above examples). The semi-verbs are only inflected for tense when they stand alone:
- lánen “I was not”
- unen “I was definitely not”
- avanen “I refused”
The particle vá is never itself inflicted for tense, since there is a distinct verb Q. ava- “I refuse, forbid” that may be fully inflected and used normally.
The three main negative prefixes are:
- la- or lá- “not, un-”
- ú- “strong not, anti-”
- ava- “(volitional) not”
I personally would avoid the negative prefix al-, since it is attested as a both negative prefix (not) and a prefix meaning “well”. However, I think it’s more likely to be recognized as a negative, so if you do you use it, I suggest only using it in that way. As noted above, for “well” I’d use the prefix mai- from the root MAY (S. mae-).
For example:
- laquetina “not said”
- úquétima “unpronounceable”
- avaquétima “forbidden to say”
- maiquetina “well-said”
This pattern can also be followed in Sindarin:
BĀ: Like Quenya, derivatives of this root can be used for volitional negation. As noted above, S. ava was fossilized as a negative adverb, functionally similar to Q. vá. The interjection S. baw (I refuse!) could only be used as an interjection or in answer to a question. The Sindarin verb for “to refuse, forbid” was boda-. The Sindarin prefix av- was sometimes used in the same way as Q. ava-, but less commonly.
LĀ: Tolkien never developed a full LĀ-paradigm for Sindarin. Almost all the attested Sindarin forms use u- for simple negation, generally as a verbal prefix. Some Neo-Sindarin authors coin a neologistic verb ᴺS. law-, and use the same triple-negative paradigm as in Quenya. I used to object to this use, but it’s become so widespread that I’ve come to accept it. It is especially popular among those who feel that Ū can no longer be used as a general negative.
Ū: If I had my way I would keep only ú as the Sindarin negative, assuming it weakened from a “strong” negative to a general purpose one, with LĀ falling out of favor. I think this is the pattern most consistent with the attest negatives in Sindarin. However, I am very much in a minority position here: most Neo-Sindarin authors have adopted LĀ (ᴺS. law) as the basis for ordinary Sindarin negations, based on notes published in PE22. I’d still retain ú as a strong negative in Sindarin, as Tolkien did.
This is essentially the Sindarin negative paradigm Fiona Jallings presents in her book: A Fan’s Guide to Neo-Sindarin.
All of this is essentially a compromise position. Rather than choosing between Ū and LĀ, I retain both with slightly different meanings. People who prefer one of over the other can still use their preferred negations and mostly be understood. Those who prefer Ū over LĀ can be considered to be speaking a Neo-Eldarin dialect where this strong negation has weakened to the point where it can used as an ordinary negative.
Addendum: When I first published this article I argued in favor for retaining ALA “good, blessed, fortunate”, but I’ve since decided it conflicts with too many of the common uses of the prefix al- in Neo-Eldarin, and I now consider it untenable. I now advocate using Q. mai- and S. mae- as replacements. I also argued against using ᴺS. law as a derivative of LĀ for negation in Neo-Sindarin, since it is completely unattested. But hardly anybody agreed with me, so I’ve abandoned that position.