Soft Mutation in Sindarin Grammar

[Originally authored by Paul Strack in 2026 as a draft entry for Eldamo]

Overview

Soft mutations are a significant feature of Sindarin grammar. Tolkien described their grammatical function at some length in a document referred to as Sindarin Mutations written in 1969 (SM69), where he said:

A feature of Sindarin grammar was the mutation of the initial consonants of words. This was due to two things: the divergence between the development of consonants medially and in absolute initial position; and the close phonetic association of words in groups that were closely connected syntactically. The initial consonants in such groups tended to develop as they would medially in an integral word (PE23/143).

And also:

... only in S. were the consonantal developments medially and in absolute initial position markedly different ... In S. the initials were treated as if in compounds. But these originally purely phonetic developments became features of grammar, and were in many points modified and formalized (PE23/144).

Thus consonantal mutations were originally a purely phonetic phenomenon, the result of different phonetic developments at the beginning versus in the interior of words. Certain closely associated words developed phonetically as if they were in a compound, so that the initial consonants of the second word developed differently then they would as an independent word. Over time, these sound changes were formalized and regularized, becoming features of Sindarin grammar.

Tolkien gave an example of this phenomenon in a 1972 letter to Richard Jeffery:

... palan-tîriel should phonetically > -thíriel [a nasal mutation], past participle “having gazed afar”; but grammatically before actual forms of verbs, the soft mutation only was normally used in later S., to avoid the confusion with other verb stems (Let/427). [The result was palan-díriel as in The Lord of the Rings].

In the aforementioned 1969 notes (SM69), Tolkien gave the following conditions as the main cases where mutations occurred:

  1. Nouns preceded by an adjectival or pronominal modifier.
  2. Adjectives following a noun or pronoun that they modified.
  3. Nouns following verbs as a direct object, or adjectives in similar position after a predicate verb like “appear, seem, become”.
  4. Verbs immediately following their subject.
  5. Proclitic words like prepositions or conjunctions.

In the last case, the preposition/conjunction can mutate the following word, and can themselves be mutated by the preceding word. Two additional cases were not mentioned in SM69, but appear regularly in Sindarin phrases elsewhere in Tolkien writings.

Definite Articles

The singular definite article i “the” induces soft mutation on a following word, but the plural definite article i(n) induces nasal mutation. See the entry on the definite article for further discussion.

Nouns Following Adjectives

According to SM69, nouns following adjectives or other modifiers are mutated. This mutation is largely hypothetical, because the usual position of adjectives in Sindarin is after the noun they modify, as noted by Tolkien in various place: “The adjectival element usually follows in Sindarin” (PE17/36) and “In S. the simple genitive was usually expressed by placing the genitival noun in adjectival position (in S. after the primary noun)” (RGEO/67).

There are, however, a few specific examples of an adjective preceding a noun in word phrases. In SM69, Tolkien mentioned that er “one, sole, only” always preceded the noun (PE23/142). There is another example from the Túrin Wrapper written around 1950:

This phrase is untranslated, but Carl Hostetter suggested il chem could mean “all hands”, where chem is a mutated plural of cam “hand” (VT50/22, 24). Regardless of its meaning, this word was originally given as unmutated cem, so the mutation appears to be c → ch. This is not a soft mutation (c → g), so very likely it is a liquid mutation induced by the l of the preceding word. It is possible that er “one” also induces liquid mutation when preceding a noun: *er cham “one (single) hand”.

If this is true, it seems that the mutations caused by adjectives preceding nouns are more conservative phonetically than adjectives following nouns, since adjectives following nouns pretty consistently undergo soft mutation regardless of the historical phonetic environment. It could be that these pre-noun adjectives function more like prepositional mutations where the exact mutation might depend on the particular word.

Regardless, I would assume these specialized mutations are limited to adjectives that always preceded the noun, such as er “one” and [maybe] il “all”. Other adjectives likely induced soft mutation on the following noun, as they often did in compounds. Compare: pêl galen “a green field” (normal order) vs. calen bêl “a green field” (abnormal order, perhaps used in poetry) vs. Calembel “Greenham” (as a compound word).

Nouns Following Nouns (Juxtapositional Genitives)

In SM69, Tolkien did not list the mutation of nouns after other nouns as one of the main grammatical mutations of Sindarin. However, he did say:

In the case of nouns and adjectives where these were placed together, if they were in any close syntactic relation, the second had a mutated initial according to phonetic development; but this was not always observed in practice ... when either a) the adjective was placed second, or b) two nouns were juxtaposed, there was hesitation; and in the Exilic period the regular usage adopted was not always the phonetic one (PE23/143).

The most common reason for noun-noun juxtaposition is genitive construction, what I refer to as the juxtapositional genitive. It is well-known that, by the 1950s and 60s, such juxtapositional genitives did not involve mutations. Compare: aran cîr lim “king [of] swift ships” with no mutation vs. aran na chîr lim with mutation due to genitive proposition na(n) (PE17/147).

However, this was not always the case. In the 1930s and 40s, it was normal for the second noun in a juxtapositional genitive to undergo soft mutation. Compare 1940s Ennyn Ðurin Aran Vória (TI/182) vs. 1950s Ennyn Durin Aran Moria (LotR/305). When Tolkien wrote in SM69 that if “two nouns were juxtaposed, there was hesitation” he may have meant either (a) that these nouns were mutated archaically or (b) that he was considering (partially) restoring the 1940s system were noun-noun combinations sometimes mutated.

See the discussion of the juxtapositional genitive for more details.

Adjectives Following Nouns

The mutation of an adjective after the noun it modifies is well established in Sindarin and its conceptual precursors. This is more or less always a soft mutation (but see below for a possible exception). Examples include:

There are also numerous examples in names, such as Cûl Veleg “Big [beleg] Load” (RC/536), Ered Wethrin “Shadowy [gwathren] Mountains” (S/106; VT42/9), or Parth Galen “Green [calen] Sward” (LotR/417). As the first few of examples illustrate, demonstratives like sen “this” and possessive pronouns mín “our” also mutate like adjectives.

Unfortunately, there are quite a few examples where adjectives do not mutate, including many from The Lord of the Rings itself. For example Imloth Melui “Lovely [melui] Flowery Vale” (LotR/866; RC/582), Fen Hollen “Closed [hollen] Door” (LotR/826; PE17/98), and [N.] Nelig Myrn “*Black [morn] Teeth” (WR/113), the last of these from Lord of the Rings drafts from the 1940s. Tolkien seems to have noted this discrepancy in SM69, as mentioned above:

In the case of nouns and adjectives where these were placed together, if they were in any close syntactic relation, the second had a mutated initial according to phonetic development; but this was not always observed in practice ... when ... the adjective was placed second ... there was hesitation; and in the Exilic period the regular usage adopted was not always the phonetic one (PE23/143).

In Common Eldarin Article (CEA) also written in 1969, Tolkien seems to have decided the mutation of m was often omitted as a way of explaining such variations in The Lord of the Rings (PE23/136 and note #9; PE23/138 note #12). However, I am of the opinion that both these 1969 notes were rationalizations after the fact, and that the real reason for the variations is that there was a period in the 1940s when Tolkien was writing The Lord of the Rings that adjective mutation was not a feature of the language. See my essay on Adjectival Mutations in Sindarin for further discussion.

Adjectives Following Other Adjectives

In SM69, Tolkien said that where a list of words appears, only the first word in the list undergoes grammatical mutation:

When full words [Tolkien’s footnote: that is all words other than unstressed pronominal and adverbial particles] were placed in a list, or in apposition, the initials of the second, third, and so on, were unchanged.

This indicates that in a list of adjectives, only the first adjective undergoes grammatical mutation as a modifier of its noun. It is possible that adjectives later in the list might still be mutated for other reasons, such as by a preceding conjunction:

Here the first adjective baran undergoes soft mutation after the noun, but later adjectives do not. In this example, calen undergoes sibilant mutation due to the influence of preceding a(h) “and”. However, Vyacheslav Stepanov has argued that “and” should be omitted in such lists based on examples in King’s Letter, so it may be equally valid to say coll varan, caran, calen “a brown, red, [and] green cloak” with no mutation at all other than the first adjective.

As a possible counterexample, consider anno ammen sír i mbas ilaurui vín “give us this day our [mín] daily bread” from a 1950s Sindarin prayer (VT44/21), where the possessive adjective mín “our” is mutated after the regular adjective ilaurui “daily”. See the next section for discussion.

Adjectives And Possessive Pronouns

There is one phrase from the King’s Letter that is especially peculiar:

Here the adjective pan “all” is mutated, but it seems to be a nasal mutation rather than the expected soft mutation. The most obvious explanation is that this mutation is induced by the preceding possessive pronoun în. This may be an exception to the previous rules, and that (a) an adjective following a possessive pronoun is still mutated even though it does not directly follow the noun that it modifies but (b) that mutation is a nasal mutation, perhaps resulting from the fact that all possessive pronouns end in n.

Another example of peculiar behavior involving possessive pronouns is mbas ilaurui vín as mentioned above, where the possessive pronoun mín is still mutated despite not immediately following the noun it modifies (VT44/21). As originally written the phrase had unmutated mín, which was then revised to mutated vín (VT44/28). These two examples imply that possessive pronouns do not behave like general adjectives. It may be that (a) possessive pronouns still mutate even if they do not directly follow the noun they modify and (b) induce nasal mutation on any adjectives that happen to follow them.

Note, however, that this hypothesis is based only on a couple of examples from the 1950s without any other supporting evidence, and must therefore be considered speculative. In particular, it may be that did Tolkien did not introduce this mutation rule for word lists until 1969. Furthermore plural adjectives underwent nasal mutations in the Early Noldorin Grammar of the 1920s (PE13/124), and mhellyn în phain could have been a transient restoration of this mutational rule instead of a specialized behavior after possessive pronouns.

Direct Objects

The mutation of direct objects following a verb was another example of grammatical mutation in SM69:

For instance: the mutation of the initial consonant of a noun following a finite verb was originally phonetic and a result of position not of its syntactic function; but it became a sign of its being the object. Hence if a noun subject was placed after a verb (a position that in S. it often adopted) it was not mutated (PE23/143).

As this statement indicates, nouns in other grammatical roles (subjects, vocatives, etc.) were not mutated when immediately following a verb. Despite Tolkien’s claim that direct objects only mutate after finite verbs, there is at least one example of direct object mutation after an infinitive/verbal noun:

Tolkien’s statement that mutation “became a sign of its being the object” also hints that a direct object may still be mutated even if not following directly after a verb. There is some evidence of this as well, the best example being:

In this example, the direct object hain precedes the verb, but is still mutated. Since it has no particular reason to be mutated in this position, being the direct object is the likeliest reason for this mutation.

As with lists of adjectives, it is also likely that only the first in a list of objects following a verb was mutated: anno annin vegil, cú, [a] thand “give me a sword [megil], bow, and shield” (as noted above, a “and” might be omitted in longer lists). However this may not be the case where the direct objects are displaced elsewhere, as possibly demonstrated by Daur a Berhael, Conin en Annûn, eglerio “Frodo [Taur?] and Samwise [Perhael], Princes of the West, praise [them]” (LotR/953).

For further discussion, see the entry on direct objects and for a “and”.

Verbal Nouns as Objects

In cases where the object of a verb is another non-finite verb (a verbal noun functioning as an infinitive), the verbal noun does not mutate. There are several clear examples:

Thus verbal nouns used as infinitives are an exception to the rule that the object of a verb mutates.

Verbs Directly After Subjects

As mentioned in SM69, mutation occurs for “verbs following their subjects, nouns or emphatic pronouns” (PE23/142). There are a couple clear examples of a verb being mutated when immediately following its subject, both from the late 1960s:

There are other examples, however, where a verb follows its subject and is not mutated:

In the first example, the prepositional phrase o Eregion intervenes between the subject and the verb teithant. In the second example, this phrase is preceded by ir Ithil ammen Eruchín “*when[?] the Moon, for us, the Children of God”, so it may be that the actual subject is Ithil “Moon” and menel-vîr is descriptive of Ithil rather than functioning as the subject of síla. In the last example, the more literal translation of silivren penna míriel is “glittering slants-down sparkling like jewels” (RGEO/64), where silivren is an adjective rather than a noun so that “glittering crystals” is a very loose translation. The actual subject appears in the next phrase: aglar elenath “glory of the starry host”.

Based on the above, I would assume that a verb mutates only when it immediately follows its subject. Where other words intervene, the verb is not mutated. However, it is also possible that the notion of subjects mutating the following verb was itself not introduced until 1968-69, and these older examples predate this grammatical rule.

Prepositions and Conjunctions

As noted in SM69, mutation occurs with “various other proclitic words, such as prepositions and connective words (e.g. and, or), before the following word” (PE23/142). Many prepositions induce soft mutation, in particular those ending in vowels. However other preposition and conjunctions induce different mutations, depending on their phonological character. For example, the prepositions na “to” and mi “in” induce soft mutation, but na(n) “of, with” induces nasal mutation while a(h) “and” induces sibilant mutation.

Furthermore, prepositions themselves may undergo mutation when they are part of a phrase modifying a preceding noun or verb. There are many specialized cases, discussed in more detail in the entry on preposition mutations.

Adverbs and Mutations

Although not directly mentioned by Tolkien in SM69, there is ample evidence that adverbs induce mutation on a following verb or adjective in Sindarin. Examples include:

The first two examples are of an adverb inducing soft mutation on a following verb. The second two examples are of an adverb inducing soft mutation on a following adjective, tiriel “gazing” [from tir- “watch”] and covannen “met” [from covad(a)- “meet”]. Given palan-diriel, it seems adverbs always induce soft mutation, even in cases where this did not match the historical phonology. The example mae govannen is more complex, since Tolkien considered a large number of possible explanations for this phrase. The earliest instance of this phrase was [N.] mai govannen “well met” from the 1940s (RS/194), where the verb was likely based on govad- “meet” = “together-go” and thus did not involve mutation at all.

There is little evidence, however, that adverbs themselves mutate after verbs. The likeliest example of adverb mutation is the phrase menel-vîr síla díriel “*heaven-jewel shines having watched [tíriel]”. Strictly speaking, tíriel is an adjective rather than an adverb, but it is known that adjectives can be function as adverbs within phrases (PE17/18), which appears to be the case here. However, there is a similar phrase silivren penna míriel “glittering slants-down sparkling like jewels” where the adverbial adjective did not mutate (LotR/238; RGEO/64).

There are a couple more examples of clear adverbial non-mutation:

In the first example, a word (ammen) intervenes between the verb and adverb, but in the second the adverb sí appears immediately after the verb, which is the position where it is most likely to mutate from a phonological perspective. Another ambiguous example is the following:

It is possible that, when Tolkien originally conceived of this phrase, hi was a mutated form of S. *si “now”, cognate to Q. sí of the same meaning. But in Tolkien’s later writings, he made it clear that the unmutated form of this adverb is actually hí “now” (PE17/27; VT49/34 note #21). It is possible that hí was unmutated even when it was first written in the late 1940s, given primitive ᴹ√khi- in notes on Demonstrative, Relative, and Correlative Stems (DRC) from 1948, where in time-reference hi [along with si] “refers only to the actual present of the time of speaking” (PE23/96-97).

Taken together, I think it is safest to assume that adverbs do not mutate when following verbs, except possibly adjectives functioning as adverbs. An adverb might mutate if it directly follows a preposition or conjunction, however, or if it is precedes an adjective but follows a noun: ost vae-garnen “a well [mae] made [carnen] city”

Conceptual Development

There are some things in Tolkien’s earlier writings that differ from the Sindarin system of grammatical mutations described above. There are not enough examples from the 1910s-1940s to compare every detail, however.

Gnomish (1910s)

The only mutation clearly described in the Gnomish Grammar of the 1910s was soft mutation (GG/7). The definite article i “the” induced soft mutation, as did the genitive article na “of the” (GG/9). Juxtapositional genitives caused soft mutation: Fôs ’Almir “Bath of Flame [Galmir]” (GG/12). Singular adjectives mutated after nouns, but plural adjectives did not: mab ’loss “white [gloss] hand” vs. mabin glossi “white hands” (GG/15).

Prepositions and other particles also seem to have induced various mutations. For example the negative particle u induced soft mutation: u ’wirn u ’wethrin “not unwelcome or welcome” (GL/47). The patronymic si(n)- “granddaughter of” induced nasal mutation: si·Pheleg “Granddaughter of Beleg” (GL/67).

Early Noldorin (1920s)

In the Early Noldorin Grammar of the 1920s, the singular article i induced soft mutation, while the plural article induced nasal mutation (PE13/120). Singular adjectives underwent soft mutation after a noun: i·mab buig “the clean hand [puig]” (PE13/124). But plural adjectives usually underwent nasal mutation: i·maib fuig(ir) “the clean hands”. Prepositions, particles, and conjunctions induced a variety of mutations. For example o “to” induced stop mutation (PE13/120, 151).

Noldorin (1930s-1940s)

Singular and plural articles continued to induce soft and nasal mutation, as they did in Sindarin of the 1950s and 60s. Juxtapositional genitives induced soft mutation in this period as was noted above, a process that was largely abandoned in Sindarin of the 1950s and 60s. There are examples of adjective mutation in the period, but as I have argued elsewhere I think there is a good chance that there was a period in the 1940s while Tolkien was writing The Lord of the Rings when adjective mutation was not a feature of the language.

Neo-Sindarin

The system I recommend for grammatical mutations in Sindarin closely follows the system Tolkien described in his document on Sindarin Mutations written in 1969 and published in 2024 (SM69: PE23/142-144). Summarizing the discussion above, I would make the following recommendations for grammatical mutations in Sindarin:

Of these rules, the last one on the mutation of verbs after subjects has only recently been widely accepted, having been made clear in SM69 first published in 2024 in PE23. As such, this rule is rarely followed in (Neo) Sindarin examples from 2024 or earlier, including many of the examples in this Eldamo lexicon. I personally am a bit hesitant to fully accept this rule, since there is still a good chance it only appeared in 1969 and may not fit with Tolkien’s earlier conception of the language. However, this rule will probably be the standard in Neo-Sindarin going forward, unless new information is published that contradicts it.


Revision #14
Created 2026-06-07 14:26:44 UTC by pfstrack
Updated 2026-06-08 19:00:35 UTC by pfstrack