• Voiceless Stops (p, t, c)
• Voiced Stops (b, d, g)
• Voiced Stop Clusters
• Ancient Nasalized Stops ([m]b, [n]d, [n]g)
• Voiceless sounds (h, wh, lh, rh)
• Preceding Nasal Effects
• Preceding Nasals in CEA
• Long Nasal Mutations
• Summary
• Conceptual Development

Nasal mutation is the second most common form of mutation in Sindarin. The best description of Sindarin nasal mutation appears in notes on the Common Eldarin Article (CEA) from 1969 (PE23/133-140), and the discussion below mostly follows the CEA system of mutations, except where that system is contradicted by examples in The Lord of the Rings (mostly for voiceless stops; see below).

Nasal mutation is the result of a preceding nasal n, and its most common use is with the plural definite article in. It also appears with certain prepositions, most notably na(n) “with, of” and (probably) an “to, for” which is the basis of the Sindarin dative. Thus i thiw “the signs (têw-plural), a Pherhael “to Samwise (Perhael)”, Taur-na-Chardhîn “Forest of [na(n)] Southern Silence (Hardhîn)”.

The nasal mutations of Sindarin are distinct from those of Welsh. In particular, the nasal mutation of voiceless stops (p, t, c) are voiceless nasals in Welsh (spelled “mh, nh, ngh”), whereas they become voiceless spirants in Sindarin (ph, th, ch). Like Welsh, the nasal mutation often modifies the preceding nasal itself, causing it to vanish or be altered. This is discussed in the section on Preceding Nasal Effects, below.

The second-best description of Sindarin nasal mutation in some brief notes from 1957, mainly contrasting it with soft mutation (PE17/147):

na (< nā) “to, towards” of space/time. with vocalic mutation. before vowel n’ ... S na, before vowels nan (nasal mutation), means “with” in sense of possessing, provided with, especially of characteristic feature. Orod na Thôn “Mount of the Pine Tree(s)”. na “to” and na “with” are therefore distinct before vowels and b, d, g, p, t, c, m, s but same before h, f, þ, r (rh), l (lh). Late forms as nan-h as for vowels, archaic nath-r, nath-l for nan-rh, nan-lh. nan|sr > nassr > nathr. nað before r (nan-r > nađr).

Here Tolkien is contrasting two prepositions, na “to” and na(n) “with”, which took the same form before most consonants (na) but the first caused soft mutation whereas the second caused nasal mutation. This produced different mutations for b, d, g, p, t, c, m, s but the same mutations (or lack thereof) for h, f, th, r, l, rh, lh. This 1957 system of mutation appears to (mostly) align with the more comprehensive description from CEA, as discussed in more detail below.

Voiceless Stops (p, t, c): The nasal mutations of voiceless stops p, t, c are the voiceless spirants ph, th, ch. These are a result of the sound change whereby voiceless stops became spirants after nasals. The nasal mutations are not simply the same as the medial developments, however, since medial nasal-spirant combinations became long voiceless nasals and ultimately long nasals, so that medial mp, nt, nc > mph, nth, nch > mm, nn, ng. Indeed, this is how the Welsh nasal mutations ended up being voiceless nasals. However, voiceless spirants are well attested as the nasal mutations of voiceless stops in Sindarin and its conceptual precursors, dating back to the Early Noldorin of the 1920s (PE13/121). Some Sindarin examples:

• Perian “Halfling” → i Pheriain “the Halflings” (LotR/953).
• têw “letter” → i thiw “the letters” (LotR/305; PE17/44).
• cair “ship” → na[n] chîr “of ships” (SA/an(d); PE17/147).

Thus, there must have been some additional phonological effect in play with the development of the nasal mutation in Sindarin. In his book Gateway to Sindarin, David Salo proposed (GS/63 §4.187) that nasals vanished before morpheme boundaries, which also helps explain the soft mutation of ancient nasalized stops and the nasal mutation of voiced stops. See the relevant phonetic entry for further discussion. Nasal mutations of voiceless stops in CEA depart subtly from the pre-1969 examples, as discussed below.

Voiced Stops (b, d, g): In CEA, the nasal mutations of voiced stops b, d, g are simple nasals m, n, ñ, with ñ often spelled ng instead.

• bâr “dwelling” → i·mair (PE23/139).
• dôr “land” → i·nuir (PE23/139).
• gôn “stone” → i·ñuin or i·nguin (PE23/139).

Some examples prior to CEA are:

• Taur-i-Melegyrn “Forest of the Great (beleg) Trees (orn-plural)” (WJ/185).
• Hauð i nenghin “*Mound of the Slain (dangen-plural)” (PE17/133); in this note dangen is derived from ✶dankĭnā rather than *ndankĭnā.
• Caras (i)Ngelelaið “City of the Trees (galadh-plural)” (PE17/60) [sic., probably a slip for Ngelaið as suggested by Christopher Gilson].

Some Noldorin examples from the 1930s and 40s indicate Tolkien also considered having nasal mutation preserve nasal-stop clusters or long nasals.

• N. Cerch iMbelain “Sickle of the Gods (balan-plural)” (Ety/KIRIK).
• N. i·nnýr “the lands (dôr-plural)” (PE22/33).

The second example is a bit peculiar, because dôr was usually derived from ancient ✶ndorē, but a nearby soft mutated definite singular form i·ðór “the land” makes it clear that in that moment Tolkien imagined it was derived from primitive simple d. Thus the example nnýr is, in fact, representative of the nasal mutation of simple d rather than ancient nasal stop [n]d.

In the Early Noldorin Grammar of the 1920s (PE13/120), however, the results were simple nasals: b, d, g → m, n, ng (the last probably representing [ŋ]). The general consensus among most Neo-Sindarin writers is the most likely nasal mutations of voiced stops are simple nasals m, n, ng or ñ, with ñ being a less ambiguous representation of isolated velar nasal [ŋ] that Tolkien usually wrote as ng. Thus:

• bereth “queen” → *i merith “the queens”.
• dû “night” → *i nui “the nights”.
• galadh “tree” → *i ñelaidh “the trees” (which Tolkien wrote as i ngelaidh).

For this to be true, the phonological origin for the nasal mutations of voiced stops would need to be different from the medial developments of nasal-stop clusters, as was the case for the previously described voiceless stops. In particular, it could be explained if the preceding nasal vanished before the new mutated nasal, a similar development to that seen in the nasal mutations of voiceless stops given above.

The cause could not be quite identical to that of voiceless stops, however, since medial nasalized clusters mb, nd become mm, nn relatively late in Sindarin’s history. It is more likely that the resulting mm, nn (and ng) simply shortened at the beginning of the word, as they did at the end. In CEA, Tolkien stated that this shortening occurred initially in all cases:

This simplification of long (‘double’) consonants was not actually confined to final position. In the spoken language they were reduced initially [emphasis added] and medially after an unstressed vowel followed by the main word-stress (PE23/140).

Voiced Stop Clusters: The nasal mutations of simple voiced stops are dependent on their being isolated between vowels, so that (for example) intervocalic n·b > m·b > m·m > ø·m. This means that, where voiced stops appear as part of clusters, they would resist nasal mutations. The preservation of voiced stop clusters can be seen in:

• gurth an Glamhoth “Death to the din-horde” (UT/39).
• Tawar-in-Drúedain “Drúadan Forest” (UT/319).
• Haudh in Gwanûr “Mound of the Twins” (LotR/1054).
• Bar-in-Gwael, untranslated (WJ/418).

The case for gw- is a bit muddled, in that this initial combination arose from ancient simple w-, a sound change that conceivably could have been inhibited by a closely preceding nasal. The (lack of) nasal mutation could have developed by analogy with other forms, and something similar happened even in compounds like dangweth “answer” = dan + gweth (PM/395). However, it seems Tolkien vacillated on this point, because in the drafts of The Shibboleth of Fëanor from 1968 he had i·Wenyn as the plural of gwanun(ig) “twin” with a vanishing g. Furthermore, in CEA Tolkien indicated that in Gwanur was a variant orthographic representation of i·ñwanur. In that system gw mutated to ñw, often orthographically represented as ngw:

gwend “maiden”, pl. in·wind. The last form is rarely found. The usual form was even in early records i·ñwind (transcribed i·ngwind). This was due to analogy with words beginning with g-, as i·ngelaið “the trees” i·nguin “the stones” ... But initial CE w- had only become gw- in S. in absolute initial position and CE im wendēi should have produced phonetically S *inwendi with differentiation of mw > nw, a much favoured sequence in S. (as in Quenya), normally from CE nw and nm. Cf. also in LR III 335, Hauð in Gwanur “Grave-mound of the Twins” (PE23/139-140).

Based on CEA, I would assume gw does not count as a “cluster” for purposes of nasal mutation and behaves like simple g (by analogy), becoming ñw (also written ngw or n-gw).

Mutations for some consonant clusters are unattested. For clusters beginning with b David Salo suggested these would still undergo nasal mutation (GS/76), but I believe that they would probably resist nasal mutation on the basis of compounds like Nimbrethil. So: *im brennyn “the lords (brannon-plural)”. It is even less clear what would happen with stop-clusters beginning with voiceless stops, since (a) they are rare and (b) do not survive medially. For clusters like pl, pr, tr, cl, I suspect they would not resist nasal mutation and would become phl, phr, thr, chl as in: *i thrync “the stakes (trunc)”.

Ancient Nasalized Stops ([m]b, [n]d, [n]g): Prior to the publication of the Common Eldarin Article (CEA), the general consensus of many Neo-Sindarin writers is that the nasals mutations of words beginning with ancient nasalized stops [m]b, [n]d, [n]g are special, in much the same that their soft mutations are special (producing nasals rather than voiced spirants). For example both David Salo (GS/76) and Thorsten Renk (PESC/120) suggested the nasal mutations of such words restored (or preserved) the ancient nasal cluster. Actual examples are somewhat contradictory, however:

• Hîr i Mbair Annui “Lord of the Westlands” (SD/129).
• dor-i-ndainn “*Land of the Nandor” (PE17/141).
• i·m(b)air en N(d)engin “the houses of the Slain” (PE17/97).

The first two examples seem to indicate that the ancient nasalized clusters mb- and nd- (from roots √MBAR and √NDAN) were retained or restored, but the third example (from roots √MBAR and √NDAK) indicates they reduced to simple nasals. As is the case with simple g-mutations, analysis of ancient ñg- clusters is complicated by Tolkien’s representation of both [ŋg] and [ŋ] by ng. Thus naur dan i ngaurhoth ← gaurhoth from root √ÑGAW (LotR/299; PE17/38-39) could be either [ŋgaurhoθ] or [ŋaurhoθ].

In CEA (written in 1969 and published in 2024), Tolkien did not explicitly describe the nasal mutations of words beginning with ancient nasalized stops, but a few appear among his examples:

• bâr “dwelling” → i·mair (PE23/139).
• dôr “land” → i·nuir (PE23/139).
• gôn “stone” → i·ñuin or i·nguin (PE23/139).

These three words were given as examples of nasal mutations of voiced stops, but the first two are known to be derived from roots with ancient nasalized clusters (√MBAR and √NDOR). This seems to indicate that in CEA, the nasal mutations of voiced stops were simple nasals in all cases, regardless of their etymological origin (whether derived from simple voiced stops or ancient nasalized clusters). As an added complication, in CEA Tolkien stated that such nasal mutations were sometimes orthographically represented as clusters:

for i·mair, i·nuir written forms were sometimes i·mbair, i·nduir to distinguish such cases from those with original initial m, n; they were not needed in the case of nouns beginning with g-, since no words had radical initial ñ in Sindarin, but nonetheless i·ñuin “the stones” was sometimes spelt i·ñguin (PE23/139).

This spelling convention helps distinguish words like i·noer “the fires” (← naur) from i·ndoer “the leagues” (← daur), though the pronunciation would be the same [i·noer]. This rule means that the occasional appearance of nd-, mb- in nasal mutations from Tolkien’s earlier writings could simply be a spelling variation rather than representative of actual pronunciation.

Voiceless sounds (h, wh, lh, rh): The nasal mutation of breath h is attested in Narn i Chîn Húrin “Tale of the Children (hên-plural) of Húrin” (WJ/160). This is identical to its soft mutation, but like the nasal mutation of voiceless stops requires some special explanation. In particular, the older form would have been *in·chīn, since initial h- in Sindarin arises only from an initial voiceless spirant ch (IPA [x]). The normal medial development of -nch- is -ng-, as noted above. Thus, like the nasal mutation of initial c to ch, the nasal mutation of initial h to ch can only be explained by a special historical phonetic development whereby nasals vanished before morpheme boundaries, as suggested by David Salo (GS/63 §4.187, 76). Similar phonetic developments were confirmed in CEA (with complications as discussed below):

Medially and initially in S. sy > ꜧ (voiceless spirantal y); and this eventually was retracted to χ (back spirant identical in sound with the product of CE kh) ... Later again when the χ (ch) derived from kh and initial sk became in absolute initial position weakened to breath h, the newer χ (ch) naturally did the same. Hence arose an apparent mutation of initial h to χ (ch), both after nasals and vowels (PE23/138).

In CEA Tolkien said that something similar happened with the nasal mutation of wh < ƕ < sw (PE23/138), and gave as an example i·chwist as the definite plural of whest “puff, breeze” (PE23/139).

As for for voiceless liquids lh, rh, in CEA Tolkien indicated that the historical phonetic developments would be insl, insr > isl, isr > iþl, iþr (PE23/137), so that the (archaic) nasal mutations would be i·thl, i·thr. But he also said that mutated forms were abandoned in later Sindarin, so that lh, rh became non-mutating (PE23/139):

• rhaw “wild beast” → definite plural †i·thraw, later i·rhaw [it is unclear why the plural is not *rhoe as in naug → noeg].
• lheweg “ear” → definite plural †i·thlewig, later i·lhewig.

Tolkien described similar nasal mutations for the preposition na(n) in Quenya Notes (QN) from 1957, likewise indicating thr mutations were archaic: “archaic nath-r, nath-l for nan-rh, nan-lh; nan|sr > nassr > nathr” (PE17/147).

Preceding Nasal Effects: As noted above, many of the nasal mutations only make sense if at some point in Sindarin’s history the nasal was lost before the mutated consonant. There is ample evidence of this for the plural definite article in, which frequently loses its n before consonants, mutated or not:

• i chîn “the children” (hên-plural)” (WJ/403).
• i mbair “the houses (mâr-plural [mb-])” (SD/128).
• i melegyrn “the great trees (belegorn-plural)” (WJ/185).
• i mírdain “the jewelsmiths (mírdan-plural)” (S/286).
• i Negyth “the Dwarves (Nogoth-plural)” (WJ/338).
• i ngaurhoth “the wolf-horde (gaur-class-plural [ng-])” (WJ/160).
• i ngelaidh [*i ñelaidh] “the trees (galadh-plural)” (PE17/60).
• i Pheriain “the Halflings (Perian-plural)” (LotR/859).
• i sedryn “the faithful (sadron-plural)” (UT/153).
• i thiw “the signs (têw-plural)” (LotR/305).
• i thuin “the pines (thôn-plural)” (PE17/81).

Note how the nasal is lost before most initial consonants even where no mutation occurs: m, n, th, s and probably f and l as well (see below). Prior to the publication of CEA, the nasal effects before f are not attested for the plural definite article in, but they are for the preposition na(n), as in:

• Arthor na Challonnas “Realm of the South-harbourage (Hallonnas)” (PE17/28) with nasal mutation.
• Arthor na Forlonnas “Realm of the North-harbourage” (PE17/28) with no nasal mutation but with lost n.

The full form in appears before vowels and consonant clusters that resist mutation, though in the case of clusters this could be merely orthographic convention rather than the actual pronunciation:

• in Edenedair “the Fathers of Men (Adanadar-plural)” (MR/373).
• in Edhil “the Eldar (Edhel-plural)” (S/286).
• in Ellath “the Elves (†Ell-class-plural)” (VT50/5).
• in Drúedain “the Woses (Drúadan-plural)” (UT/319).
• in Gwanûr “the twins (gwanunig-dual)” (LotR/1054).

Though not attested, probably the semi-vowel i (English “y”, IPA [j]) functions like a vowel, so you would get in ient “the bridges (iant)”, pronounced (in English phonetic spelling) “een yehnt” or in IPA [in jɛnt]. In cases where the nasal is preserved before consonants, it may assimilate to the sound of that consonant.

There are no clear examples of this with the plural article in, but such an assimilation can be seen in am Meril “to Rose (Meril)” in the King’s Letter to Sam and his wife Rose (SD/129). Thus perhaps *im brennyn “the lords (brannon-plural)” assuming the cluster br resists mutation. There is, however, a counterexample to assimilation from CEA: parth “field” → in·pherth (PE23/139); see below for more on the system in CEA.

Cases where the nasal is lost despite the lack of mutation can be explained by various other sound changes:

• The loss before nasals n and m can be explained by the assimilation of in to the following nasal and then shortening of that nasal.
• The loss before th and f can be explained by the same nasal loss at word boundaries seen in the nasal mutations of t and p.
• The loss before s can be explained because [ns] became [ss] which likewise shortened later at the word boundary: in·s > is·s > i·s.
• Before l, we also have a phonetic development where [nl] became [ll], so probably: in·l > il·l > i·l.

This loss of n before l is also necessary if na “to” and na(n) “with” are to be identical before l, as indicated by the 1957 note mentioned above.

One last special case is the transformation of n before r mentioned in this same 1957 note: “nað before r (nan-r > nađr)”. This was the result of the historical change whereby [nr] became [ðr] as in Caradhras “Redhorn” = caran + ras(s). It seems likely that the n in the plural definite article in went through a similar change, so you’d get idh rem “the walls (ram-plural)”. The appearance of ð before r in nasal mutations was also mentioned in CEA (see below).

Preceding Nasals in CEA: The changes to initial consonants in CEA match the discussion above in almost all cases, but there were difference in how preceding nasals were modified. Early in CEA Tolkien indicated that nasal loss before consonants was (nearly) universal, stating:

The forms of the article were:

sg. (a) e (b) e or i (c) en;
pl. (a) i (b) in (c) in.

The forms (a) were those assumed before a word having an initial consonant [emphasis added], other than g, in early (pre-record) Sindarin; (b) before g which later disappeared, after becoming spirantal ʒ; (c) before an original vowel, or one developed from CE syllabic nasals (PE23/135).

When describing the mutations for individual consonants, however, Tolkien said that the nasal was only lost before voiceless spirants derived from ancient aspirates and survived before mutated voiceless stops, saying:

The first change was one common to T. and S. and so very ancient. After the loosening of the aspirates to [voiceless] spirants a preceding nasal was lost, and the vowel before the nasal was nasalized and lengthened. The nasality was later lost (in both T. and S.). The lengthened vowel fell in with the later development of original long vowels, but was shortened in S. in the unstressed proclitic article. So that in result the plural article in > i [before spirants derived from aspirates], leaving the initial consonant unaltered. The spirantalizing of [voiceless] stops after a nasal was however very much later: a purely S. change only affecting the medial stops and not those that had become final after the loss of all original CE final vowels (-mp, -nt, -ñk remained); the nasal was therefore not phonetically lost before the new spirant (PE23/137).

This is consistent with the examples Tolkien gave in CEA (PE23/139):

• falas “beach” → i·felais.
• thoron “eagle” → i·theryn.
• heleg “ice” → i·chelig.
• parth “field” → in·pherth.
• tawar “woodland” → in·þewair.
• calar “lamp” → in·chelair.

Unfortunately these examples contradict the mutations of voiceless stops seen in The Lord of the Rings, notably Perian “Halfling” → i Pheriain (LotR/953) and têw “letter” → i thiw (LotR/305). It is not clear why Tolkien made this change or whether he would have eventually reversed himself. See my essay on Sindarin Articles and Mutations from Parma Eldalamberon #23 for further discussion.

Nasal losses before other consonans in CEA are consistent with nasal mutations see elsewhere, however, and Tolkien described the relevant sound changes in some detail (PE23/137-139). Of particular interest is the development of the nasal before l and r, where Tolkien said inl, inr > ill, iðr (PE23/137), likely with ill shortening to i·l but possibly with iðr surviving as ið·r, especially given its apparent survival for prepositions such ana(n) “of, with” as in nað·r (PE17/147). Unfortunately Tolkien gave no explicit example of a definite plural noun beginning with r in CEA to confirm this.

Long Nasal Mutations: There may be some special nasal mutations for the preposition an “to, for”, based on the example am Meril “to Rose [Meril]” (SD/129) as noted above. The earliest reference I can find this idea is Thorsten Renk’s Sindarin course (PESC/150-151), but the idea may have come from elsewhere (e.g. http://www.sindarin.de/grammatik.shtml#mut_tabelle). Unlike the nasal mutation of in, it seems the nasal of an is preserved before other nasals, albeit with assimilation. There is no good phonological justification for this difference, but there might be a morphological one. In particular, the dative preposition also appears in pronominal forms like annin “to/for me” and ammen “to/for us”; and the preserved nasals might be by analogy with such forms. Assuming this is the case, the n in the dative preposition an would likely be preserved before both original m, n and mutated m, n, ñ that originated from b, d, g:

• mellon “friend” → *am mellon “to/for a friend”.
• narn “tale” → *an narn “to/for a tale”.
• bereth “queen” → *am mereth “to/for a queen”.
• dû “night” → *an nû “to/for a night”.
• galadh “tree” → *an ñaladh “to/for a tree”.

The last of these would probably be pronounced [aŋ·galað]. These long nasal mutations for an have become popular in Neo-Sindarin because they help differentiate phrases like a mellon “and a friend” from am mellon “for a friend” (hat tip to Elaran for this last note).

There might be a similar preservation/assimilation before l as in al lam “to/for a language”. Since the change of [nl] > [ll] is no longer an active sound change in Sindarin (c.f. minlamad), it is also possible the nasal is simply preserved but not assimilated: an lam. It may depend on what the 2nd person plural dative pronoun is: allen (older assimilated) or anlen (modern not assimilated). The current consensus for Neo-Sindarin seems to be al lam and allen.

Summary: To summarize the above discussion, I would use the following for normal nasal mutation in Neo-Sindarin:

• Voiceless stops (p, t, c) become voiceless spirants (ph, th, ch).
• Voiced stops (b, d, g) become simple nasals (m, n, ñ).
• Ancient nasalized stops ([m]b, [n]d, [n]g) also become simple nasals (m, n, ñ).
• In initial clusters, however, voiced stops are preserved: bl, dr, gw, etc.
• Breath h becomes ch, and wh becomes chw.
• Voiceless liquids lh, rh resist nasal mutation in modern Sindarin, but became thl, thr in archaic speech.

In addition to the mutation of the following consonant, the preceding nasal n is altered as follows:

• The nasal is lost before consonants in most cases (i·ph, i·m), except as noted the below.
• It is preserved before vowels (in·e).
• It is preserved before clusters that resisted mutation (in·dr, in·gl).
• For the preposition an, it is also preserved before other nasals, both originals and the results of mutation (an·n, an·ñ).
• When preserved, the nasal assimilates to the following sound (im·br, am·m).
• When preceding r, the nasal n becomes dh (idh·r, adh·r).
• When preceding l, the nasal n in an (but not in) survives and becomes l (al·l).

The last four of these rules are based on some tenuous reasoning, and must be considered speculative.

Conceptual Development: There are no signs of nasal mutation in the Gnomish of the 1910s, probably because in was not the plural form of the definite article (GG/7). Instead, the form in was used (singular or plural) mainly before vowels (GL/50). By the time the Early Noldorin Grammar was written in the 1920s, however, nasal mutation was a feature of the language and it remained thereafter (PE13/120-121). These Early Noldorin nasal mutations were very similar to the ones described above except the nasal mutation of h was (archaically) a voiceless nasal nh, and h was not mutated in the “modern” form of the language. There isn’t enough information about the nasal mutations from the Noldorin of the 1930s and 40s to determine their differences from later Sindarin in any detail.