Polysyllabic nouns whose last syllable contains a short vowel are considered strong nouns. Strong nouns generally alter this final vowel, the so-called “variable vowel”, when declined into various noun cases (SD/430). Primitively, strong nouns were derived from the full-form of a triconsonantal-root or one of its simple variants, but by the time of Classical Adûnaic, the strong declensions applied to most nouns with a short final vowel (SD/425).

The strong nouns could be further subdivided into two groups (SD/430), those ending in a single consonant preceded by a short vowel (Strong I) and those ending in a short vowel preceded by one or more consonants (Strong II), for example: sulum “mast” (Strong I), lômi “night” (Strong II), zimra “jewel” (Strong II). Often the second-to-last syllable of a Strong I noun has the same short vowel as the last syllable, so that it undergoes the Adûnaic syncope when a suffix is added, losing its final vowel.

Neuter Nouns

With neuter strong nouns, the variable vowel changes when declined in the subjective and objective cases, as well as for plural nouns (SD/430-1). Since the vowels [ē] and [ō] are always long in Adûnaic, the only possible variable vowels are a, i and u.

• Subjective: The variable vowel is a-fortified, changing as follows: a ⇒ â, i ⇒ ê and u ⇒ ô. See the discussions under a-fortification and vowel-combinations for the phonetic developments leading to these changes.
• Objective: The variable vowel is changed to u. If the variable vowel is already u, there is usually no change (but see below).
• Plural: The variable vowel is changed to î. For the subjective plural, the subjective inflection -a- is added as a suffix to the plural form in the same way as it does for weak nouns. For Strong II nouns ending in a vowel, the usual glide-consonant y is used to separate the final plural î from the suffixal -a, and the result is -îya.

For the dual form of strong nouns, the usual dual suffix -at is added. For most Strong I nouns, the Adûnaic syncope also applies so that the variable vowel is lost. For Strong II nouns ending in -i or -u, the appropriate glide-consonant is inserted between the variable vowel and the suffix, y or w respectively. If the dual noun is used in the subjective case, the a in the dual suffixed is lengthened to -ât, as if it were a singular strong noun being declined into the subjective.

The full set of declensions is:

Neuter Nouns
Case	Strong I	Strong II (-i/-u)	Strong II (-a)
Normal	sulum	lômi	zimra
subjective	sulôm	lômê	zimrâ
objective	sulum (sulmu) [1]	lômu	zimru
Dual	sulmat [2]	lômiyat [3]	zimrât [4]
Dual Subjective	sulmât	lômiyât	zimrât
Plural	sulîm	lômî	zimrî
Plural Subjective	sulîma	lômîya [5]	zimrîya [5]

Some notes:

[1] For Strong I nouns where the variable vowel is a u, there is no change in the objective case. Alternately, Strong I nouns with a final vowel u sometimes form the objective with a syncope (if they allow it) and a suffixal -u: sulum → sulmu (SD/435). This formation can also occur with Strong I nouns that have the vowels a or i, but less commonly.

[2] For Strong I nouns where the second-to-last vowel is either long (târik) or followed by a consonant cluster (igmil), the Adûnaic syncope is not allowed and the variable vowel is retained in the dual (SD/430). This is probably also true when the second-to-last and last vowels are different (e.g. minal), but there are no clear examples.

[3] For Strong II nouns ending in -u instead of -i, the glide-consonant is w instead of y, as usual: ugru → ugruwat. In archaic writing, the dual suffix -at would replace the final -i/-u instead of using a glide consonant. At a later (Exilic?) stage, the dual suffix came to be regarded as -ât, so that the a was long in both the normal and subjective duals (SD/431).

[4] Strong II nouns whose final vowel is -a do not use a glide consonant. The dual suffix -at merges with the final -a to produce -ât, so that the normal and subjective duals are identical (though they were distinct in archaic writings, at the time when the dual suffix replaced the final vowel). Such nouns were so common that at later (Exilic?) stage, the dual suffix came to be regarded as -ât for all nouns (SD/431), as previously mentioned in note [3].

[5] For Strong II nouns, the final vowel in plurals became the plural affix -î. In the subjective case, the subjective inflection -a- was suffixed to this -î, with the usual glide consonant y after the î.

Gendered Nouns

Gendered nouns were usually weak, because most gendered nouns had a long vowel in their final syllable (SD/436). This less true of common nouns, more true of masculine nouns and almost always true of feminine nouns. However, there are a few gendered nouns with short final vowels that are declined as strong nouns, for example: nithil “girl” (Strong I), †naru “male” (Strong II, an archaic form of narû), raba “dog” (Strong II).

The strong gendered nouns are declined in the same way as the strong neuter nouns except for in the subjective case:

• Subjective Singular: The gendered Strong I nouns use the gendered subjective suffixes -an (common), -in (feminine) and -un (masculine), together with the usual Adûnaic syncope (SD/436). For gendered Strong II nouns, the subjective suffix is reduced to -n for all genders. This is because masculine Strong II nouns end in -u, feminine in -i and neuter in -a, so that “replacing” the final vowel with the appropriate suffix is the same as adding -n (SD/437).
• Subjective Plural: The subjective plural uses the gendered plural subjective suffix -im with the usual Adûnaic syncope for Strong I nouns. In the case of Strong II nouns, the final vowel is replaced by -îm (SD/436), or more accurately the suffix -im is added to the normal plural form ending -î.

The full set of declensions for gendered strong nouns is:

Gendered Nouns
Case	Strong I	Strong II (-i/-u)	Strong II (-a)
Normal	nithil	naru	raba
subjective	nithlin [1]	narun	raban
objective	nithul (nithlu/nithli) [2]	naru	rabu
Dual	nithlat [1]	*naruwat (†narat) [3]	rabât
Dual Subjective	nithlât [1]	*naruwât (†narât)	rabât
Plural	nithîl	narî	rabî
Plural Subjective	nithlim [1]	narîm [4]	rabîm

Some notes:

[1] The gendered Strong I nouns can undergo the Adûnaic syncope in the subjective, dual and plural subjective cases.

[2] The objective inflection -u- could either replace the variable vowel (nithul) or suppress it and be added as a suffix instead (nithlu) (SD/436). In the case of feminine nouns, the u of the objective inflection was sometimes replaced by i, because u was considered masculine, so that nithil → (objective) nithli instead of nithlu (SD/432).

[3] In his inflection charts on SD/437, Tolkien only listed the archaic dual forms for gendered Strong II nouns, such as †narat and subjective dual †narât where the dual suffix -at replaced the final vowel. It seems likely, though, that the gendered Strong II nouns went through the same development as the neuter Strong II nouns (SD/431): the final vowel came to be regarded as invariant and a glide-consonant w or y was used. These hypothetical forms marked with an asterisk (*) in the chart above.

[4] In the plural subjective, the gendered Strong II nouns replace the final vowel with a long version of the gendered plural subjective suffix -im: -îm. It seems that in this case (in fact all cases other than gendered Strong I), the suffix -im is added directly to the plural form, so that -î + -im ⇒ -îm.

Archaic Strong Ib and IIb nouns

There were some gendered nouns (and perhaps some neuter nouns as well) that were archaically declined as strong nouns despite having a long last vowel: phazân “prince”, banâth “wife”, zigûr “wizard”, anâ “human being” (SD/437). Tolkien called these declensions the Strong Ib and Strong IIb classes. These were declined as the Strong I and Strong II nouns, except that (1) the Adûnaic syncope could not occur and (2) the objective inflection -u- was inserted as a long û, since it was replacing a long vowel.

By the time of Classical Adûnaic, however, such nouns could use the regular declensions for weak-noun, and it is probably simpler to do this than to remember the details for these separate noun classes. If you are curious about the full declensions for the Strong Ib and Strong IIb nouns, see SD/437 for discussion.