The Fourth Principle: On the elucidation of the noun patterns whose first vowel is ḥolam.

1. Know that when the first vowel is ḥolam, it changes to a ḥatef qamats with singular pronominal suffixes, in the plural, in the construct state, and with pronominal suffixes—but only on the condition that the second vowel is a segol and not some other vowel. Thus with nouns such as ʾozen (“ear”), qodesh (“holiness”), and ḥodesh (“month”), and the like, one says: ʾozno, qodsho, ḥodsho; ʾoznayim, ʾoznei, ʾoznav, and so forth. But in the singular construct state they do not change, as in ʾozen Aharon (“the ear of Aaron,” Exod. 29:20), and qodesh Adonai (“holy to the LORD,” Exod. 28:36; Lev. 19:8), and similar cases.

2. But when the second vowel is not segol, as in ʾotsar (“treasury”), kokhav (“star”), ʿolel (“child”), and the like, the initial ḥolam does not change at all. As for the qamats or tsere at the end, I have already written the manner of their change in the Second Principle, Section Two. Now, I wrote concerning nouns of six vowel-points, in the Third Principle, Section Seven, that because of ḥet or ʿayin the final segol changes to pataḥ; know that the same rule applies here as well, as in ʾoraḥ (“path”) and kovaʿ (“helmet”), so one says ʾorḥo, kivʿo, and so forth. But when the middle radical is one of the gutturals, the ḥatef qamats moves under the middle radical, and the first radical takes a qamats, as from ʾohel (“tent”): va-yeṭ ʾaholah (“and he pitched his tent,” Gen. 35:21); and from toʾar (“appearance”): mah toʾaro (“what is his appearance?” 1 Sam. 28:14). Yet sometimes the first radical remains with a ḥolam, as in toʾaro mi-benei ʾadam (“his appearance more than that of men,” Isa. 52:14).