nefesh asher tiga bo vetame'ah (Levit. 22:6) ('A soul which touches it shall become defiled') The word nefesh is usually feminine and in this phrase is treated as such in both verbs tiga and vetame'ah. However the verse goes on to say velo yochal … rachatz besaro ('and shall not eat … [until] he has washed his flesh') switching to masculine. The second part of the verse views t he physical person as the subject and therefore uses the masculine gender. Indeed it continues in the masculine to the end of verse eight.

This kind of changeover - where the referent is a new subject that has not been mentioned explicitly - can also be found in the language of the Mishna. For example mi shehayu shetei kittei eidim me'idot oto (Nazir 3:7) ('Someone who had two sets of witnesses testifying') kittei ('sets') is the subject here and it is feminine. Hence shetei and me'idot are also feminine. However the Mishna moves on to speak about the (male) witnesses and uses the masculine eilu me'idim ('these testify').

Silent Alef and Pronounced Alef

ladoshem (the tetragrammaton - four letter name - with prefix Lamed read as Shem Adnut Levit. 22:21 etc.) According to the Masorah there is no Sheva under the Yud of the tetragrammaton to indicate to the reader that the Alef of the name Adnut is to be regarded as silent (can one say 'read as silent?') and this is evidence in favour of the view of R' Shabbetai Sofer who explains that this Sheva indicates the Chataf Patach of Shem Adnut (Siddur, General introduction Ch. 3). The issue of Alef in the noun adon (both holy and ordinary) is discussed in the Masorah as follows:

There are seven Alefs that are pronounced and they are (1) vaadonei (Deut. 10:17); (2) laadoneihem (Gen. 40:1);
(3) laadoneihem (Amos 4:1); (4) vaadoneinu (Kings I 1:11); (5) vaadoneinu (Psalms 135:5); (6) laadonei (Psalms 136:3);
(7) laadoneinu (Neh. 8:10) (the small 'a' indicates the Chataf Patach under the Alef) (Masorah Psalms 136; Amos 4)

It follows that elsewhere the Alef of adon will be silent in those positions where it would normally take a Chataf Patach after one of the prefixes Bet, Chaf or Lamed. There are also other cases of silent Aleph at the beginning of words, e.g. bazikim (Jer. 40:1), batar (Dan.7:6-7), leimor (throughout). R' Shabbetai Sofer wrote about this "He who fears the word of G-d will be careful about this and will also warn others, and then Hashem Yitbarach will accept our prayers and say 'Enough!' to our troubles."

R' E. Sternbuch asked: Why did the Masorah not include laadon (Micha 4:13) as having a sounded Alef? R' Shabbetai Sofer explains that laadon has a Chataf Patach under the Alef because of semichut (Siddur p. 223). This explains why the Kamatz under the Alef in the absolute (ordinary) form has changed. However, he does not address R' Sternbuch's question. The editors of R' Shabbetai Sofer's Siddur (p. 99) also mention this irregularity, but do not explain it. I would like to suggest that the 'seven pronounced Alefs' mentioned in the above Masorah are all in words of two syllables or more (Chataf vowels are not true vowels) and therefore the occurrence in Michah being a one-syllable word is of no concern to the Masorah. (The name Adoniyahu (Kings I, Ch. 1) also keeps its Chataf Patach after the prefix Lamed. Presumably these rules do not apply to proper nouns.)

It would seem that the same explanation will help us understand the difference between Elokim and Eloak. Elokim has a silent Alef after the prefixes Bet, Chaf, Lamed. It has more than one syllable. On the other hand, Eloak only has one syllable - the Patach under the Heh is known as a Patach Genuva (furtive Patach) and is also not a true vowel. Its purpose is to ease the vocal tract in its transition from other vowels to the guttural position for Heh, Chet, Ayin. It therefore does not have the standing to establish an independent syllable. As a result in the singular form Eloak, the Alef after the prefix Lamed retains its Chataf Segol and is pronounced (Job 12:4; 36:2). However, batar and leimor do not support my suggestion.