yishkon-sham he'anan (Num. 9:17) ("the cloud rested there'); yishkon he'anan (Num. 9:18) ("the cloud rested'); In the first yishkon the [o] is indicated by a Kamatz Katan. This is a short vowel; not only Ashkenazim read it as [o] but Sefaradim do too. In the second yishkon the [o] is indicated by a Cholam which is a long vowel. Why is this?

The first yishkon- is followed by a Makaf. This looks like, and functions very similarly to, a hyphen connecting it to the following word, so that yishkon- has no independent tune or stress. As a result the syllable -kon- is an unstressed closed syllable, and a short vowel - such as a Kamatz Katan - is appropriate. However, the second yishkon is not hyphenated to the following word, it is free standing. It is true that its tune - Darga - is a service tune indicating that it should be read flowing on to the following word. However this connection is not as intimate as that created by Makaf.

The Disappearing Dagesh

vayisu (Num. 10:33) ('and they traveled') The root is Nun, Samech, Ayin and with the elision of the Nun, it would be expected that the Samech receive a Dagesh. Indeed in pausal positions the same word is repeatedly found with a Dagesh in the Samech, eg vayisa'u (Gen. 35:5 and elsewhere). It seems that the Dagesh is omitted in cases where the letter has a Sheva followed by a guttural letter. Thus from the root Lamed, Kuf, Chet, ('to take') we obtain yikchu (Gen. 14:24 and elsewhere) (missing Lamed; Kuf with Sheva; followed by guttural Chet and no Dagesh), and in the pausal position yikachu (Lev. 21:7 and elsewhere) (missing Lamed; and Dagesh in Kuf). Similarly from the root Nun, Samech, Alef, (to carry) we obtain yisu (Exod. 28:43 and elsewhere) (missing Nun; Samech with Sheva; followed by guttural Alef and no Dagesh), and in the pausal position yisa'u (Lev. 20:19 and elsewhere) (missing Lamed; and Dagesh in Sin).

Segolate nouns

kizra-gad (Num. 11:7) R' A. ibn Ezra was faced with two problems: 1) there is no other occurrence of this word in Scripture 2) to regard it as a form of zera, shortened due to Semichut, does not solve the problem as zera is a Segolate noun (a group of two-syllable nouns whose second syllable is unstressed and in many cases has a Segol; however in our case it has Patach due to the occurrence of Ayin as the final letter). Generally the vowels of Segolate nouns do not change in Semichut. To deal with problem 1) R' A. ibn Ezra writes that the form z'ra is semichut ('the construct state') i.e. it is a form of the word zera ('seed'); hence kizra-gad ('like coriander seed'). To deal with problem 2) he points out that a similar form can be found in n'ta shaashu'av (Isai. 5:7) i.e. n'ta is a special shortened Semichut form of neta (again Ayin is the final letter). He thereby establishes that shortened Semichut forms may occur with Segolate nouns. It should be noted that a small number of Segolates regularly have shortened Semichut forms e.g. sheva ('seven') - sheva-esreh ('seventeen'); cheder ('room') - chadar mishkavaecha ('your bed-rooms'); sheger ('expelling') - shegar-alafaecha ('offspring').

A reader pointed out that when I quoted yavi oto last week I wrote that it is not clear who is 'brought'. The Rashbam and others speak of the sacrifice i.e. 'it'. Meshech Chochma speaks of 'the hair'.