hazeih (Num. 8:7) ('sprinkle') R' A. ibn Ezra (11th -12th cent.) writes "Rabbi Moshe [the Kohen 11th cent.] said that the first letter of the root of hazeih is Nun as in hakeh takeh [where the root is Nun, Khaf, Heh] and he is right." This suggestion, that the root of hazeih is Nun, Zayin, Heh, and that the Dagesh in the Zayin is there because of the missing Nun, is part of the general hypothisis of the Hebrew grammarians among whom R' Moshe HaKohen ibn Gikatilla and R' A. ibn Ezra were leading figures. They maintained that (nearly) all roots have three letters. Even the evidence from hakeh takeh that there is a Nun missing, depends on non-verbal forms (Sam. 2, 4:4; 9:3; Isa. 15:3; Psalms 35:15). (The Concordance lists no verbs under the root Nun, Zayin, Heh.) This view has since been almost universally accepted - even the Rashbam, grandson of Rashi, accepted it. The previous approach that roots may have less or more letters, which was propounded by R' Menahem ben Saruk and followed by Rashi, was almost totally rejected.

Not until 1854 was Mahberet Menahem edited and published - in Edinburgh, Scotland, by R' Zvi (Herschell) Filipowski - from five Medieval manuscripts. Filipowski's annotations are in Hebrew, but there is also an abstract in English which he selected and translated to demonstrate that 'the system that had been adopted in former ages in the Hebrew language [is not] inferior to the one of modern times.' He goes on to say that the authors of modern dictionaries (referring, it would seem, to R' Yona ibn Janah and R' David Kimhe) maintain that roots are exclusively of three letters. Even though this is the accepted view, he argues that the discerning Student, on perusing the present Extract, will have to pause before he can decide as to the genuineness of the modern opinion. On a thorough examination he will find, that in numerous cases, preference is to be given to the old system.

Another scholar of the nineteenth century, R' S.R. Hirsch, in his commentary on the Torah, shows that in many places, roots which have two letters in common, also have common elements of meaning. Thus he relates naria (Psalms 95:2) the root of which is Resh, Vav, Ayin to teroeim (Psalms 2:9) whose root is Resh, Ayin, Ayin. Perhaps one may say that there are roots to the roots. Hirsch makes a strong case for the view that the root of three letters (which he accepts) is not the only unit from which Hebrew is built.

What does et mean?

vesheireit et-eihav (Num. 8:26) ('he will serve [with?] his brothers') et can be derived from two different roots and can therefore have different meanings. According to the accepted (three-letter-root) system of grammar these will be unrelated. These roots are Alef, Vav, Tav, from which it is possible to decline oti ('me'), otecha ('you'), etc. accordingly et is the sign of the accusative, and Alef, Tav, Tav, (or Alef, Nun, Tav) from which it is possible to decline iti ('with me'), itecha ('with you'), etc. accordingly et means 'with'. To negate the possibility of interpreting et in this verse as meaning the sign of the accusative so that the verse means 'he shall serve his brothers' the commentators point out that it means 'with.'

R' Esriel Sternbuch inquired: Why the Qamatz?

et hapasah bemoado (Num. 9:2) In Emet LeYa'akov, R' Ya'akov Kamenetzki explains that the strongest stopping tune (in addition to the Sof Pasuq) in any Pasuq has the power to make a Segol into a Qamatz. In this Pasuq, the Tipeha is the strongest stopping tune. The same change effects Efrayim (infra 13:8).