vayomeru/vayomru, verochevo/verochvo (Exod. 15:1) ('and they said', 'and its rider'); yeminecha/yemincha (ibid 6) ('Your right hand') Is the Sheva which occurs after a long unaccented vowel (the long vowels are: Kamatz, Tzere, Chirik-followed by Yud, Shuruk, Cholam) a Sheva Na ('a sounded Sheva') or a Sheva Nach ('a silent Sheva')? Dikdukei Shai (R' S.Y. Mandelbaum, Jerusalem 1999) writes 'Rules of the Sheva Na … Rule 3. After a long unaccented vowel [the Sheva is Na]'. In a footnote he informs us that this is the opinion of R' D. Kimche (Radak), R' Eliyahu Bachur, and others. In a subsequent footnote he enumerates many early authorities who disagree with this opinion. These include the Dikdukei haTa'amim which is attributed to none other than Ben Asher, the major authority on the Masora. However, Mandelbaum does not mention the Ra'avad, author of the critiques on the Rambam. The Ra'avad is a Posek of standing and is worth quoting. Where Rambam says that one should take care to pronounce Nach and Nad distinctly (Hilchot Keriat Shema Ch. 2 Halacha 9), the Ra'avad writes:

I do not know what is wrong with sounding a Sheva Nach. If one says levavecha/levavcha sounding the Sheva of the second Bet in order to make it clearer so that it should not sound like a Vav, and similarly if one pronounces the Yud of Yisrael clearly so that it should not sound like an Alef and similar matters, one can sound the silent Shevas and may anyone who does so be blessed (ibid.).

1. A Sheva coming after a long unaccented vowel is Nach ('silent') even if that vowel has a Meteg.

2. It is permitted to pronounce a Sheva Nach as a Sheva Na - sometimes this is even a hiddur ('praiseworthy practice').

3. One should not pronounce a Sheva Na as a Sheva Nach.

It should be noted that Dikdukei Shai suggests that the above Ra'avad may be referring to levavchem (in the second paragraph of the Shema) and if that is so, the above points cannot be derived from the Ra'avad. Dikdukei Shai quotes present day authorities, concluding that Sheva in these positions is Na ('sounded').

For Torah readers

yiddemu ka'aven (Exod.15:16) Onkelos translated this 'they will be silent as a stone' - the Dalet has a Dagesh and a Sheva Na ('sounded'). If one reads the Dalet soft (without a Dagesh) and with a Sheva Nach one changes the meaning of the phrase to 'they will be similar to a stone' and thereby also changes the pronunciation!

mikkedash (Exod.15:17) ('sanctuary') Rashi writes 'the tune on it is a Zakef Gadol [indicating a pause] to separate it from the Divine Name which follows it. However, if the intention was that mikkedash should be in semichut (so that it would mean 'sanctuary of G-d') another change would be necessary - there would have to be a Patach under the Dalet. Why does Rashi only point out that the tune indicates separation? It may be that in his day the distinction between Patach and Kamatz in the pronunciation of Ashkenazim was not as great as it is today and was closer to the pronunciation of the Sefaradim.

Roots

marata (Exod. 15:23) ('to Mara' - a place) Rashi explains that the Heh at the end of mara changes to Tav and uses the word shoresh ('root') in this context. R' Z. Hanau disagrees, pointing out that Heh of a root does not change to Tav, e.g. ose-osa, yafe-yafa. It seems they had different definitions of shoresh.