Aaron2 said:
I really don't see where it does. A caster's shield has other advantages over a normal scroll. 1/2 material costs is a biggie, its worth the extra 2,000gp by it self.
Yes, it does have other advantages, but those are not pertinent to why it implies somatic components for scrolls.
"This +1 small wooden shield has a small leather strip on the back on which a spellcaster can scribe a single spell as on a scroll. A spell so scribed has only half the normal materials cost. (Experience point and component costs remain the same.) The user can cast the spell scribed on the back of the shield with no chance of arcane spell failure due to the shield."
The first sentence implies that this scribing is the same as scroll scribing (the item gives you this ability). Based on that sentence, the item has to be a spell completion item (note: nowhere in the books TMK does it state that all spell completion items have to be scrolls) since that sentence precludes other types of spell casting.
The last sentence states that although the spell is scribed, the shield's arcane spell failure chance does not apply
for this item.
This implies that arcane spell failure chance applies for other scribed or spell completion items. If they explicitly have to call out that arcane spell failure does not apply, then that's a fairly strong implication that it normally does. Why would you need the sentence at all if that is the default behavior? Spell resistance does not apply to the to hit or damage of magical weapons, but you never see is explicitly called out in a magical weapon description. There is no need.
Like I said, in and of itself, this is not overwhelmingly compelling evidence. But, when viewed with the other statements on spell completion, it does support the somatic components required position.
Aaron2 said:
Note that in the description of the item, 80% of the spells scribed on it are divine spells which don't have a spell failure chance anyway.
Again, that is not pertinent to whether spell completion items have somatic components or not.
The reason they put the divine chance so high is that arcane casters rarely (occasionally Bards) use shields. Hence, the item will probably most often be used by a divine caster, hence, they bumped up the chance of any found spell being divine (since mostly divine spell casters would have presumably used the item in the past). But, that is non-sequitur to the somatic discussion.