This is an example of the competition between conjugate (Michael) and direct addition to a conjugated carbonyl, which is itself an example of thermodynamic vs kinetic control.
You can read more about this in Clayden chapter 22, but in short,
The conjugate addition is the thermodynamic product and its formation is irreversible.
The direct addition is the kinetic product and its formation is reversible.
The conjugate addition can be favoured by:
Heating-as it is kinetically disfavoured, heating reduces the kinetic barrier to conjugate addition.
Using a soft nucelophile favours the conjugate addition, a harder nucleophile favours the direct addition.
Soft nucleophiles are ones good at attacking sp3 carbons, but not always sp2 carbons, even though in this case they would attack the sp2 alkene. They involve being polarisable, large, having a delocalised charge and having a high energy HOMO (not all of these characteristics are needed but they all help make a nucleiphile soft). Examples of soft nucleophiles include Thiolates(-SH), organnocopper reagents and deprotonated aromatic heterocycles.
Hard nucleophiles are small and densely charged, and examples include hydrides and gringards or organolithium derivatives.
There is no better example than the ones in your textbook. In the first example the copper is provided to transmetallate the gringard into an organocopper to favour the conjugate addition. Without the copper, a grigard would do a direct addition.
In the second example, the hydride is a hard nucleophile and so favours the direct addition.
Note that some nucleophiles are in between hard and soft and so if the stoichiometry allows, they will do both. A nice example is a soft hydride, like sodium borohydride.
There are chemoselective ways to reduce either the ketone or the conjugate double bond without reducing the other. Luche he reduction with NaBH4 and a catalytic Titanium (III) chloride would be used to reduce only the carbonyl. K-selectride would reduce only the double bond.