Covalent Character- A Level

Does covalent character make a bond stronger or weaker ? I am kinda confused because I read in a past paper that the theoretical lattice dissociation for AgCl is less than it is really, because theoretically, you assume all bonds are perfectly ionic when in fact there's some covalent character, so the bond is stronger and lattice dissociation is higher. However, I came across a thing that says Al2O3 has a lower melting point than expected due to its covalent character. I'm confused because surely covalent character makes it stronger, so it would have a higher melting point?? Also, I think the fact that one atom is polarising and distorts the electron cloud in the other atom is relevant in both cases, but why does it seem to make the bond stronger in AgCl but apparently weaker in Al2O3???

It's possible to do either.

If the covalent character leads to a greater degree of network bonding (giant covalent type) then this would increase the strength of the lattice and the bp. This seems to happen in the silver halides.

However, if the covalent character leads to simple covalency then this would decrease the strength of the lattice and lower the bp. This appears to be what is happening with lithium chloride.

There is no evidence for this in aluminium oxide, but one or two examination boards have ascribed the lower mp of aluminium oxide as compared to magnesium oxide as being due to a degree of covalency. If this is what your board says, go with it.