The value judgments we make (right versus wrong, beautiful versus ugly, 
interesting versus tiresome, valuable versus worthless, etc.), are hard 
to pin down and there is an old question as to whether they are human 
abstractions or have a real existence outside the human sphere.  For 
example, 2 + 2 = 5 is "wrong," so also is beating a child.  Does "wrong"
 mean the same thing is both of these cases?  Most people think not but 
in fact we don't really know -- there are schools of thought (Asian 
karmic notions being the one I'm familiar with) that would say both are 
errors in the "wrongness" sense -- they both go against something that 
consists of "right."
We can't, however, depend entirely on our feelings in these things -- 
people can be wrong about mathematical calculaitons as much as wrong 
about their behavior and what is good and bad behavior.  We instead have
 to check our calculations -- in the end see if they stand up under 
deductive scrutiny from basic principles -- axioms or postulates or 
whatever.
With behavior I think the fundamental postulate is the "golden rule" or 
concept of compassion for all sentient beings.  How does violence and 
inflicting hurt or pain on someone stand in the light of this 
postulate?  Obvious.  That doesn't mean one is a total absolutist in 
such things -- scenarios can be imagined where the failure to inflict 
pain does more harm than not -- but such scenarios are usually 
far-fetched and uncommon.
Our evolution was a different matter.  Here we managed largely by 
killing and being strongest and so on.  But evolution was simple biology
 -- in the end just a blind natural process without moral standing -- 
and is therefore misleading to use as a guide for behavior.  Out 
instincts bring us to strike out, to "defend" ourselves, to get revenge,
 to judge -- and this leads us into morally wrong behavior.
To my mind, then, any claim that it is better to inflict pain and hurt 
than not to has the burden of proof -- that the correct, moral response 
is always to avoid conflict and only with a huge preponderance of the 
evidence does one behave otherwise.