For all we know, there are three kinds of learning methods that we humans use: inductive reasoning, deductive reasoning, and repetition. We use all these three kinds of observatory techniques to conclude a particular conclusion from sets of parameters.
Dogs on the other hand, mainly use repetition as its method of learning. When a particular action or word is repeated for a certain number of times, he can then take an inductive or deductive reasoning from it. Unlike us humans, we mostly get a conclusion from inductive or deductive reasoning, before we can confirm it with repetition.
Say for example, when you repeatedly use the word ‘bad’, when your dog does something bad and when you are disciplining him. He will then take that as a keyword hinting that he will be in trouble. He does not understand the word, but because of such repetition, he can conclude that that word is signaling trouble.
Training Dog Tips 101 