Factor Labels?

Can someone give a better description of how Factor Labels work than the pipeline pilot manual? Specifically, if you are adding factor labels... what does each array index represent? What is the significance to the number of factor labels (outside of counting the number of indices)? What does one use (a character, word, previously defined variable...) for those indices...For example, in the example protocol "Read Excel Experiment" , what are A-C and X&Y supposed to represent (see below)?

     source_1    A    X

     source_2    B    X

     source_3    C    X

     source_4    A    Y

     source_5    B    Y

     source_6    C    Y

Thanks!