I've studied the valid/invalid arguments and though it's almost down, I sometimes still have mix-ups. I'm wondering if it's a good steadfast rule that, most and some only allow you to draw valid inferences if they occur first, then all the relationships ...