Huaman commented that using Miller's Law sometimes means being at a disadvantage, because "instead of giving people the impression that I'm willing to listen in order to have a real dialogue, and that I'll truly consider what they're saying, it seems to give the impression that when I'm asking questions about what led someone to a specific conclusion, I either haven't thought stuff through myself or don't have a clue what I'm talking about." And cakmpls commented about having been told that asking "why" questions "means that you doubt what the person said. A fair number of people seem to think that. Such a belief pretty much calls a halt to most meaningful conversation, IMHO."
Suppose your friend Tom says to you: "My toaster's been talking to me!" The standard response in our culture is to assume that what Tom said is false and then try to imagine/figure out what's wrong with Tom that would account for his saying something so unacceptable. As in "he's only saying that because.he's crazy" or "he's only saying that because he's trying to be annoying." The next step in that script is to respond not to what Tom said but to the conclusion we've leaped to about his reason for saying it ... as in "Oh, don't be ridiculous, toasters don't talk!" or "Look, I'm really busy here -- I don't have time for stupid jokes."
The prospects for useful communication are better if the dialogue instead goes like this:
Tom: "My toaster's been talking to me!"
You: "Hmmm. What has your toaster been saying?"
That is, Miller's Law doesn't tell you to make Tom explain why he's saying something you find unacceptable; it tells you to do that work yourself, by getting more data from Tom and then analyzing the data. You can't do that with "why" questions.
The problem with English "why" questions is that people almost always perceive them as having extra emphasis on the word "why." Even when you've tried hard to sound neutral, they perceive your question as if you'd really said "WHY are you saying that??" or "WHY are you leaving early??" The metamessage behind a neutral English why-question is only "You have a reason for X, and I'm asking you to share that reason with me, and it's appropriate for me to do that." But the metamessage behind a WHY-question adds one more item: "...and I tell you in advance that no matter what your reason is, it's not good enough."