Anonymous QST Author suggests that Part 97 requires 6 kHz AM Bandwidth Limit

[k4kyv]

Think of the midpoint of the sliding window that represents the receiver bandwidth as the carrier frequency (or in the case of a SSB receiver, place your carrier reference point at the extreme edge of the window to represent the BFO setting).  You will observe that from the time the two windows begin to overlap on one side, as the 3 kHz window passes over the 6 kHz one, until you finally exit the overlap on the other side, the reference point will have been displaced over a distance equal to the sum of the widths of the two windows.

The extreme case would be an unmodulated carrier, with zero bandwidth.  The apparent width as observed on a receiver will be exactly the width of the  receiver's bandpass.

[Pete, WA2CWA]

Since the "Doctor Is IN" column was brought up for discussion (AM Bandwidth) you might want to stop by the "Doctor Is IN" booth if you plan on attending the Dayton Hamvention this year and meet all the "on-call" Doctors. Tell them in person how you really feel.
 
From the ARRL Letter: "You've read the column in QST for years and everyone always wants to know who is behind the costume. Here's your chance to stop by the booth, ask your question and have some one-on-one time with ARRL's Technical Experts. You might even be able to submit a stumper and get it published in QST,  This year's Docs On Call will be QST Contributing Editor Ward Silver, N0AX; QEX Editor Larry Wolfgang, WR1B; ARRL Senior Technical Editor Joel Hallas, W1ZR; QST Editor Steve Ford, WB8IMY, and ARRL RF Engineer Mike Gruber, W1MG. On Saturday, automotive experts Mark Steffka, WW8MS, and Don Hibbard, W8DBH, will be on hand to answer your questions. They will be at the Doctor Booth."

This might be the Doctor working his mojo on someone with too much bandwidth::