You can check the continuity of a lamp (bulb) with a multimeter on its Ohms (Ω or 'R') range; always short the meter probes together before continuity testing just to get an idea of how the meter is calibrated at full continuity (short circuit):
- Disconnect the lampholder, if possible, from its supply wire, leaving the lamp in place
Hold or clip one of the probes to the bike's ground; I prefer to use the batt -ve or a cylinder fin
Place the other probe in contact with each wire into the lampholder(s) and note the reading*
With the lampholder still connected to earth**, resistance readings between those noted below and, say, up to 5KΩ (5000Ω) would be a reasonable indication of a poor connection within the lampholder or a grounding point/path. Where the lampholder is still grounded on the bike and the reading is high, as mentioned, leave the meter connected via clips or clip-leads and physically 'work' lamps, connections & brackets to see if the reading changes drastically, thus indicating the approximate location of a poor (resistive) connection. Connecting the +ve meter probe gradually further upstream towards its 12V source -- i.e. moving from the taillight connection behind the licence plate, then the fender connection, then terminal 'TL' at the ignition switch (follow the wiring diagram logically for each circuit) will gradually narrow down the source of a resistive connection.
*The lamps on our bikes should vary between 2.4R (Ω) for a 12V/60W(!) headlight, 4.1R for a 35W headlight, 6.9R for a 21W brake light, 28.8R for a 5W taillight and 72R for a 2W instrument lamp. These figures are accurate for a 12V lamp, and a bit anal, but for most lamps on our bikes a reading of <100R, bare lamp taken from its holder, is close enough to indicate a serviceable lamp.
** Not so the neutral light; this gets its ground from the neutral switch (except when the trans is in gear) so disconnect both wires and test the lamp & holder independently of the bike.