vasily, just measuring DC voltages around a ~500kHz oscillator circuit is not a good idea for "debugging" such a circuit since most multimeters will not show correct values if the DC is superposed by RF oscillations. Using an oscilloscope, together with a 1:10 probe (which reduces the capacitive load of the probe from ~40pF to only ~5pF) will be more helpful.
If you have no oscilloscope, you can always put an AM radio, tuned to the lowest medium wave frequency (530 to 550kHz), besides your circuit and play around with L11 to increase the volume oscillator's frequency temporarily until you get a signal in the radio. This would prove that the oscillator works, while measuring the voltage at the Emitters of Q6/Q7 is of no use. It may vary from -0.2V to +0.2V when the oscillator works (while -0.6x V will clearly indicate that the oscillator does not work), due to component tolerances.
Then you might connect your multimeter to the anode of D1 (thats the side without the ring). There you will most probably find around 0V. Tune L11 until this voltage goes down to around -4V (with all things as far away from the volume loop as possible). Then you should be able to vary this voltage by approaching your hand between 0 and ca. -4V. If you can't get this result, try shorting L10. This signal is then amplified by one half of the LM 13600 or 13700 and you will find -12V at pin 12 with 0V at D1 and +12V with -4V at D1.
At this point, everything should work fine.