Wave stacking. Simply put, it's clipping: you have two sounds with amplitude 0.6, they amount to amplitude 1.2 (well in reality it's a bit more complex than that really but, like I said, simply put), and since amplitude 1.2 goes over the clipping limit, the sound clips which sometimes is wanted but often times it's not.
To counter this, you can reduce the master volume until the maximum volume is below the clipping limit. Then the signal doesn't clip anymore because the headroom in the signals allows non-destructive amplification.
Or, you can use a compressor or a limiter to level the signals so they don't clip. What compressors do is basically lower the volume when the volume is above a certain treshold, reducing the dynamic range, and thus allowing higher overall volume without clipping. Limiters do the same, but they "look ahead" and change the volume in advance, so that the volume never goes above the treshold you set.
This way, you can make it so that when multiple notes play at the same time, they're played quieter, which makes the overall volume more level and keeps it from clipping.