Almost certainly not .... I assume you are talking about current THD values. In that case, when you "add" them together at a point of common coupling, the harmonic currents will tend to increase, perhaps in an RMS manner (although it will depend on the relative phases of each harmonic). However, your fundamental will also increase, and normally you would expect the fundamentals to be coincident in phase if power factor is sensible (and they are both sources).
If, for example you added a third source with a large fundamental and 0% THD currents, you would essentially dilute down the THD of the overall summed currents and it would reduce towards zero.
Another example is to take two identical sources with THD=2% and add them. The THD would still be 2%, but the current would be twice as big. If you could arrange it so that the fundamentals were coincident in phase (power angle etc), but the harmonics were not coincident in phase between the 2 sources, then you could even reduce the THD below 2% by adding the 2 sources.
So, its more like a weighted average of the THDs, with the weights determined by the relative fundamental current magnitudes, and if you get really lucky you might end up with an even lower THD if the harmonics are not coherent between the 2 sources.
Watch out if either ever becomes a load instead. In that case, the fundamantal magnitude gets smaller and the THD can rise. However, this might not be an issue. It depends if you are worring about absolute harmonics through a transformer (heating etc) or waveforms with excessive THD corrupting measurements or relay functionality.
Dr. Andrew Roscoehttp://personal.strath.ac.uk/andrew.j.roscoe