There is an ancient distinction between the synthetic and analytic operations of the intellect. The synthetic operation builds parts into wholes, the analytic operation breaks wholes into parts. The distinction seems to have lost its usefulness among sophisticated people, as thought becomes a mess of mush. But reductionisms flourish from this amnesia, as minds forget that one mind cannot do both operations at the same time on the same object.
So synthetic assertions always melt away under analytic scrutiny. This is normal; it says nothing about the synthetic assertion itself. You can’t see wholes with a parts-instrument; likewise, you can’t see parts with a wholes-instrument. That wholes are more than the sum of parts is not a confirmable proposition, because you can’t validate decibel measurements with a spectroscope.