Earthquakes in cities revisited

During the last twenty years, a number of publications of theoretical-numerical nature have appeared which come to the apparently-reassuring conclusion that seismic motion on the ground in cities is smaller than what this motion would be in the absence of the buildings (but for the same underground and seismic load). Other than the fact that this finding tells nothing about the motion within the buildings, it must be confronted with the overwhelming empirical evidence (e.g, earthquakes in Sendai (2011), Kathmandu (2015), Tainan City (2016), etc.) that shaking within buildings of a city is often large enough to damage or even destroy these structures. I show, on several examples, that theory can be reconciled with empirical evidence, and suggest that the crucial subject of seismic response in cities is in need of more thorough research.