We will discuss four misunderstandings often connected to use of digital traces:
1) the use of a notion of digital traces that is both too narrow and too ambitious;
2) the alternation of oblivion and paranoia on the conditions of digital traces' production;
3) the tendency to confuse digital and automatic;
4) the hope that the digital traces are easily clamped by conventional methods.
We will try to show than when these misunderstandings are avoided, digital methods can renew the vision of social sciences and help them to overcome the classic divide between qualitative and quantitative methods.