Figure 4 Nevertheless, it has often been argued that using the drag mode to test myriad cases may reduce or even eliminate any need for deductive proof (Allen, 1996; Chazan, 1993a; Hadas et al., 2000; Hoyles and Jones, 1998; Laborde, 2000). When students can generate their own empirical evidence Allen (1996) is one of the strongest critics of DGS. For him DGS tends to blur the distinction between illustration and proof. A dissection proof of the Pythagorean Theorem exemplifies this. In this DGS-based construction, students can cover the largest square by using the parts of the smaller squares, which are partitioned (Figures 4, 5 and 6). Allen argues that such an illustration cannot be considered as proof, since no previous theorems are invoked. Figure 4 The dissected Pythagoras illustration first screen