In most of the published models of the geodynamics of the eastern Mediterranean, the westward Anatolia escape is interpreted on the motion of Anatolia and Aegean area relative to Europe. Geodetic data show that the velocity increases from east to west and southwest. This contradicts basic physics, because the velocity should rather decrease moving from the energetic source (the assumed N-S Arabia indenter), and it contrasts the fact that in geodynamics inertial forces are negligible (Turcotte and Schubert, 1982). In other words, the velocity cannot increase with respect to the initial boundary forces. Therefore, in case of an indenter, plate velocity should decrease moving from east to west, which is not what we observe.

How the Aegean Sea could open if the Anatolia plate moved and is moving westward? This should have closed the basin rather opened it. In reality, Anatolia is moving westward relative to Europe (Figure 1), but not relative to Greece, which is moving southwestward even faster than Turkey (Figure 1). Anatolia is not moving westward neither with respect to Africa (Figure 4). A supposed pole of rotation of the Anatolia plate in the southeastern Levantine Sea or Egypt is figured out in a reference frame fixed to Europe (Figure 1). However those vectors do not represent the motion of Anatolia alone, but also of Arabia and Greece. Moreover vectors depict a misleading arc since the geodetic sites are not located on undeformed plate interiors but in areas undergoing regional oblique deformation, where the local stress field is deviated with respect to the main direction of plate motion in a the absolute reference frame. Examples are the left-lateral transpression at the northwestern margin of the Arabia plate generating a NW-SE oriented compression in spite of a NE directed motion of Arabia; or the right lateral transtension in the southern Aegean Sea producing a N-S extension or the left-lateral transpression in the south (e.g., the Strabo trench), generating N-S compression, contemporaneous to the southwestward motion of Greece.

The other quite popular model for the opening of the Aegean Sea and migration of the arc is the gravitational collapse of the thickened lithosphere generated by previous subduction stages. However, also this model presents some odds: 1) The system is still active, with compression in the southern arc and extension in the north Aegean where topography is even below sea-floor. 2) From the deep Ionian (3000 m below sea-level) to the Anatolia Plateau (about 1000 m above sea level) there are about 1000 km with a topography variation of 4 km, which indicates a very gentle slope (less than 1°), too low for generating active gravitational sliding of a brittle crust. Much larger topographic gradients with similar ‘unconstrained’ margins did not generate any lithospheric-scale gravitational collapse: see the arc of the Western Alps and the oceanic Provençal basin to the south.

Analogue modeling (Martinod et al., 2000) tried to mimic the tectonic evolution of the eastern Mediterranean in terms of lateral Anatolian escape and gravitational collapse. These simulations are very helpful, but in their model it is not considered that also Africa is subducting underneath the Hellenic and Cyprus arcs, and not only the Arabia plate underneath Eurasia. Moreover, apart scaling problems, neither the increase in velocity field towards the southwest nor the actual topographic gradient are not discussed. A gravitational collapse model that explains extension in the Aegean area is hard to reconcile when the topography and geodetic data relative to Africa are analyzed and combined.