Obstacles to testing Molyneux's question empirically

There have recently been various empirical attempts to answer Molyneux's question, for example, the experiments undertaken by the Held group. These studies, though intricate, have encountered some objections, for instance, from Schwenkler, who proposes two ways of improving the experiments. One is ''to re-run [the] experiment with the stimulus objects made to move, and/ or the subjects moved or permitted to move with respect to them'' (p. 94), which would promote three dimensional or otherwise viewpoint-invariant representations. The other is ''to use geometrically simpler shapes, such as the cube and sphere in Molyneux's original proposal, or planar figures instead of three-dimensional solids'' (p. 188). Connolly argues against the first modification but agrees with the second. In this article, I argue that the second modification is also problematic (though still surmountable), and that both Schwenkler and Connolly are too optimistic about the prospect of addressing Molyneux's question empirically.