It is shown which properties of a strong gravitational lens can in principle be recovered from observations of multiple extended images when no assumptions are made about the deflector or sources. The mapping between individual multiple images is identified as the carrier of information about the gravitational lens and it is shown how this information can be extracted from a hypothetical observation. The derivatives of the image map contain information about convergence ratios and reduced shears over the regions of the multiple images. For two observed images, it is not possible to reconstruct the convergence ratio and shear at the same time. For three observed images, it is possible to recover the convergence ratios and reduced shears identically. For four or more observed images, the system of constraints is overdetermined, but the same quantities can theoretically be recovered.