Overall betterness and other comparative forms of values are often taken as being representable by an ordering relation, although there are differing opinions about the exact properties that such a representation should have. In these types of order-based approaches, it is possible to define metrics on value representations and compute distances between the value representations of agents by so-called consensus measures, which describe the nearness or remoteness of the values of one agent to those of another agent or a group of agents. These measures minimally satisfy the three essential properties of coincidence, symmetry, and triangle inequality, and often more conditions are postulated. In a recent article on value disagreement, I have briefly mentioned a peculiar kind of measures that are not symmetric, hence also not distance measures, but which seem to have some ‘intuitive appeal’ and give rise to what I called perspectival disagreement. In this talk, I lay out this kind of disagreement and address the question whether it exists in real life and, if so, how important it is.