There can be few subjects that benefit more than binary stars do from complementary approaches. Consider their orbits, for an initial example. By ‘visual’ methods — by which I mean all methods of obtaining angular resolution on the sky, including the modern techniques of optical interferometry — one can determine an orbit, but with the scale known in terms of angular measure only. By spectroscopic (radial-velocity) methods one obtains the linear scale of the orbit, but with those techniques the inclination to the line of sight is indeterminate. In order to find the complete characteristics of the orbit in three dimensions — as is essential in order to find those most important data, the masses of the component stars — one needs to utilize both methods. It is, accordingly, clearly of great value to have a meeting like this that is object-oriented rather than technique-oriented; it is a long time since there was a meeting specifically devoted to binary stars, and especially in view of the great advances that have taken place in observational techniques of all sorts in recent years it is more than timely that we should have this meeting now. We are certainly very grateful to Dr. McAlister and his colleagues for arranging it and holding it at this beautiful venue.