Exploration of the unknown

Although the discovery record of radio astronomy is outstanding, the rate of new discoveries has slowed markedly in the last decade; the time is ripe for a transformational instrument. Harwit’s arguments, based on a combination of phase space and astronomical history, make it clear that the SKA must not only take a big step in sensitivity but also be technically (and, I contend, operationally) novel. The consensus that the SKA’s field-of-view, and hence survey speed, must be much larger than offered by current arrays, is a first major step in the right direction. There are, however, other generic ways in which the SKA can be designed to maximise its discovery potential. But technology alone is not enough. A holistic philosophy of maximising the the “human bandwidth” is therefore proposed. The core ideas are: parallel rather than serial access to data—combining simultaneous access to many independent groups of users (via various forms of multi-beaming) coupled with a new attitude to operation which breaks, at least in part, the present-day paradigm of “science by proposal”. Since the potential for surprise in data only increases logarithmically with the sample size, discoveries tend to be made by people getting a lot of telescope time and some SKA users of the future must also be given room to take observational risks. As the exploitation of the HST, SDSS and pulsar archives has shown, re-analysis of existing data is another highly parallelised route to exploration, free from the constraint of a proposal committee. What new observations to make, or where to seek patterns in archives, may be aided by the application of the “morphological” approach (MA), pioneered by Zwicky 50 years ago but largely forgotten by modern-day astronomers. MA is a way of forcing us to make unorthodox comparisons and hence to examine regions of multi-dimensional parameter space with the potential for providing new observational clues.