Perhaps only 5 percent of star formation occurred before z = 5 (the proportion could be higher if most of their light were reprocessed by dust). But these early stars were important: they generated the first UV and the first heavy elements; they provided the backdrop for the later formation of big galaxies and larger-scale structure. Large-scale structure may be elucidated within the next decade, by ambitious surveys (2-degree field and Sloan) and studies of CMB anisotropies; as will be the evolution of galaxies and their morphology. The later talks in this conference will highlight the exciting progress and prospects in this subject. But despite this progress, we shall, for a long time, confront uncertainty about the efficiency and modes of star formation in early structures on subgalactic scale.
I am grateful to my collaborators, especially Tom Abel, Zoltan Haiman, Martin Haehnelt, Avi Loeb, Jordi Miralda-Escudé, Piero Madau, Avery Meiksin, and Paulo Tozzi, for discussion of the topics described here. I am also grateful to the Royal Society for support.