The efforts of phonologists and phoneticians to define the notion possible word of such-and-such a language, carried out in careful detail across many languages, has led to a general picture of the phonological parts bin and tool box out of which the world's languages form their phonological systems. Although these systems exhibit a fascinating and perhaps even endless variety of superficial differences, just as as biological forms of other other sorts do, careful analysis always shows us deeper similarities, reflecting the fact that these systems are all built out of the same sort of stuff by the same sorts of processes. An excellent introduction to modern phonology, presented in breadth and also in considerable depth, can be found in Kenstowicz (1993).
Theories of phonology have improved a great deal over the past thirty years, with the result that a trained phonologist begins to work on a new language with a set of analytic tools that produce insightful analyses of many of its phonological phenomena. This process of improvement continues, as new ways of thinking shed light on old problems, and new analyses strengthen or undermine existing theories. For examples of how a fresh perspective can improve on familiar analyses while offering a convincing account of new phenomena, see Prince and Smolensky (1995).