The complete empirical picture of MLM and the linearization of morphemes
This chapter presents the representative data set of MLM phenomena which is the base for the theoretical arguments put forth in this book. After giving some general background information about the data set and how it is genetically and areally balanced, the empirical generalizations about (un)attested MLM patterns that can be drawn from this data set are discussed in detail, with a special focus on the positions in the base that are possible and frequent targets for MLM operations. Two main generalizations hold for the MLM patterns in the data set: MLM patterns show a strong edge bias and are far more frequently attested at the right edge of their base than on the left edge. How the locality restriction for MLM follows from the theoretical assumptions about morpheme linearization and especially the assumption of the RecoverableMorphemeOrderCondition proposed in this book is shown in detail.