The Mirror Alignment Principle: Morpheme Ordering at the Morphosyntax-Phonology Interface Sam Zukoff, Leipzig University (samzukoff@gmail.com) Manuscript, August 24, 2020 Abstract As codified by Baker‘s (1985) “Mirror Principle”, the linear order of morphemes within a morpholog- ically complex word generally correlates with hierarchical syntactic structure (see also Muysken 1981). Broadly speaking, morphological derivations must directly reflect syntactic derivations. While Baker uses morphological ordering as a means of demonstrating the inseparability of syntax and morphology, he does not explore in great detail the question of the formal means by which compliance with the Mirror Principle is enacted in the grammar, tacitly assuming a cyclic morphological concatenation-based system. This paper develops a new framework for morpheme ordering that derives the Mirror Principle while avoiding some of the shortcomings of cyclic morphological concatenation. The core of the proposal is an algorithm that applies at the morphology-phonology interface, called the Mirror Alignment Principle (MAP). The MAP is an algorithm that takes the hierarchical structure of morphosyntactic terminals (in the form of asymmetric c-command relations) generated by the syntax — and potentially operated on by the morphology — and dynamically translates it into a ranking of Alignment constraints (McCarthy & Prince 1993a, Prince & Smolensky [1993] 2004) in con in the phonological component. All possible morpheme orders (and phonological modifications thereof) are generated by gen, and the optimal surface order is selected by eval. Even though morpheme order in this system is computed in the phonology, the driving force behind this order resides in the syntax/morphology. This link between grammatical components generates Mirror Principle-compliant surface morpheme orders. This paper explores two case studies of classical morpheme ordering problems, demonstrating how the MAP is consistent with each case and provides new insights into various aspects of these problems. First, it will show how the MAP is consistent with the complex interaction between Mirror Principle-satisfaction and the so-called “CARP template” in the Bantu languages (Hyman 2003). Second, it will show that the MAP framework can explain ordering alternations within Arabic’s nonconcatenative verbal system. This will demonstrate that Mirror Principle-behavior can indeed be identified even in nonconcatenative morphology. Lastly, the paper situates the MAP proposal within the broader debates surrounding the phonology-morphology interface, with special attention to the MAP’s ramifications for the theory and typology of infixation. Keywords: Mirror Principle • Alignment • Morpheme ordering • Bantu CARP template • Arabic nonconcatenative morphology • Asymmetric compositionality • Infixation 1 Introduction Even before Baker‘s (1985) influential proposal of the “Mirror Principle”, it was widely recognized that the linear order of morphemes within a morphologically complex word generally correlates with hierarchical syntactic structure (see also Muysken 1981; cf. Baker 1988). * In morphologically complex words, morphemes * This paper has been in the works for a very long time. It has benefited immensely from input by too many people to count. I will here list as many as I can remember right now, but certainly there are more that deserve my thanks. Thank you to Adam Albright, Nico Baier, Mark Baker, Ryan Bennett, Kenyon Branan, Heidi Harley, Larry Hyman, Itamar Kastner, Michael Kenstowicz, Gereon Müller, Neil Myler, David Pesetsky, Ezer Rasin, Norvin Richards, Nik Rolle, Kevin Ryan, Ryan Sandell, Donca Steriade, Jochen Trommer, Matt Tucker, Martin Walkow, Tony Yates, Michelle Yuan, Eva Zimmermann, the members of Leipzig Phonology Reading Group, the members of IMoLT, the audience at NELS 47, and audiences at MIT and Berkeley. The many mistakes and bad ideas that remain are entirely my fault. 1