Variable affix order: Grammar and learning

Abstract

While affix ordering often reflects general syntactic or semantic principles, it can also be arbitrary or variable. This article develops a theory of morpheme ordering based on local morphotactic restrictions encoded as weighted bigram constraints. I examine the formal properties of morphotactic systems, including arbitrariness, nontransitivity, context-sensitivity, analogy, and variation. Several variable systems are surveyed before turning to a detailed corpus study of a variable affix in Tagalog. Bigram morphotactics is shown to cover Tagalog and the typology, while other formalisms, such as alignment, precedence, and position classes, undergenerate. Moreover, learning simulations reveal that affix ordering under bigram morphotactics is subject to analogical pressures, providing a learning-theoretic motivation for the specific patterns of variation observed in Tagalog. I raise a different set of objections to rule-based approaches invoking affix movement. Finally, I demonstrate that bigram morphotactics is restrictive, being unable to generate unattested scenarios such as nonlocal contingency in ordering.*

Limits on the extension of affixal combination: structural restrictions and processing conditions

“…Traditionally, combination of affixes has been explained by stratum-oriented models (Siegel 1974;Allen 1978;Kiparsky 1982;Giegerich 1999), by selectional restrictions of each particular affix (Fabb 1988;Plag 1999;Talamo 2010;Rodrigues 2015), by the interaction of selectional restrictions and processing constraints, by what is called the Complexity-Based Ordering hypothesis (formulated by Hay [2002] and applied, with different results, by Hay [2003] Saarinen and Hay [2014]), and by the interaction between scope, phonological subcategorization and morphotactic constraints (Caballero 2010), under the view of Optimality Theory (cf. also Ryan (2010). The evaluation of the balance between universal and language-specific factors that determine affix ordering has been the matter of debate for studies such as Sims and Parker (2015), Caballero (2010), Ryan (2010), among others.…”

“…also Ryan (2010). The evaluation of the balance between universal and language-specific factors that determine affix ordering has been the matter of debate for studies such as Sims and Parker (2015), Caballero (2010), Ryan (2010), among others.…”

The Role of Morphology in Optimality Theory

“…Morphotactic constraints determine affix positioning. These include alignment constraints (e.g., Prince 1993a, Trommer 2001) and templatic constraints (e.g., Hyman 2003, Paster 2005a, Caballero 2010, Ryan 2010.…”

Machine Learning of Inflection

“…Bigram models, a special case of n-gram models where n = 2, are frequently applied to the task of learning sequencing patterns. A recent proposal by Ryan (2010) incorporates bigram constraints (sequences of two immediately adjacent morphs) within a statistical Maximum Entropy framework (Hayes and Wilson, 2008) to learn morphotactics. The goal of Ryan's learner is to model patterns of free variation in morpheme order which are not due to phonological or semantic factors (the proposal is illustrated using data from Tagalog).…”