Trubetzkoy (1939) introduced the concept of boundary signals, and he identified a large range of them. However, he only considered very few boundaries, i.e., only morpheme and word boundaries. Chomsky & Halle (1968) introduce a feature [±boundary] and also primarily distinguish between morpheme (formative) and word boundaries. Going beyond Trubetzkoy, they assume that syntactic units are translated into phonological units but already observe that the mapping is not always perfect. In particular, the highly hierarchical syntactic structure is flattened in prosody (1968: 372).
With the introduction of Prosodic Phonology (Selkirk 1980; Nespor & Vogel 1986), a rich hierarchical structure of different prosodic units was available and the domains of the Prosodic Hierarchy were all motivated through observed application of a sandhi rule or non-application of an otherwise general rule, i.e., boundary effects. Inside words, the Prosodic hierarchy is orthogonal to the levels developed in Lexical Phonology and Morphology (Kiparsky 1982; Mohannan 1986 etc.).
Prosodic Phonology and the domains proposed have been challenged on empirical as well as theoretical grounds repeatedly. Specifically the domains beyond the word domain have been under discussion, starting with the Clitic Group. Prosodic Words nested within Prosodic Words have occasionally been proposed (Peperkamp 1997), as well as major and minor phrases or the accent phrase. Ito & Mester (2009:147) criticize the proliferation of categories in the hierarchy and introduce strictly limited recursion or rather layering of the domains above the foot (for discussion see as well Vogel 2009). This layering was subsequently extended to feet (Martínez-Paricio 2013). The probably most radical variation on this theme is the infinite recursion of the Prosodic Word (Bennett 2018). Proliferation of categories (with or without recursion) is one approach to solving problems with the hierarchy.
A different strand of research reconsiders syntactic domains, such as phases as the basis for phonological domains (Marvin 2003; Kratzer & Selkirk 2007; Newell 2008; Pak 2008; Šurkalović 2015; Jenks & Rose 2015; D’Alessandro & Scheer 2015; Sande 2017; Kastner 2019; Sande, Jenks, & Inkelas 2020). Prosodic restructuring, as alluded to by Chomsky & Halle in terms of readjustment rules and as discussed in breadth by Nespor & Vogel, still needs to be accounted for in such syntactic approaches (see, e.g., Bonet et al. 2019).
Accordingly, the most recent proposals seem to acknowledge this and develop hybrid approaches, which use syntactic structures, e.g., phases, for coarse chunking and constraints on prosodic categories for the fine tuning (Kastner 2019; Sande, Jenks, & Inkelas 2020; Bellik et al. 2023). The consequences and benefits of an integrated approach that derives phonological domains from syntactic domains such as phases, have yet to be explored.
Guiding questions:
What is a phonological domain and how can it be motivated?
How many or which phonological domains are there?
How do phonological domains interact with morphology, syntax, and pragmatics?
Which theoretical problems emerge in their formal analysis?