The computational nature of stress assignment

Abstract

While computational studies of stress patterns as phonotactics have yielded restrictive characterizations of stress (Rogers et al., 2013) with provably correct learning procedures (Heinz, 2009), an outstanding question is the nature of stress assignment as a function which assigns stress to an underlying bare string of syllables. This paper fills this gap by locating stress patterns with respect to the subsequential class of functions (Mohri, 1997), which are argued to be important for phonology in that the vast majority of phonological functions fall within the subsequential boundary (Heinz & Lai, 2013; Chandlee, 2014), with the notable exception of tone and vowel harmony (Jardine, 2016; McCollum et al., under review). The main result is that—while most, if not all quantity insensitive (QI) stress systems are subsequential functions—the same does not hold for quantity sensitive (QS) systems. Counter-intuitively, so-called default-to-opposite QS patterns are subsequential, but default-to-same QS patterns are provably not. It also supports the claim of Jardine (2016) that certain tonal patterns are non-sequential because their suprasegmental nature allows for more a more powerful computation. As stress assignment is also suprasegmental, the existence of non-sequential stress functions adds evidence for this conclusion.

Foot structure enables strict locality in phonological processes

“…Thus, subsequential functions are more expressive than ISL or OSL functions, as the requirement to adhere to input (or output) locality no longer applies. Subsequential functions represent a important divide in the space of possible functions that includes many phonological patterns and excludes many non-phonological ones (Heinz & Lai, 2013;Chandlee, 2014;Jardine, 2016;Payne, 2017;Luo, 2017;McCollum et al, 2020;Hao & Anderson, 2019;Koser & Jardine, 2020).…”

The atomic properties of stress

“…Though there are notable exceptions (Jardine 2016; Hao & Andersson 2019; McCollum et al . 2020; Koser & Jardine 2020b), most phonological processes are subsequential.…”

“…Previous work on stress from the computational perspective offers an additional perspective to the issue of what constitutes a possible stress pattern (Heinz 2007a; Rogers et al . 2013; Heinz 2014; Baek 2018; Hao & Andersson 2019; Rogers & Lambert 2019; Koser & Jardine 2020b). However, most of this previous work differs in that it studies stress patterns as phonotactics, rather than a mapping from an input to an output.…”

“…However, a sufficiently restrictive theory of iterative stress will enforce even more stringent requirements than being formally regular, as there are clearly pathological patterns that fall within the regular boundary. One example is the following ‘sour grapes’-like pattern (SG; Wilson 2003, 2006) for stress (Koser & Jardine 2020a,b). In such patterns, if a condition such as spread of a feature cannot be satisfied to the fullest possible extent, the condition is abandoned instead of applying partially:…”

Input and output locality and representation

“…A logical perspective like the one put forth in this paper presents a possible solution to this, as restricted logics can be used as a local constraint definition language (Jardine & Heinz 2016). However, this is not a trivial issue, as optimization can generate non-local patterns purely from computationally local constraints (Gerdemann & Hulden 2012;Koser & Jardine 2020).…”