Sound change vs. orthographic remapping: Visualising ‘excrescent’ <t> and <t> deletion in fifteenth-century Scots

Starting with its earliest records and into the seventeenth century (Meurman-Solin 1997), Scots displays overlapping , , , and spellings for etymologically final /xt/, /x/ and /θ/ in words such as eight, though and cloth (e.g. , ; , ; , ). The degree of variability among the spellings makes these “appear interchangeable, leading some authorities to conclude that they are just graphical variants” (Johnston 1997a:101, cf. also Macafee & Aitken 2002: §5.2). We may enquire, then, whether sound changes such as loss or excrescence of [t] have taken place, or whether we are simply dealing with “orthographic remapping” (Lass & Laing 2013: 105).

To address this issue, our study interrogates the From Inglis To Scots (FITS) database, which maps individual 15c Scots spellings onto their most likely sound values, and their etymological sources. This grapho-phonologically parsed corpus (cf. Kopaczyk et al. 2018) has been compiled on the basis of the corpus of tagged texts underpinning A Linguistic Atlas of Older Scots (LAOS – Williamson, 2008) which brings together c. 1,250 local Scots documents dating from 1380 to 1500. Using the FITS grapho-phonological display tool (a.k.a. Medusa), we conduct an examination of the phonotactic, morphotactic and lexical distributions of the target spellings, as well as their variation over time, space and text. Medusa provides diachronic mappings of etymological sounds to (a) their 15c Scots spellings and (b) their proposed 15c Scots pronunciations. The tool also generates synchronic mappings of sounds to spellings in the form of 15c graphemic substitution sets and sound substitution sets (Laing 1999 and Laing & Lass 2003: 259-260, 262-263).

Visualisations such as these help us evaluate:

a. The relative frequencies of spellings for etymological [θ], [x] and [xt] and their interchangeability in particular contexts (cf. Figs 1 & 2)

b. The evidence for [t] insertion and deletion, particularly following [f]

c. The geographic and temporal distribution of [t] insertion and deletion

d. The limits of the interchangeability of and beyond the target sequences

e. The likelihood of the proposed change from [xt] to [θ] as mutual assimilation of the source cluster (see Johnston 1997:102)

f. Whether for [θ] may be a backspelling based on a [xt] > [θ] change

While the sound values recorded initially in the FITS database are conjectured on the basis of existing literature, our method incorporates iterative, data-driven reappraisals.