Published in: Esa Penttila & Heli Paulasto (eds.), 2009. Language Contacts Meet English Dialects: Studies in Honour of Markku Filppula. Newcastle: Cambridge Scholars Publishing, 179–198. On the Geography and Date of the Merger /kw/, /hw/ > /χw/ Stephen Laker An interesting observation of Yorkshire speech was taken by a Lincolnshire informant in 1887 and is reported in Peacock’s glossary of the north Lincolnshire wapentakes of Manley and Corringham (1889, s.v. quee): Ey, thaay speäk clear different e’ Yerksheere to what we do. I mind hearin’ a woman ’at was fra that-awaays-on tellin’ on her naaibour she’d gotton a nist “why,” an’ when I went to see what it was, it was noht bud a quee-cauf. The reference is to an old northern English word for the heifer – the quey – which is pronounced differently in the Yorkshire and Lincolnshire dialects. Evidently, its Yorkshire pronunciation confused the Lincolnshire dialect speaker. The pronunciation /(h)waɪ/ bears witness to a phonological change which took place throughout the north of England and, to a limited extent, southern Lowland Scotland, 1 namely a process of frication and subsequent lenition of /kw/ > /χw/ > /hw/ > /w/. In the 1950s, when fieldwork for the Survey of English Dialects was conducted, the dialect word quey was no longer elicited from Lincolnshire informants, but it was still found in Yorkshire and other northern dialects. Many other words also evidence the change, such as the word quick “heaving; quick of the nail; active” (see maps 1–4). Nowadays, the change is best preserved in numerous place-names (see map 5). First definitive evidence for frication of /kw/ comes from northern Middle English alliterative verse, a property of which was to alliterate words with etymological /kw/ and etymological /hw/, e.g. Wherfore, wheme kyng, for what þat may come 2 (from the Old English forms: hwǣr + for “wherefore”, cwēme “queem, pleasant” and hwat “what”). Five Middle English compositions attest such alliterations, and their sources of origin have been plotted on map 6. 3 (Another North Midland 1 See LSS (I, 101, 134, 268–270, 322–323; II, 89, 192–3) and Watson (1923, 16). A similar change of /kw/ to /hw/ also occurs in Orkney and Shetland dialects (see Laker 2002, 195 note 26). However, while in dialects of especially northern England and southern Lowland Scotland /kw/ often became /(h)w/, a change in the opposite direction, i.e. /hw/ > /kw/, is not attested as in Orkney and Shetland English. It should also be borne in mind that in the Northern Isles there was a shift from Pictish to Norn and then to English, which would lead to different linguistic outcomes. Yet it is interesting to observe that the same substrate conditions which would have effectuated an early change of the initial clusters /kw/ and /hw/ to /χw/, as suggested in Laker 2002 with reference to British Celtic, probably also existed in Pictish (see Sims- Williams 1992, 48). 2 “Why, queem king, for what may come?” (Destruction of Troy line 2648). 3 For further details on how the locations of the authorial dialects have been deduced, see Turville-Petre (1988, 264–269); Gates (1969, 30), Hanna (1974, 50); Ginsberg (1992, 1–2); Hanna and Lawton (2003, xxvii–xxxv); Duggan and Turville- Petre (1989, xlii). Sir Gawain and the Green Knight (London, BL, Cotton Nero A.x) is noted for the spellings whene (lines 74 and 2492) “queen” and whyssynes (line 877) “cushions” (Tolkien and Gordon 1967). However, in the poem /kw/ is always found to alliterate with itself or with /k/, but never with /hw/. Furthermore, the poet alliterates /hw/ with itself and with /w/. In fact even when etymological /hw/ is rendered in the manuscript as <qu> it can alliterate with /w/ (255, 257, 1186, 1227). These facts indicate that in the scribe’s dialect /kw/ and /hw/ had probably merged, but not in the Gawain-poet’s dialect (Oakden 1930, 79). Indeed, it so happens that for Gawain there is enough circumstantial linguistic