British *mb

The British consonant cluster mb regularly became WCB. m(m), as can be seen in W. crwm, CB. kromm 'curved' < Br. *crumbo-.  But in several instances, it occurs in the Cumbric evidence as mb.  For example:

  • Ancrum (near Jedburgh), which contains Br. *crumbo- is recorded as Alnecrumba and Alncromb in the 12th century.
  • Cam Beck (north Cumbria) retains the sound to this day, thanks to a later re-analysis with the dialectal beck 'stream'.  The name derives from Br. *cambāco- 'crooked, bent' (cf. W. cam, CB. kamm).
  • Crummock Water, Crummock Beck (Cumbria) and Crummack Dale (West Yorkshire) are derived from Br. *crumbāco- < *crumbo-.  These are recorded as Crombocwater (14th century), Cromboc (13th century) and Crombok (12th century) respectively.
  • Cumberland, Cumbria.  Both are from the OE Cumbras 'Cumbrians' which is a borrowing from W. Cymry or its equivalent.  The former is recorded as Cumbraland in the 10th century; the latter is an early Latinisation based on the English word.  The same element occurs in Cumbrae (Cumberays 13th century) and Cummersdale (Cumbredal and Cumberdale, 12th century).
  • Crimple Beck (West Yorkshire) is supposedly from Br. *crumbo-pull- 'curved pool'.
  • Cumdivock (Cumbria) is recorded as Combeðeyfoch in the 11th century and contains the Br. cumbo-.

Crummack Dale and Crimple Beck are of limited use to this discussion, since they lie outside the Cumbric zone and could well have been borrowed before mb was assimilated in British.  Examples with OE Cumbras are also dubious since we don't know when or where this word was borrowed - compare OE cumb 'valley', an early borrowing from Br. cumbo- which also retains mb.  The same is true for OI Combrec, which was used to refer to any Brythonic language. 

Jackson (1953) concludes that the assimilation of *mb > mm was complete in WCB by the 7th century but that it lingered in the north perhaps half a century longer.  The evidence certainly suggests that mb was retained longer in Cumbric than further south, and the fact that the late names Cumwhitton and Cumwhinton contain no trace of b in early records suggests the sound did eventually become assimilated.  If Jackson's summation and dating is correct, it does point to a deviation for Cumbric, but hardly a defining one.