Halliday & Matthiessen (2014: 340, 340n):
Looked at from this point of view, the variable is not one of extension but one of causation. Some participant is engaged in a process; is the process brought about by that participant, or by some other entity? In this perspective, the lion chased the tourist relates not so much to the lion ran as to the tourist ran: ‘the tourist did some running; either the running was instigated by the tourist himself (intransitive the tourist ran), or else by some external agency (transitive the lion chased the tourist)’. Note however that the terms ‘transitive’ and ‘intransitive’ are no longer appropriate here, since they imply the extension model. The pattern yielded by this second interpretation is known as the ergative pattern. The clauses the lion chased the tourist/the tourist ran form an ergative/non-ergative pair.* …
* In the typological literature on ‘case marking’ or ‘alignment’ systems, such pairs are often referred to as ‘ergative/absolutive’, contrasting with the ‘nominative/accusative’ pair of the transitive model.