Different models of L3 acquisition have different predictions about the cross-linguistic effects of previous
learning. The ‘Cumulative Enhancement Model’ predicts that both L1 and L2 have cumulative facilitative effects on L3, but L2 the
Status Factor (L2SF) claims that L2 has stronger effects on L3 acquisition. Similarly, while the Typological Primacy Model claims
that among the already acquired languages, a typologically similar language transfers more properties to L3, the Linguistic
Proximity Model predicts a transfer of structurally similar properties from the previously learnt languages. On the other hand,
the Scalpel Model, besides considering linguistic similarity, gives importance to factors such as frequency of use and nature of
the input, language dominance, and learning context. The current paper contributes to this debate by presenting data collected
from two groups of Balochi learners of English in Balochistan (Pakistan) who were taught British English (BrE) aspiration contrast
in plosives at the allophonic level after they had acquired Pakistani English (PakE) as adult L2 learners. Thus, Balochi was L1,
PakE L2, and BrE L3 of the learners. After teaching for more than a year, the learners were tested for accuracy in VOTs for BrE
plosives. The results support the L2SF and Scalpel Model.