Users of natural sign languages such as South African Sign Language (SASL) resist the influence of phonetically based languages such as English that are spoken by hearing people. However, languages that come into regular contact interfere with each other, with the dominant language creating greater interference in the non-dominant language than vice versa. In the case of sign languages, this interference is proposed to manifest in the predominance of SVO sentence structure, mouthing (full or partial miming of spoken words), increased emotive rather than grammatical use of facial features (e.g. brow actions) and increased use of artificial signs, nonce words and fingerspelling. This study explores these interference features in a pioneer corpus project consisting of news broadcasts interpreted into SASL by two hearing interpreters (native SASL users). A discussion of how phonological, grammatical, syntactic and discourse features of SASL differ from their English counterparts is followed by a discussion of the corpus and relevant annotations. The results show high frequencies of SV sentence constructions and mouthings, and relatively high frequencies of fingerspelling, artificial signs and nonce words; brow actions are more resistant to interference.