Covering tangata whenua in Aotearoa: a big data exploration of print media and Māori

Abstract

ABSTRACTA large body of academic research documents harmful media coverage of indigenous populations across the globe. New Zealand is no exception. Aotearoa’s indigenous people, Māori, share similar damaging experiences, leading one major NZ media company to publish an apology for their historically poor depiction of tangata whenua. This paper adds to that wealth of evidence using the automated methods of machine learning to examine coverage of Māori in NZ print media. Across roughly 800,000 sentences – spanning over two decades of coverage – this research investigates print media discourses involving Māori at a mass scale while demonstrating the applicability of such tools for further research. The results replicate a collection of existing findings at large-N scale, further documenting problematic discussions of violence, political representation, and culture, among several other concerns. The novel approach also hints at complex, obscure relationships embedded within problematic language in Aotearoa’s print media, identifying notions of division - both implied and otherwise - along with notable instances of resistance.

Keywords

• Māori
• print media
• newspaper coverage
• automated topic analysis
• word embeddings
• Aotearoa New Zealand