Dear RNLDers, <br>I would be grateful for pointers to papers / literature on any aspects of working with mixed data: data where speakers code-mix, -switch, or simply use a high level of borrowings. <br>In my case, I have texts where minority lg speakers vary between heavy codemixing and a 'purer' style, depending on factors such as topic & formality/degree of self-monitoring.<br>
Issues arise as to which speakers to use, which parts of their data, how to record mixed data (and I'm sure lots of others).<br>With thanks, <br>Adrian Clynes<br>Universiti Brunei Darussalam <br> <br>