The primary aim of this study has been to find out whether the choice and frequency of pragmatic markers can be said to distinguish phatic talk (‘chats’) from informative talk. A secondary aim has been to consider the bonding effect of the pragmatic markers. Five conversational extracts from COLT (The Bergen Corpus London Teenage Language), four representing boys’ and girls’ phatic talk, and one representing informative teacher talk have been investigated. The study shows that the distinction between the two types of talk is not a matter of frequency but a matter of marker choice. The bonding effect of the markers dominates in the girls’ talk in the form of appeals for agreement and encouragement signals. In both types of talk, the pragmatic markers are successfully used to avoid conversational gaps.