In order to have a ‘representative corpus’ for discourse in Irish English, the L-CIE has adapted a framework, used also by CANCODE, which includes two axes for classification: the context-type and interaction type.
This axis of categorisation reflects the relationship that holds between the participants in the dyadic and multi-party conversations in the corpus. These types of relationships fall into five broad categories, which were identified at the outset: intimate, sociocultural, professional, transactional and pedagogic.
In this category, the distance between the speakers is at a minimum and is often related to co-habitation. Only conversations between partners or close family qualify for this category in which participants are linguistically most ‘off-guard’. At times, the ‘intimate’ category is difficult to distinguish from the ‘socio-cultural’ one. All participants in a conversation have to fall under this category for the conversation to be classified as ‘intimate’.
This category implies the voluntary interaction between speakers that seek each other’s company for the sake of interaction itself. The relationship between the speakers is usually marked by friendship and is thus not as close as that between speakers in the ‘intimate’ category.
This category refers to the relationship that holds between people who are interacting as part of their regular daily work. As such, this category only applies to interactions where all speakers are part of the professional context. Talk that is not work related but occurs between colleagues in the work-place has still been classified as ‘professional’ based on the observation that participants retain their professional relationship even when the topic of the conversation is not work related.
This category embraces interactions in which the speakers do not previously know one another. The ‘reason’ for transactional conversations is usually related to a need on the part of the hearer or the speaker. As such, conversations aim to satisfy a particular transactional goal.
This final category was set up in order to include any conversation in which the relationship between the speakers was defined by the pedagogic context. Thus a range of lectures and feedback sessions have been included. The emphasis has been on the speaker relationship rather than on the setting.
Each text has been further annotated with information about the speakers such as gender, age, occupation, education status of speakers and overall ‘topic’ and area of living.
Apart from the context-type categories, distinctions have also been made within the corpus between texts that were predominantly collaborative versus those that were non-collaborative i.e. texts in which speakers give explanations and information or relate events and tell stories. A further distinction within the collaborative texts was made between ‘collaborative ideal’ and ‘collaborative task’. Since conversations are also marked by the uni-linear transfer of information from one speaker to the other interactants, this type has been termed ‘information provision’.