Collaborative Requirements Engineering: Bridging the Gulfs Between Worlds - Intentional Perspectives on Information Systems Engineering - page 350

Common ground or mutual understanding is generated from three knowledge
sources: (a) knowledge held by the participants about each other and the topic of the
conversation; (b) knowledge about the environment in which the conversation takes
place, including artefacts in the environment; and (c) shared knowledge about social
conventions held by the participants. Conversations have tracks or topics which may
be interleaved, for example major and minor topics, and meta-discourse or conver-
sations which control the primary topic-related discourse. Meaning in conversations
has different layers; the surface layer is explicit expression, but there may also
be layers of tacit meaning which rely on deeper understanding of metaphors and
interpretation of puns, irony, jokes and fiction.
Clark's theory explains the establishment of mutual understanding via a process
of verbal and non-verbal communication integrated with action in the world, which
progresses through stages of increasing understanding (common ground) that move
towards a shared goal (the joint project). Shared understanding is helped by the
participants' knowledge of each other, past conversations, and the context, referred
to as the arena and setting. Brennan and Clark [ 3] give a set of criteria for effective
conversation which has been applied to assessing the effectiveness of computer-
mediated conversation (CMC) [ 15] :

• Co-presence: all participants share the same space and time.

• Visibility: conversers can see each other; inherent in co-presence.

• Audibility: audio communication is supported, e.g. phone conversations.

• Contemporality: messages can be generated and received in the same time

interval, i.e. synchronous communication.

• Simultaneity: communication is possible in both directions at the same time, i.e.

complete synchronous communication.

• Sequentiality: the order of message generation and receipt is preserved (asyn-

chronous, ordered).

• Reviewability: messages can be re-read, i.e. possible for text but not for audio

unless it is transcribed.

•

Revisability: messages can be edited, afforded by text but not speech.

These criteria are supplemented with concepts from Media Richness Theory [ 7]
that describe the communication channel/modality and the connectivity in com-
munication (i.e. person-to-person vs. narrowcast vs. broadcast). Representation
of participants' identity draws on concepts from Social Presence Theory [29] :
explicit visible identities,

tokens for identities, assumed covert identities, no

identity, etc.

3 RE Representations and Communication Modalities

In this section the affordances of collaborative technologies for supporting RE
are assessed using the common ground modalities framework to investigate the
types of information and knowledge that can be represented by models, scenarios,