Procedure
Analyze an excerpt of the text that you collected at the beginning of the semester or, alternatively, the excerpt of text that we provide, in terms of participant alignment, footing, and contextualization cues. 

1. What kind of interaction is going on? How do you know? How do you know whether it's friendly or hostile? How do you know whether it's a lecture or a first date? 
2. What aspects of the utterances (e.g. contextualization cues) signal shifts in participation structure (a la Goffman)? For example, when and how are the interlocutors cooperating, and when are they competing? How and why do alignments seem to shift? 
3. Are there contextual presuppositions that allow you to make sense of what is being said? How is the importance of these presuppositions conveyed? 
4. What social and interactional identities are relevant to the interaction? How are these identities being managed through the discussion? Are there grounds for claiming more than one identity per speaker? 
5. Suppose you were trying to build a system that automatically makes the kinds of judgments you made in questions 1 through 4. 
• What relevant features of the discourse might be (easily) extracted given what you know about current technologies? 
• Given your answer to the question above, would you use a rule-based, statistical, or another type of computational architecture? What are the trade-offs among the various types of architectures you can imagine? 

Provided text

[ Sample 1 ] - Four people playing a computer game (warning: strong language :-) )
[ Sample 2 ] - Statements by Presidents Bush and Yeltsin