Hi Lois. Your question is a bit vague. As Neil Bowen asks, it will depend on what you are trying to discover in that communication. Given that we don't know what your specific questions are, a general response is all that can be offered. Methodologically I guess you would look to either acquire either interviews or somehow record their interactions. This could give you a few ways to go. Interviews would give you the opportunity to work with people as informants and insiders, and ask more technical questions about their views on relationships, and work from there to discover the themes and statements they tend to make. If you have natural conversations, especially if they are video recorded (but this is unlikely), you could engage in a variety of approaches such as thematic, stance, narrative, and sequential analyses. In either case, you could then begin to look for the features you had started with (e.g. power, identity, affiliation etc.)

There are a bevvy of conceptual frameworks you could employ, that would match (or not match) your research questions.

I agree entirely with the previous respondents. However, even if you don't have any research questions - a precarious position to be in! - you could use your data to look at basic issues like who asks questions and who answers them; who tends to interrupt whom; and when either person introduces a new topic into the conversation, is it taken up and talked about or not? Looking at things like that might give you a basis for drawing conclusions about power relations within the relationship.

I agree with the above respondents. But I also have questions (hope, it helps): what do you mean "relational (!) partners"? Is this a regular natural discourse or an online discourse? There are a lot of approaches to qualitative discourse analysis. What are your research questions and what are your hypotheses?

To put it very briefly, DA helps a researcher to understand communication between partners by allowing her/him to employ linguistic determiners and extralinguistic determiners in data analysis.

If you can, please read The Discourse Reader, edited by Jaworsky and Coupland, for details.

As earlier stated by other respondents, it is important to spell out your focus. Are you interested in examining discourse structure? Dialogic positioning and authorial voices? Power relations? Your research aim will help to supply answer to your question.

All the previous respondents have made excellent points. The only addition I would make is to consider the context of the communication. Norman Fairclough might be worth reading, particularly for examining power dynamics.