Dear Helen Peterson,

You have certainly broached up a very interesting topic. In point of fact, Goffman' s (1974) concept of frame is functionally similar to the concept of scheme in cognitive psychology. Within such perspectivization, FA aims at investigating social phenomena in terms of symbolic-interpretive constructs shared by the members of a given speech community.Clearly, FA is not a replacement for DA. Although they both touch upon peoples' experiences and beliefs and the ways they conceptualize social realities, they are procedurally different. In FA, we usually begin with a particular thematic framework defining a given belief or experience, and then, try to identify a series of related themes and the subtopics which act as their tokens. Alternatively, a matrix or table is provided for each main theme representing its subtopics in rows and columns. I also refer you to Benford and Snow (2000) and D'Angelo(2002).Good luck with your research.

Best regards,

R. Biria

Dear Reza Biria,

Thank you so much for your very lucid and informative answer! This clarifies a lot about the nuances between the two different approaches. I think I will try to work with frame analysis - but as you write, I will not consider it a replacement for discourse analysis. The references are much appreciated! Thank you! Best wishes from Helen

I also recommend the Benford and Snow (2000) article from Annual Review of Sociology.

Hi Helen,

first, I suppose we are talking about critical discourse analysis (CDA), not pure discourse analysis as it is used for linguistic analyses and, given that you study policies, you want to analyse content, not language. If I'm completely wrong and you wan to to study language, I apologize in advance :).

Anyway, for me (and mainly see CDA as it was posed by Teun van Dijk) frame analysis is actually a part of CDA - you can focus on analyzing the frames present in the selected discourse because discourse is formed mainly by frames, myths, narratives, metaphors, stereotypes etc., and if you want to deconstruct (symbolical) meanings in the discourse you have analyze mentioned artefacts.  So I wouldn't put such demarcation between two methods, in fact I would not put any demarcation at all. I will give some examples to support this claim. Frames, as some cognitive artefacts created to (strategically) simplify complex problems, are often used by media to represent some social events, actors and so on. Often, these frames have normative denotation or connotation that both contribute to creation of specific discourse: squatters are framed as social deviants who pose stability risk for society and medical risk for individuals; in neoliberal social policies unemployed are framed as lazy and incompetent individuals in order to discoursively stigmatize their social position. Of course, you can omit CDA and focus purely on frames, in that case the outcome of the study would be identified frames without discourse context.

As for (dis)advantages, I cannot see how to compare CDA and frame analysis in this regard because CDA is macro-analytical method and frame analysis rather micro-analytical method (frames would be always part of a discourse). Both are qualitative methods so they are analytically demanding and, personally, I cannot think about doing frame analysis or metaphor analysis without CAQDAS that provides a tool for systematic data analysis with regard to repeating themes (discoursive formations) from which coherent frames could be identified. For CDA and frame analysis I would suggest literally anything by Teun van Dijk, for sole framing, book Framing Public Life edited by Reese, Gandy and Grant is good (the book is focused on media framing but presented ideas and procedures are portable to any data, including policy papers). 

Hopefully, this was somewhat helpful :).

Hi Jiří! Thank you so much for this informative and very clarifying answer! This is VERY helpful! And this is what I was hoping for - that it would be possible to combine the two methods like this - that they complement eachother by their focus on different analytical levels (macro/micro). However, I had not been able to figure out exactly how they complement eachother. I have not seen this spelled out this clearly anywhere else - so I'm very grateful for this! I will look at the literature you refer to. Great! Thank you!!!

I actually consider discourse analysis key to a good frame analysis. You have to code your texts according to frames, then within each frame you identify key sentences that point to the discourse of that frame. 

I am not able to find a single dominant frame especially when it comes to television news . A single story has attribution of responsibility, human interest, conflict, all at the same time. How do we then assign frames to stories?