Deconstructing Paranoia: An Analysis of the Discourses Associated with the Concept of Paranoid Delusion
David J. Harper PhD Thesis June 1999
Back to title page
Overview of criteria for evaluating qualitative research
- That there is enough material to justify the interpretation placed upon it (Hayes in Parker, 1997d).
- That the account has some face validity (Turpin et al., 1997)
- That readers (who are themselves skilled interactants and 'culturally competent' will find interpretations adequate (Potter, 1996a)
- That there is sufficient 'fit' between the data and interpretation (Elliott et al, 1996; Henwood & Pidgeon, 1992; Widdicombe, 1995); that there is evidence of an 'engagement' with the material (Stiles, 1993); that there enough evidential links with material (Turpin et al, 1997).
- That sufficient data is presented to allow readers to judge the interpretation themselves (Coyle, 1995; Harper, 1994b; Hayes in Parker, 1997d; Potter, 1996a) or to develop rival interpretations (Henwood & Pidgeon, 1994).
- That there is evidence of an interation and dialectical movement between reading and interpretation -- ie that the analysis is not just the result of a 'first pass' over the material (Stiles, 1993).
- That there is a transparency to the analysis (Hayes in Parker, 1997d; Parker, 1997c). For example, that there is a description of the 'internal process' of analysis (Stiles, 1993); that the steps in the analysis are specified (Elliott et al,1996; Sherrard, 1996); that there is enough documentation (Elliott et al, 1996; Henwood &
Pidgeon, 1992) to demonstrate a 'confirmatory trail' (eg inclusion of transcripts -- Sherrard, 1996).
- That sufficient rigour is shown.
- That the analysis is coherent in the context of previous research -- ie that it builds upon previous insights (Potter, 1996a).
- That there is evidence of reflexivity in that theory has been influenced by, and not simply imposed upon, data (Henwood & Pidgeon, 1992; Stiles, 1993).
- That the researcher practices an accountability (Turpin et al., 1997) and respect (Elliott et al., 1996) towards participants.
- That the analysis is fruitful in terms of generating new research questions (Coyle, 1995; Potter & Wetherell, 1987) and is of practical relevance and utility.
Letter to Consultant Psychiatrists informing them about the study
Letter to service-users informing them about the study
Consent form for service-users
Guidelines for interviews with service-users
Do you know what your diagnosis is?
When did you first hear of it?
Has it changed over time?
Do you agree with it?
Has your agreement with it changed over time?
How did you come to be referred to a psychiatrist?
Have you ever been said to be paranoid?
If so, why?
Do you think any of your beliefs are abnormal?
How do you think others decide that a belief is paranoid?
If you feel you are paranoid, why do you think this has occured?
Do you feel more strongly about your beliefs at particular times?
Do you know others who are paranoid?
If so, what do you think of them?
Have you ever acted on your beliefs?
What do you think about the treatment you have received?
How could it have been different?
Does medication help?
Does talking help?
What is the best way to help people considered paranoid?
What advice would you give to professionals in training about how to help people who are paranoid?
What advice would you give to people considered paranoid themselves?
Guidelines for interviews with psychiatrists and CPNs
Diagnosis in general
How would you define paranoid/persecutory delusions?
How do these delusions manifest in different diagnoses?
What do you look for when making such diagnoses?
How important are delusions in coming to a diagnosis?
Specific questions in relation to the case of patient X:
What diagnosis do they have?
Has the diagnosis changed over time?
How did they come to be referred to you? Who referred?
What did you make of them initially?
Did they have any beliefs you considered abnormal?
What made you think that?
Were you influenced by other information?
Did you attempt to check any of these out?
How did you decide?
Could others have similar beliefs and not be mentally ill?
Did patient X present an element of risk?
Did other diagnosers agree with your diagnosis?
What was the cause of the delusions?
What treatment did they receive? Did it help?
If medication did work why, if not, why not?
What do you think about talking treatments?
What comments would patient X make about the service they received?
Also questions to compare psychiatrists view of other patients with patient X:
Do patients X and why have anything in common? Are they different in any way?
For interviews with CPNs
Same but with some additional questions:
Do you remember how patient X first came to be referred?
How did they get referred to you?
What did you think of the diagnosis?
What is the best way of helping someone like this?
What is the relative importance of belief and distress?
Guidelines for interviews with GPs
What was it that prompted you to refer patient X to a psychiatrist?
Did they have any ideas or beliefs which struck you as odd?
What made you think that? Did you attempt to check it?
How important was the belief in comparison with other factors (eg family?)
Were there other things that made you concerned?
What marks people like this out compared to your other patients?
Do other patients ever have similar beliefs?
What would tend to make you suspect a belief is deluded rather than reality-based?
What did you hope a psychiatrist might be able to do?
What does patient X make of their behaviour after they've been treated?
Do you have any ideas about the cause of the delusions?
What treatment have they received? Did it help?
Biographical and interview details
Name int. location1 Age Gender Occupation int. date
Paul Dench Room A 20-30 male unemployed 01/07/94
Peter Shaw Room B 40-50 male unemployed 15/07/94
Sharon Harrison Room A 20-30 female unemployed 18/07/94
Geoff Nelson Room A 30-40 male unemployed 18/07/94
Alan Roberts Room C 40-50 male unemployed 16/12/94
Ian Berry Room C 25-35 male unemployed 21/12/94
John Stewart At home 40-50 male unemployed 31/01/95
Mike Sullivan At home 40-50 male retired 03/02/95
Don Hall At home 55-65 male unemployed 03/02/95
Dr Smith Room A 50-60 male Cons. Psych. 13/06/95
Dr Lloyd Room D 40-50 male Cons. Psych. 16/06/95
Dr James Room E 50-60 male Cons. Psych. 14/07/95
Dr Jacobs Cons. room 40-50 male GP 14/07/95
Dr Chapman Cons. room 40-50 male GP 21/07/95
Dr Cornwell Cons. room 50-60 male GP 21/07/95
Dr Howard Cons. room 35-45 female GP 28/07/95
Julie Mason Room F 25-35 female CPN 28/07/95
Terry Reid Room G 35-45 male CPN 28/07/95
Edward Jackson Room H 35-45 male CPN 02/08/95
Dr Flynn Cons. room 35-45 male GP 16/08/95
Dr Williams Room G 25-35 female SHO 18/08/95
1 Key to locations
Room A = Interview room near community staff offices on a small hospital site
Room B = Interview room in day hospital (i)
Room C = Interview room on hospital ward
Room D = Interview room near hospital office suite (i)
Room E = Meeting room in day hospital (ii)
Room F = Office in hospital office suite (ii)
Room G = Office in hospital office suite (i)
Room H = Office near Room A
Cons. room = Surgery consulting room
Form for service-users to give consent to interview professionals
Consent form for psychiatrists
Consent form for GP/CPN
List of interview combinations
Paul Dench -- Dr Smith (Cons. Psych.)
-- Julie Mason (CPN)
-- GP refused
Peter Shaw -- Dr Williams (SHO)
-- Dr Jacobs (GP)
Sharon Harrison -- Dr Smith (Cons. Psych.)
-- Dr Howard (GP)
Geoff Nelson -- Dr James (Cons. Psych.)
-- Edward Jackson (CPN)
-- GP declined (because user had, by then, left area)
Alan Roberts -- Dr Lloyd (Cons. Psych.)
-- GP/referrer inaccessible
Ian Berry -- Dr Lloyd (Cons. Psych.)
-- Terry Reid (CPN)
-- GP/referrer inaccessible
John Stewart -- Dr Smith (Cons. Psych.)
-- Dr Cornwell (GP)
Don Hall -- Dr Smith (Cons. Psych.)
-- Julie Mason (CPN)
-- Dr Flynn (GP)
Mike Sullivan -- Dr Smith (Cons. Psych.)
-- Dr Chapman (GP)
"Pause" = > 1 sec
"Long pause" = > 5 secs
"Very long pause" = > 15 secs
Time taken to transcribe interviews
Name Int length (Counter1) (mins)Tran length (mins) Ratio (int/tran)
Sharon Harrison 525 33.3 377 (6 hr 17 mins) 1/11.3
Paul Dench 713 45.2 467 (7 hr 47 mins) 1/10.3
Geoff Nelson 411 26.0 263 (4 hrs 23 mins) 1/10.1
Peter Shaw 595 37.7 458 (7 hrs 38 mins) 1/12.2
Alan Roberts 534 33.8 237 (3 hrs 57 mins) 1/7.0
John Stewart 357 22.6 172 (2 hrs 52 mins) 1/7.6
Ian Berry 355 22.5 168 (2 hrs 48 mins) 1/7.5
Mike Sullivan 662 41.9 398 (6 hrs 38 mins) 1/9.5
Don Hall 516 32.7 295 (4 hrs 55 mins) 1/9.0
Dr Smith 854 54.1 438 (7 hrs 18 mins) 1/8.1
Dr Lloyd 845 53.5 393 (6 hrs 33 mins) 1/7.4
Dr James 287 18.2 187 (3 hrs 7 mins) 1/10.3
Dr Jacobs 92 5.8 72 (1 hr 12 mins) 1/12.4
Julie Mason 414 26.2 248 (4 hrs 8 mins) 1/9.5
Terry Reid 429 27.2 254 (4 hrs 14 mins) 1/9.3
Edward Jackson 282 17.9 210 (3 hrs 30 mins) 1/11.7
Dr Howard 236 14.9 207 (3 hrs 27 mins) 1/13.9
Dr Williams 276 17.5 147 (2 hrs 27 mins) 1/8.4
Dr Flynn 147 9.3 78 (1 hr 18 mins) 1/8.4
Dr Cornwell 135 8.6 87 (1 hr 27 mins) 1/10.1
Dr Chapman 177 11.2 120 (2 hrs) 1/10.7
Don's review 638 40.4 311 (5 hrs 11 mins) 1/7.7
----------- ------ ------ ----- ----------------- ---------
22 9480 600.5 5587 93 hrs 7 mins 1/9.65*
Average length of all interviews = 27.3 minutes
Average length of user interviews (9) = 32.9 minutes
Average length of psychiatrist interviews (4) = 35.8 minutes
Average length of CPN interviews (3) = 23.8 minutes
Average length of GP interviews = 9.9 minutes
Average transcription ratio = 1/9.30*
*These figures are different due to rounding up and the different calculations. The one in the total column is the average of the total of all transcription rates. The one in the paragraph above is calculated by dividing total length in minutes transcribing by the total interview length.
1 counter on tape recorder = 3.8 seconds
Steps taken in discourse analysis
Immediately after the interviews I wrote brief memos highlighting interesting aspects of the interviews. During transcription some extracts were marked according to certain criteria (eg relevance to research questions; ability to surprise and aid in generation of theory and interpretation about the interviews and paranoia; analytically interesting -- ie that there was a lot going on in the extracts; good illustrations of dominant themes emerging etc) and also further memos were kept of emerging themes, rhetorical devices used and so on. Stenner & Stainton Rogers (1998) note how interpretation draws on the researcher's (theoretically and culturally) embedded expectations but comment that this does not preclude surprise. At the time of doing this analysis I was also involved in writing the chapters for Part I. To orient my reading I listed a number of analytic questions (see appendix 16)(1).
Following transcription, my reading of the memos and repeated reading of the transcripts aided the development of some broad and inclusive categories (see appendix 17) -- this was similar to the process described by Lewis (1995) where within-interview themes were highlighted, followed by themes occuring across interviews. I then re-read the interviews putting extracts under these broad headings. Also headings were generated from other interesting extracts (matching the criteria in step 1). Hollway (1989) and Potter & Wetherell (1987) note how these categories may be discarded after initial use since they are only provisional, a way into the analysis and a way of 'troubling' (Curt, 1994) talk about paranoia.
Following this process, I became somewhat overwhelmed by the mass of material and possible categories -- this is not unusual (see Potter & Wetherell, 1987). There was a need to select a narrower focus to structure the material. I thus selected three dominant themes for further analysis: the construction of emotions and paranoia, plausibility and medication. This choice was not entirely arbitrary. It seemed that an analysis of paranoia required an understanding of how it was brought into being and diagnosed, how it was said to be experienced and to affect action and how it was treated. In addition, there were some subsidiary reasons for these choices:
a) Categories covering a range of issues in paranoia (paranoid beliefs, actions, emotions and distress) seemed to be connected in the extracts and I wished to explore this further. Moreover, whereas the literature (and my analysis of it) had focused on intellectual factors, emotional ones also seemed important concerns for the users and professionals. Furthermore, it seemed that where distress in paranoia was discussed it was treated in an essentialist manner often to support conservative views about causation and treatment and I felt alternative readings might be possible.
b) Plausibility seemed an interesting issue since my previous work (Harper, 1991, 1994b) had revealed some of the interests at work in its construction. It was obviously an area of contested interpretation between many users and professionals and so I considered it analytically worthwhile to investigate further how different implied assumptions of the world worked to construct different versions of reality.
c) As I read the interview extracts pertaining to diagnosis, I began to notice how little attention I had paid to issues of treatment in my theorising and as I investigated further I saw how this mirrored a lack of discursive into treatment of these kinds of difficulties apart from psychotherapy. I wanted the study to have something to say about the practical treatment that service-users got and it also was a contested area and one which was fascinating since many extracts appeared to challenge the received view of medical decision-making in relation to medication.
Many of the original categories ended up being tangential to the aims of the study and/or to the analytic theme whilst others became redundant or could be collapsed into richer categories.
The interviews were re-read and extracts pertaining to these themes were listed as were relevant extracts noted earlier in the process, relevant parts of the memos and so on. At the same time a further broad set of inclusive categories was developed for each theme with extracts listed under each heading (see appendices 18 and 19 for categories pertaining to the themes in chapters 5 and 6). Although following broadly similar practices of categorisation, there were differences in the way this was done for each chapter (according to my aims for each chapter and the shape of the material) and so I will describe the process for each. The description of the analysis of chapter 5 is, however, most illustrative.
Step 4a: Chapter 4
Within chapter 4, having collected extracts under category headings, I wrote a theoretical piece, drawing on social constructionist work on emotion since this had not been a concern in Part I. The categories from the original list (see appendix 17) which seemed connected to this theme were:
I then repeatedly read through the extracts under these headings using similar analytic questions to those in appendix 16 and picked out themes relating to issues like agency and oppositions like thought/feeling, mind/body and so on. Simultaneously, I developed theoretical ideas relating to embodiment reading the current literature. As the analysis developed I realised that my analysis was developing into an examination of how the issue of whether fears were regarded as warranted was influenced by gender, race and so on. I realised that this was linked more to plausibility (the subject of chapter 5) than emotion and moved that analysis to the next chapter. I returned to the original categories and realised that an analysis structured around oppositions was more helpful: fear as present/absent; emotion/belief; belief/action; action/violence. I then worked on these categories to develop a coherent linear structure that followed a reasonably clear theoretical framework. The process of structuring an analysis is detailed. My discussion of the analysis in chapter 5 is a good illustration of how this structuring is done.
Step 4b: Chapter 5
For chapter 5 (on plausibility) I had already identified a large number of extracts pertaining to this theme and had begun to identify (eg in my memos) some rhetorical devices. Here, the analysis was more clearly driven by the extracts since the theoretical work had been done earlier (eg in chapters 2 and 3). From the initial listing (see appendix 17) a number of extracts and categories seemed relevant here: a number of the 'subject positions' extracts; 'rational' or 'irrational'/ plausibility of accounts (item 8); paranoia and positioning others/cutting others out (item 10); pathologisation of suspicion (item 12); everyday/mundane versus bizarre accounts (item 14). Re-reading these extracts, I developed a list of categories which covered most of the extracts and developed a listing of rhetorical devices used (see appendix 18). This was the start of a process of developing a coherent linear structure for the chapter and analysis. Going through these categories it became clear they could be collapsed into high-order themic categories.
The categories were systematized (ie links found between them) and were grouped under similar themes. This process reminded me of other examples of categories in the transcripts and these were then sought. An initial grouping of the categories (item numbers refer to categories listed in appendix 18) was:
i). Accomplishing insight (incorporates: 1, 5, 6)
ii). Non-rational criteria (incorporates: 3, 11, 19) plus a new one: how plausibility and rationality are gendered (which was originally in chapter 4)
iii). Rhetorical strategies of im/plausibility (incorporates: 7, 8, 12, 13, 15, 21, 24) plus: 'I checked with the patient for other evidence of delusion'; I checked with family/others and/or we investigated; how is 'oddness' achieved (incorporating 20)
iv). Rationality is not enough (incorporates: 4, 5, 8, 10, 14, 23)
v). Miscellaneous (incorporates: 9, 16, 17, 22)
I tried to develop a framework for the analysis so the categories could be grouped to tell a linear story which was relevant to the central theme of the chapter -- ie plausibility. As the categories were developed some extracts did not fit which led to a systematization of the category. There was also a process of thinning out. Some categories which seemed to be tangential were ommitted (eg the 'miscellaneous' category). An initial draft structure was then further refined.
Step 4c: Chapter 6
The theme in this chapter was hardly covered in my initial categorisation (see appendix 17) but emerged later as I thought about my own and others neglect of this important topic. As a result I went through the transcripts noting all instances where medication was discussed whilst also reading literature on materiality and embodiment and other research on medication. I then categorised these instances -- the resulting categories can be seen in appendix 19. Following a similar process to the other two chapters (although this was, in fact the first analytic chapter which was written) I developed a reading which had a linear structure. The material was most easily structured along a number of issues, some of which were oppositional in nature. This led me to return to the transcripts to find that the other issues could be represented oppositionally too.
Once an overall linear structure had been developed, the criteria for choosing extracts in each chapter were similar: which were the shortest extracts where a particular position was most clearly demonstrated? Which could be understood easiest without having to refer to the rest of that particular interview? Which gave the most diversity (in terms of professional/user; variations on the theme etc)?
Positions were briefly described in each chapter pertaining to the analytic themes. In particular they were linked to my earlier theoretical and cultural analysis and to relevant literature which could be seen to provide further evidence of these devices and discursive positions, their analytical validity and usefulness.
At points, especially in the second half of chapter 6, some extracts were chosen to develop a fuller analysis -- one which drew on many of the positions developed from the concourse. The criteria for choosing these extracts were that they were: relevant to the aims of the study; considered to be rich examples of the themes to be discussed (eg where there were multiple positions and variability between discourses occuring within the boundaries of one extract).
Questions used in analytic reading
What variations are there in the text? (Potter & Wetherell, 1987)
What patterns are present? (Potter & Wetherell, 1987; Gill, 1996)
What dominant narratives are there?
What evidence of resistance is there? (Parker, 1992)
What kind of positions are set up and taken up? (Davies & Harré, 1990; Stenner, 1993)
What kind of subject is constructed here? (Parker, 1992)
What kind of identities are created? (Parker, 1992)
What kind of objects are constructed? (Parker, 1992)
What concepts are constructed as given, obvious or taken-for-granted? (Gill, 1996; Potter, 1996b)
What type of world is constructed here?
What means of describing and defining paranoia are there?
What subjectivities do these definitions make possible? (Parker, 1992)
What kind of subjectivities do coping strategies construct? (Parker, 1992)
How is reality negotiated? (Potter, 1996b)
How is it made to appear obvious that delusions are implausible? (Edwards & Potter, 1992)
What rhetorical strategies and devices are used? (Edwards & Potter, 1992)
What oppositions are implied? (Billig et al., 1988)
What metaphors are used?
Which of these effects occur frequently?
What is not said that might have been? (Billig, 1991c; Gill, 1996; Parker, 1992)
To what problems might these responses be solutions? (Gill, 1996)
What ideological effects might these have?
Broad inclusive categories used at start of analysis of interview material
1. List of subject positions (10 pages)
2. List of how delusions can be described
3. Not trusting others/ questioning their motives/ being talked about
5. Paranoia and violence
6. Phenomenology of paranoia:
- The self as knowing
- Seeing events as intentional
- Warrants for action
- Interpreting connections between events
- Evangelising others about interpretations
- Totalizing paranoia
7. Feeling someone is after you
8. 'Rational' or 'irrational'/ plausibility of accounts
9. Pure categories/messy real life
10. Paranoia and positioning others/cutting others out
11. Professional versus lay accounts
12. Pathologisation of suspicion
14. Everyday/mundane versus bizarre accounts
16. Paranoia as symbolic
18. Paranoia and culture
20. Categories relating to particular symptoms (cf. Harper, 1991)
Broad inclusive categories aiding in the development of
discourses of plausibility
2. Checking with others - truth as intersubjective
3. Assumptions of the world
4. Belief as meaningful, not just rational
5. Hard not to believe delusions
6. Managing the self: 'It sounds weird, but...'
7. 'That might be reasonable, but not this...'
8. Judging the reasonableness of beliefs by their effects on user's life
9. There is no disagreement about the reliability of diagnosis
10. Belief may be existentially true -- the system is out to get you
11. Non-rational diagnostic criteria
12. 'It could be true, but...'
13. 'It's inherently unbelievable' -- assertion
14. Distress is more important than bizarreness
15. Hard to distinguish delusion from the truth
16. Beliefs as contestable
17. Delusions can be covered up
18. Risk to others/self
19. Diagnosis swayed by family/relatives
20. Dealing with rule-breaking and 'oddness'
21. Hard to convince others and explaining why they don't believe
22. Others may believe in your 'delusion'
23. Untestable beliefs
24. Just asserting your belief is true
Broad inclusive categories aiding in the development of
the concourse of medication discourse
1. Medication helpful
2. Medication the only treatment
3. Own choice/agency
4. Medication not targeted, crude
5. Medication optimism
6. Medication harmful
7. Medication not working
8. Medication potentially dangerous
9. Medication no good
11. Medication as a constraint
12. Changing medication
14. Doctors' views
15. How does medication work?
16. Medication mid-road
17. Medication versus talk
18. Medication not the only thing
19. Medication and diagnosis
20. Importance of users' rights
21. Medication and talk (together)
22. Medication related to relapse/recovery
23. Users' views
1. Since engaging in this process I have come across McGuire's (1997) heuristics for hypothesis generation which might have been helpful in analysis.
Back to title page