Corpus linguistics and clinical psychology: examining the psychosis - - PowerPoint PPT Presentation

Corpus linguistics and clinical psychology: examining the psychosis continuum

Dr Luke Collins | @LukeCCollins Prof Elena Semino | @elenasemino

Dr Zsófia Demjén | @ZsofiaDemjen Dr Andrew Hardie | @HardieResearch Dr Peter Moseley | @peter_moseley Dr Ben Alderson-Day | @aldersonday Prof Angela Woods | @literarti Prof Charles Fernyhough | @cfernyhough

ZD AH PM AW CF BA-D

▪ Voice-hearing ▪ Data: Interviews with spiritualists and users of mental health services ▪ The ‘psychosis continuum’ ▪ Our corpus-based approach ▪ Selected findings ▪ Summary and concluding reflections

Overview

▪

Hearing voices that others cannot hear

▪

Auditory Verbal Hallucinations (AVHs)

Voice-hearing

➢

Hallucinations are a primary diagnostic criterion for various psychotic disorders (notably, schizophrenia) according to the Diagnostic and Statistical Manual of Mental Disorders (DSM-5)

➢

AVHs are present in a range of mental health difficulties, including depression and anxiety, post-traumatic stress disorder, emotionally unstable personality disorder, and obsessive-compulsive disorder (van Os & Reininghaus, 2016).

American Psychiatric Association (2013) Diagnostic and statistical manual of mental disorders. Fifth Edition. https://doi.org/10.1176/appi.books.9780890425596 van Os, J. and Reininghaus, U. (2016) Psychosis as a transdiagnostic and extended phenotype in the general population. World Psychiatry 15(2): 118-124.

➢

AVHs also occur as a positive and meaningful experience for voice-hearers, in the absence of any need for clinical care (Baumeister, Sedgwick, Howes and Peters, 2017).

Baumeister, D., Sedgwick, O., Howes, O. and Peters, E. (2017) Auditory verbal hallucinations and continuum models of psychosis: A systematic review of the healthy voice-hearer literature. Clinical Psychology Review 51: 125–141.

➢

For example, spiritualists engage in mediumship: communication with departed spirits

Voice-hearing

▪

the terms they would use to describe their experiences

▪

the qualities of the voice-hearing experience

▪

the content of the voice-hearing experience

▪

the voices as having their own character or personality

▪

the onset of voice-hearing

▪

changes in the experience over time

▪

participants’ beliefs about/understanding of the experience.

Interview

Data

EIP Service Users Spiritualists ▪ Self-identified Spiritualists ▪ 27 participants ▪ 237 770 tokens ▪ < 9 months of using intervention services ▪ 40 participants ▪ 205 941 tokens

For example, if I was talking to you in the context of a church, so I'm hearing someone telling me something to tell you, and I'm telling you that, so I'm having the conversation with you, but also internally in my head, I'm having a conversation with the other person With that one, it's, it's not talking to me or with me, it's talking at me. It's telling me you know what I've done wrong, what should have happened. It swears a lot more than I swear, it's, it's very like an aggressive voice.

“The continuum view holds that psychotic symptoms vary along dimensions such as distress, vividness and duration in clinical and non-clinical groups” (Waters and Fernyhough, 2019: 717)

The ‘psychosis continuum’

Waters, F. and Fernyhough, C. (2019) Auditory Hallucinations: Does a continuum of severity entail continuity in mechanism? Schizophrenia Bulletin 45(4): 717-719.

Healthy Clinical Voice-hearers Healthy Voice-hearers

▪ Data: typically, interviews with members of different groups ▪ Analysis: coding for relevant phenomena and statistical comparison:

➢ Statistically significant differences, or ➢ Similarities, where no significant difference has been found.

▪ Evidence of continuity across clinical and non-clinical populations with respect to phenomenological aspects of voice-hearing, such as loudness, location and personification. ▪ Evidence of differences with respect to the interpretation and evaluation of voice-hearing experiences, and the voice-hearer’s degree of perceived control on the voices.

Previous work on the continuum

▪ Similarity/difference: – Keyness analyses at the level of semantic domains:

• Interview transcripts for one group vs. the transcripts for the other

group → candidates for differences;

• Each set of transcripts vs. oral history interviews in BNC →

candidates for similarities. ▪ Continuity/discontinuity: – Plotting the distribution of words belonging to different groups of related semantic domains in each interview for both groups: bar charts and box plots → do we observe overlaps between the two groups?

Our approach

▪ An automated tagging process whereby each token is allocated to a semantic category ▪ 21 general semantic domains; 232 more specific sub-domains

A

general and abstract terms

B

the body and the individual

C

arts and crafts

E

emotion

F

food and farming

G

government and public

H

architecture, housing and the home

I

money and commerce in industry

K

entertainment, sports and games

L

life and living things

M

movement, location, travel and transport

N

numbers and measurement

O

substances, materials,

• bjects and equipment

P

education

Q

language and communication

S

social actions, states and processes

T

time

W

world and environment

X

psychological actions, states and processes

Y

science and technology

Z

names and grammar

E1 Emotional Actions, States and Processes E3 Calm/Violent/Angry E4 Happiness and Contentment E5 Bravery and Fear E6 Worry and Confidence

E2+ Like E2- Dislike hate, can_not_stand like, love, liked E2++ Like prefer E2+++ Like favourite

UCREL Semantic Analysis System (USAS)

E2 Liking

EIP Service Users Spiritualists Oral History Interviews (BNC1994) EIP Service Users Spiritualists

Keyness comparison

Direct comparison:

▪

Highly contrastive Indirect comparison:

▪

Similarity and difference

▪

62 key ‘Spiritualist’ domains

▪

44 key ‘Service User’ domains

▪

23 shared key domains

▪

20 distinct ‘Spiritualist’ domains

▪

29 distinct ‘Service User’ domains

LL: 6.63+; Log Ratio: 1.0+

Similar and different

➢

Semantic domain could be both ‘shared’ and ‘distinct’, depending on the keyness comparison E5- Fear/Shock frightened, fear, shock, scared, terrified

▪

Shared key domain Spiritualists Service Users

▪

Key ‘Service User’ domain in direct comparison

LL LogR 74.69 1.42 341.46 2.52 61.32 1.11

Reference corpus Spiritualists Service Users

(Dis)continuity?

0.00 0.06 0.12 0.18 0.24 0.30 0.36 0.42 0.48

116DX Yan 108OC Harry Ulrik Ryan 105LT Alex 111UC Jane 117KD 110MX 130MN 113ND 126FJ Kath 115IB Xander 106II 125KT 122KX 131KN Fred 119UX Violet 123BI Will 109OT Mike Dan 102HC 104LT Iris 101BE Hugh 120MD Anthony Matt 118KC 114HT Kate Zara Bill 132DL 103TT Neil Sean 128AH 131TG Eric Grace Leah Olivia Chris Toby Brad Jade Fran Page Emma Nina Ian Gail Orla Dawn Liam Carl

E5- Fear/Shock Continuity

(Dis)continuity?

E5- Fear/Shock

Key themes

➢

The direct and indirect keyness approach identified 122 key domains

➢

We grouped domains into themes: ▪ Affect ▪ Control ▪ Meaning-making ▪ Sensory input

➢

The groupings correspond with aspect of voice-hearing that have previously been indicative of similarities/differences in the experiences of clinical/non- clinical populations (Baumeister et al., 2017)

➢

These groupings account for 36 of the key domains (29.5%)

Baumeister, D., Sedgwick, O., Howes, O. and Peters, E. (2017) Auditory verbal hallucinations and continuum models of psychosis: A systematic review of the healthy voice-hearer literature. Clinical Psychology Review 51: 125–141.

Groupings

▪ Affect: Negative emotions, Positive emotions, Negative evaluations of self ▪ Control: Disengagement, Command over, Development (of skills) ▪ Meaning-making ▪ Sensory Input: Loudness, Strength, Other senses, Cognition

➢

We plotted the dispersion of the relative frequency values for terms in these groupings

❖ This allowed to examine different realisations of ‘(dis)continuity’

Affect

▪ Negative emotions ▪ Positive emotions ▪ Negative evaluations of Self

E2- Dislike hate, hates, hated, hatred.. E4.1- Sad upset, grief, cry, depressed.. E5- Fear/Shock scared, scary, panic, fear, frightened.. E6- Worry anxiety, stress, distressing, worry..

Reference corpus comparison Service Users Both Both Service Users Direct comparison Service Users Service Users Service Users Service Users

0.00 0.20 0.40 0.60 0.80 1.00 1.20

Yan 111UC 116DX 110MX 108OC 118KC 130MN 119UX 114HT 117KD 109OT Will 126FJ 104LT 125KT 102HC Iris Kate 122KX Mike Kath 113ND 101BE 115IB 123BI 120MD 105LT 106II Xander Jane Violet 131KN Bill 103TT Alex Toby 131TG Anthony Ryan 128AH Leah Hugh Zara 132DL Chris Eric Brad Dan Fred Jade Matt Sean Olivia Emma Ian Nina Grace Harry Liam Gail Neil Dawn Orla Page Carl Ulrik Fran

Negative emotions

E2- Dislike E5- Fear/Shock E6- Worry E4.1- Sad

Overlap

Negative emotions

E2- Dislike E5- Fear/Shock E6- Worry E4.1- Sad

Overlap

Affect

▪ Negative emotions Fran: when I feel anxious or I’m feeling down or upset, the voice comes out stronger. Fran: I hate being in the house being by myself. […] Like I should be like enjoying it, you know, have the house to myself like one night! But I can’t do that because I feel too scared to you know.

Control

▪ Disengagement

A1.9 Avoiding leave_alone, avoid.. Q2.1- Speech: Not communicating shut_up, keep_quiet.. X5.1- Inattentive ignore, distract..

➢

clinical voice-hearers are more likely to try to actively ignore voices (Kråkvik et al. 2015)

➢

Increasing number of ‘relational’ therapeutic approaches

Kråkvik, B., Larøi, F., Kalhovde, A.M., Hugdahl, K., Kompus, K., Salvesen, Ø., Stiles, T.C. and Vedul-Kjelsås, E. (2015) Prevalence of auditory verbal hallucinations in a general population: A group comparison study. Scandinavian Journal of Psychology 56(5): 508–515.

Reference corpus comparison

• Service Users

Both Direct comparison Service Users Service Users Service Users

0.00 0.05 0.10 0.15 0.20 0.25 0.30 0.35 0.40

101BE 103TT 104LT 105LT 109OT 110MX 114HT 118KC 119UX Chris Emma Harry 106II 108OC 111UC 115IB 122KX 125KT Fran Gail 102HC 117KD 123BI 126FJ 131KN 131TG Grace 113ND 130MN 132DL Anthony Brad Carl Ryan 120MD 128AH Bill Dan Fred Dawn Eric Ian 116DX Hugh Jade Neil Alex Orla Kate Leah Matt Jane Page Iris Nina Mike Olivia Liam Toby Kath Ulrik Sean Violet Yan Will Xander Zara

A1.9 Avoiding Q2.1- Speech: Not communicating X5.1- Inattentive

Disengagement

Control

▪ Disengagement

Xander: if I tell him to shut up, he won’t listen, he’ll get worse! Zara: yesterday I heard it, it made us jump and look around, and then obviously I listened and I thought, just go away, leave us alone, then it went away. So it’s [...] in the past I listened, used to listen to the voice […] When it first started happening, and I was doing as I was told. But I’m not let- letting it win this time. 102HC: I was saying to her, ‘please leave me alone, God bless you, go on your way and leave me alone’, because I had to get to sleep. however I don’t want them to just disappear […] I have a connection with them now it would be nice if […] he would just sort of leave me alone

(Dis)continuity

▪ Continuity between Spiritualists and Service Users with a clinical sub-group ▪ Affect: Negative emotions, Positive emotions, Negative evaluations of self ▪ Control: Disengagement, Command over, Development (of skills) ▪ Meaning-making ▪ Sensory Input: Loudness, Strength, Other senses, Cognition ▪ Meaning-making

Meaning-making

A1.2+ Suitable relevant, appropriate A1.6 Concrete/Abstract philosophical, practical A11.1+ Important important, main A5.2+ Evaluation: True evidence, prove Q1.1 Linguistic Actions, States.. message, means ➢ Research has identified differences in the way clinical and non-clinical populations interpret their AVHs (Daalman et al., 2011)

Daalman, K., Boks, M.P.M., Diederen, K.M.J., de Weijer, A.D., Blom, J.D., Kahn, R.S. and Sommer, I.E.C. (2011) The Same or Different? A Phenomenological Comparison of Auditory Verbal Hallucinations in Healthy and Psychotic Individuals. The Journal of Clinical Psychiatry 72(3): 320-325.

Reference corpus comparison

• Spiritualists
• Spiritualists

Direct comparison Spiritualists Spiritualists Spiritualists Spiritualists Spiritualists

0.00 0.10 0.20 0.30 0.40 0.50 0.60 0.70 0.80

Ian Liam Neil Violet Will Xander Harry Carl Emma Iris Mike Orla Grace Kath Nina Zara Jane Leah Eric Bill Dan Olivia Sean Fred Alex Kate Toby Yan Brad Gail 109OT 116DX Page Dawn Ryan 106II Hugh Chris Anthony 103TT Matt 105LT 128AH 113ND Fran Jade 119UX Ulrik 130MN 110MX 132DL 118KC 126FJ 131KN 101BE 108OC 125KT 117KD 131TG 111UC 115IB 122KX 102HC 120MD 123BI 114HT 104LT

A1.2+ Suitable A1.6 Concrete/Abstract A11.1+ Important

Meaning-making

A5.2+ Evaluation: True Q1.1 Linguistic Actions, States and Process: Communication

A1.2+ Suitable A1.6 Concrete/Abstract A11.1+ Important

Meaning-making

A5.2+ Evaluation: True Q1.1 Linguistic Actions, States and Process: Communication

Sensory Input

▪ Strength

S1.2.5+ Tough/Strong strong, strengths S1.2.5++ Tough/Strong stronger 103TT: When I say a volume, I mean a strength of impression. Rather than sound volume.

Reference corpus comparison

• Spiritualists

Direct comparison Spiritualists Spiritualists

Luhrmann, T.M. (2017) Diversity within the psychotic continuum. Schizophrenia Bulletin 43(1): 27-31.

➢ Researchers have questioned whether this reflects differences in the phenomenology of the experience or a particular vocabulary associated with mediumship (Luhrmann, 2017).

S1.2.5+ Tough/Strong S1.2.5++ Tough/Strong

Strength

0.00 0.02 0.04 0.06 0.08 0.10 0.12

102HC 105LT 111UC 114HT Alex Bill Brad Carl Chris Dawn Gail Grace Harry Ian Iris Jade Jane Kate Matt Mike Neil Olivia Orla Page Sean Toby Ulrik Violet Xander Yan 106II 125KT Dan Leah 101BE 108OC 130MN 131TG Emma Fred Kath Nina Will Zara 104LT 110MX 116DX 120MD 123BI 126FJ Eric Hugh 109OT 117KD 132DL Liam 119UX Anthony 128AH 113ND 122KX 131KN 118KC 115IB 103TT Fran Ryan

26/40 Service Users (65%)

S1.2.5+ Tough/Strong S1.2.5++ Tough/Strong

Strength

Sensory Input

▪ Strength Ryan: strength of feeling, strength of character, physical strength Fran: when I feel anxious or I’m feeling down or upset, the voice comes out stronger […] and then as I’m hearing it more, and it’s coming stronger and stronger and like it’s …I can’t get used to it. Anthony: when [the voice] comes, it’s really kind of strong

Identifying (dis)continuity

▪

Keyness analysis pointed to areas of similarity/difference

▪

Dispersion revealed (dis)continuities

➢

In most cases, we observed continuity between Spiritualists and Service Users, with ‘discontinuity’ to a (clinical) sub-group ▪ Negative emotions ▪ Negative evaluations of self ▪ Disengagement ▪ Loudness ▪ Other senses

➢

Continuity in references to ‘meaning-making’

* *

Identifying (dis)continuity

➢

Need to consider both quantitative and qualitative (dis)similarities

❖ Method allows us to consider degrees of (dis)similarity ❖ We can also consider ‘degrees’ of continuity (i.e. overlap) ❖ Statistical basis for ‘discontinuity’ ❖ ‘Breaks’ can help us to develop personalised clinical interventions.

Thank you