Dr She ryl Pre ntic e Pro f. Pa ul J. T a ylo r, Dr Pa ul Ra yso n - - PowerPoint PPT Presentation

POL E S APART ? E XPL ORING T HE E XT E NT OF SIMIL ARIT Y BE T WE E N E XT RE ME AND NON- E XT RE ME ME SSAGE CONT E NT

Dr She ryl Pre ntic e Pro f. Pa ul J. T a ylo r, Dr Pa ul Ra yso n

E XT RE ME ON NON- E XT RE ME ?

• “Yo u do n’ t wa nt to b e c a ug ht re d-ha nde d… sma sh

he r o n a pa rk b e nc h. T ha t use d to b e my tric k”

• “Yo u kno w g irls in g e ne ra l a re a ll rig ht. But so me o f

the m a re b itc he s…T he b itc he s a re the type tha t…ne e d to ha ve it stuffe d to the m ha rd a nd he a vy”

PE RSPE CT I VE 1: E XT RE ME

• VS. NON-E

XT RE ME

• Gro unde d in the a ssumptio n tha t e xtre mists po sse ss

unusua l wa ys o f thinking , o r a diffe ring psyc ho -lo g ic (Me ra ri)

• L

a ng ua g e use re fle c ts this (Pe nne b a ke r)

• Autho rs ma y a c tive ly se e k to diffe re ntia te the ir rhe to ric

fro m tha t o f the ‘ e ne my’ (Cho wdhury & K re b s, 2010; Awa n, 2007)

PE RSPE CT I VE 1: E XT RE ME

• VS. NON-E

XT RE ME

• Smith, Sue dfe ld, Co nwa y, a nd Winte r (2008)
• Co mpa re d 2 te rro rist a nd 2 no n-te rro rist g ro ups
• Ang ie e t a l. (2011)
• Co mpa re d 29 vio le nt a nd no n-vio le nt me ssa g e b o a rds
• Pa yne (2009)
• Co mpa re d Al-Qa ’ ida a nd We ste rn Go ve rnme nt

na rra tive s

PE RSPE CT I VE 2: E XT RE ME & NON-E XT RE ME OVE RL AP

• E

xtre mists ha ve b e e n fo und to de mo nstra te ra tio na lity, whic h is re fle c te d in the ir rhe to ric (Sprinza k 2000; Sto ut 2009)

• Ma instre a m a nd e xtre me so urc e s ha ve b e e n fo und to

spe a k to the sa me c o nc e rns (se e Awa n, 2007)

• Po litic a l a nd e xtre me – simila r stra te g ie s to win o ve r

suppo rte rs (c f. unifying te rms)

PE RSPE CT I VE 2: E XT RE ME & NON-E XT RE ME OVE RL AP

Ma y no t b e stra te g ic o r o ve rt…

• So c io ling uistic the o ry (Jo se ph, 2004)
• So c ia l ide ntity the o ry (T

a jfe l) Ho pkins & K a ha ni-Ho pkins (2009) a rg ue a g a inst e xtre me

• vs. no n e xtre me c la ssific a tio n

PE RSPE CT I VE 2: E XT RE ME & NON-E XT RE ME OVE RL AP

• Gutma nn (2007): e xtre me lite ra ture de me a ns the o ut-

g ro up a nd na rro ws unde rsta nding

• Studie s o f pre ss disc o urse
• Disc o urse s o f xe no pho b ia , se xism & ho mo pho b ia ;

le g itimize & re me dia te e xtre mism

• Studie s o f po litic a l disc o urse
• I

n/ o ut g ro up fe a ture s; so c ia l a nd mo ra l a rg ume nta tio n; wa rfa re justific a tio n

CURRE NT ST UDY

• Aim: to e sta b lish whe the r the re is sig nific a nt o ve rla p in

c o nte nt b e twe e n a n e xe mpla r se t o f e xtre me a nd no n- e xtre me o nline me ssa g e s

MAT E RI AL S

• E

xtre me c o rpus: 250 me ssa g e s, 425,516 wo rds, writte n b y me mb e rs o f kno wn e xtre mist g ro ups

• Co unte r-e xtre me c o rpus: 250 me ssa g e s, 208,932 wo rds,

fro m Muslim c le ric s a nd British Offic ia ls

• Ma instre a m c o rpus: 250 me ssa g e s, 107,018 wo rds,

dra wn fro m fo ur ME ne ws o utle ts

PROCE DURE

• T

e xts a na lyse d via Wma trix – sub je c t to CL AWS a nd USAS ta g g ing

• Se ma ntic c a te g o ry lists re trie ve d fo r e a c h o f the thre e

c o rpo ra

• T

ra nsfe rre d to lo g -like liho o d spre a dshe e t

• Use d a n a da ptio n to ke yne ss to lo o k a t b o th simila ritie s

a nd diffe re nc e s

PROCE DURE

Totals Tag Ext Coun Main corpus1 corpus2 corpus3 LL Ext O/U Coun O/U main O/U A1.1.1 4799 2829 1175 8803 5110.57 2407.11 1285.32 99.11 U O U A1.1.1- 4 4 2.32 1.09 0.58 4.35 O U U A1.1.2 1010 454 204 1668 968.36 456.10 243.54 8.57 O U U A1.1.2- 1 1 2 1.16 0.55 0.29 3.67 U O O A1.2 4 3 7 4.06 1.91 1.02 6.33 U U O A1.2+ 89 57 18 164 95.21 44.84 23.95 5.06 U O U

• bserved frequencies

expected frequencies

T hre e c o mpa riso ns: A1: E xtre me , Ma instre a m a nd Co unte r A2a : E xtre me , Ma instre a m a nd Muslim Co unte r A2b : E xtre me , Ma instre a m a nd British Offic ia l Co unte r All use : L L = 3.84; p < 0.05

RE SUL T S: ANAL YSI S 1

Cate gor y distinc tion E xtr e me Mainstr e am Counte r Signific antly ove r use d (L L => 3.84; p < 0.05)

142 144 187

Signific antly unde r use d (L L => 3.84; p < 0.05)

200 198 155

No signific ant diffe r e nc e ac r

• ss c or

por a (L L = < 3.84; f > 10)

49 49 49

Number of semantic categories significantly over and underused, and semantic categories showing no significant difference between the extreme, counter and mainstream messages

L L

= 10.85, p < .001

L L

= 11.89, p < .001

L L

< 1

RE SUL T S: ANAL YSI S 2A

Cate gor y distinc tion E xtr e me Mainstr e am Muslim Counte r Signific antly ove r use d (L L => 3.84; p < 0.05)

166 153 120

Signific antly unde r use d (L L => 3.84; p < 0.05)

167 180 213

Number of semantic categories significantly over and underused, and semantic categories showing no significant difference between the extreme, mainstream, and Muslim authored counter-extreme messages

L L = 6.77, p < .01 L L = 13.01, p < .001 L L = 1.02, p < .5

RE SUL T S: ANAL YSI S 2B

Cate gor y distinc tion E xtr e me Mainstr e am Br itish Offic ial Counte r Signific antly ove r use d (L L => 3.84; p < 0.05)

162 147 152

Signific antly unde r use d (L L => 3.84; p < 0.05)

174 189 184

Number of semantic categories significantly over and underused, and semantic categories showing no significant difference between the extreme, mainstream and British Official authored counter-extreme messages

L L = 1.35, p < .5 L L < 1 L L < 1

SUMMARY

• Ana lysis 1: e xtre me a nd ma instre a m - c lo se a lig nme nt,

c o unte r – o ppo sing pro file

• Ana lysis 2a : Muslim c o unte r - o ppo sing
• Ana lysis 2b : British Offic ia l a lig ns with ma instre a m &

e xtre me

• L

inks to wo rk o n te rro rist pa tho lo g y

• Silke , 1998
• Me ra ri, 1999

Wha t c o uld e xpla in the re la tio nships o b se rve d?

• Ve rb a l a g g re ssio n mo de ls fro m ling uistic s a nd psyc ho lo g y

L E VI NE E T AL . (2004): MUL T I DI ME NSI ONAL VAS

• De ve lo pme nta l I

nte ra c tio nist T he o ry a nd NS

• two e mo tio na l c irc uits: se lf pro te c tio n a nd so c ia l

b e ha vio urs

• se lf pro te c tio n - ne g a tive / c o mpe titive / a g g re ssive ;

so c ia l b e ha vio urs - po sitive / c o o pe ra tive

• Hig h c o o pe ra tive ne ss + lo w c o mpe tive ne ss o r

a g g re ssive ne ss = lo w re la tio na l ne g a tivity

• L
• w c o o pe ra tive ne ss + hig h c o mpe titive ne ss o r

a g g re ssive ne ss = hig h re la tio na l ne g a tivity

ARCHE R (2014): VE RBAL AGGRE SSI ON

• Assig ns a pa rtic ula r se t o f USAS c a te g o rie s to ve rb a l

a g g re ssio n

• Vio le nt/Ang ry, I

m/po lite ne ss, (L ac k o f) Re spe c t, Damag ing and De stro ying , E valuatio n: Go o d/Bad, E valuatio n: T rue /F alse , Co mparing : Diffe re nt, Spe e c h: Co mmunic ative /Unc o mmunic ative , a nd Spe e c h Ac ts: Spe aking /No t Spe aking

• Additio n o f Warfare c a te g o ry

ARCHE R (2014): APPL I CAT I ON

• E

valuatio n: Go o d/Bad, Damag ing and De stro ying , (L ac k o f) Re spe c t, I mpo lite ne ss o ve ruse d b y the e xt. me ssa g e s

• sha re d o ve ruse o f E

valuatio n: T rue /F alse b e twe e n e xt. a nd B/ O c o unte r me ssa g e s

• sha re d o ve ruse o f Vio le nt/Ang ry b e twe e n m/ s a nd B/ O

c o unte r me ssa g e s

• o ve ruse o f Warfare in the m/ s me ssa g e s

ARCHE R (2014): APPL I CAT I ON

• o ve ruse o f Spe e c h: Co mmunic ative a nd Spe e c h Ac ts

in the ma instre a m me ssa g e s

• o ve ruse o f Calm (a n o ppo sitio n c a te g o ry to

Vio le nt/Ang ry), Anti-War a nd Spe e c h Ac ts in the Muslim c o unte r me ssa g e s

COMBI NI NG ARCHE R (2014) & L E VI NE E T AL . (2004)

• Spe e c h: Co mmunic ative /Unc o mmunic ative , a nd

Spe e c h Ac ts: Spe aking /No t Spe aking ta ke n a s ‘ ve rb a l (un)c o o pe ra tive ne ss’

• Warfare , Vio le nt/Ang ry, Damag ing and De stro ying ,

I m/po lite ne ss, E valuatio n: T rue /F alse , E valuatio n: Go o d/Bad, a nd (L ac k o f) Re spe c t a re ta ke n a s ‘ ve rb a l (no n)a g g re ssive ne ss’

L E VI NE E T AL . (2004): APPL I CAT I ON

Figure 2: Multidimensional scale of verbal aggression and cooperativeness, adapted from Levine et al. (2004). The asterisk (*) indicates author’s addition.

L E VI NE E T AL . (2004): CAUSAL CHAI N

• Be lie f Syste ms T

he o ry

• o ne ’ s se lf c o nc e pt a c ts a s a n “a nte c e de nt to

g e ne ra lize d b e lie fs a b o ut o the rs suc h a s ho stility a nd a nxie ty o r a tta c hme nt a nd c a ring ” (L e vine e t a l. 2004, 264; c iting re se a rc h b y Ha milto n a nd Mine o 1999)

L E VI NE E T AL . (2004): CAUSAL CHAI N

• e g o c e ntric sta ndpo int > ho stile b e lie fs > ve rb a lly

a g g re ssive pre dispo sitio n > a g g re ssive me ssa g e se le c tio n

• pe rso na lly e mpa the tic > a tta c hme nt/ c a re > ve rb a lly

c o o pe ra tive pre dispo sitio n > c o o pe ra tive me ssa g e se le c tio n

• Orig ina lly de ve lo pe d fo r individua ls

L E VI NE E T AL . (2004): APPL I CAT I ON

CONCL USI ON & F UT URE DI RE CT I ONS

• F

urthe r wo rk ne e de d

• K

e y finding

• I

mplic a tio ns fo r c o unte r-te rro rism re se a rc h a nd po lic y

• Me tho d ma y b e unde re stima ting simila ritie s
• E

ffe c ts o f diffe ring c o nte nt c o ding syste m o r c o mpa riso n sta tistic

T ha nk yo u! s.r.pre ntic e 1@ la nc a ste r.a c .uk