Schizophrenia as a Disorder of...

Schizophrenia as a Disorder of CommunicationGuest Editors: Margaret A. Niznikiewicz, Marek Kubicki, Christoph Mulert, and Ruth Condray

Schizophrenia Research and Treatment

Schizophrenia as a Disorder of Communication

Schizophrenia Research and Treatment

Schizophrenia as a Disorder of Communication

Guest Editors: Margaret A. Niznikiewicz, Marek Kubicki,Christoph Mulert, and Ruth Condray

Copyright © 2013 Hindawi Publishing Corporation. All rights reserved.

This is a special issue published in “Schizophrenia Research and Treatment.” All articles are open access articles distributed under theCreative Commons Attribution License, which permits unrestricted use, distribution, and reproduction in any medium, provided theoriginal work is properly cited.

Editorial Board

Celso Arango, SpainKarl-Jurgen Bar, GermanyGregor Berger, AustraliaVaughan Carr, AustraliaL. Citrome, USAMichael T. Compton, USAChristoph Correll, USAPatrick W. Corrigan, USAEmmanuel Dias-Neto, BrazilSonia Dollfus, FranceRobin Emsley, South AfricaPeter Falkai, Germany

S. Hossein Fatemi, USARohan Ganguli, CanadaPablo V. Gejman, USAMorris B. Goldman, USADavid C. Henderson, USANakao Iwata, JapanMarkus Jager, GermanyBrian Kirkpatrick, USAVeena Kumari, UKMartin Lambert, GermanyTodd Lencz, USAAnil K. Malhotra, USA

H. Y. Meltzer, USAJouko Miettunen, FinlandD. Naber, GermanySteven G. Potkin, USAJohn Daniel Ragland, USALuis San, SpainHugo Schnack, The NetherlandsSibylle Schwab, AustraliaSteven J. Siegel, USAPhilip R. Szeszko, USA

Contents

Schizophrenia as a Disorder of Communication, Margaret A. Niznikiewicz, Marek Kubicki,Christoph Mulert, and Ruth CondrayVolume 2013, Article ID 952034, 4 pages

High Order Linguistic Features Such as Ambiguity Processing as Relevant Diagnostic Markers forSchizophrenia, Daniel Ketteler, Anastasia Theodoridou, Simon Ketteler, and Matthias JagerVolume 2012, Article ID 825050, 7 pages

From Semantics to Feelings: How Do Individuals with Schizophrenia Rate the Emotional Valence ofWords?, Ana P. Pinheiro, Robert W. McCarley, Elizabeth Thompson, Oscar F. Goncalves,and Margaret NiznikiewiczVolume 2012, Article ID 431823, 12 pages

Schizophrenia as a Disorder of Social Communication, Cynthia Gayle WibleVolume 2012, Article ID 920485, 12 pages

Faulty Suppression of Irrelevant Material in Patients with Thought Disorder Linked to AttenuatedFrontotemporal Activation, S. M. Arcuri, M. R. Broome, V. Giampietro, E. Amaro Jr., T. T. J. Kircher,S. C. R. Williams, C. M. Andrew, M. Brammer, R. G. Morris, and P. K. McGuireVolume 2012, Article ID 176290, 12 pages

Cognitive Control and Discourse Comprehension in Schizophrenia, Megan A. Boudewyn,Cameron S. Carter, and Tamara Y. SwaabVolume 2012, Article ID 484502, 7 pages

Stability of Facial Affective Expressions in Schizophrenia, H. Fatouros-Bergman, J. Spang, J. Merten,G. Preisler, and A. WerbartVolume 2012, Article ID 867424, 6 pages

Hindawi Publishing CorporationSchizophrenia Research and TreatmentVolume 2013, Article ID 952034, 4 pageshttp://dx.doi.org/10.1155/2013/952034

EditorialSchizophrenia as a Disorder of Communication

Margaret A. Niznikiewicz,1,2 Marek Kubicki,2,3 Christoph Mulert,4,5 and Ruth Condray6

1 VA Boston Healthcare System, Boston, MA 02301, USA2Harvard Medical School, Boston, MA 02115, USA3 Brigham and Women Hospital, Boston, MA 02115, USA4Department of Psychiatry and Psychotherapy, University Medical Center Hamburg-Eppendorf, 20246 Hamburg, Germany5 Department of Psychiatry, LMUMunich, 80331 Munich, Germany6University of Pittsburgh School of Medicine, Pittsburgh, PA, USA

Correspondence should be addressed to Margaret A. Niznikiewicz; margaret [email protected]

Received 18 April 2013; Accepted 18 April 2013

Copyright © 2013 Margaret A. Niznikiewicz et al. This is an open access article distributed under the Creative CommonsAttribution License, which permits unrestricted use, distribution, and reproduction in any medium, provided the original work isproperly cited.

The first characterizations of schizophrenia invoked theconcept of disordered thought and broken mind as centralto its clinical presentation [1, 2]. Interestingly, Bleuler’scharacterization of schizophrenia was couched in terms offour A’s association, with its focus on disordered language,affectivity, ambivalence, and autism, all of which implicatedifferent aspects of social function [3]. Bleuler capturedmuchthat is still relevant to the study of cognitive dysfunctionin schizophrenia and, in fact, covers quite well the topicof social communication dysfunction highlighted in thisissue.The research that followed these early characterizationsfirmly established the link between abnormal brain structureand function, mediated by genetics, and many clinical andcognitive manifestations of this devastating disease [4–6].Over the last several years, great progress has been achievedin the understanding of mechanisms of schizophrenia [7–10]. And while a comprehensive theory of schizophreniais still elusive, many compelling accounts of schizophreniapathology have been put forward and generated valuableinsights [8, 9, 11–15].

Within the field of study of the cognitive dysfunctionin schizophrenia, research has focused on different aspectsof what was described recently as “cold cognition” andincluded attention, memory systems that vary in duration,capacity, and operations, as well as language and perceptualmechanisms [9]. The last few years brought a welcomebroadening of this field of study as it has been noted that

abnormalities in “hot cognition” including abnormalities inemotion and affect processing from both face and voiceare an important component of schizophrenia pathology[16, 17]. Social cognition, whose functions draw on bothprocesses of “cold” and “hot” cognition, has become a focusof intense interest with studies addressing the ability toconvey one’s attitudes and intentions and adaptively predictand interpret the attitudes and intentions of others. Finallyand importantly, it has been increasingly recognized thatimpairments in social cognition contribute to both clinicaland functional outcomes in schizophrenia [18, 19].

The aim of the current issue is to broaden the discourseon the nature of cognitive dysfunction in schizophrenia.We propose that the cognitive dysfunction in schizophreniashould be conceptualized as a disorder of communicationrather that of language itself and that communication disor-der is the core clinical deficit of schizophrenia. We believethat an array of sensory and cognitive processes and theirinteractions enable human beings to enter into meaningfulsocial communication. Thus, communication in a humansociety involves a complex set of behaviors that includeboth formal languages embodied in the rules of phonology,grammar, syntax, and semantics, as well as behaviors thatallow conveying emotional states and attitudes, and finallyand importantly successful interpretation of these behaviorsin others. They depend on effective perceptual processeson one hand and on the successful recruitment of intact

2 Schizophrenia Research and Treatment

higher order processes such as working memory, attention,inhibition, and response selection on the other hand. Wewould like to argue that neither the focus on the study oflanguage nor the study of social cognition fully captures thecommunicative difficulties that patients with schizophreniaencounter. Rather, a schizophrenia sufferer is confrontedwith a poor ability to effectively use language and a poorability to deploy other communicative devices to achievesuccessful functioning in a society, both in social and pro-fessional settings. Given the complexity of the behaviorsunder consideration, most studies on cognitive impairmentin schizophrenia tend to adopt one of the two perspectives.Thus, the studies are conducted either within the frameworkof “cold cognition” and focus on the study of language,executive function, and perception or within the frameworkof “hot cognition” and focus on the study of emotion, theoryof mind, and agency, to name a few topics.

Articles in this issue reflect both perspectives: the focuson abnormal language function as a central characteristicof schizophrenia pathology on one hand and the concep-tualization of impairment in schizophrenia as a result ofabnormal processes of social cognition on the other. M.A. Boudewyn and colleagues examine language impairmentin schizophrenia in a review paper and argue that theextent of language processing difficulties is a function of thecomplexity of a linguistic message: the more complex themessage, the more impairment will be observed. Accordingto this conceptualization, schizophrenia patients should bemost impaired in the processing of discourse that calls forthemanipulation and reconciling ofmultiple sources of infor-mation. Conversely, they should be least impaired in the pro-cessing of single words and word pairs.The authors articulatetheir proposal of language impairment within the frameworkof domain general control mechanisms as mediated at thebrain level by dorsolateral prefrontal cortex (DLPFC) andanterior cingulate cortex (ACC). They argue that it is theseimpaired mechanisms that mediate most severe languagedysfunction in schizophrenia. This proposal is distinctlydifferent from the hypothesis that language abnormality inschizophrenia is primarily rooted in abnormal processeswithin semantic memory [15]. As the authors suggest, thehypothesis of prefrontally mediated language dysfunctionpromises to be a rich source of experimental approaches thatwill test how different levels of language complexitymap ontothe degree of dysfunction in the prefrontal systems, how wellthe results of nonlinguistic tests of cognitive control correlatewith the results of language tasks that purport to rely oncognitive control functions, are how tests probing semanticmemory processes integrity compared with language tests ofcontext building in terms of observed effect sizes as testedin the same subject group and using the same methodology.It will be important to use approaches that include mul-tiple methodologies as each one—behavioral, event relatedpotential (ERP) and functional magnetic resonance imaging(fMRI)—offers unique and nonredundant pieces of evidenceon how language processes are implemented in the livingbrain and what it means for the theory of schizophrenia.

The article by S. M. Arcuri and colleagues fits very nicelywithin the proposal of examining how complex linguistic

messages are processed by brain systems.The authors explorelanguage processing in the healthy comparison subjects andpatientswith andwithout a formal thought disorder. Drawingon the tradition of ERP studies of language processingin schizophrenia, the authors examine the processing ofsentenceswith congruent and incongruent endings reasoningthat the contrast between the sentences that require integra-tion of context relative to those requiring the suppression ofinappropriate semantic material will highlight a role of brainregions involved in the inhibition of automatically primedstimuli. The results suggest that the left middle frontal cortexis activated more in the incongruent relative to congruentsentences in the healthy comparison group. When the sameincongruent/congruent contrast is analyzed in patients withthe formal thought disorder, reduced activation in the leftinferior/middle frontal gyri and in the anterior cingulate isreported. Thus, these results seem to support the idea thatindeed the brain regions involved in context maintenanceand manipulation and in the inhibition of inappropriatesemantic entries are involved in the disordered language inschizophrenia.

Finally, D. Ketteler and colleagues present a new toolto assess the nature of compromised language function inschizophrenia: higher order linguistic function test (HOLF).They also adopt the premise that more complex languageforms will create special difficulties for schizophrenia suf-ferers. However, in contrast to M. A. Boudewyn and S.M. Acruri proposals, the authors focus on ambiguity inlanguage as brought about by single words such as antonyms,synonyms, homonymy, and as well as on interpreting popularadages. While easier operations such as interpreting adagesand antonyms were not impaired in the patient group,tasks associated with more complex forms distinguishedbetween the two groups and were correlated with symptomratings as measured by PANSS scores. These results suggestthat complexity of linguistic material does not have to berelated to discourse or sentence structure but can be alsorelated to conceptual ambiguity in order to tap into linguisticdifficulties in schizophrenia patients.

The articles by A. P. Pinheiro, H. Fatouros-Bergman, andC. G. Wible are conceived within the framework of concep-tualizing schizophrenia as a disorder of social cognition. H.Fatouros-Bergman and colleagues describe a negative facialaffectivity bias in patients with schizophrenia that seemsto persist across several temporal measuring points. A. P.Pinheiro and colleagues draw attention to the fact that in spiteof evidence of affect processing difficulties in schizophrenia,the conscious ratings of emotional valence seem to be intact.The review paper by C. G. Wible advances an argumenton the centrality of social communication abnormality inschizophrenia. Rather than focusing on language as a dis-embodied communication devise, this approach situates lan-guage within the context of body-based gesture systemwhosecommunicative capabilities are richer than those possibleto be achieved within language as a semantic and syntacticsystem only. As C. G. Wible points out, a live conversationinvolves not only parsing out words and paragraphmeanings,but also relies on correct interpretation of facial expressions,and tone of voice, social salience, agency (who is doing

Schizophrenia Research and Treatment 3

the talking), and intention. An ability to anticipate anotherperson’s actions and represent another point of view, oftenreferred to as a theory of mind, is also essential for successfulcommunication. Finally, two major features of schizophreniapathology, hallucinations and delusions, involve, in additionto abnormalities in cognitive control [20, 21] and in percep-tual and attentional processes [22], also abnormalities in thesense of agency [23]. As the article argues, the brain regionthat supports most of these functions is the temporal parietaloccipital junction (TPJ) with projections to inferior frontalregions, hippocampus, and insular regions. Like the proposalput forth byM.A. Boudewyn et al., the hypothesis formulatedby C. G. Wible has a promise of generating interestingexperimental paradigms to test the role of TPJ in mediatingdifferent aspects of social communication abnormality inschizophrenia. Given the differences in the emphasis betweenthe two proposals, the importance of temporoparietal regionsarticulated by Wible, and the importance of prefrontalregions articulated in the theory of domain general controlimpairment, these two conceptualizations may be viewed astwo competing views of schizophrenia dysfunction.However,they can be also viewed as two complementary views thattogether describe the phenomenology of schizophrenia betterthan each of these theories alone. The theory of abnormalcontrol mechanisms provides a compelling account of how acomplex message system with its symbolic and multilayeredsemantics and syntax can be affected by impaired capacityto manipulate its different elements. The theory of abnormalsocial communication provides a novel perspective on howa formal language may interface with social communicativedevices (gestures, emotional facial expressions, tone of voice(prosody)) and faculties (theory of mind, sense of agency,and intention). Thus, together with the already existingtheories of abnormal processes within semantic memorylargely borne out of priming studies [15] and a theory ofperceptual dysfunction in schizophrenia relying on evidencederived from studying sensory auditory and visual processes[11], the theoretical perspectives espoused in the two reviewpapers add to the richness of theoretical conceptualizationsof cognitive dysfunction in schizophrenia.

Overall, this selection of papers is a good representa-tion of the richness of approaches applied to the study ofcommunication dysfunction in schizophrenia.They illustratethe centrality of language dysfunction in schizophrenia aswell as the importance of abnormalities in nonlanguage-based semiotic systems as contributors to the schizophreniasufferers’ inability to effectively engage with the world in anact of communication.

Margaret A. NiznikiewiczMarek Kubicki

Christoph MulertRuth Condray

References

[1] E. Kraepelin, in Dementia Praecox, B. B. E. Barclay, Ed.,Churchill Livingston, New York, NY, USA, 1919–1971.

[2] E. Bleuler, Dementia Preacox and the Group of Schizophrenias,International Universities Press, NewYork, NY, USA, 1911–1950.

[3] T. H. McGlashan, “Eugene Bleuler: centennial anniversary ofhis 1911 publication of Dementia Praecox or the group ofschizophrenias,” Schizophrenia Bulletin, vol. 37, no. 6, pp. 1101–1103, 2011.

[4] J. vanOs and S. Kapur, “Schizophrenia,”TheLancet, vol. 374, no.9690, pp. 635–645, 2009.

[5] M. E. Shenton, C. C. Dickey, M. Frumin, and R. W. McCarley,“A review of MRI findings in schizophrenia,” SchizophreniaResearch, vol. 49, no. 1-2, pp. 1–52, 2001.

[6] R.W.McCarley, M. A. Niznikiewicz, D. F. Salisbury et al., “Cog-nitive dysfunction in schizophrenia: unifying basic research andclinical aspects,” European Archives of Psychiatry and ClinicalNeuroscience, vol. 249, no. 4, supplement, pp. IV69–IV82, 1999.

[7] A. Marsman, M. P. van den Heuvel, D. W. Klomp, R. S.Kahn, P. R. Luijten, and H. E. Hulshoff Pol, “Glutamate inschizophrenia: a focused review and meta-analysis of 1H-MRSstudies,” Schizophrenia Bulletin, vol. 39, no. 1, pp. 120–129, 2013.

[8] G. Gilmour, S. Dix, L. Fellini et al., “NMDA receptors, cognitionand schizophrenia—testing the validity of the NMDA receptorhypofunction hypothesis,” Neuropharmacology, vol. 62, no. 3,pp. 1401–1412, 2012.

[9] D. M. Barch and A. Ceaser, “Cognition in schizophrenia: corepsychological and neural mechanisms,” Trends in CognitiveSciences, vol. 16, no. 1, pp. 27–34, 2012.

[10] D. J. Lodge and A. A. Grace, “Hippocampal dysregulationof dopamine system function and the pathophysiology ofschizophrenia,” Trends in Cognitive Sciences, vol. 32, no. 9, pp.507–513, 2011.

[11] D. C. Javitt, “When doors of perception close: bottom-up mod-els of disrupted cognition in schizophrenia,” Annual Review ofClinical Psychology, vol. 5, pp. 249–275, 2009.

[12] D. I. Leitman, P. Sehatpour, B. A. Higgins, J. J. Foxe, G. Silipo,and D. C. Javitt, “Sensory deficits and distributed hierarchicaldysfunction in schizophrenia,” American Journal of Psychiatry,vol. 167, no. 7, pp. 818–827, 2010.

[13] K. Wang, E. F. Cheung, Q. Y. Gong, and R. C. Chan, “Semanticprocessing disturbance in patients with schizophrenia: a meta-analysis of the N400 component,” PLoS ONE, vol. 6, no. 10,Article ID e25435, 2011.

[14] D. L. Levy, M. J. Coleman, H. Sung et al., “The geneticbasis of thought disorder and language and communicationdisturbances in schizophrenia,” Journal of Neurolinguistics, vol.23, no. 3, pp. 176–192, 2010.

[15] M. Spitzer, “A cognitive neuroscience view of schizophrenicthought disorder,” Schizophrenia Bulletin, vol. 23, no. 1, pp. 29–50, 1997.

[16] Y. S. Chung, J. R. Mathews, and D. M. Barch, “The effect ofcontext processing on different aspects of social cognition inschizophrenia,” Schizophrenia Bulletin, vol. 37, no. 5, pp. 1048–1056, 2011.

[17] J. R. Mathews and D. M. Barch, “Emotion Responsivity, SocialCognition, and Functional Outcome in Schizophrenia,” Journalof Abnormal Psychology, vol. 119, no. 1, pp. 50–59, 2010.

[18] M. Hoe, E. Nakagami, M. F. Green, and J. S. Brekke, “Thecausal relationships between neurocognition, social cognitionand functional outcome over time in schizophrenia: a latentdifference score approach,” Psychological Medicine, vol. 5, pp. 1–13, 2012.


[19] K. H. Nuechterlein, K. L. Subotnik, M. F. Green et al.,“Neurocognitive predictors of work outcome in recent-onsetschizophrenia,” Schizophrenia Bulletin, vol. 37, supplement 2, pp.S33–S40, 2011.

[20] C. D. Frith, K. J. Friston, P. F. Liddle, and R. S. J. Frackowiak,“PET imaging and cognition in schizophrenia,” Journal of theRoyal Society of Medicine, vol. 85, no. 4, pp. 222–224, 1992.

[21] P. K. McGuire, D. A. Silbersweig, I. Wright, R. M. Murray, R.S. J. Frackowiak, and C. D. Frith, “The neural correlates ofinner speech and auditory verbal imagery in schizophrenia:relationship to auditory verbal hallucinations,” British Journalof Psychiatry, vol. 169, no. 2, pp. 148–159, 1996.

[22] K. Hugdahl, E. M. Loberg, and M. Nygard, “Left temporallobe structural and functional abnormality underlying auditoryhallucinations in schizophrenia,” Frontiers in Neuroscience, vol.3, no. 1, pp. 34–45, 2009.

[23] M. Startup and S. Startup, “On two kinds of delusion ofreference,” Psychiatry Research, vol. 137, no. 1-2, pp. 87–92, 2005.

Hindawi Publishing CorporationSchizophrenia Research and TreatmentVolume 2012, Article ID 825050, 7 pagesdoi:10.1155/2012/825050

Clinical Study

High Order Linguistic Features Such as Ambiguity Processing asRelevant Diagnostic Markers for Schizophrenia

Daniel Ketteler,1 Anastasia Theodoridou,2 Simon Ketteler,3 and Matthias Jager2

1 ZADZ Zurich, Dufourstrasse 161, 8008 Zurich, Switzerland2 Research Group Clinical and Experimental Psychopathology, Department of Social and General Psychiatry,Psychiatric University Hospital Zurich, Lenggstraße 31, 8008 Zurich, Switzerland

3 Department of Neurology, RWTH Aachen University, Pauwelsstraße 30, 52074 Aachen, Germany

Correspondence should be addressed to Daniel Ketteler, [email protected]

Received 14 December 2011; Revised 16 April 2012; Accepted 16 November 2012

Academic Editor: Ruth Condray

Copyright © 2012 Daniel Ketteler et al. This is an open access article distributed under the Creative Commons Attribution License,which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited.

Due to the deficits of schizophrenic patients regarding the understanding of vague meanings (D. Ketteler and S. Ketteler (2010))we develop a special test battery called HOLF (high order linguistic function test), which should be able to detect subtle linguisticperformance deficits in schizophrenic patients. HOLF was presented to 40 schizophrenic patients and controls, focussing onlinguistic features such as ambiguity, synonymy, hypero-/hyponymy, antinomy, and adages. Using the HOLF test battery we foundthat schizophrenic patients showed significant difficulties in discriminating ambiguities, hypero- and hyponymy, or synonymycompared to healthy controls. Antonyms and adages showed less significant results in comparing both groups. The more difficulta linguistic task was, the more confusion was measured in the schizophrenic group while healthy controls did not show significantproblems in processing high order language tasks.

1. Introduction

Regarding the history of diagnostic classification of schizoph-renia, diagnostic tools and catalogues focussed on differentsymptoms to describe a complex syndrome called schizoph-renia. On the one hand, Bleuler [1] had concentrated onthe phenomenon of loosening of association to classify andexplain schizophrenian symptoms. According to Bleuler,language-based “loosening of association” is pathognomonicfor the so-called “schizophrenic symptoms complex.” On theother hand, Schneider [2] drew attention to the significanceof “core” or “first rank” symptoms first outlined by Kraepelin(specific types of hallucination and thought disorder [3]). Toovercome the at least obscure relationship between thoughtand association disorder of Bleuler’s approach, Andreasen[4, 5] shifted the focus of investigation from “thought” tothe more objectively measurable “language behaviour.” Lan-guage impairment indeed seems to be one of the “core” phe-nomenological characteristics of patients with schizophrenia[6, 7]. It seems to be clear that there are deficits in the neuralorganisation of language in schizophrenic patients [6, 8].

There is only a small number of studies focussing on highorder linguistic features and particularly on the phenomenonof ambiguity. Salisbury et al. [9] described a model of initialhyperpriming and subsequent decay of information by usingERP data investigating patients with schizophrenia. Usingevent related brain potentials and an ambiguity processingparadigm, Salisbury [10] found that schizophrenia patientsshowed the largest N400 effect to subordinate associates, withless activity to dominant meaning associates and unrelatedwords. These findings suggest a neural correlate for thedifficulties in suppressing correct word alternatives.

Several other aspects of language comprehension andproduction have been found to be abnormal in patients withschizophrenia: comprehension, attention, semantic organi-sation, reference failures, paucity of speech, or fluency [11].Covington et al. [12] discussed that thought disorder mightreflect a disruption of executive function and pragmatics.Although normal with regard to segmental phonology andmorphological organisation [13], there are obvious word-finding difficulties in patients with schizophrenia [4, 14].Disturbed language production often includes deictic terms


with no clear referents and verbs which lead to a vague andambiguous discourse. Difficulties in dealing with nonliteralexpressions were found by Corcoran and Frith [15] andLangdon et al. [16]. Particularly vague and ambiguous termsseem to irritate patients with schizophrenia. Some stud-ies investigating comprehension deficits attributed reducedcomprehension to deficient working memory [17], and somefound deficits in semantic processing [18].

Beside behavioural difficulties in solving high orderlanguage tasks there are neuroanatomic and neurofunctionalchanges, especially regarding language pathways, in patientswith schizophrenia [19–25].

Regarding the aetiology of schizophrenia from a neuroli-nguistic point of view, there might be hemispheric interac-tion difficulties particularly in processing high order linguis-tic features such as semantic ambiguities. Besides corticalabnormalities, the subcortical role of language processingwas underestimated for a long time.

Deficits of schizophrenic patients in resolving vaguemeaning probably enable test batteries for the early detectionof psychosis and might lead to a better understanding in theaetiology of schizophrenia in general. Ceccherini-Nelli andCrow [26] used a psychometric test (CLANG) to evaluatelanguage disturbances and found that language symptomslike semantic/phonemic paraphasias or poverty of speechwere superior to nuclear symptoms in discriminating ICD-10 schizophrenia from other psychoses. CLANG was origi-nally developed by Chen et al. [27]. This task highly dependson the experience of the examiner and is not appropriateto detect subtle language symptoms. We argue that theCLANG test focuses on symptoms that are more commonlyseen during episodes of acute illness. We tried to developa more elaborated test battery which is able to detect moresubtle language deficits in well-medicated patients with lowsymptom load. The so-called HOLF (high order linguisticfunction test) was presented to 40 schizophrenic patientsand controls focussing on linguistic features. HOLF is apilot test battery created by our group containing differenthigh order language features such as antonyms, homonyms,synonyms versus hyponyms, and adages. This experiment isthe first attempt to introduce our test design to the scientificcommunity. It has not been published before; however, therewas one study by Ketteler et al. [24] regarding the homonympart using fMRI with healthy individuals.

2. Method and Subjects

40 schizophrenic patients (27 male, 13 female) with a meanage of 31,54 (sd = 10,834) and 40 controls (27 male, 13female) with a mean age of 32,48 (sd = 9,081) participated inthe study.All participants were monolingual German nativespeakers and had no history of neurological disorder or ahistory of head trauma. Schizophrenia was diagnosed by anexperienced clinician using the ICD-10 criteria. To determinethe symptom load of the schizophrenic patients we usedPANSS [26] as a well-established diagnostic instrument.For the language symptoms CLANG [28] was rated by anexperienced examiner. Furthermore, age, gender, and CGI(Clinical Global Impression Score [27]) were registered in all

participants and number of previous hospitalizations in theclinical subsample.

3. Experimental Design and Procedure

A linguistic task (HOLF) was presented to 40 patients withschizophrenia and 40 controls. The first task was a warm-up task consisting of 20 antonym relation pairs, while halfof them were distractors. The second task represented other,mixed high order linguistic features, including synonymy,homonymy, and hyperonomy versus hyponomy. Each itemgroup included 20 items while 10 of them were correct, and10 of them were distractors. Individuals were instructed bythe examiner: “Please mark if the first words correlate withthe last word in the line” (for details see HOLF test in Germanlanguage attached in supplementary material available onlineat doi:10.1155/2012/825050).

Example (regarding the ambiguity task):

River Money Bank [X].

Distractors were arranged by using one or two distractors onposition one or two:

River Wind Bank [ ].

or

Door Wind Bank [ ].

Example (regarding the hyperonymy/hyponomy task).

Water Juice Drink [X].

Additionally, three classical adages were tested by giving threeanswer alternatives for each wording. HOLF has not beenpublished before, and this is the first attempt to test the effectof high order language tasks with schizophrenic patientsusing a very simple design. HOLF uses German languageitems and has not been translated to the English languageuntil now. HOLF is in an early stage of development, andfurther data concerning validity and reliability has to becollected in future studies.

4. Statistic Analysis

Anonymised data was analysed with SPSS 15.0 for Windows(Statistical Package for the Social Sciences, Lead Technolo-gies, Inc., Chicago, 2006). HOLF was analysed descriptivelywith regard to the number of items that were processedadequately. The sum score of correct answers were calculatedon total HOLF scale as well as on the five subscales antonyms,homonyms, hyperonyms, synonyms, and adages. Since thedistribution of raw scores was highly skewed with a largeproportion of subjects delivering high sum scores, nonpara-metric tests (Mann-Whitney U test, Wilcoxon rank ordertest) were performed in order to compare the two subsamples(clinical population and healthy controls).

5. Results

The sample comprised 40 patients diagnosed with schizo-phrenia according to ICD-10, F20, and 40 healthy controls.


Table 1: Statistics of HOLF, CLANG, and PANSS by subsample (patients N = 40; controls N = 40).

Scale Sample Mean SD Median Mean rank Rank sum Mann-W U Z P Alpha

HOLFPatients 68.5 13.6 72.5 21.5 860

39.5 −7.4 <0.001 0.963Controls 82.3 0.8 82.5 59.5 2380

CLANGPatients 0.4 0.4 0.2 57.5 2300

120 −7.3 <0.001 0.893Controls 0.0 0.0 0 23.5 940

PANSSPatients 1.5 0.4 1.4

Controls 1.0 0.0 1

HOLF subscales

AntonymsPatients 17.9 3.7 19 30.4 1215

395 −5.0 <0.001 0.924Controls 19.9 0.2 20 50.6 2025

HomonymsPatients 13.9 4.1 14 22.3 891

71 −7.1 <0.001 0.878Controls 19.4 0.8 20 58.7 2349

SynonymsPatients 17.0 3.1 18 23.5 941

121 −7.1 <0.001 0.845Controls 19.9 0.3 20 57.5 2299

HyperonymsPatients 17.0 4.5 19 28.1 1124

304 −5.6 <0.001 0.934Controls 19.9 0.2 20 52.9 2116

AdagesPatients 2.8 0.7 3 38.5 1540

720 −2.0 0.042 0.898Controls 3.0 0.0 3 42.5 1700

DistractorsPatients 10.2 5.6 13 29.8 1192

372 −4.4 <0.001 0.962Controls 14.7 0.5 15 51.2 2048

There were no significant differences in age and gender.The clinical subsample had a history of 4.1 previous hospi-talisations on average and a severity-of-illness-score of 4.4(moderately ill) according to Clinical Global ImpressionScale (CGI) at the time of the investigation.

Statistical measurements of the HOLF scale and subscalesas well as CLANG and PANSS are shown in Table 1. Theinternal consistency measured by Cronbach’s alpha of HOLFtotal scale and subscales was high.

Healthy controls answered the items on HOLF andCLANG mostly without mistakes, whereas patients hadsignificant problems with the task. The differences betweenthe two groups on almost all subscales were highly significantwith the exception if the adages scale.

HOLF total scale showed a significant correlation withCLANG even if controlled for psychopathology as measuredby PANSS (r = −0.396, P < 0.001, df = 76). HOLF subscalescorrelated significantly with CLANG with the exception ofadages scale. Correlation coefficients are shown in Table 2.

6. Discussion

The aim of the current study was to explore in how far schi-zophrenic symptoms are correlated with difficulties in highorder linguistic processing. By testing high order lang-uage performance using the HOLF battery we found thatschizophrenic patients showed significant difficulties in dis-criminating ambiguities, hypero- and hyponymy, and syn-onymy compared to healthy controls. Antonyms and adagesshowed less significant results comparing both groups. Thegroup differences observed for antonyms and adages wereweaker than the effects obtained for the other types of

linguistic ambiguities. If this pattern is determined to bereliable, these latter linguistic features may have less pathog-nomonic value for schizophrenia. The more difficult a lingui-stic operation was the more confusion was measured withinthe schizophrenic group while healthy controls did not showsignificant problems solving high order language tasks. Agecorrelated slightly with general performance in HOLF, butthere were more significant effects regarding the severity ofillness and times of hospitalization. One might assume thatage and times of hospitalization correlate with the chronicityof illness.

HOLF highly correlated with the standard instrumentfor scoring symptom load in schizophrenia called PANSS.We conclude that disturbed (high order) language functionmight be pathognomonic for schizophrenia. As already men-tioned by Bleuler, “loosening of association” might be thecore symptom regarding psychotic syndromes. Focussing onthe correct word alternative while discriminating ambiguousmeanings seems to be deeply disturbed in our patient group.According to our data, homonymy detection was highlyimpaired in patients with schizophrenia.

In further studies, patients with schizophrenia showedproblems in selecting context-related ambiguous meanings.They have been shown to be impaired in using the context ofsentences to determine an appropriate meaning of a homo-graphic word [29, 30]. Cohen and Servan-Schreiber [31]showed that schizophrenic patients have difficulties in pro-cessing multimeaning words, but only when disambiguatingcontext precedes the target homograph. Previous researchhas suggested that a failure in processing contextual infor-mation may account for the heterogenous clinical manifesta-tions and cognitive impairments observed in schizophrenia.


Ta

ble

2:C

orre

lati

onco

effici

ents

(pat

ien

tsN=

40;c

ontr

olsN=

40).

Cor

rela

tion

s(n=

40)

Spea

rman

-Rh

oSe

veri

tyof

illn

ess

HO

LFal

lH

OLF

dist

ract

ors

HO

LFan

tony

ms

HO

LFh

omon

yms

HO

LFsy

non

yms

HO

LFhy

per

onym

sH

OLF

adag

esC

LAN

GPA

NSS

Age

Nu

mbe

rof

hos

pita

lisat

ion

s

Cor

rela

tion

coeffi

cien

t1.

000

−.65

3∗∗

−.54

0∗∗

−.37

1∗−.

539∗

∗−.

431∗

∗−.

438∗

∗−.

266

.447

∗∗.4

47∗∗

.475

∗∗.5

08∗∗

Seve

rity

ofill

nes

sSi

g.(2

side

s)—

.000

.000

.018

.000

.006

.005

.097

.004

.004

.002

.001

N40

4040

4040

4040

4040

3940

40

Cor

rela

tion

coeffi

cien

t−.

653∗

∗1.

000

.826

∗∗.4

83∗∗

.814

∗∗.7

92∗∗

.670

∗∗.4

01∗

−.69

6∗∗−.

743∗

∗−.

338∗

−.12

8H

OLF

all

Sig.

(2si

des)

.000

—.0

00.0

02.0

00.0

00.0

00.0

10.0

00.0

00.0

33.4

31N

4040

4040

4040

4040

4039

4040

Cor

rela

tion

coeffi

cien

t−.

540∗

∗.8

26∗∗

1.00

0.3

78∗

.488

∗∗.7

62∗∗

.703

∗∗.3

00−.

598∗

∗−.

704∗

∗−.

241

−.10

1H

OLF

dist

ract

ors

Sig.

(2si

des)

.000

.000

—.0

16.0

01.0

00.0

00.0

60.0

00.0

00.1

34.5

34

N40

4040

4040

4040

4040

3940

40

Cor

rela

tion

coeffi

cien

t−.

371∗

.483

∗∗.3

78∗

1.00

0.3

48∗

.397

∗.3

70∗

.421

∗∗−.

474∗

∗−.

373∗

−.26

8−.

439∗

∗

HO

LFan

tony

ms

Sig.

(2si

des)

.018

.002

.016

—.0

28.0

11.0

19.0

07.0

02.0

19.0

94.0

05

N40

4040

4040

4040

4040

3940

40

Cor

rela

tion

coeffi

cien

t−.

539∗

∗.8

14∗∗

.488

∗∗.3

48∗

1.00

0.5

17∗∗

.255

.474

∗∗−.

446∗

∗−.

446∗

∗−.

343∗

−.08

9H

OLF

hom

onym

sSi

g.(2

side

s).0

00.0

00.0

01.0

28—

.001

.112

.002

.004

.004

.030

.586

N40

4040

4040

4040

4040

3940

40

Cor

rela

tion

coeffi

cien

t−.

431∗

∗.7

92∗∗

.762

∗∗.3

97∗

.517

∗∗1.

000

.592

∗∗.4

01∗

−.61

4∗∗−.

766∗

∗−.

131

−.09

6H

OLF

syn

onym

sSi

g.(2

side

s).0

06.0

00.0

00.0

11.0

01—

.000

.010

.000

.000

.419

.556

N40

4040

4040

4040

4040

3940

40

Cor

rela

tion

coeffi

cien

t−.

438∗

∗.6

70∗∗

.703

∗∗.3

70∗

.255

.592

∗∗1.

000

.198

−.67

4∗∗−.

603∗

∗−.

166

−.09

0H

OLF

hyp

eron

yms

Sig.

(2si

des)

.005

.000

.000

.019

.112

.000

—.2

21.0

00.0

00.3

05.5

81

N40

4040

4040

4040

4040

3940

40

Cor

rela

tion

coeffi

cien

t−.

266

.401

∗.3

00.4

21∗∗

.474

∗∗.4

01∗

.198

1.00

0−.

310

−.18

0−.

180

−.29

7H

OLF

adag

esSi

g.(2

side

s).0

97.0

10.0

60.0

07.0

02.0

10.2

21—

.051

.273

.266

.063

N40

4040

4040

4040

4040

3940

40

Cor

rela

tion

coeffi

cien

t.4

47∗∗

−.69

6∗∗

−.59

8∗∗

−.47

4∗∗

−.44

6∗∗

−.61

4∗∗

−.67

4∗∗

−.31

01.

000

.830

∗∗29

5.1

86C

LAN

GSi

g.(2

side

s)00

4.0

00.0

00.0

02.0

04.0

00.0

00.0

51—

.000

065

.251

N40

4040

4040

4040

4040

3940

40


Ta

ble

2:C

onti

nu

ed. Cor

rela

tion

s(n=

40)

Spea

rman

-Rh

oSe

veri

tyof

illn

ess

HO

LFal

lH

OLF

dist

ract

ors

HO

LFan

tony

ms

HO

LFh

omon

yms

HO

LFsy

non

yms

HO

LFhy

per

onym

sH

OLF

adag

esC

LAN

GPA

NSS

Age

Nu

mbe

rof

hos

pita

lisat

ion

s

Cor

rela

tion

coeffi

cien

t.4

47∗∗

−.74

3∗∗

−.70

4∗∗

−.37

3∗−.

446∗

∗−.

766∗

∗−.

603∗

∗−.

180

.830

∗∗1.

000

.077

−.01

1PA

NSS

Sig.

(2si

des)

.004

.000

.000

.019

.004

.000

.000

.273

.000

—.6

41.9

48N

3939

3939

3939

3939

3939

3939

Cor

rela

tion

coeffi

cien

t.4

75∗∗

−.33

8∗−.

241

−.26

8−.

343∗

−.13

1−.

166

−.18

0.2

95.0

771.

000

.395

∗

Age

Sig.

(2si

des)

.002

.033

.134

.094

.030

.419

.305

.266

.065

.641

—.0

12N

4040

4040

4040

4040

4039

4040

∗∗T

he

corr

elat

ion

issi

gnifi

can

ton

a0.

01n

ivea

u(2

side

s).

∗ Th

eco

rrel

atio

nis

sign

ifica

nt

ona

0.05

niv

eau

(2si

des)

.


Our data suggests difficulties of schizophrenic patients inprocessing multimeaning words such as ambiguities andsynonyms but also suggests difficulties in solving semantictaxonomies such as hypero- and hyponymies. Antonymswere not presented in a randomised order but in a blockdesign with several distractors. The antonym task withinour experiment might have been too simple to solve, sothese items did not show any differences between patientand control group. Therefore, we used the antonym task as awarm-up and to motivate patients to carry on with the moredifficult tasks.

CLANG [32] was a first and important step in reactivat-ing the Bleulerian idea of underlying language disturbancein schizophrenia but seems to be much too unspecific fordetecting high order language dysfunction. CLANG has tobe rated by the examiner who is always subjectively biased bythe personal impressions and diagnostic data concerning hispatient. HOLF now offers a more objective method whichis highly significantly correlated with CLANG and PANSS.PANSS indeed is a well known and established diagnostictool regarding schizophrenic symptoms.

Delusions can be considered as deviations in the capacityto attach significance to the phonological representationsthat are the primary building blocks of words. Disruptionssuch as clause boundaries or sentence endings occur gen-erally, particularly at points of attentional focus fluctuation[11, 33]. As a consequence, coherence of speech breaks down,which is fundamental for the development of formal thoughtdisorder. According to our data one might assume that themore vague or ambiguous a semantic correlation is pre-sented, the more effort has to be made by the neural systemin processing these items.

According to connectionist network models [34], eachcomponent of an utterance activates associated semanticunits which remain activated for a finite period of time [35].In schizophrenia, the speed of decay which is following thespread of activation seems prolonged. If such an inhibitoryprocess becomes impaired, patients have a greater potentialfor intrusion into later thought and speech. Althoughcontext information is important for almost all cognitivetasks, the domain of language is ideally suited for thestudy of contextual processing. Homonyms and synonymshighly depend on correct contextual priming, so deficits inprocessing these features lead to massive linguistic irritationin the schizophrenic brain as seen in our experiment.

Titone et al. [36] used a priming task by presentingsentences containing homonyms. Eighteen schizophrenicpatients were asked on lexical decisions about visual targetsrelated to the homonyms’ subordinate, respectively, domi-nant meaning. When sentences biased subordinate meaning,patients showed priming of dominant targets. The resultsalso suggest that contextual strength is an important deter-minant of when schizophrenia patients fail to inhibit contex-tually irrelevant meanings. Wentura et al. [37] investigatedpriming by using a masked repetition task and comparedformal thought-disordered patients, nonthought disorderedpatients, and healthy controls. For thought-disorderedpatients they found “hyperpriming,” whereas the othergroups showed regular priming. This result yields evidence

for a lack of inhibitory function in thought- (and thereforelanguage-) disordered patients. Furthermore, Sitnikova et al.[38] found that the N400 ERP component that is known tobe sensitive to contextual effects was attenuated in patientswith schizophrenia. This is potentially due to inadequatecontextual suppression mechanisms and/or due to increasedlevels of word-meaning activation.

“Hyperpriming” was detected in our data too while therewere many false-positive errors in the schizophrenia group.As already mentioned above, Salisbury [9, 10] described amodel of initial hyper-priming by using ERP data investi-gating patients with schizophrenia. These findings suggesta neural correlate for the difficulties in suppressing correctword alternatives.

In summary, using the HOLF test battery we foundthat schizophrenic patients showed significant difficulties indiscriminating ambiguities, hypero- and hyponymy, or syn-onymy compared to healthy controls. Antonyms and adagesshowed rather significant results in comparing both groups,so these features seem to have a less specific pathognomonicand diagnostic value regarding schizophrenia or were toosimple to solve. The more difficult a linguistic operationwas, the more confusion was measured in the schizophrenicgroup while healthy controls did not show significant prob-lems in solving higher order language tasks.

One limitation of our study might be a very low errorrate in healthy individuals. Therefore, very few errors causea highly significant difference in performance between bothgroups. On the other hand, HOLF was able to detect schi-zophrenic symptoms on a very subtle level which might bean indicator for underlying language problems, although thepatient group presented a low level of symptoms, and allpatients have been well medicated. By revitalising the Bleu-lerian focus on thought, respectively, language disorder thealmost chaotic mechanisms of psychotic experience mightbecome more understandable. Our findings might inspirethe development of early detection test batteries whichinclude high order language functions in order to detectsubtle and underlying language deficits in schizophrenicpatients. HOLF is in a very early stage of development, andmore data is needed especially regarding psychometric prop-erties such as test-retest reliability and validity.

References

[1] E. Bleuler, Dementia Praecox, or the Group of Schizophrenias(translated by J. Zincs from the Original German Edition, Deme-ntia Praecox oder die Gruppe der Schizophrenien, 1911), Inter-national University Press, New York, NY, USA, 1950.

[2] K. Schneider, Klinische Psychopathologie, Thieme, New York,NY, USA, 1950.

[3] E. Kraepelin, Dementia Praecox, Churchill Livingston, NewYork, NY, USA, 1919.

[4] N. C. Andreasen, “Thought, language, and communicationdisorders. I. Clinical assessment, definition of terms, and eval-uation of their reliability,” Archives of General Psychiatry, vol.36, no. 12, pp. 1315–1321, 1979.

[5] N. C. Andreasen, “Thought, language, and communicationdisorders. II. Diagnostic significance,” Archives of General Psy-chiatry, vol. 36, no. 12, pp. 1325–1330, 1979.


[6] L. E. DeLisi, “Speech disorder in schizophrenia: review ofthe literature and exploration of its relation to the uniquelyhuman capacity for language,” Schizophrenia Bulletin, vol. 27,no. 3, pp. 481–496, 2001.

[7] P. McKenna and T. Oh, Schizophrenic Speech, Cambridge Uni-versity Press, Cambridge, UK, 2005.

[8] M. Stephane, S. Barton, and N. N. Boutros, “Auditory verbalhallucinations and dysfunction of the neural substrates ofspeech,” Schizophrenia Research, vol. 50, no. 1-2, pp. 61–78,2001.

[9] D. F. Salisbury, M. E. Shenton, P. G. Nestor, and R. W. McCar-ley, “Semantic bias, homograph comprehension, and event-related potentials in schizophrenia,” Clinical Neurophysiology,vol. 113, no. 3, pp. 383–395, 2002.

[10] D. F. Salisbury, “Abnormal N400 responses but intact differ-ential hemispheric processing of ambiguity in schizophrenia,”Journal of Neurolinguistics, vol. 23, no. 3, pp. 240–253, 2010.

[11] A. Marini, I. Spoletini, I. A. Rubino et al., “The language ofschizophrenia: an analysis of micro and macrolinguistic abil-ities and their neuropsychological correlates,” SchizophreniaResearch, vol. 105, no. 1–3, pp. 144–155, 2008.

[12] M. A. Covington, C. He, C. Brown et al., “Schizophrenia andthe structure of language: the linguist’s view,” SchizophreniaResearch, vol. 77, no. 1, pp. 85–98, 2005.

[13] R. Chaika, Understanding Psychotic Speech: Beyond Freud andChomsky, Charles C. Thomas, Springfield, Mass, USA, 1990.

[14] P. J. McKenna, Schizophrenia and Related Syndromes, OxfordUniversity Press, Oxford, UK, 1994.

[15] R. Corcoran and C. D. Frith, “Autobiographical memory andtheory of mind: evidence of a relationship in schizophrenia,”Psychological Medicine, vol. 33, no. 5, pp. 897–905, 2003.

[16] R. Langdon, M. Coltheart, P. B. Ward, and S. V. Catts, “Dis-turbed communication in schizophrenia: the role of poorpragmatics and poor mind-reading,” Psychological Medicine,vol. 32, no. 7, pp. 1273–1284, 2002.

[17] W. M. Grove and N. C. Andreasen, “Language and thinking inpsychosis. Is there an input abnormality?” Archives of GeneralPsychiatry, vol. 42, no. 1, pp. 26–32, 1985.

[18] D. A. Kareken, P. J. Moberg, and R. C. Gur, “Proactiveinhibition and semantic organization: relationship with verbalmemory in patients with schizophrenia,” Journal of the Inter-national Neuropsychological Society, vol. 2, no. 6, pp. 486–493,1996.

[19] J. Coney and K. D. Evans, “Hemispheric asymmetries in theresolution of lexical ambiguity,” Neuropsychologia, vol. 38, no.3, pp. 272–282, 2000.

[20] M. Faust and C. Chiarello, “Sentence context and lexical ambi-guity resolution by the two hemispheres,” Neuropsychologia,vol. 36, no. 9, pp. 827–835, 1998.

[21] T. T. J. Kircher, M. J. Brammer, R. Williams, and P. K. McGuire,“Lexical retrieval during fluent speech production: an fMRIstudy,” NeuroReport, vol. 11, no. 18, pp. 4093–4096, 2000.

[22] T. T. J. Kircher, P. F. Liddle, M. J. Brammer, S. C. R. Williams,R. M. Murray, and P. K. McGuire, “Reversed lateralization oftemporal activation during speech production in thought dis-ordered patients with schizophrenia,” Psychological Medicine,vol. 32, no. 3, pp. 439–449, 2002.

[23] T. J. Crow, “Schizophrenia as failure of hemispheric domi-nance for language,” Trends in Neurosciences, vol. 20, no. 8, pp.339–343, 1997.

[24] D. Ketteler, F. Kastrau, R. Vohn, and W. Huber, “The subcor-tical role of language processing. High level linguistic featuressuch as ambiguity-resolution and the human brain; an fMRIstudy,” NeuroImage, vol. 39, no. 4, pp. 2002–2009, 2008.

[25] L. J. Seidman, W. S. Kremen, D. Koren, S. V. Faraone, J. M.Goldstein, and M. T. Tsuang, “A comparative profile analysis ofneuropsychological functioning in patients with schizophre-nia and bipolar psychoses,” Schizophrenia Research, vol. 53, no.1-2, pp. 31–44, 2002.

[26] A. Ceccherini-Nelli and T. J. Crow, “Disintegration of the com-ponents of language as the path to a revision of Bleuler’s andSchneider’s concepts of schizophrenia: linguistic disturbancescompared with first-rank symptoms in acute psychosis,”British Journal of Psychiatry, vol. 182, pp. 233–240, 2003.

[27] E. Y. H. Chen, L. C. W. Lam, C. S. Kan et al., “Language disor-ganisation in schizophrenia: validation and assessment with anew clinical rating instrument,” Hong Kong Journal of Psychia-try, vol. 6, pp. 4–13, 1996.

[28] P. Westmoreland Corson, P. Nopoulos, N. C. Andreasen, D.Heckel, and S. Arndt, “Caudate size in first-episode neuro-leptic-naive schizophrenic patients measured using an artifi-cial neural network,” Biological Psychiatry, vol. 46, no. 5, pp.712–720, 1999.

[29] T. B. Benjamin and N. F. Watt, “Psychopathology and semanticinterpretation of ambiguous words,” Journal of AbnormalPsychology, vol. 74, no. 6, pp. 706–714, 1969.

[30] L. J. Chapman, J. P. Chapman, and G. A. Miller, “A theoryof verbal behavior in schizophrenia,” Progress in ExperimentalPersonality Research, vol. 72, pp. 49–77, 1964.

[31] J. D. Cohen and D. Servan-Schreiber, “Context, cortex, anddopamine: a connectionist approach to behavior and biologyin schizophrenia,” Psychological Review, vol. 99, no. 1, pp. 45–77, 1992.

[32] Clinical Global Impressions (CGI) Scale, Psychiatric Measures,American Psychiatric Association, Washington, DC, USA,2000.

[33] B. Maher, “The language of schizophrenia: a review and inter-pretation,” British Journal of Psychiatry, vol. 120, no. 554, pp.3–17, 1972.

[34] G. S. Dell, L. K. Burger, and W. R. Svec, “Language productionand serial order: a functional analysis and a model,” Psycholog-ical Review, vol. 104, no. 1, pp. 123–147, 1997.

[35] D. E. Meyer and R. W. Schvaneveldt, “Facilitation in recogniz-ing pairs of words: evidence of a dependence between retrievaloperations,” Journal of Experimental Psychology, vol. 90, no. 2,pp. 227–234, 1971.

[36] D. Titone, D. L. Levy, and P. S. Holzman, “Contextual insen-sitivity in schizophrenic language processing: evidence fromlexical ambiguity,” Journal of Abnormal Psychology, vol. 109,no. 4, pp. 761–767, 2000.

[37] D. Wentura, S. Moritz, and C. Frings, “Further evidencefor “hyper-priming” in thought-disordered schizophrenicpatients using repeated masked category priming,” Schizophre-nia Research, vol. 102, no. 1–3, pp. 69–75, 2008.

[38] T. Sitnikova, D. F. Salisbury, G. Kuperberg, and P. J. Holcomb,“Electrophysiological insights into language processing inschizophrenia,” Psychophysiology, vol. 39, no. 6, pp. 851–860,2002.


Clinical Study

From Semantics to Feelings: How Do Individuals withSchizophrenia Rate the Emotional Valence of Words?

Ana P. Pinheiro,1, 2 Robert W. McCarley,2 Elizabeth Thompson,2

Oscar F. Goncalves,1 and Margaret Niznikiewicz2

1 Neuropsychophysiology Lab, CIPsi, School of Psychology, University of Minho, 4710-057 Braga, Portugal2 Clinical Neuroscience Division, Laboratory of Neuroscience, Department of Psychiatry, Boston VA Healthcare System,Brockton Division and Harvard Medical School, Boston, MA 02301, USA

Correspondence should be addressed to Ana P. Pinheiro, [email protected]

Received 14 December 2011; Accepted 15 March 2012

Academic Editor: Marek Kubicki

Copyright © 2012 Ana P. Pinheiro et al. This is an open access article distributed under the Creative Commons AttributionLicense, which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properlycited.

Schizophrenia is characterized by both emotional and language abnormalities. However, in spite of reports of preserved evaluationof valence of affective stimuli, such as pictures, it is less clear how individuals with schizophrenia assess verbal materialwith emotional valence, for example, the overall unpleasantness/displeasure relative to pleasantness/attraction of a word. Thisstudy aimed to investigate how schizophrenic individuals rate the emotional valence of adjectives, when compared with agroup of healthy controls. One hundred and eighty-four adjectives differing in valence were presented. These adjectives werepreviously categorized as “neutral,” “positive” (pleasant), or “negative” (unpleasant) by five judges not participating in the currentexperiment. Adjectives from the three categories were matched on word length, frequency, and familiarity. Sixteen individualswith schizophrenia diagnosis and seventeen healthy controls were asked to rate the valence of each word, by using a computerizedversion of the Self-Assessment Manikin (Bradley and Lang, 1994). Results demonstrated similar ratings of emotional valence ofwords, suggesting a similar representation of affective knowledge in schizophrenia, at least in terms of the valence dimension.

1. Introduction

Emotional abnormalities are a hallmark of schizophrenia [1–4] and are often evident in prodromal stages of this disorder[1, 5].

Recent years have seen a rapid increase in interest inemotion processing in schizophrenia. Stimuli with emotionalsalience have particular relevance for the individual and,thus, abnormalities in their processing have important con-sequences for social functioning and functional outcomesfor individuals with schizophrenia [6]. The existing studieshave explored different aspects of emotion processing in thisdisorder (see [7] for a review) including the study of (a) emo-tional perception (e.g., [8]); (b) emotional experience (e.g.,assessment of self-reported affect through the presentation ofemotionally evocative stimuli; assessment of trait differencesin emotion components) (e.g., [9]); (c) emotional expression[10]; (d) effects of emotion on cognitive processes, such as

working memory [11, 12]; (e) evaluation of the affectiveproperties of stimuli varying in valence and arousal. In termsof the conceptual framework, the latter studies represent adimensional approach to emotion. Dimensional theories ofemotion propose that emotions can be characterized along asmall number of underlying and separable dimensions, suchas valence (the overall unpleasantness/displeasure relative topleasantness/attraction of a stimulus) and arousal (the inten-sity of motivational mobilization—appetitive or defensive)[13–15]. This assumption is supported by brain functionalimaging (e.g., [16]) and event-related potential (ERP) studies(e.g., [17]) indicating differential effects of valence andarousal on brain activation and function. In contrast, adiscrete emotions approach holds that emotions may bedistinguished from one another according to a set of features[18].

The existing studies on emotion processing in schizo-phrenia suggest that the components of emotional processing


mentioned above are differentially affected by the disorder(see [7] for a review). For example, previous studies pointedto a dissociation between the subjective experience andthe expression of emotion in schizophrenia [19], based onsomewhat paradoxical findings revealing that these indi-viduals show less emotional expression, even though theyreport momentary emotional experience similar to thatof healthy controls (HC) in response to stimuli such asfilm clips, pictures, or emotional face expressions (e.g.,[20, 21]). Studies using self-report measures of emotionalexperience have additionally demonstrated that individualswith schizophrenia report experiencing feelings in a waythat is consistent with the valence of the presented evoca-tive stimuli, that is, they report negative mood states inresponse to unpleasant stimuli or positive mood states inresponse to pleasant evocative stimuli ([22–27]; see also[28] for a review). Nonetheless, other studies found thatindividuals with schizophrenia diagnosis report experiencingless positive emotion in response to happy emotional faceexpressions, in comparison with HC [25].

In particular, most studies investigating the way schizo-phrenic individuals rate affective properties of stimuli haveprovided evidence for similar evaluation of valence of af-fective stimuli in schizophrenic patients and HC (pictures[4, 20, 23, 29–31], facial expressions [32] and odors [33,34]). However, these results have not been always replicated.For example, some studies indicated that schizophrenicindividuals tend to rate pleasant stimuli as being less pleasant[35, 36], and negative stimuli as being less unpleasant[35, 36] (in both studies, the stimuli used were emotionalpictures, words, and faces). Differences in patients’ samples(e.g., gender [29]), including schizophrenia subtype, clinicalsymptoms (e.g., severity of negative symptoms and level ofanhedonia) and functional outcome measures [36–38]) aswell as differences in stimuli (e.g., level of arousal) mayaccount for the apparent discrepancies between studies.

Findings related to the assessment of arousal indicate ei-ther similarities [23] or differences [21, 36, 39] in theassessment of this dimension. For example, differences wereobserved in arousal ratings of aversive/unpleasant stimuli,with lower ratings indicated by individuals with schizophre-nia relative to HC in response to different types of stimuli,such as pictures selected from the International AffectivePictures System dataset (IAPS [39]), words selected from theAffective Norms for English Words dataset (ANEW [40]),sounds selected from the International Affective DigitizedSounds dataset (IADS [41]), or emotional faces [35]. Whencompared with patients with bipolar disorder and HC, schiz-ophrenia patients reported lower arousal for aversive stimuliwith social content [39]. Also, heightened arousal ratingswere found for pleasant pictures [21] and for neutral stimuli(pictures, words, and faces [36]). Discrepancies in thesefindings might be related to methodological differences, in-cluding sample differences (e.g., differences in anhedonialevel or in neurocognitive measures such as working memory[35]). In spite of differences in ratings of valence and arousalin some of individuals with schizophrenia, the existingstudies point to a representation of affective knowledge inschizophrenia similar to that found in HC, suggesting that

valence and arousal are also two major features of thisknowledge (see also [42]).

Besides affective abnormalities, disturbance of languageprocesses has long been reported in schizophrenia. It includesdeficits in declarative-episodic memory of verbal material[43], abnormal semantic priming effects [44, 45], and abnor-mal context processing [46]. Abnormalities were also foundin the brain network involved in semantic processing [47].Language abnormalities in schizophrenia were proposed torely both on an initial overly activated semantic network andon later inhibition difficulties indicating abnormal contextutilization [48, 49]. These semantic processing deficits donot seem to be dependent on grammatical category of aword, such as nouns, verbs, or adjectives [50]. However, it isless clear how individuals with schizophrenia process verbalmaterial with emotional valence.

Studies with healthy populations have demonstrated adifferential processing of neutral, pleasant and unpleasantverbal information [51–56] as well as an automatic process-ing of emotional word content in the sense that it is notdependent on the availability of attentional resources [57].For example, pleasant adjectives tend to be better remem-bered than unpleasant or neutral adjectives, suggesting apreferential processing of pleasant words [55, 57]. Elec-trophysiologically, the prioritized processing of emotionalverbal material is indexed by enhanced (i.e., more positiveor more negative) ERP amplitude for emotional relative toneutral words [54, 57, 58].

Studies testing affective semantic priming (a variant ofthe semantic-priming paradigm consisting of the presenta-tion of an emotional prime word before a target word withemotional meaning) reported similar affective and semanticpriming in individuals with schizophrenia when comparedwith HC [59, 60]. However, other studies revealed thatthese individuals tend to show a facilitatory priming effectfor neutral word stimuli, but not for positive or negativeword stimuli; in addition, schizophrenic individuals’ reactiontimes tended to be slower for related negative word targetsthan to unrelated negative word targets [45].

Studies on sentence processing with affective semanticcontent showed abnormalities in the interaction betweensemantic networks and emotional processing in schizophre-nia [61], as indexed by increased N400 for negative wordendings relative to both depressed and HC groups. In addi-tion, individuals with schizophrenia did not show memoryenhancement for self-referenced adjectives, contrary to HC,which may be related to poor social outcomes in thisdisorder [62]. Interestingly, phenomenological studies onauditory verbal hallucinations show that these often havenegative/derogatory semantic content [63–65], which maysuggest a relationship between clinical symptoms and pro-cessing of verbal material with negative emotional valence.

In spite of evidence suggesting abnormalities in process-ing verbal affective stimuli, it is yet not clear if abnormalitiesare related to abnormal declarative knowledge about affect. Aprevious study [42] provided evidence for similar knowledgerepresentations of verbal affective stimuli in 11 individualswith schizophrenia and 7 HC, in terms of their valence-based and arousal-based meaning. Differences were found


Table 1: Demographic and clinical characteristics of participants (mean ± SD).

VariableHealthy controls

(n = 17)Individuals with

schizophrenia (n = 16)t (df = 31)∗ P value

Age (years) 43.65 ± 11.32 48.69 ± 8.38 −1.446 .158

Gender 3 females; 14 males 4 females; 12 males NA NA

Education (years) 15.47 ± 1.81 13.00 ± 2.25 3.487 .001

Subject’s SES 2.23 ± 1.01 3.63 1.41 −2.992 .006

Parental SES 2.54 ± 0.88 3.13 ± 1.45 −1.275 .213

Onset age (years) NA 30.23 ± 11.95 NA NA

Duration (years) NA 17.15 ± 11.32 NA NA

Chlorpromazine equivalent (mg) NA 381.13 ± 247.18 NA NA

Medication type NAFirst generation = 3

Second generation = 9NA NA

Positive scale PANSS NA 23.87 ± 9.41 NA NA

Negative scale PANSS NA 22.33 ± 9.56 NA NA

General psychopathology PANSS NA 43.00 ± 14.64 NA NA

Total psychopathology PANSS NA 89.20 ± 30.17 NA NA

SES: socioeconomic status; NA: not applicable; ∗Independent sample t-test was used for group comparisons.

in the weighting of valence and arousal dimensions in atask of similarity assessment of affective word pairs: whileparticipants with schizophrenia weighted the valence andarousal dimensions equally, HC weighted more the arousalthan the valence dimension, suggesting that the relativeimportance of these dimensions may differ in individualswith schizophrenia and HC. However, in this study only16 emotion terms were used (excited, lively, cheerful,pleased, calm, relaxed, idle, still, dulled, bored, unhappy,disappointed, nervous, fearful, alert, and aroused), andthey were assessed on a 7-point Likert scale (1: extremelydissimilar; 7: extremely similar). Also, Burbridge and Barch[35] assessed emotional experience to pleasant, neutral andunpleasant stimuli in different modalities, including 75words selected from the ANEW dataset [66], varying invalence (pleasant, unpleasant, and neutral) and arousal (lowand high). Schizophrenic participants and HC were asked torate their emotional experience to the stimuli, in terms ofhow pleasant-unpleasant and aroused-calm the stimuli madethem feel. However, in this study a composite index was usedthat did not allow to investigate the separate processing ofword stimuli with emotional valence.

In this study, we compared valence ratings of adjectivesin individuals with schizophrenia and in HC. To our knowl-edge, only one previous study [42] has directly assessed howindividuals with schizophrenia assess emotional adjectives,in spite of substantial research on how they assess other typesof affective stimuli such as pictures or film clips [4, 21, 23,31, 67, 68]. However, the study of Kring et al. [42] included asmall number of adjectives that may not be representative ofthe vocabulary that depicts emotional situations in the dailylife.

In our study, we have presented a list of 184 adjectives(previously assessed as “pleasant,” “unpleasant,” or “neutral”by a group of 5 judges) to a group of 16 individuals withschizophrenia diagnosis and to 17 HC. They were asked to

assess the valence of the words on a 1–9 Likert scale [13]. Weposited that individuals with schizophrenia and HC wouldshow similar ratings of valence of adjectives, consistent withreports of preserved assessment of affective properties ofstimuli differing in valence and similar representation ofemotion in schizophrenia.

2. Materials and Methods

2.1. Participants. Sixteen subjects diagnosed with schizo-phrenia (APA, 2002) and seventeen HC participated in thestudy (see Table 1). Inclusion criteria were (a) age between18 and 50 years; (b) no history of neurological illness ortraumatic head injury, defined as loss of consciousness formore than 5 minutes and/or structural sequelae followinghead trauma; (c) no history of alcohol or drug dependencein the past five years or abuse within the last year (DSM-IV-TR; APA-2002) with diagnoses determined by the StructuredClinical Interview for DSM-IV-TR (SCID) administration[69, 70]; (d) no hearing, vision, or upper body impairment;(e) estimated intelligence quotient (IQ) of above 80 [71]; (f)no alcohol use in the 24 hours before testing; (g) an abilityand desire to participate in the experimental procedure, asdemonstrated by given written informed consent, followingHarvard Medical School (HMS) and Veterans Affairs BostonHealthcare System (VABHS) guidelines.

HC subjects were recruited from Internet and newspaperadvertisements and matched to the patients on the basis ofage, gender, parental socioeconomic status, and handedness(Table 1). For HC, additional inclusion criteria were nohistory of Axes I or II disorders as determined by SCIDinterview [69, 70]; no history of Axis I disorder in first orsecond degree family members, determined by the FamilyHistory-Research Diagnostic Criteria (FH-RDC) instrument[72]; no history of attention deficit disorder, learningdisability or developmental disorder, and no history of birth


Table 2: Psycholinguistic properties of adjectives used in the experiment.

Psycholinguistic Adjectives’ valence

measure Neutral (M ± SD) Positive (M ± SD) Negative (M ± SD)

Kucera-Francis written frequency 114.93 ± 141.98 96.92 ± 163.10 41.39 ± 51.42

Familiarity 561.24 ± 44.44 568.88 ± 41.99 548.17 ± 41.24

Concreteness 402.93 ± 51.94 342.05 ± 48.65 356.32 ± 49.45

Imageability 426.40 ± 95.41 427.97 ± 53.26 427.14 ± 51.88

Word length (number of letters) 4.92 ± 1.03 5.36 ± 1.16 5.32 ± 1.36

Word length (number ofsyllables)

1.43 ± 0.50 1.51 ± 0.51 1.60 ± 0.49

The range and direction of valence is 1 (extremely unpleasant) to 9 (extremely pleasant).

complications with resulting consequences for central ner-vous system as determined by neurodevelopmental interview[73].

The experiment was explained to each participant and allparticipants gave a written informed consent following HMSand VABHC guidelines. All were paid for their participationin the study.

2.2. Stimuli. Stimuli were 184 adjectives (see Table 3) differ-ing in emotional valence. First, a list of neutral and emotional(pleasant or positive; unpleasant or negative) adjectives wascreated. Given that the desired number of adjectives for eachvalence type could not be found in the ANEW dataset [40],we turned to published studies that have used emotionaladjectives as stimuli. Thus, we have combined words takenfrom the ANEW with words from those original studies thatpublished the lists of words as supplementary material [40,74, 75] to arrive at the final set of stimuli. Five judges (meanage ± SD = 31.4 ± 12.10 years, 3 females), all with collegedegree (mean years of formal education = 16), involved inresearch and who did not participate in the experimentaltask, categorized each word as “neutral,” “positive,” and“negative”: 60 words were categorized as neutral, 60 wordswere categorized as positive, and 64 words were categorizedas negative. Neutral adjectives described less arousing andless salient traits and states (e.g., “neutral,” “blue,” and“narrow”), while positive (e.g., “brilliant,” “famous,” “ele-gant”) and negative (e.g., “dead,” “fearful,” “sad”) adjectivesdescribed more arousing and salient affective traits and states(see also [53]).

Psycholinguistic properties were taken from the Uni-versity of Western Australia database (http://www.psych.rl.ac.uk/MRC Psych Db.html). Words in three valence cat-egories (neutral, positive, and negative) were matched fornumber of letters (four to seven letters), and number ofsyllables (one to three syllables). No difference was observedbetween the three valence types (P > 0.05). Word frequency(Kucera-Francis frequency [76]) was in the range of 1–300per million words. However, it was somewhat lower fornegative adjectives relative to both positive (P = 0.020)and neutral (P = 0.002) ones (see Table 2). Familiarity,concreteness, and imageability were also lower for negativerelative to both positive (P < 0.001) and neutral (P < 0.001)adjectives.

2.3. Procedure. Adjectives were pseudorandomized and pre-sented to each participant. Pseudorandomization was used inorder to avoid sequential presentation of more than 5 stimuliwith similar emotional valence. Words were presented inlowercase in 6 blocks in the center of a CRT computer screen,in black Arial 40-point font against white background. Eachblock consisted of 30 words. A short pause (self-paced)followed each block to minimize participants’ fatigue ordistraction. Before the presentation of a word, the fixationcross (5000 ms of duration) appeared. A word was thenpresented for 3000 ms. Following the word presentation,participants had 15000 ms to respond. They were instructedto read each word silently and to rate its emotional valenceby using a computerized version of the Self-AssessmentManikin [13]. In this system, each affective dimension isassessed on a 1–9 Likert scale: higher numbers in the valencedimension indicate evaluation as more pleasant. Mean itemratings less than 4 were classified as unpleasant, between 4and 6 were classified as neutral, and greater than 6 wereclassified as pleasant. Participants were also given a chanceof correcting their response, in case they felt they madea mistake when pressing a button. After the response, aslide appeared (with no time limit) asking participants toconfirm (by pressing button 1) or to correct their response(by pressing button 0). If subjects wanted to correct it, thetrial would restart. After the confirmation of the response,an interstimulus interval (ISI) of 1000 ms preceded the onsetof the next trial.

The rating session was preceded by a practice session.Subjects were given detailed instructions and presentedwith a block of 9 selected words that were not shownduring the actual experiment. All stimuli were presented andsynchronized through SuperLab 4.2 (Cedrus Corporation,San Pedro, CA, USA). The same software was used forrecording subjects’ responses. Data were analyzed with IBMSPSS Statistics 19 (IBM Corporation, Armonk, NY, USA).

3. Results and Discussion

Words from different categories were rated differently, assuggested by the significant effect of valence (F(2,30) =182.06, P < 0.001). No group effect or interaction involvinggroup factor were observed. Positive words were rated as


Table 3

Healthy controls(n = 17)

Individuals withschizophrenia (n = 16)

Neutral adjectives

Ample 6.12 ± 1.11 6.87 ± 1.64

Aloof 4.18 ± 0.81 4.20 ± 1.42

Blue∗ 5.18 ± 1.88 6.75 ± 1.53

Aware 6.47 ± 1.42 6.69 ± 1.62

Blank 4.71 ± 1.10 4.50 ± 1.63

Airy 5.59 ± 1.37 4.94 ± 1.69

Annual 5.29 ± 0.59 5.75 ± 1.24

Basic 5.29 ± 0.69 6.25 ± 1.65

Blond 6.12 ± 1.32 6.38 ± 1.67

Actual 5.29 ± 0.59 5.81 ± 1.52

Broad 5.41 ± 1.18 5.44 ± 1.21

Casual 6.35 ± 1.58 6.69 ± 1.49

Brief 5.59 ± 1.12 5.50 ± 1.59

Brown 5.18 ± 0.88 5.75 ± 1.84

Classic 6.59 ± 1.23 6.94 ± 2.14

Bold 6.13 ± 1.09 6.20 ± 1.61

Common∗ 5.29 ± 0.85 6.20 ± 1.32

Close 5.76 ± 1.52 5.25 ± 1.48

Civil 6.44 ± 1.31 6.75 ± 1.48

Central∗ 5.24 ± 0.56 6.13 ± 1.46

Dense 4.65 ± 1.22 4.06 ± 2.11

Constant 5.71 ± 1.05 6.19 ± 1.64

Compact 5.53 ± 1.46 5.63 ± 1.41

Daily∗ 5.53 ± 1.12 6.75± 1.69

Cubic 5.06 ± 0.24 5.20 ± 0.86

Curly 5.47 ± 1.12 6.38 ± 1.67

Equal 6.18 ± 1.70 6.81 ± 1.64

Deep 5.06 ± 1.03 5.33 ± 2.02

Dry 4.71 ± 1.49 4.94 ± 2.32

Direct 6.65 ± 1.54 7.13 ± 1.50

Green 5.94 ± 1.25 6.80 ± 1.90

Long 4.76 ± 1.75 5.75 ± 2.35

Herbal 6.18 ± 1.33 5.88 ± 2.00

Loud 3.71 ± 1.21 3.50 ± 1.75

Large 5.35 ± 0.86 5.44 ± 2.56

Exact 6.31 ± 1.40 7.19 ± 1.56

Full 5.88 ± 1.62 5.73 ± 2.12

Lay 5.65 ± 1.37 5.81 ± 2.40

High 6.06 ± 1.82 6.19 ± 2.29

Flat 4.82 ± 1.07 4.19 ± 1.60

Red 5.24 ± 0.75 5.38 ± 1.63

Mutual 6.35 ± 1.17 6.38 ± 2.16

Plural 5.00 ±0.00 5.63 ± 1.26

Main 5.12 ± 0.33 5.75 ± 1.34

Overt 5.24 ± 0.97 4.50 ± 1.55

Quiet 6.00 ± 1.66 6.06 ± 1.44

Plain 4.88 ± 0.34 4.94 ± 1.24


Table 3: Continued.



Raw 4.65 ± 1.11 3.69 ± 1.82

Mild 5.47 ± 0.80 5.75 ± 1.48

Near 5.47 ± 0.87 5.81 ± 1.22

Smooth 6.94 ± 1.25 6.50 ± 1.90

Slim 6.24 ± 1.30 6.81 ± 1.97

White 5.82 ± 2.04 6.44 ± 1.55

Tall 5.94 ± 1.52 7.00 ± 1.86

Tiny 4.53 ± 1.01 5.00 ± 1.59

Yellow 5.65 ± 1.46 5.13 ± 2.36

Wild 5.53 ± 1.01 5.13 ± 1.67

Sharp 5.94 ± 1.68 5.94 ± 2.38

Small 4.94 ± 0.66 5.38 ± 2.06

Thick 4.76 ± 0.56 5.50 ± 1.97

Positive adjectives

Calm 7.23 ± 2.02 7.33 ± 1.40

Beautiful 8.29 ± 0.77 7.47 ± 2.13

Adorable 7.82 ± 0.95 7.25 ± 1.88

Clean 7.82 ± 1.07 7.00 ± 2.07

Alive 8.12 ± 1.17 7.93 ± 2.02

Capable 7.18 ± 1.19 6.75 ± 1.95

Brave 7.71 ± 1.31 7.19 ± 1.38

Blessed 8.12 ± 1.11 7.38 ± 2.03

Confident 8.12 ± 0.99 7.75 ± 1.06

Bright 7.71 ± 1.26 7.63 ± 1.31

Erotic 6.71± 2.11 6.53± 1.64

Elegant 7.94 ± 1.20 7.73 ± 1.39

Famous 6.94 ± 1.52 7.00 ± 1.75

Gentle 7.82± 1.19 7.56± 1.41

Genial∗ 6.88 ± 1.54 4.81 ± 1.64

Cute 7.53 ± 1.18 7.40 ± 1.30

Friendly 8.18± 0.81 7.94± 1.18

Free 7.94 ± 1.25 7.81 ± 2.10

Fabulous 7.94± 1.03 7.69 ± 1.25

Funny 8.18 ± 1.01 8.06± 1.00

Gifted 7.82 ± 1.01 8.00 ± 1.21

Joyful 8.06 ± 1.25 7.81 ± 1.17

Happy 8.12 ± 1.27 8.00 ± 1.26

Good 7.53 ± 1.01 7.50 ± 1.32

Honest 7.82 ± 0.95 8.13 ± 1.36

Inspired 7.71 ± 1.16 7.56 ± 1.21

Grateful 8.00 ± 1.17 7.63 ± 1.96

Keen 6.41 ± 1.54 6.38 ± 1.67

Hopeful 7.82 ± 1.33 6.75 ± 2.41

Glad 7.35 ± 1.27 7.44 ± 2.03

Loyal 8.00 ± 1.00 7.94 ± 1.24

Magical 7.18 ± 1.19 7.47 ± 1.30

Perfect 7.59 ± 1.06 7.81 ± 2.07

Precious 7.41 ± 1.06 7.53 ± 1.51


Table 3: Continued.



Merry 7.94± 0.83 7.50 ± 2.03

Lovely 8.06 ± 0.75 7.13 ± 2.36

Lucky 7.53 ± 1.46 7.69 ± 1.49

Kind 7.82 ± 1.38 7.75 ± 1.48

Loved 8.47 ± 0.72 8.07 ± 1.10

Nice 7.82 ± 1.01 7.69 ± 1.40

Protected 7.06 ± 1.98 7.31 ± 1.30

Secure 7.59 ± 1.12 7.75 ± 1.29

Safe 7.6 5± 1.32 7.94 ± 1.29

Right 6.59 ± 1.54 7.25 ± 1.29

Pretty 7.47 ± 1.07 7.31 ± 1.49

Proud 7.00 ± 1.37 7.50 ± 1.41

Romantic 8.12 ± 1.05 7.94 ± 1.34

Satisfied 7.53 ± 1.23 7.25 ± 1.39

Sexy 7.88 ± 1.17 7.25 ± 2.05

Relaxed 7.71 ± 1.10 7.44 ± 1.55

Terrific 7.71 ± 1.10 7.50 ± 1.59

Super 7.59 ± 1.18 7.33 ± 1.59

Special 7.47 ± 1.23 7.25 ± 2.14

Soft 6.47 ± 1.18 6.63 ± 1.86

Tender 6.65 ± 1.32 6.75 ± 1.81

Strong 7.41± 1.66 7.00 ± 1.63

Vigorous 6.53 ± 2.10 5.88 ± 1.96

Wise 7.94 ± 1.20 7.63 ± 1.54

Useful 7.24 ± 1.30 7.44 ± 2.06

Wealthy 7.65 ± 1.17 7.63 ± 1.59

Negative adjectives

Bloody 1.71 ± 0.92 2.56 ± 2.00

Bored 3.18 ± 1.29 4.00 ± 2.13

Alone 3.82 ± 1.78 3.25 ± 1.61

Clumsy 3.76 ± 1.30 3.44 ± 1.59

Coarse 4.06 ± 1.39 4.44 ± 1.67

Blind 2.35 ± 1.22 2.44 ± 2.06

Abnormal 3.00 ± 1.37 3.50 ± 1.83

Bad 2.29 ± 1.05 3.19 ± 2.17

Angry 2.24± 0.90 2.75 ± 1.77

Afraid 2.24 ± 0.90 2.19 ± 1.28

False 3.94 ± 2.36 3.81 ± 1.76

Cruel 1.71 ± 0.77 2.50 ± 2.10

Dirty 2.65 ± 1.58 3.19 ± 1.60

Dull 3.06 ± 1.14 3.25 ± 1.53

Cynic 4.29 ± 1.83 4.06± 1.57

Enraged 2.29 ± 1.31 2.63 ± 1.67

Crazy 3.35 ± 1.46 3.25 ± 1.95

Dumb 3.29 ± 1.40 3.00 ± 1.55

Dreadful∗ 2.00 ± 1.12 3.31 ± 2.09

Dead∗ 1.29 ± 0.69 2.38 ± 2.03

Ill 2.82 ± 1.78 3.50 ± 2.13

Furious∗ 2.12± 0.78 3.44 ± 2.31


Table 3: Continued.



Hostile 2.18 ± 1.19 2.38 ± 1.36

Inferior 2.71 ± 1.26 3.00 ± 1.67

Foolish 3.00 ± 1.22 3.50 ± 1.59

Insane 2.47± 1.84 2.50 ± 1.26

Helpless 2.29 ± 1.05 2.94 ± 1.73

Impure 3.94 ± 1.30 3.63 ± 1.67

Guilty∗ 2.00 ± 0.71 3.00 ± 1.67

Fearful 2.00 ± 1.06 2.44 ± 1.09

Monstrous 3.24 ± 1.60 3.81± 2.10

Lost 2.41± 0.87 2.56± 1.15

Morbid 2.71± 1.61 3.19 ± 2.10

Odd 4.00 ± 1.32 3.94 ± 2.08

Lazy 2.94 ± 0.75 3.88 ± 1.93

Malign 3.12 ± 1.36 3.69 ± 1.70

Lonely 2.12 ± 1.27 2.56 ± 1.41

Mad 2.65 ± 1.06 2.81 ± 1.33

Sinful 2.29 ± 1.36 2.94 ± 2.02

Shamed 2.18 ± 1.42 3.38± 1.75

Rejected 2.06 ± 1.20 3.00 ± 1.83

Poor 2.47 ± 1.37 2.94 ± 1.53

Selfish 2.00 ± 0.94 2.94 ± 1.34

Sick 1.82 ± 0.95 2.19 ± 1.22

Sad 2.35 ± 0.93 2.93 ± 1.62

Odious 4.12 ± 1.45 4.25 ± 1.57

Scared 2.06 ± 1.20 2.88 ± 2.00

Rude 2.12 ± 0.99 2.75 ± 1.18

Useless 2.24 ± 1.03 2.94 ± 1.53

Terrible 2.00 ± 0.79 2.31 ± 1.20

Unhappy 2.18 ± 0.88 2.69 ± 1.30

Upset 2.35 ± 0.79 2.56 ± 1.31

Weak 2.94 ± 1.30 3.31 ± 1.70

Tough 6.06 ± 1.78 5.38 ± 1.78

Wicked 2.94 ± 1.89 2.88 ± 1.82

Tense 3.12± 1.22 4.00 ± 1.93

Ugly 2.35 ± 1.27 3.56 ± 2.10

Stupid 2.47 ± 1.37 2.69 ± 1.40

Terrified 1.76 ± 0.90 2.44 ± 2.13

Wrong 3.12 ± 1.27 2.75 ± 1.13

Violent 1.94 ± 1.56 2.38 ± 1.20

Tragic 1.94 ± 1.09 2.13 ± 1.25

Mean 2.35 ± 1.11 2.63 ± 1.89

Jealous 2.59 ± 1.18 2.69 ± 1.35∗P < .05.

higher in valence, followed by neutral, and finally by negativewords (P < 0.001 for all comparisons).

However, independent t-tests examining group differ-ences for particular items showed differences in ratings for

a subset of words. In general, some of the words that werepreviously categorized as “neutral” by a group of volunteerswere rated more positively by individuals with schizophreniawhen compared with HC, such as “blue” (P = 0.013; SZ =


10

8

6

4

2

0Neutral Negative Positive

HCSZ

Figure 1: Ratings of adjectives previously categorized as “neutral,”“positive,” and “negative” by healthy controls and schizophrenicindividuals.

6.75; HC = 5.18); “basic” (P = 0.036; SZ = 6.25; HC = 5.29);“common” (P = 0.026; SZ = 6.20; HC = 5.29); “central” (P =0.025; SZ = 6.13; HC = 5.24); “daily” (P = 0.020; SZ = 6.75;HC = 5.53); “plural” (P = 0.049; SZ = 5.63; HC = 5.00)

In addition, the word “genial,” previously categorized as“positive” by a group of volunteers, was rated was rated lesspositively by individuals with schizophrenia relative to HC(P = 0.001; HC = 6.88; SZ = 4.81).

Finally, some of the words that were previously catego-rized as “negative” by a group of volunteers were rated lessnegatively by individuals with schizophrenia when comparedwith HC: “dreadful” (P = 0.030; HC = 2.00; SZ = 3.31);“dead” (P = 0.046; HC = 1.29; SZ = 2.38); “furious” (P =0.033; HC = 2.12; SZ = 3.44); “guilty” (P = 0.031; HC =2.00; SZ = 3.00); “shamed” (P = 0.038; HC = 2.18; SZ =3.38); “selfish” (P = 0.026; HC = 2.00; SZ = 2.94); “ugly”(P = 0.052; HC = 2.35; SZ = 3.56).

Given that the words “genial,” “furious,” and “selfish”have lower frequency values than the other words for whichthe groups’ ratings differed (Kucera-Francis frequency = 5,8, and 8, resp.), we have tested for the effects of yearsof education on differences in valence ratings of thesespecific words. Therefore, we have included education as acovariate in our ANOVA. No significant effect was found(“genial;” P = 0.131; “furious;” P = 0.937; “selfish;” P =0.423).

These findings are consistent with previous studies re-porting similar evaluation of valence of affective stimuli inindividuals with schizophrenia and HC (e.g., [4, 9, 20, 21,31]), supporting the hypothesis that the representation ofemotion in schizophrenia is similar to controls (at least interms of the valence dimension) (Figure 1).

However, the current results should be considered in thecontext of such limitations, as the small sample size and thefact that schizophrenic individuals were medicated. A largersample including more women with schizophrenia diagnosiswill allow, as well, the investigation of potential genderdifferences in the representation of affective knowledge forverbal stimuli [77, 78]. Also, given that findings on emotional

experience (that include ratings of stimuli with affectiveproperties) are more variable than findings on emotionalexpression, possibly due to differences in the stimuli used(e.g., face expressions, pictures, odors) and to differencesin patients’ samples (e.g., schizophrenia subtype, gender,clinical symptoms), replication of these findings is neededwith different schizophrenia subtypes and clinical symptoms(e.g., positive versus negative symptomatology) (see [7] for areview of emotional response deficits in schizophrenia).

Future studies could extend the current findings byexploring how schizophrenic individuals assess the arousal ofaffective verbal material. This could be done by incorporatinga 1–9 scale for arousal ratings (from not arousing to extreme-ly arousing) as suggested by Bradley and Lang [13], allowingfor the study of group differences in the intensity of stimuli.

Additionally, since deficits in emotion processing arealready observed in the prodromal stage of the disorder [5], itwould be interesting to explore the representation of affectiveknowledge and its effects on processing of verbal affect-related stimuli in prodromal and first-episode schizophrenicindividuals in comparison with HC and chronic schizophre-nia. This would allow a better understanding of possiblechanges in emotional processing before the frank onset ofpsychosis and in the first stages of the disease, particularlyin terms of ratings of stimuli’s affective properties. Futurestudies should address these issues and questions.

4. Conclusions

This study aimed to investigate how schizophrenic individ-uals rate the valence of adjectives, when compared withhealthy controls. Results indicated similar ratings of emo-tional valence of words, providing support for a similarrepresentation of affective knowledge related to words inschizophrenia, at least in terms of the valence dimension.Therefore, these findings further suggest that the process ofextracting emotional information from semantically emo-tional words is similar in individuals with schizophrenia andhealthy controls, increasing the confidence in self-reports ofaffect in this clinical group.

Acknowledgments

This work was supported by a postdoctoral Grant (SFRH/BPD/68967/2010) and by the PTDC/PSI-PCL/116626/2010Grant from Fundacao para a Ciencia e a Tecnologia-FCT(Portugal) both awarded to A. P. Pinheiro., and by the Na-tional Institute of Mental Health-NIMH (RO1 MH 040799grant awarded to R. W. McCarley.; RO3 MH 078036 grantawarded to M. Niznikiewicz). A special acknowledgment toall the participants in this study.

References

[1] A. Aleman and R. S. Kahn, “Strange feelings: do amygdalaabnormalities dysregulate the emotional brain in schizophre-nia?” Progress in Neurobiology, vol. 77, no. 5, pp. 283–298,2005.


[2] J. Edwards, H. J. Jackson, and P. E. Pattison, “Emotion recogni-tion via facial expression and affective prosody in schizophre-nia: a methodological review,” Clinical Psychology Review, vol.22, no. 6, pp. 789–832, 2002.

[3] G. R. Kuperberg, D. A. Kreher, A. Swain, D. C. Goff, andD. J. Holt, “Selective emotional processing deficits to socialvignettes in schizophrenia: an ERP study,” SchizophreniaBulletin, vol. 37, no. 1, pp. 148–163, 2011.

[4] B. Rockstroh, M. Junghofer, T. Elbert, G. Buodo, and G. A.Miller, “Electromagnetic brain activity evoked by affectivestimuli in schizophrenia,” Psychophysiology, vol. 43, no. 5, pp.431–439, 2006.

[5] L. K. Phillips and L. J. Seidman, “Emotion processing in per-sons at risk for schizophrenia,” Schizophrenia Bulletin, vol. 34,no. 5, pp. 888–903, 2008.

[6] M. F. Green, D. L. Penn, R. Bentall et al., “Social cognitionin schizophrenia: an NIMH workshop on definitions, assess-ment, and research opportunities,” Schizophrenia Bulletin, vol.34, no. 6, pp. 1211–1220, 2008.

[7] A. M. Kring and E. K. Moran, “Emotional response deficits inschizophrenia: insights from affective science,” SchizophreniaBulletin, vol. 34, no. 5, pp. 819–834, 2008.

[8] S. L. Kerr and J. M. Neale, “Emotion perception in schizophre-nia: specific deficit or further evidence of generalized poorperformance?” Journal of Abnormal Psychology, vol. 102, no.2, pp. 312–318, 1993.

[9] A. M. Kring, S. L. Kerr, D. A. Smith, and J. M. Neale, “Flataffect in schizophrenia does not reflect diminished subjectiveexperience of emotion,” Journal of Abnormal Psychology, vol.102, no. 4, pp. 507–517, 1993.

[10] W. F. Flack Jr., J. D. Laird, and L. A. Cavallaro, “Emotionalexpression and feeling in schizophrenia: effects of specific ex-pressive behaviors on emotional experiences,” Journal of Clin-ical Psychology, vol. 55, no. 1, pp. 1–20, 1999.

[11] K. Becerril and D. Barch, “Influence of emotional processingon working memory in schizophrenia,” Schizophrenia Bulletin,vol. 37, no. 5, pp. 1027–1038, 2011.

[12] U. Habel, K. Pauly, K. Koch et al., “Emotion-cognition interac-tions in schizophrenia,” World Journal of Biological Psychiatry,vol. 11, no. 8, pp. 934–944, 2010.

[13] M. M. Bradley and P. J. Lang, “Measuring emotion: the self-assessment manikin and the semantic differential,” Journal ofBehavior Therapy and Experimental Psychiatry, vol. 25, no. 1,pp. 49–59, 1994.

[14] C. Osgood, G. Suci, and P. Tannenbaum, The Measurement ofMeaning, University of Illinois, Urbana, Ill, USA, 1957.

[15] J. A. Russell, “A circumplex model of affect,” Journal of Person-ality and Social Psychology, vol. 39, no. 6, pp. 1161–1178, 1980.

[16] F. Dolcos, K. S. Labar, and R. Cabeza, “Dissociable effects ofarousal and valence on prefrontal activity indexing emotionalevaluation and subsequent memory: an event-related fMRIstudy,” NeuroImage, vol. 23, no. 1, pp. 64–74, 2004.

[17] B. Rozenkrants, J. K. Olofsson, and J. Polich, “Affective visualevent-related potentials: arousal, valence, and repetition ef-fects for normal and distorted pictures,” International Journalof Psychophysiology, vol. 67, no. 2, pp. 114–123, 2008.

[18] P. Ekman, “An argument for basic emotions,” Cognition andEmotion, vol. 6, no. 3, pp. 169–200, 1992.

[19] L. H. Sweet, M. Primeau, C. G. Fichtner, and G. Lutz, “Disso-ciation of affect recognition and mood state from blunting inpatients with schizophrenia,” Psychiatry Research, vol. 81, no.3, pp. 301–308, 1998.

[20] R. J. Hempel, J. H. Tulen, N. J. Van Beveren, H. G. VanSteenis, P. G. Mulder, and M. W. Hengeveld, “Physiological

responsivity to emotional pictures in schizophrenia,” Journalof Psychiatric Research, vol. 39, no. 5, pp. 509–518, 2005.

[21] W. P. Horan, J. K. Wynn, A. M. Kring, R. F. Simons, and M. F.Green, “Electrophysiological correlates of emotional respond-ing in schizophrenia,” Journal of Abnormal Psychology, vol. 119,no. 1, pp. 18–30, 2010.

[22] E. S. Herbener, C. Rosen, T. Khine, and J. A. Sweeney, “Failureof positive but not negative emotional valence to enhancememory in schizophrenia,” Journal of Abnormal Psychology,vol. 116, no. 1, pp. 43–55, 2007.

[23] E. S. Herbener, W. Song, T. T. Khine, and J. A. Sweeney, “Whataspects of emotional functioning are impaired in schizophre-nia?” Schizophrenia Research, vol. 98, no. 1-3, pp. 239–246,2008.

[24] S. W. Quirk and M. E. Strauss, “Visual exploration of emotioneliciting images by patients with schizophrenia,” Journal ofNervous and Mental Disease, vol. 189, no. 11, pp. 757–765,2001.

[25] F. Schneider, R. C. Gur, R. E. Gur, and D. L. Shtasel, “Emo-tional processing in schizophrenia: neurobehavioral probes inrelation to psychopathology,” Schizophrenia Research, vol. 17,no. 1, pp. 67–75, 1995.

[26] F. Schneider, U. Weiss, C. Kessler et al., “Differential amygdalaactivation in schizophrenia during sadness,” SchizophreniaResearch, vol. 34, no. 3, pp. 133–142, 1998.

[27] H. Takahashi, M. Koeda, K. Oda et al., “An fMRI study ofdifferential neural response to affective pictures in schizophre-nia,” NeuroImage, vol. 22, no. 3, pp. 1247–1254, 2004.

[28] A. S. Cohen and K. S. Minor, “Emotional experience inpatients with schizophrenia revisited: meta-analysis of labo-ratory studies,” Schizophrenia Bulletin, vol. 36, no. 1, pp. 143–150, 2010.

[29] E. A. Heerey and J. M. Gold, “Patients with schizophreniademonstrate dissociation between affective experience andmotivated behavior,” Journal of Abnormal Psychology, vol. 116,no. 2, pp. 268–278, 2007.

[30] R. J. Hempel, J. H. Tulen, N. J. van Beveren, P. G. Mulder,and M. W. Hengeveld, “Subjective and physiological responsesto emotion-eliciting pictures in male schizophrenic patients,”International Journal of Psychophysiology, vol. 64, no. 2, pp.174–183, 2007.

[31] S. F. Taylor, K. L. Phan, J. C. Britton, and I. Liberzon,“Neural response to emotional salience in schizophrenia,”Neuropsychopharmacology, vol. 30, no. 5, pp. 984–995, 2005.

[32] D. J. Holt, A. P. Weiss, S. L. Rauch et al., “Sustained activationof the hippocampus in response to fearful faces in schizophre-nia,” Biological Psychiatry, vol. 57, no. 9, pp. 1011–1019, 2005.

[33] C. I. Rupp, W. W. Fleischhacker, G. Kemmler et al., “Olfactoryfunctions and volumetric measures of orbitofrontal andlimbic regions in schizophrenia,” Schizophrenia Research, vol.74, no. 2-3, pp. 149–161, 2005.

[34] F. Schneider, U. Habel, M. Reske, I. Toni, P. Falkai, and N.J. Shah, “Neural substrates of olfactory processing in schiz-ophrenia patients and their healthy relatives,” Psychiatry Re-search, vol. 155, no. 2, pp. 103–112, 2007.

[35] J. A. Burbridge and D. M. Barch, “Anhedonia and the experi-ence of emotion in individuals with schizophrenia,” Journal ofAbnormal Psychology, vol. 116, no. 1, pp. 30–42, 2007.

[36] E. C. Dowd and D. M. Barch, “Anhedonia and emotional expe-rience in schizophrenia: neural and behavioral indicators,”Biological Psychiatry, vol. 67, no. 10, pp. 902–911, 2010.

[37] G. P. Strauss, D. N. Allen, S. A. Ross, L. A. Duke, and J.Schwartz, “Olfactory hedonic judgment in patients with


deficit syndrome schizophrenia,” Schizophrenia Bulletin, vol.36, no. 4, pp. 860–868, 2010.

[38] G. P. Strauss and E. S. Herbener, “Patterns of emotional ex-perience in schizophrenia: differences in emotional responseto visual stimuli are associated with clinical presentation andfunctional outcome,” Schizophrenia Research, vol. 128, no. 1-3,pp. 117–123, 2011.

[39] S. R. Aminoff, J. Jensen, T. V. Lagerberg, O. A. Andreassen, andI. Melle, “Decreased self-reported arousal in schizophreniaduring aversive picture viewing compared to bipolar disorderand healthy controls,” Psychiatry Research, vol. 185, no. 3, pp.309–314, 2011.

[40] M. M. Bradley and P. J. Lang, “Affective norms for Englishwords (ANEW): instruction manual and affective ratings,”Tech. Rep., The Center for Research in Psychophysiology,University of Florida, Gainesville, Fla, USA, 1999.

[41] M. M. Bradley and P. J. Lang, “International affective digitizedsounds (IADS): stimuli, instruction manual and affectiveratings,” Tech. Rep., The Center for Research in Psychophys-iology, University of Florida, Gainesville, Fla, USA, 1999.

[42] A. M. Kring, L. F. Barrett, and D. E. Gard, “On the broadapplicability of the affective circumplex: representations ofaffective knowledge among schizophrenia patients,” Psycholog-ical Science, vol. 14, no. 3, pp. 207–214, 2003.

[43] P. G. Nestor, M. O. Kimble, B. F. O’Donnell et al., “Aberrantsemantic activation in schizophrenia: a neurophysiologicalstudy,” American Journal of Psychiatry, vol. 154, no. 5, pp. 640–646, 1997.

[44] D. H. Mathalon, B. J. Roach, and J. M. Ford, “Automaticsemantic priming abnormalities in schizophrenia,” Interna-tional Journal of Psychophysiology, vol. 75, no. 2, pp. 157–166,2010.

[45] S. L. Rossell, J. Shapleske, and A. S. David, “Direct andindirect semantic priming with neutral and emotional wordsin schizophrenia: relationship to delusions,” Cognitive Neu-ropsychiatry, vol. 5, no. 4, pp. 271–292, 2000.

[46] M. A. Niznikiewicz, B. F. O’Donnell, P. G. Nestor et al., “ERPassessment of visual and auditory language processing inschizophrenia,” Journal of Abnormal Psychology, vol. 106, no.1, pp. 85–94, 1997.

[47] M. Kubicki, R. W. McCarley, P. G. Nestor et al., “An fMRIstudy of semantic processing in men with schizophrenia,”NeuroImage, vol. 20, no. 4, pp. 1923–1933, 2003.

[48] M. Niznikiewicz, “Future directions for examining semanticmemory in schizophrenia spectrum disorders,” Clinical EEGand Neuroscience, vol. 39, no. 2, pp. 95–98, 2008.

[49] M. Niznikiewicz, M. S. Mittal, P. G. Nestor, and R. W. McCar-ley, “Abnormal inhibitory processes in semantic networks inschizophrenia,” International Journal of Psychophysiology, vol.75, no. 2, pp. 133–140, 2010.

[50] S. L. Rossell and R. A. Batty, “Elucidating semantic disorgan-isation from a word comprehension task: do patients withschizophrenia and bipolar disorder show differential process-ing of nouns, verbs and adjectives?” Schizophrenia Research,vol. 102, no. 1-3, pp. 63–68, 2008.

[51] M. Bayer, W. Sommer, and A. Schacht, “Emotional wordsimpact the mind but not the body: evidence from pupillaryresponses,” Psychophysiology, vol. 48, no. 11, pp. 1554–1562,2011.

[52] E. Bernat, S. Bunce, and H. Shevrin, “Event-related brain po-tentials differentiate positive and negative mood adjectivesduring both supraliminal and subliminal visual processing,”International Journal of Psychophysiology, vol. 42, no. 1, pp. 11–34, 2001.

[53] C. Herbert, T. Ethofer, S. Anders et al., “Amygdala activationduring reading of emotional adjectives—an advantage forpleasant content,” Social Cognitive & Affective Neuroscience,vol. 4, no. 1, pp. 35–49, 2009.

[54] C. Herbert, M. Junghofer, and J. Kissler, “Event related poten-tials to emotional adjectives during reading,” Psychophysiology,vol. 45, no. 3, pp. 487–498, 2008.

[55] C. Herbert, J. Kissler, M. Junghofer, P. Peyk, and B. Rockstroh,“Processing of emotional adjectives: evidence from startleEMG and ERPs,” Psychophysiology, vol. 43, no. 2, pp. 197–206,2006.

[56] R. J. Maddock, A. S. Garrett, and M. H. Buonocore, “Posteriorcingulate cortex activation by emotional words: fMRI evidencefrom a valence decision task,” Human Brain Mapping, vol. 18,no. 1, pp. 30–41, 2003.

[57] J. Kissler, C. Herbert, I. Winkler, and M. Junghofer, “Emotionand attention in visual word processing: an ERP study,”Biological Psychology, vol. 80, no. 1, pp. 75–83, 2009.

[58] J. Kissler, C. Herbert, P. Peyk, and M. Junghofer, “Buzzwords:early cortical responses to emotional words during reading:research report,” Psychological Science, vol. 18, no. 6, pp. 475–480, 2007.

[59] S. L. Rossell, “Affective semantic priming in patients withschizophrenia,” Psychiatry Research, vol. 129, no. 3, pp. 221–228, 2004.

[60] T. Suslow, C. Roestel, T. Droste, and V. Arolt, “Automatic pro-cessing of verbal emotion stimuli in schizophrenia,” PsychiatryResearch, vol. 120, no. 2, pp. 131–144, 2003.

[61] H. Klumpp, J. Keller, G. A. Miller, B. R. Casas, J. L. Best,and P. J. Deldin, “Semantic processing of emotional wordsin depression and schizophrenia,” International Journal ofPsychophysiology, vol. 75, no. 2, pp. 211–215, 2010.

[62] P. O. Harvey, J. Lee, W. P. Horan, K. Ochsner, and M. F. Green,“Do patients with schizophrenia benefit from a self-referentialmemory bias?” Schizophrenia Research, vol. 127, no. 1–3, pp.171–177, 2011.

[63] R. E. Hoffman, M. Varanko, J. Gilmore, and A. L. Mishara,“Experiential features used by patients with schizophrenia todifferentiate ‘voices’ from ordinary verbal thought,” Psycholog-ical Medicine, vol. 38, no. 8, pp. 1167–1176, 2008.

[64] F. Laroi, “How do auditory verbal hallucinations in patientsdiffer from those in non-patients?” Frontiers in HumanNeuroscience, vol. 6, p. 25, 2012.

[65] T. H. Nayani and A. S. David, “The auditory hallucination: aphenomenological survey,” Psychological Medicine, vol. 26, no.1, pp. 177–189, 1996.

[66] P. J. Lang, M. M. Bradley, and B. N. Cuthbert, “Technicalmanual and affective ratings,” in International Affective PictureSystem (IAPS), Center for Research in Psychophysiology, Uni-versity of Florida, Gainesville, Fla, USA, 1999.

[67] F. Schneider, R. C. Gur, K. Koch et al., “Impairment in thespecificity of emotion processing in schizophrenia,” AmericanJournal of Psychiatry, vol. 163, no. 3, pp. 442–447, 2006.

[68] N. Y. Seiferth, K. Pauly, T. Kellermann et al., “Neuronalcorrelates of facial emotion discrimination in early onsetschizophrenia,” Neuropsychopharmacology, vol. 34, no. 2, pp.477–487, 2009.

[69] M. B. First, R. L. Spitzer, M. Gibbon, and J. B. M. Williams,Structured Clinical Interview for DSM-IV Axis II PersonalityDisorders (SCID-II, Version 2.0), Biometrics Research Depart-ment, New York State Psychiatric Institute, New York, NY,USA, 1995.


[70] M. B. First, R. L. Spitzer, M. Gibbon, and J. B. M. Williams,Structured Clinical Interview for DSM-IV Axis I Diagnosis-Patient Edition (SCID-I/P, Version 2.0), Biometrics ResearchDepartment, New York State Psychiatric Institute, New York,NY, USA, 2002.

[71] D. Wechsler, Wechsler Adult Intelligence Scale: Administrationand Scoring Manual, The Psychological Corporation, SanAntonio, Tex, USA, 3rd edition, 1997.

[72] N. C. Andreasen, J. Endicott, R. L. Spitzer, and G. Winokur,“The family history method using diagnostic criteria: reliabil-ity and validity,” Archives of General Psychiatry, vol. 34, no. 10,pp. 1229–1235, 1977.

[73] S. V. Faraone, L. J. Seidman, W. S. Kremen, J. R. Pepple, M.J. Lyons, and M. T. Tsuang, “Neuropsychological functioningamong the nonpsychotic relatives of schizophrenic patients: adiagnostic efficiency analysis,” Journal of Abnormal Psychology,vol. 104, no. 2, pp. 286–304, 1995.

[74] P. P. Allen, L. C. Johns, C. H. Fu, M. R. Broome, G. N.Vythelingum, and P. K. McGuire, “Misattribution of externalspeech in patients with hallucinations and delusions,” Schizo-phrenia Research, vol. 69, no. 2-3, pp. 277–287, 2004.

[75] L. C. Johns, S. Rossell, C. Frith et al., “Verbal self-moni-toring and auditory verbal hallucinations in patients withschizophrenia,” Psychological Medicine, vol. 31, no. 4, pp. 705–715, 2001.

[76] H. Kucera and W. N. Francis, Computational Analysis ofPresent-Day American English, Brown University Press, Provi-dence, RI, USA, 1967.

[77] A. M. Kring and A. H. Gordon, “Sex differences in emotion:expression, experience, and physiology,” Journal of Personalityand Social Psychology, vol. 74, no. 3, pp. 686–703, 1997.

[78] E. H. Nasser, N. Walders, and J. H. Jenkins, “The experienceof schizophrenia: what’s gender got to do with it? A criticalreview of the current status of research on schizophrenia,”Schizophrenia Bulletin, vol. 28, no. 2, pp. 351–362, 2002.


Review Article

Schizophrenia as a Disorder of Social Communication

Cynthia Gayle Wible

Laboratory of Neuroscience, Department of Psychiatry VA Boston Healthcare System, Harvard Medical School,Psychiatry 116A, 940 Belmont Street, Brockton, MA 02301, USA

Correspondence should be addressed to Cynthia Gayle Wible, [email protected]

Received 20 December 2011; Revised 23 February 2012; Accepted 21 March 2012

Academic Editor: Margaret A. Niznikiewicz

Copyright © 2012 Cynthia Gayle Wible. This is an open access article distributed under the Creative Commons AttributionLicense, which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properlycited.

Evidence is reviewed for the existence of a core system for moment-to-moment social communication that is based on theperception of dynamic gestures and other social perceptual processes in the temporal-parietal occipital junction (TPJ), includingthe posterior superior temporal sulcus (PSTS) and surrounding regions. Overactivation of these regions may produce theschizophrenic syndrome. The TPJ plays a key role in the perception and production of dynamic social, emotional, and attentionalgestures for the self and others. These include dynamic gestures of the body, face, and eyes as well as audiovisual speech andprosody. Many negative symptoms are characterized by deficits in responding within these domains. Several properties of thissystem have been discovered through single neuron recording, brain stimulation, neuroimaging, and the study of neurologicalimpairment. These properties map onto the schizophrenic syndrome. The representation of dynamic gestures is multimodal(auditory, visual, and tactile), matching the predominant hallucinatory categories in schizophrenia. Inherent in the perceptualsignal of gesture representation is a computation of intention, agency, and anticipation or expectancy (for the self and others). Theneurons are also tuned or biased to rapidly detect threat-related emotions. I review preliminary evidence that overactivation ofthis system can result in schizophrenia.

1. Introduction

Is there a system for dynamic moment to moment socialcommunication in the brain? Social perception has nowbeen extensively studied in humans and nonhuman primatesand is defined as follows: “. . .the initial states of evaluatingthe social communicative intentions of others by analysis ofeye-gaze direction, facial expressions, body movements, andother types of biological motion.” [1]. Recent evidencesuggests that such a system does exist and that abnormalactivity in this system may produce the symptoms andcognitive deficits that comprise the syndrome of schizophre-nia. A posterior system will be described whose activitymay correspond to or underlie the perceptual experienceof conversing and interacting with others. Until recently,communication has primarily been studied through lan-guage research. This line of research has focused mainly onthe structure of the representation and the neural basis oflanguage input (graphemes, phonemes), the internal lexi-cal/semantic representation, and language output (writing

and speech production) and has made significant progressin understanding language.

2. Communication Involves Not OnlyLanguage Processing but Also SocialCognitive Functions

Communication with another individual involves a set ofrepresentations or processes that are outside the scope oftraditional language research. Many of these representationalsystems are studied under the rubric of social cognition.Social communication (in humans) necessitates the dynamicperception of speech, social-emotional cues, and the produc-tion of the communicative message (narrative, intonation,and gestures).

The speech signal that is used to communicate is (pri-marily) both auditory and visual in nature. In naturalisticsettings, the perception of speech occurs simultaneouslywith the dynamic visual perception of gestures, especially


of the face. The audiovisual speech signal not only conveysmeaning through words but also contains information aboutemotion (facial expressions and prosody). Perception duringconversation also necessitates a host of other functions thatare crucial for communication. During the rapid interplay ofconversation, the ability to anticipate other’s thoughts andactions is also essential. Theory of mind is used; this is theability to “think about thoughts” and to represent other’spoint of view. The representation of agency (who is doingwhat) and intention, both in other’s and in one’s self, areimportant. The correct deployment of attention and workingmemory are needed as well as the ability to follow or trackother’s behavior and speech. A narrative of the conversationmust be built up and maintained; this is an understanding ofthe context and the story or message that is being conveyedby the individual words and sentences. During conversation,meaning is built up not only from words but also from bodyand facial gestures as well as intonation (prosody). Recentevidence suggests that there may be a set of brain areas whosefunctionality matches the representations and processesneeded during moment to moment social communication[2–6]. The temporal parietal occipital junction or TPJ(especially in the right hemisphere) is referred to as the“social brain area” [7] because it has been implicated innumerous neuroimaging studies of social cognitive functionssuch as theory of mind, empathy, social attention, andother functions [2, 3]. Figures 1 and 2 show the functionalarchitecture of the TPJ. It has been hypothesized that the TPJis involved in lower-level processes associated with empathy,theory of mind, agency, and attentional orienting to salientstimuli and that these processes are crucial to higher-levelsocial function [2] (see Figure 2). The TPJ includes theinferior parietal region as well as the posterior superiortemporal sulcus (PSTS) and gyrus (note that only the righthemisphere is shown, but the left TPJ region is also involvedin these functions). The functional territory for the PSTS isoften thought to extend inferiorly to at least the posteriormiddle temporal gyrus. The notion that the TPJ is crucialfor social interaction/communication has now been testeddirectly in human subjects. A recent naturalistic functionalmagnetic resonance imaging (FMRI) study of live face-to-face interaction within an MRI scanner showed strongevidence that the TPJ is a core region for moment to momentsocial interaction [4]. Medial prefrontal regions have alsobeen hypothesized to play a role in social functioning anddata on this point will be reviewed (predominately in thelater section of the paper).

3. General Characteristics and AnatomicalConnections of the TPJ

The PSTS has a major role in the perception of dynamicgestures. A basic function of the inferior parietal region is inthe formation of intentions for action and this region (as wellas the intraparietal sulcus) contains mirror representationsthat are active during either the perception or production ofan action [8]. The intraparietal sulcus separates the inferiorand superior parietal lobe. Portions of the intraparietal

Figure 1: Summary figure of the overlap of functional regionsin the TPJ (inferior parietal and PSTS); this region is involvedin eye gaze (red); audiovisual speech (orange); self-representation(yellow); theory of mind (green); emotional perception of facesand prosody (dark blue and purple). Rerepresentation of data is,respectively, from [37–40, 47].

Reorienting Empathy

Agency Theory of mind

Figure 2: Results of a quantitative meta-analysis of 70 functionalneuroimaging studies. Activation likelihood estimation maps inthe right temporoparietal junction are shown. Colors code theprobability of activation with brighter yellow indicating the highestactivation (reprinted with permission from [2]).

sulcus such as the lateral intraparietal region or LIP arethought to contain a dynamically constructed saliency mapthat is used to guide gestures such as saccades [8–10].The TPJ and surrounding regions project to the inferiorfrontal region and are closely anatomically associated withhippocampal and insular regions [5, 6, 11]. Interoceptiveand homeostatic body signals activate the insula (includingthirst, sensual touch, sexual arousal, warmth, heartbeat, andbladder distension) [12]. Subjective ratings of these bodysignals correlate most strongly with activation in the anteriorinsula and this region has been hypothesized to underlieemotional awareness [12]. Craig describes the anterior insulafor “feeling emotion and homeostatic and body feelings”


relative to the anterior cingulate’s limbic motor function orrole in initiating or motivating behavior.

4. Social Cognition Is Abnormal inSchizophrenia

It is well known that social cognitive deficits form a majorpart of the schizophrenic syndrome. Social cognitive deficitsare considered to be a major determinant of functionaloutcome for schizophrenic individuals [13]. Social respon-siveness is affected in schizophrenia; the negative symptomsinclude asociality as well as a lack of facial movement, facialexpression, eye contact, and vocal inflection (prosody). Infact the four most prevalent so-called negative symptomsof schizophrenia are a lack of expressive gestures, a lackof vocal inflection and social inattention, as well as ageneral impairment in attention. Schizophrenic individualsalso present with theory of mind deficits and these deficitsare related to a lack of expressiveness [14, 15]. It is alsoimportant to note that internal experiences of emotion andrelated sensations (perhaps related to insular function) donot seem to be as abnormal in schizophrenia as emotionalresponsiveness [16, 17].

5. Building Links between the Syndromeof Schizophrenia and Brain Dysfunction:a Parsimonious Account of Positive andNegative Symptoms as Stemming from theSame, Similar, and/or Adjacent SocialCognitive Modules within the TPJ

These negative symptoms and social cognitive deficits couldclearly be described as problems with social communicativerepresentations or systems. However, I will argue that most,if not all, of the schizophrenic syndrome can be understoodunder the rubric of social communication. Recent advancesin understanding this system will be described and used toreframe the symptoms of schizophrenia and to then linkthem to brain function. It is proposed that the overactivationof the functional modules in the TPJ that subserve socialcommunication produces the syndrome of schizophrenia.Direct evidence for TPJ involvement in schizophrenia willbe presented near the end of the paper. The formulationpresented here is new and remains to be systematicallytested. However there are many lines of evidence that willbe described below here in after. For example, schizophrenicsymptoms are evoked by stimulation to the TPJ [18–21],electrical brain activity within the TPJ coincides with thepresence and absence of symptoms [22], and the disruptionof TPJ activity alleviates symptoms [23–28].

6. Reframing the Schizophrenic Syndrome andEvaluating the Evidence for the Involvementof the TPJ in Social Processing

Understanding aspects of this system at the neural (sin-gle cell) and systems neuroscience levels may provide a

basis for the understanding the disparate symptoms andcognitive deficits found in the schizophrenic syndrome.Several principles were used to guide the construction ofthe current framework. Neuroimaging alone cannot provideevidence for a region’s involvement in a particular functionand taken alone constitutes relatively weak evidence for alink between a function and a brain region. Lesion studiesmust also show that the function is impaired when theregion is damaged. In evaluating human lesion data it isimportant to analyze each case individually [29]. The extentof involvement of other damaged regions should be takeninto account. Data from patients with brain trauma fromhead injury or with epilepsy should be viewed with cautionas these patients often have diffuse damage that may bemissed in a radiologic examination. Converging evidencefrom several sources provides the strongest basis for positinga relationship between brain and function. For example, ininterpreting functional imaging data, it is also important thatother sources show convergence with the theory, includingbrain stimulation, transcranial magnetic stimulation (TMS),and data from neuronal recordings in humans or animals.Data from single case studies with circumscribed brainlesions that contradict a specific theoretical position must beaccounted for and are a serious indictment of the theoreticalposition [29]. For each functional domain discussed here inafter, I will cite neuroimaging and lesion data to show thatthe TPJ is a core substrate for the function (as well as datafrom the other techniques as discussed previously). Whensingle neuron data are described, I will cite evidence showingsimilar findings from human studies in order to provideexplicit links between cellular properties and systems levelfunction (a detailed description of all the studies cited herecan be found in [5, 6]).

7. The Phenomenology of the SchizophrenicSyndrome Reframed

For brevity, this paper will describe the most prominentsymptoms and their relationship to brain function; a detailedaccount for the other symptoms can be found in [6].Auditory hallucinations of voices are the most prevalentsymptom [30]. Patients have the feeling that someone isactually there and speaking. Auditory hallucinations canalso take the form of hearing conversations between voices.Patients will sometimes report that it seems like there isactually another person or persons in their head. Some typesof auditory hallucinations may be due to a misattributionof the patient’s thoughts to an outside source (abnormalagency attribution for the voice). Visual hallucinations ofpeople in action are the next most prominent category ofhallucination [31]. The most common delusions, delusionsof persecution and delusions of reference, are characterizedby the feeling of being followed, watched, or that peopleare secretly communicating using gestures or clothes [32].The prominent positive symptoms of schizophrenia revolvearound the perception or feeling of an entity that is speaking,communicating, watching, following, observing, or spying.Schizophrenia is also characterized by abnormal agency


judgments; patients can feel as though their actions andthoughts are controlled by an outside force or entity (e.g.,delusions of control, thought insertion, thought broadcast-ing). Therefore, the prominent hallucinations and delusionsmay be reframed as the misperception of a social entityor entities or overactivation of representations in the TPJthat subserve many aspects of dynamic social processing.Throughout the paper I will review properties of dynamicsocial perception and interaction that may be important forunderstanding the schizophrenic syndrome. For example,recent advances in cognitive neuroscience show that theperception of a social entity is accompanied by an intrinsicand automatic perception of agency, intention, and socialprediction [33–36]. These functions are intrinsic to therepresentation of dynamic action at the single neuron level[33, 34, 36]. It has also been shown that self and otherprocessing overlap and the feeling of a presence can alsostem from abnormal activation of the self-representation[19–21, 37]. The prominent negative symptoms are relatedto abnormal social responding and a lack of social expressionor response primarily in the face and voice. The regionsthat are used to perceive dynamic social interaction arealso used in the formation of intentions for the productionof social, emotional, and attentional gestures as well asfor theory of mind [2, 7, 38–42]. This reframing of thesyndrome is parsimonious in that it postulates that thesymptoms can be understood as the integrated action of oneregion (as opposed to widespread abnormalities or complexinteractions between anterior and posterior regions).

In the following sections, I will describe the functionalarchitecture (the arrangement or juxtaposition of functionalregions) as well as properties of the neuronal representationof dynamic social action that can be used to understand theschizophrenic syndrome. The functional contributions of theTPJ have been described in several disparate subfields of neu-roscience and cognitive neuroscience that are conceptuallyoverlapping but motivated by the investigation of differentpsychological functions [2, 5, 6]. In describing these findings,I will also show how individual schizophrenic symptoms canbe reframed and organized under the rubric of TPJ function.

8. The Communicative Signal Is Based on TPJFunction, Is Multimodal (Especially Auditoryand Visual), and Carries an AutomaticComputation of Agency, Intention, andSocial Prediction

The neural code used by the TPJ reflects the fact thatcommunication occurs not only through auditory languagecomprehension but also through gestures that can occursimultaneously with the speech signal. A basic and promi-nent function of the TPJ is to represent dynamic gestures,especially in lateral temporal and inferior parietal regionscentered on the PSTS in humans [42]. Lesions and repetitivetranscranial magnetic stimulation (rTMS) to these regionsimpair biological motion perception [43, 44]. The homol-ogous regions in the monkey have neurons that are multi-modal and respond to the sight, sound, and somatosensory

aspects of biological motion [41, 45]. Single-cell recordingsin monkeys show that audiovisual representation is themost predominant type of representation followed by visual-tactile representation [45]. The PSTS combines informationabout biological motion, social/biographical informationand speech [41, 46]. In humans, the representation of audio-visual speech occupies a large portion of the TPJ territory(see Figure 1 [47]). These relative proportions match the factthat hallucinations of voices and the feeling that someoneis speaking predominate in schizophrenia followed by visualhallucinations of people and then by tactile hallucinations.Hearing voices also activates the TPJ [48].

Multimodal gesture representations have properties thatallow the neurons to participate in more conceptual func-tions such as the computation of agency, the ability to keeptrack of other’s behavior, and the ability to predict theactions of others (social prediction) [33, 34, 49–51]. Forexample, repetition suppression in the monkey STS createsa situation in which neurons that are tuned to sequences ofdynamic gestures are maximally responsive to gestures thatare about to occur [33]. Single-cell recordings in monkeyshave shown that these functions are automatically computedand are an inherent part of the gesture representation atthe single neuron level. Neuroimaging has confirmed thatactivity in the PSTS in humans is also modulated by theperceived intentionality of the gesture or movement [35].In addition, the human STS was found to be a region inwhich interactions occur between observed action and thereafferent motor-related copy of that action [52].

Hence, overactivation of this dynamic gesture systemcould produce the erroneous perception of speech, visualhuman action (visual hallucinations of people), or tactile hal-lucinations. Direct evidence for this supposition comes fromthe fact that cortical stimulation of the TPJ (in nonpsychoticindividuals) and lateral posterior temporal lobe can result invisual hallucinations of action scenes involving people [18].Cortical stimulation of the TPJ can also cause feelings oflimbs or body parts changing shape in nonpsychotic indi-viduals (e.g., somatosensory hallucinations and delusions[19]). The overactivation of gesture representations couldbe graded so that some patients might experience frankhallucinations and others might experience a feeling thatthere are persons acting, watching, or communicating withthem. In addition, this overactivation would be predicted toproduce a feeling of agency or of an agent with intentionsand would be expected to interfere with social prediction andsocial interaction.

Activity in these PSTS/TPJ multimodal neurons thatencode gestures (especially audiovisual gestures) has beenshown in monkeys and humans to provide rapid feedbackexcitation (within 30 msec) to unimodal regions and to resultin gamma synchronization [53]. This phenomenon has beendocumented in several studies (reviewed in [5, 6]). Hence,erroneously activated audiovisual speech representationsmay cause the aberrant activation of auditory cortex andthe gamma synchronization of TPJ and primary auditoryregions. This would result in increased attention and activityin auditory channels.


9. The TPJ Is a Core Substrate forTheory of Mind

The TPJ is a core neural substrate for theory of mindor the ability to understand and represent another’s pointof view or thoughts [39, 54, 55]. Neuroimaging studiesof theory of mind have shown several foci including theTPJ, the orbital frontal cortex, amygdala, anterior insula,and medial frontal regions [56]. The medial prefrontalcortex or the anterior paracingulate cortex—approximatelycorresponding to Brodmann area (BA) 9/32—had beenthought to be a key region for theory of mind function[56]. However, previous reports of an association betweentheory of mind, social processing, and the medial prefrontalcortex have relied on patients with traumatic brain injuryor epilepsy. These types of patient often have diffuse braindamage that is difficult to detect in radiologic exams (see Birdet al. (2004) [57] for a critical discussion of these studies).There have now been several reports of circumscribedlesions to the medial prefrontal cortex where the patientsdo not show theory of mind impairments [57–60]. Bird etal. (2004) [57] present a carefully documented case of apatient whose injury is circumscribed and coincides withthe region identified in several fMRI studies to be involvedin theory of mind (see Figure 3). This individual did showimpairments in planning and memory as well as a tendencyto confabulate but did not show theory of mind deficitson any of a battery of tasks [57]. In another study, threepatients with anterior cingulate damage were tested ontheory of mind, social situation processing, and motivationaldecision making. Two of the patients had selective anteriorcingulate damage (one also had temporal lobe damage).These two patients were not impaired in theory of mind,motivational decision making, or social situation processing[58].

TPJ lesions do impair theory of mind [60]. In addition,recent neuroimaging studies have shown that the (right) TPJhas a response profile that is more selective for theory ofmind than other regions such as the posterior cingulate andthe medial prefrontal cortex. The right TPJ was selectivelyinvolved in the attribution of mental states rather thanreasoning about general socially relevant facts about a person[54]. The ability to make inferences about mental statesor beliefs and intentions during moral judgments was alsoimpaired when TMS was applied to the right TPJ [61].Neurons in the monkey STS that respond to dynamic actionor biological motion are tuned to respond to physical or vocalactions that are about to occur or to anticipate action [33,49]. Aberrant activity in the TPJ would interfere with theoryof mind computations resulting in abnormal or slowedsocial behavior. This abnormal activity would also interferewith dynamic social response and the tracking and henceunderstanding of other’s social behavior. These impairmentsare core components of the schizophrenic syndrome whichincludes symptoms such as asociality and social inattentionas well as theory of mind deficits as described previously[13–15, 62, 63].

Figure 3: Figure from Bird et al., 2004, showing the lesion siteof a patient with anterior cingulate damage who did not showtheory of mind impairments. Peak activation foci related to theoryof mind from ten different neuroimaging studies are shown (boxes)superimposed onto the lesion site. The figure shows that a lesionin the region of the anterior cingulate activated by theory of mindtasks does not cause impairments on these tasks (reprinted withpermission from [57]).

10. Emotional Perception and Reaction

Emotional gesture perception and reaction for the face andbody as well as prosody perception are dependent on theTPJ, presumably because of its role in voice and gesturerepresentation (e.g., eyes and face). Emotional perceptionand reaction are linked; the perception of facial expressionsproduces a subtle version of the expression on the viewer’sface [64]. The perceptual circuits that are used to perceiveother’s actions are also used to represent and plan ourown actions, and several regions are active during bothobservation and imitation of dynamic gestures [7, 64, 65].Single neurons in the monkey PSTS respond to emotionalgestures [49]. Human neuroimaging studies also show thatthe TPJ is involved in emotional gesture and prosodyperception [66, 67]. Lesions within the primary/secondarysomatosensory cortex, supramarginal gyrus (part of inferiorparietal cortex/TPJ), and insula (adjacent and medial tothe TPJ) result in a deficit in the ability to perceive facialemotion and prosody [40]. The supramarginal gyrus maybe especially important for the creation of the intentionsor plans for emotional and prosodic reaction. Hence, theoverexcitation of emotional gesture perceptual circuits couldinterfere with reaction or with the formation of intentionsfor facial emotional reaction. This would result in socialunreactivity or asociality as well as a lack of facial movement,eye contact, vocal inflection, and facial expression; these areall prominent negative symptoms of schizophrenia.

11. Eye Gestures, Social Attention, and Agency

Movements of the eyes are especially important in socialcognition and communication; they not only convey emo-tion but also convey the focus of attention. Dynamicrepresentations of emotional gesture and attention are usedin the perception of more conceptual aspects of social


cognition [35, 68]. The TPJ and especially the posterior STSand superior temporal regions are core components of asystem for the representation and perception of eye gaze[38, 69, 70]. The TPJ was shown to be selectively involvedin representing joint or social attention in a neuroimaginginvestigation of live social interaction [39]. The perceptionof gaze direction and the control of attention via gaze areimpaired with TPJ damage [70, 71]. Aberrant activity in thissystem could produce a deficit in the tracking of others eyegestures resulting in abnormal social attention and hencesocial reaction and understanding.

As discussed previously, the perception of intentionalityand agency are an inherent component of the perception ofdynamic gestures, including eye gaze [35, 72–74]. For exam-ple, in the monkey, neurons in the STS combine informationabout gestures of the arm (e.g., reaching) with informationabout the direction of gaze (the focus of attention) tocompute agency or intentionality [34]. Overactivation ofgaze and other gesture representations would produce thequalia not only of a presence (of being watched) but also ofa presence with intentions. This assumption is corroboratedwith data; direct cortical stimulation of the TPJ can cause thefeeling of a “shadowy presence” [20].

12. Gesture Neurons Are Tuned to RapidlyDetect Threat

Another property of gesture representation in the TPJ isthat the neurons in this neural circuit are tuned to rapidlydetect emotionally negative or socially threating gestures.The amygdala is most often cited with regard to this function.However, amygdala damage does not cause a deficit in thedetection of dynamic body expressions of fear [75, 76]. Thismeans that there is another brain region that is the coresubstrate for this function. Single neuron recording in themonkey STS showed that the neurons are tuned or biasedto detect potentially threatening biological motion [41].Event-related potential (ERP) and neuroimaging studies inhumans have confirmed that the TPJ is used to perceivethreating or emotionally negative gestures [41, 77]. Hence,although this region is generally used to perceive dynamicgestures, there is a neural tuning for the detection of socialthreat. Hence, overactivation of these gesture representationsmight produce the feeling of being in a socially threatingsituation. This feeling could be a component of delusions ofpersecution and other symptoms.

13. Self-Representation, Embodiment,and Agency

The perception of self-made gestures from our own visu-ospatial perspective and the corresponding auditory, visual,tactile, and vestibular information resulting from actionsmay provide the concrete scaffolding that is used to createand sustain the self-representation or the feeling of being anembodied, thinking, and acting person [37]. There are alsomore fixed body schemas that are presumably constructedthrough experience and that are used in the dynamic

representation of the self (these are also located in the vicinityof the TPJ—in the intraparietal sulcus and the secondarysomatosensory cortex; see [6] for a summary). Our sense ofself is so evident and constant that it may be surprising tolearn that it is actually a perceptual construct. However, thisconstruct can break down and be demonstrably associatedwith certain neural circuits in the same way as other psycho-logical functions such as memory and language. Changes inthe feeling of embodiment can be affected by manipulatingvisual-somatosensory percepts that are represented withinthe TPJ, by stimulation of the TPJ, or by lesions in the TPJ[37, 78]. For example, the experimental manipulation ofvisuospatial perspective and simultaneous tactile stimulationcan cause the feeling of inhabiting a different body or bodyswapping [79]. Stimulation or lesion in the TPJ region cancause out-of-body experiences and other changes in thefeeling of embodiment [19]. For example, a patient (Susan)with right temporoparietal damage was reported who had a“Capgras syndrome for her mirror image” [80] (see pages73–75). This patient used sign language and was observedcommunicating with the person in the mirror (her ownreflection). She believed that there was this other “Susan” inthe mirror who had the same background, appearance, andage and went to the same school as she did (but she did notknow her in school). The other Susan was judged to be a bitslower and not as good at sign language as the real Susan.

The regions and representations that are used to repre-sent the self overlap with those that are used to representothers (see detailed discussion in [6, 37, 81]). The regionsthat are thought to compute the self-representation are alsoactivated by the perception of dynamic visual (gesture), audi-tory, or somatosensory stimulation from others, includingthe TPJ, somatosensory cortex, secondary somatosensorycortex, the extrastriate body area, the intraparietal sulcus,and ventral premotor regions [42, 82]. For example, thesecondary somatosensory area responds to touch regardlessof whether it was felt on one’s own skin or seen onanother [83]. Most of the functional components thathave been identified with self-representation in the TPJ aremodulated by auditory, visual, and tactile inputs and henceoveractivation of these channels would be expected to disturbthe self-representation [6].

Abnormal feelings of agency and of a presence couldalso occur as a result of a disturbance of the TPJ self-representation (reviewed in [5, 6]). The bulk of the experi-mental evidence shows that abnormal activity in TPJ regionsand vestibular cortex (insula/inferior parietal lobe) canproduce these phenomena [21]. Heautoscopy, the experienceof an alternate self or Doppelganger, is associated with eitherabnormal activity or damage in the TPJ [21]. Visual hallu-cinations of people are the second most prominent type ofhallucination in schizophrenia and may arise from abnormalactivation of the self-representation. When the body doubleis seen (visual hallucination), the visual appearance may ormay not mirror the person’s own image, but the imitation ofbody gestures by the double can produce the feeling that thedouble contains the real mind. The experience of the self canbe perceived to arise from the body double or can alternatebetween the physical body and the body double. This type


of hallucination can also be experienced as a feeling of apresence or multiple presences of closely projected doublesthat are not visible [21]. This feeling of a presence hasbeen induced by electrical stimulation of the PSTS region[20]. Aberrant activation of the self-representation wouldbe expected to interfere with feelings of embodiment andagency for one’s thoughts and actions as is seen in symptomssuch as delusions of control and could also produce thequalia (feeling of a presence) associated with delusions ofpersecution, delusions of jealousy, and delusions of referenceor visual hallucinations.

Several neuroimaging studies of schizophrenia haveshown that abnormal agency processing in patients is relatedto overactivation in (especially the right) IP or TPJ (reviewedin [5, 84]). For example, one study of agency processingfound abnormal activity in the right inferior parietal regionin schizophrenic subjects and this overactivation was alsopositively correlated with the Schneiderian score [85].

14. The Medial and Orbital Frontal Regions,Social Cognition, and Self-Representation

Impairments in self-awareness related to abnormal activityin midline structures have been hypothesized to play a rolein the social processing problems that have been identifiedin schizophrenia [86]. Several central and medial regionsof the brain, including those in the frontal lobe (medialorbital, the ventromedial prefrontal cortex, the anteriorcingulate cortex, dorsomedial prefrontal cortex), have beentheorized to be involved in self-referential processing andself-reflection [87, 88]. The evidence for this suppositioncomes mainly from neuroimaging studies where centralmidline regions have been found to show activity in taskssuch as the recall of personal information, the assessmentof self-personality traits, appearance, attitudes, or feelings[87, 88]. Gillihan and Farah (2005) [89], in a critical reviewof this literature, noted that confounding factors were notadequately controlled in many of these studies. A recentstudy used TMS to disrupt processing and to probe thebrain response to self-related stimuli with non-self-relatedstimuli. Superior performance on self-related items (the self-reference effect) was found. TMS to left and right parietalcortex suppressed the self reference effect, but no effectson the self-reference effect were found with TMS to themedial prefrontal lobe [90]. It was discussed previously howpatients with circumscribed and well-documented medialfrontal lesions do not present with a deficit in social situationprocessing nor have deficits in self-related processing beenreported [57, 58, 91]. Confabulation is known to occur withfrontal lobe damage [80]. The current framework assumesthat delusions are not confabulations but rather explanationsof bizarre sensory phenomena as suggested by BrendanMaher (2006) [92].

The dorsal anterior cingulate has previously been associ-ated with action monitoring or the monitoring of responseconflict and in cognitive control (see review in [93]). Thisfunction is thought to be used in social situations [87, 93, 94].However, there are now several reports of individuals with

circumscribed lesions of this area. These individuals have notbeen found to have deficits in tasks used to measure responseconflict such as the Stroop nor in behavioral measuresof cognitive control; see for example, [91]. In fact, recentevidence suggests that the dorsal anterior cingulate responseto errors may be more related to levels of negative affectduring task performance than to response conflict [95].These results are consistent with the view that the dorsalanterior cingulate plays a role in the control of autonomicresponses that accompany cognitive effort (see review in[91]). The dorsal anterior cingulate may be involved inresponse selection based on reward contingencies; data fromsingle unit studies in monkeys and in humans and furtherlesion data from humans undergoing cingulotomy supportthis hypothesis (see review in [91]). The dorsomedial aspectof the prefrontal cortex may have a role in empathy, inrepresenting emotional pain in others, and/or in the abilityto use emotional pain or negative consequences to constrainbehavior [96]. Although empathy and theory of mind scoresare correlated, the precise relationship between them isunclear [57]. A recent FMRI study found that the anteriormiddle cingulate cortex responded to both emotionally andphysically painful events, while the dorsomedial prefrontalcortex responded selectively to emotional pain [96]. A lackof empathy does not seem to be central to schizophrenia,although it is a symptom of the manic phase of bipolardisorder.

Orbital and ventral medial frontal lesions also do notcause theory of mind deficits [97] but rather result in anacquired sociopathy [98]. Orbital and ventral medial frontallobe damage has generally been reported to result in a lackof empathy, euphoria, irresponsibility, a lack of concernfor the future, and a lack of concern for social rules [97–99]. When tested explicitly on measures of social judgmentand responding, individuals with orbital and inferior medialfrontal lesions present with an intact knowledge of socialrules and the ability to judge the outcomes [98]. However,their autonomic responses to socially meaningful stimulimay be abnormal and it is has been hypothesized thatthese patients fail to activate somatic (body emotional)states in response to social stimuli [99]. Another reportsurmised that orbital damage results in a reduced ability touse the expectation of negative emotional reactions (anger)to constrain behavior [97]. Recently it has been reportedthat orbitofrontal lesions cause a diminished sensitivity tovarying reward magnitudes [100]. These data are consistentwith single unit recordings in monkeys showing neuronalresponses that were associated with various aspects ofreward computed from experience with behavioral choices[101]. It is claimed that damage to the orbital and ventralmedial prefrontal cortex leads to an inability to “developa coherent model of one’s own self” and hence emotionalliability and aberrant social functioning [87]. However, thefunction of orbital and ventral medial frontal cortices hasbeen repeatedly shown to be related to an inability to usenegative emotions to constrain behavior or to competentlycompute and use reward or punishment outcomes. Thissyndrome seems more closely aligned with mania wheredecision making is impaired and individuals often do risky


things without regard to the emotions of others (manicsymptoms include euphoria, lack of empathy, impulsiveness,lack of concern for consequences of behavior and for socialrules) and manic patients sometimes confabulate as is seenin patients with frontal lobe lesions [80].

In summary, medial prefrontal regions have been shownto be activated during tasks assessing theory of mind andself related processing; this presupposes a role for theseregions in social processing. Anterior cingulate regions showcoactivation with the TPJ in many neuroimaging studiesand hence abnormal activity in the TPJ could be expectedto affect activity in this region. However, as yet there aremany competing theories concerning the role of this partof the system and hence it is unclear what the functionalconsequences of abnormal activity would be and what formthey would take (overactivation, lack of activation).

15. Direct Evidence for TPJ Involvementin Schizophrenia

It is hypothesized that the TPJ (and primary efferent areas)should show slightly elevated baseline activity interposedwith epochs of hyperactivation that correspond to the abnor-mal qualia (hallucinations and delusions) and responsivedeficiencies (negative symptoms) seen in schizophrenia [5,6, 102]. Although it is not often acknowledged (perhapsbecause of the almost axiomatic belief in frontal lobeabnormalities), there is strong evidence for TPJ involvementin psychosis and schizophrenia. Activity has been recordedin the TPJ that corresponds to the experience of symptoms,activity has been found that is correlated with symptoms,and recent large-scale morphometric studies show volumereductions in the TPJ in schizophrenia. Torrey [103] wrotean extensive review of the evidence for inferior parietalinvolvement in schizophrenia that includes an account ofthe relationship between prodromal symptoms and parietalfunction (we will not reiterate that evidence here). TheTPJ region is a relatively anatomically variable region andhence volume abnormalities could be relatively difficultto detect; the inferior parietal region is one of the lastregions in the brain to be myelinated and develops intolate adolescence [104]. However, recently two large-scalemorphometric studies of schizophrenic volume reductionshave been published. Both of these studies show corereductions in TPJ volumes [105, 106]. Direct evidence forabnormal activity in the TPJ during psychosis comes frombrain recordings; right inferior parietal activity was related adelusional state in a magnetoencephalography (MEG) study[22]. When the abnormal activity subsided, the delusionssubsided. Correspondingly, interference with TPJ activityalleviates schizophrenic symptoms. At first TMS was appliedto the TPJ to treat auditory hallucinations; at least 6published studies have now shown that TMS applied to theTPJ alleviates auditory hallucinations and some reports showreductions in other symptoms as well [23–28]. For example,in one study schizophrenic subjects with treatment-resistantauditory hallucinations were given daily rTMS treatmentsfor 10 days and showed a reduction in the frequency of

auditory hallucinations and an improvement in many othersymptoms [23].

Data from our previously published studies have consis-tently supported TPJ involvement; schizophrenic symptomssuch as auditory hallucinations and thought disorder werecorrelated with levels of FMRI activity in the inferior parietaland superior temporal sulcus in several studies [107–109].One of the most systematic FMRI symptom capture studiesof schizophrenia was one of a single schizophrenic subjectwho heard voices for approximately 26 seconds and thenno voices for approximately 26 seconds; this periodicity isrelatively optimal for an FMRI design [110]. FMRI activationwas detected surrounding the PSTS approximately 3 secondsprior to the conscious detection of the auditory hallucina-tion. Activation of the inferior parietal and inferior frontalregions followed. The PSTS activation (which extended intothe superior and middle temporal regions) was persistentthrough the entire experience. This coincides with theunusual extended neural responses (e.g., 7 seconds) recordedin monkey STS neurons that are activated by dynamicmultimodal gestures [49].

TPJ over-activation could stem from a number of causes,but it is hypothesized that in schizophrenia the sourceis most often from over-activation of the hippocampalsystem [111]. The activity of the hippocampus shows ahigh association or correlation with activity in the TPJ (so-called default mode or cortical hub activity) and evidencefrom epileptic individuals shows that abnormal hippocampalactivity underlies active psychosis and that the TPJ regionis also involved [6]. Note that resting state abnormalitieshave been repeatedly found in schizophrenia and that aseed placed within the hippocampus will produce activity inmost of the “default mode” regions [112–115] (see [6] for adetailed discussion of hippocampal involvement).

16. Summary

In summary, the regions that make up the TPJ form a coresystem for the perception and production of emotional faceand body gestures as well as prosody. This region showsfunctional activity that indicates that it is preferentiallyinvolved in the perception of narrative (versus words orsentences) [3]. This area is a core component for theperception of social attention or of the process of jointattention (the deployment of attention according to socialand communicative cues). This area is also the core regionin the brain for theory of mind or understanding other’sintentions and thoughts. The neural activity in this system isbiased to detect or anticipate future multimodal social acts,or in other words to anticipate speech and actions. Inherentin the representation of multimodal gestures is a coding ofintention and agency. Even though this system is used for theperception and production of biological motion or gestures,there is a neural bias or tuning for the rapid identificationof social threat and parts of the TPJ are core componentsof fear perception (reviewed in [5, 6]). The TPJ not onlyrepresents other’s actions but may also be a core area forrepresenting the self. This is the only region of the brain that


produces verbal memory deficits when damaged and henceis the neural substrate for verbal working memory (reviewedin [5, 6]). Since neural activity in this region correspondsto social perception and joint attention, over-activation ofthis region could result in the erroneous perception ofbeing within a dynamic social situation. Over-activation ofthis region would have widespread consequences for manydomains including social perception, social reactivity, andattention.

When the consistency and weight of the evidence isconsidered, the characteristics of TPJ function more closelymatch the symptoms of schizophrenia. Hippocampal activityhas been consistently shown to be abnormal in schizophreniawhich is highly correlated with both TPJ and anterior cingu-late/paracingulate activity [113–115]. The characteristics ofmedial prefrontal function (especially from the lesion data)may match those of bipolar disorder more closely than thoseof schizophrenia.

Conflict of Interests

The authors declare that this paper was conducted in theabsence of any commercial or financial relationships thatcould be construed as a potential conflict of interests (6,2950-2967).

Acknowledgments

This work was supported by an NIMH Grant 1 R01MH067080-01A2 and by the Harvard Neuro-Discovery Cen-ter (formally HCNR). It was funded also by the BiomedicalInformatics Research Network (U24RR021992), NationalInstitute of Mental Health. The author would like to thankIsrael Molina for help with the paper.

References

[1] T. Allison, A. Puce, and G. McCarthy, “Social perceptionfrom visual cues: role of the STS region,” Trends in CognitiveSciences, vol. 4, no. 7, pp. 267–278, 2000.

[2] J. Decety and C. Lamm, “The role of the right temporopari-etal junction in social interaction: how low-level computa-tional processes contribute to meta-cognition,” Neuroscien-tist, vol. 13, no. 6, pp. 580–593, 2007.

[3] E. Redcay, “The superior temporal sulcus performs a com-mon function for social and speech perception: implicationsfor the emergence of autism,” Neuroscience and BiobehavioralReviews, vol. 32, no. 1, pp. 123–142, 2008.

[4] E. Redcay, D. Dodell-Feder, M. J. Pearrow et al., “Live face-to-face interaction during fMRI: a new tool for social cognitiveneuroscience,” NeuroImage, vol. 50, no. 4, pp. 1639–1647,2010.

[5] C. G. Wible, A. P. Preus, and R. Hashimoto, “A cognitiveneuroscience view of schizophrenic symptoms: abnormalactivation of a system for social perception and communica-tion,” Brain Imaging and Behavior, vol. 3, no. 1, pp. 85–110,2009.

[6] C. G. Wible, “Hippocampaltemporal-parietal junction inter-action in the production of psychotic symptoms: a frame-work for understanding the schizophrenic syndrome,” Inpress.

[7] J. Decety and J. Grezes, “The power of simulation: imaginingone’s own and other’s behavior,” Brain Research, vol. 1079,no. 1, pp. 4–14, 2006.

[8] J. Gottlieb, “From thought to action: the parietal cortex as abridge between perception, action, and cognition,” Neuron,vol. 53, no. 1, pp. 9–16, 2007.

[9] J. T. Klein, R. O. Deaner, and M. L. Platt, “Neural correlatesof social target value in macaque parietal cortex,” CurrentBiology, vol. 18, no. 6, pp. 419–424, 2008.

[10] K. G. Thompson and N. P. Bichot, “A visual salience map inthe primate frontal eye field,” Progress in Brain Research, vol.147, pp. 251–262, 2005.

[11] T. Brandt, P. Bartenstein, A. Janek, and M. Dieterich,“Reciprocal inhibitory visual-vestibular interaction. Visualmotion stimulation deactivates the parieto-insular vestibularcortex,” Brain, vol. 121, no. 9, pp. 1749–1758, 1998.

[12] A. D. Craig, “How do you feel—now? The anterior insula andhuman awareness,” Nature Reviews Neuroscience, vol. 10, no.1, pp. 59–70, 2009.

[13] D. C. Javitt, “When doors of perception close: bottom-upmodels of disrupted cognition in schizophrenia,” AnnualReview of Clinical Psychology, vol. 5, pp. 249–275, 2009.

[14] M. Brune, M. Abdel-Hamid, C. Sonntag, C. Lehmkamper,and R. Langdon, “Linking social cognition with socialinteraction: non-verbal expressivity, social competence and“mentalising” in patients with schizophrenia spectrum disor-ders,” Behavioral and Brain Functions, vol. 5, article 6, 2009.

[15] N. C. Andreasen, C. A. Calarge, and D. S. O’Leary, “Theoryof mind and schizophrenia: a positron emission tomographystudy of medication-free patients,” Schizophrenia Bulletin,vol. 35, no. 5, p. 1030, 2009.



[18] W. Penfield and P. Perot, “The brain’s record of auditory andvisual experience: a final summary and discussion,” Brain,vol. 86, no. 4, pp. 595–696, 1963.

[19] O. Blanke, S. Ortigue, T. Landis, and M. Seeck, “Stimulatingillusory own-body perceptions,” Nature, vol. 419, no. 6904,pp. 269–270, 2002.

[20] S. Arzy, M. Seeck, S. Ortigue, L. Spinelli, and O. Blanke,“Induction of an illusory shadow person,” Nature, vol. 443,no. 7109, article 287, 2006.

[21] P. Brugger, O. Blanke, M. Regard, D. T. Bradford, and T.Landis, “Polyopic heautoscopy: case report and review of theliterature,” Cortex, vol. 42, no. 5, pp. 666–674, 2006.

[22] R. Ishii, L. Canuet, M. Iwase et al., “Right parietal activationduring delusional state in episodic interictal psychosis ofepilepsy: a report of two cases,” Epilepsy and Behavior, vol.9, no. 2, pp. 367–372, 2006.

[23] S. H. Lee, W. Kim, Y. C. Chung et al., “A double blindstudy showing that two weeks of daily repetitive TMS overthe left or right temporoparietal cortex reduces symptomsin patients with schizophrenia who are having treatment-refractory auditory hallucinations,” Neuroscience Letters, vol.376, no. 3, pp. 177–181, 2005.


[24] E. Poulet, J. Brunelin, B. Bediou et al., “Slow transcranialmagnetic stimulation can rapidly reduce resistant auditoryhallucinations in schizophrenia,” Biological Psychiatry, vol.57, no. 2, pp. 188–191, 2005.

[25] R. E. Hoffman, R. Gueorguieva, K. A. Hawkins et al., “Tem-poroparietal transcranial magnetic stimulation for auditoryhallucinations: safety, efficacy and moderators in a fiftypatient sample,” Biological Psychiatry, vol. 58, no. 2, pp. 97–104, 2005.

[26] A. Vercammen, H. Knegtering, R. Bruggeman et al., “Effectsof bilateral repetitive transcranial magnetic stimulationon treatment resistant auditory-verbal hallucinations inschizophrenia: a randomized controlled trial,” SchizophreniaResearch, vol. 114, no. 1–3, pp. 172–179, 2009.

[27] O. Rosenberg, Y. Roth, M. Kotler, A. Zangen, and P. Dannon,“Deep transcranial magnetic stimulation for the treatmentof auditory hallucinations: a preliminary open-label study,”Annals of General Psychiatry, vol. 10, no. 1, article 3, 2011.

[28] R. Jardri, B. Lucas, Y. Delevoye-Turrell et al., “An 11-year-oldboy with drug-resistant schizophrenia treated with temporo-parietal rTMS,” Molecular Psychiatry, vol. 12, no. 4, article320, 2007.

[29] A. Caramazza, “On drawing inferences about the structureof normal cognitive systems from the analysis of patterns ofimpaired performance: the case for single-patient studies,”Brain and Cognition, vol. 5, no. 1, pp. 41–66, 1986.

[30] D. Silbersweig and E. Stern, “Functional neuroimaging ofhallucinations in schizophrenia: toward an integration ofbottom-up and top-down approaches,” Molecular Psychiatry,vol. 1, no. 5, pp. 367–375, 1996.

[31] J. Gauntlett-Gilbert and E. Kuipers, “Phenomenology ofvisual hallucinations in psychiatric conditions,” Journal ofNervous and Mental Disease, vol. 191, no. 3, pp. 203–205,2003.

[32] M. Startup and S. Startup, “On two kinds of delusion ofreference,” Psychiatry Research, vol. 137, no. 1-2, pp. 87–92,2005.

[33] D. I. Perrett, D. Xiao, N. E. Barraclough, C. Keysers, andM. W. Oram, “Seeing the future: natural image sequencesproduce anticipatory neuronal activity and bias perceptualreport,” Quarterly Journal of Experimental Psychology, vol. 62,no. 11, pp. 2081–2104, 2009.

[34] T. Jellema, C. I. Baker, B. Wicker, and D. I. Perrett, “Neuralrepresentation for the perception of the intentionality ofactions,” Brain and Cognition, vol. 44, no. 2, pp. 280–302,2000.

[35] J. P. Morris, K. A. Pelphrey, and G. McCarthy, “Perceivedcausality influences brain activity evoked by biologicalmotion,” Social Neuroscience, vol. 3, no. 1, pp. 16–25, 2008.

[36] D. I. Perrett, P. A. Smith, D. D. Potter et al., “Visual cells inthe temporal cortex sensitive to face view and gaze direction,”Proceedings of the Royal Society B, vol. 223, pp. 293–317, 1985.

[37] O. Blanke and S. Arzy, “The out-of-body experience: dis-turbed self-processing at the temporo-parietal junction,”Neuroscientist, vol. 11, no. 1, pp. 16–24, 2005.

[38] L. Nummenmaa, L. Passamonti, J. Rowe, A. D. Engell, andA. J. Calder, “Connectivity analysis reveals a cortical networkfor eye gaze perception,” Cerebral Cortex, vol. 20, no. 8, pp.1780–1787, 2010.

[39] L. Young, D. Dodell-Feder, and R. Saxe, “What gets the atten-tion of the temporo-parietal junction? An fMRI investigationof attention and theory of mind,” Neuropsychologia, vol. 48,no. 9, pp. 2658–2664, 2010.

[40] R. Adolphs, H. Damasio, and D. Tranel, “Neural systemsfor recognition of emotional prosody. A 3-D lesion study,”Emotion, vol. 2, no. 1, pp. 23–51, 2002.

[41] A. Puce and D. Perrett, “Electrophysiology and brain imagingof biological motion,” Philosophical Transactions of the RoyalSociety B, vol. 358, no. 1431, pp. 435–445, 2003.

[42] K. A. Pelphrey, J. P. Morris, C. R. Michelich, T. Allison,and G. McCarthy, “Functional anatomy of biological motionperception in posterior temporal cortex: an fMRI study ofeye, mouth and hand movements,” Cerebral Cortex, vol. 15,no. 12, pp. 1866–1876, 2005.

[43] J. Billino, D. I. Braun, K. D. Bohm, F. Bremmer, and K.R. Gegenfurtner, “Cortical networks for motion processing:effects of focal brain lesions on perception of differentmotion types,” Neuropsychologia, vol. 47, no. 10, pp. 2133–2144, 2009.

[44] E. D. Grossman, L. Battelli, and A. Pascual-Leone, “RepetitiveTMS over posterior STS disrupts perception of biologicalmotion,” Vision Research, vol. 45, no. 22, pp. 2847–2853,2005.

[45] C. Bruce, R. Desimone, and C. G. Gross, “Visual propertiesof neurons in a polysensory area in superior temporal sulcusof the macaque,” Journal of Neurophysiology, vol. 46, no. 2,pp. 369–384, 1981.

[46] M. I. Gobbini and J. V. Haxby, “Neural systems for recogni-tion of familiar faces,” Neuropsychologia, vol. 45, no. 1, pp.32–41, 2007.

[47] T. M. Wright, K. A. Pelphrey, T. Allison, M. J. McKeown,and G. McCarthy, “Polysensory interactions along lateraltemporal regions evoked by audiovisual speech,” CerebralCortex, vol. 13, no. 10, pp. 1034–1043, 2003.

[48] K. V. Kriegstein and A.-L. Giraud, “Distinct functionalsubstrates along the right superior temporal sulcus for theprocessing of voices,” NeuroImage, vol. 22, no. 2, pp. 948–955,2004.

[49] N. E. Barraclough, D. Xiao, C. I. Baker, M. W. Oram, and D.I. Perrett, “Integration of visual and auditory information bysuperior temporal sulcus neurons responsive to the sight ofactions,” Journal of Cognitive Neuroscience, vol. 17, no. 3, pp.377–391, 2005.

[50] M. S. Beauchamp, “Statistical criteria in fMRI studies ofmultisensory integration,” Neuroinformatics, vol. 3, no. 2, pp.93–113, 2005.

[51] M. S. Beauchamp, B. D. Argall, J. Bodurka, J. H. Duyn,and A. Martin, “Unraveling multisensory integration: patchyorganization within human STS multisensory cortex,” NatureNeuroscience, vol. 7, no. 11, pp. 1190–1192, 2004.

[52] M. Iacoboni, L. M. Koski, M. Brass et al., “Reafferent copiesof imitated actions in the right superior temporal cortex,”Proceedings of the National Academy of Sciences of the UnitedStates of America, vol. 98, no. 24, pp. 13995–13999, 2001.

[53] A. A. Ghazanfar, C. Chandrasekaran, and N. K. Logothetis,“Interactions between the superior temporal sulcus andauditory cortex mediate dynamic face/voice integration inrhesus monkeys,” Journal of Neuroscience, vol. 28, pp. 4457–4469, 2008.

[54] R. Saxe and A. Wexler, “Making sense of another mind: therole of the right temporo-parietal junction,” Neuropsycholo-gia, vol. 43, no. 10, pp. 1391–1399, 2005.

[55] J. Scholz, C. Triantafyllou, S. Whitfield-Gabrieli, E. N. Brown,and R. Saxe, “Distinct regions of right temporo-parietaljunction are selective for theory of mind and exogenousattention,” PLoS ONE, vol. 4, no. 3, Article ID e4869, 2009.


[56] C. D. Frith and U. Frith, “How we predict what other peopleare going to do,” Brain Research, vol. 1079, no. 1, pp. 36–46,2006.

[57] C. M. Bird, F. Castelli, O. Malik, U. Frith, and M. Husain,“The impact of extensive medial frontal lobe damage on“Theory of Mind” and cognition,” Brain, vol. 127, no. 4, pp.914–928, 2004.

[58] A. Baird, B. K. Dewar, H. Critchley, S. J. Gilbert, R. J.Dolan, and L. Cipolotti, “Cognitive functioning after medialfrontal lobe damage including the anterior cingulate cortex:a preliminary investigation,” Brain and Cognition, vol. 60, no.2, pp. 166–175, 2006.

[59] L. J. Bach, F. Happe, S. Fleminger, and J. Powell, “Theoryof mind: independence of executive function and the roleof the frontal cortex in acquired brain injury,” CognitiveNeuropsychiatry, vol. 5, no. 3, pp. 175–192, 2000.

[60] D. Samson, I. A. Apperly, C. Chiavarino, and G. W.Humphreys, “Left temporoparietal junction is necessary forrepresenting someone else’s belief,” Nature Neuroscience, vol.7, no. 5, pp. 499–500, 2004.

[61] L. Young, F. Cushman, M. Hauser, and R. Saxe, “The neuralbasis of the interaction between theory of mind and moraljudgment,” Proceedings of the National Academy of SciencesUnited States, vol. 104, no. 20, pp. 8235–8240, 2007.

[62] Y. Sarfati, M. C. Hardy-Bayle, C. Besche, and D. Wid-locher, “Attribution of intentions to others in people withschizophrenia: a nonverbal exploration with comic strips,”Schizophrenia Research, vol. 25, no. 3, pp. 199–209, 1997.

[63] Y. Sarfati, M. C. Hardy-Bayle, E. Brunet, and D. Wid-locher, “Investigating theory of mind in schizophrenia: influ-ence of verbalization in disorganized and non-disorganizedpatients,” Schizophrenia Research, vol. 37, no. 2, pp. 183–190,1999.

[64] U. Dimberg, M. Thunberg, and K. Elmehed, “Unconsciousfacial reactions to emotional facial expressions,” PsychologicalScience, vol. 11, no. 1, pp. 86–89, 2000.

[65] L. Carr, M. Iacoboni, M. C. Dubeaut, J. C. Mazziotta, andG. L. Lenzi, “Neural mechanisms of empathy in humans:a relay from neural systems for imitation to limbic areas,”Proceedings of the National Academy of Sciences of the UnitedStates of America, vol. 100, no. 9, pp. 5497–5502, 2003.

[66] J. Narumoto, T. Okada, N. Sadato, K. Fukui, and Y. Yonekura,“Attention to emotion modulates fMRI activity in humanright superior temporal sulcus,” Cognitive Brain Research, vol.12, no. 2, pp. 225–231, 2001.

[67] D. I. Leitman, D. H. Wolf, J. D. Ragland et al., “‘It’s notwhat you say, but how you say it’: a reciprocal temporo-frontal network for affective prosody,” Frontiers in HumanNeuroscience, vol. 4, article 19, 2010.

[68] S. Baron-Cohen, S. Wheelwright, and T. Jolliffe, “Is there a“language of the eyes”? Evidence from normal adults, andadults with autism or Asperger Syndrome,” Visual Cognition,vol. 4, no. 3, pp. 311–331, 1997.

[69] M.V. Lombardo, B. Chakrabarti, E.T. Bullmore et al., “Sharedneural circuits for mentalizing about the self and others,”Journal of Cognitive Neuroscience, vol. 22, no. 7, pp. 1623–1635, 2010.

[70] T. Akiyama, M. Kato, T. Muramatsu, F. Saito, S. Umeda,and H. Kashima, “Gaze but not arrows: a dissociativeimpairment after right superior temporal gyrus damage,”Neuropsychologia, vol. 44, no. 10, pp. 1804–1810, 2006.

[71] N. Boddaert, N. Chabane, H. Gervais et al., “Superior tem-poral sulcus anatomical abnormalities in childhood autism: a

voxel-based morphometry MRI study,” NeuroImage, vol. 23,no. 1, pp. 364–369, 2004.

[72] K. A. Pelphrey, R. J. Viola, and G. McCarthy, “When strangerspass: processing of mutual and averted social gaze in thesuperior temporal sulcus,” Psychological Science, vol. 15, no.9, pp. 598–603, 2004.

[73] R. Saxe, D. K. Xiao, G. Kovacs, D. I. Perrett, and N. Kan-wisher, “A region of right posterior superior temporal sulcusresponds to observed intentional actions,” Neuropsychologia,vol. 42, no. 11, pp. 1435–1446, 2004.

[74] J. Schultz, H. Imamizu, M. Kawato, and C. D. Frith,“Activation of the human superior temporal gyrus duringobservation of goal attribution by intentional objects,”Journal of Cognitive Neuroscience, vol. 16, no. 10, pp. 1695–1705, 2004.

[75] A. P. Atkinson, A. S. Heberlein, and R. Adolphs, “Sparedability to recognise fear from static and moving whole-bodycues following bilateral amygdala damage,” Neuropsychologia,vol. 45, no. 12, pp. 2772–2782, 2007.

[76] C. Cristinzio, K. N’Diaye, M. Seeck, P. Vuilleumier, and D.Sander, “Integration of gaze direction and facial expressionin patients with unilateral amygdala damage,” Brain, vol. 133,no. 1, pp. 248–261, 2010.

[77] J. Grezes, S. Pichon, and B. De Gelder, “Perceiving fear indynamic body expressions,” NeuroImage, vol. 35, no. 2, pp.959–967, 2007.

[78] E. L. Altschuler and V. S. Ramachandran, “Last but not least:a simple method to stand outside oneself,” Perception, vol. 36,no. 4, pp. 632–634, 2007.

[79] V. I. Petkova and H. H. Ehrsson, “If I were you: perceptualillusion of body swapping,” PLoS ONE, vol. 3, no. 12, ArticleID e3832, 2008.

[80] T. E. Feinberg, Altered Egos: How the Brain Creates the Self,Oxford University Press, 2001.

[81] M. Tsakiris, “My body in the brain: a neurocognitive modelof body-ownership,” Neuropsychologia, vol. 48, no. 3, pp.703–712, 2010.

[82] K. J. Wheaton, J. C. Thompson, A. Syngeniotis, D. F.Abbott, and A. Puce, “Viewing the motion of human bodyparts activates different regions of premotor, temporal, andparietal cortex,” NeuroImage, vol. 22, no. 1, pp. 277–288,2004.

[83] C. Keysers, B. Wicker, V. Gazzola, J. L. Anton, L. Fogassi, andV. Gallese, “A touching sight: SII/PV activation during theobservation and experience of touch,” Neuron, vol. 42, no.2, pp. 335–346, 2004.

[84] S. A. Spence, D. J. Brooks, S. R. Hirsch, P. F. Liddle, J. Meehan,and P. M. Grasby, “A PET study of voluntary movementin schizophrenic patients experiencing passivity phenomena(delusions of alien control),” Brain, vol. 120, no. 11, pp.1997–2011, 1997.

[85] C. Farrer, N. Franck, C. D. Frith et al., “Neural correlates ofaction attribution in schizophrenia,” Psychiatry Research, vol.131, no. 1, pp. 31–44, 2004.

[86] D. J. Holt, B. S. Cassidy, J. R. Andrews-Hanna et al.,“An anterior-to-posterior shift in midline cortical activityinschizophrenia during self-reflection,” Biological Psychiatry,vol. 69, no. 5, pp. 415–423, 2011.

[87] G. Northoff, A. Heinzel, M. de Greck, F. Bermpohl, H.Dobrowolny, and J. Panksepp, “Self-referential processing inour brain—a meta-analysis of imaging studies on the self,”NeuroImage, vol. 31, no. 1, pp. 440–457, 2006.

[88] L. T. Rameson, A. B. Satpute, and M. D. Lieberman,“The neural correlates of implicit and explicit self-relevantprocessing,” NeuroImage, vol. 50, no. 2, pp. 701–708, 2010.


[89] S. J. Gillihan and M. J. Farah, “Is self special? A critical reviewof evidence from experimental psychology and cognitiveneuroscience,” Psychological Bulletin, vol. 131, no. 1, pp. 76–97, 2005.

[90] H. C. Lou, B. Luber, A. Stanford, and S. H. Lisanby, “Self-specific processing in the default network: a single-pulse TMSstudy,” Experimental Brain Research, vol. 207, no. 1-2, pp. 27–38, 2010.

[91] L. K. Fellows and M. J. Farah, “Is anterior cingulate cortexnecessary for cognitive control?” Brain, vol. 128, no. 4, pp.788–796, 2005.

[92] B. A. Maher, “The relationship between delusions andhallucinations,” Current Psychiatry Reports, vol. 8, no. 3, pp.179–183, 2006.

[93] D. M. Amodio and C. D. Frith, “Meeting of minds: themedial frontal cortex and social cognition,” Nature ReviewsNeuroscience, vol. 7, no. 4, pp. 268–277, 2006.

[94] G. Northoff, “Self and brain: what is self-related processing?”Trends Cognitive Science, vol. 15, no. 5, pp. 186–187, 2011,Author reply 187–188.

[95] R. P. Spunt and M. D. Lieberman, “An integrative modelof the neural systems supporting the comprehension ofobserved emotional behavior,” NeuroImage, vol. 59, no. 3, pp.3050–3059, 2012.

[96] E. G. Bruneau, A. Pluta, and R. Saxe, “Distinct roles of the“Shared Pain” and “Theory of Mind” networks in processingothers’ emotional suffering,” Neuropsychologia, vol. 50, no. 2,pp. 219–231, 2012.

[97] R. J. Blair and L. Cipolotti, “Impaired social responsereversal. A case of ’acquired sociopathy’,” Brain, vol. 123, no.6, pp. 1122–1141, 2000.

[98] J. L. Saver and A. R. Damasio, “Preserved access and process-ing of social knowledge in a patient with acquired sociopathydue to ventromedial frontal damage,” Neuropsychologia, vol.29, no. 12, pp. 1241–1249, 1991.

[99] A. R. Damasio, D. Tranel, and H. Damasio, “Individualswith sociopathic behavior caused by frontal damage fail torespond autonomically to social stimuli,” Behavioural BrainResearch, vol. 41, no. 2, pp. 81–94, 1990.

[100] J. Kim, E. Yang, L. Overhalser et al., “Orbitofrontal cortexlesions disrupt adaptation of choice-behavior to varyingaffective context in economic decision-making,” In press.

[101] J. W. Kable and P. W. Glimcher, “The neurobiology ofdecision: consensus and controversy,” Neuron, vol. 63, no. 6,pp. 733–745, 2009.

[102] D. H. Ffytche, R. J. Howard, M. J. Brammer, A. David,P. Woodruff, and S. Williams, “The anatomy of consciousvision: an fMRI study of visual hallucinations,” NatureNeuroscience, vol. 1, no. 8, pp. 738–742, 1998.

[103] E. F. Torrey, “Schizophrenia and the inferior parietal lobule,”Schizophrenia Research, vol. 97, no. 1–3, pp. 215–225, 2007.

[104] N. Geschwind, “Disconnexion syndromes in animals andman,” Brain, vol. 88, no. 2, pp. 237–294, 1965.

[105] R. A. Honea, A. Meyer-Lindenberg, K. B. Hobbs et al.,“Is gray matter volume an intermediate phenotype forschizophrenia? A voxel-based morphometry study of patientswith schizophrenia and their healthy siblings,” BiologicalPsychiatry, vol. 63, no. 5, pp. 465–474, 2008.

[106] J. M. Segall, J. A. Turner, T. G. Van Erp et al., “Voxel-basedmorphometric multisite collaborative study on schizophre-nia,” Schizophrenia Bulletin, vol. 35, no. 1, pp. 82–95, 2009.

[107] S. D. Han, P. G. Nestor, M. Hale-Spencer et al., “Functionalneuroimaging of word priming in males with chronicschizophrenia,” NeuroImage, vol. 35, no. 1, pp. 273–282,2007.

[108] C. G. Wible, K. Lee, I. Molina et al., “FMRI activitycorrelated with auditory hallucinations during performanceof a working memory task: data from the FBIRN consortiumstudy,” Schizophrenia Bulletin, vol. 35, no. 1, pp. 47–57, 2009.

[109] R. I. Hashimoto, K. Lee, A. Preus, R. W. McCarley, and C.G. Wible, “An fMRI study of functional abnormalities inthe verbal working memory system and the relationship toclinical symptoms in chronic schizophrenia,” Cerebral Cortex,vol. 20, no. 1, pp. 46–60, 2010.

[110] B. R. Lennox, S. Bert, G. Park, P. B. Jones, and P. G. Morris,“Spatial and temporal mapping of neural activity associatedwith auditory hallucinations,” The Lancet, vol. 353, no. 9153,p. 644, 1999.

[111] S. A. Schobel, N. M. Lewandowski, C. M. Corcoran etal., “Differential targeting of the CA1 subfield of the hip-pocampal formation by schizophrenia and related psychoticdisorders,” Archives of General Psychiatry, vol. 66, no. 9, pp.938–946, 2009.

[112] A. G. Garrity, G. D. Pearlson, K. McKiernan, D. Lloyd,K. A. Kiehl, and V. D. Calhoun, “Aberrant “default mode”functional connectivity in schizophrenia,” American Journalof Psychiatry, vol. 164, no. 3, pp. 450–457, 2007.

[113] J. L. Vincent, A. Z. Snyder, M. D. Fox et al., “Coherent spon-taneous activity identifies a hippocampal-parietal memorynetwork,” Journal of Neurophysiology, vol. 96, no. 6, pp. 3517–3531, 2006.

[114] R. L. Buckner, J. Sepulcre, T. Talukdar et al., “Corticalhubs revealed by intrinsic functional connectivity: mapping,assessment of stability, and relation to Alzheimer’s disease,”Journal of Neuroscience, vol. 29, no. 6, pp. 1860–1873, 2009.

[115] M. Focking, P. Dicker, J. A. English, K. O. Schubert, M.J. Dunn, and D. R. Cotter, “Commonproteomic changesin the hippocampus in schizophrenia and bipolar disorderand particular evidence for involvement of cornu ammonisregions 2 and 3,” Archives of General Psychiatry, vol. 68, no. 5,pp. 477–488, 2011.


Research Article

Faulty Suppression of Irrelevant Material inPatients with Thought Disorder Linked to AttenuatedFrontotemporal Activation

S. M. Arcuri,1, 2 M. R. Broome,1, 3 V. Giampietro,4 E. Amaro Jr.,2 T. T. J. Kircher,5

S. C. R. Williams,4 C. M. Andrew,4 M. Brammer,4 R. G. Morris,6 and P. K. McGuire1

1 Department of Psychosis Studies, Institute of Psychiatry, King’s College London, London, UK2 NIF, Functional Neuroimaging, Lab/LIM 44, Instituto de Radiologia, University of Sao Paulo, Sao Paulo, Brazil3 Warwick Medical School, University of Warwick, Coventry UK and Psychosis Clinical Academic Group, Coventry, UK4 Department of Neuroimaging, Institute of Psychiatry, King’s College London, London, UK5 Marburg University, Marburg, Germany6 Department of Psychology, Institute of Psychiatry, King’s College London, London, UK

Correspondence should be addressed to S. M. Arcuri, [email protected]

Received 15 November 2011; Accepted 23 December 2011

Academic Editor: Christoph Mulert

Copyright © 2012 S. M. Arcuri et al. This is an open access article distributed under the Creative Commons Attribution License,which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited.

Formal thought disorder is a feature schizophrenia that manifests as disorganized, incoherent speech, and is associated with a poorclinical outcome. The neurocognitive basis of this symptom is unclear but it is thought to involve an impairment in semanticprocessing classically described as a loosening of meaningful associations. Using a paradigm derived from the n400 event-related,potential, we examined the extent to which regional activation during semantic processing is altered in schizophrenic patients withformal thought disorder. Ten healthy control and 18 schizophrenic participants (9 with and 9 without formal thought disorder)performed a semantic decision sentence task during an event-related functional magnetic resonance imaging experiment. Weemployed analysis of variance to estimate the main effects of semantic congruency and groups on activation and specific effectsof formal thought disorder were addressed using post-hoc comparisons. We found that the frontotemporal network, normallyengaged by a semantic decision task, was underactivated in schizophrenia, particularly in patients with FTD. This network isimplicated in the inhibition of automatically primed stimuli and impairment of its function interferes with language processingand contributes to the production of incoherent speech.

1. Introduction

Bleuler’s work in psychosis continues to be highly influentialin furthering understanding of the signs and symptoms ofschizophrenia [1]. Nevertheless, one of his primary concep-tual contributions in understanding schizophrenia, “distur-bance of associations” [2, 3], remains to be explained interms of underlying neural basis. In turn, Bleuler’s ideas wereinfluenced by Jung’s word association task [4]. Regardless ofthe psychological or affective mechanisms that may influencethe production of speech, Jungian word association is bynature a semantic association test. Interpretations of thebroader meanings of “split mind” and “association”, arising

from the psychoanalytical field, are not incompatible withan inbuilt characteristic of this test, which taps into theconcept of semantic priming [5], extensively investigatedin schizophrenia (see Minzenberg et al. [6] for a review insingle-word semantic priming in schizophrenia). However,some studies do not take into account specific symptomsproposed by Bleuler, such as formal thought disorder, as anunderlying factor that would interfere with task performance(e.g., [7]), mixing up performance of patients with a rangeof distinct symptoms. Of note, there is a line of investigationsuggesting that formal thought disorder (FTD) is associatedwith hyperactivation of the semantic network [8–14]. Thesestudies mainly employ semantic tasks based on word-pair


semantic priming paradigms. It is important to note thatsuch kind of studies, although provided evidence for alteredsemantic processing associated with FTD, may fail to tapinto the type of communication impairment clinically seenin FTD patients. The latter manifests itself in linguistic unitsthat range from large thoughts, such as whole periods withseveral sentences embedded, to small utterances, such aswords within the same sentence [15], or even a single word(e.g., neologism).

Bleuler also suggested that, within its symptoms, FTD ispossibly the closest to the neural substrate of schizophrenia[2]. Biological evidence for a neural basis of FTD has beenprovided by methods employing functional measures ofbrain activity. EEG studies employing the n400, an event-related potential (ERP) that is a neural marker of semanticprocessing [16–18], demonstrated a reduction in n400 am-plitudes and increased latency in schizophrenia during a sen-tence processing task [19–21], consistent with impairment insemantic integration, particularly associated with FTD [22–26].

Functional neuroimaging of schizophrenia suggestedthat FTD is associated with altered resting activity in themedial temporal cortex [27] and with altered activity (brainmetabolism or Blood Oxygen Level Dependent-BOLD-activity) during tasks that engage language processing inthe dorsolateral prefrontal, bilateral anterior cingulate, andlateral temporal cortices [28–32]). FTD has also been linkedto reductions in grey matter volume in the inferior frontal[33, 34] and the temporo-parietal cortex [35]. Using a modelof information processing, MacDonalds III and collaboratorsdemonstrated diminished activity in the middle frontal gyrusin schizophrenic patients with disorganization symptoms(includes FTD) associated with context processing deficits[31]. Collectively, the studies particularly implicate theprefrontal, middle frontal, temporal, and anterior cingulatecortices in the pathophysiology of FTD in schizophrenia.

In summary, FTD is a symptom of schizophreniathat manifests itself as disorganized, incoherent speech.Andreasen [36] has operationalized studying this symptomby creating a scale that decomposes the concept of “looseningof meaningful associations” into measurable items. Thus,this clinical manifestation of schizophrenia, which can beobjectively measured, is thought to involve impairment insemantic processing, deficits in executive functioning, andaltered brain activity in the left frontal, temporal, andanterior cingulate cortices.

The objective of the present study was to use event-related fMRI to examine the extent to which regional activa-tion during semantic processing is altered in schizophrenicpatients with FTD. Methodological issues which were nottackled by some of the functional imaging studies mentionedabove (small sample sizes, block design fMRI task and anovert production of speech) were carefully addressed. Par-ticularly, we devised a task likely to engage areas associatedwith linguistic processes with heavier demands [37] using aparadigm derived from semantic decision procedures usedin the ERP studies that originally correlated FTD with n400abnormalities in schizophrenia. We sought to engage thetop down modulation of semantic processing by requiring

suppression or inhibitory mechanisms to take place, neces-sary to process semantic incongruent information. Activa-tion of the left inferior frontal and left middle temporal [38,39], right anterior cingulate [38], and bilateral precuneus[39] was found in studies that required the generation ofsemantically incongruent endings to complete a previouslyprimed incomplete sentence stem. Therefore, these areas areactivated in tasks requiring suppression of endings (mean-ings) automatically activated by sentences.

We hypothesized the following:

(1) Incongruent sentences would engage left inferiorfrontal, left middle temporal, right anterior cingulate,and bilateral precuneus more than congruent sen-tences.

(2) FTD patients with schizophrenia would showdiminished activation relative to controls andpatients without FTD in:

(a) areas previously demonstrated to show aberrantactivity in this group: left prefrontal (middleand inferior), left temporal, and bilateral ante-rior cingulate cortex and

(b) areas associated with manipulation of incon-gruent material (i.e., the above and also theprecuneus).

2. Methods

2.1. Subjects. Ten healthy adult volunteers and 18 patientsmeeting DSM-IV criteria for schizophrenia [40] took part inthe study. Acutely psychotic patients with either high or lowlevels of positive FTD were recruited from the South Londonand Maudsley NHS Trust. Controls were recruited from thesame geographical area through local advertisement.

All subjects were dextral [41], males with National AdultReading Test Revised [42] IQ ≥ 80, and native speakersof British English. Exclusion criteria for controls were aprevious history of a neurological or psychiatric disorder,substance dependence, or a medical disorder that couldaffect the brain. Exclusion criteria for patient were anotherDSM-IV axis I diagnosis and age of onset of schizophreniaprior to 18 years of age. Subjects received oral and writteninformation about the procedures and gave written consentto participate, with £30 in return for participating. Theproject was approved by the Research Ethical Committee ofthe Institute of Psychiatry.

Psychopathology was assessed using the Scale for theAssessment of Positive Symptoms (SAPS) [36] and the Scalefor the Assessment of Negative Symptoms (SANS) [43].Positive FTD was assessed on the basis of the corresponding8 items in the SAPS (derailment; incoherence; illogicality;circumstantiality; pressure of speech; distractible speech;tangentiality; clanging).

In each patient, the score on each FTD SAPS item (range0–5) was summed to yield a total FTD score (range 0–40).The FTD score was then used both as a classifier to splitthe patients sample in two subgroups (as described below)and as a continuous measure (see Figure 2(b)), without


distinctions within the patients. Thus, patients who hada FTD score 0–4 were classified as non-FTD and thosewith scores ranging from 5–40 were classified as FTD. Thepatients sample was split into two groups on this basisproducing a subgroup with (FTD, n = 9 and a subgroupwithout FTD (Non-FTD, n = 9). There were no significantdifferences between these subgroup in SAPS total score,SANS total score, scale for the Assessment of Global Func-tioning (GAF) [40] score, antipsychotic medication dose (inchlorpromazine equivalent), duration of illness, number ofpsychiatric admissions, or age at first admission (Table 1).There were also no differences in the score for negative FTD,as defined by the SANS items (poverty of speech, poverty ofcontent of speech, thought blocking, and increased latency ofresponse).

The two patient subgroups and the healthy controls hadsimilar demographic characteristics (Table 2).

2.2. Task

2.2.1. Sentence Stems. Eighty sentence stems (i.e., last wordmissing) were used with the degree of constraint previouslydefined in a large sample drawn from the same populationas the control subjects in the present study [44]. Forty stemshad a high semantic constraint (HCt) (cloze probability >0.94) and 40 low semantic constraint (LCt) (cloze probability< 0.34). For example, “He posted the letter without a. . ..”,is a HCt stem (completed by the word “stamp” by 96% ofsubjects), while “She couldn’t imagine anyone less. . .” is a LCtstem.

2.2.2. Target Words. Two types of words were presented astarget stimuli, taken from completions that had originallybeen produced for the stems [44]. These words wereeither congruous with the sentence stem (i.e., semanticallyrelated) or incongruous (semantically unrelated). To avoidphonological priming [45], words with the same initialphoneme as the most frequent congruent completion (bestcompletion) for that stem were excluded.

2.3. Experimental Paradigm

2.3.1. fMRI Stimuli. Sentence stems were presented onto ascreen subtending a visual angle of approximately 1◦ (textheight) by 7◦, (stem length). Stems were presented for 2.5 sec.After an interval of 0,7 sec during which the screen was blank,the target word appeared (Figure 1). During the inter stimu-lus interval, subjects were instructed to fixate on an asteriskfor at least 13.7 sec. Eighty trials of 20.4 seconds durationwere presented, divided into 5 runs of 16 trials each. Trialtype per run was pseudorandomized to control for ordereffects between subjects. Subjects were instructed to readthe sentence stem, then decide if the target completed thestem in a sensible way or not making their choice using oneof two buttons on a button box (for accuracy and responsetime recordings). Prior to scanning, all subjects underwent atraining session to make sure they were able to understandand perform the task.

2.3.2. fMRI Data Acquisition. Gradient-echo planar MRimages were acquired using a quadrature head coil in a 1.5Tesla GE Signa System (General Electric, Milwaukee, WS,USA). Head movement was minimised by foam paddingand a supporting band across the forehead. A gradient echoEPI axial acquisition (TR 1700 milliseconds-ms, TE 40 ms,FA 90◦, matrix 64 × 64, FOV 24 cm, thickness 7 mm, gap0.7 mm, 192 volumes) was used to collect 12 slices parallelto the intercommissural (AC-PC) plane. The total numberof images acquired (in 5 runs) was 960 with just underfive minutes between them. Structural images were acquiredusing gradient echo IR EPI sequence (TR 1600 ms, TE 40 ms,TI 180 ms, matrix 128 × 128, FOV 20 cm, thickness 3 mm,gap 0.3 mm, NEX 8, 43 slices). This latter image was used fornormalisation to a standard template.

2.4. Analysis. This experiment is part of a larger study [46]in which a series of experiments were designed to test thesemantic network and executive functions in FTD patients.The present experiment manipulated semantic constraintsand congruencies. In this paper, we only addressed effects ofsemantic congruency.

2.5. Online Behavioral Data. Repeated Measures (SPSS,General Linear Model) analyses were performed for accuracyscores and reaction times (RT) using semantic congruencyas within-subject factors. A “yes” response for a semanticallyincongruent word or “no” response to a congruent word wasconsidered inaccurate. The percentage of correct responsesacross trials of the same condition was used as within-subjectfactor.

2.6. Neuroimaging Data. Imaging data were analysed withxbam v3.4, Department of Biostatistics and Computing,Institute of Psychiatry, King’s College London [47]. Moredetails about the software can be obtained at http://www.brainmap.it/.

2.6.1. Movement Estimation and Correction. The mean signalintensity over all time points (960) was averaged to create atarget image (average image intensity at every voxel across alltime points) and the sum of absolute differences in grey scalevalues between the voxels of each observed image (at eachtime point) and its corresponding base image was computed.A rigid body registration search algorithm was used toestimate the extent of translation and rotation (3 rotations +3 translations), minimising the total difference betweenmatch and base images. The match images were realignedrelative to the base image by tricubic spline interpolation andthe realigned T2∗-weighted time series were regressed on theconcomitant and lagged time series of estimated movementat each voxel [48]. Residual time series resulting from the laststage of this procedure were thus uncorrelated with estimatedrigid motion in 3D.

2.6.2. Data Analysis. A nonparametric procedure [49] wasadopted that avoids assumptions about the distribution ofstatistics under the null hypothesis. First, the experimental


Table 1: Clinical characteristics of the Patient Groups.

Group Non-FTD FTD P value

Positive FTD 0 (1) 0–2 17 (11) 5–20 <0.001a

Negative FTD 5 (7) 0–9 6 (6) 3–14 0.546a

SAPS (sum) 23 (16) 8–50 14 (10) 6–26 0.471b

SANS (sum) 19 (9) 1–34 27 (13) 11–52 0.588b

GAF 31 (13) 21–51 30 (14) 20–40 0.546a

Chlorpromazine equivalent in (mg/day) 700 (967) 133–1800 517 (550) 320–1200 0.772a

Duration of illness (in years) 6 (12) 0.5–17 11 (14) 8–28 0.190a

Hospital admissions

Number 2 (4) 1–15 6 (9) 1–20 0.142a

Age at 1st 25 (9) 19–60 26 (11) 15–31 0.222a

aMann-Whitney [median (interquartile range) min and max], bANOVA [Mean (SD) and range], cFischer Exact.

Table 2: Demographic Description of the Sample.

GroupsAGE Median (interquartile range);

min and maxNart IQ Mean (SD); range

Years of full-time education median(interquartile range); min and max

Controls 35 (10); 24–54 108.3 (13.5); 89–126 14.3–(5.3); 9–24

NON-TD 38 (21); 24–63 101.4 (12.7); 86–122 11.0 (1.5)∗; 9–15

TD 33 (16); 23–55 99.2 (10.7); 82–115 11.0 (3)∗; 9–12

Analysis results P = 0.222a P = 0.267b P = 0.029a

aKruskal Wallis, bone-way ANOVA, ∗no difference between TD and non-TD patients for years of education.

design was convolved with two poisson functions chosento model the haemodynamic delays of 4 and 8 seconds(see Figure 1). The best fit (weighted sum) of these twoconvolutions to the observed time series at each voxel wasthen computed by least squares analysis. The use of the twoconvolutions within the experimental design allows the timedelay between stimulus onset and peak bold response to varybetween these limits, which encompass the normal rangeof haemodynamic delays. Using the parameters of the leastsquares model fit, the sums of squares of deviations fromthe mean image intensity over the whole time series dueto the model and the residuals were calculated. The ratioof these sums of squares of residuals (model/residuals) wasthen calculated (and called SSQ ratio). The appropriate nulldistribution of this statistic was obtained by repeating themodel fitting procedure after wavelet-based permutation ofthe time series [47] twenty times at each voxel and combiningthe data over all voxels. For any desired type I error rate,the appropriate critical threshold value of SSQ ratio could beobtained from this distribution. Any value of SSQ ratio thatexceeded this threshold was deemed to indicate the presenceof a voxel responding to the experimental paradigm.

In order to construct group images, the observed and“null” SSQ ratio maps for each experimental conditionwere then transformed into Talairach space [50]. Medianactivation maps for each condition were computed aftersmoothing with a Gaussian filter (FWHM 7.2 mm). Threemaps were produced: (1) condition “b” versus “a”, (2)condition “c” versus “a”, and (3) condition “b” versus “c”,

where “a” means baseline and “b” and “c” the experimentalconditions tested.

The main effect of semantic congruency was assessedby contrasting the 40 congruent trials (20 low congruent +20 high congruent) and 40 incongruent trials (20 lowincongruent + 20 high congruent) [conditions “b” and “c”maps]. In order to probe the analysis to the main cognitiveprocess under investigation, we chose the 3rd TR in an event-related design, in which the presentation both of the sentencestem and the final word has already taken place, as shown inFigure 1.

Differences in the responses between conditions wereobtained at each voxel by regression of the linear model F =a0 + a1P + e, where the vector of responses in all subjects ata voxel, P is a dummy coding vector expressing the particularcontrast between experimental conditions of interest, ande is the vector of random errors. Values of a∗1 were testedfor statistical significance by randomly reallocating the dataof each voxel between conditions, thus realising the nulldistribution under the hypothesis of no difference betweenresponses to different experimental conditions. The analysesreported in this study were carried out using cluster-levelstatistics to avoid harsh multiple comparison correctionsrequired if each voxel is tested individually [51]. Initially,maps of a∗1 were thresholded retaining only clusters with aprobability of type I error = 0.0001, for main effects of thetask in healthy controls. However, the patients group showeda diminished overall strength of activation relative to healthycontrols. We thus decided to report results from mapsthresholded at a voxel/cluster probability of type I error =


Sentencepresentation:

4 s 8 s

Word presentation:

Max 3.5 s

Star presentation:

13.7 s

Stamp ∗

2.5 s Gap

0.7

s

1 TR = repetition time = acquisition of 1 brain volume (1.7)

1 rial = 12 TRS or 12 x acquisition of 1 brain volume (with 12 slices)

= 1 s1.7 3.4 5.1 6.8 8.5 10.2 11.9 13.6 15.3 17 18.7 20.4

= 1 brain slice

He posted the letter

without a

t

Figure 1: fMRI task design.

0.05/0.01, adjusted to get less than 1 false positive per map,to facilitate interpretation of the results, particularly thedifferences found between controls and patients and betweenpatients themselves (FTD and Non-FTD). The integral of theSSQ ratio values for each of the resulting three-dimensionalclusters was then tested against the distribution of clusterSSQ ratio integrals generated after random reallocation ofdata between conditions at each voxel (see above). Mapsreflecting the contrast between conditions (i.e., congruentversus incongruent) in each group were analysed withANOVAS, producing 2 maps, one showing regions moreactivated in the first condition and the other showing thereverse.

To minimize the potentially confounding effect of differ-ences in task performance on brain activation, all incorrecttrials were excluded from the image analysis. Thus theneuroimaging data were derived from trials in which subjectscorrectly decided whether the final word was or not seman-tically congruent with the previous sentence stem.

2.6.3. Data Analyses to Compare Groups. To investigate be-tween group differences due to effects of semantic congruen-cy, the images resulting from the contrast between semanticcongruent versus incongruent trials were directly comparedbetween groups, first contrasting all groups in a 3-GroupANOVA with cluster probability of type I error set at 0.01.This analysis produced a range of maps in which one groupshowed greater activation than the other groups and vice-versa. These maps are presented in the results section below(Figure 2 and Table 3).

Further analyses were conducted to investigate the natureof the results arising from the semantic congruency analysis:we wanted to determine whether differences in activationmaps observed between controls and patients could be dueto a specific pattern of activation associated with FTD or not.

Therefore, separate post-hoc ANOVAS contrasting FTD withcontrols and FTD with Non-FTD were performed.

3. Results

3.1. Online Behavioral Responses

3.1.1. Accuracy. We did not observe an effect of semanticcongruency in performance accuracy (F = 0.65, P = 0.43,df = 1). Equally, there was no significant interaction be-tween congruency and group (F = 0.46, P = 0.64, df = 1).Post- hoc tests (Tukey) demonstrated that there was a signif-icant contrast between FTD and controls [mean Difference =−0.128; Std error = 0.397; P = 0.009; CI: −0.227 to −0.029],reflecting impairment seen in incongruent trials (Figure 4)with FTD showing the worse performance (more mistakesthan both other groups). Non-FTD showed an intermediatelevel of performance, making more mistakes than FTD andfewer relative to controls, without significant differences inboth contrasts.

3.1.2. Reaction Times. In all groups, responses to congruenttrials were faster than incongruent ones (see Figure 5), withan observed significant effect of semantic congruency onreaction times (F = 14.17, P = 0.001, df = 1) and no signif-icant interactions between semantic congruency and group(F = 0.84, P = 0.442, df = 2). A robust main effect of groupon RT was found (F = 34.10, P < 0.001, df = 2). Non-FTDpatients were slower than FTD and controls in both con-ditions (congruent and incongruent trials). Controls werefaster than FTD in both conditions. Post-hoc contrastsdemonstrated significant differences in RT between all pairscontrasted, that is, controls versus FTD [mean difference =−0.442; std error = 0.800; P < 0.001; CI: −0.639 to −0.244],controls versus NON-FTD [mean difference = −0.672; std


ch2bet

(a)

Formal thought disorder score

0 5 10 15 20 25 30

Peak

SSQ

R2 linear = 0.325

0.05

0.025

0

−0.025

−0.05

(b)

Figure 2: (a) 3-Group ANOVA: semantic congruency (congruent versus incongruent) trials. Comparison of healthy controls (n = 10)versus schizophrenic patients with formal thought disorder (FTD, n = 9) versus schizophrenic patients without formal thought disorder(Non-FTD, n = 9). Attenuated activation in the left inferior frontal gyrus was observed in FTD relative to Non-FTD and to controls.The latter showed greater activation in this region also relative to non-FTD. The left side of the brain images corresponds to the rightside of the brain. The superior part of the brain images corresponds to the anterior brain region. (b) Scatter plot of the activity in the leftinferior frontal cortex as a function of the score in formal thought disorder (FTD) within schizophrenic patients (n = 18). We observe thatFTD higher scores in a sample of psychotic schizophrenic patients are negatively correlated with activity in the left inferior frontal cortex(x = −43.33, y = 14.81, z = 14.85).

error = 0.085; P < 0.001; CI: −0.880 to −0.463], and FTDversus NON-FTD [mean difference = −0.229; std error =0.088; P = 0.036; CI: −0.446 to −0.013].

4. Neuroimaging Results

4.1. Control Subjects

4.1.1. Congruent and Incongruent Sentences (Table 3). Con-gruent trials were associated with increased activation rel-ative to incongruent trials in the dorsal portion of the leftinferior frontal gyrus, the orbital portion of the right inferiorfrontal gyrus, the anterior cingulate bilaterally, the rightcaudate nucleus, the left posterior middle temporal gyrus,the left parahippocampal gyrus, the right precuneus, and theright inferior frontal gyrus. Conversely, incongruent trialswere associated with more activation than congruent trials in

the left middle frontal gyrus, the right medial frontal gyrus,and the left superior frontal gyrus.

4.1.2. Between Group Differences: Controls versus Non-FTDversus FTD (Table 3). The contrast between congruent andincongruent sentences was associated with greater activationin the dorsal portion of the left inferior frontal cortexin controls than Non-FTD patients, which in turn hadincreased activation relative to FTD patients in the sameregion (Figure 2(a)). In this analysis, we did not find regionsin which the patients groups showed more activation thancontrols.

4.1.3. Activation Associated with FTD. POST-HOC contrastsof FTD patients versus controls and FTD versus Non-FTD patients confirmed attenuated activation in the leftinferior frontal cortex in the FTD group. Relative to controls,FTD patients showed less activation in the dorsal portion


Table 3: The semantic decision match fmri task: neuroimaging results: 3D cluster sizes, talairach coordinates (of peak voxels) and cluster Pvalues for main effects of the semantic congruency.

(a) Within group analysis: congruent versus incongruent trials in healthy controls (n = 10). Displayed only clusters > or = 6 in size

Contrast Size Tal(x) Tal(y) Tal(z) Probability Cerebral region

Congruent >Incongruent

2189 43.33 14.81 14.85 0.000656 Left inferior frontal gyrus, dorsal portion

312 46.94 29.63 −7.15 0.000656 Right inferior frontal gyrus, orbital portion

181 0.00 11.11 31.35 0.000656 Bilateral anterior cingulate gyrus

3 14.44 11.11 25.85 0.008530 Right anterior cingulate gyrus

15 10.83 22.22 −1.65 0.000656 Right caudate

7 −36.11 −48.15 3.85 0.001969 Left middle temporal gyrus

6 −21.67 −29.63 −12.65 0.001969 Left parahippocampal gyrus

6 21.67 −66.67 36.85 0.004593 Right precuneus

Incongruent >Congruent

3094 −32.50 37.04 25.85 0.000678 Left middle frontal gyrus

15 3.61 44.44 25.85 0.004749 Right medial frontal gyrus

14 −21.67 55.56 9.35 0.000678 Left superior frontal gyrus

(b) Between Groups Analyses:. 3-Group ANOVA: Semantic Congruency (congruent versus incongruent) trials. Comparison of healthy controls (n = 10)versus schizophrenic patients with Formal Thought Disorder (FTD, n = 9) versus schizophrenic patients without formal thought disorder (Non-FTD,n = 9)

Results Size Tal(x) Tal(y) Tal(z) Probability Cerebral Region

Controls > NTD > TD 79 −43.33 14.81 14.85 0.007784 Left inferior frontal gyrus, dorsal portion

(c) Post-hoc 2 group comparisons: semantic congruency (congruent versus incongruent trials) schizophrenic patients with formal thought disorder(FTD, n = 9) versus healthy controls (n = 10) and FTD versus schizophrenic patients without formal thought disorder (Non-FTD, n = 9)

Results Size Tal(x) Tal(y) Tal(z) Probability Cerebral Region

Controls > TD

172 −43.33 14.81 14.85 0.000284 Left inferior frontal gyrus, dorsal portion

95 28.89 −70.37 −18.15 0.000851 Right cerebellum, posterior lobe, declive

92 50.56 25.93 14.85 0.002553 Right inferior frontal gyrus, dorsal portion

67 −57.78 −37.04 3.85 0.011211 Left middle temporal gyrus

42 −25.28 −62.96 36.85 0.012766 Left Precuneus

35 −39.72 −62.96 −18.15 0.012482 Left cerebellum, posterior lobe, declive

25 −10.83 −92.59 −7.15 0.012482 Left lingual gyrus

NON-FTD > TD185 −43.33 14.81 20.35 0.000256 Left middle frontal gyrus

127 46.94 18.52 25.85 0.001282 Right middle frontal gyrus

37 0.00 18.52 36.85 0.008720 Bilateral anterior cingulate

NON-FTD < TD 81 7.22 −55.56 20.35 0.003912 Right posterior cingulate

of the inferior frontal cortex bilaterally, the left middletemporal gyrus, precuneus, and lingual gyrus and in thecerebellum bilaterally (Figure 3(a)). No areas were relativelymore activated in the FTD than the control group.

Relative to Non-FTD patients, FTD patients showed lessactivation in the left middle frontal gyrus and bilaterallyin the anterior cingulate gyrus (Figure 3(b) and Table 3).Conversely, the FTD subgroup showed more activation thanthe Non-FTD subgroup in the right posterior cingulate gyrusand the left cerebellum (Figure 3(c)).

5. Discussion

5.1. Behavioral Results. Semantic congruency influencedRTs, with minimum values in trials with congruous endings,

as expected (see Figure 5). Therefore, we observed an effectof inhibition (or interference) associated with processing ofincongruent sentences in all groups. The patients made moreerrors than controls, particularly those with FTD. This wasmost evident on incongruent trials where patients with FTDperformed much worse than both patients with no FTD andcontrols, showing deficiency in processes where inhibition isexpected.

5.2. Neuroimaging Results

5.2.1. Semantic Processing in Healthy Controls. We found thatincongruent sentences activated the left middle frontal cortexmore than congruent ones, an area associated with suppres-sion when the same type of semantic material was employed


(a) (b)

(c)

Figure 3: Post-hoc 2 group comparisons: semantic congruency (congruent versus incongruent trials). (a) healthy controls (n = 10) versusschizophrenic patients with formal thought disorder (FTD, n = 9) (b, c) FTD versus schizophrenic patients without formal Thought Disorder(Non-FTD, n = 9). (a) Controls show greater activation relative to FTD in the inferior frontal gyrus bilaterally, dorsal portions, in the leftmiddle temporal gyrus (BA 22), in the left lingual gyrus, in the left precuneus, and in the cerebellum (posterior lobe) bilaterally. (b) Non-FTD patients showed greater activation of bilateral middle frontal gyrus and bilateral anterior cingulate, relative to FTD. (c) FTD patientsshow greater activation relative to Non-FTD patients in the right posterior cingulate. The left side of the brain images correspond to the rightside of the brain. The superior part of the brain images correspond to the anterior brain region.

[39] and with inhibition of stereotyped responses [52]. How-ever, in our study, we observed greater activity for congruentthan incongruent trials in healthy controls in the rightprecuneus, contrary to our hypothesis. Therefore, our firsthypothesis was partially confirmed. A possible explanationfor these differences observed between ours and Allen etal.’s study may be the fact that their task required overtproduction of a final completing word for visually presentedpriming sentence stems, whilst ours only required making adecision upon a visually presented final word.

The general pattern of activation we found was expectedbecause the semantic task used placed high demands onexecutive control, such as maintaining competing items onhold whilst checking for semantic appropriateness [53] andinhibiting unnecessary items. Thus, in healthy controls, weobserved more activation for congruent than incongruenttrials in the right inferior (orbital) and left inferior (dorsal)cortices, left posterior temporal cortex, bilateral anteriorcingulate, and left precuneus. Our findings are in line withdata supporting an important role of the left hemisphere in


0.95

0.85

0.75Controls NON-FTD FTD

CongruentIncongruent

Mea

np

erce

nta

geof

corr

ect

resp

onse

s(s

tan

dard

erro

r)

Figure 4: Semantic decision match fMRI task: performanceaccuracy in semantic congruent and incongruenet trials. Control(n = 10) versus NON-FTD (n = 9) versus FTD (n = 9).

Controls NON-FTD FTD0.5

0.7

0.9

1.1

1.3

1.5

1.7

1.9

Mea

nre

acti

onti

mes

(sec

onds

)(s

tan

dard

erro

r)

CongruentIncongruent

Figure 5: Semantic decision match fMRI task: reaction times insemantic congruent and incongruent trials. Control (n = 10) versusNON-FTD (n = 9) versus FTD (n = 9).

the final integration of semantic aspects present in sentencesand texts [54] and recent meta-analytical studies showingactivity in these regions associated with semantic processing[55–57].

5.2.2. Semantic Processing in FTD Schizophrenic Patients. Wealso hypothesized that FTD patients would show decreasedactivation in the prefrontal (middle and inferior), temporal,and cingulate cortices. Relative to controls, FTD patientsshowed a diminished activation in dorsal inferior frontal cor-tex (bilaterally) and in the left middle temporal gyrus whenprocessing incongruent relative to congruent sentences. Wepredicted that these differences would also be evident whencontrasting patients with and without FTD. We found thatfor the incongruent/congruent contrast, the FTD subgroupshowed less activation than Non-FTD patients bilaterallyin the left inferior/middle frontal gyri and the anteriorcingulate gyri. Thus, our second hypothesis was confirmed.The differential brain activity observed in FTD patientswhile performing a task, without mistakes, seem to bearising from an early point of the semantic processing, sincethese attenuated activations were observed in areas in whichcontrols showed increased activation for congruent trialsrelative to incongruent ones and not the other way round

(contrary to our initial prediction). Although this may seema contradictory finding, the current study confirms alteredbrain activity in these areas specifically associated with FTDin schizophrenia [28–32].

Manipulation of incongruent sentences requires over-coming a prepotent response, that is, suppression of auto-matically activated word(s) after a primed sentence stem[36]. The latter process needs to be maintained “active” fora longer period. Activity in the DLPFC has been associatedwith working memory and contextual information in exper-iments requiring heavier semantic processing demands [31,58]. These group differences in inferior frontal and leftmiddle temporal responses reflected less activation of theseareas in patients with FTD relative to both controls andpatients without FTD. This differential pattern of response issimilar to that evident in their behaviour, where patients withFTD performed markedly worse than both other groups ontrials involving semantically incongruent sentences, makingmore mistakes and also having a shorter RT than Non-FTDpatients. Relative to Non-FTD, FTD also had less activationin the anterior cingulate cortices, which have been suggestedby several authors to participate in demanding processes,such as monitoring conflict of response [59], engagementof cognitive control [60], top-down inhibitory modulation[61], detection of mnemonic competition, and retrieval in-duced forgetting [62].

A final point should be made about the activity in theprecuneus. This region has been demonstrated to be involvedin semantic processing [39, 57, 63, 64] episodic memoryretrieval [65], retrieval-induced facilitation [66], and do-pamine regulation of working memory [67]. FTD patients,but not Non-FTD patients, had less activation in the leftprecuneus than controls. Despite some subtle inconsistenciesin hemispheric localization (not rare in fMRI studies ofschizophrenia), this finding suggests a problem in episodicmemory retrieval that might have contributed to the FTDgroup worse task performance. Thus, FTD patients showedrobust diminished activation in areas implicated in thesuppression of irrelevant material as well as detection andresolution of mnemonic conflict [68], selecting context-appropriate meanings in the presence of competing mean-ings [69], and differences between impaired and facilitatedinformation [66]. Interestingly, Arcuri [46] found that FTDpatients produced significantly more expected words thancontrols, when overtly producing anomalous endings forsentences of a high degree of constraint, in a modifiedversion of the Hayling Task [70]. That study included thescanned FTD patients (n = 9) from the present studybut had a larger sample of FTD patients (n = 21). Thus,FTD patients who showed diminished activation in areasimplicated in aspects of executive control, when decidingwhether a final target word is semantically appropriate fora previously presented sentence, also had difficulties ininhibiting automatically activated words that followed sen-tences of high semantic constraint, when they were asked toopenly generate semantically incongruent completions forthose sentences.

The differences between FTD and the Non-FTD patientsare not attributable to sociodemographic or other clinical


differences as they are matched in all respects other thanthe severity of positive FTD. Moreover, although the FTDpatients performed the task more poorly than Non-FTDpatients and controls, the image analysis was restricted tocorrect trials, indicating that the differences in activationwere not simply secondary to the more impaired task perfor-mance in this subgroup. The differences in regional responsesmay thus be linked to the pathophysiology of incoherentspeech. The prominent role of the inferior frontal cortexin language processing and particularly semantic processingis consistent with this suggestion. However, the fact that aqualitatively similar but less severe reduction in activationwas evident in patients with no FTD indicates that findings inthis area are also associated with the disorder of schizophre-nia, independent of the presence of this particular symptom.

Liddle et al. [27] proposed that the disorganisation syn-drome (disorders of the form of thought and innappropriateaffect) is associated with a differential pattern of brain activ-ity in frontal, temporal, and cingulate cortices relative to thetwo other syndromes (psychomotor poverty and reality dis-tortion). In our sample, although we focused specifically onthe symptom of FTD, it was significantly correlated withinappropriate affect, that is, patients with FTD had higherscores in innappropriate effect as well, pointing to the samedirection of a distinct syndrome of disorganisation withinschizophrenia.

5.3. Limitations. The sample was small, and larger studiesshould be done to confirm our results and possibly explainthe limitations in the observed data, which might have con-tributed to a possible lack of power in our analyses.

6. Conclusion

The frontotemporal network normally engaged by a seman-tic decision task was underactivated in schizophrenia, partic-ularly in patients with FTD. This network is implicated in theinhibition of automatically primed stimuli and impairmentof its function interferes with language processing and con-tributes to the production of incoherent speech.

Acknowledgments

The authors are grateful for all the participants for theircooperation, particularly the patients. All authors con-tributed critically to paper writing. S. M. Arcuri designedand performed the experiment, recruited, interviewed, andtested participants (healthy controls and patients), anal-ysed the data, and interpreted the results. M. R. Broomerecruited patients and contributed to experiment perfor-mance. V.Giampietro and M. Brammer designed the fmrianalysis software and contributed to fmri data analysis. E.Amaro Jr. contributed to experimental design and to event-related fmri expertise for data collection and analysis. T. T. J.Kircher contributed to recruitment and experiment perfor-mance in healthy control participants. S. C. R. Williams andC. M. Andrew provided input for the experimental design

and fmri procedure. R. G. Morris contributed to the exper-imental design and offline data analysis and interpretationof results. P. K. McGuire provided critical input for testing apsychiatric population, data analysis, and interpretation. Theauthors thank Dr. Tamara Russell for helpful comments. Thiswork is part of Dr Arcuri’s Ph.D. scholarship CAPES (BEX20–74/95-1) Brasilia, Brazil.

References

[1] V. Peralta and M. J. Cuesta, “Eugen bleuler and the schizophre-nias: 100 Years after,” Schizophrenia Bulletin, vol. 37, no. 6, pp.1118–1120, 2011.

[2] S. Heckers, “Bleuler and the neurobiology of schizophrenia,”Schizophrenia Bulletin, vol. 37, no. 6, pp. 1131–1135, 2011.

[3] M.-C. Hardy-Bayle, Y. Sarfati, and C. Passerieux, “The cogni-tive basis of disorganization symptomatology in Schizophre-nia and its clinical correlates: toward a pathogenetic approachto disorganization,” Schizophrenia Bulletin, vol. 29, no. 3, pp.459–471, 2003.

[4] A. Moskowitz and G. Heim, “Eugen Bleuler’s DementiaPraecox or the Group of Schizophrenias (1911): a centenaryappreciation and reconsideration,” Schizophrenia Bulletin, vol.37, no. 3, pp. 471–479, 2011.

[5] M. Spitzer, “Associative networks, formal thought disorderand schizophrenia. On the experimental psychopathology ofspeech-dependent thought processes,” Nevernartz, vol. 64, no.3, pp. 147–159, 1993.

[6] M. J. Minzenberg, B. A. Ober, and S. Vinogradov, “Semanticpriming in schizophrenia: a review and synthesis,” Journal ofthe International Neuropsychological Society, vol. 8, no. 5, pp.699–720, 2002.

[7] J. D. Ragland, R. C. Gur, J. Raz et al., “Effect of schizophreniaon frontotemporal activity during word encoding and recog-nition: a PET cerebral blood flow study,” American Journal ofPsychiatry, vol. 158, no. 7, pp. 1114–1125, 2001.

[8] M. Spitzer, U. Braun, L. Hermle, and S. Maier, “Associativesemantic network dysfunction in thought-disordered schizo-phrenic patients: direct evidence from indirect semantic prim-ing,” Biological Psychiatry, vol. 34, no. 12, pp. 864–877, 1993.

[9] M. Spitzer, I. Weisker, M. Winter, S. Maier, L. Hermle, andB. A. Maher, “Semantic and phonological priming in schizo-phrenia,” Journal of Abnormal Psychology, vol. 103, no. 3, pp.485–494, 1994.

[10] M. Spitzer, “A cognitive neuroscience view of schizophrenicthought disorder,” Schizophrenia Bulletin, vol. 23, no. 1, pp.29–50, 1997.

[11] M. Kiefer and M. Spitzer, “Time course of conscious andunconscious semantic brain activation,” NeuroReport, vol. 11,no. 11, pp. 2401–2407, 2000.

[12] M. Kiefer, U. Martens, M. Weisbrod, L. Hermle, and M.Spitzer, “Increased unconscious semantic activation in schizo-phrenia patients with formal thought disorder,” SchizophreniaResearch, vol. 114, no. 1–3, pp. 79–83, 2009.

[13] T. E. Goldberg, M. S. Aloia, M. L. Gourovitch, D. Missar,D. Pickar, and D. R. Weinberger, “Cognitive substrates ofthought disorder, I: the semantic system,” American Journal ofPsychiatry, vol. 155, no. 12, pp. 1671–1676, 1998.

[14] J. G. Kerns and H. Berenbaum, “Cognitive impairments asso-ciated with formal thought disorder in people with schizo-phrenia,” Journal of Abnormal Psychology, vol. 111, no. 2, pp.211–224, 2002.

[15] T. H. McGlashan, “Eugen bleuler: centennial anniversary ofhis 1911 publication of dementia praecox or the group of


schizophrenias,” Schizophrenia Bulletin, vol. 37, no. 6, pp.1101–1103, 2011.

[16] G. McCarthy, A. C. Nobre, S. Bentin, and D. D. Spencer,“Language-related field potentials in the anterior-medial tem-poral lobe: I. Intracranial distribution and neural generators,”The Journal of Neuroscience, vol. 15, no. 2, pp. 1080–1089,1995.

[17] A. C. Nobre and G. McCarthy, “Language-related field poten-tials in the anterior-medial temporal lobe: II. Effects of wordtype and semantic priming,” The Journal of Neuroscience, vol.15, no. 2, pp. 1090–1098, 1995.

[18] P. J. Holcomb, “Semantic priming and stimulus degradation:implications for the role of the N400 in language processing,”Psychophysiology, vol. 30, no. 1, pp. 47–61, 1993.

[19] P. F. Mitchell, S. Andrews, A. M. Fox, S. V. Catts, P. B.Ward, and N. McConaghy, “Active and passive attention inschizophrenia: an ERP study of information processing in alinguistic task,” Biological Psychology, vol. 32, no. 2-3, pp. 101–124, 1991.

[20] M. A. Niznikiewicz, B. F. O’Donnell, P. G. Nestor et al.,“ERP assessment of visual and auditory language processing inschizophrenia,” Journal of Abnormal Psychology, vol. 106, no. 1,pp. 85–94, 1997.

[21] S. Andrews, A. M. Shelley, P. B. Ward, A. Fox, S. V. Catts, andN. McConaghy, “Event-related potential indices of semanticprocessing in schizophrenia,” Biological Psychiatry, vol. 34, no.7, pp. 443–458, 1993.

[22] J. Adams, S. F. Faux, P. G. Nestor et al., “ERP abnormalitiesduring semantic processing in schizophrenia,” SchizophreniaResearch, vol. 10, no. 3, pp. 247–257, 1993.

[23] M. Kutas and S. A. Hillyard, “Brain potentials during readingreflect word expectancy and semantic association,” Nature,vol. 307, no. 5947, pp. 161–163, 1984.

[24] D. Salisbury, “N400 to lexical ambiguity and semantic incon-gruity in schizophrenia,” International Journal of Psychophysi-ology, vol. 75, no. 2, pp. 127–132, 2010.

[25] M. Kostova, C. Passerieux, J. P. Laurent, and M. C. Hardy-Bayle, “N400 anomalies in schizophrenia are correlatedwith the severity of formal thought disorder,” SchizophreniaResearch, vol. 78, no. 2-3, pp. 285–291, 2005.

[26] J. P. Laurent, M. Kostova, and C. Passerieux, “N400 and P300modulation as functions of processing level in schizophreniapatients exhibiting formal thought disorder,” InternationalJournal of Psychophysiology, vol. 75, no. 2, pp. 177–182, 2010.

[27] P. F. Liddle, K. J. Friston, C. D. Frith, S. R. Hirsch, T. Jones,and R. S. J. Frackowiak, “Patterns of cerebral blood flow inschizophrenia,” British Journal of Psychiatry, vol. 160, pp. 179–186, 1992.

[28] P. K. McGuire, D. J. Quested, S. A. Spence, R. M. Murray, C. D.Frith, and P. F. Liddle, “Pathophysiology of “positive” thoughtdisorder in schizophrenia,” British Journal of Psychiatry, vol.173, pp. 231–235, 1998.

[29] T. T. J. Kircher, E. T. Bulimore, M. J. Brammer et al., “Dif-ferential activation of temporal cortex during sentence com-pletion in schizophrenic patients with and without formalthought disorder,” Schizophrenia Research, vol. 50, no. 1-2, pp.27–40, 2001.

[30] T. T. J. Kircher, P. F. Liddle, M. J. Brammer, S. C. R. Williams,R. M. Murray, and P. K. McGuire, “Neural correlates of formalthought disorder in schizophrenia: preliminary findings froma functional magnetic resonance imaging study,” Archives ofGeneral Psychiatry, vol. 58, no. 8, pp. 769–774, 2001.

[31] A. W. MacDonald III, C. S. Carter, J. G. Kerns et al., “Specificityof prefrontal dysfunction and context processing deficits to

schizophrenia in never-medicated patients with first-episodepsychosis,” American Journal of Psychiatry, vol. 162, no. 3, pp.475–484, 2005.

[32] I. H. Park, H. J. Park, J. W. Chun, E. Y. Kim, and J. J. Kim,“Prefrontal functional dissociation in the semantic network ofpatients with schizophrenia,” NeuroReport, vol. 19, no. 14, pp.1391–1395, 2008.

[33] M. Nakamura, P. G. Nestor, J. J. Levitt et al., “Orbitofrontalvolume deficit in schizophrenia and thought disorder,” Brain,vol. 131, no. 1, pp. 180–195, 2008.

[34] J. Sun, J. J. Maller, L. Guo, and P. B. Fitzgerald, “Superiortemporal gyrus volume change in schizophrenia: a review onRegion of Interest volumetric studies,” Brain Research Reviews,vol. 61, no. 1, pp. 14–32, 2009.

[35] H. Horn, A. Federspiel, M. Wirth et al., “Structural andmetabolic changes in language areas linked to formal thoughtdisorder,” British Journal of Psychiatry, vol. 194, no. 2, pp. 130–138, 2009.

[36] N. C. Andreasen, Scale for the Assessment of Positive Symptoms(SAPS), The University of Iowa, Iowa City, Iowa, USA, 1984.

[37] C. Van Petten, “Words and sentences: event-related brainpotential measures,” Psychophysiology, vol. 32, no. 6, pp. 511–525, 1995.

[38] D. A. Nathaniel-James, P. Fletcher, and C. D. Frith, “Thefunctional anatomy of verbal initiation and suppression usingthe Hayling Test,” Neuropsychologia, vol. 35, no. 4, pp. 559–566, 1997.

[39] P. Allen, A. Mechelli, K. E. Stephan et al., “Fronto-temporalinteractions during overt verbal initiation and suppression,”Journal of Cognitive Neuroscience, vol. 20, no. 9, pp. 1656–1669,2008.

[40] American Psychiatric Association, Diagnostic and StatisticalManual of Mental Disorders, American Psychiatric Associa-tion, Washington, DC, USA, 4th edition, 1994.

[41] M. Annett, “A classification of hand preference by associationanalysis,” The British journal of psychology, vol. 61, no. 3, pp.303–321, 1970.

[42] H. E. Nelson and J. R. Willison, National Adult Reading Test(NART): Test Manual, NFER-Nelson, Windsor, UK, 2nd edi-tion, 1991.

[43] N. C. Andreasen, Scale for the Assessment of Negative Symptoms(SANS), The University of Iowa, Iowa City, Iowa, USA, 1984.

[44] S. M. Arcuri, S. Rabe-Hesketh, R. G. Morris, and P. K.McGuire, “Regional variation of cloze probabilities for sen-tence contexts,” Behavior Research Methods, Instruments, andComputers, vol. 33, no. 1, pp. 80–90, 2001.

[45] D. Norris, J. M. McQueen, and A. Cutler, “Bias effects infacilitatory phonological priming,” Memory and Cognition,vol. 30, no. 3, pp. 399–411, 2002.

[46] S. M. Arcuri, Neural and cognitive studies of thought disorderin schizophrenia, Ph.D. thesis, King’s College, University ofLondon, London, UK, 2003.

[47] E. Bullmore, C. Long, J. Suckling et al., “Colored noise andcomputational inference in neurophysiological (fMRI) timeseries analysis: resampling methods in time and wavelet do-mains,” Human Brain Mapping, vol. 12, no. 2, pp. 61–78, 2001.

[48] K. J. Friston, S. Williams, R. Howard, R. S. J. Frackowiak, andR. Turner, “Movement-related effects in fMRI time-series,”Magnetic Resonance in Medicine, vol. 35, no. 3, pp. 346–355,1996.

[49] E. Bullmore, M. Brammer, S. C. R. Williams et al., “Statisticalmethods of estimation and inference for functional MR imageanalysis,” Magnetic Resonance in Medicine, vol. 35, no. 2, pp.261–277, 1996.


[50] J. Talairach and P. Tournoux, Co-Planar Stereotaxic Atlas of theHuman Brain, Thieme, New York, NY, USA, 1993.

[51] E. T. Bullmore, M. J. Brammer, S. Rabe-Hesketh et al., “Meth-ods for diagnosis and treatment of stimulus-correlated motionin generic brain activation studies using fMRI,” Human BrainMapping, vol. 7, no. 1, pp. 38–48, 1999.

[52] H. Kadota, H. Sekiguchi, S. Takeuchi, M. Miyazaki, Y. Kohno,and Y. Nakajima, “The role of the dorsolateral prefrontal cor-tex in the inhibition of stereotyped responses,” ExperimentalBrain Research, vol. 203, no. 3, pp. 593–600, 2010.

[53] J. M. Novick, I. P. Kan, J. C. Trueswell, and S. L. Thompson-Schill, “A case for conflict across multiple domains: memoryand language impairments following damage to ventrolateralprefrontal cortex,” Cognitive Neuropsychology, vol. 26, no. 6,pp. 527–567, 2009.

[54] F. Homae, R. Hashimoto, K. Nakajima, Y. Miyashita, and K.L. Sakai, “From perception to sentence comprehension: theconvergence of auditory and visual information of languagein the left inferior frontal cortex,” NeuroImage, vol. 16, no. 4,pp. 883–900, 2002.

[55] J. R. Binder, R. H. Desai, W. W. Graves, and L. L. Conant,“Where is the semantic system? A critical review and meta-analysis of 120 functional neuroimaging studies,” CerebralCortex, vol. 19, no. 12, pp. 2767–2796, 2009.

[56] M. Vigneau, V. Beaucousin, P. Y. Herve et al., “Meta-analyzingleft hemisphere language areas: phonology, semantics, andsentence processing,” NeuroImage, vol. 30, no. 4, pp. 1414–1432, 2006.

[57] W. W. Graves, R. Desai, C. Humphries, M. S. Seidenberg, andJ. R. Binder, “Neural systems for reading aloud: a multipara-metric approach,” Cerebral Cortex, vol. 20, no. 8, pp. 1799–1815, 2010.

[58] M. D’Esposito, B. R. Postle, and B. Rypma, “Prefrontal corticalcontributions to working memory: evidence from event-related fMRI studies,” Experimental Brain Research, vol. 133,no. 1, pp. 3–11, 2000.

[59] D. M. Barch, T. S. Braver, F. W. Sabb, and D. C. Noll, “Anteriorcingulate and the monitoring of response conflict: evidencefrom an fMRI study of overt verb generation,” Journal ofCognitive Neuroscience, vol. 12, no. 2, pp. 298–309, 2000.

[60] J. G. Kerns, J. D. Cohen, A. W. MacDonald III, R. Y. Cho,V. A. Stenger, and C. S. Carter, “Anterior cingulate conflictmonitoring and adjustments in control,” Science, vol. 303, no.5660, pp. 1023–1026, 2004.

[61] P. H. Ghatan, J. C. Hsieh, K. M. Petersson, S. Stone-Elander,and M. Ingvar, “Coexistence of attention-based facilitationand inhibition in the human cortex,” NeuroImage, vol. 7, no.1, pp. 23–29, 1998.

[62] B. A. Kuhl, N. M. Dudukovic, I. Kahn, and A. D. Wagner,“Decreased demands on cognitive control reveal the neuralprocessing benefits of forgetting,” Nature Neuroscience, vol. 10,no. 7, pp. 908–914, 2007.

[63] T. T. J. Kircher, M. Brammer, N. T. Andreu, S. C. R. Williams,and P. K. McGuire, “Engagement of right temporal cortexduring processing of linguistic context,” Neuropsychologia, vol.39, no. 8, pp. 798–809, 2001.

[64] G. I. de Zubicaray, F. O. Zelaya, C. Andrew, S. C. R.Williams, and E. T. Bullmore, “Cerebral regions associatedwith verbal response initiation, suppression and strategy use,”Neuropsychologia, vol. 38, no. 9, pp. 1292–1304, 2000.

[65] A. E. Cavanna and M. R. Trimble, “The precuneus: a review ofits functional anatomy and behavioural correlates,” Brain, vol.129, no. 3, pp. 564–583, 2006.

[66] M. Wimber, K. H. Bauml, Z. Bergstrom, G. Markopoulos,H. J. Heinze, and A. Richardson-Klavehn, “Neural markersof inhibition in human memory retrieval,” The Journal ofNeuroscience, vol. 28, no. 50, pp. 13419–13427, 2008.

[67] P. Fusar-Poli, O. D. Howes, P. Allen et al., “Abnormal fron-tostriatal interactions in people with prodromal signs of psy-chosis: a multimodal imaging study,” Archives of General Psy-chiatry, vol. 67, no. 7, pp. 683–691, 2010.

[68] B. A. Kuhl, I. Kahn, N. M. Dudukovic, and A. D. Wagner,“Overcoming suppression in order to remember: contribu-tions from anterior cingulate and ventrolateral prefrontal cor-tex,” Cognitive, Affective and Behavioral Neuroscience, vol. 8,no. 2, pp. 211–221, 2008.

[69] C. Metzler, “Effects of left frontal lesions on the selection ofcontext-appropriate meanings,” Neuropsychology, vol. 15, no.3, pp. 315–328, 2001.

[70] P. W. Burgess and T. Shallice, “Response suppression, Initia-tion and strategy use following frontal lobe lesions,” Neuropsy-chologia, vol. 34, no. 4, pp. 263–272, 1996.


Review Article

Cognitive Control and Discourse Comprehension inSchizophrenia

Megan A. Boudewyn,1 Cameron S. Carter,1, 2 and Tamara Y. Swaab1

1 Department of Psychology and Center for Mind and Brain, University of California, Davis, One Shields Avenue,Davis, CA 95616, USA

2 Department of Psychiatry and Behavioral Sciences, Imaging Research Center, and Center for Neuroscience,University of California, Davis, Davis, CA 95616, USA

Correspondence should be addressed to Megan A. Boudewyn, [email protected]

Received 16 November 2011; Accepted 17 January 2012

Academic Editor: Margaret A. Niznikiewicz

Copyright © 2012 Megan A. Boudewyn et al. This is an open access article distributed under the Creative Commons AttributionLicense, which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properlycited.

Cognitive deficits across a wide range of domains have been consistently observed in schizophrenia and are linked to poorfunctional outcome (Green, 1996; Carter, 2006). Language abnormalities are among the most salient and include disorganizedspeech as well as deficits in comprehension. In this review, we aim to evaluate impairments of language processing in schizophreniain relation to a domain-general control deficit. We first provide an overview of language comprehension in the healthy humanbrain, stressing the role of cognitive control processes, especially during discourse comprehension. We then discuss cognitivecontrol deficits in schizophrenia, before turning to evidence suggesting that schizophrenia patients are particularly impaired atprocessing meaningful discourse as a result of deficits in control functions. We conclude that domain-general control mechanismsare impaired in schizophrenia and that during language comprehension this is most likely to result in difficulties during theprocessing of discourse-level context, which involves integrating and maintaining multiple levels of meaning. Finally, we predictthat language comprehension in schizophrenia patients will be most impaired during discourse processing. We further suggestthat discourse comprehension problems in schizophrenia might be mitigated when conflicting information is absent and strongrelations amongst individual words are present in the discourse context.

“There is no “centre of Speech” in the brain any more than there is a faculty of Speech in the mind.The entire brain, more or less, is at work in a man who uses language”

William JamesFrom The Principles of Psychology, 1890

“The mind in dementia praecox is like an orchestra without a conductor”Kraepelin, 1919

1. Introduction

Impaired cognition across a wide range of cognitive domainsis a pervasive feature of schizophrenia and is connectedto poor functional outcome for patients [1, 2]. Of thecognitive deficits that have been observed in schizophreniapatients, language abnormalities are among the most salientand include disorganized speech as well as deficits in

comprehension. However, there is no general consensus as towhether the cognitive impairments seen in schizophrenia canbe attributed to a single disrupted mechanism, to multipledisrupted systems, or to low-level perceptual deficits.

Accounts of language deficits in schizophrenia mightbe divided into theories that focus on irregularities insemantic memory structure and functioning, and those thatemphasize deficits in the ability to effectively use context [3].


Accounts that posit abnormalities in semantic memory inschizophrenia, such as the exaggerated spread of activation ina semantic network [4], are largely based on studies involvingthe processing of words in isolation. In contrast, accounts ofhigh-level language processing have focused on impairmentsin the ability to build and maintain context as the root oflanguage deficits in schizophrenia. Such deficits have beenattributed to problems with general cognitive processes ofenhancement and suppression in a prominent theory ofdiscourse processing, the structure-building framework [5].Kuperberg [6] has proposed an imbalance between semantic-memory-based processing systems and combinatorial mech-anisms in schizophrenia, such that patients often rely to agreater extent on semantic-memory-based processing at theexpense of the structured build-up of context. The interac-tion between the semantic-memory-processing stream andthe combinatorial stream has been proposed to be influencedby cognitive control mechanisms [7]. According to this view,control mechanisms should therefore be recruited whenintegration demands are high.

In the remainder of this theoretically-oriented reviewpaper we seek to increase understanding and highlight novelhypotheses related to impairments in aspects of languageprocessing in schizophrenia. We will begin with a discussionof language comprehension in the healthy human brain andwill highlight the potentially important role of cognitive con-trol processes, especially during discourse comprehension.We will then review the literature on cognitive control deficitsin schizophrenia. Finally, we will review evidence suggestingthat schizophrenia patients are particularly impaired atprocessing meaningful discourse as a result of impairedcontrol functions. We conclude that schizophrenia patientsare impaired in discourse comprehension when controldemands are particularly high, for example, when differentaspects of language input conflict with one another.

2. Cognitive Control and High-Level LanguageProcessing in Healthy Participants

Outside of the laboratory, natural language processingalmost always involves a rich signal, featuring multiple con-nected sentences that each generates a meaningful contextand also must be linked to the overall meaning providedby discourse context. For example, a sentence such as“Yesterday he went to the bank” may be interpreted asdescribing a person visiting a financial institution, yet thatinterpretation would be incorrect if the preceding discoursewas about a person going to the side of a river. In short,discourse context is a complex web of information includingword-level, sentence-level, and message-level meaning, aswell as syntactic structure, previously stored backgroundinformation, and other context information (e.g., speech-accompanying gestures used by the speaker).

In order to construct a coherent meaning representa-tion, language users must integrate these various sourcesof information, and some prominent models of languagecomprehension predict that this involves specific controlmechanisms. As briefly discussed earlier in this review, one

theory that proposes an important role of control mech-anisms during discourse comprehension is the structure-building framework, which emphasizes suppression ability ascritical to language comprehension [8–10]. According to thisframework, incoming language input is mapped onto mentalstructures via the enhancement of relevant information andsuppression of irrelevant information. When input fits wellwith the current structure, it will be incorporated into thatstructure. This process is referred to as mapping, whichinvolves adding on to the developing representation ofthe text. Integration of new information into an existingrepresentation places fewer demands on cognitive resourcesthan the initial construction of a discourse representation;therefore, incoming input that is coherent with the previouscontext leads to faster processing. When input is lesscoherent with the existing structure, the comprehender willconstruct a new substructure in order to accommodate thenew information into the overall representation (shifting).The creation of a new substructure has been labeled“shifting,” and it requires additional processing, which canslow down comprehension. A failure to effectively suppressirrelevant information that is activated by the incominginput (e.g., the financial-institution meaning of the word“bank” in the example above, when context-irrelevant) mayresult in the construction of excessive substructures [8–10]. The cognitive processes of mapping and shifting arethought to rely on control networks; inefficient mapping,and particularly shifting, has been argued to result inpoor discourse comprehension in healthy adults, since thiscould lead to lingering activation of discourse-irrelevantinformation [8–10]. Excessive shifting could also result inthe disorganized discourse representations that have beenobserved in schizophrenia patients [5].

Another model of language processing that recognizesa role of cognitive control mechanisms during languageprocessing is the memory, unification, and control (MUC)model [11]. Briefly, according to the MUC model, languageprocessing requires activation and retrieval of semantic andsyntactic memory representations stored in left temporalcortex. Unification refers to the construction of a meaningrepresentation involving multiple words (or sentences),using information about the words that were activated andretrieved. Unification processes are proposed to be mediatedby the left inferior frontal cortex (Broca’s complex). Finally,control processes are assumed to be mediated by the dor-solateral prefrontal cortex (DLPFC) and anterior cingulatecortex (ACC) and are recruited when attentional control isnecessary, for example, during turn-taking in conversationand during the selection of situation-appropriate languagein bilinguals [11]. (Other theories also suggest a direct roleof cognitive control during processing of other aspects of thelanguage input (for reviews, see [12, 13].)

In sum, prominent theories of high-level language pro-cessing concur that comprehension is supported by controlmechanisms. In addition, in the case of incoherent discourse,one or more sources of information may conflict (e.g.,activation of the financial-institution meaning of the wordbank would conflict with discourse-level context in whichthe side-of-the-river meaning is promoted), placing even


greater demands on these control mechanisms to supportcomprehension.

3. Cognitive Control and Schizophrenia

Cognitive control is somewhat of an umbrella term in thatit refers to domain-general processes that govern processingin other systems, including the allocation of attentionalresources, conflict detection and resolution, the maintenanceof task-relevant information, and inhibition. According tothe Guided Activation Model, control mechanisms mediat-ing top-down processing are based in the prefrontal cortex[14]. Specifically, the PFC maintains and represents task-relevant context, including goals, which serves to guideprocessing in relevant neural regions associated with taskexecution [14]. Deficits in this ability to maintain contextin order to guide processing would be expected to resultin a wide range of impairments across cognitive domains.Several theories postulate that impaired cognitive controlmight account for the widespread deficits in cognitionseen in schizophrenia [15–18]. As noted above, alternativeexplanations suggest that deficits in sensory or perceptualsystems may account for the cognitive deficits seen inschizophrenia (e.g., [19]). In the current paper, we adopt theapproach outlined by Lesh and colleagues [18]; namely, thata single underlying deficit in prefrontally mediated controlfunctions is at the heart of the broad range of cognitivedeficits that have been observed in schizophrenia patients.

A detailed discussion of cognitive dysfunction inschizophrenia, and the evidence that suggests deficits incontrol mechanisms are the basis for this dysfunction, isbeyond the scope of this review. However, it is importantto note that there is substantial evidence suggesting adomain-general control deficit in schizophrenia patients andspecifically showing reduced DLPFC activity accompaniedby impaired behavioral performance in patients duringtasks that are demanding of cognitive control (see [18]for a review), though a recent meta-analysis of executive-function tasks also showed that reduced DLPFC activationis present in schizophrenia whether performance is impairedor matched [20]. One salient example of empirical evidencefor a control deficit in schizophrenia comes from a studyby MacDonald and colleagues [21]. They presented never-medicated, first-episode schizophrenia patients, as well asnever-medicated nonschizophrenia psychosis patients andhealthy controls, with the AX-CPT task in an fMRI session[21]. The AX-CPT is a task that is specifically designed totax cognitive control functions: subjects are instructed topress a button every time they see an “X” that immediatelyfollows an “A” and to otherwise withhold their response.This type of trial (AX trial) is highly frequent, accountingfor 70% of total trials; this leads to the tendency to respondto an X and to anticipate making a response after anA. Therefore, false alarms often occur in response to AYtrials (where “Y” stands for any letter except “X”) as thecontext of “A” leads to the expectation of an upcomingresponse. Finally, BX trials (where “B” stands for any letterexcept “A”) are least frequent and are most demanding on

the controlled maintenance of context, because participantsmust use the context of having just encountered a “B” inorder to correctly inhibit their response to the “X”. Comparedto controls and nonschizophrenia psychosis patients, never-medicated schizophrenic patients made more errors on BXtrials [21]. fMRI results for BX trials for which controlsresponded appropriately showed increased activity in theDLPFC and also the posterior parietal areas compared tothe less demanding AX trials. For the patient group, thissame contrast showed that DLPFC activity was significantlyreduced compared to controls and that reduced DLPFCactivity was linked to increased disorganization symptomsin the schizophrenic individuals [21]. These results areconsistent with several other studies that have also foundreduced DLPFC activity corresponding to poor performanceon control tasks in schizophrenia patients (e.g., [22–24]) aswell as the results of a recent meta-analysis [20]. Importantly,the reduction in DLPFC activity found for schizophreniapatients compared to controls is most pronounced undercontrol-demanding task conditions; when control demandsare relatively low, DLPFC activation for patients has beenshown to approximate that of controls (e.g., [23]) or evento exceed that of controls (e.g., [25, 26]).

4. Cognitive Control, High-Level Language, andSchizophrenia

Language dysfunction is a hallmark of schizophrenia, leadingto the production of disorganized speech as well as deficitsin language comprehension (for reviews, see [6, 27, 28]).However, schizophrenia is far from a homogenous disorder;symptoms vary across individuals and symptoms withinan individual can vary over time as well. When present,common language phenomena observed in schizophre-nia patients include tangentiality (jumping from topic totopic without providing obvious links in response to aquestion), derailment (disjointed speech that slips fromtopic to topic), incoherence (incomprehensible speech), andpoverty of speech (reduction in the quantity of speech)[29, 30]. Although many of the features of language dys-function in schizophrenia are observed in production atthe discourse level (e.g., tangentiality and derailment), thebulk of the research on real-time language comprehensionin schizophrenia patients has focused on the word- andsentence-levels of processing.

A prominent area of research on language comprehen-sion in schizophrenia has been the influence of word-levelmeaning associations on the processing of incoming words(i.e., semantic priming paradigms). Results from these stud-ies suggest that the activation and retrieval of stored meaningrepresentations of words is relatively intact, and some studieseven show larger-than-normal effects of semantic priming(e.g., [7, 31–33]). Interestingly, under “automatic” primingconditions (e.g., when targets follow closely after primes),schizophrenia patients show normal or exaggerated semanticpriming effects but show smaller or absent effects whencontrolled processing is required, such as evaluation of therelation between prime and target [3, 33]. This pattern of


results has been interpreted as an indication of semanticmemory dysfunction in schizophrenia, specifically involvinga faster and more extensive propagation of activation insemantic memory [4, 28, 34].

At the sentence level, there is also a great deal ofwork showing various deficits in schizophrenia comparedto healthy adults (e.g., [35–37]). Much of this work canbe summarized as suggesting a deficit in the build-upand maintenance of context: compared to healthy controls,schizophrenia patients have been shown to be unable tobenefit from linguistic context in a word-monitoring task[35] and to have difficulty using sentence context to selectthe context-appropriate meaning of an ambiguous word[36]. Sitnikova and colleagues [36] presented schizophreniapatients and healthy controls with sentences that biasedtowards either the dominant meaning of an ambiguousword (e.g., bridge as an architectural structure) or thesubordinate meaning (bridge as a card game). The secondclause of the sentences also contained a word that wassemantically associated to the dominant meaning of theambiguous word that had appeared earlier in the sentence(e.g., . . . because the river had rocks in it, following either“diving was forbidden from the bridge . . .” or “The guestsplayed bridge . . .”). Sitnikova and colleagues [36] measuredthe electrical activity of the brain as healthy controls andpatients read the sentences. Event-related potentials (ERPs)were extracted for critical words that were either consistentor not with the preceding sentence contexts. Of particularrelevance to this study is an ERP effect that has been labeledthe N400. The N400 is a negatively deflecting ERP waveformthat is modulated by semantic fit, such as relatednessto previous words and congruence or predictability givenprior context (for a review, see [38]). Controls showed adecreased N400 amplitude for target words like river whenthey were appropriate given the previous sentence context(diving was forbidden from the bridge . . .) compared towhen they were inappropriate (the guests played bridge . . .).However, patients’ N400 effects to context-appropriate andcontext-inappropriate target words were indistinguishable,suggesting that patients were unable to benefit from contextto suppress the irrelevant meaning of the ambiguous word(e.g., river) [36]. In contrast, the patient and control groupsdid not differ in their response to unambiguous context-congruent compared to incongruent words (e.g., . . . the riverhad rocks/cracks in it . . .), such that both groups showed alarger N400 amplitude in response to incongruent words(cracks) than to congruent words (rocks) [36]. This setof results is suggestive of a specific impairment when thecontrolled use of linguistic context is required, such as tosuppress context-inappropriate meanings of words, ratherthan of an overall lack of sensitivity or attention to languagecontext.

Further evidence supporting deficits in the use of lin-guistic context in schizophrenia comes from a 2006 studyfrom Kuperberg and colleagues. Patients and controls werepresented with syntactically well-formed sentences that con-tained animacy violations (e.g., for breakfast the eggs wouldonly eat toast and jam). Previously, a P600 effect was foundin healthy adults when comparing these types of sentences

to those containing no animacy violation (e.g., for breakfastthe boys would only eat toast and jam) [27]. Semanticviolations typically elicit N400 effects, whereas P600 effectshave traditionally been linked to syntactic manipulations (see[38] for a review). The so-called “semantic P600” found inhealthy adults has been interpreted as reflective of conflictbetween the syntax-dictated sentence meaning (e.g., thatthe eggs were eating) and the aggregate meaning based onrelations among individual words (e.g., that eggs and eatare typically combined such that the eggs are being eaten)[27, 39]. Interestingly, schizophrenia patients show a reducedsemantic P600 effect relative to controls [37]. This pattern ofresults suggests that when strong semantic relations amongstindividual words are in conflict with sentence-level meaningas dictated by syntax, schizophrenia patients appear to beoverly influenced by the word-level relations.

In contrast, very few studies have looked at onlinediscourse comprehension in schizophrenia. As discussed ear-lier, several theories of high-level language comprehensionpredict that demands on cognitive control are high duringdiscourse processing, as discourse is a rich and multilevelsignal containing a great deal of information to maintain andintegrate, all with the potential to generate conflict. Givena model of cognitive dysfunction in schizophrenia based oncontrol deficits [18], clear difficulties in language processingmight be expected at the discourse level.

Indeed, offline studies of memory have shown thatpatients do not benefit from discourse organization to thesame extent as controls, manifesting as a lack of improvedtext recall with increased text coherence [40–43]. Further,a recent electrophysiological study presented schizophreniapatients and healthy adults with three-sentence passages inwhich the individual sentences were highly related to oneanother, intermediately related, or unrelated [44]. For exam-ple, a highly related passage might describe two characters ashaving an argument in the first sentence, as hitting each otherin the second sentence, and as having bruises the next day inthe third sentence. In order to understand the third sentenceof this type of passage, no inference is necessary: the contextof the previous two sentences directly states a cause for thebruising. An intermediately related passage, however, mightonly mention in the second sentence that the characters wereupset. In this case, an inference is necessary in order to builda coherent representation of the meaning of the passage.Finally, in an unrelated passage, the third sentence would becompletely incongruent with the previous two sentences. Inresponse to these types of passages, controls showed contexteffects on the N400 to critical target words (e.g., bruises)in the final sentences, with the greatest reduction in theN400 waveform for highly related passages (the “easiest”condition, when no inference was needed), followed byintermediately related passages, and finally by the unrelatedcondition. In contrast, ERP results in the patient group didnot distinguish among conditions in the N400 time window[44]. However, both the control and patient groups showed asimilar pattern of behavioral responses to the stimuli, ratinghighly related passages as “very related”, followed by theintermediately related passages as slightly less related, andthe unrelated passages as unrelated. In addition, although


the electrophysiological response for the patient group didnot differentiate among relatedness conditions in the N400time window, patients did show a difference between relatedpassages and unrelated passages in a later time window(700–1000 ms after reading the critical target word); incontrast, the control group did not show a differenceamongst conditions in this late time window [44]. Thissuggests that, although patients did not show differentialN400 responses depending on relatedness condition, theywere attending to the task and may have attempted tointegrate the targets in the unrelated condition at a laterprocessing point compared to controls. The generation ofinferences such as in the intermediately related exampleabove is often necessary in order to properly understanddiscourse context, as language input often does not directlycontain all of the information needed to comprehend themessage. Therefore, the patient group’s failure to differentiateamongst degrees of causal relatedness in the same timewindow as the control group suggests that schizophreniapatients have difficulty generating inferences to construct acoherent discourse representation on the same time scale ashealthy adults [44].

Consistent with this pattern is another recent ERP studythat presented patients and controls with five-sentence pas-sages, each of which contained a noun that could serve as thereferent of a noun in the fourth sentence but varied in termsof semantic fit [45]. For example, upon encountering outfit(The night before work, Lisa ironed the outfit) in the fourthsentence, possible referents mentioned in previous sentenceswould be suit (context appropriate and lexically associated),costume (context inappropriate but lexically associated), orring (context inappropriate and lexically unassociated). Thecontrol group showed a graded ERP response throughout theN400 time window, showing the smallest amplitude for thecontext-appropriate/associated condition, followed by thecontext-inappropriate/associated condition, and finally thecontext-inappropriate/unassociated condition. In contrast,in the early portion of the N400 time window (300–400 ms),the schizophrenia group distinguished between globallyappropriate and inappropriate semantic fit (reduced ampli-tude for the context-appropriate referent in comparison toeither context-inappropriate referent, irrespective of lexicalassociation). However, in the later portion of the N400 timewindow (400–500 ms), the pattern switched so that patientsdistinguished between locally associated and unassociatedsemantic fit (reduced amplitude for both associated condi-tions, regardless of context appropriateness, compared to theinappropriate-unassociated condition) [45]. In other words,the controls seemed able to benefit from the combination ofcontext-level fit and word-level fit, whereas the schizophreniapatients seemed to be toggling between sensitivity to globaland local fit [45].

In summary, the literature on high-level language com-prehension in schizophrenia patients shows that they aremost impaired when control demands are highest, as whenthe use of context is needed to constrain word meaning[36] or construct a meaning at odds with the associativerelations amongst individual words [37]. These results aresuggestive of a role for a cognitive control deficit in abnormal

language processing in schizophrenia, leading patients tofail to suppress context-irrelevant information as well asmaintain linguistic context in order to guide the processingof incoming words. This pattern of results is much in linewith the type of context-maintenance deficits seen in patientswhen performing the AX-CPT task described above, inwhich patients have difficulty maintaining the context ofhaving just seen a “B” in order to respond correctly bysuppressing a response to an ensuing “A” (e.g., [21]). Further,studies of discourse comprehension, a level of languageprocessing that places relatively high demands on controlmechanisms, have shown that schizophrenia patients areimpaired at building coherent discourse representations [44]and are unable to effectively integrate global discourse-level context with local word-level context [45]. Thesestudies of online discourse comprehension are supported byseveral offline studies showing that schizophrenia patientsare unable to make use of discourse coherence when recallingdiscourse content [40–43].

The evidence reviewed above shows that deficits indiscourse comprehension in schizophrenia can be accountedfor by a domain-general deficit in cognitive control. Asmentioned in the introduction, several theories convergeon control mechanisms as related to the range of cognitivedeficits seen in schizophrenia patients [5, 6, 39]. Therefore,an important question concerns specifically what aspects ofcontrol affect discourse comprehension in schizophrenia. Wesuggest that the current evidence is consistent with the ideathat discourse comprehension deficits in schizophrenia resultfrom (1) deficits in the controlled maintenance of context-relevant information, which may be mediated by dysfunc-tions of the DLPFC and (2) deficits in the ability to resolveconflicting information in the language input, and theinability to monitor the incoming input for conflict, whichmay be mediated by dysfunction of the ACC. Followingfrom the Guided Activation Model discussed above, in whichprefrontal regions engaged in context maintenance guideactivation in the neural regions responsible for task exe-cution, prefrontal dysfunction in schizophrenia is expectedto result in processing and integration difficulties in theperisylvian-language network. Specifically, we suggest thatdeficits in the maintenance of context-relevant informationwill lead to impoverished discourse representations that areheavily reliant on word-level meaning relations. This kindof loosely integrated representation may suffice, providedincoming input does not introduce conflict. Therefore,discourse comprehension in schizophrenia patients shouldbe most successful when the words contained in the contextare related in meaning but will be more difficult for thesepatients when the semantic fit amongst individual words andsentences is poor.

5. Conclusions

In summary, we conclude that prefrontally mediated cog-nitive control mechanisms are impaired in schizophreniaand that during language comprehension this will mostlikely impact the integration and maintenance of context,


which involves (especially in the case of high-level languageprocessing) multiple levels of meaning. As context accumu-lates, demands on control mechanisms governing integrationand maintenance are increased, resulting in discourse-levelprocessing as a particularly demanding level of processingin terms of control. Further, as context accumulates, thepotential for multiple aspects of the input to conflictincreases; conflicting aspects of context maximally tap intocontrol processes. Two main predictions are generated fromthis hypothesis that may be tested as part of a future researchagenda using methods from the cognitive neuroscience oflanguage: (1) schizophrenia patients will be most impairedduring processing at discourse level; (2) the construction ofcoherent discourse representations will be most successfulwhen supported by strong relations amongst individualwords, and patients will be most impaired when aspects ofthe context conflict with each other.

Stemming from these predictions are several implicationsfor future approaches to the study of high-level languagecomprehension in schizophrenia. First, more research oncomprehension at the discourse-level is needed, with aparticular emphasis on studies that deliberately manipulatesources of conflict and competition within discourse context.As noted above, sources of conflict, such as word-levelambiguity, are not special challenges to the comprehen-sion system only found in laboratory settings, but insteadare rather commonplace during natural language compre-hension. Likewise, other challenges to the comprehensionsystem, such as the generation of inferences in order toconstruct coherent discourse representations, are also quitecommon and place demands on cognitive control resources.Therefore, approaches that target those aspects of languagethat are most demanding of control are likely to be fruitful inthe study of language deficits in schizophrenia. In addition,future emphasis on cognitive neuroscience techniques suchas EEG and fMRI, particularly those that utilize connectivityanalysis, will be of use in determining how maintenanceand control operations in prefrontal cortex mediate languageprocessing in schizophrenia.

References

[1] M. F. Green, “What are the functional consequences ofneurocognitive deficits in schizophrenia?” American Journal ofPsychiatry, vol. 153, no. 3, pp. 321–330, 1996.

[2] C. S. Carter, “Understanding the glass ceiling for functionaloutcome in schizophrenia,” American Journal of Psychiatry,vol. 163, no. 3, pp. 356–358, 2006.

[3] G. R. Kuperberg, D. A. Kreher, and T. Ditman, “What canevent-related potentials tell us about language, and perhapseven thought, in schizophrenia?” International Journal ofPsychophysiology, vol. 75, no. 2, pp. 66–76, 2010.

[4] M. Spitzer, U. Braun, L. Hermle, and S. Maier, “Associa-tive semantic network dysfunction in thought-disorderedschizophrenic patients: direct evidence from indirect semanticpriming,” Biological Psychiatry, vol. 34, no. 12, pp. 864–877,1993.

[5] M. A. Gernsbacher, K. A. Tallent, and C. M. Bolliger, “Dis-ordered discourse in schizophrenia described by the structure

building framework,” Discourse Studies, vol. 1, no. 3, pp. 355–372, 1999.

[6] G. R. Kuperberg, “Language in schizophrenia part 2: what canpsycholinguistics bring to the study of schizophrenia and viceversa?” Linguistics and Language Compass, vol. 4, no. 8, pp.590–604, 2010.

[7] G. R. Kuperberg, T. Deckersbach, D. J. Holt, D. Goff, and W. C.West, “Increased temporal and prefrontal activity in responseto semantic associations in schizophrenia,” Archives of GeneralPsychiatry, vol. 64, no. 2, pp. 138–151, 2007.

[8] M. A. Gernsbacher, “The structure-building framework: whatit is, what it might also be, and why,” in Language Compre-hension as Structure Building, Lawrence Erlbaum Associates,Hillsdale, NJ, USA, 1990.

[9] M. A. Gernsbacher and M. E. Faust, “The mechanism ofsuppression: a component of general comprehension skill,”Journal of Experimental Psychology: Learning, Memory, andCognition, vol. 17, no. 2, pp. 245–262, 1991.

[10] M. A. Gernsbacher, “Two decades of structure building,”Discourse Processes, vol. 23, no. 3, pp. 265–304, 1997.

[11] P. Hagoort, “The memory, unification, and control (MUC)model of language,” in Automaticity and Control in LanguageProcessing, A. S. Meyer, L.R. Wheeldon, and A. Krott, Eds.,Psychology Press, New York, NY, USA, 2007.

[12] J. M. Novick, J. C. Trueswell, and S. L. Thompson-Schill,“Cognitive control and parsing: reexamining the role ofBroca’s area in sentence comprehension,” Cognitive, Affectiveand Behavioral Neuroscience, vol. 5, no. 3, pp. 263–281, 2005.

[13] J. M. Novick, J. C. Trueswell, and S. L. Thompson-Schill,“Broca’s area and language processing: evidence for the cog-nitive control connection,” Linguistics and Language Compass,vol. 4, no. 10, pp. 906–924, 2010.

[14] E. K. Miller and J. D. Cohen, “An integrative theory ofprefrontal cortex function,” Annual Review of Neuroscience,vol. 24, pp. 167–202, 2001.

[15] E. Callaway and S. Naghdi, “An information processing modelfor schizophrenia,” Archives of General Psychiatry, vol. 39, no.3, pp. 339–347, 1982.

[16] D. L. Braff, “Information processing and attention dysfunc-tions in schizophrenia,” Schizophrenia Bulletin, vol. 19, no. 2,pp. 233–259, 1993.

[17] J. D. Cohen and D. Servan-Schreiber, “Context, cortex, anddopamine: a connectionist approach to behavior and biologyin schizophrenia,” Psychological Review, vol. 99, no. 1, pp. 45–77, 1992.

[18] T. A. Lesh, T. A. Niendam, M. J. Minzenberg, and C. S. Carter,“Cognitive control deficits in schizophrenia: mechanisms andmeaning,” Neuropsychopharmacology, vol. 36, no. 1, pp. 316–338, 2011.

[19] E. C. Dias, P. D. Butler, M. J. Hoptman, and D. C. Javitt,“Early sensory contributions to contextual encoding deficitsin schizophrenia,” Archives of General Psychiatry, vol. 68, no. 7,pp. 654–664, 2011.

[20] M. J. Minzenberg, A. R. Laird, S. Thelen, C. S. Carter, and D. C.Glahn, “Meta-analysis of 41 functional neuroimaging studiesof executive function in schizophrenia,” Archives of GeneralPsychiatry, vol. 66, no. 8, pp. 811–822, 2009.

[21] A. W. MacDonald, C. S. Carter, J. G. Kerns et al., “Specificityof prefrontal dysfunction and context processing deficits toschizophrenia in never-medicated patients with first-episodepsychosis,” American Journal of Psychiatry, vol. 162, no. 3, pp.475–484, 2005.

[22] D. M. Barch, C. S. Carter, T. S. Braver et al., “Selective deficitsin prefrontal cortex function in medication-naive patients


with schizophrenia,” Archives of General Psychiatry, vol. 58, no.3, pp. 280–288, 2001.

[23] W. M. Perlstein, C. S. Carter, D. C. Noll, and J. D. Cohen,“Relation of prefrontal cortex dysfunction to working memoryand symptoms in schizophrenia,” American Journal of Psychi-atry, vol. 158, no. 7, pp. 1105–1113, 2001.

[24] J. H. Yoon, M. J. Minzenberg, S. Ursu et al., “Associationof dorsolateral prefrontal cortex dysfunction with disruptedcoordinated brain activity in schizophrenia: relationship withimpaired cognition, behavioral disorganization, and globalfunction,” American Journal of Psychiatry, vol. 165, no. 8, pp.1006–1014, 2008.

[25] S. G. Potkin, J. A. Turner, G. G. Brown et al., “Workingmemory and DLPFC inefficiency in schizophrenia: the FBIRNstudy,” Schizophrenia Bulletin, vol. 35, no. 1, pp. 19–31, 2009.

[26] P. C. Fletcher, P. J. McKenna, C. D. Frith, P. M. Grasby, K. J.Friston, and R. J. Dolan, “Brain activations in schizophreniaduring a graded memory task studied with functional neu-roimaging,” Archives of General Psychiatry, vol. 55, no. 11, pp.1001–1008, 1998.

[27] G. R. Kuperberg, T. Sitnikova, D. Caplan, and P. J. Hol-comb, “Electrophysiological distinctions in processing con-ceptual relationships within simple sentences,” Cognitive BrainResearch, vol. 17, no. 1, pp. 117–129, 2003.

[28] G. R. Kuperberg, “Language in schizophrenia part 1: anintroduction,” Linguistics and Language Compass, vol. 4, no.8, pp. 576–589, 2010.

[29] N. C. Andreasen, “Thought, language, and communicationdisorders. I. Clinical assessment, definition of terms, andevaluation of their reliability,” Archives of General Psychiatry,vol. 36, no. 12, pp. 1315–1321, 1979.

[30] N. C. Andreasen, “Thought, language, and communicationdisorders. II. Diagnostic significance,” Archives of GeneralPsychiatry, vol. 36, no. 12, pp. 1325–1330, 1979.

[31] D. M. Barch, D. Servan-Schreiber, S. Steingard, J. D. Cohen, S.S. Steinhauer, and D. P. Van Kammen, “Semantic priming inschizophrenia: an examination of spreading activation usingword pronunciation and multiple SOAs,” Journal of AbnormalPsychology, vol. 105, no. 4, pp. 592–601, 1996.

[32] D. A. Kreher, P. J. Holcomb, D. Goff, and G. R. Kuperberg,“Neural evidence for faster and further automatic spreadingactivation in schizophrenic thought disorder,” SchizophreniaBulletin, vol. 34, no. 3, pp. 473–482, 2008.

[33] D. A. Kreher, D. Goff, and G. R. Kuperberg, “Why all theconfusion? Experimental task explains discrepant semanticpriming effects in schizophrenia under ”automatic” condi-tions: evidence from event-related potentials,” SchizophreniaResearch, vol. 111, no. 1–3, pp. 174–181, 2009.

[34] T. C. Manschreck, B. A. Maher, J. J. Milavetz, D. Ames, C. C.Weisstein, and M. L. Schneyer, “Semantic priming in thoughtdisordered schizophrenic patients,” Schizophrenia Research,vol. 1, no. 1, pp. 61–66, 1988.

[35] G. R. Kuperberg, P. K. McGuire, and A. S. David, “Reducedsensitivity to linguistic context in schizophrenic thoughtdisorder: evidence from on-line monitoring for words inlinguistically anomalous sentences,” Journal of Abnormal Psy-chology, vol. 107, no. 3, pp. 423–434, 1998.

[36] T. Sitnikova, D. F. Salisbury, G. Kuperberg, and P. J. Holcomb,“Electrophysiological insights into language processing inschizophrenia,” Psychophysiology, vol. 39, no. 6, pp. 851–860,2002.

[37] G. R. Kuperberg, T. Sitnikova, D. Goff, and P. J. Holcomb,“Making sense of sentences in schizophrenia: electrophysio-logical evidence for abnormal interactions between semantic

and syntactic processing,” Journal of Abnormal Psychology, vol.115, no. 2, pp. 251–265, 2006.

[38] T. Y. Swaab, K. Ledoux, C. C. Camblin, and M. Boudewyn,“ERPs and language processing,” in Oxford Handbook of Event-Related Potential Components, S. J. Luck and E. S. Kappenman,Eds., chapter 14, Oxford University Press, New York, NY, USA,2011.

[39] G. R. Kuperberg, “Neural mechanisms of language compre-hension: challenges to syntax,” Brain Research, vol. 1146, no.1, pp. 23–49, 2007.

[40] J. S. Lawson, A. McGhie, and J. Chapman, “Perception ofspeech in schizophrenia,” The British Journal of Psychiatry, vol.110, pp. 375–380, 1964.

[41] R. Levy and A. E. Maxwell, “The effect of verbal context on therecall of schizophrenics and other psychiatric patients,” BritishJournal of Psychiatry, vol. 114, no. 508, pp. 311–316, 1968.

[42] B. A. Maher, T. C. Manschreck, and M. E. Rucklos,“Contextual constraint and the recall of verbal material inschizophrenia: the effect of thought disorder,” British Journalof Psychiatry, vol. 137, no. 1, pp. 69–73, 1980.

[43] T. C. Manschreck, B. A. Maher, J. E. Rosenthal, and J. Berner,“Reduced primacy and related features in schizophrenia,”Schizophrenia Research, vol. 5, no. 1, pp. 35–41, 1991.

[44] T. Ditman and G. R. Kuperberg, “The time course of buildingdiscourse coherence in schizophrenia: an ERP investigation,”Psychophysiology, vol. 44, no. 6, pp. 991–1001, 2007.

[45] T. Ditman, D. Goff, and G. R. Kuperberg, “Slow and steady:sustained effects of lexico-semantic associations can mediatereferential impairments in schizophrenia,” Cognitive, Affectiveand Behavioral Neuroscience, vol. 11, no. 2, pp. 245–258, 2011.


Clinical Study

Stability of Facial Affective Expressions in Schizophrenia

H. Fatouros-Bergman,1, 2, 3 J. Spang,4 J. Merten,4 G. Preisler,2 and A. Werbart2

1 Karolinska Institutet, Department of Clinical Neuroscience, Center for Psychiatric Research, 171 77 Stockholm, Sweden2 Department of Psychology, Stockholm University, 106 91 Stockholm, Sweden3 Psychiatric Research Centre, 701 16 Orebro, Sweden4 Universitat des Saarland, 66 123 Saarbrucken, Germany

Correspondence should be addressed to H. Fatouros-Bergman, [email protected]

Received 13 November 2011; Revised 9 January 2012; Accepted 17 January 2012

Academic Editor: Marek Kubicki

Copyright © 2012 H. Fatouros-Bergman et al. This is an open access article distributed under the Creative Commons AttributionLicense, which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properlycited.

Thirty-two videorecorded interviews were conducted by two interviewers with eight patients diagnosed with schizophrenia. Eachpatient was interviewed four times: three weekly interviews by the first interviewer and one additional interview by the secondinterviewer. 64 selected sequences where the patients were speaking about psychotic experiences were scored for facial affectivebehaviour with Emotion Facial Action Coding System (EMFACS). In accordance with previous research, the results show thatpatients diagnosed with schizophrenia express negative facial affectivity. Facial affective behaviour seems not to be dependent ontemporality, since within-subjects ANOVA revealed no substantial changes in the amount of affects displayed across the weeklyinterview occasions. Whereas previous findings found contempt to be the most frequent affect in patients, in the present materialdisgust was as common, but depended on the interviewer. The results suggest that facial affectivity in these patients is primarilydominated by the negative emotions of disgust and, to a lesser extent, contempt and implies that this seems to be a fairly stablefeature.

1. Introduction

Studies of facial behaviour have shown a clear reduction offacial expressiveness and of facial affective expressiveness inpatients diagnosed with schizophrenia [1–11]. The reductionof facial expressiveness is especially prominent in the upperface [12] and has also been observed while patients are con-fronted with emotional stimuli [6, 10, 12, 13] as well aswhen they are imitating emotional stimuli [14]. Interestingly,most findings point at the conclusion that reduction of facialexpressiveness is not correlated with impaired emotionalexperience [8, 10, 11]. The patients’ capacity for emotionalrecognition is also reduced [11], a very robust findingaccording to a recent review [15]. However, facial emotionalexpressiveness is also reduced in patients with depression[6, 12], but patients with schizophrenia are still distinguishedin this respect from patients with depression, Parkinson’sdisease, and right hemisphere brain damage [10] since theirdiminished expressiveness is more prominent. Furthermore,

patients with schizophrenia have been found to limit theirfacial affective repertoire to mainly negative affective expres-sions [7, 11, 16], an observation that may be present evenbefore the clinical onset of psychosis [17]. Contempt wasfound to be the most frequent affect shown by these patients[5, 7, 18] and they showed significantly less happinesscompared to healthy subjects [5]. Healthy subjects, on theother hand, interacting with each other, usually show avariety of both negative and positive affects, and the mostfrequent affect shown by them is usually happiness [16].Overall facial activity has been reported as 2.3 times higherin interactions amongst healthy subjects compared to healthysubjects interacting with inpatients with schizophrenia [5, 7,19].

1.1. Objectives. The objective was to replicate previousfindings regarding negative facial affectivity in schizophreniaand to study the stability of these facial affective expressions.One focus was to examine whether facial affective behavior


Table 1: Patient characteristics.

Gender AgePsychosisdiagnosed

Schizophreniadiagnosed

Type of schizophreniaMedication interv.

A1, B1, C1

Medication interviewA2

M 27 1992 1997 Paranoid 295.30 Risperidone Risperidone

M 40 1987 1989 Paranoid 295.30 Olanzapine Risperidone

M 32 1991 2000 Undiff. 295.90 Perphenazine Perphenazine

M 52 1967 1971 Paranoid 295.30 Zuclopenthixol Zuclopenthixol

M 36 1994 1994 Paranoid 295.30 Olanzapine Olanzapine

F 48 1985 1985 Paranoid 295.30 Zuclopenthixol Zuclopenthixol

F 52 1972 1982 Paranoid 295.30 Ziprasidone None

M 27 1998 1998 Catatonic 295.20 Olanzapine Olanzapine

is dependent on temporality. The other focus was to testwhether the patients’s affective behavior is person dependent.The following hypotheses were tested.

(1) There are no substantial changes in the type of affectsthat the patients display across several interviewoccasions.

(2) The patient’s affects are not altered by the change ofinterview partner.

2. Methods

2.1. Subjects. Eight patients diagnosed with schizophreniaaccording to DSM-IV (1994) agreed to participate in thepresent study. Patients’ characteristics at the onset of thestudy are presented in Table 1. All the included patientscame from a psychiatric clinic with integrated in- andoutpatient care, had at least one past inpatient admissionsbut were not in an acute psychotic state at the time of theinterviews and either retired or unemployed. The patientsfirst received a letter with information about the study, andthey participated after having given their informed consent.The research project has been approved by the Committee ofEthics, Orebro County Council (605/00, 29/03).

2.2. Procedures

2.2.1. The Interviews. Thirty-two interviews were conductedin a clinical setting by two female psychologists both intheir early thirties and experienced in working with patientsdiagnosed with schizophrenia. An open interview model [20]was employed similar to interviews used in clinical settings.The interviews focused on the patients’ own problemformulation and background description to their currentproblems. The interview content centred often on psychoticexperiences as well as difficulties in the patient’s life. Thenarratives often included stories about mental suffering andreferences to the personal crisis stemming from psychoticexperiences and hospitalization. The interviewer adjustedthe questions to the patient’s answers. The interview wasdesigned to encourage mutuality in the interaction andelicit the patient’s own thoughts and perspectives. Eachpatient was interviewed three times by the first interviewer(interview A1, interview B1, and interview C1). The interval

between interviews A1, B1, and C1 was one week. Whenthese interviews were finished the patients were interviewedanother time by the second interviewer (interview A2). Eachinterview lasted about 45 minutes. Verbatim transcripts ofthe interviews included pauses and sighing.

2.2.2. Video Recording. The interviews were recorded bythree remote-control cameras. One camera focused on thepatient’s face, one on the interviewer’s, and one on the wholescene. The camera uptakes were synchronised and displayedon a single screen provided with time code. These were storedon DVD-discs and replayed including slow motion on acomputer using Power-DVD version 3.0 software.

2.2.3. Coding. The Emotion Facial Action Coding System(EMFACS) [21], a coding system for video data, was usedto evaluate the patient’s and interviewers’ facial affectivebehaviour. EMFACS is based on the Facial Action CodingSystem (FACS) [22] a coding system for registering allvisible facial mimic movements, called Action Units (AUs).Two sequences, 3 minutes each, where the patients weredescribing their psychotic experiences were selected fromeach interview. These narratives seemed emotional and wereassumed to trigger facial affectivity. First, the interviews werereplayed several times and all visible mimic movements (AU)were coded for the patient and interviewer. In a secondstep the AUs were interpreted with an emotional dictionarysoftware. This software compares the combinations of AUswith the EMFACS dictionary and registers seven affects:anger, contempt (AU 14, unilateral AU 12 or AU 14), disgust(AU 9 or AU 10), sadness, fear, surprise, and happiness(AU 6 + AU 12). Positive affects were operationalised ashappiness and surprise and negative affects as contempt,disgust, anger, sadness and fear. When referring to happinessonly Duchenne smiles were coded (AU 6+12) and not justsocial smiles (just AU 12). Both AUs and affects are briefand last approximately for milliseconds or seconds. A totalof 7.834 visible muscular movements (AU) were coded,corresponding to 1.805 facial affective expressions. The retestreliability for EMFACS has been reported to vary from 0.89to 1.00 and intercoder agreement from 0.87 to 1.00 [5]. Thevalidity has been confirmed in ethnological and ethologicalstudies [23, 24]. In the present study two independent codersat the University of the Saarland in Saarbrucken, Germany,


Table 2: Descriptive statistics of affects for patients (N = 8) in interviews A1, B1, C1 and interview A2.

Anger Contempt Disgust Fear Sadness Surprise Happiness

A1Mean 2.63 5.00 14.25 0.25 0.00 0.13 0.63

SD 5.04 8.21 15.04 0.71 0.00 0.35 0.74

B1Mean 2.25 7.75 14.00 0.38 0.13 0.38 0.88

SD 2.96 9.92 10.01 0.74 0.35 0.52 1.46

C1Mean 3.38 4.50 17.88 0.00 0.13 0.13 2.00

SD 5.40 5.40 14.82 0.00 0.35 0.35 2.51

A2Mean 3.63 14.13 11.38 0.00 0.50 0.13 1.63

SD 3.02 19.42 14.32 0.00 1.07 0.35 3.16

performed the coding. They did not have command of thelanguage spoken in the interviews since they had a differentlanguage background than the patients and the interviewersand were blind to the diagnoses of the patients. All interviewsfor each patient were coded by the same coder. As a reliabilitycheck, a sample of the interviews (6 minutes of interviewtime) was presented to both coders who scored an excellentintercoder agreement (κ > 0.80) on rating AU.

2.2.4. Statistics. The statistical package SPSS, version 15,was used. Descriptive statistics were used to summarisethe frequency and variation of the facial affects displayed.Within-subjects ANOVAs were performed to test whetherthe amount of affects varied across the interviews. Theaffects served as dependent variable with interview as thewithin-factor having three levels (interview A1, interview B1,interview C1). Wilcoxon tests for repeated measures wereemployed to test whether the patients displayed identicalaffects for both interviewers. The independent variable wasthe interviewer and the depended variable the affects shownby the patients. Finally Mann-Whitney U test was used tocompare the interviewers’ facial affectivity.

3. Results

Overall, facial affectivity in patients was predominantlynegative. Disgust, contempt and anger predominated (seeFigures 1 and 2 and Table 2 for descriptive statistics).

As the standard deviation in Table 2 shows, there wasconsiderable variation among patients regarding the fre-quency of affects. Furthermore a closer examination of thepatients’ affects in data from interviews A1, B1, and C1 revealsthat there was a considerable range in total affects displayed(see Figure 3) contributing to the high standard deviations.However, all patients displayed mostly negative affects.

The ANOVAs of data from the 24 interviews (interviewA1, interview B1, interview C1) with interviewer 1 revealedno significant effect among patients for any affect, (allF(2,14) ≤ 2.246, P ≥ 0.143) indicating that the amount offacial affectivity displayed by the patients did not varythroughout the interview series. Error variance for the pa-tients’ four most frequent affects was, anger (F(2,14) = 0.640),contempt (F(2,14) = 1.103), disgust (F(2,14) = 0.585), andhappiness (F(2,14) = 2.246).

Contempt23%

Disgust59%

Anger11%

Sadness0%

Fear1%

Surprise1%

Happiness5%

Figure 1: Percentage of patients’ facial affective expressions in 24interviews (interview A1, interview B1, interview C1) with inter-viewer 1.

Anger12%

Contempt45%

Disgust37%

Sadness2%

Fear0%

Surprise0%

Happiness4%

Figure 2: Percentage of patients’ facial affective expressions in 8interviews (interview A2) with interviewer 2.

Figure 4 shows the average number of affects for the fouraffects most frequently observed in the patients. The errorbars indicate standard error of the mean. Happiness wasinfrequent and not observed in 3 patients, which contributedto a lower variance for this affect. Disgust appears to increase,although not significantly.

The patients’ affects in interview A1 and A2 were thencompared. Although both disgust and contempt dominatedthe patients’ facial affects, the relative frequency differed. Forinterviewer 2, the patients displayed more contempt thandisgust; yet when interacting with interviewer 1 the same


0

20

40

60

80

100

120

1 2 3 4 5 6 7 8

Patients

Am

oun

t of

aff

ects

ContemptDisgust

AngerHappiness

Figure 3: Amount of affects per patient (N = 8) in interviews A1,B1, C1.

An

ger

A

An

ger

B

An

ger

C

Con

tem

pt A

Con

tem

pt B

Con

tem

pt C

Dis

gust

A

Dis

gust

B

Dis

gust

C

Hap

pin

ess

A

Hap

pin

ess

B

Hap

pin

ess

C

0

5

10

15

20

Mea

n±

SE

Figure 4: Average number of affects for the patients in 24 interviews(interview A1, interview B1, interview C1) with interviewer 1; errorbars indicate standard error of the mean.

patients displayed more disgust than contempt. However,the other affects displayed towards both interviewers didnot differ significantly (all P ≥ 0.18), except for the higheramount of contempt (z = 2.207, N-Ties = 6, P = 0.027, twotailed) shown to interviewer 2. Thus, with the exception ofcontempt, the second hypothesis was retained.

Both interviewers’ facial affective expressions were alsocompared with each other in the A interviews. With theexception of sadness, which bordered on significance (U =14.5, N1 = 8, N2 = 8, P = 0.06, two tailed), there wasno significant difference (all P ≥ 0.10) between the facialaffective expressions of the two interviewers.

4. Discussion

The patients showed elevated amounts of negative facialaffectivity. Patients most frequently displayed either disgustor contempt, followed by anger. These findings are inaccordance with previous research which found that negativefacial affectivity is to be expected in the facial activity ofpatients with schizophrenia [5, 7, 18]. However, contrary toprevious findings [5] the most frequent affect in the present

study was disgust as well as contempt. Due to the smallnessof the sample this should be interpreted with cautiousness.Previous research found disgust to be the second mostfrequent affect after contempt [5].

Regarding the stability of facial affective expressions wefound no substantial and systematic changes in the amountof affects the patients displayed at the various interviews,which suggests that affects appear stable and seem fairlyindependent of interview occasion. The patients’ affects werenot altered by the change of interview partner, with theexception that contempt was more frequently displayed withthe first interviewer. No conclusions can be drawn about thereason for this finding, since we cannot relate the resultsto changes in the pathological state of the patients due tolack of corresponding data. Though, we do not believe thatthis finding is due to differences in the interviewers’ facialaffective expressiveness, since we also analyzed their facialaffectivity and found no significant differences. We thenknow that their facial affective behaviour is similar to eachother.

Most of the patients in the present study remained onthe same antipsychotic medication at both data collections.To our knowledge there are no studies examining the impactof antipsychotic medication on facial affective behaviour inschizophrenia. Studies examining the effects of antipsychoticmedication on facial expressiveness in general show contra-dicting results [3, 4, 6, 25, 26]. However, a newly publishedreview claims that diminished expression in schizophrenia isobserved independently of medication [10]. Nonetheless, ifthere is an impact on facial expressiveness, it is improbablethat the type of facial affective expression should be selec-tively affected. While it is possible that the medicated patientsmay show reduced amounts of facial affective behaviour incomparison with unmedicated patients, it seems unlikelythat the quality of affects shown, whether it was disgust orhappiness, was affected.

As the sample was nonrandom and small, caution isadvised in generalising the results to the total population ofpatients diagnosed with schizophrenia. This type of in-depthinvestigations does not allow big samples thereby limiting thepower. On the other hand, the results are strengthened bythe repeated measurement procedure with the same patientsparticipating at several occasions. However, a limitation wasthat interviewer 2 only performed one interview (A2) thatwas compared with interview A1. However, since the analysisof interviews A1, B1, and C1 was performed first and showedthat facial affectivity in the patients seemed stable overtime, it was therefore assumed that interview A2 would berepresentative of a hypothetical series A2, B2 and C2 andcould be used alone without jeopardizing validity. Naturallytranscription and use of full series would have permittedmore robust comparisons.

Another limitation is the use of only two interviewersboth female. Facial affectivity may vary across genders. Alsothe uneven gender distribution among the patients mayraise questions with regard to results generalizability sincefacial expressivity has been found to be lower in healthymales than in females. However, findings in patients withschizophrenia have been contradicting regarding this matter


[9] and a newly published review [10] found only 3 studiesthat have been examining gender differences and emotionalexpressiveness in schizophrenia. Two of them found womento be more expressive and one found no differences.

Another limitation that may be crucial to assess the sta-bility of facial expressivity is the lack of the patient’s symptomprofiles and psychopathological state at the moment of theinterviews. It is, for example, possible that facial affectivitymay vary depending on whether the patients have positive ornegative symptomatology (Positive and Negative SyndromeScale, PANSS). Given the small sample size it was not possibleto analyse neither this aspect, nor gender-related interactionsbetween patients and interviewers.

The interpretation of the facial behavior as emotionsis based on the EmFACS dictionary by Paul Ekman, thathas been updated and validated by two of the authors (J.Merten, J. Spang) during the last 20 years, but not published.However, the authors (J. Merten, J. Spang) have conductedseveral decoding studies with various combinations of facialexpressions and have selected those with high recognitionaccuracy. Parts of this work are published in Merten (2005)[27].

The facial affective expression of contempt may involvefeelings of superiority as well as disdain for others [28].Disgust may be connected to feelings of aversion, oftenfor body products, but may also be evoked by sociomoralviolations [29]. Disgust is an emotion initially connected toevolutionary purposes, as it leads to avoidance of exposure topotentially harmful situations. It can be triggered by offensivesensations, such as smells, tastes, body products, contact withdeath but also by sociomoral violations [29]. Disgust mayalso be triggered by observing other people suffering or bytheir injuries, although this tendency diminishes with closerelationship to the person suffering [28]. Being the object forothers disgust is connected with shame. During the creationof intimate relationships, disgust is suspended and in theprocess of establishing personal intimacy individuals shareconfessions involving experiences or actions that otherwisewould have been the source of disgust and avoidance [28, 30].

Facial expressions may be expressions of underlyingemotional states. However, it is important to emphasizethat there is no isomorphic relation between facial affectiveexpressions and inner emotional states. We may, for example,smile when we are happy but also when we are feelinguncertain or ashamed. Facial affective expressions do alsoserve social purposes; they communicate social motivesand are driven by social intents. They are important toolsin the social interaction; they are present since infancyserving communicative purposes [31, 32]. Reduction offacial expressiveness has been found to related to poorerinterpersonal relationships having implications in othersocial domains [10]. Nonverbal behavior and facial affectiveexpressions signals motives such as the readiness to affiliateattack or continue current interaction and are stronglyrelated to the communicative features of the interaction,such as the topic of discussion. Facial affective expressionsmay also illustrate the affective components in the narrative.Consecutively, the aversive content of the interviews withfocus on psychotic experiences may partly explain the

prevalence of “contempt” and “disgust,” since those facialdisplays are connected to the current social context and alsoto the verbal content of the interaction and the questions ofthe interview.

Therefore, the disgust displayed in the present studymight be a reflection of the disgust felt for the sometimesawful experiences connected to the patients’ sickness andsuffering. As low self-esteem frequently can be observedin schizophrenia [33–35], facial affective expressions ofdisgust and contempt may also be seen as an expressionof self-disgust or self-contempt. Low self-esteem, feelingsof shame and guilt are frequently found in first-episodepsychosis [36]. The connection between low self-esteem andthe facial affective expressions of disgust and contempt inschizophrenia could therefore be a possible next step forfuture research.

To summarise, negative facial affectivity, mainly disgustand contempt, seems to dominate the facial expressionsof patients with schizophrenia in clinical interviews whilespeaking about psychotic experiences. In the present studythis finding seems to be a fairly stable feature independent oftime and interviewer. Due to the small sample these findingsneed to be further replicated.

Acknowledgments

This work was supported by grants from the PsychiatricResearch Centre in Orebro County Council (offered a fundeddoctoral position to the first author); the Committee ofResearch in Orebro County Council (Grant nos. 93/04,171/04, 7/04, 170/04); the Research and Development Unitin Orebro County Council; Karlakliniken in Orebro CountyCouncil; Marcus and Amalia Wallenberg Foundation (MAW2000.0055); the Lars Hiertas Minne Foundation and theCentre for Psychiatric Research and Education StockholmCounty Council and Karolinska Institutet.

References

[1] M. A. Aghevli, J. J. Blanchard, and W. P. Horan, “The expres-sion and experience of emotion in schizophrenia: a study ofsocial interactions,” Psychiatry Research, vol. 119, no. 3, pp.261–270, 2003.

[2] A. M. Kring and J. M. Neale, “Do schizophrenic patients showa disjunctive relationship among expressive, experiential, andpsychophysiological components of emotion?” Journal ofAbnormal Psychology, vol. 105, no. 2, pp. 249–257, 1996.

[3] A. M. Kring and K. S. Earnst, “Stability of emotionalresponding in schizophrenia,” Behavior Therapy, vol. 30, no.3, pp. 373–388, 1999.


[5] E. Steimer-Krause, R. Krause, and G. Wagner, “Interac-tion regulations used by schizophrenic and psychosomaticpatients: studies on facial behavior in dyadic interactions,”Psychiatry, vol. 53, no. 3, pp. 209–228, 1990.

[6] F. Tremeau, D. Malaspina, F. Duval et al., “Facial expressive-ness in patients with schizophrenia compared to depressed


patients and nonpatient comparison subjects,” The AmericanJournal of Psychiatry, vol. 162, no. 1, pp. 92–101, 2005.

[7] R. Krause, E. Steimer-Krause, and H. Hufnagel, “Expressionand experience of affects in paranoid schizophrenia,” EuropeanReview of Applied Psychology, vol. 42, pp. 131–140, 1992.

[8] C. G. Kohler, E. A. Martin, N. Stolar et al., “Static posedand evoked facial expressions of emotions in schizophrenia,”Schizophrenia Research, vol. 105, no. 1–3, pp. 49–60, 2008.

[9] G. Simons, J. H. Ellgring, K. Beck-Dossler, W. Gaebel, and W.Wolwer, “Facial expression in male and female schizophreniapatients,” European Archives of Psychiatry and Clinical Neuro-science, vol. 260, no. 3, pp. 267–276, 2010.


[11] F. Tremeau, “A review of emotion deficits in schizophrenia,”Dialogues in Clinical Neuroscience, vol. 8, no. 1, pp. 59–70,2006.

[12] W. Gaebel and W. Wolwer, “Facial expressivity in the course ofschizophrenia and depression,” European Archives of Psychiatryand Clinical Neuroscience, vol. 254, no. 5, pp. 335–342, 2004.

[13] I. Falkenberg, M. Bartels, and B. Wild, “Keep smiling! Facialreactions to emotional stimuli and their relationship to emo-tional contagion in patients with schizophrenia,” EuropeanArchives of Psychiatry and Clinical Neuroscience, vol. 258, no.4, pp. 245–253, 2008.

[14] B. L. Schwartz, J. Mastropaolo, R. B. Rosse, G. Mathis, and S.I. Deutsch, “Imitation of facial expressions in schizophrenia,”Psychiatry Research, vol. 145, no. 2-3, pp. 87–94, 2006.

[15] C. G. Kohler, J. B. Walker, E. A. Martin, K. M. Healey, andP. J. Moberg, “Facial emotion perception in schizophrenia: ameta-analytic review,” Schizophrenia Bulletin, vol. 36, no. 5,pp. 1009–1019, 2010.

[16] J. Merten and R. Krause, “What makes good therapists fail?”in Nonverbal Behavior in Clinical Settings, P. Philippot, R. S.Feldman, and E. J. Coats, Eds., Oxford University Press, NewYork, NY, USA, 2003.

[17] E. F. Walker, K. E. Grimes, D. M. Davis, and A. J. Smith,“Childhood precursors of schizophrenia: facial expressions ofemotion,” The American Journal of Psychiatry, vol. 150, no. 11,pp. 1654–1660, 1993.

[18] F. Schwab, J. Merten, and R. Krause, “Expressiveness indyadic communication,” in The (Non) Expression of Emotionsin Health and Disease, A. Vingerhoets, F. van Bussel, andJ. Boelhouwer, Eds., Tilburg University Press, Tilburg, TheNetherlands, 1997.

[19] R. Krause, E. Steimer, C. Sanger-Alt, and G. Wagner, “Facialexpression of schizophrenic patients and their interactionpartners,” Psychiatry, vol. 52, no. 1, pp. 1–12, 1989.

[20] A. Werbart and S. Levander, “Understanding the incompre-hensible: private theories of first-episode psychotic patientsand their therapists,” Bulletin of the Menninger Clinic, vol. 69,no. 2, pp. 103–136, 2005.

[21] W. V. E. Friesen, EMAFCS-7.(Unpublished manual), 1984.

[22] P. Ekman and W. V. Friesen, Manual for the Facial ActionCoding System (FACS), Consulting Psychologists Press, PaloAlto, Calif, USA, 1978.

[23] P. Ekman, E. R. Sorenson, and W. V. Friesen, “Pan-culturalelements in facial displays of emotion,” Science, vol. 164, no.3875, pp. 86–88, 1969.

[24] H. Hufnagel, E. Steimer-Krause, and R. Krause, “Mimischesverhalten und erleben bei schizophrenen patienten und beigesunden,” Zeitschrift fur Klinische Psychologie, vol. 20, no. 4,pp. 356–370, 1991.

[25] F. Schneider, H. Ellgring, J. Friedrich et al., “The effectsof neuroleptics on facial action in schizophrenic patients,”Pharmacopsychiatry, vol. 25, no. 5, pp. 233–239, 1992.

[26] A. Troisi, G. Spalletta, and A. Pasini, “Non-verbal behaviourdeficits in schizophrenia: an ethological study of drug-freepatients,” Acta Psychiatrica Scandinavica, vol. 97, no. 2, pp.109–115, 1998.

[27] J. Merten, “Culture, gender and the recognition of the basicemotions,” Psychologia, vol. 48, no. 4, pp. 306–316, 2005.

[28] P. Ekman, Emotions Revealed: Understanding Faces and Feel-ings, Weidenfeld & Nicolson, London, UK, 2003.

[29] J. Haidt, P. Rozin, C. McCauley, and S. Imada, “Body, psyche,and culture: the relationship between disgust and morality,”Psychology and Developing Societies, vol. 9, no. 1, pp. 107–131,1997.

[30] P. Rozin and A. E. Fallon, “A perspective on disgust,”Psychological Review, vol. 94, no. 1, pp. 23–41, 1987.

[31] C. E. Izard, C. A. Fantauzzo, J. M. Castle, O. M. Haynes, M.F. Rayias, and P. H. Putnam, “The ontogeny and significanceof infants’ facial expressions in the first 9 months of life,”Developmental Psychology, vol. 31, no. 6, pp. 997–1013, 1995.

[32] D. N. Stern, The Interpersonal World of the Infant: A View fromPsychoanalysis and Development Psychology, Basic Books, NewYork, NY, USA, 2003.

[33] D. Freeman, P. Garety, D. Fowler et al., “The London—EastAnglia randomized controlled trial of cognitive-behaviourtherapy for psychosis IV: self-esteem and persecutory delu-sions,” British Journal of Clinical Psychology, vol. 37, no. 4, pp.415–430, 1998.

[34] T. Lecomte, M. Cyr, A. D. Lesage, J. Wilde, C. Leclerc, and N.Ricard, “Efficacy of a self-esteem module in the empowermentof individuals with schizophrenia,” Journal of Nervous andMental Disease, vol. 187, no. 7, pp. 406–413, 1999.

[35] P. H. Silverstone, “Low self-esteem in different psychiatricconditions,” British Journal of Clinical Psychology, vol. 30, no.2, pp. 185–188, 1991.

[36] R. Miller and S. E. Mason, “Shame and guilt in first-episode schizophrenia and schizoaffective disorders,” Journalof Contemporary Psychotherapy, vol. 35, no. 2, pp. 211–221,2005.

Schizophrenia as a Disorder of...

Documents

Transcript of Schizophrenia as a Disorder of...