MASTERTHESIS,15hpMEDIAANDCOMMUNICATIONSTUDIESINTERNATIONALCOMMUNICATIONSPRING2016

Supervisor:FredrikStiernstedtExaminer:AndersSvensson

2016

ExercisingpowerthroughCSRcommunicationonFacebook

Insightsfromtheoilindustry

ELYSEVANZANDVOORT

-1-

Corporations are increasingly using social media as a tool for communicating Corporate Social

Responsibility. Marketing researchers have conducted ample research on the topic, however, a

communicationperspectiveismissing.Inordertofillthisgapandgainanuancedunderstandingofhow

corporations are communicating CSR and potentially enacting power relations, this thesis focuses on

linguisticelementsinCSR-relatedFacebookposts.AcontentanalysiswasperformedontheFacebook

pages of three oil corporations, covering a total of 120 posts. Results demonstrate that all three

companies aim for engagement with the audience, using various semantic and sensory interactivity

elements, and maintaining an informal writing style. Despite the latter seemingly contradicting the

assumption that corporations are enacting power, there are elements that support this claim. The

corporationsimplementaconstraintofcontentintheirpoststhroughtheuseofabstractwriting,and

includeaconstraintonpositionsthroughthenarrativestylesofaccountingandadvertisement,which

offer limited encouragement for participation. Regarding rhetoric, the ethos included in the posts

carriedsignificantinterconnectionwiththeCSRtopicsdiscussed,throughwhichthecompaniesseemed

to enforce constraints of content. Shell and Total emphasize certain environmental issues,while not

focusingonother impacts.BPdoesnotnottargetenvironmental issuesandmainlyhighlightspositive

socialimpact.Althoughtheenactmentofpowerisnotpresentineachofthelinguisticstructuresofthe

posts,thereareelementsthatindicatethepresenceofpowerrelations,whichcouldoffergroundwork

forfurtherresearch.

Keywords:CorporateSocialResponsibility;communication,socialmedia;discourse,power;oilindustry

JÖNKÖPINGUNIVERSITYSchoolofEducationandCommunicationBox1026,SE-55111Jönköping,Sweden+46(0)36101000

Masterthesis,15creditsCourse:MediaandCommunicationSciencewithSpecializationinInternationalCommunicationTerm:Spring2016

ABSTRACT

Writer(s):

ElysevanZandvoort

Title: ExercisingpowerthroughCSRcommunicationonFacebookSubtitle:Language:

InsightsfromtheoilindustryEnglish Pages:63

-2-

TableofcontentsTableofcontents...........................................................................................................................-2-

1.Introduction...............................................................................................................................-4-

2.PerspectivesonCSRandcommunication.................................................................................-5-

2.1CorporateSocialResponsibility...........................................................................................................-6-

2.2CSRcommunicationstrategies...........................................................................................................-7-

2.3Socialmediaasapersuasivetool........................................................................................................-8-

3.Aimandresearchquestions......................................................................................................-9-

4.Previousresearch....................................................................................................................-10-

4.1CSRcommunicationthroughsocialmedia........................................................................................-10-

4.1.1WhysocialmediaiseffectiveforCSRcommunication...............................................................-10-4.1.2UseofsocialmediaforCSRandrecommendations...................................................................-12-4.1.3CSRonsocialmediaandcorporatereputation..........................................................................-13-

4.2Gapsinresearchfield.......................................................................................................................-14-

5.Theory......................................................................................................................................-14-

5.1Socialmedia......................................................................................................................................-14-

5.2Onlinecommunication,poweranddiscourse..................................................................................-16-

5.2.1Communication.........................................................................................................................-16-5.2.2Powerrelations..........................................................................................................................-19-5.2.3Discourse...................................................................................................................................-20-

5.3Styleandrhetoric..............................................................................................................................-22-

5.3.1Style...........................................................................................................................................-23-5.3.2Rhetoric.....................................................................................................................................-24-

6.Method....................................................................................................................................-26-

6.1Datacollection..................................................................................................................................-27-

6.2.1Companies.................................................................................................................................-27-6.2.2Dataselection............................................................................................................................-28-

6.2Dataanalysis.....................................................................................................................................-29-

6.2.1Textualstyles.............................................................................................................................-30-6.2.2Rhetoric.....................................................................................................................................-33-

6.3Qualityofthestudy..........................................................................................................................-35-

7.Results......................................................................................................................................-36-

-3-

7.1Textualstyles....................................................................................................................................-38-

7.2Rhetoric............................................................................................................................................-42-

7.3Summary...........................................................................................................................................-44-

8.Discussion................................................................................................................................-45-

8.1Textualstyles....................................................................................................................................-46-

8.2Rhetoric............................................................................................................................................-48-

8.3CSRtopics.........................................................................................................................................-49-

8.4Limitationsandfutureresearch........................................................................................................-50-

8.5Conclusion........................................................................................................................................-51-

References...................................................................................................................................-52-

AppendixA:Codebook................................................................................................................-59-

-4-

1.Introduction

From2014to2015,theamountofsocialmediaaccountsworldwideincreasedby12%,whichisequalto

anestimated222millionextra socialmediausers, and this number continues to grow (WeAre Social,

2015). From a corporate perspective, this means that a broader audience than before can now be

reachedthroughsocialmedia.Hence,companiesareincreasinglyusingsocialmedia(Kesavan,Bernacchi

& Mascarenhas, 2013) and are looking into new possibilities for corporate strategies. A considerably

growingstrategyisCorporateSocialResponsibility(Singh,2016).Fromnowoninthisthesis,Iwillusethe

abbreviation CSR for Corporate Social Responsibility. Coombs and Holladay (2012) define CSR as “the

voluntary actions that a corporation implements as it pursues its mission and fulfills its perceived

obligationstostakeholders,includingemployees,communities,theenvironment,andsocietyasawhole”

(p. 8). For the communication of CSR initiatives to the public, the appropriate tools should be chosen

(Fieseler,Fleck&Meckel,2009).Socialmediaisoneofthepossibletools.

Marketing scholars have found socialmedia, specifically Social Networking Sites (from now on

calledSNSinthisthesis),tobeaneffectiveplatformforcommunicatingCSR.Theystatethatitprovidesa

wayofmeetingstakeholders’expectations(Castelló,MorsingandSchultz,2013;Du,Bhattacharya&Sen,

2010; Dutot et al., 2016; Kesavan, Bernacchi&Mascarenhas, 2013) and generating positive corporate

reputations (Capriotti&Moreno,2007;Curley&Noormohamed,2014;Duet al., 2010;Kesavanet al.,

2013;Moretti& Tuan, 2013). As themajority of the researchers state, this ismainly attributed to the

social media characteristics of providing peer-to-peer communication (Korschun & Du, 2013; Schau,

Muñiz & Arnould, 2009) and including user-generated content (Korschun& Du, 2013; Tench& Jones,

2013).However,accordingtoCarrandHayes(2015),amyriadofscholarstendtosimplifythedefinition

ofsocialmedia,or thedefinition is toodiscipline-oriented, inthiscasemarketing-oriented.Duetothis,

previous research on communicating CSR through SNS is, according to communication research

standards,notsufficientlytheoreticallygrounded.

From a communication perspective, Castells (2009) states that the emergence of digital

communication enabled companies to reach the audience through various channels, anyplace and

anytime.Oneofthemaindevelopmentsisthat,asopposedtotraditionalmedia,socialmediaenablesa

more direct targeting of the audience,meaning that companies can adapt their communicationmore

specifically to the audience they aim to address (Castells, 2009). However, Castells (2009) also

emphasizesthatsocialmediaprovidesameansfortheregularcitizentogainamorepowerfulpositionin

-5-

dialogues, being able to reach a significantly larger audience than before. Therefore, social media

introducedanewmeanstoexercisepower,whichcorporationshavesignificantlyembracedintheirCSR

communication.Inthecurrentcapitalistsystem,CSRisgenerallyadoptedinamannerthatbenefitsthe

economicgrowthofacompany (Sandoval,2014).Hence, throughthecommunicationofCSRpractices,

companies aim to influence andpersuade the audience’ perceptionof the company (Ihlen, 2011). The

power relations that are exercised through CSR communication online are embedded in discourse, as

statedbyFairclough(1989),whodescribesdiscourseaslanguageusebasedonthecontextthediscourse

takesplacein.Inordertoidentifythesepowerrelations,thisstudywilltakeamediaandcommunication

approachandwillexploretheCSRdiscoursebyfocusingonthelinguisticstructuresoftheCSRdiscourse.

MarketingscholarspreviouslyestablishedthesuccessofCSRcommunicationthroughthesocial

media genre of SNS (Capriotti&Moreno, 2007; Curley&Noormohamed, 2014; Castelló,Morsing and

Schultz, 2013; Du, Bhattacharya & Sen, 2010; Dutot et al., 2016; Kesavan, Bernacchi &Mascarenhas,

2013;Moretti&Tuan,2013),which iswhythedata for this researchwascollected fromtheFacebook

pagesofthethreemostactiveoilandgascompaniesonFacebook:Shell,TotalandBP.Thesecompanies

belongtoaso-calledcontroversial industry(Du&Vieira,2012).WoolfsonandBeck(2005)considerthe

oilindustrytobecontroversial,duetothefactthattheindustryhasbeeninvolvedinpracticesthatmight

have a negative impact on a social, ethical and environmental level. Bhattacharya and Sen (2004)

emphasize that controversial companies face an enhanced challenge in communicating their CSR

strategy,becausestakeholderswilllikelyhaveaninitialapprehensiveapproachtoanyCSRinitiative.

InordertodevelopacomprehensiveunderstandingofCSRcommunicationthroughsocialmedia,

this thesis firstly elaborates on the concept of CSR, related strategies and using social media as a

persuasive tool for CSR. Thereafter, the aim and research questions of this study are presented, and

previousresearchonusingsocialmediaasatoolforCSRcommunicationareexplored.Followingisthe

theoretical background onwhich this study is built and themethod thatwas selected to conduct the

research. Finally, the results arepresented, and thediscussion sectionexpandsonhow the results are

linkedtoexistingtheoryonsocialmedia,discourseandpowerrelations.

2.PerspectivesonCSRandcommunication

To develop a comprehensive background, this thesis firstly addresses the concept of CSR, CSR

communicationstrategiesandusingSNSasatoolforpersuasion.

-6-

2.1CorporateSocialResponsibility

AcrosspreviouslyconductedresearchonthetopicofCSR,theconcepthasbeendefinedinvaryingways.

AsstatedbyCoombsandHolladay(2012),conceptualizingCSRisdifficult,duetothefactthatthereisnot

onedefinitionthatcomprisesCSRinitsentirety.However,basedonpreviousresearch,theyproposethe

following:“CSRisthevoluntaryactionsthatacorporationimplementsasitpursuesitsmissionandfulfills

its perceived obligations to stakeholders, including employees, communities, the environment, and

societyasawhole”(Coombs&Holladay,2012,p.8).Dutotetal.(2016)alsoemphasizethevoluntaryact

of CSR. Definitions given by Reilly & Hynan (2014) and the World Business Council for Sustainable

Development(1999)addtothisbynotonlyattendingtoperceivedobligationsandexpectations,butalso

toworkingtowardsgeneraldevelopmentof thesituationofstakeholders, includingemployeesandthe

entiresociety.TheEuropeanUnion(2011)givesamoreconcisedefinitionwhichrelatestotheprevious

ones: “CSR is the responsibility of enterprises for their impacts on society”. To conclude the

conceptualization, Nwagbara and Reid (2013) state that CSR concerns social, economic and

environmentalimpactsonthecommunity.Thisisknownasthe‘triplebottomline’,asalsomentionedby

CoombsandHolladay(2012).

What CSR inherently entails is that companies have the responsibility of acting socially

responsible towards society. Due to this, there is the need for corporations to adhere to seemingly

opposing goals of economic growth on the one hand and responding to expectations regarding social

responsibility on the other (Sandoval, 2014). However, as Sandoval (2014) problematizes, current CSR

theories do not take into consideration the actual relations between economic objectives and society.

Essentially,companieshavethemainaimofeconomicgrowth,whichisnotindirectlinewithgenerating

positive social consequences. Therefore, Sandoval (2014) proposes a new categorization for relating

social and economic objectives of companies, based onWolfgang Hofkirchner’s work (2002, 2003, as

cited in Sandoval, 2014). This categorization affiliates economic growth and social expectations in four

ways: reductionism, projectionism, dualism and dialects. In general, experts take a reductionist,

projectionistordualistapproachtotherelationshipbetweensocialandeconomicobjectives (Sandoval,

2014).Reductionismreferstoincorporatingsocialissuesinthebusinessstrategyofthecorporationand

using it in such a way that it can expand business. Dualism entails a maintained focus on economic

growth, with socially responsible activities on the side, such as philanthropism. Predominantly, these

-7-

approaches entail including social responsibility in the daily practices of a corporation, ensuring that it

does not negatively influence the economic growth by including social responsibilities in marketing

strategies, for example. Projectionism on the other hand focuses on subordinating the economic

objectives to social responsibility,making social responsibility themore important goal. However, this

approach,liketheformertwo,maintainstheassumptionthatthechallengeofincorporatingCSRcanbe

solved within the capitalist system, without debating capitalism entirely (Sandoval, 2014). Capitalism

focusesonprofitsandnotonsocialissues,meaningthatexcludingothereconomicsystemsautomatically

excludes an inherent interest in social consequences for society. Sandoval (2014) proposes the fourth

approach,dialects,asacounterargumenttoCSRwithinthecapitalistsystem.Thisapproachemphasizes

theactualoppositionbetweenandeconomicgrowthandsocial responsibility inaway that itdoesnot

categorizeCSRasaneconomicissue.Thisapproachcaninstigateotherperspectivesonthestructureof

thecurrenteconomyandthepositionofCSRwithinthissystem.

2.2CSRcommunicationstrategies

As discussed by Sandoval (2014), whether preferred or not, the most common approach to the CSR

debate remains communicating and spinning CSR practices in such a way that they directly benefit

economicgrowth.ThisisrealizedbyusingcertainCSRcommunicationstrategies.

Coombs and Holladay (2012) discuss the proactive vs. reactive communication strategy. A

proactive strategy entails implementing CSR initiative in the business plan before stakeholders have

voiced their expectations, while a reactive strategy occurs when CSR projects are taken on after

stakeholders show their opinion on the social responsibility of the company. In general, a proactive

approach is considered to lead to positive responses and an enhanced corporate reputation, while a

reactiveapproachdoesnot(Becker-Olson,Cudmore&Hill,2006).Nonetheless,Wagner,LutzandWeitz

(2009) found that thereareways–abstractnessand inoculation– tocommunicateCSRmessages that

reducethenegativityofreactiveCSR.Whencorporationscommunicateconcretemessagesonnegative

topicsthathavebeenconcretelydiscussedbyotherparties, it ismentallysimplifiedfortheaudienceto

comparethesemessagesandthereforegainarathernegativestance.However,whencorporationslimit

themselves to abstract reactive communication, the corporate message become less comparable to

those of other parties that were initially communicated, and the audience will develop less negative

emotions.Secondly,inoculationentailsthatbypresentingnegativeaspectstotheaudienceinanuanced

-8-

tonebeforeotherpartieshavepublishedthesameaspectsmorestrongly,whilesimultaneouslyrefuting

these,thedegreeofnegativityintheaudience’sopinionscanbedecreased.Notwithstanding,Leeetal.

(2009) state that, despite efforts to increase a positive attitude towards reactive CSR, the negative

responsesaremaintained.

AnotherapproachtocommunicatingCSRischoosingasoftversusahardstrategy(Tixier,2003).

This is built around the decision between communicating CSR through traditional communicationwith

relatively little involvement (soft approach) or through newmedia, technologies and advertising (hard

approach).LookingintoSNSasatoolforpersuasion,thehardapproachseemstobecomingincreasingly

favorable.SeeingthatthisthesisexploresCSRcommunicationthroughSNS,solelythehardapproach is

applicable.

2.3Socialmediaasapersuasivetool

With the rise of the Internet, there is a growing amount of channels for marketing strategies that

businessescanutilize(Gass&Seiter,2016;Nahai,2012;Simons&Jones,2011).Persuasionisarelevant

onlinepracticethatisalsousedinCSRcommunication.Internetisaninherentlyfasttypeoftechnology,

enabling persuaders to reach their target audience faster and with higher precision (Simons & Jones,

2011).Theyalsohave theability togain informationabout theneedsand likesof theaudience,which

helpsthemtoadapttheirmessagesmorespecifically.Asignificantchannelforreachingtheaudienceis

SNS.Overtime,individualshavetakenaratherapprehensivestancetowardsregularadvertisingandare

no longer strongly influenced by traditional marketing strategies. According to Friestadt and Wright

(1994),thisisanaturalphenomenon,coinedthechange-of-meaningprinciple.Whenpeopleareexposed

toacertainmarketingstrategyforalongerperiodoftime,theygettounderstandthestrategyandgivea

differentmeaning to it.Thisunderstandingmight result ina lesseffectivestrategy,which iswhysocial

networks have become an undeniable asset in persuading the audience (Gass & Seiter, 2016). Social

circlesentailadegreeofsocialinfluencethatpeoplearesensitive(Gass&Seiter,2016).Word-of-mouth

isdirectlyrelatedtosocialinfluenceandreferstothespreadingofamessagethroughsocialconnections

(Nahai,2012).WhenSNSusersfollowotherusers,andshareandliketheirmessages,thiscanbeseenas

a form of influence. Individuals are influenced by their friends’ and other connections’ messages and

posts,andmightcarrytheseon(Gass&Seiter,2016).

-9-

AccordingtoNahai(2012),socialmediaaffordspeopletomakedecisionsbasedon“reputational

capital”,inwhichpeoplemakedecisionsinbuyingandusingproductsandservicesbasedoninformation

theyhavegainedthroughsocialnetworks.Thisistheunfavorableaspectofusingsocialmediaastoolfor

persuasion,becauseithasresultedinanincreaseinaccountabilityforcompanies,meaningthattheyare

directlyandpubliclyheldaccountablebyconsumersthroughsocialmedia.Thishascompelledbusinessto

go online, and represent and potentially defend their brand. Perloff (2010) elaborates on the online

presence of companies, by focusing on the creation of online communities for persuasion. Instead of

sending out standardized messages to the audience, businesses are now instigating dialogues and

respondingonline,withtheaimofattractingacommunityaroundtheirbrandandpersuadeconsumers

intopositiveattitudes. AsCapriotti (2011) states, this is especially applicable for communicating social

responsibility.Thereisaswitchfromthe‘persuasive/informativediscourse’to‘dialogicdiscourse’inCSR

communication through SNS (Capriotti, 2011). This decreases the original type of persuasion and it

somewhatequalizesthepowerrelationshipsbetweencompaniesandstakeholders.

Thisthesiswillnotsomuchemphasizetheconceptofpersuasioninrelationtosocialmedia,but

will focus more on the enactment of power, which is closely related to persuasion. Therefore,

understandingsocialmediatobeapossibletoolforpersuasionisessentialincomprehendingitsdynamic

andinteractivecharacter.

3.Aimandresearchquestions

As aforementioned, this study explores CSR communication through SNS by taking a communication

approach and critically analyzing CSR-related posts on Facebook. The aim is to gain a more nuanced

understandingofCSRcommunicationonline,byanalyzingtheconceptsofsocialmediacommunication,

discourse,andpower.AsFairclough(1989)states,powercanbeexercisedthroughdiscoursebychoicein

language use, dependent on the context it takes place in. Focusing on linguistic elements in the CSR

communication on Facebook will lead to a comprehensive understanding of how corporations are

communicatingCSRandenactingpowerrelationsthroughdiscourseonline.Thelinguisticstructuresthat

willbeinvestigatedinthisresearcharestylisticandrhetoricalelements.

Previous researchon the topichasbeenpredominantly conductedbymarketing scholars,who

have proven to take a rather simplistic approach to the concept of social media (Carr and Hayes,

2015). This study will use thorough knowledge about socialmedia from amedia and communication

-10-

studies perspective and will therefore contribute to current literature on using SNS as tool for CSR

communication.

Theresearchquestionsthatthisstudyposesare:

RQ1HowaretextualstylesusedinCSRcommunicationthroughSNS?

RQ2TowhatextentisrhetoricpresentinCSRcommunicationthroughSNS?

4.Previousresearch

4.1CSRcommunicationthroughsocialmedia

Research from the area of communication studies that focuses on social media as a tool for CSR

communication is highly scarce andwas not detected during the review of previous literature for this

thesis.However,thereisasubstantialamountofmarketingresearchonthetopic.Elaboratingonthisis

relevantforgaininganunderstandingofpreviousresearchagainstwhichthisstudyplacesitself.

4.1.1WhysocialmediaiseffectiveforCSRcommunication

Overtheyears,extensivemarketingliteraturehasbeenconductedonusingsocialmediaasatoolforCSR

communication. One of the aims that a number of studies chose,was to investigate the reasonswhy

socialmedia can function as auseful tool forCSR communication. The first study tobediscussed that

took this direction is a study conducted by Kesavan, Bernacchi andMascarenhas (2013). By reviewing

previousliteratureandusingexamplesofCSRinitiativesexecutedbylargecorporations,theyfoundthat

social media is a powerful tool that should be adopted by corporations for communication with

stakeholders, and above all for communicating CSR. They concluded that social media for CSR is an

effective strategy inmeeting the needs of stakeholders all over the world, because it combinesmass

communicationwithpersonaldialogue.

CurleyandNoormohamed(2014)proposedanaimthatwassimilartoKesavanetal.’saim.Their

methodwasalsosimilar,becausetheyreviewedpreviouslyconductedresearchtoseewhysocialmediais

an effective tool for communicating CSR. They discussed CSR, social media as a marketing tool, and

combinedthetwotodrawaconclusion.ThestudyshowsthatinformingstakeholdersaboutCSRonsocial

mediamotivatesdialogue,loyalconsumerism,andcreatesapositivereputationforthecompany.

-11-

Castellóetal. (2013)elaboratedon the implicationsof socialmedia forCSRcommunicationby

focusingonthepolyphoniccharacteristicsofsocialmedia.Theyfoundthatcorporationswillnotfindone

commonopinionorcommonexpectationsonsocialmedia,butinsteadsocialmediaoffersaplatformfor

a myriad of expectations and needs of stakeholders. They conclude that corporations should

acknowledgethisfactandfindawaytointegratemultitudeviewsandperspectivesintheirCSRinitiatives

totryandmeettheexpectationsofthepublic.Then,theycanbuildastrongandbeneficialrelationship

withtheirstakeholders.

Jiménez-ZarcoandBicho(2015)tookadifferentpointofviewandaimedtoresearchifandwhy

socialmedia is effective from the perspective of stakeholders. Based on surveys thatwere distributed

among various stakeholders, their findings included that stakeholders perceive social media as an

importantmannerofcommunicatingCSR.Theyfeelsocialmediaisatrustworthynetworkandconsumers

mention that theirbuyingbehaviorcanbe influencedbysocialmediamessagingofsocially responsibly

companies.

NwagbaraandReid(2013)alsoincorporatedtheaimofthepreviouslydiscussedresearches,but

added how social media can be used efficiently. They used a literature review in their study to draw

conclusions on how CSR communication is shaped by social media. They found that because of the

increaseofsocialmediause,companiesshouldpaysignificantattentiontotheirpresenceonsocialmedia

networks andmake use of these networks to the full extent. Their recommendations for socialmedia

usageinrelationtoCSRincludetakingintoaccount:insight,acceleration,depth,issuemanagementand

democratization.

Moretti and Tuan’s study (2013) is strongly in line with that of Nwagbara and Reid (2013),

becauseitaimedtoresearchbothwhyandhowsocialmediaisaneffectivetool.However,theyused4

case studies of Italian companies as their methodology. Their findings showed that in essence, CSR

communication on social media is more beneficial for the corporate reputation than for the actual

performance of the business. Nonetheless, they also state that the direct communication with

stakeholdersmightcertainly influence theirbehavior innotonlyconsuming,butalso investing,or they

mightattractthemasfutureemployees.

Lee, Oh & Kim (2013) conducted a network analysis of Twitter accounts of Fortune 500

companiesandalso focusedonbothwhysocialmedia isaneffectiveplatformandhow it canbeused

efficiently.They found that socialmediaoffersusefulopportunities tocommunicatewithstakeholders.

Hence, this shouldbedoneextensively. It appears that communicatingCSRespecially strengthens and

-12-

increasesacompany’ssocialfollowing,morethanothertypesofcontent.Theyconcludethat,becauseof

socialmedia,socialresponsibilityhasbecomemoreattractiveforcompaniesthanbefore.However,they

donotstresstheneedforasocialmediapresencetothesameextentasotherresearchhasdone.

4.1.2UseofsocialmediaforCSRandrecommendations

AsNwagbaraandReid(2013),MorettiandTuan(2013)andLeeetal. (2013)partiallydid,anumberof

studies focusedexclusivelyonhowcompaniesare currentlyusing socialmediaasa tool andhow they

should be using it. Reilly and Hynan (2014) looked into current CSR communication practices of

companies and gave recommendations on how CSR should be communicated. The method of this

researchwasdifferentfromthosepreviouslydiscussed,becauseitconcernedanempiricalresearchthat

used16publiclytradedcompaniesfromdifferentsectorsbyanalyzingin-depththesocialmediaactivity,

and the formal organizational communication in the form of reports. Results showed that so-called

‘green’companiesrefertoCSRactivitiesmoreandhaveastrongersocialmediapresencethannon-green

companies.ElaboratingonthesocialmediapresenceinrelationtoCSR,ReillyandHynan(2014)gavethe

following recommendations: research CSR initiative of successful companies in the same industry; use

definite terms and go into detail when reporting about CSR; do not present the company in a more

sustainablewaythanit is;safeguardthecompany’sreputation;adheretoacorrectwayofsocialmedia

usage;stayup-to-dateaboutmediainnovationsanddevelopmentswithinthefield.

FieselerandFleck(2013)tookasimilarapproachandresearchedhowsocialmediahaschanged

CSRforcompanies.TheyconductedasocialnetworkanalysisoftheCSRblogsofthreeITcompaniesand

found that it is inevitable for companies nowadays not to get involved in social media. Web 2.0 has

becomeaprominentaspectof societyandoffers theopportunityofactivedialoguewithstakeholders.

TomaselliandMelia’sresearch(2014)was in linewiththepreviousone,because it looked intocurrent

successfulCSRpracticesandhowthiscanhelpdevelopfutureCSR.Theyusedcasestudyanalysistostudy

3companiesthathaveproventobesuccessfulintheirCSRcommunication,byfocusingonhowtheirCSR

practiceswerecommunicatedonalltypesofmedia.Theirfindingsshowedthatsocialmediastandsoutin

itseffectivenessasopposedtotraditionalmedia.Theyhighlightedthatcompaniesshouldpayattention

to social media and its characteristics of stakeholder participation, sharing and the two-way

communicationbetweencorporationsandstakeholders.

-13-

AlsoColleoni(2013)tookthisapproachbyconductingasocialnetworkanalysisof7CSRTwitter

accountsoflargecorporations.Theaimwassimilartothoseofthetwoaforementionedstudies,namely

researching how companies are currently using social media to communicate CSR and what

recommendations can be given. The study found that it is still challenging for companies tomeet the

expectationsofstakeholders.ItisrecommendedthatbylookingattrendingtopicsintheareaofCSRand

to investigate what CSR information stakeholders are searching for, a better response towards

stakeholders’needscanbedeveloped.

WardandWylie(2014)lookedintocurrentCSRpracticesofIrishcompaniesonsocialmedia.They

foundthat,basedonstakeholdertheory,companiesshouldusesocialmediatorespondtostakeholder

expectations in the field of social responsibility. They conclude that Irish companies do not use social

mediaforCSRpracticestotheextentthatwouldbeadvisable.

4.1.3CSRonsocialmediaandcorporatereputation

Pavitt (2012) took amore specific approach and aimed to research how CSR communication through

socialmedia specifically influences the reputation of companies. She researched previous literature to

elaborateontheabilityforstakeholderstovoicestrongopinionsandcomplaintsandhowthisinfluences

corporatereputation.Shefoundthatbecauseofinterconnectivityonsocialmedia,companiesareforced

to look into theirCSRand improve their social reputation. Socialmedia challenges the transparencyof

corporations and companies will not be able to keep up with competitors without responding to

stakeholders’needsandexpectationsthattheyvoiceonline.

Dutotetal.(2016)supportedthefindingsofPavitt’sstudybyconductingasocialmediacontent

analysisof4Frenchcompanies.Theyfoundthatcompaniesthatusesocialmediaasatoolseemtohavea

betteronline reputation thancompanies thatdonotorona lessactive level.However, theamountof

social activity does not have a significant impact on the reputation unless the social activity is directly

linkedtoCSR.

WilburnandWilburn (2015) looked specifically at strategies companies canusewhen theyare

beingcalledoutontheirsocialresponsibilitypracticesbystakeholdersandtheircorporatereputationis

diminished. They found that using third parties to validate their initiatives, using certificates and

partnering with nonprofit organizations are all good ways of responding to negative responses from

-14-

stakeholders.Theyconcludethatsocialmediahascausedtheneedforcompaniestocontinuallyattend

tostakeholdersneedsandexpectations.

4.2Gapsinresearchfield

Themajority of the aforementioned research has been conducted bymarketing scholars. Amyriad of

scholars,fromdifferentdisciplinesandspecialties,haveaimedtodefinesocialmediaasaconcept(Carr&

Hayes, 2015). As Carr and Hayes (2015) state, these definitions are often too focused on one or two

aspectsofsocialmedia,oraretoobrieftobeapplicableasanoverarchingdefinition.Thisalsoshowsin

theliteratureonCSRcommunicationthroughsocialmedia,inwhichingeneralmarketingscholarsgivea

rathersimplisticconceptualizationofsocialmediaandonlinecommunication.

ThisresearchwillfollowthestudiesonhowsocialmediaisusedasatoolforCSRcommunication

(Colleoni, 2013; Fieseler& Fleck, 2013; Reilly&Hynan, 2014, Tomaselli&Melia, 2014,Ward&Wylie,

2014),anddoesnotsomuch focuson thequestionof ‘is socialmediaasuccessful tool?’,which is the

othermaintopicformarketingstudiesinthisfield.Byusingknowledgeaboutsocialmedia,discourseand

power relations from amedia and communication scholar perspective, this studywill add to previous

researchonhowCSRiscommunicatedthroughSNS.

ThisstudywillalsofillanempiricalgapbyinvestigatingFacebookpagesofthreelargeoilandgas

companies.Analyzing these typeofcompanies,whichareconsideredcontroversial, througha focuson

theirFacebookpageshasnotbeendonethusfarinthefieldofCSRcommunicationonSNS.DuandVieira

(2012)conductedananalysisof theCSRon thewebsitesofoilandgascompanies,butdidnot include

socialmediaintheirstudy.

5.Theory

5.1Socialmedia

WiththeemergenceoftheInternet,newmediahavearisen.Newmediathatcanbedefinedas“mediaat

theturnofthe20thand21stcenturieswhicharebothintegratedandinteractiveandusedigitalcodeand

hypertext as technical means” (Van Dijk, 2012, p. 11). A new media field that allows for digital

communication issocialmedia.Thesemediaconnectpeopletosocialspacesandcanbe interactiveon

bothapersonalandmasslevel.

-15-

Conceptualizing social media is a demanding task. As Carr and Hayes (2015) state in their

researchondefiningsocialmediaasaconcept,thereisacurrentlackofanoverarchingdefinition.Some

definitionsrevolvearoundsocialmediamessaginganditsnature,focusingmostlyontheuser-generated

characterofsocialmediaandtherelevanceofdialogue(e.g.,Kaplan&Haenlein,2010).VanDijk(2012)

hasasimilarfocusonthedefinitionofsocialmediabydefiningitasapplicationsontheInternetthrough

whichcontentcanbeshared,whileemphasizingthetwo-waycommunicationthatcantakeplace.Other

definitions focus mainly on the technological aspect of social media, emphasizing devices, and social

mediaasatool(e.g.,O’Reilly,2005).Specificstudyfieldshavealsoaimedtodefinesocialmedia,suchas

Kent (2010) from a public relations perspective, who limited the definition to “any interactive

communication channel that allows for two-way interaction and feedback” (p. 645). However, the

majorityofthesedefinitionsleaveoutrelevantcommunicativeaspectsofsocialmedia,whichiswhyCarr

andHayes (2015)proposethefollowingnewdefinition:“Socialmediaare Internet-basedchannels that

allow users to opportunistically interact and selectively self-present, either in real-time or

asynchronously,withbothbroadandnarrowaudienceswhoderivevalue fromuser-generatedcontent

andtheperceptionof interactionwithothers” (p.50).This isanattempttoconceptualizesocialmedia

without being too broad, to technology-oriented or too discipline-specific. This study will use this

definitionasthedefiningbaseofsocialmedia.

Social media can be categorized in different genres, which include corresponding separate

channels. These genres are comprised of “content creation and publishing tools, social network sites,

contentsharingsites,collaborativeproductionssites,virtualworldsandadd-ons”(Lietsala&Sirkkunen,

2008,p.28).Thisexcludesgenresthatarebasedonone-to-onecommunication,becauseofthelackof

‘media’ in one-to-one communication networks. The tools used for creating and publishing content

includemainlyblogs,butalsopodcastsandwikis(Lietsala&Sirkkunen,2008).Blogsareonlinewebsites

thatusuallyrevolvearoundanauthor,orblogger,witharatherpersonalstyleofwriting.Thecontentcan

varyfromplaintext,tophotosorvideos.Ontheotherhand,wikis includeuser-generatedcontentthat

canalsobealteredandaddedtobyotherusers.Lastly,podcastsarecomprisedsolelyofvideoandaudio

material that is postedonline. Thenext genre is comprisedofwebsites that focuson sharing content.

ExamplesoftheseareFlickrandYoutube(Lietsala&Sirkkunen,2008).Onthesewebsites,userscanpost

respectively photos and videos, and comment on posted content. The third genre to be discussed is

virtual realities,whichenableusers toparticipate in virtualworlds.Here, the focus layson creatingan

online life that includes new experiences, enjoyment and the ability to socialize (Lietsala& Sirkkunen,

-16-

2008). Another socialmedia genre involves networks that provide thepossibility towork on the same

contenttogetherwithotherusers.BlufftonTodayisanAmericanwebsitethatisaclearexampleofthis

genre, as statedby LietsalaandSirkkunen (2008).Anewswebsite that letsuserspostnews itemsand

offers the possibility to work on news items together. In the same sense, Wikipedia might be a

collaborative production as well. However, as Lietsala and Sirkkunen (2008) state, Wikipedia focuses

specifically on publishing content and not somuch on the collaborative process ofworking towards a

desiredgoal,whichinthecaseofBlufftonTodaywouldbeacomprehensiveoverviewofthenews.The

lastgenre todiscussarewebsitesaimedatsocialnetworking,calledSNS.Thesewebsitesarebasedon

user interactionandusuallyentailcreatingaprofilethatgivesarepresentationofthepersonorgroup.

ExamplesareMySpace,FacebookandOrkut (Lietsala&Sirkkunen,2008).Thedatausedfor this thesis’

analysisiscomprisedofsolelysocialnetworkingcontentfromthewebsiteFacebook.

5.2Onlinecommunication,poweranddiscourse

As aforementioned, this thesis aims to develop amore nuancedunderstanding of CSR communication

through SNS by focusing on the interconnectedness between communication, power and discourse

online. These concepts will be elaborated on hereafter, after which an account will be given of the

theoretical notions of style and rhetoric. These notions are analyzed in an attempt to explore

communication,poweranddiscourse.

5.2.1Communication

Themedia landscapehasundergone several transformations that interact togetherandhave shapeda

new type of communication that is taking place online (Castells, 2009). The first transformation that

Castells(2009)elaboratesonistechnologicaltransformation.Inthecontemporarysociety,individualsare

able to communicate through varying channels, andenabledbydigitalization, there is awide rangeof

toolsthatcanbeused.Communicationcantakeplacethroughcomputer-mediatednetworksonbotha

localandgloballevel.Forcorporations,anewrangeofpossibilitieshasopenedupincommunicatingtheir

CSR practices to the public. Where mass media, that used to stay rather local, used to be the main

channelofcommunicationwiththepublic,newmedianowprovidecommunicationwithpeopleallover

theworld.Thesecondtransformationisofanorganizationalandinstitutionalnature,wherethesenders

are themedia and the receivers are the perceived audience (Castells, 2009).Media conglomerates all

-17-

over the world are using digital communication as a new channel for their strategies. They combine

traditionalandnewmedia.Theyoperategloballyaswellaslocallybymaintainingamessagethatcanbe

sentonamass levelwhilepersonalizingtheirmessagestospecificaudiences (Castells,2009).Although

Castells(2009)focusesonmediacorporations,thisorganizationalandinstitutionaltransformationcanbe

appliedtoamyriadofcorporations.Withtheriseofnewmedia,thesecompaniescanintegratealltypes

of media in their operations and use these however they see fit. Whether this is sending out mass

messages,oradaptingmessages to theaudience tosuchanextent that theyseempersonalmessages,

butaremass.Duetothis,CSRcommunicationcanbeadaptedtospecificaudiencesandthereisamore

substantial platform for reaching the audience through multiple channels at once. The third

transformation is theculturaldimensionofonlinecommunication.Castells (2009)states that therehas

beenadevelopmentof a global culture, aswell as adiversificationof cultural identity. Individuals feel

more connected to a global community, while simultaneously feeling close to sub-cultures that they

selectbasedontheirpersonal interests.Thelatterhaspromptedextensivenetworksofcommunication

thateachcoverarelativelimitedfieldofinterest,butwhichconnectasignificantamountoflike-minded

individuals.Tryingtodevelopoverarchingcommunicationforboththeglobalcommunityaswellasthe

sub-culturesishighlychallengingandmightleadtomiscommunication(Castells,2009).Corporationsalso

needtotakethisintoaccountwhenadaptingtheirCSRcommunicationtoaudiences.Theycanchooseto

clearlydeterminetheaudienceandunderstandwhataudiencesarepossiblyexcluded,orcompaniescan

choosetokeepthemessagesmoreapproachabletoeveryone.

Castells(2009)hascoinedthenewformofinteractivesocietal,ormass,communicationas‘mass

self-communication’. This entails potential interaction frommany to many, instead of communication

fromone toone,many tooneorone tomany.AsCastells (2009) states,his typeofcommunication is

mass,becausethereisthepossibilityofreachingawideaudience.Howeveritisalsoacommunicationof

theself,becausethemessagesthataresentareself-created,theaudiencetowhichitisdirectedisself-

selected,andinformationusedforthemessageisself-retrieved.VanDijk(2012)supportstheconceptof

massself-communicationandemphasizesthecombinationofinterpersonalandmasscommunicationon

socialmedia.Ashasbeentoucheduponpreviously in this theoretical framework, theconceptofself is

highlyrelevantinnewmediacommunication.Castells(2009)emphasizesthateveryindividualcanchoose

to createonlinemedia content, and include in that contentanything theywant.The selectionofwhat

contentisconsumedisalsospecifictotheself:anindividualchooseswhattheyreadorwatchandhow

theyinteractwiththatcontent,fromallthatisavailableontheInternet.Thelatterisparticularlyrelevant

-18-

for corporations communicating their CSRon socialmedia. The followers that they haveon Facebook,

havedeliberatelychosentofollowtheiractivitiesandtheirupdates.Thismeansthattheaudiencethey

arereachingwiththeircommunicationisratherone-sidedandlimited.It is interestingtolookintohow

theyarecommunicatingwiththeseaudiences,byfocusingoninteractivityandformality,forexample.

Thedisseminationof the Internet,Wi-Fi and thedigitizationof communicationhas reachedall

aspectsofdailylifeandhasthereforecreatedso-calledhorizontalnetworksofcommunication,whichcan

beaccessedanywhereatanytimewithamyriadoftools(Castells,2009).Usershavetheconstantoption

ofcreating theirownmasscommunication.Focusingspecificallyon theuseof socialmedia, individuals

participateinonlinenetworksforavarietyofreasons.Fenton(2012)elaboratesontheopennessofsocial

media being seen as a trigger for individuals to connect to society. These people might feel they

disappearinthetumultuousnessofmoderndaylifeandtheInternetgivesthemthespacetofeellikea

relevantpartofthecommunity.Corporationscouldusethisincreatingacommunitywheresolidarityis

keyandwhereindividualscanfeelpartofsomething.AnexampleofthisinCSRcommunicationwouldbe,

aiming to strive as a community for a betterworld. However, Fenton (2012) emphasizes that SNS are

inherentlyexpressiveandcommunicative,andnotsomuchrelatedtocivicengagement.Therefore,the

study field of communication should predominantly perceive SNS as related to the former. Thismight

counteracttheaimofengagingtheaudienceintheirpractices,butbecausethereisstillacertainamount

ofengagementtakingplaceonSNS,thiscanbedisregarded.

AsVanDijk (2012) states, individualsoftenenter SNSwith theaimof findinga so-called social

space, like Facebook, in which they can interact with others. The desire to be social is essentially a

personal, individual desire. Thishasbeendeveloped intonetwork individualizationbyWellman (2000),

whichCastells(2001)adoptedaswellinhisworkontheInternetgalaxy.VanDijk(2012)agreeswiththis

concept,statingthatitmeansthat“theindividual isbecomingthemostimportantnodeinthenetwork

societyandnotaparticularplace,groupororganization”(p.181).Thissupportsthegeneralprocessof

individualizationwhichhasbeen continuous inmodernday society.However, anetwork in its essence

seemstodefytheideaofindividualization,seeingthatitaimstoconnectindividualsandnotkeepthem

apart. Therefore, corporations aiming to create CSR-related dialogues on SNS might benefit from the

socialmotivationofuserstojoinFacebook.

Thenetworkofsocialconnectionsthatusersmaintainonlinecanbeaccessedanyplace,anytime

(VanDijk,2012). Individualshaveawiderrangeofsocialopportunitiesatthesametimeandthismight

lead to individuals spending an increasing amount of time online. This way, the online reality that is

-19-

created becomes intertwined with actual reality and daily life seems to exist both online and offline

(Castells, 2009; Fenton, 2012). A consequence of this is that the lines between real life and online

communicationfadeanditbecomesmoredifficultforpeopletounderstandwhatisrealandwhatisnot.

As Van Dijk (2012) states, individuals often use the Internet to build an identity, one that might be

significantly different from the identity that they show in real life. Moreover, companies use the

interpersonalcharacterofsocialmediatospecificallytargetaudiences,meaningthattheirmessagesmay

seem interpersonal but are actually part ofmass communication (Castells, 2009; VanDijk, 2012). This

together with the fact that sharing personal information is somewhat necessary for successful social

mediacommunication,makessocialmediaanattractivecommunicationchannelforbusinesses(VanDijk,

2012).Incommunicatingbusinessstrategies,suchasCSR,theaudiencecannowbereachedanyplaceat

anytime.

5.2.2Powerrelations

Whenlookingintotheemergenceofmassself-communicationandthetransformationscommunication

hasgonethrough,thereisaredthreadthatshowstheinfluenceofpowerrelationsineveryaspectofthe

development of digital communication, including social media communication (Castells, 2009). This

power is especially coming frommultinationalmedia corporations. They use digital communication in

theirday-to-daypracticesand introducetheirowncontentonamultitudeofcommunicationoutletsto

affectandinfluenceusers.However, it isnotaone-dimensionaltypeofcorporateinfluence.AsCastells

(2009)states,mediacorporationsareoperatingonagloballevel,onwhichtheyhavetotakeintoaccount

boththeglobalcommunityaswellasthediversificationofonlineculturalidentities.Therefore,theyare

creating strategies that target specific audiences, whilemaintaining the possibility to reach the entire

world.Thisshowsthatthereisnotsolelyonepowerrelationatwork,butratheramultitudeofdynamic

powerrelations.Moreso,eachindividualisuniqueandhas,throughthecurrentcommunicationsystem,

the choice in what content they consume. This enables a ‘creative audience’, which interprets and

processeseachmessageinitsownuniqueway(Castells,2009).Hence,theactualinformationtheytake

from the message differs per individual, which is why the power of media corporations is, although

significant,notall-encompassing.ThisissomethingthatcorporationsalsofaceincommunicatingCSR.It

canbestatedthatthereistheneedformessagestobeformulatedinsuchamannerthatitcanreachthe

majorityofthepeopleinapositiveway.

-20-

WiththeInternet,thepublic’sawarenessofcorporateandinstitutionalcontrolhasincreasedand

theemergenceofcitizengroupsthatvouchforlesscorporateinfluenceonthemediaisnotable(Castells,

2009;Fenton,2012).Juris(2008)agreeswiththisbyhighlightingthattheInternet itselfcanbeusedby

theaudiencetofightcorporateinfluencebyemphasizingtheirownculture,activitiesandbyunitingwith

like minded individuals. However, it is inevitable to take into account that an individual’s mind and

specifically the thought process is shaped by its surroundings, taking everything in, consciously and

subliminally (Castells,2009).With thedigitalizationof communication, stimuli frommediacorporations

are consistently present in daily life and communication becomes a channel of influencing individuals’

minds. Moreover, Fenton (2012) emphasizes the fact that the chance of encountering online spaces

whereusershavethecompletefreedomtocreateandprotecttheirownculturewithoutinfluencefrom

organizations and institutions is significantly small. Users might have little opportunity to defend

themselves against thepowerof corporations thatCastells (2009) talks about. Fenton (2012) refers to

political movements, which individuals can only express freely in online spaces, when there is no

suppressionorcontrol.Andevenifthesepoliticalmovementsexistprominentlyonline,thisisnotenough

forregularcitizenstosignificantly influencecurrentpoliticalsituations(Fenton,2012).Fromthiscanbe

deduced that online social communities, although relatively open, do not directly relate to a more

democraticsociety.Thiscanalsoberelatedtosocialresponsibility,becausecitizenshavemoreplatform

tocallcompaniesout fortheirCSRpractices.However,corporationscanusethemultitudeofchannels

thatisavailabletoreachaudiencesandimposetheirgoodbehavioronthem.

5.2.3Discourse

The means through which individuals, corporations and institutions can manifest power, is through

discourse(Fairclough,1989).AsSimpsonandMayr(2010)state,discoursecanbedescribedaslanguage

inaction.Itdoesnotsolelyrefertolinguistics,butratheroperatesinbroaderareasandisdependenton

situational,socialorculturalcontexts.Fairclough(1989)explainsthisbyconstruinglanguageindiscourse

asasocialstructure,aspartofhowsocietyuseslanguage.Inthisnotion,languageisasocialprocessand

itisdependentonotherelementsofsocietythansolelylanguage.Linguisticstructuresinpeople’suseof

languageistoacertainextentdeterminedbysociety,meaningthatthewaypeoplespeakisinfluencedby

socialphenomena(Fairclough,1989).Thisisalsothecasetheotherwayaround,seeingthatlanguageis

notsolelydeterminedbysocialinfluences.Rather,itisconstructedbyindividualsina,toacertainextent,

-21-

uniqueway.Dueto this,asFairclough (1989)states, language ispartofsocialconstructsand language

use can equally influence social phenomena. However, linguistic and social structures are not

commensurate.Linguisticconstructsremainapartofsociety,withtheabilityofequallyinfluencingsocial

phenomena (Fairclough, 1989). In the relationship between language and society, Fairclough (1989)

emphasizesthenotionofsocialorder.Socialorderisdefinedas“suchastructuringofaparticularsocial

‘space’ into various domains associated with various types of practice” (Fairclough, 1989, p. 29). In

discourse, people experience the social order that discourse takes place in. This means that the way

discoursesarebuiltisdependentonthesocialorderandrelationsofpowerbetweenthecommunicators

(Fairclough, 1989). In CSR communication through Facebook, potential discourse between the

corporations and the followers is therefore dependent on potential power relations between the two.

AccordingtoFairclough(1989),discoursecanbeusedtoexecutepowerrelationsbyfunctioningasthe

spacetodoso,byusing language inacertainmannerthat isbasedonthecontext thediscoursetakes

placein.Inthecaseofthisresearch,thecontextrevolvesaroundcorporationscommunicatingtheirCSR

practicestotheirfollowersonFacebook.

Powerrelationsoccuraroundconstraints;thedominantparticipantinthediscoursedetermines

andlimitsinfluenceoftheotherparticipants(Fairclough,1989).Theseconstraintscantakeplaceonthree

different levels,asFairclough (1989)describes.Firstly, themorepowerfulparticipantcanconstrain the

content of the discourse, and determines which topics and information are communicated. This is

especially the case in media discourse, as Fairclough (1989) explains. In mass media, there is a clear

distinction between producers and receivers, and the producers determine what topics are

communicated. However, in the time Fairclough (1989) wrote his book, social media was not yet

introduced,meaningthathisconceptionofproducerandreceiver inmediadiscoursedoesnotentirely

matchthecurrentmedialandscape.Regularcitizenscanjustaseasilybecomeproducersandtheoriginal

producers can become receivers of the citizens’ communication. In CSR communication through

Facebook,thismeansthatthefollowershavetheabilitytocomment,shareandpostaboutthetopicsthe

corporations are communicating. Moreover, taking into account Castells (2009) notion of mass self-

communication, it is relevant to highlight that the content users are consuming is to a certain extent

determined by the self. The consumers themselves decide what content they watch. As previously

mentioned,thispossibilityincreasesthepoweroftheregularcitizensandsomewhatequalizesthepower

tensionbetweencorporationsandconsumers.However,thenotionofdominantparticipantremains. In

anytypeofdiscourse,thereisamoredominantparticipantcreatingconstraintsinthepowerrelations.

-22-

The second level of constraint Fairclough (1989) identifies, revolves around the dominant

participant regulating the type of relationship that exists within discourse. In CSR communication on

Facebook, the corporation seems to determine that they are the communicating actor. They

communicate their information to the follower, and the follower is ought to take the position of the

receiver.Thisisatensioninthepowerrelationbetweenthecompanyandthefollower.

Lastly, themostpowerful participantdetermines theposition theotherparticipant can take in

thediscourse(Fairclough,1989).Acorporationcanconstrainthefollowersinwhatpositiontheyareable

totakeinthediscourse.UsingFacebookautomaticallymeansthatthefollowerscansolelycommenton

postsorwritetheirownpostontheirowntimeline.Theyarenotabletotakea leadingposition inthe

discourse.Throughcertainlanguageuse,thesepositionscanbeconstrainedevenmoresubstantially.By

using,forexample,formallanguage,thecorporationswouldnotnecessarilyencourageparticipationand

engagementofthefollower.Thecompanywouldremaindistantfromthefollowersandemphasizethe

powerdistance.

FromtheaforementionedcanbededucedthatcorporationsthatcommunicateCSRthroughSNS,

are able to exercise power through their discourse online. They can coordinate the topics that are

addressed,thepositiontheuserhasinthecommunicationofthesetopics,andtowhatextenttheuseris

invitedtoengageandparticipate.Thisisdeterminedthroughthediscoursetheycreateonline,orinother

wordstheuseoflanguagedependentonthecontextitoccursin(Fairclough,1989).

5.3Styleandrhetoric

As clarified by the previous elements of this thesis, power can be exercised through discourse. This

discourseentailslanguageasasocialpractice,meaninglanguageisusedbasedoncertaincontexts.Inthis

thesis, the analyzed discourse takes place in the context of corporations communicating their CSR

practices to their followers on SNS. Therefore, to analyze the language in this discourse equals

investigatinghowcorporationsexecutepower relations throughCSRcommunicationonFacebook.The

linguisticelementsthatareanalyzedinordertodoso,arestylisticandrhetoricalstructures.

-23-

5.3.1Style

There are different approaches to categorizing textual styles of texts. One of the more generic

approachesisanarrativecategorization.Humphreys,Gill,KrishnamurthyandNewbury(2013)identified

in their study on Twitter content the following narrative styles: accounting, commentary, information

seeking,contentsharingandresponse.Anaccountingstyleentailstheusergivinganaccountofhisorher

currentactivities,orinformationregardinghisorherownstatus.Acommentarystyle,however,refersto

a post inwhich the user comments on an event or subject, giving an opinion or expressing a feeling.

Whenuserssharecontentfromothersourcesthantheirown,thenarrativestyleisofacontent-sharing

nature.Furthermore, inpostswhereusersaskedorseemedtobe lookingforspecific information,they

usedainformation-seekingstyle.Lastly,contentcanbecategorizedasresponsestyle,whenitspecifically

entailsaresponsetocontentofanotheruser.Toacertainextent,alloftheseaccountingstylescanbe

attributedtoSNSingeneral.ThecategorizationofpoststylesonFacebookdevelopedbyCvijikj,Spiegler

andMichahelles(2011)hassomesimilaritieswithHumphreysetal.(2013).Theyidentifiedthestylesof

information, designed question, statements, advertisements, competitions, questionnaires and

announcements. Information,designedquestionand statements resemble respectively theaccounting,

information-seeking and commentary style as stated by Humphreys et al. (2013). The other styles are

mostly relatedtoorganizationaluseofFacebookand less to individuals.Lookingatnarrativestylescan

helpexplorethenatureoftheCSRcommunication.Arethecompaniesinclinedtowardssolelyupdating

their followers, or are they more invested in engaging and including their followers by posting

competitionsandquestionnaires?

Another type of textual style that can be analyzed is related to language use. Halliday (1978)

identified the concept of register, which entails a variety of language use based on the situation the

discourse takesplace in,alsodefinedby researchersas languagestyle.Analyzing the languageusecan

helpidentifyhowcorporationstalktotheiraudiences,thereforeviewtheiraudiences,andaboutpower

relations that they are aiming to exercise by creating either distance or closeness. Registers can be

categorizedondifferentlevels.Theycanbecategorizedbroadly,intermsofwrittenandspokenstyle,but

canalsobecategorizedmoredetailed(Biber,1995).Leckie-Tarry(1995)describesthefollowingregisters:

“interpersonalorideational,concreteorabstract,meaningasprocessormeaningasproduct,verbalstyle

or nominal style, informal or formal, implicit structures or explicit structures, unstable or stable,

-24-

propositional or lexical, situation-dependent or situation-independent, low hierarchization or high

hierarchization, and adult or child” (p. 64). Originally, formality was the most common method of

categorization(Biber&Finegan,1994),butotherclineshavebeenidentifiedoverthepastdecades.

Focusingonthedegreeof interactivity inatext isanotherwayofexploringtextualstyles.Sohn

(2013)developedamodelwhich identifies interactivityonthreedifferent levels:semantic,sensoryand

behavioral. These dimensions are based on the different experiences of an individual during a social

interaction.Firstly,interactioncantakeplaceonasemanticlevel,whenthetextthatiswrittenorspoken

speaks to people in such a way that they feel interacted with. This can be measured by looking at

elements suchas “personalizedmessages (e.g. ‘goodbye’or ‘thank you’) anduser-friendliness” (Sohn,

2014, p. 862). Lastly, sensory interaction is present when the text is designed to attract and engage

people. Measuring the sensory dimension is carried out by analyzing the appearance of the text and

determining whether it is, for example, “eye-catching, animation, colorful, sensation, and attractive

design” (Sohn, 2014, p. 862). Behavioral interaction is stimulated when the text includes elements of

“usercontrol,feedbackpossibilityandcustomizability”(Sohn,2014,p.862).Specificallymeasuringthese

interactivitydimensionsonsocialnetworkingprofiles,doesnotonlyoffer insightintotheuserinterface

of the social media, but will also allow for an exploration of the user interface in relation to the

experiencestheuserhasonsocialmedia(Sohn,2014).Thismightleadtoabetterunderstandingofhow

corporationscanusesocialmediatoengageandpersuadetheaudienceinfavouroftheirCSRinitiatives.

5.3.2Rhetoric

Aristotle identified three notions of using rhetoric for persuasion: logos, pathos and ethos (Corbett &

Connors, 1999). Logos refers to the use of reasoning, in which argumentation and logic is used to

persuade.Pathosontheotherhandentailstheuseofemotionalcuestoplayintopeople’semotionsand

persuadethem.Lastly,ethos involvesappealingtotheaudiencethroughpresentationofcharacterand

trustworthiness.However,asParker (1972) states, rhetoric isnot solely related topersuasion.There is

also a certain degree of manipulation that it revolves around. Human relationships are, to a certain

extent,manipulative at thebaseand thismanipulation is intertwinedwithpower.Wanting toexercise

power isequal towanting tomanipulate (Parker,1972).Moreover,possessingpower is somewhat the

possibility ofmanipulatingothers. Rhetoric ismanipulation in the sense that it is usually thepreferred

measure of exercising manipulation, or indirectly power (Parker, 1972). In this manipulation through

-25-

rhetoric,theconceptofadherencecanbeimportant.Thecommunicatoraimstoadheretothepublicto

its viewsand ideas.This isalso thecase forcompanies thatcommunicate theirCSRpractices.Through

rhetoric, they canaim toadhere theaudience to their ideasand tomake them in linewithwhat they

believe.

In relation to CSR, rhetoric encompasses the “textual strategies” that companies use to

communicatetheirpracticestothepublic,andpossiblymanipulatetheiraudience(Ihlen,2011,p.147).

The rhetorical approach can be used to explore how corporations communicate their CSR, and what

methodstheyusedtoensurethattheirpracticesappeargenuineandhonest.Somescholars,mainlyfrom

themarketingfield,conceptualizerhetoricinrelationtoCSRastheoppositeof‘reality’(Grushina,2008;

Ihlen,2011).However,fromacommunicationperspective,thisisnotentirelyaccurate.Althoughrhetoric

isusedasamethodforpersuasion,italsooffersanunderstandingofhowrealityisconstructedandhow

tomakesenseoftheworldthroughdiscourse(Grushina,2008;Ihlen,2011).AccordingtoIhlen(2011,p.

152),scholarsingeneralfocusoneitherofthosetwoconceptions,whichhecallsthecriticalandpractical

perspective.Criticalscholarsrefertorhetoricasbeingtheoppositeofrealityandperceiverhetoricasa

possibility to cover up dubious CSRpractices. The practical scholars, on the other hand, focus on how

rhetoricisusedinpracticeandacknowledgethattheyoftenhelpshapeanunderstandingofreality.

HigginsandWalker(2012)expandonlogosinrelationtoCSRcommunicationbyfocusingonthe

apparentlogicinsteadoftheactuallogic.Itisnotnecessaryforcorporationstobelogic;itissufficientfor

theirreasoningtoappearrational.Inidentifyinglogos,argumentationisahighlyrelevantnotion(Higgins

&Walker,2012).Well-arguedcommunicationcanappeartotheaudienceaslogicalandcanhelpinthe

act of persuasion. Presenting justifications, making claims, and providing data and examples, are also

mannersofpersuadingtheaudiencethroughtheuseoflogos.

Contrarily,byusingpathos, thecommunicatoraims toappeal to theemotionsof theaudience

(Corbett & Connors, 1999). In CSR communication, this usually relates to activating emotions and

specifically demonstrating identification with the audience (Higgins & Walker, 2012). The company

intendstomaketheaudiencefeellikeheorsheunderstandsthesituation,andtheneedsanddesires,of

thepublic.Besides identification, thecorporationcanusemetaphors tomake the informationand the

initiativesmorerelatabletotheaudience(Higgins&Walker,2012).

AccordingtoIhlen(2011),thethirdnotionofrhetoric,namelyethoscandebunklogicalreasoning

in CSR communication. This can be accomplished by showing good character and therefore directly

impacting the audience’s perspective on the credibility of the company. Ihlen (2011) identifies five

-26-

differentethosstrategies thatcompaniesadopt to improveormaintaincredibility.Theycanclaimthat

theyhaveimprovedtheworld,cleaneduptheirownact,careabouttheaudience,ortheycanfocuson

joiningtheCSRmovementandotherpartiespraisingthemfortheirpractices(Ihlen,2011,p.157).

6.Method

Themethod chosen for this study is content analysis.Neuendorf (2002) defines content analysis as a

“summarizing,quantitativeanalysisofmessagesthatreliesonthescientificmethod(includingattention

to objectivity-intersubjectivity, a priori design, reliability, validity, generalizability, replicability, and

hypothesistesting)andisnotlimitedastothetypesofvariablesthatmaybemeasuredorthecontextin

whichthemessagesarecreatedorpresented”(p.10).

According to Neuendorf (2002), the first important aspect of content analysis that can be

derived from this definition revolves around summarizingmessages in a quantitativeway. In general,

therearethreemaintypesofmessages:verbal,nonverbalandmediated.Inthisresearch,themessages

analyzed are mediated, namely messages in the form of Facebook posts. Regardless of the type of

message,theaimindoingcontentanalysisistocomprehendhowmessagesarecarriedout(Neuendorf,

2002). The aim of this research is to explore the power relations enacted in CSR communication on

Facebookbyanalyzingtheactualdiscoursetakingplace,which fits theobjectiveofcontentanalysis in

understanding how the communication takes place.Hansen andMachin (2013) emphasize the actual

countingdone in contentanalysis tobeable to formconclusionsabout themessages sentwithin the

areaofanalysisandhowthisissignificantonawider,socialscale.

Secondly,contentanalysisshouldhaveascientificmotivationandshould includethescientific

method in the research (Neuendorf,2002).Wrench,Thomas-Maddox,PeckRichmondandMcCroskey

(2013)stresstheneedforsubstantiatingtheuseofcontentanalysisinempiricalresearch.Thisresearch

stems from the scientific motivation of aiming to explore CSR communication on Facebook from a

communication perspective and the intent of gaining a comprehensive understanding of CSR

communicationonSNSingeneral.

Lastly,Neuendorf (2002) focuseson thedifferent fields that contentanalysis canbeused for.

Initially,itwasamethodspecificallydesignedforanalyzingtexts,whichmatchesthedesiredmethodof

thisresearch.However,therangehasbroadenedandcontentanalysiscannowbeusedtoanalyzealso

-27-

“transcribed speech, verbal interactions, visual images, characterizations, nonverbal behavior, sound

events,oranyothertypeofmessage”(Neuendorf,2002,p.24).

6.1Datacollection

Datawas collected from the Facebook pages of oil companies. This type of companywas chosen for

analysis,because itbelongs to thecontroversialoilandgas industry (Du&Vieira,2012).This industry

has a negative reputation considering environmental and social issues, and as Bhattacharya and Sen

(2004)state,havemorechallengesthanothertypesofcompaniesincommunicatingCSR.Thisresearch

offersinsightsintohowthesecorporationscommunicatetheirCSRpractices,beinginthepositionthey

arein.

6.2.1CompaniesThespecificcompaniesthatwereselectedareShell,TotalandBP.Thesecompaniesseemtohavethe

largestFacebookEnglish-speakingfollowing(figure1)and,afterageneral inquiryofthepage,seemto

post frequentlyon the topicof sustainability.Due to these characteristics, these companiesprovidea

convenientdatasampleforthisresearch.Moreover,thesecorporationsarelikelytoprovideinteresting

insights into practices of the oil and gas industry. Each of them owns a place in the 10 largest oil

companies in theworld, fillingthe6th,10thand5thpositionrespectively (Forbes,2016).Shellproduces

slightlylessoilthanBP,buthasthehighestvaluewith$216billion.Totalproducestheleastoil,butits

worthlaysinbetweentheothertwocompanieswith$136billion.BPisthemostsubstantialproducerof

the companies. However, its current value is not the highest with $121 billion, which is a significant

decrease from the$200billionvaluebefore theoil spill in theGulfofMexico in2010 (Forbes,2016).

ConsideringthemainissuesthatthecompaniesseemtoaddressintheirCSRstrategy,itcanbestated

that Shell predominantly aims at: safety, environmental performance, aiding theworld’s transition to

lowerCO2emissions,andworkingwithcommunities tounderstand their concerns (Shell,2016).Total

hasnotpublishedasustainabilityreportfor2015ontheirwebsite.However,intheirsustainabilityreport

of2014,theyprioritizedtheirinitiativesinthefollowingorder:safety,economicflexibility,energyneeds,

climatechange,promotionofnaturalgas,adaptingtotheworld’stransitiontolowerCO2emissions,and

innovation(Total,2015).Finally,BPfocusesitsCSRstrategyonthefollowingkeyissues:energychallenge

and climate change, their own operations, their own activities, safety, environment, and society (BP,

-28-

2016).Theissuestheypredominantlydiscussintheirsustainabilityreportaretheirownoperationsand

society,withdevotingrespectively7and6partsoftheirreporttotheseissues.Initially,thesepriorities

establish that Shell and Total both seem to significantly value safety and climate change, and Total

includes a substantial economic aspect in their report. BP seems to be less concerned with the

environmentandmoresowithsocietyandimprovingtheirownoperations,insteadofcontributingona

globallevel.Whetherthisisrelatedtotheoilspillin2010remainsunknown.Nevertheless,duetotheir

similarities and differences, BP seems to slightly deviate from the comparable strategies of Shell and

Total. These apparent discrepancies and similarities in CSR strategies offer an interesting data set to

furtheranalyzetheirCSRcommunication.

Figure1

6.2.2DataselectionThechoiceforanalyzing3companieswasbasedonthefactthatitfacilitatesananalyticalcomparison.It

offerstheabilityofcomparingandcontrasting3differentcompanieswithinthesameindustry,whichin

this case leads to comprehensive insights on how current companies in the controversial oil and gas

industryare communicatingCSR.AllCSR-relatedposts thatappearedon theFacebookpagesof these

threecompaniesoverthelastyear-fromApril30th,2015tillApril30th,2016-wereanalyzed.Thistime

periodwas chosen,because it gives anactual overviewof theCSRpostson the Facebookpages. The

eventualdatasetinvolved22postsforShell,58forTotal,andBPposted40posts,whichleadstoatotal

of 120 posts. The data was collected by making screenshots of all the CSR-related posts of every

company’sFacebookpage(forexample,seefigure2).

CSR-relatedpostswereselectedbasedonthetypologyofCSRcommunicationcontentasstated

byCoombsandHolladay(2012).TheycategorizethetypesofCSRcommunicationcontentinsocialand

environmental. Social CSR content encompasses “human rights, children, the immediate community,

labor rights,education,and/or thosewith financialormedicalneeds” (CoombsandHolladay,2012,p.

16) and environmental content concerns “waste reduction, sustainable forest harvesting, recycling,

-29-

noise reduction, restoration of indigenous plant life, and/or the sustainability of a manufacturing

process” (Coombs and Holladay, 2012, p. 16). For this research, Facebook posts that cover either of

thesetopics,wereconsideredtobeCSR-relatedposts.

Figure2

6.2Dataanalysis

SPSSwasusedastoolfortheanalysisofthedata.Afterinputofthevariables,frequencytablesandcross

tabulationswere run todisplayandanalyze the resultsof the study.Before theactual analysis, apilot

studywasconductedon10ofeachFacebookpage’sposts. Thishelped in theprocessof finalizing the

variablesforeach item,bydeterminingwhetherthevariablescouldbecodedsuccessfullyandwhether

allofthevariablestogetherprovidedthedesiredinformation.

Toanswerthefirstresearchquestion,‘howaretextualstylesusedinCSRcommunicationthrough

SNS?’, the following stylistic elementswere identified:narrative categorization, formality, abstractness,

andinteractivity.Thesecondresearchquestion,‘whatrhetoricispresentinCSRcommunicationthrough

SNS?’,wasansweredbasedontherhetoricalstructuresofethos,logosandpathos.Hereafterwillfollow

an account of these different elements and the variables that are assigned to these elements. For a

detailedguideoncodingeveryelement,seethecodebook(appendixA).

-30-

6.2.1Textualstyles

CSRtopics

CoombsandHolladay(2012)identifiedseveralformsofCSRthatcanbecommunicatedbycorporations.

These types of CSR are divided into social topics and environmental topics. Social: “human rights,

children, the immediate community, labor rights, education, and/or those with financial or medical

needs” (Coombs and Holladay, 2012, p. 16). Environmental: “waste reduction, sustainable forest

harvesting, recycling,noisereduction, restorationof indigenousplant life,and/or thesustainabilityofa

manufacturingprocess”(CoombsandHolladay,2012,p.16).Identifyingthesetopicsishighlyrelevantin

ordertodeterminewhattypeofCSRinitiativesaoilandgascompanycommunicates.Dotheylimitthe

communicatedinitiativestoenvironmentalcauses,whichtheyhaveasignificantnegativeimpacton(Du

&Vieira, 2012)?Or do they choose to take away the attention from these issues and decide to focus

moreothercauses?

Thedata selectionof this researchwasbasedon theaforementioned topics,meaning that the

postswereselectedonwhethertheycontaininformationaboutsocialissuesorenvironmentalissuesthat

the company seemed to commit to. The coder coded if the post is predominantly covering social or

environmental issues,afterwhichthespecific issueasdeterminedbyCoombsandHolladay(2012)was

identified.Afterapilot study, somealterationsweremade to thevariables.Fromthesocial topics, the

‘immediate’ community was changed to ‘specific community’, due to the fact that all the three

investigated companies are global companies, meaning that there is no clear immediate community.

Afterthepilot-testitshowedthatthecompaniessporadicallyaimtohelpaspecificcommunity,whichis

whyimmediatecommunitywaschangedtospecificcommunity.Alsovariable4wasaltered,from‘labor

rights’to‘employmentandlaborrights’,duetothefactthatthepilotstudyshowedanumberofposts

combininglaborrightsandemploymentintheCSRcommunicationofBP.

Narrativestyle

InordertocategorizeFacebookpostsinnarrativesandexplorewhattypeofnarrativecontentoilandgas

companies prefer to post online, the models developed by Humphreys et al. (2013) and Cvijikj et al.

(2011)werecombined.Asaforementioned inthetheorysectionofthisthesis,Humphreysetal. (2013)

established a model that is comprised of the following narrative styles on Twitter: “accounting,

commentary, information seeking, content sharing and response” (p. 422). Twitter is an inherently

-31-

different socialnetworking sitewithadifferentuser-interface thanFacebook. The style ‘response’was

not included in thisanalysis,becausethere isnooptiononFacebooktoresponddirectly tocontentby

postingapostonyourownpage.Thepostshouldbepostedonthepageofthepersonororganization

thattheresponseismeantfor.Moreover,Humphreysetal.(2013)studiedtheTwitterprofilesofalltypes

ofusers,notcompanies.Afterthepilotstudy,thestyleofcontentsharingwasexcludedaswell,because

theinvestigatedcompaniesdidnotseemtoshareothercompanies’content.

Cvijikjetal.(2011)specificallylookedintobusinessuseofFacebookandwasthereforehigherin

comparabilitywiththisresearch.Theircategorizationwascomprisedof:“information,designedquestion,

statements, advertisements, competitions, questionnaires and announcements” (Cvijikj et al., 2011, p.

812).Information,designedquestionandstatementscorrespondedrespectivelywithHumphreysetal.’s

(2013)categoriesofaccounting, informationseekingandcommentary.Afterthepilotstudyandabrief

review of the Facebook pages, the category of questionnaires did not seem to apply. However, the

advertisements,competitionsandannouncementsstylesweredeemedappropriatefortheanalysis.The

final categorization therefore was comprised of: accounting, information-seeking, commentary,

advertisement,competitionandannouncement.

Interactivity

Inrelationtopowerandonlinediscourse,interactivityplaysanotablepart.Jarrett(2008)statesthatwith

the emergence of interaction online, individuals are enabled to take a stronger stance against power

relations.ThisisinlinewiththeassumptionsofCastells(2009)andFenton(2012)regardingindividuals’

position against power through social media. By analyzing the degree of interaction that oil and gas

companiesencourageontheirFacebookpage,potentialpowerrelationscanbeexplored.

The level of interactivity on Facebook was measured based on a model developed by Sohn

(2013), which divides online expected interactivity in three main dimensions: sensory, semantic and

behavioral.Sensory interactivity refers toastimulationof thesenses. Incaseofonlinecommunication,

this entails vision and hearing, because the other senses cannot be stimulated online. Semantic

interactivityincludeslanguagethatispleasantforthereaderandduetowhichthereaderfeelsinteracted

with.Behavioral interactivityontheotherhandfocusesmainlyonthepossibilitiesthecontentgivesfor

customizing and responding. This level of interactivitywas excluded from this research, because every

Facebook post affords the same degree of behavioral interaction (e.g. commenting, voting, sharing).

-32-

BasedonSohn(2013),thevariables ‘image’, ‘video’and‘interactivetext’wereusedto identifysensory

activity. ‘Personalized messages’, ‘user-friendly language’, and ‘identification of the audience’ were

attributedtosemanticinteractivity.

Besidesthenarrativecategorizationandinteractivity,thereareotherstylesthatarerelevanttodiscussin

relationtoCSRcommunication.Leckie-Tarry(1995)compiledarangeofregisters,orvarietiesoflanguage

uses, to classify discourse: “interpersonal or ideational, concrete or abstract, meaning as process or

meaning as product, verbal style or nominal style, informal or formal, implicit structures or explicit

structures,unstableorstable,propositionalorlexical,situation-dependentorsituation-independent,low

hierarchizationorhighhierarchization,andadultorchild” (p.64).Asmentionedbefore inthis thesis,a

register is relevant in relation toCSRcommunication,because itdescribescertainvarieties in language

useasbeingdependentonthesituationalcontext.Inthiscase,theaimistoanalyzeCSRcommunication

onFacebook,whichconstitutesaspecificsituation.Theregistersthatwerechosenfortheanalysiswere

formalityvsinformalityandconcretevsabstract.

Formality

The first register tobediscussed is formality.According toFairclough (1989),powercanmanifest itself

through formal language. This is due to the fact that it limits the content that the receiver of the

communication isexposed,orcanbeexposed, to. Italso limits the individuals thatare invited into the

dialogueanditcreatesadegreeofpoliteness.AsFairclough(1989)states,politenessstemsfromdistance

inpower,meaningthatformallanguagecouldimplyahigherdegreeofpowerdistance.

Todetermineaformalorinformalwritingstyleforeachposts,amodeldevelopedbySheikhaand

Inkpen (2012)was used. Thismodelwas developed to satisfy the need of a concrete classification for

identifying formal and informal texts online. Sheikha and Inkpen (2012, p. 7) established the following

specificcharacteristicsofformalstyle:impersonalstyle,passivevoice,complexwordsandsentences,no

contractions,notmanyabbreviations,businessand technicalvocabularies,politewords,objectivestyle

with facts and references, no vague expressions or slang. And they stated the following informal style

characteristics: personal style, active voice, short simple words and sentences, contractions,

abbreviations, phrasal verbs,words that express rapport and familiarity, subjective stylewith opinions

andfeelings,vagueexpressionsandslang.

-33-

Abstractness

The second register used for this research is abstractness of discourse content. For the analysis of

abstractnessof theFacebookposts, the researchofWagneretal. (2009)wasusedasa startingpoint.

Wagner et al. (2009) states that concrete communication includes details, statistics and third-party

sources. Abstract information, on the other hand, gives a rather superficial account of an issue or

situation,withoutpresentingdetails,statisticsorothersources.Asaforementionedinthetheorysection

of this thesis, CSR companies generally tend to account for their positive impacts indetail, or inother

words, concrete (Wagneret al., 2009).On theotherhand,negativepractices are communicated in an

abstract way, with significantly less details while possibly trying to frame it as less negative. By

communicating an issue abstractively, a substantial amount of information is left out. This might be

informationthatcouldnegativelyalteranindividual’sopinionsignificantly.

Oil and gas companies have had significant negative impacts on the environment and social

situations(Du&Vieira,2012).Therefore,exploringtowhatextentoilandgascorporationsuseconcrete

or abstract styles of communicating their CSR practices to the public,might give insight intowhether

corporations prefer to provide the public with a deeper or more superficial understanding of the

initiatives.

6.2.2Rhetoric

ToexplorerhetoricinCSRcommunication,allthreedimensionsofrhetoricwereanalyzed:logos,pathos

andethos.

Logos

As aforementioned, logos refers to argumentation and logical reasoning to appeal to the audience

(Corbett&Connors,1999).HigginsandWalker(2012,p.198) identified intheirresearchonrhetoric in

CSR communication, the following aspects of logos in CSR discourse: “argumentation, logic,

warrants/justification, claims, data and evidence/examples”.Due to the limitations of content analysis,

notallofthesenotionscouldbeanalyzed.Firstly,claimswouldsomewhatoverlapwiththeethosthatis

researchedandmakingadistinctionbetweenlogos-orientedclaimsandethos-orientedclaimswouldbe

highly complicated and expectedly too time-consuming for this research. Secondly, it is difficult to

-34-

quantify logic and argumentations by assigning specificword choices to the concepts; the possibilities

would be endless. Moreover, argumentation and logic are both concepts prone to subjectivity. In a

quantitative analysis, it is desirable tomaintain asmuch objectivity in the coding process as possible.

leavingittointerpretationofthecoderwouldleadtoasignificantdegreeofsubjectivity.Therefore,the

notions that were possible to quantitatively explore where selected: warrants/justifications, data and

evidence/examples.

Pathos

Pathosinvolvestheuseofemotionalcuestoaimtopersuadetheaudience(Corbett&Connors,1999).In

relationtoCSRcommunication,HigginsandWalker(2012)acknowledgedidentificationandmetaphorsas

themain identifiersofpathos.Themetaphor isanelement thatcanbequantitatively investigatedand

was therefore used as an element in the analysis of this study. However, identification is a more

complicated and subjective concept. A feeling of identification in the text can differ from person to

person. By using the research of Higgins andWalker (2012) as a base, the coder focused on cultural

referencesandwordsrelatedtoafeelingofcommunitysuchas‘we’and‘sharing’,andemotion-related

words suchas ‘proud’ and ‘happy’. Thisway, this thesis attempted toquantitativelyoperationalize the

conceptofpathos.

Ethos

Ethosreferstoimpliedethicsthatanindividualorinstitutioncommunicates(Ihlen,2011).Byimplyinga

certaindegreeofcharacter inthecommunication,thereceivergainsafeelingofunderstandingofwho

the communicator is and its trustworthiness and reliability. In theCSR communicationof controversial

industries, suchas theoilandgas industry, trustworthinesscanbechallengingtoachieve (Du&Vieira,

2012).Therefore,ethosprovestobearelevantassetinthesuccessfulcommunicationofCSRpractices.

As aforementioned, Ihlen (2011) found five ethos strategies that corporations generally apply when

communicatingCSRtothepublic: theyclaimthat theyhave improvedtheworld,cleaneduptheirown

act,careabouttheaudience,ortheyfocusonjoiningtheCSRmovementandotherpartiespraisingthem

fortheirpractices(Ihlen,2011,p.157).Duetothefactthatthesenotionsweredevelopedfortheanalysis

of CSR communication, there was no need for alteration of the variables. The Facebook posts were

matchedwitheitherofthesestrategies.

-35-

6.3Qualityofthestudy

Todeterminethequalityofthestudy,therewasanumberofaspectsthathadtobetakenintoaccount.

Firstly, the coding of the data was executed by one coder. Various elements that were analyzed are

subjecttointerpretationandwerecodedbasedonthepersonalpreferences,opinionandintuitionofthe

coder. The reliability of the research remains unexplored. Including solely one coder in the research,

limitedpotentialreliabilitytestingtoachievingstability,whichinvolvesthedesignoftest-rest.Thisentails

a repetition of the coding procedure, usually after a certain amount of time has elapsed, to explore

possibleintraobserverinconsistencies(Krippendorf,2013).Thelimitedtimeframeinwhichthisresearch

tookplacemadethisunachievable.

Thevalidityoftheresearchreliesparticularlyonfacevalidity,whichentailsthattheresearch is

plausibleand“makessense”,asKrippendorff(2013)coins it.Fairclough(1989)hasfoundpowercanbe

exercised through discourse, in other words, language use depending on the context. Therefore,

investigating linguistic structures in CSR communication on Facebook makes sense with the aim of

exploringtheenactmentofpower.Contentanalysisregularlycountsonfacevalidity,duetothefactthat

analysisoftextsentailshowtextsareinterpreted,relyingoncommonsense(Krippendorff,2013).

Thevariablesforthe items4,5,6,and15arenotmutuallyexclusive,meaningthataFacebook

post could possibly be coded with multiple variables. In these cases, the codebook was specified

thoroughly and the coder was instructed to choose the predominant variable. This proved to be no

significant issue inpracticeanddueto the fact thatsolelyonecoderwascoding, theuniformityof the

codingremained.However,thisisanaspectofthestudythatcouldbeimprovedupon.

Anotheritemthatcontainedasubjectiveaspectwaspathos.Themainnotiontoidentifypathos

was stated to be identification (Higgins & Walker, 2012). One person can feel that the corporations

identifieswithhisorherneedsandsituation,whileotherpeoplemightnotagree.Byfocusingonspecific

signalwords,itwasattemptedtoidentifyidentification.

Lastly, thedataof thestudywassomewhatunequallydivided. In thetimeframeofayearthat

wasanalyzed,Shellposted22postasopposedtoTotalandBPwhoposted58and40CSR-relatedposts

respectively.Duetothis,thetrustworthinessoftheShelldatasetisarguablylessthanthatoftheother

twocorporations.

-36-

7.Results

Thissectionwillelaborateontheresultsgainedfromthefrequencytablesandcrosstabulationscreated

aftercodingtheFacebookpostswithSPSS.InordertogainaninitialunderstandingoftheCSRpresence

thatthecompanieshaveestablishedforthemselves,theCSRtopicsoftheFacebookpostswereexplored.

Analysis of the Facebook pages of the three oil companies reveals that all three companies have a

differentapproachinthetypeofCSRinitiativestheyprefertocommunicate.

Ascanbeseeninfigure3,themajorityofthepostspublishedbyShellinvolvesenvironmentalCSRissues,

none of their posts revolve solely around social initiatives, and a little over one-third includes both

environmental and social practices. Contrarily, BP addresses exclusively social CSR initiatives in a vast

majority of their CSR-related Facebookposts. Total positions itself in themiddle anddedicates a small

majorityoftheirFacebookpostsonenvironmentalandclosetoone-thirdonsocialissues.

Alsonoteworthyiswhatspecificenvironmentalandsocialtopicsthecompaniesconcentrateon.

Concerning environmental issues, therewere solely two topics that the companies seemed to address

(figure 4). Shell posts predominantly about environmental issues, but does not necessarily focus on

reducing their own negative impact. They mainly address improving the sustainability of existing

manufacturingprocesses,andtheyemphasizeinnovation.ThesameappliestoTotal,whichalsodevotes

themajority of their posts to improving the sustainability of production,whilewaste reduction comes

second.BPontheotherhand,postssignificantlymoreonwastereduction.

Figure3CSRtopic

28%

53%

19%

0%

Total0%

59%

32%

9%

Shell

72%

23%

5%

0%

BPAmerica

-37-

Thesocial issuesaremorevaried,ascanbeseen infigure5.Shell refers intheirpostsonsocial topics,

mostlytohelpingspecificcommunities.Totalalsofocusesoncommunities,butspecificallytheonesthat

areexperiencing financialandmedicalneeds,or theiremployees.Asmentionedbefore,ofall theCSR-

postspublishedbyBP,threequarterisaimedatsocialissues.Thesesocialissuesrevolvearoundcreating

employment and providing education opportunities, which both aim at providing for the immediate

community.

This initial analysis shows that companies choose to address the negative connotations around oil

companies differently by focusing on contrasting CSR practices. Shell and Total prefer communicating

environmentalissues,andmainlyfocusonthesustainabilityofmanufacturingprocesses.BPontheother

Figure5socialtopics

14%

72%

0%0% 14%

Shell0%

7%

33%

19%

41%

Total

0%

15%

29%

32%

24%

BPAmerica

15%

85%

Shell

26%

74%

Total

73%

27%

BPAmerica

Figure4environmentaltopics

-38-

hand,prioritizessocialoverenvironmental issuesandhasavarietyofsocial issuestheyaddress intheir

posts.However,educationandemploymentcountforthemajorityoftheirsocialCSRposts

7.1Textualstyles

Corporations can choose from a range of linguistic styles to communicate their CSR practices and to

possiblyexercisepowerrelationsbetweenthemandtheirfollowers.Togainagenericunderstandingof

howcorporationsarecommunicatingtheCSRmessages,narrativestyleswereinvestigated(figure6).

Ingeneral,thevastmajorityoftheFacebookpostsarewritteninanaccountingstyle,meaningthatthe

posts mainly cover current events, developments, and updates on the status of the practices of the

company. Solely Total includes another notable narrative style, with one-fourth of its Facebook posts

revolvingaroundadvertisements.ShellandBPalsopostsomeadvertisements,butthisaccounts forno

morethanone-tenthoftheirpostsLastly,regardingannouncementsoffutureeventsoractivities,BPand

Total respectivelypublish2and1post, counting for ameagerpartof theirposts. Therefore, it canbe

concluded that accounting is the preferred narrative style, with Total significantly engaging in

advertisementsaswell.

Lookingbeyondnarrativestylesandfocusingmoreontheactualcommunication,theinteractivity

of the Facebookpostswas investigated. Regarding sensory interactivity, as presented in figure7, Shell

and BP predominantly choose videos over images, while Total opts for images in the majority of the

cases.Videosprovideconsiderablymoreroomforcommunicatinginformation,becauseofwhichitmight

be argued that using videos is preferredwhen the companywants to disclosemore information than

Figure6narrativestyles

0%

10%

20%

30%

40%

50%

60%

70%

80%

90%

100%

Accounting Commentary Advertisement Announcement

Shell Total BPAmerica

-39-

whatispresentedinthetextualpartofthepost.Nevertheless,thecompaniesfocusonthetextsaswell,

with all three companies using interactive text to appeal to the audience in the vastmajority of their

posts.Thisincludeselementssuchasexclamationmarks,hashtagsandlinks.Fromthiscanbededucted

thatthecorporationsoptforusingthesensesofthefollowerstocreateinteraction.

Anotherdimensionofinteractivityisthesemanticdimension(figure8).Throughaddressingtheuserwith

certainwords, theuser can feelmoreengagedwithandperceive thepostas interactive (Sohn,2013).

Thistypeofinteractivitywasencouragedsignificantlylessthansensoryinteractivity.Themajorityofthe

Facebookpostswaswritten in user-friendly language to ensure the followers understood the content,

andbothShellandBPpersonalizedthemessagessubstantiallybydirectlyaddressingthereader.Totaldid

use user-friendly language in nearly all the posts, but did not pertain to a majority of personalized

messages:58.6%oftheirFacebookpostsdidnotcontainanypersonalizedelements.Hence,Totalseems

totakeamoredistantapproachtotheirfollowersthantheothertwocompanies.Moreover,noneofthe

corporationsregularlyidentifiedthefollowersintheposts.Besidesoneortwotimes,noneoftheposts

clarified who they were posting the message for, due to which the posts lost interaction with the

followers.

22%

33%

45%

BPAmerica

42%

14%

44%

Total

9%

43%

48%

Shell

Figure7sensoryinteractivity

-40-

To deepen the understanding of exercising potential power relations throughCSR communication, the

notionofformalitymightberelevant,aspicturedinfigure9.BothShellandBPdidnotincludeanyformal

languageintheirFacebookposts,andbothmaintainan informalstyleof languagewithanactivevoice,

simplewordsandsentences,wordsthatexpressfamiliarity,andopinionsandfeelings.EventhoughTotal

didnotusea substantial amountof formal language in theirpostseither,however, theydidusemore

than the other two companies with 10.3% of the posts being written formally. Due to the unproven

reliabilityoftheanalysis,itisquestionablewhetherthisresultissignificant.However,assumingthatitis,

itcanbearguedthattheseresultsareinlinewiththedegreeofpersonalizedmessagesthatTotalused,

asmentioned before. Shell and BP used an extensive amount of personal elements in their Facebook

posts, whichmight be connected to their informal writing style. However, Total did not use asmuch

personalizedmessages,andmightthereforebeexpectedtohavea less informalwritingstyle.Although

notsubstantial,thiswasconfirmedbytheresultsoftheformalityanalysis.

0%

20%

40%

60%

80%

100%

Shell Total BPAmerica

Formal Informal

Figure9formality

70%1%

29%

Total

57%

1%

42%

BPAmerica

54%

2%

44%

Shell

Figure8semanticinteractivity

-41-

By focusing on abstractness, this research further investigated if the corporations go into detail when

communicating their CSR initiatives and practices (figure 10). All of the posts published by Shell are

writtenabstractly,without significantdetail about the issues that arediscussed. Furthermore, the vast

majorityoftheBPpostswereabstractandTotalalsopredominantlypublishedtheirpostsinanabstract

style.Hence,allthreecompaniesprefernottospeaksignificantlydetailedabouttheCSRpracticesthey

areundertaking.Abstractnessisnotnecessarilydirectlyrelatedtoanyofthepreviouslydiscussednotions

ofnarrativestyle,interactivityandformality.However,itdoesbuildfurtherontheapparenteffortofthe

companiestonotdirectlyaddressthefollowersandcommunicatewiththem.Bymaintainingasuperficial

andabstractstyle, thefollowersdonotgainadetailedandcomprehensiveunderstandingofthe issues

that are focused on. This way, the corporations seem to aim at communicating in a rather shallow

mannerwiththeirfollowers.

Ascanbeconcludedfromtheaforementionedresults,thereisaratherimbalanceintowhatextentthe

corporations determined to closely communicatewith their followers. Shell and BP opt for postswith

general information and limited invitation for participation,whilemaintaining an informalwriting style

and interactivity on both a sensory and semantic level. Total prefers general information, limited

encouragement of participation, and sensory and semantic interactivity aswell. However, Total is less

inclinedtopromotethesemanticinteractivityelementofpersonalmessages,andslightlylessinclinedto

include an informal writing style. To investigate if personal messages and informal writing style are

related,amultivariateanalysiswasconducted(figure11).

Figure10abstractness

-42-

This analysis shows that in the case of Total, personalized messages seem to partially relate to the

formalityoftheFacebookposts.Ifthepostswerewritteninaformalstyle,personalizedmessageswere

not present in any of the posts. However, when the writing style was informal, the presence of

personalized messages was close to equally divided. Therefore, it can be argued that the use of

personalizedmessages istoacertainextentrelatedtotheformality intheFacebookposts.Seeingthat

semantic interactivity includes personalized messages, this means that formal messages are less

interactive and encourage less engagement from the audience than informal messages. All three

companiespredominantlyuseinformalmessages,whichdemonstratesthatthecompaniesdonotreject

engagementwiththeaudienceinthemajorityoftheFacebookposts.

7.2Rhetoric

InordertoexplorerhetoricalaspectsoftheCSR-relatedFacebookposts,allrhetoricaldimensionswere

analyzed:logos,pathosandethos.Byfocusingonlogos,theaimistoidentifyhowthecorporationsmade

use of rational argumentation to appeal to the audience. In this case this was analyzed by adopting

justifications,dataandexamples,asfigure12presents.

Theanalysis shows that theextensivemajorityof thepostspublishedby the companiesdidnot adopt

specificlogospractices.Shellonlyuseddata,justificationsandexamples,inafifthoftheirposts.BPused

slightlymore,byfocusingonjustificationsandexamplesinaround35%oftheirposts,andondatainone-

Figure12logos

Figure11personalizedmessagesxformality

-43-

fourth.Totalontheotherhand,usedsignificantlylessjustificationsandexamples,butdidincludedatain

more than half of their messages. Based on this, it can be argued that for CSR communication, oil

corporationsdonotsignificantlyrelyonreasoningandargumentationtoappealtotheaudience.

Theuseofpathosishighlyunequallydividedwhencomparingthethreecorporations(figure13).

Pathos refers to appealing to the audiencewith emotional cues. One of these emotional cues can be

metaphors,duetothefactthatmetaphorsmakeincomprehensibletopicsmoreengagingandrelatable.

Shellengages inusingmetaphors innearly two-thirdof thepublishedFacebookposts.BPontheother

hand,optsformetaphorsinamere10%oftheirmessages.Totalpositionsitselfinthemiddlewithalittle

over one-fourth of their posts including metaphors to appeal to the audience. Another manner of

appealing to the audience’s emotions, is byusingnotions that the followers can relate to. BP includes

elementsto identifywiththeaudience inthree-quartertheirpublications,whileShellandTotalengage

significantly less in this,withTotalonlyaddingelementsof identification in less thanone-fourthof the

posts.Ashaspreviously stoodout in these results,Total seems to takemoredistantapproach to their

followersthantheothertwocorporations.Theyaimtoappealtotheaudiencebyincludingmetaphors,

butseemtomaintaindistancebyusingsignificantly lesselements tomaketheaudience feel identified

with.

Finally, certain ethos-related elements seem to be present in all of the Facebook posts (figure 14). A

majority of the posts published by Shell are related to improving the world, and caring about the

audience comes secondwith a little under one-fourth of the posts. This is the opposite for BP: a vast

majorityoftheirpostsrevolvesaroundcaringfor theaudience,while15%focusesoncleaninguptheir

own act and a mere 12.5% on improving the world. Total is positioned in between the other two

corporations with themost substantial amount of posts devoted to improving the world, after which

followtheirclaimsaboutcaringfortheaudienceandcleaninguptheirownact.

Figure13pathos

-44-

ShellandTotalseemtotakeasimilarapproachinappealingtotheaudiencebyfocusingpredominantly

ontheirinitiativestoimprovetheworldandmakeitabetterplacetoliveforpeopleallovertheworld.

ThisisinlinewiththeCSRtopicsthecompaniesprefertoaddress,whichhavebeenanalyzedpreviously.

Both Shell and Total address mostly environmental topics, in which they focus on sustainability of

manufacturingprocessesandsignificantlylessonwastereduction.Thisdemonstratesthattheyseemto

communicate lessaboutcleaningup theirownact,and focusmainlyon theirobjective to improve the

world.BPontheotherhand,hasamorecommunity-orientedapproachandtheirpostsmainly revolve

around improving specific aspects of people’s life. This shows in the CSR topics they address, with

educationandemploymentbeingthemainissuesBPfocuseson.

7.3Summary

To summarize the findings, both Shell and Total mainly focus on environmental issues in their CSR

communication,aimingatthesustainabilityofmanufacturingprocesses.Ontheotherhand,BPprioritizes

socialissuesandfocusespredominantlyonissuesdirectlyrelatedtothedevelopmentofthecommunity,

suchaseducationandemployment.

Considering textual styles, all three companies prefer an accounting style for their Facebook

posts. Solely Total uses another style in a considerable amount of posts: advertisements in close to a

quarter. Both of these narrative style encourage little engagement of the audience. Focusing on

Figure14ethos

0%10%20%30%40%50%60%70%80%90%

100%

claimstoimprovethe

world

claimstocleanuptheirownact

claimstocareabouttheaudience

focusesonjoiningCSRmovement

noneoftheabove

Shell Total BPAmerica

-45-

interactivity,bothsensoryinteractivityandsemanticinteractivitytakeplaceonallthreeFacebookpages.

However,consideringsemanticinteractivity,noneofthecompaniesoptedinmorethan2%oftheposts

for identifying the specific target audience. Furthermore, Total seems to be less inclined to use

personalizedmessages,withusingthissemanticelementin29%oftheposts,asopposedto42%and44%

oftheothercompanies.RegardingtheformalityoftheFacebookposts,allcompaniesprefertomaintain

informal language.Totalopts forslightlymoreformal languagethantheothertwocompanies,butthis

difference is limited to such an extent that it might not be reliable. After an analysis of a possible

correlation between formality and personalized messages, it can be argued that there might be a

correlation between the lack of personalized messages and the slightly larger percentage of formal

languageinTotal’sFacebookposts.Lastly,inabstractness,allcompaniesoptforabstractnessasopposed

toconcretenessinavastmajorityofposts.

Fromtheanalysisoftherhetoricalelementscanbedeductedthatthereisasignificantimbalance

intheuseoflogosandpathosamongthecompanies.WhereBPpreferstoincludethelogospracticesof

justificationsandexamples,ShellusesallnotionsequallyandTotalhasapreferenceforusingdataforthe

argumentation intheCSRmessages. Itseemsthatnoneofthecompanieschoosestofocuson logosto

appealtotheaudience,duetothefactthat,besidesTotalusingdata inaslightmajorityoftheirposts,

neither of the other two companies used logos in the majority of their content. Focusing on pathos

demonstratesthatTotalequallyusedmetaphorsandelementstoidentifywiththeaudienceinaminority

oftheirposts,whileShellsignificantlyusesmetaphorsandBPidentificationwiththeaudience.Inregards

toethos,all companies seemed to include somenotionofethos. Shell andTotalprefer topresent the

majorityoftheirpracticesasmeansofimprovingtheworld,whileTotalpredominantlyadoptstheclaim

ofcaringabouttheaudience.

8.Discussion

InanefforttoexplorehowoilcorporationsusetextualandrhetoricalstructurestocommunicateCSRand

exercisepower relations throughFacebook, Iperformeda contentanalysisona totalof120Facebook

posts spread over three oil companies: Shell, Total and BP. The research questions proposed for this

research were: RQ1 How are textual structures used in CSR communication through SNS? RQ2What

rhetoricispresentinCSRcommunicationthroughSNS?Toanswerthefirstresearchquestion,thetextual

stylesofnarratives,interactivity,formalityandabstractnesswereanalyzed.Inordertoanswerthesecond

-46-

research question, the rhetorical dimensions of logos, pathos and ethos were explored. After data

analysis, these researchquestionswereanswered toacertainextent,andthiswillbeelaboratedon in

thisconcludingsectionofthethesis.

8.1Textualstyles

Narrative styles refer to the narratives that are present in Facebook posts, which can influence how

companiesencourageinteractivity.Asbecameevidentfromtheresults,allthreecompaniesincorporate

narrative styles that do not seem to involve direct engagement: accounting and, to a certain extent,

advertisement.Inaaccountingpost,thecompanydoesnotdirectlyinvitetheaudiencetoparticipatein

thecommunication.Thesameappliestoapostwrittenasanadvertisement,whichisusedtopresenta

serviceorproduct,andtopersuadetheaudienceintogainingapositiveattitudetowardsit.Cvijikjetal.

(2011) support the claim of accounting and advertisement posts not generating engagement and

participation,bystatingthattheaccountingstyledoesnotencouragecommenting.Inordertogeneratea

substantialnumberofcomments,postswritteninaninformation-seekingorcompetitionstyleshouldbe

published,dueto the fact that thesestylesaredirectly related toengagementof the followers.Seeing

thatallthreecorporationsprefernarrativestylesthatdonotinvolvedirectengagement,itseemsthatthe

investigatedcompaniesdonotspecificallyaimtoinstigateadialoguewiththefollowers,butrathergive

one-sidedaccountsoftheirpractices.However,Cvijikjetal. (2011)alsofoundthattheaccountingstyle

generatesmorelikesthanotherstyles,meaningthatfollowersingeneralseemtoappreciatethistypeof

posts. This could be regarded as a type of engagement. Nonetheless, it does not necessarily invite

dialogue and this could interfere in enacting power. As Hammond, Anderson and Cissna (2003) state,

dialogueandpowerareintertwined.Byenteringintodialogue,thepowerrelationsbetweentwoparties

areenacted.Thecorporationsdonotseemtoaimfordialogue,andthereforemightabstainthemselves

toacertainextentfromexercisingpower.

Tocontinueonthetopicofengagement,bothsensoryandsemanticinteractivitywereexplored.

As previously mentioned, all elements of sensory interactivity (video, image, interactive text) were

presentinallthreeFacebookpages,interactivetexttoalargerextentthantheothertwoelements.This

mightindicatethatthecompanies,regardlessofthenarrativestyle,aimtoengageandinteractwiththeir

followers.Regardingsemanticinteractivity,itwasfoundthatnoneofthecompaniesoptinmorethan2%

oftheirpostsforidentifyingthespecificaudiencetheyareaddressingthepostto.Thiscouldberelatedto

-47-

the absenceofdialogue that the corporations aremaintaining in their CSR communication,whichwas

previouslymentionedintheanalysisofthenarrativestyles.Itcanbearguedthatthecorporationsintend

toengagetheaudience,butdonotnecessarilyaimforittofeeltargetedenoughtoenterthedialogue.

Byrefrainingfromdialogue,thecompaniesaredeprivingthemselvestoacertainextentoftheexecution

of power relations (Hammond et al, 2003), but they also ensure that the audience does not feel

encouragedtoparticipateinthecommunicationandvoicetheirpossiblynegativeopinions.

AnalyzingtheregistersofformalityandabstractnessintheFacebookpostsrevealedthatthevast

majority of the posts of all three companieswerewritten in an informal and abstract style. From the

aforementioned analysis of narrative styles and interactivity, and the related seemingly apprehensive

attitude of the corporations towards dialogue, it would be expected that the corporations use a

substantialamountofformallanguageintheirposts.Similarly,ifcorporationswouldaimexercisepower

through CSR communication, they would likely opt for formality, as derived from Fairclough (1989).

Fairclough (1989) strongly links formality to the notion of power relations. Formal language use

encompassesconstraintsonallthreelevels:content,relationsandsubjects(Fairclough,1989).Itrestricts

contentthatcanbeincludedinthediscourse,itlimitspotentialcasualrelationshipsanditdeterminesthe

position the other participants are allowed to take.Using formal language is directly related to power

distance,meaningthatacommunicatorusingthistypeoflanguageaimstocreatedistanceandincrease

its dominant position as opposed to the other participant (Fairclough, 1989). Nevertheless, the

corporationspreferaninformalwritingstyle.Thismightbeinlinewiththepreviouslystatedassumption

that thecorporationsaimtoengagewiththeiraudience,butdonot intendtostartanactualdialogue.

Using an informalwriting style does not necessarily invite dialogue, but one could state that due to a

morelimitedpowerdistance,thiswritingstylecontributestoappealingtotheaudience.

The high degree of abstractness in the posts is not in accordancewithWagner et al.’s (2009)

researchonabstractnessinCSRcommunication.TheystatethatabstractmessagesaboutCSRinitiatives

tendto include issuesthatmighthaveanegative impactonthereputationofthecompany.Therefore,

whenitconcernsmessagesthatlikelyenforceapositiveimage,themessagesarewrittenconcretely,with

more details and information. However, the vastmajority of all the Facebook posts are abstract,with

littletext,whileaimingforgeneratingapositivereputation.Throughtheuseofelementsthatencourage

sensoryinteractivity,suchasvideosandimages,thecorporationsaddinformationtothetextualpartof

the post. Yet, the degree of information in the videos was taken into account when coding the

abstractnessof theposts,meaning thateven including theseelements, themajorityof thepostscould

-48-

notbedescribedasconcrete.Thisabstractnessinthepostsmightbeattributedtoenactmentofpower,

asopposedtotheinformalstyleoftheposts.Inrelationtopotentialpowerenactment,itcanbeargued

that abstract languageuse is related to the constraint on content and subject, as statedby Fairclough

(1989). Restraining the available knowledge for the followers might imply that these followers are

constrainedinthecontenttheygetaccessto,andinthepositiontheyareabletotakeinthediscourse.

Withlimitedinformation,thefollowerswouldknowtoolittleaboutthetopictotakepartinthediscourse

andtakeaninformedstance.Ontheotherhand,socialmediaisrelatedtoanewtypeofcommunication,

so-calledmass self-communication (Castells, 2009). Thismeans that all the contentpublishedonline is

restricted,directed,andcreatedbytheself.Furthermore,usersdeterminethemselveswhatcontentthey

consume. Individuals that follow Facebook pages of oil and gas companies, will likely have an initial

interestintheindustry.Theydeliberatelyexposethemselvestothecontentproducedbythecompanies

and are, through this deliberation, expected to be motivated and engaged to participate in the

communication.

8.2Rhetoric

Elaboratingontheuseoflanguage,thisresearchfocusedonrhetoricalstructuresintheFacebookposts.

Therhetoricaldimensionsof logos,pathosandethoswereanalyzedandresultsshowedthattheuseof

rhetoricalelementswassignificantlydiverseamongthecompanies.BesidesTotal,whichusedthelogos

elementofdata inaslightmajorityoftheirposts,thecompaniesdidnotessentiallyuseargumentation

and logical reasoning to appeal to the audience. As Higgins and Walker (2012) state, corporations

communicating their CSR strategies might use logos with the aim of appearing rational and well-

constructed in their practices. However, Higgins and Walker (2012) studied CSR reports, which are

essentially reportswith facts and figures about theCSRpracticesof the companies. Thenatureof this

typeofchannelforcommunicationissignificantlydifferentfromsocialmedia.Facebookisinherentlynot

amediathatincludesstatisticsandfactsinitscommunication,itisdesignedforconnectingpeople(Van

Dijk,2010).ThismightexplainthelackoflogosinthecommunicationofCSRonFacebook.

However, the notion of pathos,which entails appealing to the emotions of the audience,was

used significantlymore. The results show that all three companiesusepathos in adifferentway. Shell

includesasignificantamountofmetaphors,whileBPprefersreferringtoissuestheaudiencecanidentify

with,andTotalusesbothelementsequally.Aspreviouslymentionedinthetheorysectionofthisthesis,

-49-

HigginsandWalker(2012)relatepathostoCSRinthewaythatcompaniesaimtomaketheaudiencefeel

likeitunderstandstheirneedsanddesires.AsDuandVieira(2012)state,oilandgascompanieshavehad

anotoriousinfluenceonenvironmentalandsocialconcerns.Duetothis,itislikelythattheaudiencehas

encountered the negative impacts of their practices themselves and have taken a rather negative and

emotional stancetowards thecompanies.Accepting thisassumptiontobe true, thepathoselementof

usingtopicsthattheaudiencedirectlyrelatesto,isdirectlyrelatedtothis.BPisthecorporationwiththe

mostwell-knownenvironmentalscandalandit isalsothecompanythatusestopicsto identifywiththe

audiencethemost,whichcouldberelatednotions.Theuseofmetaphors isa relatively ‘softer’pathos

method(Higgins&Walker,2012),whichismainlyadoptedbyShell.

Through the adoption of ethos, all three companies aimed to increase the trustworthiness of

theircharacter towards theaudience.As Ihlen (2011)states,ethos tendsto trump logosandpathos in

thecommunicationofCSR.Thereasoningandargumentationofaninitiativemightbeinaccurate,orthe

initiativemight not appeal to the emotions of the audience, the corporation can still appear credible

throughtheuseofethos.Ihlen(2011)identifiescertainethosstrategiesthatcorporationspredominantly

useincommunicatingCSR.Ofthesestrategies,ShellandTotalfocusedoncommunicatingtheirclaimof

improving the world, while Total predominantly emphasizes how they care for the audience. This is

directlyinlinewiththeinitialinquiryofthecorporations’CSRreports.ThisinquiryshowedthatShelland

Totalfocusonenvironmentalissues,whileBPseemstofocusmoreonthesocialinfluenceithasonthe

direct community. It can be argued that the use of ethos is related to power. Through certain use of

language to increasecredibilityof thecorporations’ character, theoil companiesaim to shift attention

fromtheirnegativeimpactstotheirpositiveimpacts(Ihlen,2011).

8.3CSRtopics

Anotherelementof theFacebookposts thatwasanalyzed,and likelyprovestobe influential regarding

CSR communication and related power relations, is the CSR topic that is communicated. These topics

revolvearoundenvironmentaland/orsocialissues.Thisiscloselyrelatedtotheethosstrategiesthatthe

companiesadopt.ShellandTotalpredominantlypostonenvironmental issues,butfocus intheseposts

on increasingthesustainabilityof theirmanufacturingprocesses,andnotsomuchonwastereduction.

Thismight suggest that the companies are refraining fromelaborating on themost apparent negative

impactofoilcorporations;wasteintheformofoilspillingandCO2emissions.Alongtheselines,itcanbe

-50-

argued that BP aims to significantly exclude environmental issues, by predominantly focusing on the

positivesocial impactof thecorporation.Thismayormaynotbe related to theBPoil spillof2010, in

whichatonepoint42,000gallonsofoilspilledonadailybasis(Robertson&Kaufman,2010).Eitherway,

thisphenomenoncanberelatedtopowerrelationsandconstraints,asdescribedbyFairclough(1989).As

aforementioned,exercisingpowerrelationsthroughdiscourseincludesvariouslevelsofconstraints.One

oftheseconstraintsinvolvescontent.InthecaseofShellandTotal,itcanbearguedthatthecompanies

predominantly limit their environmental content to the initiatives that cause them to have a positive

influence on the environment, instead of focusing on decreasing their negative impact. BP limits their

content predominantly to social issues, without significantly including environmental topics. These

constraints incontentdemonstrateadegreeofpowerthecompaniesareexercising inrelationtotheir

followers.

8.4Limitationsandfutureresearch

Therearecertainlimitationsthatshouldbetakenintoaccountregardingtheconductedresearch.Firstly,

the sampleof the study is rather small. This is especially the case for Shell, ofwhichnomore than22

postswereanalyzed.Aresearchthatanalyzesamoresubstantialamountofdatawilllikelyproducemore

significantresultsandwouldbeabletoeithersupportorrejectthefindingsfromthisresearch.Secondly,

thefindingsaresolelybasedontheFacebookpostspublishedbythecorporations.Asmentionedinthe

introductionof this thesis,Facebook isaplatformwhichoffersdialogueandtwo-sidedcommunication,

enablingboththeproducerandreceiverofcommunicationtoexercisepower.Itwouldbeinterestingto

analyzethecommentsontheFacebookpostsandexploretowhatextentfollowerscommentoncontent

and how the companies interact with the followers. This would give a more all-encompassing

understandingofthepowerrelationsbeingexercisedinCSRcommunicationthroughFacebook.Thirdly,

the discussion of the results showed possible correlations between the BP oil spill and the CSR

communicationstrategyBPhasseemedtoadopt.Futureresearchcouldconductanin-depthanalysisof

thebackgroundof thecompanies togain interestingdataonhow importanteventsandCSRstrategies

arerelated.

-51-

8.5Conclusion

Inshort,thisstudywassetouttoexplorelinguisticstructuresinCSRcommunicationthroughFacebook

andhasidentifiedthetextualelementsofnarratives, interactivity,formalityandabstractness,andthe

rhetorical elements of logos, pathos and ethos, in the posts of 3 oil corporations. Taking a

communication approach, the research sought to investigate how potential power relations are

exercisedthroughCSRcommunicationonSNS.Resultsshowedthatcertain linguisticstructures inthe

posts refer to power enactment, while other structures do not support this claim. This power was

mainlyexercisedthroughconstraintsoncontentand,toacertainextent,constraintsonthepositionthe

followerscantake.Nevertheless,asignificantdegreeofinformalityandfamiliaritywasmaintainedwith

the followers, aiming to decrease the power distance between the corporations and their audience.

With this study, I hope to lay groundwork for potential further research on the enactment of power

throughCSRcommunicationonSNS.

-52-

References

Ali,I.,Jiménez-Zarco,A.I.,&Bicho,M.(2015).UsingsocialmediaforCSRcommunicationandengaging

stakeholders.DevelopmentsinCorporateGovernanceandResponsibility,7,165-185.

doi:10.1108/s2043-052320150000007010

Bazarova,N.N.,Taft,J.G.,Choi,Y.H.,&Cosley,D.(2012).Managingimpressionsandrelationshipson

Facebook:self-presentationalandrelationalconcernsrevealedthroughtheanalysisoflanguage

style.JournalofLanguageandSocialPsychology,32(2),121-141.

doi:10.1177/0261927x12456384

Becker-Olsen,K.L.,Cudmore,B.A.,&Hill,R.P.(2006).Theimpactofperceivedcorporatesocial

responsibilityonconsumerbehavior.JournalofBusinessResearch,59(1),46-53.

doi:10.1016/j.jbusres.2005.01.001

Bhattacharya,C.,&Sen,S.(2004).Doingbetteratdoinggood:When,why,andhowconsumersrespond

tocorporatesocialinitiatives.CaliforniaManagementReview,47(1),9-24.

doi:10.2307/41166284

Biber,D.(1995).Dimensionsofregistervariation:Across-linguisticcomparison.Cambridge,United

Kingdom:CambridgeUniversityPress.

Biber,D.,&Finegan,E.(1994).Sociolinguisticperspectivesonregister.NewYorkCity,NY:Oxford

UniversityPress.

Bitzer,L.(1968).Therhetoricalsituation.PhilosophyandRhetoric,1(1),1-14.

Bitzer,L.(1992).Therhetoricalsituation.PhilosophyandRhetoric,25,1-14.Retrievedfrom

http://www.jstor.org/stable/40237697

BP.(2016).Sustainabilityreport2015.Retrievedfrom

https://www.bp.com/content/dam/bp/pdf/sustainability/group-reports/bp-sustainability-report-

2015.pdf

Busch,T.,&Shepherd,T.(2014).Doingwellbydoinggood?NormativetensionsunderlyingTwitter's

CorporateSocialResponsibilityethos.Convergence:TheInternationalJournalofResearchinto

NewMediaTechnologies,20(3),293-315.doi:10.1177/1354856514531533

Capriotti,P.(2011).CommunicatingCorporateSocialResponsibilitythroughtheInternetandsocial

media. InØ.Ihlen,J.Bartlett,&S.May(Eds.),Thehandbookofcommunicationandcorporatesocial

responsibility(pp.358-378).Chichester,UnitedKingdom:Wiley-Blackwell.

-53-

Capriotti,P.,&Moreno,Á.(2007).Corporatecitizenshipandpublicrelations:Theimportanceand

interactivityofsocialresponsibilityissuesoncorporatewebsites.PublicRelationsReview,33(1),

84-91.doi:10.1016/j.pubrev.2006.11.012

Carr,C.T.,&Hayes,R.A.(2015).Socialmedia:defining,developing,anddivining.AtlanticJournalof

Communication,23(1),46-65.doi:10.1080/15456870.2015.972282

Castelló,I.,Morsing,M.,&Schultz,F.(2013).CommunicativedynamicsandthepolyphonyofCorporate

SocialResponsibilityinthenetworksociety.JournalofBusinessEthics,118(4),683-694.

doi:10.1007/s10551-013-1954-1

Castells,M.(2001).TheInternetgalaxy,reflectionsontheInternet,BusinessandSociety.Oxford,United

Kingdom:OxfordUniversityPress.

Castells,M.(2009).Communicationpower.Oxford,UnitedKingdom:OxfordUniversityPress.

Colleoni,E.(2013).CSRcommunicationstrategiesfororganizationallegitimacyinsocialmedia.Corporate

Communications:AnInternationalJournal,18(2),228-248.doi:10.1108/13563281311319508

Coombs,W.T.,&Holladay,S.J.(2012).ManagingCorporateSocialResponsibility:Acommunication

approach.Malden,MA:Wiley-Blackwell.

Corbett,E.P.,&Connors,R.J.(1999).Classicalrhetoricforthemodernstudent(4thed.).NewYork,NY:

OxfordUniversityPress.

Curley,C.B.,&AbgrabNoormohamed,N.(2013).SocialmediamarketingeffectsOnCorporateSocial

Responsibility.JournalofBusiness&EconomicsResearch,12(1),61.

doi:10.19030/jber.v12i1.8379

Cvijikj,I.P.,Spiegler,E.D.,&Michahelles,F.(2011).Theeffectofposttype,categoryandpostingdayon

userinteractionlevelonFacebook.IEEEThirdInternationalConferenceonPrivacy,Security,Risk

andTrust,810-813.doi:10.1109/passat/socialcom.2011.21

Denef,S.,Bayerl,P.S.,&Kaptein,N.A.(2013).Socialmediaandthepolice:tweetingpracticesofbritish

policeforcesduringtheAugust2011riots.ProceedingsoftheSIGCHIConferenceonhuman

factorsincomputingsystems-CHI'13,3471-3480.doi:10.1145/2470654.2466477

Du,S.,Bhattacharya,C.,&Sen,S.(2010).MaximizingBusinessReturnstoCorporateSocialResponsibility

(CSR):TheRoleofCSRCommunication.InternationalJournalofManagementReviews,12(1),8-

19.doi:10.1111/j.1468-2370.2009.00276.x

Du,S.,&Vieira,E.T.(2012).StrivingforlegitimacythroughCorporateSocialResponsibility:insightsfrom

oilcompanies.JBusEthics,110(4),413-427.doi:10.1007/s10551-012-1490-4

-54-

Dutot,V.,LacalleGalvez,E.,&Versailles,D.W.(2016).CSRcommunicationsstrategiesthroughsocial

mediaandinfluenceone-reputation.ManagementDecision,54(2),363-389.doi:10.1108/md-01-

2015-0015

EuropeanUnion.(2011,October25).CorporateSocialResponsibility:anewdefinition,anewagendafor

action.RetrievedMarch25,2016,fromhttp://europa.eu/rapid/press-release_MEMO-11-

730_en.htm

Fairclough,N.(1989).Languageandpower.London,UnitedKingdom:Longman.

Fenton,N.(2012).MisunderstandingtheInternet.J.Curran&D.Freedman(Eds.).Abingdon,United

Kingdom:Routledge.

Fieseler,C.,&Fleck,M.(2013).Thepursuitofempowermentthroughsocialmedia:structuralsocial

capitaldynamicsinCSR-blogging.JournalofBusinessEthics,118(4),759-775.

doi:10.1007/s10551-013-1959-9

Fieseler,C.,Fleck,M.,&Meckel,M.(2009).CorporateSocialResponsibilityintheblogosphere.Journalof

BusinessEthics,91(4),599-614.doi:10.1007/s10551-009-0135-8

Forbes.(2016,March30).World'sBiggestOilCompanies-2016.RetrievedMay22,2016,from

http://www.forbes.com/pictures/hefj45fim/no-1-gazprom/

Friestad,M.,&Wright,P.(1994).Thepersuasionknowledgemodel:Howpeoplecopewithpersuasion

attempts.JournalofConsumerResearch,21(1),1.doi:10.1086/209380

Gass,R.H.,&Seiter,J.S.(2016).Persuasion,socialinfluence,andcompliancegaining(5thed.).NewYork

City,NY:Routledge.

Grushina,S.(2008).CorporateSocialResponsibilitythroughthelensofcommunicationtheoryand

research.PaperpresentedatthemeetingofNationalCommunicationAssociation,SanDiego.

Halliday,M.A.(1978).Languageassocialsemiotic:Thesocialinterpretationoflanguageandmeaning.

Baltimore,MA:UniversityParkPress.

Hammond,S.C.,Anderson,R.,&Cissna,K.N.(2003).Theproblematicsofdialogueandpower.InP.J.

Kalbfleisch(Ed.),Communicationyearbook27(pp.125-157).Mahwah,NJ:LawrenceErlbaum

Associates.

Higgins,C.,&Walker,R.(2012).Ethos,logos,pathos:Strategiesofpersuasioninsocial/environmental

reports.AccountingForum,36(3),194-208.doi:10.1016/j.accfor.2012.02.003

Humphreys,L.,Gill,P.,Krishnamurthy,B.,&Newbury,E.(2013).Historicizingnewmedia:acontent

analysisofTwitter.JournalofCommunication,63(3),413-431.doi:10.1111/jcom.12030

-55-

Ihlen,Ø.(2011).RhetoricandCorporateSocialResponsibility.InØ.Ihlen,J.Bartlett,&S.May(Eds.),The

handbookofcommunicationandCorporateSocialResponsibility(pp.147-166).Chichester,

UnitedKingdom:Wiley-Blackwell.

Jarrett,K.(2008).Interactivityisevil!AcriticalinvestigationofWeb2.0.FirstMonday,13(3).

doi:10.5210/fm.v13i3.2140

Juris,J.S.(2008).Networkingfutures:Themovementsagainstcorporateglobalization.Durham,NC:Duke

UniversityPress.

Kaplan,A.M.,&Haenlein,M.(2010).Usersoftheworld,unite!Thechallengesandopportunitiesofsocial

media.BusinessHorizons,53,59–68.doi:10.1016/j.bushor.2009.09.003

Kent,M.L.(2010).Directionsinsocialmediaforprofessionalsandscholars.InR.L.Heath(Ed.),Handbook

ofpublicrelations(2nded.,pp.643–656).ThousandOaks,CA:Sage.

Kesavan,R.,Bernacchi,M.D.,&Mascarenhas,O.A.(2013).WordofMouse:CSRcommunicationandthe

socialmedia.InternationalManagementReview,9(1),59-67.

Korschun,D.,&Du,S.(2013).HowvirtualCorporateSocialResponsibilitydialogsgeneratevalue:A

frameworkandpropositions.JournalofBusinessResearch,66(9),1494-1504.

doi:10.1016/j.jbusres.2012.09.011

Krippendorff,K.(2013).Contentanalysis:Anintroductiontoitsmethodology(3rded.).ThousandOaks,

CA:SagePublications.

Leckie-Tarry,H.(1995).Languageandcontext.D.Birch(Ed.).London,UnitedKingdom:Pinter.

Lee,K.,Oh,W.,&Kim,N.(2013).Socialmediaforsociallyresponsiblefirms:analysisofFortune500’s

TwitterprofilesandtheirCSR/CSIRratings.JournalofBusinessEthics,118(4),791-806.

doi:10.1007/s10551-013-1961-2

Moretti,A.,&Tuan,A.(2013,October).SocialmediaasalaboratoryofvaluecreationintheCSRfield.

PaperpresentedatSmartLife-Dall'innovazioneTecnologicaalMercato,Milan.

Nahai,N.(2012).Websofinfluence:Thepsychologyofonlinepersuasion.Harlow,UnitedKingdom:

Pearson.

Neuendorf,K.A.(2002).Thecontentanalysisguidebook.ThousandOaks,CA:SagePublications.

Nwagbara,U.,&Reid,P.(2013).CorporateSocialResponsibilityCommunicationintheageofnew

media:Towardsthelogicofsustainabilitycommunication.ReviewofInternationalComparative

Management,14(3),400-414.

-56-

O’Reilly,T.(2005).WhatisWeb2.0:Designpatternsandbusinessmodelsforthenextgenerationof

software.O’ReillyMedia.Retrievedfrom

http://oreillynet.com/pub/a/oreilly/tim/news/2005/09/30/what-is-web-20.html

Pavitt,H.(2012).NoPlacetoHide:NewtechnologicaladvancesinWeb2.0andSocialMediamayforce

organisationstoimprovetheirCorporateSocialResponsibility.SocialAlternatives,31(2),22-26.

Parker,D.H.(1972).Rhetoric,ethicsandmanipulation.Philosophy&Rhetoric,5(2),69-87.

Perloff,R.M.(2010).Thedynamicsofpersuasion:communicationandattitudesinthe21st

century(4thed.).NewYorkCity,NY:Routledge.

Reilly,A.H.,&Hynan,K.A.(2014).Corporatecommunication,sustainability,andsocialmedia:It'snot

easy(really)beinggreen.BusinessHorizons,57(6),747-758.doi:10.1016/j.bushor.2014.07.008

Robertson,C.,&Kaufman,L.(2010,April25).Oilleakscouldtakemonthstostop.NewYorkTimes.

Retrievedfrom

http://www.nytimes.com/2010/04/26/us/26rig.html?rref=collection%2Ftimestopic%2FGulf%20of

%20Mexico%20Oil%20Spill%20(2010)&action=click&contentCollection=timestopics®ion=stream&

module=stream_unit&version=latest&contentPlacement=1314&pgtype=collection&_r=0

Sandoval,M.(2014).Fromcorporatetosocialmedia:Criticalperspectivesoncorporatesocial

responsibilityinmediaandcommunicationindustries.Abingdon,UnitedKingdom:Routledge.

Schau,H.J.,Muñiz,A.M.,&Arnould,E.J.(2009).Howbrandcommunitypracticescreatevalue.Journal

ofMarketing,73(5),30-51.doi:10.1509/jmkg.73.5.30

Sheikha,F.A.,&Inkpen,D.(2012).Learningtoclassifydocumentsaccordingtoformalandinformal

style.LinguisticIssuesinLanguageTechnology,8(1),1-29.Retrievedfrom

http://journals.linguisticsociety.org/elanguage/lilt/article/download/2844/2844-5756-1-PB.pdf

Shell.(2016).Sustainabilityreport2015.Retrievedfromhttp://reports.shell.com/sustainability-

report/2015/

Simons,H.W.,&Jones,J.G.(2011).Persuasioninsociety(2nded.).NewYorkCity,NY:Routledge.

Simpson,P.,&Mayr,A.(2010).Languageandpower:Aresourcebookforstudents.London,United

Kingdom:Routledge.

Singh,J.(2016).StudyofrelationbetweenCSRandemployeeengagementinhotelindustry.InL.K.

Guliani&S.A.Rizwan(Eds.),CorporateSocialResponsibilityinthehospitalityandtourismindustry

(pp.63-68).IGIGlobal.

-57-

Sohn,D.,&Choi,S.M.(2013).Measuringexpectedinteractivity:Scaledevelopmentandvalidation.New

Media&Society,16(5),856-870.doi:10.1177/1461444813495808

Tench,R.,&Jones,B.(2015).Socialmedia:theWildWestofCSRcommunications.SocialResponsibility

Journal,11(2),290-305.doi:10.1108/srj-12-2012-0157

Tixier,M.(2003).Softvs.hardapproachincommunicatingonCorporateSocialResponsibility.

Thunderbird InternationalBusinessReview,45(1),71-91.doi:10.1002/tie.10056

Tomaselli,G.,&Melia,M.(2014).TheroleofinteractivetechnologiesforCSRcommunication.Journalof

InternationalScientificPublications,8(1),324-340.Retrievedfrom

http://www.gruppohera.it/binary/hr_responsabilita_sociale/ricerche/the_role_of_interactive_tec

h_for_CSR_communication_JOURNAL_PUBLICATION.1410508275.pdf

Total.(2015).Sustainablegrowthreport2014.Retrievedfrom

http://www.total.com/sites/default/files/atoms/files/rapport_croissance_durable_va.pdf

VanDijk,J.(2012).TheNetworkSociety(3rded.).ThousandOaks,CA:SAGEPublicationsInc.

Wagner,T.,Lutz,R.,&Weitz,B.(2009).Corporatehypocrisy:overcomingthethreatofInconsistent

CorporateSocialResponsibilityperceptions.JournalofMarketing,73(6),77-91.

doi:10.1037/e621072012-143

Ward,A.M.,&Wylie,J.(2014).Wordofmouse:usingsocialmediatoinformyourCorporateSocial

Responsibilitystrategy.AccountancyIreland,46(6),58-60.

WeAreSocial.(2015).Digital,social&mobileworldwidein2015.Retrievedfrom

http://wearesocial.com/uk/special-reports/digital-social-mobile-worldwide-2015

Wellman,B.(2000).Changingconnectivity:AfuturehistoryofY2.03K.SociologicalResearchOnline,4(4).

doi:10.5153/sro.400

Wilburn,K.,&Wilburn,R.(2015).Socialpurposeinasocialmediaworld.JournalofGlobalResponsibility,

6(2),163-177.doi:10.1108/jgr-07-2015-0011

Woolfson,C.,&Beck,M.(2005).CorporateSocialResponsibilityfailuresintheoilindustry.Amityville,NY:

BaywoodPublishingCompany,inc.

WorldBusinessCouncilforSustainableDevelopment.(1999).CSR:Meetingchangingexpectations.

Retrievedfrom

http://www.wbcsd.org/pages/edocument/edocumentdetails.aspx?id=82&nosearchcontextkey=tr

ue

-58-

Wrench,J.S.,Thomas-Maddox,C.,PeckRichmond,V.,&McCroskey,J.C.(2013).Quantitativeresearch

methodsforcommunication:Ahands-onapproach(2nded.).NewYork,NY:OxfordUniversity

Press.

-59-

AppendixA:Codebook1.Item

Code Criteria

1.Numberofpost Firstis‘1’,secondis‘2’,thirdis‘3’,etc.

2.Facebookpage

Code

1.Shell

2.Total

3.BP

3.CSRtopic

Code Criteria

1.Social Thepostfocusesmainlyonhelpingcommunities,individualpeopleandemployees

2.Environmental Thepostfocusesmainlyonimprovingtheenvironment,wildlifeandfightingclimatechange

3.Bothtopics ThepostcontainsinformationonbothsocialandenvironmentalissuestheyfocusonwiththeirCSRstrategy

4.Cannotbecoded

4.CSRsocialtopic(onlycodeifcoded‘1’foritem3)

Code Criteria

1.Humanrights Thepostcontainsinformationoninitiativesthatfightforhumanrights,suchashunger,peace,watersupply,etc.

2.Children Thepostcontainsinformationoninitiativesthathelpchildrenspecifically,butdoesnotfocusoneducation

3.Specificcommunity Thepostcontainsinformationoninitiativesthatfocusesonhelpingacommunityandtheemphasisisspecificallyonthatcommunity

4.Employmentandlaborrights

Thepostcontainsinformationoninitiativesthatfocusonbetterworkenvironmentsorwithcreatingand/ormaintainingjobs

5.Education Thepostcontainsinformationoninitiativesthatspecificallyfocusoneducationaldevelopments.

6.Thosewithfinancialormedicalneeds

Thepostcontainsinformationoninitiativesthatfocusonhelpingpeoplethathavefinancialandmedicalneedsandtheinformationemphasizestheseneeds.

7.Cannotbecoded

5.CSRenvironmentaltopic(onlycodeifcoded‘2’foritem3)

Code Criteria

1.Wastereduction Thepostcontainsinformationoninitiativesthatmainlyaimtoreducewastesuchasoilspillingandco2emissions

-60-

2.Sustainableforestharvesting

Thepostcontainsinformationoninitiativesthataimtosustainablyharvestforests

3.Recycling Thepostcontainsinformationoninitiativesthatreusewasteforproductionofnewproductsandmaterials

4.Noisereduction Thepostcontainsinformationoninitiativesthataimtoreducetheamountofnoiseacertainareaisexposedto

5.Restorationofindigenousplantlife

Thepostcontainsinformationoninitiativesthataimtorestoreindigenousflora

6.Sustainabilityofamanufacturingprocess

Thepostcontainsinformationoninitiativesthataimtoimprovecurrentprocesses,focusingontheinnovationinsteadofthepotentialwastereduction

7.Cannotbecoded

TEXTUALSTYLES6.Narrativestyle

Code Criteria

1.Accounting Thepostreportsoncurrentevents,newinformationorotheralterationsinthestatusofthecompany

2.Information-seeking

Thepostasksforparticularinformationthroughposingaquestion

3.Commentary Thepostrepresentsafeeling,emotionorattitudetowardssomething

4.Advertisement Thepostiswrittenaroundanadvertisement,focusesonaspecificproductpossiblyavailabletothereader

5.Competition Thepostaddressesreaderstoparticipateinacompetition

6.Announcement Thepostmakesanannouncementofupcomingeventsandactivities

7.Cannotbecoded

7.Sensoryinteractivity(image)

Code Criteria

1.Image Thepostcontainsanimage

2.Noimage Thepostdoesnotcontainanimage

3.Cannotbecoded

8.Sensoryinteractivity(video)

Code Criteria

1.Video Thepostcontainsavideo

2.Novideo Thepostdoesnotcontainavideo

3.Cannotbecoded

-61-

9.Sensoryinteractivity(interactivetext)

Code Criteria

1.Interactivetext Thepostcontainstextthatiswrittenwithcapitalletters,and/orcontainshashtags,linksandexclamationmarks

2.Nointeractivetext Thepostdoesnotcontaintextthatiswrittenwithcapitalletters,and/orcontainshashtags,linksandexclamationmarks

3.Cannotbecoded

10.Semanticinteractivity(personalizedmessage)

Code Criteria

1.Personalizedmessages

Thepostcontainsmessageslike‘goodbye’,‘thankyou’,‘readmore’and‘findout’thatseemtoaddressthereader

3.Nopersonalizedmessages

Thepostdoesnotcontainmessageslike‘goodbye’,‘thankyou’,‘readmore’and‘findout’thatseemtoaddressthereader

4.Cannotbecoded

11.Semanticinteractivity(user-friendlylanguage)

Code Criteria

1.User-friendlylanguage

Thepostiswritteninarathereasystyle,withoutmanycomplicatedwordsandjargon,makingtheinformationunderstandableforthereader

3.Nouser-friendlylanguage

Thepostiswritteninarathercomplicatedstyle,withcomplicatedwordsandjargon,makingtheinformationhardtounderstandforthereader

4.Cannotbecoded

12.Semanticinteractivity(identifyingtheaudience)

Code Criteria

1.Identifyingtheaudience

Thepostidentifiestheaudienceitaddressestheinformationtobyclearlystatinganddescribingwhomtheyarewritingto

2.Noidentifyingtheaudience

Thepostdoesnotidentifytheaudienceitaddressestheinformationto

3.Cannotbecoded

13.Formality/informality

Code Criteria

1.Formality Thepostincludesimpersonalstyle,passivevoice,complexwordsandsentences,nocontractions,notmanyabbreviations,businessandtechnicalvocabularies,politewords,objectivestylewithfactsandreferences,novagueexpressionsorslang

2.Informality Thepostincludespersonalstyle,activevoice,shortsimplewordsandsentences,contractions,abbreviations,phrasalverbs,wordsthatexpressrapportandfamiliarity,subjectivestylewithopinionsandfeelings,vagueexpressionsandslang

3.Cannotbecoded

-62-

14.Concrete/abstract

Code Criteria

1.Concrete Thepostcontainsdetails,informationfromdifferentsourcesandreferences,possiblymorepositivenews.Includepoststhatcontainavideothatgivesdetailedinformation,butexcludepostifitgivesalink.

2.Abstract Thepostonlycontainsgeneralinformationandfewdetails.Possiblytryingtoframenegativenewsasanissuethatisactivelybeingworkedon,forexample

3.Cannotbecoded

RHETORIC15.Logos(warrants/justifications)

Code Criteria

1.Warrants/justifications

Thepostjustifiesoneofthecorporation'sactionsand/orinitiativesbyexpandingonhowandwhytheactionorinitiativewasundertaken

2.Nowarrants/justifications

Thepostdoesnotjustifythecorporation'sactionsand/orinitiatives

3.Cannotbecoded

16.Logos(data)

Code Criteria

1.Data Thepostincludesdatatoelaborateontheinitiativethatiscommunicatedbythecorporationandmakethepostmoresubstaniated

2.Nodata Thepostdoesnotincludedatathatelaboratesontheinitiativethatiscommunicatedbythecorporation

3.Cannotbecoded

17.Logos(evidence/example)

Code Criteria

1.Evidence/example ThepostcontainsexamplesabouttheCSRinitiativethatisbeingcommunicated.DonotcodeifthepostwritesaboutsolelyonespecificCSRinitiativeanddoesnotrelatethistoabiggerallencompassinggoal.E.g.thepostaddresseshelpingpeoplebycreatingjobsandgivingstudentsgrantsdoesqualifyasexamples,butifthepostonlyaddressesthattheyaregivinggrantstostudents,thisdoesnotqualify.

2.Noevidence/example

ThepostdoesnotcontainexamplesabouttheCSRinitiativethatisbeingcommunicated.

3.Cannotbecoded

-63-

18.Pathos(metaphors)

Code Criteria

1.Metaphors Thepostincludesmetaphorstoexpandon,andexplain,theinformationabouttheinitiativeoractiontaken

2.Nometaphors Thepostdoesnotmetaphorstoexpandon,andexplain,theinformationabouttheinitiativeoractiontaken

3.Cannotbecoded

19.Pathos(identification)

Code Criteria

1.Identificationwiththeaudience

Thepostincorporateselementstomaketheusersfeelliketheycanidentifytothecorporations,mostlythroughculturalreferences(e.g.sports,love,friendship,health)orwordssuchas‘we’and‘sharing’,andemotion-relatedwordssuchas‘proud’and‘happy’.Donotcodeifitsays‘we’asin‘wethecompany’.

2.Noidentificationwiththeaudience

Thepostdoesnotincorporateelementstomaketheusersfeelliketheycanidentifytothecorporations.

3.Cannotbecoded

20.Ethos

Code Criteria

1.Claimstoimprovetheworld

Thepostfocusesonhowthecompany’spracticesmaketheworldabetterplace,howtheworldneedstheseinitiativesand/orhowlivinganywhereintheworldwillimproveduetotheirpractices

2.Claimstocleanuptheirownact

Thepostpresentshowtheyhavereducedtheirownnegativeimpactontheworld,e.g.decreaseinCO2emissions

3.Claimstocareaboutaudience

Thepostmentionsspecificneedsoftheaudience,addresseshowtofulfillthoseneeds,mightaskforfeedback

4.FocusesonjoiningCSRmovement

ThepostmentionshowthecompanyparticipatesinCSR,mightfocusonhowtheymeetrequirementsofinternationalsustainabilityorganizationsand/orwhatorganizationstheyhavejoinedtobelongtotheCSRmovement

5.Focusesonthirdpartiespraisingthemfortheirsocialresponsibility

TheposthighlightshowthecompanyhavebeenpraisedbythirdpartiesfortheirCSRpracticesthroughe.g.rankings.

6.Noneoftheabove 7.Cannotbecoded