1

HaveICTinnovationsturnedtheprincipleof“organizationalchoice”intotheprincipleof“infinitefreedom”?

PaperpresentedattheconferenceoftheSociotechnicalRoundtableSeptember7th-9th2016inSanFransisco

Dr.Eric-HansKramerFacultyofMilitarySciences

NetherlandsDefenceAcademyPOBox900024800PABredaTheNetherlandserichans@xs4all.nl

AbstractInnovationsinInformationandCommunicationtechnologyhavecreatedpossibilitiesinorganizationsthatpreviouslywereunthinkable.Intherealmofmilitarystudiestheso-called chaoplexic paradigm suggests that these possibilities are so path-breaking thatthey constitute a paradigm-shift in thinking about organizations. Inspired by criminaland jihadistnetworks, thechaoplexicparadigmclaimsthatmilitaryunits“in the field”can acquire far reaching autonomy as a result of a continuously updated situationalawareness,whichisenabledbysophisticateddata-linkswithotherunits.Subsequently,thenetworkofsuchmilitaryunitsisthoughttobeextremelyflexibleandagilebecauseitdoesnotneed to relyonpredeterminedhierarchicalpathsof coordination. “Order” insuchnetworksisthoughtto“emerge”asaresultofself-organizingunitsthatcoordinatewith other units. This paper discusses whether this philosophy truly constitutes aparadigm shift in thinking about organizations. Themain claim is that this is not thecase.BasedonadiscussionofacasestudyofamilitarytaskforceandananalysisbasedonNormal Accidents Theory, it is argued that the chaoplexic paradigm overlooks theproblem of “hidden” interconnections between units. All in all, it is argued that thechaoplexicparadigmunderestimatestheimportanceofhierarchyinorganizations.Itisfurthermorearguedthattheideasofthechaoplexicparadigmcanonlyworkifmilitaryunitsare“perfectinformationprocessors”withoutboundedrationalityandinextremelysmall networks. Alternatively, a perspective is worked out which emphasizes thevulnerabilityofnetworksoforganizationalunitsindynamicallycomplexenvironments.Accordingtothisperspectiveoperatorsarecontinuouslyconfrontedtowiththeneedtobalance the limitations of given structure with the most recent demands from theoperationalenvironment.

2

HaveICTinnovationsturnedtheprincipleof“organizationalchoice”intothe

principleof“infinitefreedom”?ItisindisputablethatinthelastdecadesinnovationsinInformationandCommunicationTechnology (ICT) have had groundbreaking implications for organizations. Moreover,these implications are not restricted to a particular domain. ICT innovations haveinfluenced organizations across the board. One notable development in the last twodecades has been that, due to ICT innovations, organizations are better able to adoptnetwork-forms. ICT innovations enable communications and coordination betweengroups and departments that are geographically widely dispersed. One particularlycompellingsuggestionisthatsuchICTinnovationsmightbeabletomake“bureaucracy”and functional “silos” a thing of the past.More specifically, the suggestion is that theorganizational problems associated with organizations build up on the principle of“functionalconcentration”(lackofcontrolovercentralprocesses,lackofflexibility,lackof innovationpotential, poor quality ofworking life;De Sitter, 2000)might solvedbyestablishingdirectICTlinksbetweendepartments.

Sometimesthisideaismorethanameresuggestionbutisbroughtforwardasapath-breakingphilosophyoforganizing.Intherealmofmilitarystudiesthisphilosophypoints to the potential of “robustly networked systems” and is usually referred to asNetwork Centric Warfare (NCW) or Network Enabled Capabilities (NEC). Thisphilosophypointsoutthatnetworksarepotentiallyextremelyflexibleandagilebecausethey do not rely on predetermined hierarchical paths of coordination. According toBousquet(2008,p.916),theseideasareinspiredbycriminalandjihadistnetworks.Theideaisthatunits“inthefield”canacquireautonomyasaresultacontinuouslyupdatedsituationalawareness,whichisenabledbysophisticateddata-linkswithotherunits.Thewhole systemofmilitary units needs to become a fully “robustly” connectednetwork(everynodeisconnectedtoeveryother)andissometimescalledan“edgeorganization”because of its radically decentralized character. The intended result is a collection of“swarming”unitsthat“self-synchronize”andsubsequentlycoordinationbetweenunitsisexpectedto“emerge”.Bousquet(2008,p.916)observesthattheseideasarecurrentlysoinfluentialthattheyessentiallyconstituteanewparadigminthinkingaboutmilitaryorganizationsandwarfare,whichhecalls theparadigmof “chaoplexicwar”.Bousquet(2008,p.923)onthisideaofchaoplexity:“Chaosisseennolongerassimplyathreattoorderwhichmustbeavertedatallcosts,butastheveryconditionofpossibilityoforder.Thekeynotionshereare thoseofnon-linearity, self-organizationandemergence, andthe pivotal technological figure is that of the network, the distributed model ofinformation exchange perhaps best embodied by the internet.” The very reason suchideas are considered valuable in this research community is that traditionalorganizationsareconsideredtoorigidtobeabletorespondtoveryturbulent,uncertainand dynamic environments, while the “edge organization” essentially solves thisproblem.

Theideasbehindthechaoplexicparadigmaresoradicalthattheymakeaclean

3

breakwithexistingideaswithinorganizationsciences.Assuchtheyarerightfullycalleda “paradigm”: if they work as the paradigmatic example indicates, we are left withorganizations thatrelyonprinciples thatare fundamentallydifferentcompared to theones we usually apply for understanding organizations Because the chaoplexicparadigm argues that robust networks are better able to deal with dynamicenvironments because they are far more agile it essentially argues in favor of aparadigm shift regarding some key ideas in organization-studies.. In essence, thechaoplexicparadigmseemstosuggestthatasaresultofICTinnovationshierarchycanbemadesuperfluous.Eversincethe1950’sandtheworkofHerbertSimonhierarchyisconsideredtobeacorefeatureofsystems(Simon,1962).Hierarchyreferstoanestingof subsystems within systems (Simon, 1997). Present ideas about organizationsgenerallyclaimthathierarchy–or“design”–isafunctionalpropertyofsystemsbecauseasystemwithouthierarchywouldbeovercomeby“informationoverload”.However,ifeverynodeisconnectedtoeveryothernode, theentireorganizationbecomesflatandmutual adjustment is the only relevant coordination principle. If “order” emerges insuchsystems, itactually implies thatcoordinationcostsassociatedwithhierarchycanbereducedtozero.

This paper focuses on the paradigm-shifting core of the chaoplexic paradigm. Iwillbequiteskepticaloftheclaimsofthechaoplexicparadigm.Whileitiscertainlytruethatnetwork formsare for certainorganizations in someenvironments very valuableand can be considered an important organizational innovation made possible bydevelopmentsinICT,Iwillarguethatitgoestofartothinkthathierarchyingeneralhasbecomesuperfluous.Thispaperwillofferacounterexampleagainstthisargumentandwilloffera theoreticalcritiqueofNCWandNEC.Boththeexampleandthetheoreticalcritique originate from existing research into the organizational vulnerabilities ofmilitary taskforces. On the basis of this example and the theoretical critique thechaoplexicparadigmwillbeanalyzed.ThechaoplexicparadigmIn the early years of the sociotechnical tradition, Eric Trist identified the principle of“organizationalchoice”(Trist,et.al.1963).Coreofthisprincipleistheideathatwhenitcomestoadoptingtechnologicalinnovations,organizationshaveachoice.Theinventionof the conveyor belt does not enforce a particular organizational form on theorganization. Instead, theprincipleoforganizationalchoicestates that it ispossible toadopt innovative technologyby implementing it inanorganizational formthat fits thepurposes of the organization. The chaoplexic paradigm seems to have turned thisprinciple on its head: instead of enforcing a kind ofmechanistic organizational form,technological innovations in the realm of ICT appear to offer absolute liberation: theinfinite freedom to continuously change organizational forms given the particularchallengesataparticulartime.So justasorganizationalchoice implies liberationfromtechnologicaldeterminism,thechaoplexicparadigmseemstoclaimthatICTinnovationsofferliberationfromtheinevitablyconstrainingeffectsof“choice”.Alsowithintherealmof Science and Technology Studies, this is a somewhat unusual position. According to

4

Bijker (2006) there are generally three positions in the debate of how technologyinfluences society (i.c. also organizations). In this debate, technological determinismclaimsthattechnologyhasdeterminingeffectsonsociety;socialvoluntarismarguesthatsociety determines technological developments and social constructionism argues thattechnology and society mutually influence each other. The chaoplexic seems to –somewhat paradoxically – claim that technological development determines absolutefreedominorganizationalforms.

Essenceofthechaoplexicparadigmisthe ideaisthatICTcanbeusedtocreate“robustlynetworkedorganizations”inwhicheachnodeisconnectedtoallothernodes.Subsequently,thisICTinfrastructureenablesrapidinformationsharingbetweennodes,which furthermoreenableverygooddecentralizedSituationalAwareness.This idea isspecificallyworkedoutinconceptssuchasNCWandNEC(Alberts&Hayes,2003).ThebasicsofNCW&NECarestatedasfollowsbyCebrowski(2005,p.7):

• “Arobustlynetworkedforceimprovesinformationsharing

• Information sharing enhances the quality of information and shared situational

awareness.

• Shared situationalawareness enables collaborationand self-synchronization, and

enhancessustainabilityandspeedofcommand.

• These,inturn,dramaticallyincreasemissioneffectiveness”

Self-synchronization” refers to the coordination of activities between units at thedecentralizedpositions,whichimpliesthatcommunicationrunsacrosstheorganization,rather than followinghierarchical lines. In thatcasecontrol isbelieved toemerge asaresult of local interactions (Alberts & Hayes, 2003). Or to put it differently, controlemergesasaresultofmassiveprocessofmutualadjustment.Atkinson&Moffat(2005,p.40-41)explainthisphilosophyasfollows:

ThenatureofNetworkCentricWarfareforsuchfutureInformationAgeforcescan

be outlined as: within a broad intent and constraints available to all forces, the

local force units self-synchronize under mission command in order to achieve

overallintent.Thisprocessisenabledbytheabilityoftheforcesinvolvedtorobustly

network.Wecandescribesuchassystemas loosely coupled tocapturethe localfreedomavailabletotheunitstoprosecutetheirmissionwithinanawarenessofthe

overall intent and constraints imposed by higher level command. (…). In this

process,informationistransformedin“sharedawareness”,whichisavailabletoall.

Thisleadstounitslinkingupwithotherunits,whichareeitherlocalinapsychical

senseorlocalthroughaninformationgridorintranet(self-synchronizaton).Thisin

turnleadstoemergentbehaviorandeffectsinthebattlespace”.

Aswillbecomeclearer later, thekeyword in thisquote is “looselycoupled”.The ideaapparently is that the ideas work if there is “loose coupling” between units: if theoperationsofoneunitisonlyweaklyconnectedtoothers.Proponentsofthechaoplexic

5

paradigmarguethattheseideasaresupportedbycomplexityscience(Gray,2002).Theidea is that control “emerges” in robustly networked systems as a result of localinteractions. This is based on the “order for free” idea from complexity science(Waldrop, 1992). The general idea is that in Complex Adaptive Systems (CAS) orderdoes not need to be designed. Kramer& Van Bezooijen (2016) discuss this linkwithCAS:

Waldrop (1992) discusses the example of bakeries in large cities. Together these

bakeriesareresponsibleforasignificantpartoffoodsupply.Takentogetherthese

bakerieshandletheissueoffoodsupplyremarkablyadequate.Yet,noonetellsthe

individualbakerieshowmuchbreadtoproduce,norhowtoorganizetheirlogistic

supply. Incontrast,makingthe foodsupply incities theresponsibilityofacentral

officethatissuesdetailedorderstobakerieswouldbearecipefordisaster.Soina

sense,foodsupplyinlargecitiesisacomplicatedprocessthatis“orderedforfree”.

Sotheideaofthechaoplexicparadigmisthat“order”canemergeinamilitarytaskforcein a likewise manner. General McCrystal in his book Team of teams (2015) refers toelementsofthisphilosophyinordertodescribethemostrecentinnovationsintheUSArmy. The image that emerges from this book is that the US Army started atransformation towards thisphilosophy. In the forewordof thebook,Walter Isaacsonclaimsthattherelevanceofthisbookliesexactlyinthisrealm(Isaacson,2015,p.viii):

“Management models based on planning and predicting instead of resilient

adaptation to changing circumstances are no longer suited to today’s challenges.

Organizationsmustbenetworked,notsiloed,inordertosucceed.

AboutthedevelopmentsintheUSArmyIsaacsonstates(2015,p.viii-ix):

“Oneconclusiontheyreachedwasthatagilityandadaptabilityarenormallylimited

tosmallteams.(…).“Theprimarylessonthatemerged,andisdetailedinthisbook,is

the need to scale the adaptability and cohesiveness of small teams up to the

enterprise level. This involves creating a team of teams to foster cross-silo

collaboration.”

Subsequently, in the book McCrystal alludes to (but not specifically mentions) thechaoplexic philosophy at differentmoments and concludes it has been successful. Forexample(2015;p.164):

“We knew that forging the neural network that would facilitate our emergent

analysis of complexproblemswas vital forour long-term success, sowedesigned

prepackaged communication bundles that our teams could take into the field,

wherevertheywereintheworld.LikeNASA,weinvestedinbandwidthtoenableus

toreacheverycomponentofour forceandourpartners, fromausterebasesnear

6

the Syrian border to CIA headquarters at Langley, Virginia. Satellite dishes, fromsmalltohuge,connectedtheforce.Securevideoteleconferences,chatrooms,aWebportal,ande-mailbecamekeyarteriesofourcirculatorysystem.Technicallyitwascomplex, financially it was expensive, but we were trying to build a culture ofsharing:anymemberof theTaskForce,andanyof thepartnerswe invited,couldeventually dial in to theO&I securely from their laptops and listen through theirheadphones.”

OneofthemainmessagesofMcCrystalisthattheimplementationofthisphilosophyhasintegral effects. For example, it requires a different kind of leadership compared totraditionalorganizationsandadifferent“culture”.

To qualify the claims of the chaoplexic paradigm somewhat in advance of theactualargumentof thispaper, the followingpassage fromaNATOpublication isquiteinteresting. In apublicationof theNATOdefense college, El Fartasi&DeVivo (2016)startfromthebeliefthatworkingwithsophisticateddata-linksinnetworkedsystemsisalreadyareality,althoughthequotebelowsuggeststhatitisnotquiteusedinthewaythechaoplexicparadigmsuggests:

NATOhasalwaysmaintainedatechnicaladvantageoveritspotentialadversaries.Through its Alliance Ground Surveillance System (AGS), consisting of five GlobalHawk Unmanned Vehicle (UAVs) and Ground Control Stations, the Alliance willhavea21stcenturysystemthatwillsupporttheCommander’scomprehensivereal-time and near real-time situational awareness at strategic distance and withinhours,ifnotminutes,ofacrisisarising.AGSwillgathermassiveamountsofdatatoenablesituationalawarenessoflandandmaritimeenvironments;butwhatuseareaGlobalHawkanditsgroundstations,withouttheITinfrastructurethroughwhichits data is communicated for possible use by decision-makers?NATO’s IT enablessurveillanceandintelligencegathering,servingasNATO’seyesintheskyinlightofincreasedairspacesecuritybreachesonitsborders.Ensuringcontinuoussituationalawarenessinspecificareasofinterest,duringcriticalperiodsofheightenedtension,is therefore of the utmost importance. Without the current IT architecture tosupport operational capability, both the Alliance and nations risk makinguninformed decisions with incomplete and fragmentary data, with results thatcouldleadtopotentiallyseriousunintendedconsequences.

Thisquoteactuallysuggeststhatdatalinksarenecessaryforsituationalawarenessofacentralcommanderwhoisakeydecisionmaker.Thisideaisactuallytheveryoppositeofwhat the chaoplexic philosophy proposes and it is also different from the spirit ofMcCrystal’sargument.Thisis,however,aclassicalcontradictioninorganizationstudies:willsophisticatedtechnologybeusedtofosterself-organization,ortofostercentralizedcontrol?Thechaoplexicparadigmarguesinfavoroftheformer,butthelattertendencydoes not seem to be totally absent in its implementation. Apparently, the newtechnologyalsofeedsintotheolddreamofperfectcentralization.

7

Thechaoplexic“imageoforganization”The chaoplexic ideas are particularly interesting from the perspective of the

sociotechnicaltradition.Theyappeartobeabletoovercometheproblemsofoneofthe

main enemies of the sociotechnical tradition: the large scale specialized bureaucratic

organization built upon the ideas of functional concentration (concentrating a single

specialism in a single department). Furthermore, they aim to fight traditional

bureaucracy by advocating radical decentralization and self-organization, the very

topicsthathavebeenembracedbythesociotechnicaltraditionfordecades.Soinasense

the ideas of the chaoplexic paradigmmight appear as a validationof some core ideas

from the sociotechnical tradition. At the same time, they might seem to change the

nature of these sociotechnical ideas quite considerably. After all, the sociotechnical

tradition has developed alternatives for traditional bureaucracies. The chaoplexic

paradigmseemstosuggest thatmerelyconnectingeveryunit toeveryotherunitwith

sophisticateddatalinksistheonlydesigneffortittakesforachievingthesociotechnical

ideals.Inawaytherefore,itmightappearthatthechaoplexicparadigmisessentiallyan

innovationthatmakes(sociotechnical)organizationaldesignatopicofpasttimes.

In essence, the chaoplexic paradigm is based upon a different “image of

organization”comparedtothe“hierarchyparadigm”.Thisistheimageof“organizations

asantcolonies”(Dekker2011,p157).Kramer&Moorkamp(2016)describethisimage

asfollows:

“This image focuses on the complicated organized behaviour that is displayed byant colonies,while no single ant has designed the colony, or understands it in itsentire complexity. The organized behaviour of the colony is a macro-levelphenomenonthatemergesfromthebehaviourofinteractingants.”

Sidney Dekker uses this image of the ant colony to emphasize the idea of organizedcomplexity(2011,p.157):

“Itiscomplexbecausetherearealargenumberofcomponents,and,asaresult,adense throng of mini-programs running and interacting, cross influencing eachother”.

ThiscomplexityisorganizedaccordingtoDekker(2011,p.157):

“Butwhatitproducesisnotdisorganized.Ratheritisorganized(...)asanamazingemergent product of the complex interactions between a multitude of simplerentities”.

Sotheideaofemerging“organizedcomplexity”replacestheideaoforganizationdesign.

Kramer&Moorkamp(2016)emphasizefurthermore:

8

“Dekkerconsiderscomplexsystemsbothresilientandvulnerablebyvirtueoftheircomplexity,whichhedeemsaparadox.”

This idea of vulnerability of complex systems is unavailable in the ideas of NCW andNEC, but Dekker offers no analysis of how resilience and vulnerability relate1. Morespecifically,Dekkerstatesthat(2011,p.157):

“Becausetheyconsistofcomplexwebsofrelationships,andbecausealotofcontrolis distributed rather than centralized, complex systems can adapt to a changingworld”.

However,

“Complexity opens up a way for a particular kind of brittleness. Their opennessmeansunpredictablebehaviour”(Dekker2011p153).

So if structure can be understood bymeans of the idea of “organized complexity” anorganizationisprincipallyseenasan“infinitelymalleable”tissuethatcancombineandrecombine elements in infinite ways in order to adapt to the newest environmentalcontingencies, without the burden of a previously designed hierarchical framewhichseverely limits themalleabilityof the tissue. Shedding thisburden is specificallywhatcreates the agility of the tissue. While the chaoplexic paradigm emphasizes theadaptability and resilience of “organizations as ant colonies”, others adhering to thisimageoforganizationalsoemphasizevulnerability.Counter-example:amilitarytaskforceinanexpeditionarymissionThecounter-examplethatisofferedhereisacasestudyofamilitarytaskforce.Thiscasestudy was developed at the Netherlands Defense Academy by Moorkamp, and isdescribedindetailinMoorkamp&Kramer(2014).Thecasedescriptionaimstoexpressthe idea that, although the conditions of the taskforce seem to be very fit for thechaoplexic paradigm, the organizing principles behind the chaoplexic paradigm couldnotpossiblyhaveworked.

In the last twodecades, the focusof theNetherlandsArmedForceshas shiftedfrom preparing for all kinds of Cold War scenarios, to what is called “expeditionarymissions”.ThesearemissionsinwhichtheArmedForcesgoon“expedition”–goabroad– in the context of peacekeeping, peace enforcing missions, post conflict rebuildingmissions, etc. In this time period, The Netherlands Armed forces have contributed tomissionsinforexampleBosnia,Iraq,AfghanistanandMali.Typicalforsuchmissionsisthattaskforcesaredeployed.Taskforcesaretemporaryconstructionsthataredeployed

1Intheabsenceofsuchananalysis,thisparadoxisunconvincing.Itratherseemslikeapredeterminedanswertoeverypossiblequestion.Isthereanaccident?Itiscausedby“organizedcomplexity”.Istherenoaccident?Itiscausedby“organizedcomplexity”.Remarkablyresilientbehavior?Itiscausedby“organizedcomplexity”.

9

for the purposes of a singlemission. Basically, a taskforce is designedby establishingwhatkindsof“assets”or“capabilities”areexpectedtobenecessaryinagivenmission.TheseassetsandcapabilitiesaresubsequentlyselectedfromthestandingarmyinTheNetherlands.Itwillbeobviousthatsuchtaskforcesneedtofindadesignthatworksinaparticular environment. As such, expeditionary taskforces are underdevelopedorganizations.

Onthebasisofananalysisofresearchintosuchexpeditionarytaskforces,Kramer&VanBezooijen(2016) identify twomainorganizationalchallenges. Inthe firstplace,typicallyexpeditionarytaskforcesfaceenvironmentswithmanyunknowns.ThisiswhatKramer (2007) refers to as “dynamic complexity”. These environments constantlychangeunder the influenceofdifficult to interpretdynamics. In thesecondplace,asaconsequenceoftheseenvironmentalconditions,taskforcescontinuouslycalibratedtheirinternal design to fit the conditions that they found themselves in. In the case of theDutch Uruzgan mission, Kramer et.al (2012) concluded that: “(…) previous to thedeploymentoftheDutchtroopstheAfghanprovinceUruzganwasbasicallyawhitespotonthemap.Furthermore,duringdeploymentthelocalcircumstancesprovedtobeverychangeable; the force of the Taliban opposition could vary significantly. Operationalunitswereoftenthefirsttoexperiencechangesintheenvironmentandactedthereforein a general sense as explorers.With this inmind, it is quite impossible to design anorganizationina“one-shotdesigneffort”thatcandealwithallcurrentandfuture–asyetunknown–contingencies.”

AspartofhisPhDproject,Moorkampstudiedsafetymanagementstrategiesinaparticularunit thatoperatedanUnmannedAerialVehicle(UAV)aspartofTaskForceUruzgan(TFU)(Moorkamp&Kramer,2014;Moorkampetal2015).TFUwasataskforcethatwasdeployedby theNetherlandsArmedForces, inorder tostartupa rebuildingprocessintheAfghanprovinceofUruzganfrom2006to2010.Thisparticularunit,107AerialSystemsBattery(107ASBt),wasaddedtoTFUfordoingAerialReconnaissanceand as such they contributed to the intelligence organization in TFU. 107ASBtwas aunitthat,atthetime,wasareconnaissanceplaneoftheArtillery.ThatmeantthattheyweretrainedasapartoftheArtillery.Oneparticularconsequencewasthattheyusedtotraininarestrictedairzone,becauseobviouslyanArtillerybattalioncannottrainwithall kinds of air-traffic flying over. A consequence was however, that they were notfamiliarwithoperating in an areawithother air spaceusers such as inUruzgan.Theconsequencewasthatpartoftheself-organizingactivitiesof107ASBtwerefocusedon“discovering, establishing and refining rules and procedures for interaction, such asrules foravoidingcollision in theair, or ‘‘deconfliction”” (Moorkampet. al.2015,p.6).Thisamountedbasicallytolearninganentirenewtradeduringdeployment(“whatdoesacontroltowermeanwhentheysayX,YorZ”).Thislearningprocesspartlywaspartlyatrialanderroraffair,withsomenearmisses,mostnotablyaUAVnearly flying intoanApache helicopter after launch (“not a near miss, but a near hit”, according to oneoperator). Moreover, it is very questionable if the operators of 107 ASBt eventuallydeveloped a full understanding of the whole spectrum (the exact meaning of“deconfliction”, “air traffic control”, Restricted Air Zone’s (ROZ), etc), which was

10

evidencedbythemcontactingthecontroltoweronhowtooperateinaROZinsteadofa

ForwardAirController(FAC-er).MorespecificallyMoorkampet.al(2015,p.6)conclude

that:

“Themajorityoftheself-organizingstrategiesof107ASBt’soperatorswereaimed

attryingtosolveproblemswithregardtotheamountandtypeofassignmentsand

air space users within TFU were a major source of problems for 107 ASBt’s

operators.WithinTFU,107ASBtwaslocatedinanextensivenetworkofotherArmy

andAir Forceunits. This organizational configurationdiffered substantially from

theconfigurationinwhichtheyweretraininginTheNetherlands.InUruzgan,107

ASBt was gradually included in TFU’s complicated primary process that varied

quicklyduetothedynamiccomplexmissionareainUruzgan.”

Apartfromalackofknowledgeof107ASBtonhowtooperateintaskforceconditions,

therewasalackofunderstandingonthepartoftherestofTFUonhowtoexactlyusea

UAV.ThatmadeintegrationoftheUAVintoTFUsignificantlymoredifficult.Thecontrol

towerdidnotknowtheycouldn’tordertheUAVtodepartataspecificmomentbecause

their enginewas not suited for that, and they didn’t know the UAV’sweren’t able to

deconflict on the basis of the prevailing “see and avoid” strategy (because as anUAV

doesn’thaveapilotandthisUAV’scamerawaspointedtotheground).Intelligenceunits

did not know that UAV’s couldn’t be deployed to search for a red car because their

camerawasblackandwhite.Furthermore,thedifferentunitsintheTFU-networkwere

surprised to findout that theUAVwasn’t able to contact theparticipantswithwhom

theiroperationsinterconnected,becausetheirradiowasnotsuitedforthat.Afterall,the

UAVwasnevermeantto interconnectwiththekindsofunits inTFU.Theseareonlya

coupleofexamples.GiventhatpersonnelinTFUrotatedquitefrequently,differentunits

continuedtoorganizemeetingstomakeclearwhatthespecificsoftheUAVare,whatit

canandcannotdo,etc.Allinall,Moorkampet.al(2015,p.8)conclude:

“(…) 107ASBt’s operators developed numerous self-organizing strategies in their

attempt to create safe and controlled UAV operationswithin TFU. In effect, they

were constantly designing and redesigning structures of both their own unit and

TFUinordertodevelopsafeoperationswithinTFU.Theresultssectionhasshown

that107ASBtandotherunitswithinTFUweredifferentiatedsubstantiallyand,as

such,107ASBtlackedintegrationintoTFU’sproductionandcontrolstructure.Asa

consequence,107ASBt’soperatorsstartedimprovisingandpioneeringwithinTFU

in order to develop such integration. Along the way, they succeeded in tackling

variousproblemsthattheyencounteredwhileoperatingwithinTFU.”

Theproblemsinself-organizingtheirwayoutoftheinteractionstheywerepartsuggest

that operators in 107 ASBt suffered from “information overload”. The network was

initiallytoocomplextounderstandhowtheiroperationsrelatedtootherparticipantsin

11

the TFU network. Only after being some time in themission area, and routineswere

workedout,thesituationstabilizedsomewhatfor107ASBtinTFU.

Istheexpeditionarytaskforcearealcounter-example?The example of the expeditionary taskforce is considered here a counter-example

against the chaoplexic paradigm. It suggestswhy a network confrontedwith dynamic

conditions does notwork like the chaoplexic paradigm specifies. In the first place, is

necessarytoarguewhythisexampleisparticularlysuitable.ItisclearthattheTFU,with

107 ASBt as a part, typically faced the conditions forwhich the chaoplexic paradigm

seems to bemeant. Furthermore, their tendency to keep changing their organization

design seems to refer to a need for an ability to constantly change form. TFU was

essentiallyanunderdevelopednetwork.Operatorsworkedhardtodevelopthenetwork

into an organized system that isworkable. They found out latent inputs that affected

their particular node and tried to explicate these and somehow organize this latent

input. By organizing the interaction with other “nodes” around them, they build

structure,and inawayonecanunderstand thatascontrol thatemerges throughself-

organization. As such the example seems suitable to discuss in the context of the

chaoplexicparadigm.

Against thisbackgroundit isclearthatwhatwasobservable inthecase isa far

cryfromwhatthechaoplexicparadigmformulatesasideal.Whatappearsfromthiscase

isthat“control”didnotemergeasaresultofinteractionsbetweennodes,andthatthis

resultedinasuper-agiletaskforce.Onthecontrary,operatorsstruggledtounderstand

thesystemtheywerepartofandtheinterconnectionsthatinfluencedtheirwork.They

did not have perfect situational awareness, which would have enabled “self

synchronization”.All inall,TFUwasnotawell-organizedsystem(asevidencedbythe

safetyincidents).Whatisobservableinthecaseisindeedself-organization,butnotthe

kindofself-organizationthatisatriumphofimaginativeproblemsolving,basedonswift

coordinationbetweenunits.Instead,thekindofself-organizationthatisclearfromthe

caseresemblesmoreadesperateattemptbyoperatorstokeepaveryimperfectsystem

afloat.Theconclusionsfromthiscasearethereforetheoppositeoftheclaimsmadeby

thechaoplexicparadigm.

However,itwouldbetooeasytostopatthisconclusion.Ifthiscaseisinterpreted

as a falsification of the chaoplexic paradigm, two quite obvious possible counter-

argumentsappeartoberelevant.Inthefirstplace,107ASBtappearsasanexampleofa

quite exotic unit thatwas uniquely unfit to be integrated in TFU.Not onlywere they

forced to operate in conditions that differed fundamentally from the way they were

trained to operate – integrating a UAV in a mission was a first for the Netherlands

ArmedForces–soitmightbeexpectedthatthereweresomeproblems.Inthesecond

place, TFU was apparently not at all built up on the basis of the principles of the

chaoplexicparadigm.Sinceunitssometimes foundona trialanderrorbasis that their

operations interconnected, TFU was apparently not “robustly networked” and

apparently lacked sophisticated data-links between units, which made information

sharingandmutualadjustmentproblematic.Ifanythingthesetwopointsappeartobein

12

support of the chaoplexic paradigm. Furthermore, apparently there was a process ofself-organization going on in TFU thatwas aimed atworking out an underdevelopedtaskforce structure, and all in all there weren’t that many incidents (although therecertainlyweresome).SodespitethefactthatthewayTFUoperatedwasafarcryfromthe idealof thechaoplexicparadigm,still itwasself-organizationand the tremendouseffortofoperationalpersonnelthatkeptthingsgoinginTFU.

While both points are valid, they are not considered valid counterargumentsagainsttheparticularpointthatismadehere.Inthefirstplace,Moorkampdidvalidatethe107ASBtcaseandindeedfoundcomparabledynamicsbothwithinTFUasinothermissions (Moorkamp& Kramer, 2014;Moorkamp et al, 2015), so although 107 ASBtmighthavebeenunfit,itwasn’tuniquelyunfit.Moreimportantisthequestionif“robustnetworking”withsophisticatedICTconnectionswouldhavesolvedtheissuesofcontrolinTFU.ThevulnerabilityoftheTFUnetworkwasnotcausedbyalackof“information-flow”,orabotchyinformationflowbetweenunits(althoughpartlytheystruggledwithfor example incompatible radios). It was a lack of understanding of the differentreciprocalinterdependencesbetween“nodes”inthenetwork.Theybasicallyfoundouttheseinterdependencesbytrialanderror.Onlywhentheproblemoccurred,theyfoundoutthattheiractivitiesinterconnectedwithotherunitsinthetaskforce.Eveniftheyhadperfectmeans of communications, and couldhave contacted all other unitswheneverthey liked, they wouldn’t know what to ask them before they understood how theiractivitiesinterconnected.Inotherwords,anissueoforganizing(“howdomyactivitiesinfluenceX andhowareX’s activities influencingme”) is of a different order than anissue of informing (“what is X doing at thismoment”). Situational awareness ismorethan“acquiringinformation”.Beingconsciousofhowtheoperationsofaparticularunitmight affect the operation of another requires an (sometimesdeep) understanding oftheotherunit.This cannotbe reduced to a sort general knowledgeof the specifics ofparticularunits,but itcanbeamatterofdetail.Mereinformationprocessingdoesnotsolve the problem that a particular UAV cannot launch at predetermined specificmoment because it needs to wait unit the engine is sufficiently warmed up. So thecontrol tower learned that it needed tounderstand specific traitsof the engineof theUAV inorder to coordinatewith themeffectively,while theUAV-unitneeded to learnthatsuchdetailswereimportanttothecontroltowerinthefirstplace.TheargumentagainstthechaoplexicparadigmThe expeditionary taskforce as studied above is considered a counterexample here,because it pointsout a specificproblemof large interconnectednetworks.That is theissue of “hidden” and “not-understood” interconnections between elements of thetaskforce.AsTFUwasbasicallyanunderdevelopednetwork,partsoftheexpeditionarynetworkwereinterconnectedinawayoperatorsdidnotrealizeuntilreal interactionsemerged.Subsequentlyoperatorstriedto“structure”ortoorganizetheseinteractions,without being aware if their local solution to the negative effects of a particularinterconnection,affectedotherthetaskforcebeyondtheirperspective.Inthatway“self-organization”might be “emergent”, but not every “emergent pattern” is by definition

13

good(Kuipers,et.al,2010).Essentiallytherefore,theTFUexampleshowstheproblemofcontrolling (hidden) interconnections between “nodes” of the network with possiblenegativeconsequencesfortheoperationofthenetworksasawhole.

Thepossiblenegativecascadingeffectsofinterconnectionsbetween“nodes”inasystemhas been described byNormalAccidents Theory (NAT), developed by Perrow(Perrow,1999).NAT’sessentialpoint is thataparticularkindof“cascading”accidentscanbe “normal” in someorganizations.More specifically, such “normalaccidents” canpotentially occur in organizations that operate in disruptive environments because inthatcaseelementsinteractinawaythatcannotbefully“designed”.Ifthatisthecase,asystem is characterized by complex interaction. If there is, furthermore, a significantinterdependence between elements, a disturbance has the potential to resonatethroughoutasystem(tightcoupling).Organizations thatarecharacterizedbycomplexinteractionandtightcouplingarethereforeparticularlyvulnerabletocascadingfailures.

Van Bezooijen & Kramer (2014) and Kramer & Van Bezooijen (2016) usePerrow’sNATtocriticizethechaoplexicparadigm.Theirargumentisthatthechaoplexicparadigm is specifically focused on creating complex interaction and tight coupling.Giventhatmilitarytaskforcesaremeanttooperate inturbulentconditionsthatcreatecomplexinteractions,VanBezooijen&Kramer(2014,p.14)conclude:

“(…) IT is implemented inmilitaryoperations tomake themmoreagile, enablingmilitary units to react quicker to environmental change by removing time slack.Being quicker is a form of tighter coupling. In other words, designing for agile,emerging collaborations between networked units without direct supervisiondirectlyinfluencesthepotentialfornormalaccidentstohappen.”

SowhilepreviouslyAtkinson&Moffatclaimedthattheideasofthechaoplexicparadigmwork if there is loose coupling,VanBezooijen&Kramer (2014) claim that inmilitarytaskforces,buildupofspecializedunits,aretightlycoupled,particularlyiftheattemptistocreatemoreagileorganizations:morespeedmeans that the interrelationsbetweenunitsbecometighter.Furthermore, the tendency inArmies tocreatespecializedunits,also creates tight coupling. In armies an element is still thought to be the basicspecializedmilitaryunit: an infantryplatoon, anApache squadron, anArtillery-canon,an UAV, etc. These basic elements are called “assets” and the basic thought is thatmilitarytaskforcerequiresacollectionofassets.Basically,an“asset”consistsofacreworganized around some technological system or a basic collection of soldiers. Whiletraditional armies offer battalion and brigade structures which combine theseinterdependentunits into largerorganizationalclusters, thechaoplexicparadigmaimstokeepbasicelementsandclaimsthatthecoordinatedactionbetweeninterdependentelementscanbeestablishedbyconnectingthemthroughICT.

Whatisnotconsideredbythechaoplexicparadigm,isthatbyorganizingaround“assets”whicharespecializedunits, itbasicallychooses for theprincipleof functionalconcentration (De Sitter, 2000). The effect is that if the taskforcewants to conduct acertain task it needs to assemble different units,which leads to structural complexity

14

(moreinterconnections)inthenetworkasawhole,andtightcouplinginsteadofloose

coupling2.

If the chaoplexic paradigm depends on “loose coupling” it should invest in

creatingbasic“nodes”thatarelooselycoupledtothenetworktheyarepartof.However,

this point never enters the discussion. Ina sense, thechaoplexicparadigmattempts toconservesilos,withaclaimed“magical”solutionfortheproblemsofcoordinatingbetweensilos(controlwillemerge).Allinall,Kramer&VanBezooijen(2016)concludethat:

“The chaoplexic paradigm is considered here as a theory that combines anopportunistic reading of complexity science with an improbable idea aboutdesigningmilitary taskforces.This certainly isnoargumentagainst theuseof IT.Instead, theanalysis shows thatbeingable toharvest thepotential benefits of ITrequiresasophisticatedunderstandingofthewayorganizationalsystemsworkindynamicallycomplexenvironments.”

Perrow calls such discussions about network centric operations “bulleted boy scout

homilies”,thatpromiseafutureofsuper-agility,iftheArmedForceswouldjustinvestin

X,YorZ(Perrow,2004).

HierarchyandthechaoplexicparadigmNAT can be seen as a theory from the existing “hierarchy-paradigm” in organization

studies. It is thereforenotvery surprising thaton thebasisof sucha theory it canbe

claimedthatthenewparadigmiswrong.Inessence,theexistingparadigmrestsonthe

ideathat“hierarchy”isanindispensibleaspectoflargerorganizations.JamesThompson

(2008,p.59)claimsregardinghierarchythat:

“Itisunfortunatethatthistermhascometostandalmostexclusivelyfordegreesofhighnessand lowness, for this tendstohidethebasicsignificanceofhierarchy forcomplexorganizations.Eachlevelisnotsimplyhigherthantheonebelow,butitisamoreinclusiveclustering,orcombinationofinterdependentgroups,tohandlethoseaspects of coordination which are beyond the scope of any of its components”(emphasisinoriginal).

Simon (1962) argues that hierarchy simplifies systems by reducing the amount of

information that needs to be processed. This is what Simon (1997) calls a “nearly

decomposablesystem”:itisasysteminwhichtheinternalcomplexitywithinanodeis

greaterthanthecomplexityofthenetworkofnodes.Bystrivingformutualadjustment

withinunitscoordinationcostsareminimized(Thompson,2008,p.57).Thecoreofthis

“hierarchy paradigm” is that an organization confronted with a dynamically complex

2Ifalargeandcomplexnetworkofunitsis“robustly”networked,theresultingnetworkbecomes

extremelycomplex.Moreparticularly,ataskforcelikeTFUwith49unitswouldonthebasisoftheformula

n(n-1)/2have1176relationsbetweenunitsiftheywouldberobustlynetworked.However,ifthe

taskforcewouldconsistofjust4units,therewouldonlybe6interconnections.

15

environmentneedtoreducetheuncertaintyofthisenvironmentbymeansofhierarchy,

i.e.creatinganorganizationalstructure.JamesThompsonreferstotheassumptionthat

structureisthevehiclebywhichorganizationsachieveboundedrationalityandclaims

that(2008,p54):

“Bydelimitingpossibilities,controloverresourcesandothermatters,organizationsprovide theirparticipatingmemberswithboundarieswithinwhichefficiencymaybeareasonableexpectation.Butifstructureaffordsnumerousspheresofboundedrationality, itmustalso facilitate thecoordinatedactionof these interdependentsystems”(emphasisinoriginal).

Thisassumptionininlinewithmostrecentformulationsinsystem’stheoryinthesense

that systems need to “reduce complexity” before they can deal with a dynamically

complex environment (Blom, 1997; Christis, 1998; Kramer 2007). Although the

chaoplexicparadigmisbasedoncomplexityscience,whichisalsoavariationofsystems

theory,thisparticularassumptionisnotsharedbythechaoplexicparadigm.Infact,the

chaoplexic paradigm seems to suggest that systems can become as complex as their

environment3. The assumption that systems in complex environments need to “limit

possibilities”inorderto“reducecomplexity”mightforsomebecounter-intuitive.Many

will point to the opposite: hierarchies are actually so restricting that they limit the

abilitytodealwithuncertainty.

Thishasinfactalwaysbeenamainpointofthesociotechnicaltradition.Itisno

coincidence that Phil Herbst called his 1976 book about the sociotechnical tradition

Alternativestohierarchies(Herbst,1976).Thehistoryofthesociotechnicaltraditionhasbeen about how to overcome the restricting characteristics of “classical” bureaucratic

andmechanicalstructures.Ifthereisonemessagecentraltothesociotechnicaltradition

itisthathierarchyisoften“theproblem”ratherthanthe“solution”andcertainlynotthe

solutiontodealingwithdynamiccomplexity.Thismightexplainalotoftheallureofthe

ideas of the “chaoplexic paradigm”: ICT-innovations might finally able to beat the

archenemyofthesociotechnicaltradition.However,ifthe“hierarchy-paradigm”claims

that “limiting possibilities” is the way to deal with uncertainty, the claim is not that

uncertaintycanbetotallytransformedintocertainty.Instead,theideaisthatinorderto

be not overwhelmed by the infinite complexity of the environment, a system needs

assumptions,i.e.somedegreeofselectivityinthewaytheenvironmentisengaged.This

inevitably leads to the position that there is no perfect way to deal with dynamic

complexity. Selectivity is inevitable, but is also inevitably blinding and hierarchy is

indeed inevitably constraining. This is a core problem of organizing according to the

hierarchyparadigmthatcannotbeovercomebyamagicalICTsolution.

The essential difference between the paradigms becomes obvious. A hierarchy

“limits possibilities”, while previously it was argued that the chaoplexic paradigm is

based on an idea that an organization is a tissue that is “infinitely malleable”. One

3ThisisthereasonwhyKramer&VanBezooijen(2016)claimedthatthechaoplexicparadigmisbasedon

anopportunisticreadingofcomplexityscience.

16

paradigm states that hierarchy is necessary as a means to be able to deal with

uncertainty.Accordingtothehierarchy-paradigmanorganizationwouldateverypoint

be overloaded with possibilities if there were no hierarchy. The other states that

hierarchycanbeabolishedbecause it in factprohibitsdealingwithuncertainty. Inthe

lattercase,no“organizationalchoice”wouldbenecessary:anorganizationcankeepall

possible structures open and let the “right” one emerge out of interactions between

nodes.Theconclusionhereisthattheideasofthechaoplexicparadigmcouldonlywork

if lowerlevelunitsare“perfectinformationprocessors”without“boundedrationality”.

ThisseemstobeaveryunrealisticassumptionandthatiswhyIconsiderthecoreclaims

ofthechaoplexicparadigmunconvincing.

ThevulnerabletaskforceAttheendofthispaper,someideaswillbediscussedthatseempromisingasawayto

understandnetworkedorganizationssuchasTFU.Coreofthisperspectiveistheideaof

vulnerability.Theprevious leadstotheconclusionthatorganizations incertain–very

turbulent – environments are by definition imperfect. On the one hand they need to

“limitpossibilities”becauseof “boundedrationality”and “informationoverload”while

on the other hand they need openness. The very problem of organizations in

dynamicallycomplexenvironmentsisthatontheonehandtheyhavenochoicebutto

“limitpossibilities”becauseof “boundedrationality”and “informationoverload”while

on the other hand they have no choice but to be open (see also Kramer, 2007).

Accordingtotheperspectivehere,thisdialecticbetweenlimitationandopennessisthe

essential problem of organizations that creates their inherent vulnerability in very

turbulentenvironments.

This very dilemma is clear in JamesThompson’s discussion of the problems of

whathecalledthesyntheticorganization.Thissyntheticorganizationthat isaccordingtoThompsonanadhocorganizationthatemergestodealwithnaturaldisasters(2008,

p.52).Soinfact,itisnotunliketheexpeditionarytaskforcediscussedabove.Thompson

claims of this synthetic organization that it is not efficient: some resources are not

deployed,andsomeareemployedatcross-purposes(2008,p.53).Heclaimsthatthisis

thecasebecause(2008,p.53):

“Perhaps the overriding reason is that the synthetic organization mustsimultaneouslyestablishitsstructureandcarryonoperations.Underconditionsofgreatuncertainty,itmustlearnthenatureandextentoftheoverallproblemtobesolvedandthenatureandlocationofrelevantresources.Atthesametimeismustassembleandinterrelatethecomponents,anditmustdoallthiswithoutthebenefitof established rulesor commonlyknowchannelsof communication.The syntheticorganizationcannottakeinventorybeforeswingingintoaction.(…).Thesyntheticorganization for disaster recovery is inefficient by technological or economicstandards because it must order the actions of its components in a situation ofinterdependence and in the face of uncertainty as to where and how thatinterdependence exists. It can be presumed that efficiencywould be higher if the

17

synthetic-organization headquarters knew in advance either the extent of theproblemtobesolvedorthefullarrayofresourcesavailabletoit.”

Sothe“syntheticorganization”certainlyisstructured,butisfarfromideallystructured,or even “finished”. It is in fact underdeveloped andworks out its structure as it goesalong.Itisquiteinterestingtonotethat–inapublicationthatisgenerallyconsideredtobeoneofthemostfundamentalonesinorganizationscience–specificallydiscussestheveryorganizationalissuesthatwereobservableintheexpeditionarytaskforcediscussedabove. What is discernible in this quote of Thompson is that in such organizationsstructureandprocessbecomedirectlyrelated.Kramer&Moorkamp(2016)claimabouttheclassicaldifferencebetweenstructureandprocess:

“Whilestructurereferstofixed–noun-like–characteristicsoforganizations,suchas their structural design, process refers to the way activities – verbs – withinsystemsareorganized.”

Syntheticorganizationsneedto“swingintoaction”beforetheyhavefullyworkedoutastructurethatwouldworkinaparticularenvironment.Thatmeansthatoperators“act”(process) without fully understanding the interconnections in the organization(structure). In these cases it is local problem solving (“process”) that creates fixedrelations in an organization (“structure”). The idea of “the organizational skeleton”developed by Weick specifies this relation between structure and process (Weick,2005). He uses a definition by Bate and Pyewho claim that: “design is a bare bonesframework on which a more organic, emergent, social structure develops as peopleinteract,argue,fallout,cometogether,andotherwisemanagetheirdaytodaysituation”(Bate and Pye 2000, p.199). In developing local problem solving strategies (process),operators need to understand how their behavior is limited by the behavior of otherunitsthatact(structure).IfyouwanttoflyanUAVinUruzgan,itisimportanttotrytounderstandhowthebehavioroftheUAVisrelatedtootherflyingunitsinthesamearea,even though there might not be an existing structure in place that specifies yourinteractionswitha control tower. Sooperators in such systemsdon’twork “within”astructure,theireverydaywork(i.e.“normalwork”)isspecificallyorientedatbalancingstructureandprocessandchangingeachofthemasyougoalong.Theireverydayworkis partly “designing”. Because of the sheer complexity of this task, such systems willalwaysbevulnerable:actsofoperatorsinresponsetoenvironmentalcontingenciescanthreatenthefunctionalintegrationoftheentirenetwork.Kramer&Moorkamp(2016):

“When this perspective on organizations is used it is obvious that they areinherentlyvulnerable.Theirdesignisprincipallyoutdatedagainstthebackgroundof a dynamically complex environment. In such environments, organizations areseenas principally unsettled. Functional integrationneeds to be established formmoment to moment, and the resulting coherence is not a sort of system-widesymphoniousharmony,butonethatisgoodenoughtosurvivetothenextmoment.

18

In normal work, the conflict between on the one hand the environmentaldisruptions,andontheotherhandtheconflictwiththeexistingstructuraldesignofanorganizationbecomesvisible.Thesolutions foundbyoperatorscanpotentiallyleadtovulnerabilitybecausetheywereunawareofcertaindependenciesbetweendifferentprocesses,becausetheyprovokeacertainreactionfromtheenvironmentandassuchsettingthestageforfurtherenvironmentaldisruptions.”

Kramer&Moorkamp(2016)statethatthebestsuchorganizationscandoistocaptureanabilitytoreflectonthepossibilitiesandlimitationsoftheirownorganizations,sinceany “design” is based on pre-existing ideas and in hyperturbulent conditions any“design” can become outdated very quickly. In other words, they need to be able to“organize doubt” directed at their own organizations (Kramer, 2007). Kramer &Moorkamp(2016):

“Thisoutdatedcharacterofdesignisvisibleinprocess:operatorsthataimtokeepasystemafloatareessentially involvedwiththis issue.Forthisreason, it isclaimedhere that systems need to organize their ability to continuously reflect on theirorganization:theyneedto“organizedoubt”directedattheirownorganization.”

“Organizing doubt” would in the case of TFU mean that the taskforce would haveunderstoodwhatkindsoforganizationalproblemstheywouldhaveencounteredwhentheydeployed49unitsinaverydynamicenvironment.Giventheacknowledgementofthe inherent vulnerability of the taskforce, they could have created all kinds ofstructures,perhapsa“parallellearningmechanisms”(Shani&Doherty,2003),inordertodealwithorganizationalissues,insteadofstumblingfromoneissuetothenext.ConclusionanddiscussionThe ironic side to this paper is that sociotechnical ideas are used to claim that themilitaryorganizationshouldtakehierarchymoreseriously.Evenmoreironicperhapsisthatsociotechnicalideas,whichhavetraditionallybeenusedtobreakdownrepressivehierarchicalstructures,areusedheretoclaimthatitisimportantforstructurestolimitpossibilities.Theconclusionofthispaperisthatthechaoplexicparadigmdidnotcreateaparadigm shift: our thinking about organizationsdoesneed to change radically as aresult of ICT innovations. Although the ideas of complexity science are very inspiringrenewing in organization studies (Morgan, 1997), a “one on one” translation into amodelthatdescribeshoworganizationscansuddenlyworkasagileintegratednetworkswithout hierarchy seems a bridge too far. That might work for very loosely coupledsystemssuchasbakeries,butisunlikelytoworkforveryintegratednetworksthathavenotimetoloose.ThetheoryofCASwasneveraboutthat4.4Indeed, ifevolutioncanbeunderstoodintermsoforderthatemergesfromchaos, itshouldbekept inmindthatnoteveryvariationsurvivesinbiologicalevolution.Infact,outofamassofvariationsonlysomesurvive,quiteunliketheimageofthesupernetworkwithouthierarchy,whichisbasicallyatheoryabouttrialswithouterror.

19

Thequestioniswhereallofthisleavesus.Althoughthecoreoftheideasofthechaoplexic paradigm is rejected here, that does not mean that the sociotechnicaltradition can happily ignore ICT innovations. A sociotechnical perspective will arguethat there is “organizational choice”.New technologydoesnot forceanorganizationalstructureuponus.Instead,wehavea“choice”:wecandevelopsmartwaysoforganizingtomakethemostof thenewtechnology. Itshouldbeemphasizedthat“organizationalchoice”shouldnotbeunderstoodasimplying“socialvoluntarism”(theideathatsocietyor the organization determines technological development). Instead, the idea thattechnologicaldevelopmentsandorganizationmutuallyinfluenceeachotherseemstobepreferable.Forthesociotechnicaltradition,thismeansthatalthoughICTinnovationsdonot determine the way organizations will be structured, they create possibilities inorganizational forms that previously were unthinkable. It is up to the sociotechnicaltraditiontotheorizeabouthowtotakeupthesetechnologicaldevelopmentsandtoseehowtheycanbeusedtocreateorganizationsthatareboth“humane”and“effective”.

Regardingthesepossibilities,itisindisputablethatICTinnovationshavenotonly“improved” existing ways of coordinating, they also have created new possibilitiesbeyond the scope of the traditional ways of coordinating. They have enabledmutualcoordinationbetweenorganizationalpartsinawaythatwouldhavebeenimpossibleinthe past (offshore software development is just one example). That means that –technologically–theyhaveenabledthecreationofnetworksinawaythatwouldhavebeen impossible before. The chaoplexic paradigm might have argued that such newnetworkscanbemadesoflexibleandmalleablethattheycanbedeployedinextremelyturbulentenvironmentsandfindtheirwayontheirown.Theimplicationofthispaperisthat–althoughtheremightbefewerandfewertechnologicallimitationincreatingsuchnetworks, there is in fact an organizational problem associated with such networks.Kramer&Moorkamp(2016)havecalledthesenetworks“vulnerable”asaresultoftheirnot fully developed character and the hidden interconnections between parts of thenetwork. This vulnerability will surface increasingly as environments becomeincreasingly turbulent.Thisemphasison the threatsof interconnection innetworks isgaining traction in international publications. It will be a main focus in Moorkamp’sforthcoming dissertation and it is a central issue in Roe & Schulman (2016) recentpublication on managing interconnections in critical infrastructures. So a broaderimplicationof this paper is that if technology createsneworganizational possibilities,thesociotechnicaltheorizingshouldfocusonhowtomakethemostofthistechnology,without the associated hazards. In case of ICT-technology and network-forms, thatmeans that sociotechnical theorizing should focus on developing an understanding oftheorganizationaldynamicsof“network-forms”ReferencesAlberts,D.S.&Hayes,R.E.(2003).PowertotheEdge:CommandandControlinthe

InformationAge.WashingtonDC:CCRPPublicationSeries

20

Atkinson,S.R.&Moffat,J.(2005).TheAgileOrganization.WashingtonD,C:CCRPPublicationSeries.

Bate,P.,R.Khan,&Pye.A.(2000).Towardsaculturallysensitiveapproachtoorganizationstructuring:Whereorganizationdesignmeetsorganizationdevelopment.OrganizationScience11(2):197–211

Bijker,W.(2006).Whyandhowtechnologymatters.In:TheOxfordHandbookofContextualPoliticalAnalysis.Oxford:OxfordUniversityPress.

Bezooijen,vanBJAandKramerEH(2014).MissioncommandintheInformationAge:ANormalAccidentsPerspectivetoNetworkedMilitaryOperations.In:JournalofStrategicStudies

Blom,T.(1997).Complexiteitencontingentie.EenkritischeinleidingtotdesociologievanNiklasLuhmann.Kampen:KokAgora.

Bousquet, A. (2008). Chaoplexic warfare or the future of military organization. In:InternationalAffairs.Volume84,Issue5,pages915–929

Cebrowski,A.K.(2005).TheImplementationofNetwork-CentricWarfare.DepartmentofDefense.OfficeofForceTransformationOfficeoftheSecretaryofDefense

Christis,J.(1998).Arbeid,organisatieenstress.Eenvisievanuitdesociotechnischearbeids-enorganisatiekunde.Amsterdam:HetSpinhuis.

DekkerS(2011).DriftintoFailure.Fromhuntingbrokencomponentstounderstandingcomplexsystems.AshgatePublishingLimited,Farnham

ElFartasi,N.&DeVivo,D.(2016).Resilientforhowlong?InformationTechnologyWarfareinthe21stcentury:TheAlliance’sinvisiblethreat.In:VoxCollegii.NATODefenseCollege

Gray, C.S. (2002).Strategy forChaos.Revolutions inMilitaryAffairsand theevidenceofmilitaryhistory.London:FrankCasspublishers

Herbst, P. (1976). Alternatives to hierarchies. Leiden: Martinus Nijhof Social SciencesDivision

Isaacson,W. (2015). Foreword. In:McCrystal, S. (2015).TeamofTeams.New rules ofengagementforacomplexworld.London:PortfolioPenguin

KuipersH,VanAmelsvoortP,KramerEH(2010).Hetnieuweorganiseren.Alternatievenvoordebureaucratie.Leuven:Acco

Kramer EH (2007). Organizing doubt. Grounded Theory, Army Units and Dealing withdynamiccomplexity.CopenhagenBusinessUniversityPress,Copenhagen

Kramer,E.H.Waard,deE.Graaff,deM.(2012).TaskForceUruzganandexperimentationwithorganizationdesign.In:Meulen,vander,J.Soeters,J.Beerens,R,Vogelaar,A.MissionUruzgan.Amsterdam:AmsterdamUniversityPress

Kramer,E.H.&Bezooijen,B.J,A.van.(2016;Inpress)Thelimitsofself-organization.In:The Royal Swedish Academy of War Sciences Proceedings and Journal(forthcoming)

Kramer,E.H.&Moorkamp,M.(2016;inpress).Understandingorganizationalvulnerabilityinmilitarytaskforces.In:Beeres,R.,Bakx,G.,DeWaard,E.J.&Rietjens,B.OrganizingforSafetyandSecurityinMilitaryOrganizations,NLARMS,NetherlandsAnnualReviewofMilitaryStudies.NewYork:SpringerVerlag

21

(forthcoming)McCrystal,S.(2015).TeamofTeams.Newrulesofengagementforacomplexworld.

London:PortfolioPenguinMoorkamp,M.,Kramer,E.H.(2014).VeiligheidvanexpeditionairorganiserenbinnenTask

ForceUruzgan.NederlandseDefensieAcademie,BredaMoorkampM,WyboJL,KramerEH(2015).PioneeringwithUAVsatthebattlefield:the

influenceoforganizationaldesignontheemergenceofsafety.In:SafetyScience(publishedonlinein2015).

Morgan,G.(1997).Imagesoforganization.London:Sage;NewEdition.PerrowC(1999).NormalAccidents.Livingwithhigh-risktechnologies,2ndedition.

PrincetonUniversityPress,PrincetonPerrowC(2004).APersonalNoteonNormalAccidents.Organization&Environment17:

9-14Roe,E.&Schulman,P.(2016).ReliabilityandRisk.TheChallengeofManaging

InterconnectedInfrastructures.Stanford:StanfordUniversityPress.SimonHA(1962).TheArchitectureofComplexity.In:ProceedingsoftheAmerican

PhilosophicalSociety,106m467-482Simon,H.A.(1997).Thesciencesoftheartificial.Cambridge(MA):MITpress.3rdedition.DeSitter,L.U.(2000).Synergetischproduceren:HumanResourcesMobilisationinde

produktieeeninleidinginstructuurbouw.UitgeverijVanGorcum,Assen.Shani,R.&Doherty,P.(2003).Learningbydesign:buildingsustainableorganizations.

Oxford,UK:Blackwellpublishing.Thompson,J.D.(2008).OrganizationsinAction.SocialScienceBasesofAdministrative

Theory.NewBrunswick:TransactionPublishers.Originallypublishedin1967Trist,E.L.,Higgin,G.W.,Murray,H,Pollock,A.B.(1963).Organizationalchoice.

Capabilitiesofgroupsatthecoalfaceunderchangingtechnologies:theloss,re-discovery&transformationofaworktradition.London:TavistockPublications

Waldrop,M.(1992).Complexity.Theemergingscienceattheedgeoforderandchaos.NewYork:Simon&SchusterWaard,DeE.J.&Kramer,E.H.(2010).Expeditionaryoperationsandmodularorganizationdesign.In:Soeters,J.;Fenema,vanP.&Beeres,R.Managingmilitaryorganizations.London:Routledge

Weick,KE(2004).RethinkingOrganizationalDesign.In:BolandjrRJ,CollopyF(eds)ManagingasDesigning.StanfordUniversityPress,Stanford