Armin Grunwald: Technik fur die Gesellschaft von

morgen. Moglichkeiten und Grenzen gesellschaftlicher

Technikgestaltung

Campus Verlag, Frankfurt/New York, 2000,

314 pp (ISBN 3-593-36750-5), Euros 34.90

Michael Decker

‘‘Objectives’’, not ‘‘Introduction’’ is thetitle of the first chapter of A. Grun-wald’s new book Technik fur dieGesellschaft von morgen – entirelyconsistent for an author whose writingshave always reflected a theory of actionguided by aims and objectives. The aimof his new book is to offer furtherclarifications and suggestions regard-ing the following core issues (p. 43):

1. Does the notion of shaping tech-nology with a social perspective stillhave a meaning in a modern, de-centralized and differentiated soci-ety, and what can this meaning be?

2. What can it mean in today’s societyto be active in shaping technologywith a social perspective in alegitimate manner?

3. What can it mean in the(post)modern, pluralistic society toshape technology with a socialperspective according to criteria ofrationality?

By laying down these objectives, theauthor commits himself to a certainstructure of the argument developed inthis book. At the same time, Grunwaldavoids subjecting his thinking to anyimmediate pressure to act with regardto some imminent social decisionsconcerning technology development.Instead, he gives a space of its own tohis reflexive–analytical theorizing onthe questions above. Even if a variety oftechnological developments exert anactual pressure to act and decide, atheoretical discussion of the (types of)problem(s) presented by develop-ments, and of the strategies for dealingwith them, surely is indeed desirable, atleast in the sense of a division of labor.However, Grunwald is right to add thatsuch a discussion will be particularlyhelpful if a rigorous orientation of thepractice of technology assessment andshaping prevents it from evaporatinginto the ethereal realm of ‘‘puretheory’’ (p. 42).

Published online: 20 November 2002� Springer-Verlag 2002

M. DeckerEuropaische Akademie GmbH, Wilhelmstrasse 56,53474 Bad Neuenahr-Ahrweiler, GermanyE-mail: Michael.Decker@dlr.de

Poiesis Prax (2003) 1: 231–238 DOI 10.1007/s10202-002-0023-0

Book Review

231

The tasks to be mastered before theobjectives can be achieved are charac-terized throughout, like any scientificreflection on technology, by a set ofstructural problems, which, in practicaltechnology assessment, all too oftenmanifest themselves as traps on bothsides of the methodical path: (i) on theone hand, there is the danger of one-sidedly regarding the problem from theperspective of the participant, over-emphasizing the intended results of atechnology in particular. On the otherhand, too much emphasis may be puton the spectator perspective mainlyconcerned with the factual conse-quences of the same technology. (ii)Considering the unquestionable (andunquestioned by Grunwald) intrinsicmomentum of technology develop-ment, there is the danger of a fatalisticattitude towards it, perceiving thedevelopment as being determined, asit were, by nature (the ‘‘evolutionisticfallacy’’). Conversely, especially amongthose who further the development bytheir research, there is the risk of beingbiased towards an excessive optimismwith regard to their power of shapingthe development (the ‘‘planners’ opti-mistic fallacy’’). (iii) Finally, there lurksto one side of the path the trap of the‘‘descriptivist fallacy’’, which elevatesan actual social development – includ-ing what is de facto desired in a societyat any one time – to the binding stan-dard while, to the other side, there liesthe gaping trap of the normativistfallacy of moralistically formulatingdictates and prohibitions lacking asufficient, normative basis and/or abasis in communal customs and pur-poses. Such offers of guidance, as far asthey have not evolved at least fromconcrete social practice, end up asmere empty pleas.

Grunwald, conscious of these traps,focuses initially, in an analytical ap-proach, on established forms of creat-ing models to describe technologicaldevelopments. The technogeneticforms essentially look at the process of

producing technology. Evolution-theo-retical forms of model creation com-plement this approach with the effortof explaining the chronological dy-namics of technical products, as well astheir social integration. In contrast tothese forms of model creation, thesocio-technological forms are of adeterministic nature: they assume thatthe technological development isdetermined by sociological rules ofprogression, which the model creatorsinterpret in analogy to the rules ofprogression in natural systems.

Relying on work done byM. Gutmann and P. Janich, Grunwaldarrives at a critique of thoseapproaches by applying a model-theo-retical analysis. Models neither evolvefrom nature, nor do they emergespontaneously, but they are instru-ments created by the ‘‘model builder’’to help him in his efforts to understandthings. Hence, according to Grunwald,the concept of models should bereconstructed on (at least) threecornerstones: a model of something isdesigned by somebody for some purpose(p. 61). Consequently, the creation of amodel involves decisions being madeby the model designer according to hisor her objectives. For instance, thesubject area to be modelled has to bedefined in accordance with the aims ofthe model-building exercise, and mustbe described in a language consideredas conducive to these aims. The prob-lem with these models is that theydescribe the technological developmentex post, from a spectator perspective,before they formulate assertions aboutthe future development of the techno-logical; extrapolating, so to speak, fromthe past to the future. This leaves openthe question of why these descriptionsshould be of normative relevance forex ante matters of shaping develop-ments. Grunwald rightly points outthat the spectator perspective is inad-equate here, and that the participantperspective must be taken into accountexplicitly: past actions teach us about

232

future actions only if they are recon-structed from the agent’s perspective.This emphasis on the participant’sperspective is a central issue of Me-thodical Culturalism (D. Hartmann, P.Janich), Grunwald’s philosophical ori-gin.

He accordingly sets action-theoreti-cal modelling, guided by practice andbased on the rational explanation ofactions, against the established formsof model creation – although the par-ticipant perspective is paired (in con-trast to Grunwald’s older work) withthe spectator perspective. While thefirst is appropriate for shaping devel-opments, the latter is needed for thereflection in terms of sociological the-ories (p. 87), as a corrective to themodelling of technological develop-ments from the participant perspectiveand to avoid individualist distortions.This hand-in-hand of the participantand spectator perspectives is demon-strated by the example of how cir-cumstantial evidence is treated incourt, where the statements of theagents (delinquent, victim, witnesses)are combined with evidence andinterpretations.

Entirely in line with his approach,understanding the modelling of tech-nology shaping as a flexible planningprocess, Grunwald suggests an aims-guided incrementalism between theplanning-optimistic and planning-sceptical views. The mission of tech-nology assessment (TA) (p. 105) wouldbe to work out such flexible planning oftechnology development, guided bylong-term objectives and, at the sametime, showing short-term flexibility.Accordingly, the intention to shapetakes centre stage in the first part of thetrilogy ‘‘Social Shapability of Technol-ogy’’.

The shaping of technology from asocial perspective has to be legitimizedas such. As much as there is agreementabout this point, opinions diverge onthe question of how this legitimizationcan be achieved. Through closer

inspection of some of the core intu-itions with respect to the concept oflegitimation, Grunwald endeavours toclassify the various agents in theshaping of technology and to put themunder a critical spotlight (Sect. 3.3). Inthe process, the state is rehabilitated asthe principal agent of a social shapingof technology. The difficulties andproblems of arriving at socially legiti-mized statements are discussed inSect. 3.2. Although one cannot getaround the legitimization by proce-dures (J. Habermas) – for in situationsof having to make a decision undertime pressure, ‘‘discussing through’’ aproblem is surely out of the question –there still is the need to realize an ar-gumentive foundation for the decision,or it would not be accepted as legiti-mate even if the procedure has beenfollowed correctly. Especially for thelosers in a decision (every decisionproduces losers), legitimization has toensure the acceptability of the result. Inthis context in particular, proceduresare often suggested that are supposedto assure the acceptance of the decisionby providing the participation ofcitizens, laypeople or stakeholders(individuals concerned). For a repre-sentative selection of the participantsensures that potential losers also havetheir part in the discussion. In contrastto the assumption in participatorymodels of legitimization, (mere factual)acceptance cannot contribute to thelegitimization – the more rigorous cri-terion of acceptability must be invoked(C.F. Gethmann). Likewise, to be re-jected is the often heard argument thatthe social shaping of technology mainlytakes place in industry, in its labora-tories and factories: engineers andmanagers may be dealing with theconcrete shaping of a technology,however, they legitimize their decisionsvis-a-vis the managing board and theshareholders, but not for society as awhole. Thus, shaping technology froma social perspective is incumbent onthe state, which has to act mainly by

233

regulating the shape of boundary con-ditions.

The fourth and last chapter on theway to reasoned knowledge and legiti-mate guidance, takes us to argumentiverationality, explicitly introduced as aconcept of reflection, which ensuresthat ‘‘the possibility of descriptive andnormative assertions being transsub-jective becomes a topic’’ (p. 194). Theaim is – always from the participantperspective – to create reliable struc-tures for the social agents, where thecriteria for ‘‘acting rationally’’ arerooted in society itself. Normative rec-ommendations for action ought to op-timize (not guarantee!) the prospects ofsuccess. Argumentive rationality is re-alized through integrative, transparentresearch into technology. Science isassigned the task of developing – in aproblem-oriented and therefore, inmost cases, interdisciplinary manner –integrated and coherent recommenda-tions for political action. Integration isunderstood to be achieved not just by‘‘delivering’’ interdisciplinary researchof a high quality (see Sect. 4.2.3.3),which is difficult in itself, but also byproviding a definition of the problem(Sect. 4.2.3.1). Such a definition alwaysinvolves the transdisciplinary problemthat criteria not connected to scienceplay a crucial role. Grunwald identifiesthese extra-scientific aspects in partic-ular as a field of application of partic-ipative methods.

Is normativeness, as a vital ingredi-ent of a social shaping of technology,something that is constituted and le-gitimized? Concerning this question,Grunwald’s argument starts from theconstitution of normative and, in par-ticular, evaluative rationality. Firstly,he introduces predeliberative consent(basically C.F. Gethmann’s ‘predis-coursive consent’) as the basis of con-sultation itself, thereby making itobvious that absolute objectivity is notwhat one strives for (p. 261). Thispredeliberative consent provides the

space of reasons (principles) fromwhich one can proceed, via if–thenchains of arguments, to deciding con-troversial issues. Subsequently, argu-mentive rationality (separated intoinstrumental, cognitive and evaluativerationality) is described as a socialconstruct based on predeliberativeagreements. Here the possibility ofcriticism, which is heard only after theactual deliberation has taken place, isafforded a central role. The possibilityof criticism has to be rooted in socialpractice, as the de facto acceptance ofthe counterfactual. This also consti-tutes the basis of Grunwald’s thoughtson the social coherence, which is citedas a corrective for long-term guidance(p. 281). Innovations are ‘‘breaches ofcoherence’’ in this context, and initiallylead to a conflict. Such short-term in-coherence will transform into a (new)coherence in the future, if the newtechnology succeeds in becoming con-nected or ‘‘encultured’’ (p. 286). Final-ly, taking up the title of the book, theshapability of technology is presentedas a concept of reflection encoding theinquiry into the possibilities and limitsof shaping technology (p. 291). Theactual activity of the participants ispresented as their basic decision for thedesire to shape technology. The con-ditions to make possible such shapinghave to be created by society itself.

The practitioner’s perspectiveOne of the essential bonuses of thebook has to be that the ‘‘possibilitiesand limits of the social shaping oftechnology’’ referred to in the subtitleare not discussed by just workingthrough possible methods, and theirlimits and potentials. What is underscrutiny is the shapability of technolo-gy from a social perspective itself,which does not, however, mean thatthere is nothing to be learned bypractitioners here. Quite the opposite:the book is at its most gripping if thepractice-oriented reader keeps asking

234

himself the question, ‘‘how can Ioperationalize this?’’ at every turn.The practitioner is well-advised toavoid in his procedures the variousfallacies outlined in the first chapter.The graphic description and, mostimportantly, the recurrent reference tothese ‘‘traps’’, for instance in the cri-tique of modelling theories (Sect. 2.2),or the possibility of criticism(Sect. 4.3.4), show how ‘‘omnipresent’’they are. Furthermore, it is shown in allthe clarity one could wish for that oneshould aim for the legitimization ofresults. The practitioner should alwaysask the question: ‘‘How will I ‘sell’ myresults to the loser?’’ (p. 118). One ofthe principal merits of the book is thatit points out that both socio-politicaland rational arguments are requiredhere, and thereby confronts the misin-terpretations and misunderstandings ofthe procedure and argumentationmonists respectively. Advice on how tooperationalize this endeavor is given inan indirect fashion only, through thecritical discussion of the problemsconcerning legitimization (Sect. 3.3).However, this critique is very detailed,e.g. for participative approaches(Sect. 3.3.4). The same is true for theissue of rationality. The reader is of-fered advice on what to look out for inoperationalization: the proceduresmust have been accepted and the re-sults do not need to be unequivocal inevery detail. Within so-called rational-ity corridors, there is space for variousoptions.

In the following, three representativesections will be examined in more de-tail, with regard to their content of in-formation concerning the potential foroperationalization, which is of partic-ular relevance for the practitioner: (1)How do I arrive at the definition of aproblem? (Sect. 4.2.3.1); (2) what isthe basis of the discussion?(Sect. 4.3.3); and finally (3), what arethe requirements to be met by results?(Sect. 2.4.2).

Problem definition and problemdescription (Sect. 4.2.3.1)In accordance with the philosophicalschool of methodical constructivism(P. Lorenzen) of which Grunwald is adescendant, all research basically isproblem-oriented research, whichusually requires interdisciplinary co-operation and is fraught with one dif-ficulty right from the start – the explicitdefinition and description of theproblem. Grunwald points out in greatdetail that concepts used in the socialdiscussion bear not just descriptionsbut also judgements (p. 221). He alsoemphasizes that the concrete formula-tion of the problem already pointsproblem-oriented research in a certaindirection (p. 222). He finally points outthat from a diffuse definition of a topic,various problem definitions can begenerated for ‘‘scientific treatment’’.Classification is a necessity, but also adanger, since it can exclude relevantpoints and dictate priorities amongpoints regarded as relevant (p. 223).For Grunwald, the way in which toclassify is the pre-empirical componentof integrative research. He proposes tolet individuals concerned take part inthe problem definition for such ‘‘deci-sions relevant to science but governedby extra-scientific criteria’’ (p. 223). InSect. 3.4.1, it is further clarified thatpeople concerned should help extendthe database by contributing their‘‘local knowledge’’ (p. 179). His call forparticipation has to be a surprise, evenif it comes under the proviso that moreresearch is needed on this issue,because Grunwald strongly criticizedparticipative procedures in some of hisearlier publications. Participatory TA(PTA) is rated as a tool of conflictmanagement in decision situations ofthe NIMBY type (not in my back yard),and as a weapon against politicalapathy amongst the population. Thequestion remains what to do with thedata gained from participation. Evenafter consulting the individuals or

235

groups concerned, it will still be nec-essary to classify and decide on what isrelevant, and those classifications anddecisions should be based on rationalarguments. But are the decisions to bemade by scientists or by citizens? Towhat extent should problem-orientedresearch projects be bound by thesedecisions on relevance? In an initialtreatment, the problems surroundingthe legitimization of participative dis-course remain unsolved, even if suchdiscourse ‘‘only’’ serves to define aproblem. A question less targeted onthe potential for operationalizationwould be: what is ‘‘local knowledge’’?Grunwald has stated in earlier workthat local knowledge is limited to re-gional technology issues. But what isthe difference between this local expertknowledge and scientific knowledge?Does it still make sense to distinguishbetween this ‘‘layman’s expertise’’ andscientific knowledge? One would havehoped that Grunwald, who is a philos-opher of science and justifiably worriesabout maintaining the quality of in-terdisciplinary research, had offered anin-depth treatment of the status of localknowledge from a philosophy of sci-ence point of view – especially as hedoes not limit the duty of participation(p. 230) in the pre-empirical defini-tions for integrative research toregional issues. In this respect, onecertainly looks forward to futurepublications.

The constitution of normativerationalityFrom the perspective of the practitio-ner, two point lists take centre-stagehere. The first of them explains thepredeliberative consent as ‘‘the funda-ment on which arguments can be built’’(p. 262). For instance, a moderator ofan interdisciplinary group of scientistswill have to take care that such prede-liberative consent does exist betweenthe participants of the discussion.One essential aspect is that the

predeliberative consent has to be de-veloped collectively by the participants.Assessment catalogues have to be laiddown, social practices reconstructedand rationality standards acknowl-edged. Grunwald explains that differentpredeliberative consents can lead todifferent results, and finally, predelib-erative agreements are shown to behelpful through giving guidance andcreating a relative stability; which,however, can be changed in reflectionon the practice. If one invited a scien-tist to contribute to an expert groupand enclosed the list on predeliberativeconsent – with a polite request to takenote – the scientist would probablyadmit to himself that he has neverobserved any of the points, at least notknowingly. The reason is that theseaspects are not at the centre of every-day scientific work. Within any scien-tific discipline, the methods by whichscientific assertions are given weightare known. The same catalogues ofreasoning are applied ‘‘automatically’’,and everyone has an ‘‘instinct’’ forwhen modifications are necessary orworthwhile. Predeliberative consentonly becomes relevant in an interdis-ciplinary setting, because then themethods will be debated, and ‘‘auto-matisms’’ and ‘‘instincts’’ are notshared anymore. Grunwald himselfseems to have noticed that, with a viewto the practice shaping technology, histreatment of predeliberative consentrequires further explanation, notwith-standing his lucent writing. Therefore,he adds a list of pragmatic precondi-tions (p. 265), which can be calledupon as a catalogue of criteria for theproject executive. Points 1 (a numberof options have to be available fordecision) and 3 (symmetry of partici-pants) can be secured ex ante withrelative ease. Point 2, demanding thereadiness to learn on the part of theparticipants, i.e. their acceptance of therisk of discourse (result openness), canhardly be affirmed ex ante. Initially,each participant will regard his

236

argument as so powerful that he ratesthis risk as acceptable. During the dis-course, however, the risk to the indi-vidual can increase to such an extentthat exit from the discourse becomesthe only option. At first sight, theconfidence of success (point 4) willalways be given, since without it theindividual participant would notcollaborate in the project. Closerinspection, however, shows that thetheoretical possibility of a rationalityassessment is difficult to operational-ize.

What are the requirements to be metby results? (2.4.2)Aims-oriented incrementalism isintroduced as the form of flexibleplanning to apply in the shaping oftechnology. It is positioned between theextremes of utter optimism and hos-tility towards planning. By linking it toother chapters (list on p. 103), the cleardescription of this aims-oriented in-crementalism as a model for flexibleplanning is made the centrepiece of thebook. This is hardly surprising, sincethe theory of planning is one of thepriorities of Grunwald’s scientific work.

Here the practitioner finds answersto the question of what the results of atechnology assessment should looklike. On the one hand, directions andlong-term objectives have to be de-fined. These objectives can then beinvoked as criteria for the adequacy ofimpending decisions on a course ofaction, for any action must constitute astep towards achieving the objective. Atthat stage, it is crucial that the objec-tives themselves are subject to changes.In accordance with the usual scheme ofproject organization, Grunwald pro-poses milestones as the tool for oper-ationalizing the necessary reflectionaccompanying the process. Thesemilestones can be the results of short-term, aims-oriented action. Such re-sults supplement the knowledge base,which is the foundation of shapingtechnology, and offer a welcome

opportunity to scrutinize the objectivesof the respective planning process.Hence, the results of a discussion aboutthe concrete shape of a technologicaldevelopment must comprise a rea-soned definition of the objectives of theplanning process, as well as anticipatecertain development stages, at which areflection on what has been achieved sofar shall take place and the resultingchanges to the flexible planning shallbe implemented.

In this way, Grunwald succeeds inadding tangible substance to thecatchphrase ‘‘technology assessmentaccompanying developments’’. Thecombination of long-term objectivescorridors and short-term decisions onaction, for which actual acceptancemust be achieved in any case, appearsto be plausible, although it is far fromobvious that the results of a TA processshould be of this form. Quite often themere comparison of options and theirconsequences are considered as theresult of a TA study, without evenmentioning a concrete recommenda-tion for one of these options for action.To achieve something like ‘‘value neu-trality’’ is an explicit aim. It may besensible to remind the participants ofthe discourse in advance that the de-velopment of a flexible plan is part ofthe agenda for the project. Misunder-standings during the project may beavoided if this requirement weresupported by predeliberative consent.

General commentsIn his preface, Grunwald points outthat this book ‘‘lives off’’ his previousscientific work, joining the mosaic andtrying to close gaps that may appear init. Readers who know his work will notfail to miss this fact. Nevertheless,Grunwald made an excellent job of re-structuring this patchwork. He has alsosucceeded in finding the path betweencompleteness and deliberate omission.The structure ‘‘across’’ the chapters isparticularly remarkable. The leadinghypotheses and basis differentiations

237

from the first chapter are taken upagain whenever it is appropriate. At thebeginning of chapters two, three andfour, the crucial concepts and theirorigin are discussed (model, p. 61, le-gitimization, p. 120, pragmatic ratio-nality, p. 196). At the end of each ofthese chapters, the author revisits thecore issue, ‘‘social shapability of tech-nology’’, summarizing everything saidso far about this issue and relating it tothe next topic to be treated. The re-construction is always followed by acritique of the approaches under sci-entific discussion and by Grunwald’sown proposal of how to deal with theissue in hand. Thus, each chapter canbe read on its own, while the identicalstructure of each individual chaptersafely guides the reader through, andmakes him or her appreciate, Grun-wald’s chain of arguments.

By providing this ‘‘service’’, Grun-wald creates a wide circle of prospec-tive readers for this book, includingthose who shape technology throughpolitical decisions (see Sect. 3.3.1, therole of the state in technology) and,hence, have a professional interest inshapability. The structured explana-tions and detailed critique of existingapproaches make this book suitableeven for newcomers to the field of TAand for the interested public. Thereader gains a valuable overview and isencouraged, through the comprehen-sive literature list, to follow up on anytopic in the scientific discourse. Thisbook should be required reading, forthe agents of TA in all its variations,regardless of the fact that the role ofparticipation and ‘‘local knowledge’’remains to be specified in more detail.

238